Steel Water Storage Tanks Design Construction Maintenance and Repair PDF

Source: Steel Water Storage Tanks: Design, Construction, Maintenance, and Repair
CHAPTER 1
Tank History, Typical
Configurations,
Locating, Sizing,
and Selecting
Ira M. Gabin, P.E.
Dixon Engineering
Richard A. Horn, P.E.

CB&I
The rapid development and expansion of public water supply systems

at the beginning of the 20th century led to the establishment of public
health standards for drinking water systems. An area of major con-
cern for these systems was the storage facilities. Early steel reservoirs
and standpipes were of riveted construction. Modern welded-steel
reservoirs can be built to very large capacities with either domed or
column-supported roofs.
In the 1970s, it became common for smaller-capacity reservoirs
and standpipes to use bolted construction technology, originally de-
veloped for industrial and agricultural uses. Prefabricated panels and
bolted connections reduced erection costs and made these structures
popular in rural areas. The advent of factory-applied ceramic coat-
ings reduced future maintenance costs, adding to the tanks’ attrac-
tiveness to water supply systems with limited financial resources.
Bolted tanks with diameters greater than 30 ft (9 m) are often built
with low-maintenance aluminum geodesic domed roofs, a technology
1
Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)
Copyright © 2010 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
Tank History, Typical Configurations, Locating, Sizing, and Selecting
2 Chapter One
FIGURE 1-1 Geodesic dome on bolted-steel reservoir.
commonly found on wastewater plant storage tanks as well. Figure

1-1 shows a geodesic dome on a bolted reservoir.
The earliest elevated storage tanks were constructed of wood in
the manner of water refilling stations for steam-powered trains. Some
were built on stone or brick columns. Limitations as to size and dura-
bility, as well as public health concerns, led to steel becoming the
material of choice for elevated tanks. Most steel elevated tanks con-
structed before 1950 were riveted, their legs consisting of opposed
channels connected by latticework bracing. Roofs on most small tanks
and many larger ones were the familiar cone or “witch’s hat” design
(Fig. 1-2). Some larger elevated tanks had hemispherical or ellipsoidal
roof designs.
Welded construction became the industry norm by the early 1950s
and remains the standard for most elevated tanks. Legged tanks con-
tinued to be built in great numbers; however, the lattice legs replaced
tubular sections. Many larger-capacity legged tanks were of the radial
arm design shown in Fig. 1-3. These have been phased out in favor of
the toroelliptical legged tank style.
Early prototypes of single-pedestal tanks were developed in the
1940s and became a common alternative to legged tanks by the 1950s.
The more efficient shape of these structures provided the advantage
of lower maintenance costs. In the 1960s, the fluted-column single-
pedestal design was introduced, which provided a usable area in the
column for pumping equipment, storage, offices, and other municipal
uses.
Legged tanks continue to be built primarily in sizes up to 1 million
gallons (mil gal) (3.8 million liters [ML]) as a lower-cost alternative
to single-pedestal or fluted-column tanks. Single-pedestal tanks are
widely specified from very small to large capacities. Larger capacities

Ta n k H i s t o r y, C o n f i g u r a t i o n s , L o c a t i n g , S i z i n g , S e l e c t i n g 3
FIGURE 1-2 Witch’s

hat roof design.
FIGURE 1-3 Legged tank with radial arm design.

4 Chapter One
(0.75 to 2 mil gal [2.8 to 7.6 ML] or more) are generally single-pedestal
or fluted-column tanks. Some fluted-column tanks have even larger
capacities.
In the late 1980s, composite-tank technology combined a concrete
pedestal with the steel-bowl geometry of the fluted-column tank. This
addressed one of the concerns of the fluted-column design—the large
steel surface area and resulting higher repainting costs. Built generally
to hold 0.75 to 2 mil gal (2.8 to 7.6 ML) of water, composite tanks are
now in use throughout the United States and Canada.
Other materials and technologies are available for specialized ap-
plications. However, the steel, glass-lined steel, concrete, and com-
posite tanks discussed in this chapter comprise the large majority of
tanks currently in use and being specified for new construction.
Reservoirs
A reservoir is a ground-supported, flat-bottom cylindrical tank with a
shell height less than or equal to its diameter. Reservoirs are one of the
most common types of water storage structure. They are used as a part
of the distribution system as well as to hold treated water for pumping
into the distribution system. Of the three types of steel water tanks, a
reservoir, because of its low height, is generally the most economical
to fabricate, erect, and maintain. See Figs. 1-4 and 1-5 for a photo and
a cross-sectional view of a welded-steel reservoir; see Figs. 1-6 and 1-7
for a photo and a cross-sectional view of a bolted-steel reservoir. Table
1-1 gives typical sizes of welded-steel reservoirs, and Table 1-2 gives
capacities of glass-coated, bolted-steel reservoirs and standpipes.
Storage reservoirs for potable water are covered by roof structures,
which may be either column supported or self-supporting. Standard
tank accessories may include shell and roof manholes, screened roof
vents, inside or outside ladders, and connections for pipes as required.
Standpipes
Standpipes are ground-supported, flat-bottom cylindrical storage
tanks that are taller than their diameter. They are usually built where
there is little elevated terrain and where extra height is needed to cre-
ate pressure for water distribution. See Figs. 1-8 and 1-9 for a photo
and a cross-sectional view of a welded-steel standpipe and Figs. 1-10
and 1-11 for a photo and a cross-sectional view of a bolted-steel stand-
pipe. Table 1-3 gives capacities and sizes of typical welded-steel stand-
pipes.
Standpipe systems are often designed so that the water in the tank,
until it reaches a certain low level, maintains the system pressure.
When that low level is reached, pumps come on, valving is changed,

FIGURE 1-4 Welded-steel reservoir. (Photo: Gay Porter DeNileon, AWWA)
Approved ladder, cage,

Roof manholes platform, or safety devices
Roof vent
12 in. (0.3 m) complying with Occupational
Overflow Safety and Health Act
3⁄4 in. (19 mm)
pipe
Capacity level
Weir box
Roof rafters
(optional) Column support
Inlet–outlet
Splash (optional) Tank bottom Shell manholes
Column
pad Base elbow crowned (two required)
bases
or valve pit at center
Sand pad
Compacted backfill
Crushed rock or gravel Concrete foundation
FIGURE 1-5 Cross-sectional view of welded-steel reservoir. (Source: AWWA

Manual M42, Steel Water-Storage Tanks)

6 Chapter One
FIGURE 1-6 Bolted-steel reservoir, glass fused to steel.
Approved ladder, cage, and platform

complying with Occupational Safety
and Health Act
Roof manway Gravity

ventilator
Internal
overflow
funnel
Overflow
Floor sloped toward pipe
outlet pipe
Inlet–outlet Splash
pad
24-in. (0.6-m) round Grade

access door level
FIGURE 1-7 Cross-sectional view of bolted-steel reservoir. (Source: AWWA


Capacity Range of Sizes Available

Diameter Height to Diameter Height to
(US gal) (m3 ) (ft [in.]) TCL (ft [in.]) (m) TCL (m)
50,000 189 19 [3] 24 [0] 5.9 7.3
60,000 227 21 [0] 24 [0] 6.4 7.3
75,000 284 23 [6] 24 [0] 7.2 7.3
100,000 379 23 [6] 32 [0] 7.2 9.8
27 [0] 24 [0] 8.2 7.3
125,000 473 26 [0] 32 [0] 7.9 9.8
30 [3] 24 [0] 9.2 7.3
150,000 568 28 [6] 32 [0] 8.7 9.8
33 [0] 24 [0] 10.0 7.3
200,000 757 33 [0] 32 [0] 10.0 9.8
38 [3] 24 [0] 11.7 7.3
250,000 946 37 [0] 32 [0] 11.3 9.8
42 [9] 24 [0] 13.0 7.3
300,000 1,136 40 [6] 32 [0] 12.3 9.8
46 [9] 24 [0] 14.3 7.3
400,000 1,515 46 [6] 32 [0] 14.2 9.8
54 [0] 24 [0] 16.5 7.3
500,000 1,893 46 [6] 40 [0] 14.2 12.2
52 [0] 32 [0] 15.9 9.8
60 [6] 24 [0] 18.4 7.3
600,000 2,271 51 [0] 40 [0] 15.6 12.2
57 [0] 32 [0] 17.4 9.8
750,000 2,839 57 [0] 40 [0] 17.4 12.2
64 [0] 32 [0] 19.5 9.8
1,000,000 3,785 66 [0] 40 [0] 20.1 12.2
74 [0] 32 [0] 22.6 9.8
1,500,000 5,678 80 [6] 40 [0] 24.5 12.2
90 [6] 32 [0] 27.6 9.8
2,000,000 7,571 93 [0] 40 [0] 28.4 12.2
104 [6] 32 [0] 31.9 9.8
3,000,000 11,356 114 [0] 40 [0] 34.7 12.2
127 [6] 32 [0] 38.9 9.8
4,000,000 15,142 131 [6] 40 [0] 40.1 12.2
147 [6] 32 [0] 44.9 9.8
5,000,000 18,927 147 [0] 40 [0] 44.8 12.2
165 [0] 32 [0] 50.3 9.8
7,500,000 28,391 180 [0] 40 [0] 54.9 12.2
201 [6] 32 [0] 61.4 9.8
10,000,000 37,854 233 [0] 32 [0] 71.0 9.8
208 [0] 40 [0] 63.5 12.2
Source: AWWA Manual M42, Steel Water-Storage Tanks.

Note: TCL = top capacity level.
TABLE 1-1 Capacities and Sizes of Typical Welded-Steel Water-Storage

Reservoirs
7
8
Nominal Height (ft)∗
Nominal 15 19 24 28 33 38 43 47 52 57 61 66 70 75 79 84 89 93 98 102 107 112 116 121
Diameter
(ft)∗ Capacity in Thousands of Gallons†
14 16 22 27 32 37 44 49 54 59 65 70 75 80 86 91 96 101 107 112 117 122 128 133 139
17 24 31 39 47 54 63 70 78 86 93 101 108 116 123 131 139 146 154 161 169 177 184 192 199
20 33 43 53 64 74 86 96 106 117 122 137 148 158 168 179 189 199 210 220 230 241 251 261 272
25 54 71 88 105 122 142 159 176 193 210 227 244 261 278 296 313 330 347 364
31 81 107 132 158 183 212 238 263 289 320 340 365 391 416 442
36 114 149 185 220 256 292 327 363 398 434 469 505
42 151 199 246 294 341 388 436 483 531 578
50 218 286 355 423 491 559 628 696
62 326 326 428 530 632 734 836
70 421 553 685 816 948
81 567 744 921 1,099
90 691 906 1,122 1,337
101 874 1,147 1,420
120 1,247 1,637
∗
To convert feet to meters, multiply by 0.3048.
†
Capacity in thousands of gallons. To convert gallons to cubic meters, multiply by 0.0037854.


TABLE 1-2 Capacities of Glass-Coated, Bolted-Steel Reservoirs and Standpipes

FIGURE 1-8
Welded-steel
standpipe with
decorative
pilasters.
and water is pumped from the lower portion of the standpipe into the
system.
As with reservoirs, steel standpipes are covered with a roof struc-
ture and may be provided with ornamental trim. Standard accessories
may include shell and roof manholes, roof vent(s), a fixed outside
ladder, and connections or pipes as required. Inside ladders are not
recommended in locations where freezing weather can be expected.
Roof Designs for Reservoirs and Standpipes

The emphasis on making steel water reservoirs and standpipes at-
tractive as well as functional has led to the development of a wide
variety of roof designs. Alternative roof styles for welded tanks in-
clude conical, toriconical, umbrella, dome, and ellipsoidal designs.
Some are column supported; others are self-supporting. Bolted-steel
tanks are usually provided with conical roofs or may be furnished
with an aluminum geodesic dome. Column-supported roof struc-
tures are not usually used on steel standpipes taller than 50 ft (15 m).
Whichever design is selected, it is particularly important to design
any rafters, trusses, columns, stiffeners, and connections to minimize
potential corrosion sites. All interfaces and connections of such mem-
bers should be analyzed for their corrosion potential, and protective

10 Chapter One
Roof vent
Roof plate
Roof manholes
Overflow pipe
Capacity level Weir box

(optional)
Painter’s trolley rail
Approved ladder, cage,

platform, or safety devices
complying with Occupational
Safety and Health Act
Tank bottom
crowned at center
Inlet– outlet
(optional)
Base elbow or Splash
Shell manholes valve pit pad
(two required)
Concrete
foundation Sand pad
Crushed rock or gravel
Compacted backfill
or undisturbed soil
FIGURE 1-9 Cross-sectional view of typical welded-steel standpipe.

(Source: AWWA Manual M42, Steel Water-Storage Tanks)
coatings should be applied to all surfaces deemed necessary from a

cost/benefit standpoint.
Column- and Rafter-Supported Cone Roofs

The column- and rafter-supported roof (Fig. 1-12) is generally the most
economical for a reservoir. The roof has a minimum slope for adequate
drainage and provides easy access to the manhole for interior inspec-
tion. Column loads are spread to a safe limit by column bases, and
concrete footings under the columns are not usually required.
A modification of this design incorporates a transition from the
shell plate to the roof plate that is a smooth curve rather than a sharp
break. This transition, or knuckle plate, is a dished or rolled section that
usually requires a stiffener at the rafter attachment point (Fig. 1-13).

FIGURE 1-10 Bolted-steel standpipe.
Approved ladder, cage, and platform

complying with Occupational Safety
and Health Act
Roof walkway
and guard rail
Internal
overflow
funnel
Gravity
Roof access Top
ventilator
elbow
Overflow
pipe
Floor sloped toward

outlet pipe
Inlet–outlet
Splash
(optional)
pad
24-in. (0.6-m) round Grade

access door level
FIGURE 1-11 Cross-sectional view of bolted-steel standpipe. (Source: AWWA


12 Chapter One

(US gal) (m3 ) (ft [in.]) TCL (ft [in.]) (m) TCL (m)
50,000 189 14 [9] 40 [0] 4.5 12.2
60,000 227 16 [2] 40 [0] 4.9 12.2
75,000 284 18 [0] 40 [0] 5.5 12.2
100,000 379 19 [0] 48 [0] 5.8 14.6
125,000 473 21 [3] 48 [0] 6.5 14.6
150,000 568 23 [3] 48 [0] 7.1 14.6
200,000 757 24 [10] 56 [0] 7.6 17.1
250,000 946 27 [9] 56 [0] 8.5 17.1
300,000 1,136 28 [5] 64 [0] 8.7 19.5
400,000 1,514 32 [10] 64 [0] 10.0 19.5
500,000 1,893 34 [7] 72 [0] 10.5 21.9
600,000 2,271 37 [10] 72 [0] 11.5 21.9
750,000 2,839 42 [6] 72 [0] 12.9 21.9
1,000,000 3,785 46 [4] 80 [0] 14.1 24.4
1,500,000 5,678 56 [9] 80 [0] 17.3 24.4
2,000,000 7,571 65 [6] 80 [0] 20.0 24.4
2,500,000 9,464 69 [10] 88 [0] 21.3 26.8
3,000,000 11,356 76 [6] 88 [0] 23.3 26.8
4,000,000 15,142 84 [6] 96 [0] 25.8 29.3
5,000,000 18,927 94 [6] 96 [0] 28.8 29.3

Note: TCL = top capacity level.
TABLE 1-3 Capacities and Sizes of Typical Welded-Steel Standpipes
Self-Supporting Dome Roof and Umbrella Roof

Steel self-supporting roofs are constructed of plates that are butt
welded, lap welded, or lap bolted. They are supported directly on
the top angle and shell plate. This type of roof is used where an un-
cluttered interior and smooth exterior appearance are desired. Dome-
roof sections are pressed to form a spherical shape. Umbrella roofs are
formed to a radius in one direction only, forming chords like the cloth
between the spines of an umbrella (Fig. 1-14).
Structural stiffeners may be used internally on large-diameter
roofs to avoid excessive plate thickness on welded or bolted tanks.
Sometimes steel trusses may be used to support the roof, but these

12 in. (0.3 m) Vent 3 ⁄16-in. (4.7-mm)

Top angle
3⁄4 in. (19 mm) lap-welded roof plate
Girders are Capacity

required when level
more than one
Channel column is used One or more
rafters
supporting
columns
Butt-welded
tank shell
1⁄4-in. (6.4-mm)
Column
base lap-welded
bottom plate
FIGURE 1-12 Tank with column- and rafter-supported cone roof. (Source:
AWWA Manual M42, Steel Water-Storage Tanks)
should be avoided if possible, because they may create corrosion

problems. In addition, the trusses should be kept above the water
line to prevent damage by ice and accelerated rates of corrosion.
A modification of the self-supporting dome is the toriconical roof.
This consists of a rolled or pressed knuckle and a higher-pitched self-
supporting center.
Aluminum dome roofs are sometimes erected on bolted-steel or
welded-steel tanks. These aluminum domes are usually constructed
Knuckle plate 3/16 in. (4.7-mm)

12 in. (0.3 m)
¾ in. (19 mm) lap-welded roof plate
Capacity
Channel level
Radius rafter
One or more
supporting
columns
Butt-welded
tank shell
Column ¼-in. (6.4-mm)
base lap-welded
bottom plate
FIGURE 1-13 Column- and rafter-supported roof with knuckle. (Source: AWWA

14 Chapter One
3 ⁄16-in. (4.7-mm) minimum thickness

lap- or butt-welded Vent
roof plate Cap plate
Top
angle
Capacity level
Sp 1.2 0 D
= .8
he D . m
ric . m in.
0
al ax
ra .
di
us
Butt-welded
tank shell
1⁄4-in. (6.4-mm)
lap-welded
bottom plate
FIGURE 1-14 Self-supporting dome roof or umbrella roof. (Source: AWWA

of triangulated space truss (geodesic) panels. The dead weight of these

domes is usually 3 lb/ft2 (143 N/m2 ) or less, compared with 3.8 lb/ft2
(181 N/m2 ) for a bolted-steel roof and 7.6 lb/ft2 (364 N/m2 ) for a
welded-steel roof.
Self-Supporting Ellipsoidal Roof

The self-supporting ellipsoidal roof is not a true ellipse, but it is formed
with two radii yielding major- and minor-axis proportions of approxi-
mately 2:1. The transition from shell to roof is a smooth unbroken curve
(Fig. 1-15). This roof design is suitable for large- and small-diameter
reservoirs and standpipes. On tanks 50 ft (15 m) in diameter or less,
the roof is usually free of internal structural members. Larger-diameter
tanks usually have radial and circumferential stiffening members or
rafters, which may be subject to corrosion problems if they are not
properly designed or maintained.
Self-Supporting Cone Roof
An inexpensive and very functional type of roof for small-diameter
reservoirs and standpipes is the self-supporting cone roof without in-
ternal structural members. This roof is usually too steep to walk on.
Access to manholes and vents by a roof ladder or steps and handrail
should be provided. All means of access should be designed individ-
ually and installed to comply with current standards.

1⁄4-in. (6.4-mm) Area above Vent

minimum thickness capacity level
and butt welded in area may be lap welded
filled with water
le
ck
u
Capacity level
Kn
Butt-welded
tank shell
1⁄4-in. (6.4-mm)
lap-welded
bottom plate
FIGURE 1-15 Self-supporting ellipsoidal roof. (Source: AWWA Manual M42,

Steel Water-Storage Tanks)
Elevated Tanks
An elevated steel water tank has two primary components: the tank it-
self and its supporting structure. Such tanks are ordinarily used where
there is insufficient elevated terrain to ensure distribution of water at
suitable pressure by gravity. These tanks are of welded construction.
Elevated tanks can be categorized into several different types. The
various diameters and head ranges for the tanks described in the re-
maining figures and tables in this chapter are only representative and
may vary with individual fabricators. Specific diameter/head range
combinations should be determined by the tank fabricator within the
limits indicated in the tables. Height should be specified by the pur-
chaser as the dimension between the top of the foundation and the top
capacity level of the tank. Further dimensions, which are a function of
the fabricator’s standard, should not be specified. To minimize cost,
desired operating ranges should be specified to fall within standard
available tank dimensions. However, individual operating needs may
dictate nonstandard operating ranges.
Multiple-Column Elevated Tanks

Small-Capacity Elevated (Double-Ellipsoidal) Tanks
The small-capacity multiple-column elevated (or double-ellipsoidal)
tank has a cylindrical sidewall, an ellipsoidal bottom and roof, and a

16 Chapter One
FIGURE 1-16
Double-ellipsoidal
tank. (Photo: Gay
Porter DeNileon,
AWWA)
top capacity level (TCL) in the roof several feet or meters above the
top of the cylindrical shell. Although in the past they were constructed
in capacities up to 1 mil gal (3.8 ML), today, double-ellipsoidal tanks
are typically constructed only in capacities of 200,000 gal (760,000 L)
or less. See Figs. 1-16 and 1-17 for a photo and a cross-sectional view
of a small-capacity elevated (double-ellipsoidal) tank. Table 1-4 gives
capacities and sizes of typical double-ellipsoidal elevated tanks.
Medium-Capacity Elevated Tanks

For medium-capacity multiple-column elevated tanks, the toroellip-
soidal design provides a lower initial cost by using the strength of steel
most efficiently. The features used (torus bottom and ellipsoidal roof)
cause the central riser to support, as well as contain, a considerable
portion of the stored water, while the major portion of the steel bottom
acts as a membrane in tension. These tanks usually have a capacity
between 200,000 gal (760,000 L) and 500,000 gal (1.9 ML). See Figs.
1-18 and 1-19 for a photo and a cross-sectional view of a medium-
capacity elevated tank. Table 1-5 gives capacities and sizes of typical
medium-capacity elevated tanks.

Diameter
Head
range
Balcony or
stiffening girder
Purchaser to specify
As required
6 in. (0.15 m) min.
FIGURE 1-17 Cross-sectional view of double-ellipsoidal tank. (Source: AWWA

Large-Capacity Multiple-Column Elevated Tanks

Large-capacity elevated tanks (>500,000 gal [>1,893 m3 ]) provide eco-
nomical service for communities that need to store a substantial vol-
ume of water. Lower operating and pumping costs are ensured be-
cause of the low head range, which achieves minimum variation of wa-
ter pressure throughout the system. See Figs. 1-20 and 1-21 for a photo
and a cross-sectional view of a large-capacity elevated tank. Table 1-6
gives capacities and sizes of typical large-capacity elevated tanks.
Pedestal Elevated Tanks

Small-Capacity Single-Pedestal Tanks
The single-pedestal spherical tank is widely favored for smaller-
capacity tanks when appearance is a concern. The gracefully flared
base contains sufficient space for pumping units and other operating
equipment, a feature common to all pedestal-type vessels.

18 Chapter One

Diameter Head Diameter Head
(US gal) (m3 ) (ft) Range (ft) (m) Range (m)
25,000 95 18–20 12.5–15.5 5.5–6.1 3.3–4.7
30,000 114 18–20 15.0–16.5 5.5–6.1 4.6–5.0
40,000 151 22–23 15.0–17.0 5.7–7.0 4.6–5.2
50,000 189 22–24 18.0–20.0 6.7–7.3 5.5–6.1
60,000 227 22–25 19.0–23.0 6.7–7.6 5.3–7.0
75,000 284 26–30 16.0–24.0 7.9–9.1 4.9–7.3
100,000 379 23–30 20.0–25.0 3.5–9.1 6.1–7.6
125,000 473 30–32 23.0–28.0 9.1–9.7 7.0–8.5
150,000 568 32–34 24.5–29.5 9.7–10.4 7.5–9.0
200,000 757 36–38 28.0–29.5 11.0–11.6 8.5–9.0
TABLE 1-4 Capacities and Sizes of Typical Double-Ellipsoidal Elevated Tanks
FIGURE 1-18
Medium-capacity
welded-steel
elevated tank.
(Photo: Gay Porter
DeNileon, AWWA)

Head
Range
Balcony or
stiffening girder
As required
6 in. min.
FIGURE 1-19 Cross-sectional view of medium-capacity, torus-bottom welded-

steel elevated tank. (Source: AWWA Manual M42, Steel Water-Storage Tanks)
Ladders to the container and roof are inside to protect against

unauthorized access. These tanks are usually constructed in capacities
of 200,000 gal (760,000 L) or less. See Figs. 1-22 and 1-23 for a photo and
a cross-sectional view of a small-capacity single-pedestal tank. Table
1-7 gives capacities and sizes of typical small-capacity single-pedestal
tanks.
Small-capacity elevated tanks are also constructed as various com-
binations of cones and cylinders. An alternative design is shown in
Fig. 1-24.
Large-Capacity Single-Pedestal Tanks

The tubular supporting pedestal gives the large-capacity single-
pedestal tank a distinctively contemporary look. Large capacities (0.2
to 2 mil gal [0.76 to 7.6 ML]) are provided by this low-head-range
spheroidal tank design. See Figs. 1-25 and 1-26 for a photo and a
cross-sectional view of a large-capacity single-pedestal tank. Table 1-8

20 Chapter One

(US gal) (m3 ) (ft) TCL (ft [in.]) (m) TCL (m)
200,000 757 36–38 28 [30] 11.0–11.6 8.5–9.1
250,000 946 38–40 28 [33] 11.6–12.2 8.5–10.1
300,000 1,136 43–45 28 [31] 13.1–13.7 8.5–9.4
400,000 1,514 46–50 30 [36] 14.0–15.2 9.1–11.0
500,000 1,893 50–56 29 [38] 15.2–17.1 8.8–11.5
600,000 2,271 51–0 40 [0] 15.6 12.2
57–0 32 [0] 17.4 9.8
750,000 2,839 56–65 34 [45] 17.1–19.8 10.4–13.7
1,000,000 3,785 64–65 45 [46] 19.5–19.8 13.7–14.0
TABLE 1-5 Capacities and Sizes of Typical Medium-Capacity Elevated Tanks
FIGURE 1-20 Large-capacity elevated tank. (Photo courtesy of Landmark

Structures, Inc.)

Diameter
Head
range
As required
6 in. (0.15 m) min.
FIGURE 1-21 Cross-sectional view of large-capacity, multicolumn elevated

tank. (Source: AWWA Manual M42, Steel Water-Storage Tanks)

500,000 1,893 60–65 24–25 18.3–19.8 7.3–7.9
600,000 2,271 65–70 24–25 19.8–21.3 7.3–7.9
750,000 2,839 70–76 25–30 21.3–23.2 7.6–9.1
1,000,000 3,785 75–87 25–35 22.9–25.5 7.6–10.7
1,500,000 5,678 91–98 30–35 27.7–29.9 9.1–10.7
2,000,000 7,571 105–106 34–36 32.0–32.3 10.4–11.0
2,500,000 9,464 108–117 39–41 32.9–35.7 11.9–12.5
3,000,000 11,356 119–127 35–40 36.3–38.7 10.7–12.2
TABLE 1-6 Capacities and Sizes of Typical Large-Capacity Welded-Steel

Elevated Tanks

22 Chapter One
FIGURE 1-22
Spherical
single-pedestal
tanks give pleasant
silhouette. (Photo:
Walter Baas,
AWWA)
gives capacities and sizes of typical large-capacity single-pedestal

tanks.
Modified Single-Pedestal Tanks

The attractive modified single-pedestal tank has a central support col-
umn (usually fluted to give structural rigidity) that encloses the riser
pipe, overflow pipe, and access ladder to the tank roof. The support
column may be constructed of steel or concrete. The space within the
column can provide multistory usable floor space for pumping, stor-
age, and office facilities. Although available in all capacities, these
tanks are not usually constructed in capacities less than 500,000 gal
(1.9 ML). See Figs. 1-27 and 1-28 for a photo and a cross-sectional view
of a modified single-pedestal tank. Table 1-9 gives capacities and sizes
of typical modified single-pedestal tanks.
Composite Elevated Tanks

Composite elevated tanks are of an attractive design that uses the
best design features of steel and concrete. Concrete, which is excellent
for compression loads, is used as the support column for the steel
bowl. The concrete has the advantage of requiring either no painting

Diameter
Head
range
As required
6 in. (0.15 m) min.
FIGURE 1-23 Cross-sectional view of small-capacity spherical single-pedestal

tank. (Source: AWWA Manual M42, Steel Water-Storage Tanks)

25,000 95 19–20 15–17 5.8–6.1 4.6–5.2
30,000 114 20–21 15–18 6.1–6.4 4.6–5.5
40,000 151 21–23 19–22 6.4–7.0 5.8–6.7
50,000 189 23–24 19–23 7.0–7.3 5.8–7.0
60,000 227 24–26 22–24 7.3–7.9 6.7–7.3
75,000 284 25–28 23–27 7.9–8.5 7.0–8.2
100,000 379 29–30 25–30 8.8–9.1 7.6–9.1
125,000 473 31–33 27–32 9.4–10.0 8.2–9.7
150,000 568 33–34 30–34 10.1–10.4 9.1–10.4
200,000 757 36–38 36–38 11.0–11.6 11.0–11.6
TABLE 1-7 Capacities and Sizes of Typical Small-Capacity Single-Pedestal

Tanks
23
24 Chapter One
FIGURE 1-24 Alternative single-pedestal tank design.
or a low-cost exterior coating for aesthetic purposes. The steel bowl

construction is similar to that found on the fluted-column tanks; the
bowl can be built with either a cone or a domed roof. The most common
designs use a domed concrete floor with a steel liner. Commonly built
to store between 750,000 gal and 2 mil gal (2.8 and 7.6 ML), these tanks
provide many of the benefits of a fluted-column tank with significantly
less area that requires painting, thereby reducing maintenance costs.
The diameter of the concrete column is generally somewhat smaller
(30 to 60 ft [9 to 18 m]) than for a fluted-column tank, so the area in the
column for other uses is reduced. See Figs. 1-29 and 1-30 for a photo
and cross-sectional view of a composite elevated tank.
Locating, Sizing, and Selecting a Water Tank

Locating, sizing, and selecting a water-storage tank involve the eval-
uation of several design considerations and require an awareness of
zoning and other regulations. The purpose of this section is to dis-
cuss these considerations and to provide the reader with a checklist
to work through in the effort to arrive at a reasonable solution.

FIGURE 1-25 Large-capacity single-pedestal elevated tank. (Photo courtesy of

Tnemec/STI/SPFA)
Locating a Water Tank

Generally, locating tanks depends on where people are living now and
where future neighborhoods will be built within the area served by
the water system. In addition, numerous other conditions can signif-
icantly influence the choice of a suitable site and therefore the overall
cost of the tank project. Answers to the following basic questions must
be determined and considered when selecting a location for a new
water-storage tank.
Hydraulics
r What are the maximum and minimum pressures that you
want to provide the end users?
r Is it better to pump or use gravity flow to provide the needed
pressure?
r What are the local utility costs of pumping during daily and
peak demand periods?
Proximity to Users
r Where is the growth in the community taking place now and
projected to be in the future?
r Is land available in the area of future growth?

26 Chapter One
Diameter
Head
range
Access
tube
As required
6 in. (0.15 m) min.
FIGURE 1-26 Cross-sectional view of large-capacity single-pedestal elevated

tank. (Source: AWWA Manual 42, Steel Water-Storage Tanks)

200,000 757 40–42 27–30 12.2–12.8 8.2–9.1
250,000 946 43–47 25–32 13.1–14.3 7.6–9.7
300,000 1,136 46–48 30–33 14.0–14.6 9.1–10.1
400,000 1,514 50–53 30–40 15.2–16.1 9.1–12.2
500,000 1,893 55–60 30–40 16.3–18.3 9.1–12.2
750,000 2,839 64–66 38–42 19.5–20.1 11.6–12.3
1,000,000 3,785 74–78 35–40 22.5–23.8 10.7–12.2
1,250,000 4,732 76–80 40–45 22.9–24.4 12.2–13.7
1,500,000 5,678 85–90 45–50 25.9–27.4 13.7–15.2
2,000,000 7,571 90–95 50–55 27.4–29.0 15.2–16.3
TABLE 1-8 Capacities and Sizes of Typical Large-Capacity Single-Pedestal

Tanks

FIGURE 1-27 Folded-plate design of a modified single-pedestal tank support.

(Photo courtesy of Tnemec/STI/SPFA)
Acquiring Land
r What is the cost of the tank site being considered? Is the land
even available?
r What is the cost of connecting water mains and permanent
electrical power at each site being considered?
Zoning
r Is a zoning map available, and are the potential sites zoned
to allow a tank project?
Federal Aviation Administration (FAA)
r Would the FAA allow a tank at the required height to be built
on the potential site?
r Are obstruction lights or FAA painting required on the tank
at the potential site?
Size of Site
Is the site large enough for
r Erection equipment, steel storage, staging operations, ground
assembly, and crane operations with a safe and adequate dis-
tance for items that may be dropped from the tank during
erection?

28 Chapter One
Diameter
Head
range
As required
Fluted
column
FIGURE 1-28 Cross-sectional view of a modified single-pedestal tank.

250,000 946 41–43 29–31 12.5–13.1 8.8–9.4
300,000 1,136 43–45 29–31 13.1–13.7 8.8–9.4
500,000 1,893 49–64 30–39 14.9–19.5 9.1–11.9
750,000 2,839 63–65 37–40 19.2–19.8 11.3–12.2
1,000,000 3,785 73–78 35–42 22.2–23.8 10.7–12.8
1,250,000 4,732 76–80 40–45 22.9–24.4 12.2–13.7
1,500,000 5,678 85–87 39–46 25.9–26.5 11.9–14.0
2,000,000 7,571 97–102 38–46 29.6–31.1 11.6–14.0
2,500,000 9,464 107–110 43–45 32.6–33.5 13.1–13.7
3,000,000 11,356 109–120 40–45 33.3–36.6 12.2–13.7
TABLE 1-9 Capacities and Sizes of Typical Modified Single-Pedestal Tanks

FIGURE 1-29 Composite elevated tank.
r Maintenance of the tank and piping after completion?

r Abrasive blasting and painting of the tank now and in the
future?
Topography
Does the site have—or can it be made to have—good drainage
to ease construction operations and minimize standing water
around the completed tank?
Access to Site
r Is the site accessible on public roads by concrete and large
semitrailer tractor rigs?
r Is there an access road or temporary easement to the site?
Will permission be given to build a road? Who will pay for the
road? Will it be a permanent or temporary road? If temporary,
will it be necessary to remove it at the end of the project?
Soil Conditions
r Is the soil bearing strength at the bottom of the tank founda-
tion adequate to support the tank without requiring expen-
sive deep foundations?

30 Chapter One
Upper roof
cone
High water
line
Steel bottom Access tube

plate
Low water line Lower cone
Concrete
support dome
Concrete column
Note: Not to scale.
FIGURE 1-30 Cross-sectional view of composite elevated tank.
r Where is the water table? Will the foundation need to be de-

watered during construction?
r Is the earth firm enough to support construction equipment
during normal weather conditions or will gravel, crane mats,
and other earth-stabilizing methods be required?
Hazards and Construction

r Are there power lines or other obstructions above or beside
the site or proposed access road that would interfere with the
safety of site traffic, construction, painting, or maintenance
operations? Will a power line be closer than 40 ft (12.2 m)
from the tank?
r Are there underground obstructions such as gas lines, sewers,
or buried electrical or telephone lines? Were there mines or
burial grounds on this site?

r If pile driving is required, will it disturb or cause failure of or
damage to neighboring foundations or other structures?
r Will pile driving, excavation, steel erection, or abrasive blast-
ing cause noise unacceptable to a neighbor such as a school,
hospital, or nursing home?
r Will the tank be in an area frequented by small children or
vandals and, if so, could this be mitigated by site fencing?
Environmental Assessment
r Has an environmental assessment been completed on the site?
r What agencies, forms, and permits may be required, and how
long will approvals take?
NIMBY (Not in My Backyard)
r Will the tank obstruct the view of historical landmarks or
other items of concern to the citizens?
r How sensitive are the neighbors to having a tank in close
proximity?
Determining answers to these questions can help you to better
analyze and compare costs of alternate sites, so you can select the
most desirable location for your new tank. Additionally, you will want
to understand and consider the following criteria during your site
selection.
Hydraulics
Other issues that affect site selection include the required pressure
at hydrants and residences, the required site elevation, compatibility
with the distribution system, the geographic size and location of the
demand area, and the tank’s proximity to the water supply. Rules of
thumb for required water pressure are shown in Table 1-10. Check the
local standards or codes for more specific requirements.
One hundred sixty-two US and Canadian water utilities re-
sponded to an AWWA network modeling survey that requested the
actual minimum and maximum distribution system pressures that
Pressure
Location (psi/kPa) Comments
At hydrants 35/241 20-psi (0.138-kPa) minimum at
during fire flow other fire hydrants not directly
conditions serving the fire
Residential 50–75/ Higher pressures may need to
0.345–0.517 use a pressure-reducing valve
TABLE 1-10 Required Water Pressure

32 Chapter One
Min. psi Max. psi

100 100
4% > 60 psi
14% > 170 psi
10% 50–59 psi
80 80 10% 150–169 psi
28% 40–49 psi

12% 130–149 psi
60 60
62%
Percent
Percent
25% 110–129 psi
55%
34% 30–39 psi
40 40
18% 90–109 psi
20 20
23% 20–29 psi 15% 70–89 psi
< 20 psi 6% < 70 psi

0 1% 0
Minimum pressure Maximum pressure
FIGURE 1-31 Pressure ranges for utilities.
they provided. Figure 1-31 shows the percentage of utilities in each

pressure range. If the pressures provided are more than 75 psi (0.517
kPa), it may be necessary to provide a pressure-reducing valve to pre-
vent home appliances from being overpressurized.
The required pressure can be provided through pumping, gravity
flow, or a combination of the two. How pressure is provided depends
on the sites available and the type of tank to be used.
Pumping with Ground Storage Tanks

Pumping will be required if a ground storage tank is used where the
topography is relatively flat throughout the service area and a higher-
elevation site is unavailable.
Gravity Flow with Ground Storage and Elevated Tanks

The required pressure can also be obtained by building a ground stor-
age tank on a hill or at higher elevation above the demand area so that
gravity flow provides the pressure, much like a water cooler.
An elevated tank provides the required pressure by raising the
water storage height up to an elevation above the demand area so
that gravity can provide the pressure. Costs can be lessened if the
elevated tank is also constructed on a hill site or at higher elevation.
This not only lessens the necessary height of the tank but also can
reduce its cost.

Gravity Flow and Pumping with a Standpipe

Water could also be stored in a standpipe (a tall cylindrical tank) where
the topography is relatively flat throughout the service area and a hill
or higher-elevation site is unavailable. In a full standpipe, the upper-
most one-third of water stored provides effective pressure for gravity
flow. If the tank is two-thirds full, the upper half of the water would
provide emergency pressure. In a tank only one-third full, the water
provides little or no pressure (i.e., ineffective pressure) and would
have to be pumped to be used.
Much like with an elevated tank, costs can be saved if the stand-
pipe is constructed on a higher-elevation site or hill. This not only
lessens the necessary height of the tank but also can reduce its cost.
Gravity Flow Height Calculations

Following is an example of how to calculate the minimum height at
which to store water to provide an assumed minimum pressure for
residential use through gravity flow. (Check your local standards or
codes.)
Height (for 50 psi [345 kPa] minimum)
50 psi [345 kPa]
=
(62.4 lb/ft3/144 sq in./sq ft)
= 115.4 ft [35.1 m] (≈ 115 ft [≈ 35 m])
or
Height (for 50 psi [345 kPa] minimum)
50 psi [345 kPa]
= = 115.4 ft [35.1 m]
(62.4 lb/ft3/144 sq in./sq ft)
To this calculated height, add the additional height required to

meet the friction loss of the water in the distribution piping.
Alternatively, one can use a conversion chart to find the required
height at which to store the water to provide the pressure needed.
Figure 1-32 shows how various types of tanks provide this pressure
using gravity flow.
Pumping Versus Gravity Flow

Pumping If a site with an increased elevation of at least 115 ft (35 m)
above the service area cannot be found, the only option with a ground
storage tank is to use pumping to provide the required pressure. If you
are going to pump, you should be aware that water demand varies
throughout the day. As such, you will have to use a variable-speed
pump.
A typical water usage graph (Fig. 1-33) shows the filling of a tank
during the night and early morning hours when demand is low. The

34 Chapter One
Reservoir
115 ft (35 m)
Elevated tank Standpipe
FIGURE 1-32 Providing pressure using gravity flow.
tank is emptied during the day; water demand peaks sometime be-
tween 5 p.m. and 9 p.m.
Electric utilities charge more for electricity during their peak de-
mand period (see sample rates in Fig. 1-34). By overlaying the sample
electric rates on the water usage graph (Fig. 1-35), one can see that the
peak demands for electricity and water occur about the same time of
day. Using these data, one can make the following calculations:
r Peak demand (5 p.m. to 9 p.m.) pumping costs: $0.1175/kWh

average utility cost to pump half of the daily water demand
to end users.
Peak demand
Constant pumping rate Emptying tank
Filling tank
Usage rate
Midnight 6:00 A.M. Noon 6:00 P.M. Midnight

Time
FIGURE 1-33 Typical water usage.

$0.14
$0.12
Average cost per kilowatt-hour
$0.10
$0.08
$0.06
$0.04
$0.02
$0.00
Midnight 3:00 A.M. 6:00 A.M. 9:00 A.M. Noon 3:00 P.M. 6:00 P.M. 9:00 P.M. Midnight
Time
FIGURE 1-34 Sample electric rates.
r Nonpeak demand pumping costs: $0.1080/kWh average util-

ity cost to pump the other half of the daily water demand to
end users.
r Tank filling costs: $0.0675/kWh average daily utility cost to
fill the tank by pumping.
In this case, utility costs during peak demand are almost 75 percent
more than the cost of the average rate used to fill the tank, while even
nonpeak costs are about 60 percent more. These calculations should
be modified for your system using your local daily water usage and
utility rates. Regardless of the local factors, pumping during peak
$0.14
Peak demand
$0.12
Average cost per kilowatt-hour
Emptying
tank
$0.10
Constant pumping rate
$0.08
Filling tank
$0.06
Usage rate
$0.04
$0.02
$0.00
Midnight 6:00 A.M. Noon 6:00 P.M. Midnight
Time
FIGURE 1-35 Higher rates during peak demand.

36 Chapter One
electricity rates to meet peak water demand is usually more expensive

than gravity flow and can become quite costly over time.
Additionally, if you lack sites with hills or higher elevations and
choose to pump to meet the pressure and water demand, incorporate
the following initial and lifetime costs into your present-value analysis
as follows:
r The additional daily costs of pumping over and above gravity
flow (peak and nonpeak)
r The added cost of a variable-speed pump (usually required;
larger than the constant-speed pump used at night for a
gravity-flow tank)
r Cost of a backup pump or pumps
r Cost for additional piping and controls for the backup
pump(s)
r Cost of backup generator
r Expense of enlarging the pump building to house the addi-
tional pumps and piping
r Cost to maintain and replace all of these as needed.
Often, when these additional costs are considered, it is most likely

that the extra initial costs to provide gravity flow may actually be a
more cost-effective solution over time.
Gravity flow One can save these peak-demand electricity costs by peak
shaving. To peak shave, start by locating a ground storage tank on the
side of a hill, or build an elevated tank or standpipe. A smaller pump
can then be used to pump the water up into the tank during the night
and early morning at a constant rate when electricity rates are much
lower. Then, during the demand period, water can be provided at the
needed pressure by using gravity flow. This avoids the much higher
electricity rates during this time period and allows use of a smaller,
less costly pump.
Because of these advantages, gravity flow is the preferred method
of providing water pressure. If possible, place the tank on a hill or
elevate it to take advantage of this method.
The ideal location: For any type of storage tank, the ideal location
is on a hill that is in the middle of the demand area and is owned by
the community.
Proximity to Users
When choosing a site for a new water-storage tank, the prospective
tank owner should consider the growth in residential demand (single-
family, multifamily, and high-rise structures) and commercial demand
(industry, schools, and hospitals). A new residential development on

the north side of the service area and a new tank on the south side
would result in very little water pressure for residents of the new
development. The ideal situation is to construct a new water-storage
tank in the service area before the area experiences population growth
and buildup. This way, you have a better chance to get the right piece
of land at the right time and at the right price.
Acquiring Land
When acquiring land, the prospective tank owner must consider the
availability and suitability of the land for a tank project; the costs for
the land, required support utilities, and the length of connections to
the existing distribution system; and the surrounding conditions.
NIMBY (not in my back yard!) One of the biggest issues that a water
utility can face when attempting to locate a new water-storage tank is
the public concern of NIMBY! Despite these concerns, even the most
appearance-conscious communities can agree to a mutually beneficial
solution to this stumbling block. The following are some successful
approaches to be used in overcoming public concern:
r Encourage community involvement: When choosing the style of

the tank, let the citizens express their concerns and provide in-
put. In some communities, citizen groups have used contests
to select the color scheme of the tank exterior or the lettering
and logo design.
r Educate the citizenry: Explain the reasons the new tank is
needed and the beneficial effect it will have on them person-
ally (for example, improved water pressure and fire protec-
tion). Demonstrate how improved fire flow will affect insur-
ance rates, assure them of the safety record of water-storage
tanks, and explain the anticipated maintenance cycle.
r Help the public visualize the completed tank: Using an artist’s con-
ception, computerized renderings, and a digital photograph
of the site, compile an image that shows the community what
the finished water-storage tank will look like.
Zoning Regulations
Once a site has been located, check on the zoning of the selected site
to ensure that it is currently zoned for this use or can be rezoned.
Obtaining proper zoning for a water tank is typically more difficult in
a residential area than in an industrial area or in an area near public
facilities such as schools, government property, and airports.
Often, schools are built in the areas of population growth, and
the school yard may make a good site for a tank. There are many
aesthetically pleasing tank styles that limit access.

38 Chapter One
FAA Considerations
Forms must be completed and filed with the Federal Aviation Ad-
ministration (FAA) to establish whether a tank can be built on the
chosen site at the required height. The FAA is concerned about any
obstruction to its airspace 200 ft (61 m) above ground level and
any obstruction within an approach pattern to an airport runway.
Lengths of approach pattern vary depending on the size of the air-
port, the length of the runway, and the direction of the runway, as
follows:
r Large airport: No obstruction that exceeds a 100:1 surface
within 20,000 ft (6.1 km) of an airport having at least one
runway >3,200 ft (>975 m).
r Small airport: No obstruction that exceeds a 50:1 surface within
10,000 ft (3 km) of an airport whose longest runway is <3,200
ft (<975 m).
r Heliport: No obstruction that exceeds a 25:1 surface within
5,000 ft (1.5 km) of a heliport.
If the FAA determines that the tank will be in the approach pattern,
the tank may have to be equipped with aviation lighting or painted
in a special aviation warning paint scheme. The most common of the
aviation paint schemes is the red-and-white checked pattern found
on tanks near airports.
The following circulars, forms, and information regarding ob-
struction evaluation and airport airspace analysis are available on the
FAA Web site (https://oeaaa.faa.gov):
r For information on proposed tank construction projects, con-

sult “Proposed Construction or Alteration of Objects that may
Affect the Navigable Airspace” (Advisory Circular 70/7460-
2K).
r Standards for marking and lighting tanks and other structures
are provided in “Obstruction Marking and Lighting” (Advi-
sory Circular 70/7460-1K).
r “Notice of Proposed Construction or Alteration” (Form 7460-
1) should be completed by the tank owner before the site is
even purchased and certainly prior to construction. The form
can now be completed and submitted online at the FAA Web
site. Information required includes latitude, longitude, loca-
tion marked on a US Geological Survey (USGS) map, elevation
of site (mean sea level), and the greatest height of any part on
the tank, including handrails or antennas upon completion.
Once the FAA has reviewed the information on the form, it
will make a determination on the proposed tank and location

and post it online. The determination may be one of the fol-

lowing:
r A tank can be built on this site at the height requested.
r A tank cannot be built on the proposed site at all.
r A tank can be built on this site but not at the height re-
quested.
r A tank can be built on this site at the height requested,
but will require an obstruction light and/or obstruction
marking.
r A tank can be built on this site but not at the height re-
quested and will require an obstruction light and/or ob-
struction marking.
r “Supplemental Notice of Actual Construction or Alteration”
(Form 7460-2) is usually completed by the tank contractor. It
must be submitted 30 days prior to the start of tank erection
and requires information similar to that requested on Form
7460-1, except for the following:
r Indicate the start and completion dates of the construction.
r Must indicate the greatest height of the tank or equipment
during construction. Often the tank contractor uses a der-
rick with a boom to erect the tank. The height of the derrick
and boom may actually exceed the maximum height of the
tank on completion. The FAA will want to know this and
may actually require the tank contractor to install an ob-
struction light at the tip of the boom to alert pilots to the
tank’s location.
Size of Site
Tank constructors recommend that the distance from the edge of the
tank to the site boundary be a minimum of 50 to 75 ft (15.24 to 22.86 m).
A tank may be constructed on a smaller site, but it will require extra
handling and planning to stage materials in a disciplined sequence.
Eliminating space constraints enables the tank contractor to build the
tank more efficiently and can reduce costs up to a point.
Take into consideration the space needed for the following:
r Material storage during construction
r Erection and painting operations
r Support facilities such as pump houses, valve vaults, and
parking areas
r Future maintenance and repainting
r Placement of tank at safe distance from private property and
utilities.

40 Chapter One
Topography
The tank site can have a major influence on the cost of construction
and on design details for the foundation. The site should allow good
drainage away from the foundation(s), provide a level working sur-
face for construction, and have some type of erosion protection. Stand-
ing or ponding water on the site can add dewatering costs to the project
and may even require changes to the foundation design, leading to
added costs. Consider these added costs when evaluating sites.
Access to Site
Access to the tank site is an important aspect of site selection. Devel-
opers and residents often want the tank to be located in the back of
the development, away from the streets or even in off-road remote lo-
cations. This poses a problem getting the large trucks and equipment
required for construction to the tank site.
Other things that must be considered when assessing site ac-
cess are the distance from paved roads, permanent versus temporary
roads, accessibility by large trucks, and securing temporary easements
for site access during construction, if needed. The best site access is
via a permanent road up to the tank. The most economical means
of achieving this is to put the tank access road in with the original
subdivision roads.
Soil Conditions
A full soil investigation should be conducted before the final site is
chosen and certainly before it is purchased. The soil assessment will
determine whether the soil is adequate to support the tank and its
contents and what type of foundation must be designed. Some sites
may require deep foundations (piles or drilled piers) that could add
significant costs to the design and construction of the tank.
The soil investigation will provide needed information about the
following issues:
r Soil bearing capacity (how much of a load the can soil support)
r Site classification for seismic design
r Excessive or uneven settlement
r Water table elevations
r Rock elevations if present
r Site history
r Substrata conditions
r Slope stability
The depth at which the required soil bearing is obtained to support

the foundation along with the slope stability has implications for the

size of the site required. For example, with a 1:1 slope stability and
a 50-ft (15.24-m)-diameter foundation with the required soil bearing
15 ft (4.57 m) down, the minimum size of the hole for the foundation
would be 15 ft + 50 ft + 15 ft = 80 ft (4.57 m + 15.24 m + 4.57 m =
24.38 m). To this, one would have to add room for digging equipment
and room to store the excavated material on site.
The results of the soils investigation can affect the design and
costs of both the foundation and tank to such an extent that one could
actually save money on the overall project by paying more for a site
with better soil conditions. It is prudent to make the site purchase only
after you have received the results of the soil investigation.
Obstructions/Hazards
Obstructions that must be avoided include overhead power lines,
underground utilities, and existing structures. OSHA (Occupational
Safety and Health Administration) and many tank contractors spec-
ify safe minimum work distances required from power lines depend-
ing on what voltage the lines carry. Construction hazards may in-
clude abrasive blasting, painting, pile-driving vibration, noise, and
fire.
Waves and energy produced by AM antennas comprise one of
the least understood obstructions. AM antennas are typically the tall,
slender, red-and-white antennas that do not have dishes or whip an-
tennas hanging off of them; the entire structure acts as the broadcast
antenna. On the electromagnetic spectrum, AM waves are the longest
waves generated and can be from 656 to 1,968 ft (200 to 600 m) long.
These long waves carry energy. Metal objects used in tanks or tank
construction such as rebar, steel plate, and even crane lines can act
as receiving antennas that collect and store the AM wave energy. If a
grounded worker touches these energized metal objects, the collected
energy is released, possibly shocking the worker and making the work
site unsafe. Whether the AM antenna has any effect on your tank site
depends on how far the antenna is from your tank, what power it is
broadcasting at, and whether it is a directional or nondirectional an-
tenna. At the Federal Communications Commission (FCC) Web site
(www.fcc.gov/mb/audio/amq.html), one can insert the latitude and
longitude of the centerline of the tank (also used in the FCC submit-
tal) and use the “Stations within a Radius” input. The Web site will
indicate if any AM antennas are present. If so, station details will
indicate whether the transmission location is directional or nondirec-
tional. Problems can be present for distances up to 0.6 mile (1 km) for
nondirectional and 1.9 miles (3 km) for directional antennas. If you
encounter an AM antenna that might be a problem and are seriously
considering the site in question, you may want to hire a specialist to
further examine the situation.

42 Chapter One
If the expense and risks of dealing with these obstructions and

hazards adds enough costs to your project, you might be better off
paying more for a site that is free of such obstructions and hazards.
Environmental Issues
Environmental issues that come into play during tank construction in-
clude the protection of vegetation, wildlife, wetlands, and floodplains;
historical landmarks and burial sites; and local wind and snow. Most
states require that a permit request be submitted to the US Environ-
mental Protection Association (USEPA) before construction to identify
any such environmental issues.
Sizing the Tank

Demand
Tank capacity One of the main purposes of a water-storage tank is to
provide storage to meet the water demands of the area it will service.
As a rule of thumb, you can determine your new water-storage tank
capacity by making the following calculation:
Average daily usage (peak and nonpeak) + fire flow requirements
+ added capacity to offset maintenance or pipe breaks
+ additional capacity for future demand = tank capacity
Current average daily use This is the amount of water used on average
in a 24-hour period. Calculate this by determining the average water
usage currently per person and multiply this by the number of people
that the new area currently serves.
r Peak demand: Peak demand typically occurs between 5:00 p.m.

and 9:00 p.m. and is usually half of the current average usage.
r Off-peak demand: This comprises the other half of the average
current daily usage.
Fire flow demand To the current average daily usage add an additional
one-half to one-third of the current average daily usage. This figure
varies depending on the local codes and standards. One should also
check the requirements of the Insurance Service Organization (ISO)
(www.iso.com) and other local standards and codes.
r Maintenance and piping breaks: As a contingency measure, con-
sider adding 10 percent, plus or minus, to provide extra stor-
age if the service area distribution piping has leaks.
r For future demand, project the future population for the ser-
vice area and then multiply that by the current average daily
water use in gallons (liters) per person. An alternate method

ire
Washin ampsh
g ton New H nt
North Vermoia
Minnesota
Montana ine
an Ma
Dakota nsylv
Orego
n Pen
n
South Massachusetts
i sc
ichiga
Idaho
ork
New Y
on
Dakota Rhode Island
Wyoming
si n
Connecticut
Indiana M
Iowa New Jersey
Nevad Nebraska
a OOhhiioo District of Columbia
Utah Illinois Delaware
Ca Colorado Maryland
lifor Kansas ia
nia Missouri Kentu
cky Virgin West
North Virginia
Ten nes see Carolina
Arizona New
as
Oklahoma South
ns
Mexico
Alabama
Missssippi
ka
Carolina
Ar
Georgia Water withdrawals
Lo
Texas
uis
in milion gallons
Hawaii
ian
per day
Flo
a
0 to 2,000
r id
Alaska
a
U.S.Virgin Islands 2,000 to 5,000
5,000 to 10,000
10,000 to 20,000
Puer to Rico 20,000 to 52,000
Source: US Geological Survey Circular 1268
FIGURE 1-36 Average daily water usage per capita.
would be to check with the US Geological Survey to learn what

the average daily usage is per person by state (Fig. 1-36).
Local standards and codes related to tank capacity should be con-

sulted and complied with.
Turnover Tanks sized to meet peak demand must also have adequate
turnover when demand for water is not at a peak. Unused water can
become stagnant, generating unwanted tastes and odors. In cold cli-
mates, lack of turnover can cause tank icing. Water turnover problems
can be solved by filling the tank to a lower capacity that matches the
reduction in demand or by adding a recirculation system. Addition-
ally, several mixing systems are available that can create a more uni-
form residual chlorine content, reduce stagnation, and help prevent
the generation of unwanted tastes and odors.
Volume/standard capacities For elevated tanks, the most economical

storage is achieved by selecting a standard capacity and head range
on the basis of the recommendations of the tank contractor. Typical
capacity ranges of elevated tanks are given in Tables 1-8 and 1-9.
The largest-capacity elevated tank built to date is 4 mil gal (15,142
m3 ). It may be possible to build larger capacity tanks, but they would
be the first of their kind.
Reservoirs and standpipes are more flexible in their height/
diameter limitation. It was once thought that reservoirs could only
be constructed in height increments of 8 or 10 ft (2.44 or 3.05 m). Steel

44 Chapter One
is now readily available in made-to-order heights (and widths) in ad-

dition to these. An economical tank can be built to whatever diameter
and height is required (Tables 1-1 and 1-3).
Diameter and Height Selection

For a ground storage tank, three major factors influence the selection
of the most economical diameter and height. Of the following three
factors, soil bearing and earthquake design usually have the biggest
influence.
Soil bearing The tank foundation and ultimately the soil must sup-
port the weight of both the water and the tank. Each cubic foot of water
weighs 62.4 lb/ft3 . The calculation of the weight of a 1-ft2 column of
water from the bottom of the tank to its top capacity height can give
one an idea of the weight that must be supported. A sample calcula-
tion for a 40-ft (12-m) column of water would be 40 ft × 62.4 lb/ft3 =
2,496 lb/ft2 or about 2,500 lb/ft2 . So, if a 2,500-lb/ft2 soil bearing is not
available at the tank site, various foundation types could be evaluated
to support the column height of water needed. Deep foundations or
large mats may increase costs to the extent that it may actually be more
economical to either change the height of the tank or evaluate other
sites with higher soil bearing values.
Earthquake Typically, the taller and thinner the tank, the more that
earthquake may affect the design.
Wind The taller and wider a tank, the more wind may affect the
design.
Here are some examples of diameters (D) and heights (H) of
ground storage tanks with typical design conditions that might make
them more economical:
r D = H: For decent soil bearing values of 4,000 psf/ft2 , with

low earthquake factors and typical 90-mph wind design, a
tank in which diameter is equal to height may be the most
economical shape for small- and medium-size tanks.
r D < H: For soil bearing values greater than 4,000 psf/ft2 , with
low earthquake factors and typical 90-mph wind design, a
tank in which diameter is less than height may be the most
economical shape. In these tanks, there are fewer costs in the
bottom and roof and more costs in the shell.
r D > H: For soil bearing values less than 4,000 psf, with high
earthquake factors and winds greater than 90 mph, a tank
in which diameter is greater than height may be the most
economical shape. In these tanks, there are more costs in the
bottom and roof and fewer costs in the shell.

It is important that you call your local tank contractor to help

determine the most economical diameter and height combination for
the design conditions at your site.
Selecting the Tank

Ground/Elevated Storage
Your first decision is whether to build a ground storage tank or an el-
evated tank. If higher-elevation sites are available that could provide
the required pressure through gravity flow without the need for run-
ning a lot of water main to reach the site, the more economical choice
would probably be a ground storage reservoir. If site elevations are
not available within the water system area or are not high enough to
provide the required pressure through gravity flow, an elevated tank
or standpipe would probably be a better choice. Using a reservoir and
pumping to meet the pressure and daily water demand will add daily
pumping and peak demand charges throughout the life of your tank.
See the previous section on hydraulics.
Again, it is important that you call your local tank contractor to
provide budget pricing for various tank options to evaluate the initial
and lifetime costs of your new storage tank.
Aesthetics/Appearance
The aesthetic appeal of a new water-storage tank is often one of the
most talked-about elements of tank selection. The public may want
a tank that will blend into its surroundings, or be a highly visible
landmark for the community, or match the system’s existing tanks.
The tank owner and security personnel may want to place the tank
on a more visible site that can be readily secured and monitored. This
decision must be handled on a case-by-case basis.
Ornamental Tanks
Highly stylized ornamental tanks can provide community or com-
pany identity and advertisement, be more aesthetically pleasing, or re-
solve NIMBY issues. Unique, decorative tanks have been constructed
in many areas and, although more costly to construct, they are often
landmarks in which the community takes pride.
Economics
Although the initial cost of constructing a tank has a significant eco-
nomic impact, the tank’s operating cost, reliability, and maintenance
requirements must also be considered.
Special Needs
Sometimes communities have special needs or desires; for example, a
community may want to house the fire department in the base of the

46 Chapter One
tank. Multiuse tanks can be constructed to match the community’s

needs.
Liability
To limit liability, tank owners seek methods to control access. Some
styles of tanks—such as single-pedestal spheroid, fluted-pedestal, and
composite elevated tanks—do not have exterior ladders, thereby effi-
ciently limiting access. On legged or ground tanks, ladder guards can
be installed that limit access to the ladders.
Life-Cycle Costs
Anticipated need for and scheduling of tank repainting and mainte-
nance are important considerations. The style of tank, its surface area,
and the type of surface all directly influence maintenance costs.

CHAPTER 2
Selecting and
Specifying
Appurtenances
William B. Harper, P.E., Andre Harper,
and Krista L. Harper, P.E.
Harper and Associates
Appurtenances, or accessories, for tanks are vital for the function,

operation, and maintenance of the tank system. Appurtenances are
covered in separate sections of American Water Works Association
(AWWA) D100, Standard for Welded Carbon-Steel Tanks for Water
Storage; AWWA D103, Standard for Factory-Coated Bolted-Steel
Tanks for Water Storage; and AWWA Manual M42 Manual of Water
Supply Practices, Steel Water-Storage Tanks; and in regulatory docu-
ments such as those issued by the Occupational Safety and Health
Administration (OSHA). The majority of the appurtenances and at-
tachments described for steel water-storage tanks are required by law,
code, and industry standards to make the tank a safe and functional
facility. Other accessories are optional and may be specified by the
owner to improve the facility’s function or appearance.
Accessory items on the tank structure should generally be located
where they are readily accessible from a fixed ladder or platform sur-
face. An exception to this is the location of stub overflows, which,
when used, are purposely located away from ladders to avoid ladder
icing. Specific tank accessories required by AWWA standards should
be shown but not detailed on the bid drawings unless a specific detail
is required, because each manufacturer has proprietary components
that fulfill the intent of the standards. Such details may cause problems
if bidders are required to provide another constructor’s proprietary
47
Selecting and Specifying Appurtenances
48 C h a p t e r Tw o
apparatus. Accessory details and orientation should be developed and

included in the shop drawing submittals after the contract is awarded.
The governing requirement should be that the accessories meet the
minimum requirements of the regulations, referenced AWWA stan-
dards, and the intent of the specifications. If the owner elects to in-
clude components and operating systems in excess of those specified
by AWWA standards, the engineer should provide specifications and
details that clearly define the components required and the scope of
related work.
Ground-Supported Tanks
Welded-Tank Shell Manholes
For safety and ease of interior access during construction activities and
maintenance inspections, at least two opposing shell manholes are
required on welded ground-supported tanks for ventilation during
interior coating operations. On tanks more than 100 ft (30.5 m) in
diameter, it may be desirable to have three or more shell manholes,
keeping the maximum circumferential spacing at 100 ft (30.5 m) or less.
AWWA D103 requires only one shell manhole on bolted tanks because
a tank panel can be removed to provide additional ventilation, but the
specifier may elect to provide additional manholes.
Sizes and Types

Typically, shell manholes are 24 or 30 in. (610 or 760 mm) in diam-
eter to accommodate ventilating equipment and allow easy egress.
Manholes larger than 30 in. (760 mm) in diameter are uncommon and
may require special design details for structural integrity. Single-bolt
inward-opening shell manholes or outward-opening shell manholes
with hinged covers are standard (Figs. 2-1 and 2-2). Outward-opening
covers may require reinforcing plates on the shell, whereas inward-
opening manholes usually accomplish their reinforcement through
heavy plate necks. The heavy plate neck also provides the gasket sur-
face to the cover. The inward-opening cover must be hinged to ensure
proper operation. If the tank will be subject to severe icing conditions,
an inward-opening manhole may not be desirable.
Welded-Tank Flush Manholes

If specified by the purchaser, flush-type cleanouts (Fig. 2-3) shall be
furnished for ground-supported tanks. Flush rectangular manholes
(rectangular manholes mounted flush with the bottom of the tank)
having a minimum length of 24 in. (610 mm) in the short direction
and a maximum length of 48 in. (1,220 mm) in the long direction are
also available. Such manholes are useful when a tank interior is be-
ing cleaned. Refer to AWWA D100 and American Petroleum Institute

3/8
A A
Front elevation
Typical
3/8
Exploded side elevation
Section A–A
1/4
Typical
FIGURE 2-1 Inward-opening shell manhole detail. (Source: AWWA Manual M42,
Tank shell
plate
1/4
5 in. (127 mm) min.
See Detail A
Machine flange
face full width
Weld before
machining
er
Detail A
et
m
ia
Roll reinforcing
d
C.
plate to exterior 9 in.

B.
tank shell radius (229 mm)
1/4-in. (6.4-mm)
hole on horizontal
centerline
Front elevation
A
Tank shell plate
Floor assembly plate Side elevation
A
Bolt holes
equally spaced 1/4
to straddle centerline
See Detail A
Isometric blowout Section A–A
FIGURE 2-2 Outward-opening shell manhole detail. (Source: AWWA Manual

M42, Steel Water-Storage Tanks)

Shell
Section A–A
3 ft (0.9 m)
Bottom plate
Nearest horizontal weld
15 in. 46 each 1-in.-

Reinforcement plate
(381 mm) min. (25.4-mm)-diameter
bolts, equally spaced
(see API 650, Table 3-11)
Note: Refer to API Section 650, Figure 3-8 Flush-Type Cleanout Fittings
and Tables 3-11, 3-12, and 3-13
FIGURE 2-3 Flush-type cleanout.
(API) Standard 650 for details and design requirements. Although the
flush-type manhole is permitted in the AWWA standards, its use is not
recommended in high-seismic regions where the additional stiffness
of the reinforcing may cause stress concentrations or buckling in an
earthquake.
Reinforcing
The shell plates where the manholes are located shall be reinforced to
comply with AWWA D100 Section 3.13.2.5, and all portions of the
manholes (including reinforcing of the neck, the bolting, and
the cover) shall be designed to withstand the weight and pressure
of the tank contents.
Bolted-Tank Shell Manholes

One manhole, unless otherwise specified, shall be provided in the first
ring of the tank shell at a location to be designated by the purchaser.
If any manhole cover weighs more than 50 lb (23 kg), a hinge shall be
provided.
Size and Shape

Manholes may be either circular, 24 in. (610 mm) in diameter; square,
24 × 24 in. (610 × 610 mm); or elliptical, 18 × 22 in. (457 × 559 mm),
minimum size. Flush rectangular manholes with a minimum length

Selecting and Specifying Appurtenances 51
of 24 in. (610 mm) in the short direction and a maximum length of 48

in. (1,220 mm) in the long direction are also acceptable.
Flush Manholes
Flush rectangular manholes (rectangular manholes mounted flush
with the bottom of the tank) having a minimum length of 24 in. (610
mm) in the short direction and a maximum length of 48 in. (1,220
mm) in the long direction are also available. Such manholes are useful
when a tank interior is being cleaned. Refer to AWWA D100 and API
Standard 650 for details and design requirements.
Reinforcing
The shell plates where the manholes are located shall be reinforced to
comply with AWWA D103 Section 3.11, and all portions of the man-
holes (including reinforcing of the neck, the bolting, and the cover)
shall be designed to withstand the weight and pressure of the tank
contents.
Pipe Connections
The number of tank-bottom or shell-piping connections should be
kept to a minimum. The earlier practice was to use a common in-
let/outlet drain connection through the tank bottom or on the tank
shell (Figs. 2-4 and 2-5). If a bottom connection is used, a removable
section of pipe 6 to 8 in. (150 to 200 mm) long may extend above the
outlet at floor level to serve as a silt stop. The drainpipe shall be re-
cessed to aid in draining the tank. Recent requirements concerning
minimum and maximum detention time during which the water re-
mains in the tank may require separate inlet and outlet connections.
Baffles and flow diverters are also used to control detention time.
1/4 (typ.)
FIGURE 2-4 Recessed inlet–outlet pipe bottom connection detail. (Source:

AWWA Manual M42, Steel Water-Storage Tanks)

Plan
Elevation
FIGURE 2-5 Nonrecessed inlet–outlet bottom connection. (Source: AWWA

Piping connections through the tank bottom or shell are normally

furnished in steel pipe welded or bolted to the shell or bottom. Ductile-
iron or cast-iron pipe connections must pass through a mechani-
cal joint–type connection that is welded or bolted to the steel tank
bottom.
Pipe connections shall be of the size specified by the purchaser and
are usually attached to the tank bottom. The point of attachment shall
be designated by the purchaser. Connections to the tank or piping
furnished by the tank constructor shall be made by the purchaser.
Silt Stop
If a removable silt stop is required, it shall be at least 4 in. (102 mm)
high, and the fitting or piping connection shall be flush with the tank
floor when the stop is removed. If a removable silt stop is not required,
then the fitting or connecting pipe, or both, shall extend above the floor
at least 4 in. (102 mm).
Drain Sump
To facilitate a more efficient and expedient removal of residual water
remaining on the tank bottom after draining the tank, a new drain

and a sump 3 ft (914 mm) in diameter by 6 in. (152 mm) deep shall be
installed. The drain sump shall be covered with a slip-type or hinged,
grated cover that can be easily removed or opened during mainte-
nance intervals.
Shell Connections
Shell connections are permitted as long as the purchaser makes ade-
quate provisions to protect the pipe from freezing or vandalism and
provides adequate pipe flexibility to account for shell rotation and de-
flections of the shell when the tank is filled and drained. These include
sample taps, disinfection fittings, and fire-hose fittings for the interior
and exterior.
Piping Flexibility
Special piping flexibility to accommodate seismic movements and set-
tlement in the piping system shall be provided to protect the connec-
tion to the shell. AWWA D100 defines the distance from the shell
intersection that through-the-tank-bottom piping connections may be
located on unanchored tanks designed for seismic conditions. Bot-
tom connections shall comply with AWWA D100 Section 13.5.1 as a
minimum. Underbottom connections are not recommended on tanks
in high-seismic zones. Tank sliding or uplift may impose additional
stress on the connection and tear the bottom. Sidewall connections,
which can readily be inspected after a seismic event, are preferred.
Overflows
A properly sized overflow is essential to protect the tank structure
from excessive water levels caused by rapid variations in distribution
system conditions. Exterior overflows are recommended. In colder
climates, ice buildup on an internal overflow may become a problem
and eventually break the overflow pipe. Overflow waters should be
directed beyond the exterior perimeter of the tank to prevent damage
to the tank grade or foundation during overflow. Most state stan-
dards recommend that the overflow on elevated tanks be extended
down the side of the tank to within approximately 12 to 24 in. (305
to 610 mm) above grade. Extending the overflow pipe prevents water
discharged from the pipe from freezing on the tower structure and
damaging it. In addition, most governing agencies require an air gap
between the overflow tank piping and final drainage system to pro-
tect against backflow. Figure 2-6 shows one type of overflow-pipe air
gap. Most states require a screen or flap/gate arrangement over the
end of the pipe connected to the tank and a removable grate on the
bottom portion of the pipe. The valve shall be a flanged passive check
opening with 2 in. (51 mm) of water and shall be able to withstand zero
backpressure. As distribution systems and pumping capacities are in-
creased, the vent and overflow capacities of existing tanks should be

FIGURE 2-6
Overflow air break
with flap valve.
(Source: AWWA
Manual M42, Steel
Water-Storage
Tanks)
reevaluated to ensure their adequacy to relieve potential pressure or

vacuum conditions in the tank. Overflows should be easily accessible
for maintenance, repair, and inspection.
Welded-Tank Overflow
An overflow protects the tank from overpressure, overload, and pos-
sible catastrophic failure should the pumps or altitude valve fail to
shut off when the tank is filled to capacity. A properly operated tank
should not overflow during normal operation. An overflowing tank
is an emergency, and the malfunction causing the overflow should be
determined and corrected as soon as possible.
The tank shall be equipped with an overflow of the type and
size specified by the purchaser. If a stub overflow is specified, it shall
project at least 12 in. (305 mm) beyond the tank shell. If an overflow
to ground is specified, it shall be placed down the side of the tank
shell and supported at proper intervals with suitable brackets. The
overflow discharge shall be located such that it will not be obstructed
by snow or ground clutter. The overflow to the ground shall discharge
over a drainage inlet structure or a splash block. It shall originate at the
top in a weir box or other appropriate type of intake. A top stiffener
shall not be cut or partially removed. The overflow pipe and intake

shall have a capacity at least equal to the pumping rate as specified by

the purchaser. Where a side opening–type overflow is used, the head
shall be not more than 6 in. (152 mm) above the lip of the overflow
and in no case more than 12 in. (305 mm) above the top capacity level.
The overflow pipe shall terminate at the bottom with an elbow. Un-
less otherwise specified by the purchaser, the overflow pipe shall be
steel pipe with screwed or welded connections if less than 4 in. (102
mm) in diameter, or flanged or welded connections if 4 in. (102 mm)
in diameter or larger. The purchaser shall specify the maximum flow
rate in gallons or liters per minute, for which the overflow shall be de-
signed. Overflow pipes may be either internal or external as specified
by the purchaser. Minimum external overflow pipe thickness shall
be 3/16 in. (4.76 mm). Internal overflow pipes are not recommended
when tank usage and climatic conditions are such that ice may dam-
age the overflow pipe or its attachments. When specifying an internal
overflow pipe, the purchaser should consider the consequences of
an overflow failure, which can empty the tank of its contents. An in-
ternal overflow pipe shall be at least 0.25 in. (6.35 mm) thick. The end
of the overflow pipe may be covered with a coarse, corrosion-resistant
screen equivalent to 3/8 in. (9.5 mm) or larger mesh. The end of the
overflow pipe may also be covered with a flap valve or other protective
cover as specified by the purchaser.
Bolted-Tank Overflow
The tank shall be equipped with an overflow of the type and size spec-
ified by the purchaser. If a stub overflow is specified, it shall project
at least 12 in. (305 mm) beyond the tank shell. If an overflow to the
ground is specified, it shall be brought down the outside of the tank
shell and supported at proper intervals with suitable brackets. The
overflow to the ground shall discharge over a drainage inlet struc-
ture or a splash block. It shall terminate at the top in a weir box or
other appropriate intake. A top stiffener shall not be cut or partially
removed. The overflow pipe and intake shall have a capacity at least
equal to the pumping rate as specified by the purchaser, with a wa-
ter level not more than 6 in. (152 mm) above the weir. The overflow
pipe shall terminate at the bottom with an elbow. Unless otherwise
specified by the purchaser, the overflow pipe shall be steel pipe with
screwed or welded connections if smaller than 4 in. (102 mm) in di-
ameter, or flanged or welded connections if 4 in. (102 mm) in diameter
or larger. The external overflow pipe shall have a minimum thickness
of 3/16 in. (4.76 mm). The purchaser shall specify the maximum flow
rate, in gallons or liters per minute, for which the overflow shall be de-
signed. Internal overflows are not recommended but may be provided
if specified by the purchaser. The internal overflow pipe shall have a
minimum thickness of 0.25 in. (6.35 mm).

Ladders
Safe access must be provided for authorized personnel who need to
reach upper shell areas and the top of the tank facility.
Exterior Vertical Ladders
Exterior ladders, cages, and platforms designed to meet OSHA stan-
dards are recommended (Fig. 2-7). Either the ladder should terminate
Flare out to
join top hoop
or platform
Tank shell
1/4
1 ft (0.3 m) to tank bottom 8 ft (2.4 m) to tank bottom

Ladder elevation Cage elevation
FIGURE 2-7 Exterior caged ladder details. (Source: AWWA Manual M42, Steel
Water-Storage Tanks)

at least 8 ft (2.4 m) above grade, or a solid, locking door, provided

to discourage unauthorized access to the tank, should be installed on
the lower 10 to 20 ft (2.4 to 6.1 m) of the exterior ladder. Certain areas
will require a locking door and anti-climb screening at the bottom of
the ladder cage to discourage unauthorized access. The exterior lad-
der, roof hatch opening, and interior ladder (if specified) should be
located close together to reduce the movement necessary by a climber
on the tank roof. The ladder can be specified as painted carbon steel,
galvanized steel, or stainless steel. If stainless-steel ladders are used,
insulation (dielectric connections) must be included that separates the
stainless steel from the carbon-steel tank, and all stainless-steel com-
ponents must be coated to prevent corrosion from occurring on the
carbon-steel tank.
Exterior Circular Stairway

When specified, the exterior stairway shall be of a semicircular de-
sign meeting API Standard 650 Section 3.8.9, Platforms, Walkways,
and Stairways (Fig. 2-8). The stairway shall be at a specified location
terminating at the exterior roof hatch. Minimum stair width shall be
30 in. (762 mm). An expanded metal security enclosure with a hinged
gate and lock system shall be installed around specified ground-level
termination.
Hand rail
to match
roof
hand rail 42 in. (1.06 m)
Roof
1
2.5 in. × 2.5 in. × 3/8 in. 1 1/4 in. × 6 in. × 6 in. (6.4 mm
(63.5 mm × 63.5 mm × × 152 mm × 152 mm) plate
0.38 mm) angle, two each
Tie into platform Shell
Platform 8 in.
(203 mm)
30 in.
1.5-in. (38-mm) (762 mm)
pipe handrail min. 9 in.
along outer (228.6 mm)
stair perimeter Stair treads 3/16 in.
(0.1875 mm) plate,
shape as shown
30 in. (762 mm)
Plan view wide (typ. 53)
0.25 in. × 12 in. (6.3 mm ×
Wearpads 1/4 in. × 6 in. 305 mm) FB stairway
(1.3 mm × 152.4 mm) runner inside and outside
Stairway brace
FIGURE 2-8 Exterior circular stairway. (Note: FB = flat bar.)

Interior Ladders
Because of accelerated rates of corrosion and the potential for ice
buildup in areas where freezing temperatures occur, ladders inside
the tank container are not recommended. Ice buildup on an interior
ladder can impose loads on the tank wall plates that are sufficient
to pierce or rupture the tank container. Even in temperate climates,
corrosion can damage interior ladders, making them unsafe. The use
of stainless-steel ladders must include insulation (dielectric connec-
tions) separating the stainless steel from the carbon-steel tank, and all
stainless-steel components must be coated to prevent corrosion of the
carbon-steel tank.
Ladders are installed inside dry risers and access tubes. There they
are not subjected to corrosive conditions, and the access doors may
be locked to deter access. In general, all interior ladders shall meet
design criteria noted herein for exterior ladders.
Minimum requirements Minimum requirements for ladders, hatches,

and so forth can be found in OSHA 29 Code of Federal Regulations
(CFR) Part 1910, Occupational Safety and Health Standards, General
Industry Standards.
Welded-Tank Ladders
Exterior tank ladder The contractor shall furnish a tank ladder on the
outside of the shell beginning 8 ft (2.4 m), or as specified, above the
level of the tank bottom and located to provide access to the roof
manway. The minimum clear width of step surface for rungs shall
be 16 in. (406 mm), and rungs shall be equally spaced 12 in. (280
mm) on center. The perpendicular distance from the centerline of the
rungs to the tank wall shall not be less than 7 in. (178 mm). Rung
size shall not be less than 3/4 in. (19 mm) in diameter or equivalent
section. The maximum spacing of supports attaching the ladder to
the tank shall not exceed 10 ft (3 m). The minimum design live load
shall be two loads of 250 lb (113.6 kg) each concentrated between
any two consecutive attachments to the tank. Each rung in the ladder
shall be designed for a single concentrated load of 250 lb (113.6 kg)
minimum. The design loads shall be considered to be concentrated
at such a point or points as will cause the maximum stress in the
structural ladder member being considered. Side rails may be of any
shape having section properties adequate to support the design loads
and providing a means of securely fastening each rung to the side rail
so as to develop the full strength of the rung and to lock each rung to
the side rails.
Interior tank ladder Inside tank ladders are not recommended for cold
climates where ice may form. If an inside ladder is required, it shall

comply with the requirements for exterior ladders outlined in the

previous paragraph.
Roof ladder For tanks with roofs, unless otherwise specified, the man-
ufacturer shall furnish access to roof hatches and vents. Such access
shall be reached from the outside tank ladder. A roof ladder is not
required on portions of standpipe or reservoir roofs having a slope
less than 2 inches per 12 inches of rise (2/12). A roof ladder shall be
provided on roofs having a slope greater than 2/12. For roof slopes
from 2/12 to 5/12, there shall be a nonskid walkway and a single
handrail. For a roof slope greater than 5/12, a ladder or stairway shall
be provided.
Bolted-Tank Ladders
Exterior ladders, cages, and platforms designed to meet OSHA stan-
dards are recommended. Either the ladder should terminate at least
8 ft (2.4 m) above grade, or a solid, locking door, provided to dis-
courage unauthorized access to the tank, should be installed on the
lower 8 to 20 ft (2.4 to 6.1 m) of the exterior ladder. Certain areas will
require a locking door and anti-climb screening at the bottom of the
ladder cage to discourage unauthorized access. The exterior ladder,
roof hatch opening, and interior ladder (if specified) should be located
close together to reduce the movement necessary by a climber on the
tank roof.
Exterior tank ladder The constructor shall furnish a tank ladder on

the outside of the shell beginning 8 ft (2.4 m), or as specified, above
the level of the tank bottom and at a location to be designated by the
purchaser, preferably near one of the manholes. The side rails shall
not be less than 2 × 3/8 in. (51 × 9.5 mm), with a spacing between
them not less than 16 in. (406 mm). The nonskid rungs shall not be
less than 3/4-in. (19-mm) round or square bars spaced 12 in. (305 mm)
apart on centers.
Interior tank ladder Inside tank ladders are not recommended for cold
climates where ice may form. If an inside ladder is required, the side
rails shall not be less than 2 × 3/8 in. (51 × 9.5 mm), with a spacing
between them of not less than 16 in. (406 mm). Rungs shall not be less
than 3/4-in. (19-mm) round or square bars spaced 12 in. (305 mm)
apart on centers.
Roof ladder For standpipes and reservoirs with roofs, unless other-
wise specified, the constructor shall furnish access to roof hatches and
vents. Such access shall be reached from the outside tank ladder. Refer
to AWWA D103 Section 5.4 for minimum requirements for roof lad-
ders based on the slope of the roof.

Minimum requirements Minimum requirements for ladders, hatches,

and so forth can be found in OSHA 29 CFR Part 1910.
Ladder Safety Devices

If federal or local laws or regulations require a safety cage, rest plat-
forms, roof-ladder handrails, or other safety devices, the purchaser
shall so specify. None of these devices are advisable on the submerged
portion of interior ladders in low-temperature climates.
Ladder platforms For tanks with a total height in excess of 20 ft (6.1 m)

or as required by OSHA, ladders with offset rest platforms (secondary
platforms) are required every 20 ft (6.1 m) or as required by OSHA for
the ladder assembly. This length may be extended to 30 ft (9.1 m) if an
approved safety cage is used.
If the ladder extends up the exterior of the tank, the ladder should
be equipped with a locked guard to prevent unauthorized access to
the tank exterior and roof.
Conduits, fixtures, pipes, valves, and other items should not in-
terfere with the safety of the ladders or platforms.
Roof safety railings In addition to exterior ladder safety devices, the

most commonly installed safety items are safety railings at the roof
where the exterior ladder terminates (Fig. 2-9). These railings protect
personnel on the roof near the roof hatch. New handrailing shall be
fabricated and installed to form a totally enclosed work area around
the roof hatch. Existing pipe sections may be used, provided they
meet current OSHA requirements. A self-closing hinged gate shall be
provided at the exterior ladder opening. Handrailing layout shall be
verified in the field before handrails are fabricated. All safety railings
installed on the tops of tanks and ladders should comply with mini-
mum OSHA requirements or local building codes. Consult with the
applicable agency in charge of tank location to determine the latest
safety requirements. Total-perimeter handrails are not recommended
in areas of high snow load.
Ladder safe-climbing rails or cables In lieu of intermediate platforms,

approved safe-climbing rails or cables may be used. Figure 2-10 shows
a typical safe-climbing rail. Some tank owners may desire supplemen-
tary rest platforms in addition to the safety rail. Climbing rails can be
galvanized or stainless steel. Painted carbon-steel rails are not recom-
mended.
Safety belts and sleeves should be furnished for the ladder safety
devices. Safety sleeves should be checked for proper operation along
the full height of the rails or cables. Any coating, deviations, or
obstructions that prevent the free operation of the sleeve should be
removed. Special dismount sections are available to ensure that the

Exterior Interior
ladder ladder
Hatch
3 ft
(0.9 m)
Self-closing
gate
To roof
Typical handrailing layout
Self-closing
hinges
Plan view
Typical handrail post bottom

Partial
1.5-in. (38-mm) Gate elevation See Detail A
standard black pipe
1/4 in. (6.3 mm)
1/4 in. (6.3 mm) 1.5-in. (38-mm)-
diameter pipe
6-in. (152-mm)-square pad

42 in.
Detail A (1.06 m)
4 in. × 1/4 in. (102 mm

× 6.3 mm) plate
Rail elevation
FIGURE 2-9 Safety rail enclosure. (Note: For roof slope 2/12 or less, windy or
wet conditions may require additional safety lines for areas outside
enclosure.)
climber does not fall from the tank when the climber is dismounting
onto the roof.
Roof fall prevention cable assemblies To prevent personnel who are

working or walking on the tank roof from falling over the edge of the
roof, furnish and install stainless-steel cable assemblies as shown in
Fig. 2-11. Cable shall be stainless-steel aircraft cable of 7 × 19 construc-
tion.
Cable collar around center vent shall be 0.25 in. (6.35 mm) in diam-
eter placed in a plastic sleeve to prevent paint damage and connected

FIGURE 2-10 Safe-climbing rail for an outside ladder. (Source: AWWA Manual
M42, Steel Water-Storage Tanks)
with stainless-steel cable clamps. The cable shall allow a maximum of

2 in. (51 mm) greater diameter than the center vent neck at the roof
level. The cable shall rest freely on the roof after installation.
The stainless-steel cables that prevent personnel from falling shall
be 3/16 in. (4.76 mm) in diameter and shall be fabricated with a swaged
loop at the center vent attaching point and a swaged carabiner double-
locking attachment at the hold-down point adjacent to the roof hatch
work area. Each personnel cable shall terminate 42 in. (1,067 mm) from
the outer perimeter of the roof when attached to center vent cable.
A hold-down eye shall be welded at the attachment point de-
scribed previously. The size of the eye and pad shall be as determined
by the civil engineer and the structural engineer.
Roof Openings
At least two roof openings are required for personnel access and ven-
tilation during maintenance and rehabilitation activities on welded-
steel tanks.
Primary Opening
The first (primary) roof opening should be located near the tank side-
wall close to the exterior ladder. The previous minimum size for this

1/4-in. (6.3-mm) cable with

a maximum 2 in. (51 mm)
greater diameter than Center vent
the center vent. Use
swaged loop to connect Cable loop construction
the two ends of cable. ble See Detail A
ca
d
de
t en
ex
lly in. )
Fu 42 06 m
.
(1
Attachment pad
Provide three carabiners compatible See Detail B
with the 1/2-in. (12.7-mm) round bar.
Pad is for attachment when not in
use. Attach carabiners to cable with
a swaged loop as per Detail A.
6 in. × 6 in. × 1/4 in. Cable at

Round bar 1/2 in. × 10 in. (152 mm × 152 mm center vent
(12.7 mm × 254 mm) with × 6.4 mm) carbon-steel
large bend as shown pad for attachment Swaging tool used
4 in. at this point to
(101.6 mm) secure the cable
1/4 in. 3 in.
(6.3 mm) (76 mm) to itself
Existing wood roof
Safety cable attachment

Detail B
Install at exit to roof Typical cable loop
from enclosure area Detail A
FIGURE 2-11 Safety cable system.
roof opening was 15 × 24 in. (380 × 610 mm), but OSHA now requires
a 30-in. (760-mm) square or round opening with a hinged cover and
locking hasp to facilitate access to the tank interior. With the advent
of diving inspections in tanks, which necessitate the use of a rubber
raft for inspecting the underside of the roof, it may be prudent to use
larger roof hatch assemblies. These can range in size from 48 to 60
in. (1,220 to 1,524 mm) and can be constructed of aluminum covers.
A curb at least 4 in. (100 mm) high and a 2.in. (50-mm) downward
cover overlap are mandatory on any roof opening to prevent rain or
snowmelt from entering the tank (Fig. 2-12). Bolted and gasketed roof
manways without the curb and overlap are allowed on bolted tanks.
Secondary Opening
The second roof opening should be located near the tank center or
180◦ circumferentially from the primary opening. Its diameter should
be at least 20 in. (500 mm). If the center vent is of adequate size, is
not obstructed, and has a removable cover, the vent may suffice as the
secondary opening. The secondary opening, whether the center vent
or a separate opening, should be designed with a removable cover to

3⁄16 3⁄16
Cover elevation
3⁄16
Typical
3⁄16
Curb elevation
Plan view
FIGURE 2-12 Roof manhole assembly details. (Source: AWWA Manual M42,
accommodate a bolted ventilation fan. All removable covers should

be secured to the roof of the tank using hinges or chains. An additional
safety rail may be required between the secondary roof openings and
the edge of the roof.
Roof manholes should be equipped with locks to prevent unau-
thorized entry into the tank. Tanks that have access tubes leading
to the roof should have their roof manholes properly latched to pre-
vent them from blowing open in a strong wind, and any access doors
to the tank ladders should be locked. Shell manholes should be prop-
erly sealed to prevent leakage.
Additional Roof Openings

Cathodic protection handholes/covers, suspension insulators, and refer-
ence cell access Where cathodic protection is to be installed in the
tank interior, access handholes with covers and suspension insula-
tors shall be provided through the roof at locations specified by the
designer of cathodic protection. A fitting shall also be provided for
the cathodic-protection reference cell to be suspended into the tank.
Access openings shall also be provided at designated locations to ac-
commodate the electrical conduit for impressed-current systems. If
alternate power sources are specified, they can be either wind- or
solar-powered units mounted on the roof at specified locations.
Liquid level indicator fittings If a gauge board consisting of a float and a
target board is present, holes for designated fittings shall be provided
on the roof at locations specified by the designer.
Inlet stop/start controls Where a probe or transducer system con-
trols the water level and advises operators of low water levels, probes
requiring waterproof flanged entries at the top of the tank shall be
installed at locations specified by the designer.

Sampling hatch If specified by the owner, access through the roof for
a sampling hatch shall be provided at the location specified by the
designer. A suitable cover shall be provided over the access entry.
Overflow inspection hatch If specified by the owner, access through
the roof shall be provided to enable observation and inspection of the
interior overflow opening. A suitable cover shall be provided over the
access entry.
Vents
For closed-top tanks, venting must be provided to safeguard against
excess pressure or vacuum buildup during the maximum inflow or
outflow of water. Structural failures of tanks can be caused by inade-
quate venting. When the vents are being sized, the area to which the
overflow pipe contributes should not be considered part of the venti-
lation area. A minimum of one vent is required; this should be located
near the center of the roof. For larger-diameter tanks, several vents
should be located around the periphery as well as at the center of the
tank to facilitate crossflow ventilation.
The most common forms of tank vents are the mushroom, pan
(Fig. 2-13), and 180-degree types. Vents with pressure- and vacuum-
releasing pallets are recommended. A clog-resistant vent is shown in
Fig. 2-14. All vents should be screened to protect against the entry
of birds, animals, and insects. The screening should be stainless steel
or some other type of corrosion-resistant material. Some health au-
thorities require that shields be installed to keep dirt and debris from
blowing into the tank. In areas of snow buildup, the vents should be
protected or elevated to prevent them from being clogged by snow.
Special vent designs may be necessary to prevent vents from clogging
or freezing over, based on local conditions and operations.
(Outside diameter)
3/16
C
ov
er
A A
.
am
di
nt
Ve
Tank ro
of
di
am
(Inside diameter)
et
er
(Outside diameter)
(Hole in roof)
Plan view Section A-A
FIGURE 2-13 Pan deck vent detail. (Note: diam. = diameter.) (Source: AWWA

re
su
pres 1/2 - no. 13 × 15 flattened
Air expanded metal bird screen
Pres
PTFE gaskets (typical) sure
palle
Vacuum t
pallet
Screen
(brass material is normal)
Support bars
Air vacuum
Carbon–steel
body
Roof
Install vent vertical 5 +
FIGURE 2-14 Typical clog-resistant vent detail. (Note: Pallets should be

removed during coating to prevent clogging of the screens. Periodic
inspection and maintenance are required to keep in proper working condition.
PTFE = polytetrafluoroethylene.) (Source: AWWA Manual M42, Steel
Many older riveted tanks do not have vents; instead, they have
finial balls that provide limited or no ventilation area. These finial balls
should be replaced with vents when maintenance or repair work is
done on the tank. As distribution systems and pumping capacities are
enlarged, the vent and overflow capacities on existing tanks should
be reevaluated. Tanks have failed because of pressure or vacuum re-
sulting from inadequately sized or improperly maintained vents and
overflows. The maximum withdrawal rate is usually assumed to be
either the value that occurs when the pipes at grade level break or the
maximum rate pumped from low-elevation reservoir tanks.
Controls and Devices for Indicating Water Level

An indicating system of some type should be provided on the tank so
that operators can easily determine the water level. The most common
devices used to measure water level are gauge boards and pressure
transducer readouts.
Each form of water-level indication has advantages and disad-
vantages. Cost and the need for direct or remote reading, ease of
maintenance, and performance in adverse weather conditions should
all be considered when selecting an indicating system.
Gauge Boards
Gauge boards are normally composed of a float and target board on
which water level indication is accomplished by noting the position
of a target against a gauge board on the outside of the tank. The target

marker is controlled by a cable attached to the float. As the float rises,

the target marker is lowered. At capacity or full conditions, the tar-
get will be at the low end of the gauge board, making it potentially
accessible to vandalism or snow buildup. Half-travel gauge boards
are recommended to protect the target and float from vandalism and
high winds. Float-type systems such as these are not recommended
in freezing climates.
Altitude Valves
In many water distribution systems, altitude valves are used to con-
trol the water level in tanks for which the high water level is at a lower
elevation than the pressure gradient of the system. Even some small
one-tank systems have been designed with an altitude valve on the
tank inlet/outlet line. Both of these are examples of improper use of al-
titude valves. Altitude-control systems can be designed and installed
with timers that force the altitude valve to open, allowing water to
flow into and out of the tank and ensuring more frequent turnover.
Altitude valves may malfunction even in good weather. Freezing
weather increases the likelihood of malfunction, with frozen pressure-
sensing lines giving the altitude valve false signals. This usually causes
the tank to overflow, but it may also cause the valve to remain closed,
keeping the water in the tank static. Putting electrical heat tape and
insulation on the control piping or heating the altitude valve enclosure
minimizes these problems.
Remote Readings
A pressure transducer in the tank can indicate the water level at a
remote readout some distance from the tank facility. The pressure
transducer must be installed so that it is completely isolated from all
inlet and outlet openings. Pressure transducers are sensitive enough
to sense pressure changes created by water movement through a line
that would cause a false reading. The pressure transducer can also
control flow in and out of the tank by actuating pumps or valves.
Inlet Stop/Start Controls

A water utility may install a probe or transducer system to control the
water level and to advise operators of low water levels. Probes require
waterproof flanged entries at the top of the tank. In addition, for radio
or wired telemetry equipment, an insulated conduit from the tank
top to ground level must be installed to carry the electrical signal. If
probes are used in tanks that are subject to icing conditions, the probe
system should be designed to prevent damage from freezing.
Pressure Gauges
If freeze protection is provided, economical Bourdon pressure gauges
may be connected directly to the tank or riser.

Emergency Fill/Withdraw Connections

Some tanks may require provisions for emergency filling or with-
drawal. Typically, such provisions are needed when tanks are located
in remote areas where firefighters may withdraw directly from the
tank to fill pumper trucks or use stored water directly. Where required,
the emergency valves and connections should be designed to match
the emergency facilities of the agency that will use them. The design
must avoid cross-connections between the emergency system and the
potable water system. Connections should be protected from freezing
and vandalism, and the tank venting and overflow systems should be
sized for these unusual fillings and withdrawals.
Cold-Weather Operations
Designing Tanks for Cold Weather
Proper design of a tank will prevent most freezing problems and, if
freezing does occur, will allow personnel to follow operating proce-
dures to easily deal with it.
Inside appurtenances Tanks located in an area where the lowest one-

day mean temperature (LODMT) is −5◦ F (−15◦ C) or colder should
not be equipped with inside ladders or overflow pipes. As ice forms
and moves up and down, it can exert tons of force on ladders and
pipes, tearing them loose from their supports and possibly ripping
or punching holes in the container. The resulting leak will occur at
a very inopportune time. If an inside overflow pipe is broken, the
tank will rapidly lose all water down to the break, creating a large icy
area on the ground below. If the vent is plugged with ice or snow, the
tank roof may collapse when water evacuates the tank rapidly. It is
acceptable to equip a tank with inside ladders and overflow pipes if
the tank is known to have a high turnover of warm water. A ladder and
overflow can also be installed at the center of the tank and supported
by the access tube, as in single-pedestal tanks and extremely large
column-type tanks. The use of interior girders, roof bracing, painter’s
rails, or virtually any other protrusion below the high water line or
within an area affected by floating or suspended ice is a poor design
practice for areas with an LODMT of −20◦ F (−29◦ C) or colder. Certain
local conditions or tank usage patterns may cause equally severe icing
problems in warmer areas.
External features In addition to standard appurtenances and acces-

sories discussed herein, several design issues for the tank exterior are
significant for cold-weather operation.
Roof opening location Risers or inlet pipes should be directly below

roof vents or manholes, or an auxiliary opening should be provided.
This arrangement will facilitate thawing the tank if required. No pipe


G
E-max. opening H-max. opening

B-pipe
thread U-pipe
(Normal thread thread T-pipe
engagement) thread
V
Screwed models J-lift

have B-pipe threads
Padlock
(by others) F C-max. opening
(Normal thread (normal thread engagement)
A-pipe thread engagement)
Flanged models have

K-150 LB ASA mounting flange
L-holes; M-dia. on S bolt circle
Size A B C D E F G H J K L M S T U V
2 in. 1 ⁄ 2 9 2 ⁄ 5 ⁄ 3 ⁄ 4 ⁄ 8 ⁄ ⁄ 3 4 ⁄ 6 1 ⁄ 1 ⁄ 5 ⁄
3 in. 2 ⁄ 3 8 ⁄ 2 ⁄ 5 ⁄ 4 ⁄ 5 ⁄ 7 ⁄ ⁄ 4 8 ⁄ 7 ⁄ 2 ⁄ 4 ⁄
4 in. 3 4 8 ⁄ 3 ⁄ 6 ⁄ 4 ⁄ 6 7 ⁄ 1 ⁄ 6 8 ⁄ 9 ⁄ 3 3 ⁄ 4 ⁄
FIGURE 2-15 Double-seating, internal-closing drain valve. (Note: lb = pounds;

1 in. = 25.4 mm; 1 lb = 0.4 kg; ASA = American Standards Association [now
ANSI].) (Source: American Water Works Manual M42, Steel Water-Storage
Tanks)
opening should have a protective discharge cap that would preclude

the dropping of a probe into the inlet pipe to thaw the ice blockage.
Any gratings over piping should not be so restrictive as to prevent
thawing lines or pipes from being lowered into the pipe. Gratings
also conduct heat and promote freezing.
Static water projections Unless it is heated and insulated, piping
that extends from the tank or from other piping should not contain
static water. Drain valves extending on nipples will easily freeze.
Drain valves should be of the double-seating, internal-closing type
(Fig. 2-15).
Additional outlets Side outlets on the riser pipe for use in pumping
in steam or warm water to thaw the riser may be included in the
design. These outlets should be plugged at the pipe outer diameter to
eliminate an unheated projection.
Frostproof vents Vents should be designed to avoid freezing over or to
provide for pressure or vacuum relief. Some tank manufacturers have
proprietary designs for this purpose. A unique freezing problem may
occur when frost freezes solid over the fine screen in the vent and over-
flow. This type of freezing usually occurs on the fine screen designed
to keep insects out of the tank. Such freezing prevents the exchange

of air into the tank, resulting in a vacuum in the tank that can collapse
(implode) the tank until there is a structural rupture to break the vac-
uum. Preliminary research indicates that fiberglass screen material is
resistant to freezing.
Systems to Prevent Freezing

Various types of systems or equipment can be used to prevent tank
freezing. The following information is included as the design may
include one or more of the noted items.
Heating Heating a community water supply tank is usually not eco-
nomically feasible, though industrial sprinkler tanks for fire protection
have been heated for many years. However, new insurance rate struc-
tures and better community water supply systems have allowed many
factories either to dismantle the fire protection tank or to discontinue
heating it. In many cases, the insurance savings no longer offset the
heating energy costs.
Air bubblers Air-bubbler systems have been used successfully in
ground storage tanks and in elevated tanks with large risers. A bub-
bler system is shown in Fig. 2-16. Research on the use of these systems
Riser
Air compressor
Aeration line
Aeration line Riser
Air compressor
FIGURE 2-16 Tank riser bubbler system. (Source: AWWA Manual M42, Steel

indicates that a high-pressure compressor should be used that exerts

just enough pressure to overcome air-line friction, orifice friction, and
the hydrostatic head related to the water depth. There must also be an
influx of warmer water, because the bubbling action tends to remove
all the heat from a confined volume of water. The compressor should
be equipped with an air filter and water and oil traps to minimize the
likelihood that contaminants will be pumped into the tank.
Circulating pumps Circulating pumps that do not heat the water have
been successful on tanks with small-diameter (6 to 12 in. [150 to
300 mm]) riser pipes in Iowa, Minnesota, North Dakota, and South
Dakota. A circulating system is shown in Fig. 2-17. A relatively small
(1.5-hp [1.1-kW]) pump draws water from the base elbow, pulling
water down the insulated riser or from the connecting pipe. The pump
discharges water into a line 1 in. (25 mm) in diameter that enters the
Bowl of tank
Drip ring
Insulated riser
Circulating line
Foundation
1.5-hp circulating pump
FIGURE 2-17 Pumped circulation system for small riser pipes. (Source: AWWA

riser at the base of the tank and discharges into the tank container.
This creates circulation in the riser.
Insulation In many tank designs, it is a common practice to insulate

riser pipes. Many ground storage tanks and a few elevated tanks have
also had their exteriors sprayed with urethane foam to insulate them.
This yields a rough appearance, however, and there may be problems
maintaining adhesion between the foam and the steel. No matter how
thick the insulation, the stored water will eventually freeze without
heat input.
Additional Accessories and Exceptions

Any additional accessories required to be furnished shall be specified
by the purchaser. Exceptions to the provisions of this section may be
specified by the purchaser to suit special situations.
Elevated Tanks
Steel Riser
In localities where freezing temperatures do not occur, the purchaser
may specify a small steel riser. In other locations and unless a small
pipe is specified, a steel riser not less than 36 in. (910 mm) in outside
diameter (OD) shall be furnished. Where the riser pipe supports a
considerable load, the riser diameter and thickness shall preferably
be determined by the constructor.
Cold Climates
The minimum riser diameter of 36 in. (910 mm) shall be increased in
cold climates unless the riser is heated to prevent freezing. The proper
diameter depends on the extent of the tank’s use and the temperature
of the water supplied. In extremely cold climates, a minimum diameter
of 72 in. (1,830 mm) is recommended.
Manhole
Large-diameter risers shall contain a manhole about 3 ft (0.91 m) above
the base of the riser. The manhole shall not be less than 12 × 18 in.
(305 × 457 mm), and the opening shall be reinforced or the riser plate
so designed that all stresses are provided for around the opening.
Safety Grill
A safety grill is intended to prevent a person from falling down the
riser and shall be exempt from the design loads specified in AWWA
D100 Section 3.1.6. When a safety grill is used in the top of the riser
during erection, it shall be removed if the tank is located in climates
where freezing is likely to occur. When grills are left in place, they

shall be provided with a hinged door that is at least 18 × 18 in. (457 ×

457 mm).
Expansion Joint
Where the riser is not load bearing, flexibility to accommodate differ-
ential movements of the tank and riser foundation must be included.
This flexibility may be provided by an expansion joint or by riser lay-
outs that have sufficient offset to be axially deformed without over-
stressing the riser, tank, or foundation.
Pipe Connection
The pipe connection shall be of the size specified by the purchaser,
and it is usually attached to the riser bottom at a point designated by
the purchaser. Connections to the tank or piping furnished by the tank
constructor shall be made by the purchaser.
Silt Stop
If a removable silt stop is required, it shall be at least 6 in. (152 mm)
high, and the fitting or piping connection shall be flush with the riser
floor when the stop is removed. If a removable silt stop is not required,
the connecting pipe shall extend at least 6 in. (152 mm), and preferably
about 2.59 ft (789 mm), above the riser floor.
Inlet Protection
In risers 36 in. (910 mm) in diameter or larger, the inlet pipe shall be
protected against the entry of foreign materials dropping from above.
This shall be done by terminating the inlet pipe or the top of the silt-
stop pipe with a tee, with the “run” of the tee placed horizontally, or
by placing over the silt-stop or inlet pipe a circular plate 8 in. (203 mm)
larger in diameter than the pipe and located horizontally above the
end of the pipe or silt stop at a distance equal to the diameter of the
pipe. The circular plate shall be attached to the pipe, silt stop, or riser
bottom with a suitable bracket or welded bars. Adequate clearance
shall be provided between the ends of the elbow or from the edge of
the circular plate to the wall of the riser pipe to permit proper flow
of water through the inlet pipe. Pipe connections to the riser shell are
permitted, as long as adequate protection against freezing has been
provided.
Overflow
An overflow protects the tank from overpressure, overload, and pos-
sible catastrophic failure should the pumps or altitude valve fail to
shut off when the tank is filled to capacity. A properly operated tank
should not overflow during normal operation. An overflowing tank

is an emergency, and the malfunction causing the overflow should be

determined and corrected as soon as possible.
The tank shall be equipped with an overflow of the type and
size specified by the purchaser. If a stub overflow is specified, it
shall project at least 12 in. (304 mm) beyond the tank shell. For tanks
equipped with balconies, the overflow shall be extended to discharge
below the balcony. If an overflow to ground is specified, it shall be
placed down the tank shell and supported at proper intervals with
suitable brackets. The overflow shall be located such that it will not
be obstructed by snow or ground clutter. It shall terminate at the top
in a weir box or other appropriate type of intake. The top angle shall
not be cut or partially removed. The overflow pipe and intake shall
have a capacity at least equal to the inlet rate as specified by the pur-
chaser, with a head not more than 6 in. (152 mm) above the lip of
the overflow and in no case more than 12 in. (304 mm) above the top
capacity level, where a side opening–type overflow is used. The over-
flow pipe shall terminate at the bottom with an elbow, which shall
be directed away from the foundation. Unless otherwise specified by
the purchaser, the overflow pipe shall be steel pipe, with screwed or
welded connections if less than 4 in. (102 mm) in diameter, or with
flanged or welded connections if 4 in. (102 mm) or larger in diameter.
Overflows may be either internal or external as specified by the pur-
chaser. Minimum external overflow pipe thickness shall be 3/16 in.
(4.8 mm). Internal overflows are not recommended when tank usage
and climatic conditions are such that ice damage may occur to the
overflow or its attachments. When specifying an internal overflow,
the purchaser should consider the consequences of an overflow fail-
ure, which can empty the tank of its contents. Internal overflow pipe
shall be at least 0.25 in. (6.35 mm) thick. The end of the overflow may
be covered with a coarse, corrosion-resistant screen equivalent to 3/8
in. (9.5 mm) or larger mesh or with a flap valve, as specified by the
purchaser.
Ladders
Tower Ladder
A tower ladder shall be furnished with side rails no less than 2 in. ×
3/8 in. (51 mm × 9.5 mm), with a spacing between side rails of not less
than 16 in. (406 mm) and with nonskid rungs not less than 0.75 in. (19
mm) round or square, spaced 12 in. (305 mm) apart on centers. The
tower ladder shall extend from a point 8 ft (2.4 m) above the ground
up to and connecting with either the horizontal balcony girder or the
tank ladder, if no balcony is used. The ladder may be vertical but shall
not in any place have a backward slope.

Outside Tank Ladder

In all cases, a ladder shall be provided on the outside of the tank shell
connecting either with the balcony or with the tower ladder, if no
balcony is included. The outside tank ladder shall have side rails not
less than 2 in. × 3/8 in. (51 mm × 9.5 mm), with a spacing between the
side rails of not less than 16 in. (406 mm) and rungs not less than 0.75
in. (19 mm) round or square, spaced 12 in. (305 mm) apart on centers.
The tank ladder may be attached to the roof ladder.
Roof Ladder
Unless otherwise specified, the constructor shall furnish access to roof
hatches and vents. Such access shall be reached from the outside tank
ladder or riser ladder on pedestal tanks according to the following:
r For slopes 5/12 or greater, a ladder or stairway shall be pro-

vided.
r Slopes less than 5/12 and greater than 2/12 shall be provided
with a single handrail and nonskid walkway.
r Slopes 2/12 or less do not require a handrail or nonskid sur-
face.
Ladder Requirements
Minimum requirements for ladders, hatches, and so forth can be found
in OSHA 29 CFR Part 1910. Note: Regardless of the access protection
provided to tank roof hatches and vents, weather conditions on tank
roofs are extremely variable, and workers and their supervisors are
expected to exercise good judgment in matters of safety. Among other
things, this may include the use of safety lines when windy, icy, or
other hazardous conditions exist.
Ladder Safety Devices

If safety cages, rest platforms, roof-ladder handrails, or other safety de-
vices in excess of OSHA requirements are stipulated by the purchaser
or by state or local laws or other regulations, the purchaser shall so
specify. None of these devices are advisable on the submerged portion
of interior ladders in low-temperature climates.
Roof Openings
Above Top Capacity Level
An opening shall be provided above the top capacity level. It shall
have a clear dimension of at least 24 in. (610 mm) in one direction and
15 in. (381 mm) in the other direction and shall be provided with a
suitable hinged cover and a hasp to permit locking. The opening shall

have a curb of at least 4 in. (102 mm) high, and the cover shall have a
downward overlap of at least 2 in. (51 mm).
Tank Center
An additional opening with a removable cover having an opening
dimension or diameter of at least 20 in. (500 mm) and a neck at least
4 in. (102 mm) high shall be provided at, or near, the center of the
tank. This opening may also be used for the attachment of exterior
paint rigging. Where conveniently accessible to an outside balcony
or platform, a shell manhole may be substituted for the additional
opening. If properly designed, the shell manhole may be placed below
the top capacity level.
Vent
If the tank roof is of tight construction, a suitable vent shall be fur-
nished above the top capacity level, which shall have a capacity to
pass air so that at the maximum flow rate of water either entering or
leaving the tank, excessive pressure will not be developed. The over-
flow pipe shall not be considered a tank vent. Warning: An improperly
vented tank may cause external pressures to act on the tank that can
cause buckling even at a low-pressure differential.
Location
One tank vent shall always be located near the center of the roof, even
if more than one tank vent is required. For tanks with centrally located
access tubes, a reasonable offset of the vent is permissible. The vent
shall be designed and constructed to prevent the entrance of birds or
animals.
Screening
When governing health authorities require screening against insects, a
pressure-vacuum screened vent or a separate pressure-vacuum relief
mechanism shall be provided that will operate if the screens frost
over or become clogged with foreign material. The screens or relief
mechanism shall not be damaged by the occurrence and shall return
automatically to operating position after the blockage is cleared. Note:
The purchaser should clean the screens and check the pallets or relief
mechanism for operation at least once a year, but preferably each
spring and fall.
Additional Accessories and Exceptions

Any additional accessories required to be furnished shall be specified
by the purchaser. Exceptions to the provisions of this section may be
specified by the purchaser to suit special situations.

Tank Mixing Systems

Recently, many improvements have been made to the efficiency of
mixing systems that allow homogeneous mixing of all water each time
there is a fill and/or draft cycle (Fig. 2-18). New designs are needed that
accommodate these systems, which include appurtenances and acces-
sories not previously used. Because of differences among systems, the
following overview is offered to assist in determining appurtenances
and accessories to be used on a specific tank mixing design.
Placing the inlet and outlet nozzles of water-storage reservoirs
on opposing sides of the tank shell was for years considered to be
the optimum arrangement to achieve good water blending. Adding
an interior elbow to direct the water in a circular flow pattern was
considered a good practice to help avoid short-circuiting through the
outlet. In recent years, the industry standard practice has included the
installation of a hydraulic tank mixing system (TMS). A typical TMS
consists of an arrangement of piping placed internally in the reser-
voir to allow a homogenous mixing of all water in the reservoir each
time there is a fill and/or draft cycle. The TMS consists of a manifold
system in which both inflow and outflow waters pass through two
sets of properly sized and placed check valves. One set of valves is
used during the fill cycle, and the other is used to drain the tank. The
manifold can be a common inlet/outlet system connected to one shell
nozzle designed per API Standard 650 or to a bottom-flanged nozzle
Blind flange
2
Note
See
Manifold
Inlets Plan-outlet cross
Shell
Outlet cross Shell
Inlets
Existing outlet
Manifold
Existing inlet
FIGURE 2-18 Mixing system layout. (Notes: [1] Modification to inlet reservoir
may be required in order to install mixing system. Detail of existing
penetration must be provided. [2] Angle to be 30 degrees for water depth
30 ft [9 m] or below. Angle to be 45 degrees for water depth above
30 ft [9 m].)

connected to a 90-degree elbow. Alternate designs may include sepa-

rate inlet and outlet manifolds positioned either horizontally or verti-
cally depending on the physical height and diameter of the reservoir.
The size of the manifold piping and the placement, number, and
size of the inlet and outlet valves can be determined through calcu-
lations resulting from hydraulic studies and/or computational fluid
dynamics (CFD) modeling.
The types of materials used to construct the TMS manifolds
is largely a function of budget limitations. Materials may include
polyvinyl chloride (PVC), high-density polyethylene (HDPE), fusion-
bonded epoxy, liquid epoxy, or cement-mortar-lined (CML) and
epoxy-coated carbon- or ductile-steel pipe and fittings. At the upper
end of the scale, 316-grade stainless-steel pipe and fabricated fittings
can give a lifetime of maintenance-free service. The choice of materials
also depends on whether the TMS is being installed in a new tank (in
which case the parts can be placed inside the shell before the roof goes
on) or whether the project involves retrofitting an existing reservoir
with limited access.
Water Sampling Stations

As with TMSs, water sampling stations have become more sophisti-
cated in recent years. Because systems differ, the following overview
is offered to assist in determining appurtenances and accessories to
be used for a specific sampling design.
Testing the quality of the water in water-storage reservoirs is
becoming increasingly important. The use of chloramines for disin-
fection requires that the chlorine residuals in the water reservoir be
checked far more often. Placing water sampling points at various lo-
cations and levels in the reservoir enables the sampling technician to
also check the water for possible problems resulting from stagnation
and/or thermal stratification. These problems usually result from pe-
riods of low water usage—that is, in the off-peak season or when there
are rapid changes in the weather. They can also result from incomplete
mixing of the reservoir contents when fresh water is introduced to wa-
ter that has been in the tank for an extended period of time.
It is also important for the system operator to regulate the fill/
draw cycle to match periods of high and low consumption. Numerous
variables enter into the formulations that keep these ratios at optimum
conditions, and by sampling the water for pH, chlorine residual, and
bacteriological levels, the operator gains valuable information to help
monitor and control proper conditions.
The size of the reservoir determines the quantity, size, and location
of the sampling points. Small reservoirs with high turnover rates may
only require one sampling port, while in most reservoirs with a single
center roof-support column, it is suggested that at least three collection

points be positioned vertically in a fixed position near the shell. This

usually provides sufficient sampling data. The sampling points should
be set at elevations within the lower, middle, and upper-third of the
content levels. Direct sunshine on the tank shell, especially during
summer months, can raise the temperature of the water near the shell,
thus making the water sample misleading. Therefore, it is suggested
that the sampling points be placed at least 3 ft (0.9 m) away from the
shell, preferably on the north side of the reservoir. Water samples taken
at these points yield consistent levels and are more representative
of the tank water. For larger reservoirs, additional sampling points
from the middle and intermediate points should be considered. It is
possible to locate sampling points at variable levels through lines that
are attached to brackets connected to the intermediate roof-support
columns. The final number, size, and location of sampling points must
be carefully planned to allow sampling from all points within the
reservoir. Lastly, in some cases it might also be beneficial to locate a
sampling point directly in front of the inlet water nozzle to collect data
from the incoming water. This allows the operator to compare data
from the fresh water with data from the stored water.
Generally, lines 0.75 in. (19 mm) in diameter permit sufficient sam-
pling from all points within the reservoir. It is imperative that the lines
be constructed of noncorrosive materials such as PVC or 316-grade
stainless steel. For the convenience of the system operator, it is best to
have all sampling lines pass through the shell and into one secured
enclosure that is mounted to the shell at chest height. Samples can
be taken by opening control valves that are also made of 316-grade
stainless steel. These valves should be conveniently located within the
enclosure and positioned such that there is ample room beneath them
for the collection bottles. For security purposes, the enclosure should
be locked when not in use.
Antennas
The AWWA Steel Tank Committee has noted that the wireless commu-
nication industry has been installing antennas on ground and elevated
tanks at an ever-increasing rate. The major tank contractors have all
but forfeited these installations to non-tank constructors. Guidelines
were added to the commentary for AWWA D100–05 (see Appendix
A) to provide the owners and their consultants with additional infor-
mation when addressing these accessories. The guidelines consider
functional, structural, future maintenance, and safety issues having
to do with antenna and communication installations.
Health and Safety

Recommended safety precautions for radio frequency (RF) exposure
of personnel maintaining the tank should be reviewed with the wire-
less carriers.

Precautions shall be taken to prevent water contamination. Access

to the tank interior water compartment should not be permitted.
The paint system should be checked for hazardous metals. Where
hazardous metals are found in the paint system, the environment,
potable water, and workers must be protected from contamination.
Fall protection should be provided for workers. This may consist
of a safety rail around the installation or anchor points on the tank roof
where safety lines can be attached. The addition of auxiliary ladders
or safety lines for access to new equipment should be considered.
Antenna cables should be supported at regular intervals (about 4
ft [1.2 m] on center) in exposed locations. Cables should be attached to
ladders, as they present a safety hazard. Cable ladders or other com-
mercially available cable support systems are available and should be
installed separately on the tank.
Manholes and other access ports should not be obstructed by
the cable routing. Where space is limited (e.g., small-diameter 36-in.
[915-mm] access tubes in pedestal tanks), cables should be fitted to
the access tube well to maximize clearance.
Cables routed along balconies and platforms should be routed so
as not to obstruct access. Consideration should be given to provide
auxiliary painters-scaffold supports if the antenna installation renders
the existing system unusable. Antenna cables should be raised off the
tank surface to permit painting behind them.
Bibliography
Harper, W. B. 1986. Designing a More Corrosion-Free Water Storage Tank.
In Proc. 1986 AWWA Annual Conference, Washington, D.C.; Denver, CO.:
AWWA.
Matchett, B. 2006. Introduction to Improved Water Sampling Stations for Steel
Water Storage Tanks. NACE International, Channel Islands Section semi-
nar “Build a Tank in a Day,” Oxnard, CA.
Matchett, B. 2007. Introduction to Improved Tank Mixing Systems for
Water Storage Tanks. NACE International, Channel Islands Section semi-
nar “Build a Tank in a Day,” Oxnard, CA.

CHAPTER 3
Controlling
Corrosion
Mike Bauer and Joe Davis
Tnemec
Anthony D. Ippoliti
Sherwin Williams
Jeff Rog
Corrpro
The Nature of Corrosion

The corrosion of steel in aqueous solutions is an electrochemical pro-
cess in which a current flows and a chemical reaction occurs. Corrosion
is a natural process that follows the laws of science. All metals possess
inherent levels of electrical energy. They must maintain those levels
to remain stable and thus not be subject to the degradation process of
corrosion. These levels of electrical energy are measurable, and met-
als can be either more or less reactive in various environments on the
basis of their inherent levels of energy or electrical potentials. Metals
with higher levels of electrical energy tend to be more reactive, and
metals with lower levels of electrical energy tend to be less reactive
(more noble). Figure 3-1 illustrates how the various metals’ electrical
potentials compare.
The corrosion cell comprises four basic elements: anode, cathode,
electrolyte, and closure path (Fig. 3-2).
The anode is the metal that corrodes—that is, metal ions leave
its surface and enter the electrolyte solution. The cathode is a metal
from which no metal ions enter the solution. The electrolyte may be
any solution, such as drinking water, that is capable of conducting
81
Controlling Corrosion
82 Chapter Three
FIGURE 3-1 How

electrical Active
potentials of • Magnesium alloy
various metals • Zinc
compare.
• Aluminum alloy
• Cadmium
• Mild steel (new)
• Mild steel (old)
• Cast iron
• Stainless steel
• Copper, brass, bronze
• Titanium
• Gold
Noble
electricity. The closure path, also called the return current path, is
the electrical conductor (usually metal) that connects the anode and
the cathode. If any one of these elements is missing, corrosion does
not occur. For example, coating stops corrosion from occurring by
providing a barrier to the current that flows between the metal and
the electrolyte.
A dry-cell battery is a corrosion cell. When the battery’s anode
(zinc) and cathode (carbon) are connected through a closure path (the
lightbulb), the potential difference between the zinc and the carbon
Metallic path
Anode
Cathode
+
H OH +
Electrolyte
FIGURE 3-2 Elements of the corrosion cell.

Controlling Corrosion 83
Tank wall
(conductor)
Water
(electrolyte)
Anodic
area
(corrodes)
Cathodic
area
(protected)
FIGURE 3-3 Anode and cathode in a steel water-storage tank.
produces a current flow. The current continues to flow until the zinc
anode is consumed by the corrosion process.
It is important to consider why the current flows in the direction
it does. The direction of flow is determined by the metals selected
for the dry-cell battery’s case and center post. If the center post was
magnesium instead of carbon, the current flow would be reversed:
The magnesium center post would be the anode (which corrodes),
and the zinc base would be the cathode (which does not corrode).
The current can also be forced to flow in the opposite direction
if the standard carbon/zinc battery is connected to an outside cur-
rent source instead of the lightbulb. In this situation, the anode and
the cathode would also be reversed—that is, the battery case would
become the cathode and would be protected from corrosion.
In a steel water-storage tank, some portion of the metal will be the
anode and some portion will be the cathode (Fig. 3-3).
Which area takes on which function depends on impurities in
the metal; surface conditions; oxygen concentrations in the water; the
presence of any dissimilar metals; stresses caused by manufacturing,
heat, or concentrated structural loads; and/or several other factors. At
the anode, metal ions leave the surface, enter the water, and combine
with oxygen to form rust. Electrons released from the anode travel
through the metal to the cathode. At the cathode, an ion exchange
occurs, but no metal is lost and no corrosion occurs.
The presence of ladders, mixing systems, baffling systems, floats,
or other accessories made of stainless steel that are electrically

84 Chapter Three
continuous with the carbon-steel tank causes accelerated corrosion

in steel exposed at holidays (voids) in the coating. In such cases, the
stainless-steel components are the cathodes and the exposed-steel por-
tions of the tank are the anodes. Care must be taken when designing
such accessories to eliminate the galvanic or dissimilar metal corro-
sion between the metals of different electrical potentials. Methods
for addressing the corrosion caused by dissimilar metals are cathodic
protection, using a homogenous metal (coated carbon steel), using a
nonmetallic material such as fiberglass-reinforced plastic, coating the
stainless steel to minimize the cathodic surface areas the carbon steel
is reacting with, and/or electrical isolation of the dissimilar metals.
A corrosion engineer should review situations in which dissimilar
metals are used in steel water tank fabrication to determine the most
effective solution(s) for controlling corrosion.
Principles of Cathodic Protection

Cathodic protection systems are used to prevent or retard the cor-
rosion that would naturally occur in a steel water tank. These sys-
tems prevent or slow corrosion by altering the electrochemical envi-
ronment so that the submerged tank shell becomes the cathode of a
corrosion cell. Since the cathode of a cell does not corrode, the sub-
merged metallic tank shell is protected. There are two basic types of
cathodic protection systems: impressed-current systems and galvanic
systems.
Impressed-Current Systems
In an impressed-current system of cathodic protection, an outside
source of electrical power forces current into anodes submerged in the
storage tank’s water. The current flows from the anodes through the
water (electrolyte) and onto the submerged walls of the tank, making
the tank itself the cathode of the corrosion cell. An impressed-current
cathodic protection system (Fig. 3-4) consists of a manual or automatic
alternating current/direct current (AC/DC) converter (i.e., a rectifier),
feeder wires, and anodes inside the tank. The DC output voltage is
typically adjusted and controlled automatically to account for a wide
range of variables. To prevent damage to the coating, care must be
exercised to ensure that the polarized voltage does not exceed a max-
imum value as noted in the industry standards; otherwise the coating
may be damaged. Because excessive current output may damage the
coating, manually controlled rectifiers without automatic adjustment
and potential limiting capabilities are typically not recommended for
coated steel. The precise maximum negative voltage is dependent on
the characteristics of the coating and other factors.

FIGURE 3-4
Impressed-current
system.
Galvanic Systems
In a galvanic system (Fig. 3-5), a block of specially selected metal
called a sacrificial anode is immersed in the electrolyte and electrically
connected to the metal of the tank. The metal of the sacrificial anode is
selected so that it will become the anode of the corrosion cell, with the
steel tank being the cathode. Magnesium is the most common anode
material employed for corrosion control in potable water. The anodes
are typically of an extruded-rod type that are either suspended in the
water from the roof of the tank or suspended in the lower portion
of the tank supported from the sidewalls or by supports on the tank
floor.
These anodes are fabricated with a copper lead wire connected
to the core of the anode and then attached to the steel tank. When
the connection is made and the anodes are submerged in the water,
the current flow from anode to cathode (steel tank) begins; thus, the
magnesium corrodes and the steel is protected. Galvanic systems have
become increasingly popular because no electrical current is required.
Output of the sacrificial anodes may be monitored by using refer-
ence electrodes permanently installed in the tank below the surface of
the water. The anode lead wires and reference electrode lead wires may
be run into a test station installed at ground level to facilitate routine

86 Chapter Three
FIGURE 3-5 Galvanic system.
monitoring. A test station may also be equipped with a rheostat to

control the output of the anodes by altering the circuit resistance. In
the absence of a test station, routing testing is accomplished with a
portable reference cell.
Protective Coatings
Cathodic protection is normally used in conjunction with a well-
coated tank surface. The coating reduces the rate of anode consump-
tion and power use. Coatings typically have microscopic voids that
expose the metal to the water and allow metal loss if cathodic protec-
tion is also not in place. The ideal corrosion control system combines
a good dielectric coating (metallic coatings are not dielectric) and
a properly designed, installed, and maintained cathodic protection
system.

Exterior Corrosion of Tank Bottom

Cathodic protection systems are usually designed to protect the inte-
rior wetted surfaces of a water-storage tank. In some cases, however,
the exterior of a tank bottom or shell is in contact with corrosive soils.
This is typically the case for flat-bottom tank styles such as reser-
voirs and standpipes. In those situations, proper selection of the tank
base material or backfill may reduce corrosion, or a separate cathodic
protection system can be designed to protect exterior surfaces in con-
tact with soil (refer to American Water Works Association [AWWA]
Manual M27, External Corrosion: Introduction to Chemistry and Control,
for details). In cases where cathodic protection will be used to pro-
tect such surfaces, impressed-current systems are generally recom-
mended. Sacrificial anodes can also be used, depending on factors
including the size of the tank floor to be protected, soil or sand resis-
tivity, and whether the surface to be protected is coated or bare.
Cathodic Protection Design

In designing a cathodic protection system, the engineer must con-
sider the quality of the protective coating, tank geometry, surface area,
obstructions, geographic location, temperature, turbulence, and the
chemical composition of the water stored in the tank. Among items
to be specified in the design are the AC/DC converter or alternative
power source, the anode materials, and the anode configuration and
suspension.
Automatically Controlled AC/DC Converter (Rectifier)

The protective-current demands in a water-storage tank continuously
change because of variations in water chemistry and temperature,
fluctuations in water level, coating deterioration, and polarization ef-
fects. Automatically controlled impressed-current cathodic protection
systems (Fig. 3-6) are typically used in water-storage tanks to adjust
for these variations.
Reference electrodes are used to continuously monitor the pro-
tective level and control the cathodic protection current delivered to
the structure by the system. Separate control circuits are used for riser
pipes and other areas of the tank that may have different localized
conditions.
Anode Materials and Design Life

Impressed-current anodes are typically made of mixed-metal oxide or
platinum that can provide a nominal minimum 10- to 20-year life. The
design life of the anode system is based on the anticipated protective-
current requirements, condition of the coating, percentage of bare steel
to be protected, and the known consumption rates for the selected

88 Chapter Three
FIGURE 3-6 Automatically controlled AC/DC converter (rectifier).
anode material. The approximate consumption rate is 17 to 20 lb/amp

year for magnesium anodes. For platinized niobium, platinized ti-
tanium, or mixed-metal oxides on titanium, the approximate con-
sumption rate is 0.00008 to 0.0013 lb/amp year (0.036 to 0.59 g/amp
year).
Anode Configuration and Suspension

Distribution of current from any anode configuration is affected by
the geometric shape of the tank, obstructions within the tank, interior
coating, and chemical characteristics of the water.
The anode system may be installed vertically or horizontally. Ver-
tically suspended anodes (Fig. 3-7a ) are installed by hanging the an-
ode from an electrically insulated device at the tank roof adjacent to
holes cut in the roof. Horizontally suspended anodes (Fig. 3-7b) are
positioned below the normal water level attached to the tank shell
or access tube. In elevated tanks with an inlet/outlet riser pipe 30 in.
(76 cm) in diameter or larger, a vertically suspended anode is used to
provide protection within the riser.
When holes are cut in the roof, the finished installation must be
watertight to eliminate openings for insects and runoff to enter the
tank.
For tanks subject to icing, either vertical anode systems with exten-
sible elements or horizontal suspension systems designed to minimize

FIGURE 3-7a Vertical system reservoir tank.
ice damage to the anodes should be considered. These suspension sys-

tems can provide year-round protection and may eliminate the need
for annual anode replacement due to ice damage.
Maintenance of Cathodic Protection System

Following the installation of a cathodic protection system, a tank-
to-water potential profile is performed to ensure that the system
is providing optimal corrosion control. The level of corrosion con-
trol achieved by the cathodic protection system can be determined
through electrical testing. Corrosion is under control when a copper/
copper sulfate reference electrode is placed adjacent to, but not touch-
ing, the submerged tank surface and a polarized tank-to-water poten-
tial of −0.850 V or less is measured.
Automatic rectifiers continuously monitor the tank-to-water po-
tential being maintained by the system and make adjustments to

90 Chapter Three
FIGURE 3-7b Horizontal system reservoir tank.
control corrosion. Personnel responsible for operating and maintain-

ing the cathodic protection system should refer to the designer’s in-
structions to fully understand their responsibilities. They should con-
sult with the manufacturer if necessary regarding the equipment’s
operation and make certain that all responsible personnel are fa-
miliar with its operation. A successful cathodic protection corrosion
control system will continuously operate according to established
criteria.
Annual inspection of the cathodic protection system by the man-
ufacturer or by a qualified corrosion engineer is recommended. At
a minimum, this inspection should include an overall examination
of the entire cathodic protection system, replacement of all defective
parts, a potential profile survey, a physical check of the anode place-
ment, and a written report. Various annual service plans are available
from the cathodic protection companies or other service organizations.
Cathodic protection systems should be regularly tested and inspected
to ensure that they provide the maximum level of corrosion control to
the surfaces of the submerged steel tank.

Paint (Coating) Basics

Paint is a generic term used to describe a protective or decorative
coating that can be applied to a surface. Protective coatings and linings
applied to potable water tanks are formulated to protect the substrate
from corrosion on tank interiors as well as for corrosion protection
and aesthetic value on tank exteriors. The many types of protective
coatings available vary in their intended use, formulations, methods
of application, and in how they dry or cure.
Components of Paint
Most paints are made up of three primary components—the solvent,
which is incorporated into the formula to lower viscosity and allow
the painter to get the paint out of its container and onto the substrate;
the resin or binder, which binds the material together and, more than
any other component, determines the physical properties and perfor-
mance of the cured film; and the pigment, which can provide color,
hiding, or any number of other desirable properties in the film (e.g.,
gloss control, sag resistance, or added film strength). The combination
of the solvent and the resin is called the vehicle (Fig. 3-8). The resin
binder and the pigments make up the protective dried film after the
solvent evaporates. Most paints also contain additives, which will also
be covered in this section.
Volatile solvents
Vehicle
Resins (binders)
Pigments
FIGURE 3-8 Primary components of paint: solvent, resin, and pigment.

92 Chapter Three
Organic Solvents
Most paints contain volatile organic compounds (VOCs). These or-
ganic solvents are incorporated into paint formulas to lower viscosity
and allow the painter to get the paint out of its container and onto
the substrate. Water acts as a solvent for some types of paint, but it
is not an organic solvent and therefore is not considered a VOC. Or-
ganic paint solvents generally fall into one or more of the following
classifications:
Active solvent: An active solvent is a true solvent for the resin
(binder) portion of the formula. It dissolves the resin and keeps
it in solution.
Diluent: Although not a solvent for the resin portion of a partic-
ular paint formula, a diluent can still be used in conjunction
with a true (active) solvent, without causing precipitation or
incompatibility (“kick-out”).
Latent (auxiliary) solvent: This is not a true solvent, but com-
bined with an active solvent, it increases the strength (solvency
power) of the active solvent.
Solvent Properties That Determine Use

Before deciding what solvent or combination of solvents can be used
in a particular paint formula, two important solvent properties must
be considered:
Evaporation rate Some solvents are more volatile than others: The
greater the volatility, the faster the evaporation rate. Because of its
effect on application properties, it is important to consider the evap-
oration rate when selecting a solvent or solvent blend for a coating
formula. For example, the use of a “fast” solvent may be appropriate
if the coating is typically spray applied, but the result may not be a
smooth, continuous film if the coating is applied by brush or roller.
Flash point Flash point is defined as the temperature at which the

vapor directly above a liquid will ignite when exposed to a spark or
an open flame. The faster the evaporation rate of the solvent, the lower
the flash point. The U.S. Department of Transportation defines paints
with a flash point below 100◦ F (38◦ C) as flammable and paints with a
flash point above 100◦ F (38◦ C) as combustible. A label picturing a red
flame must be affixed to containers holding flammable liquids with a
flash point below 100◦ F (38◦ C).
Solvent Types or Families

Aliphatic hydrocarbons Derived from the distillation of crude oil,
aliphatic hydrocarbons are considered weak solvents for most resin
types. In a limited number of paint formulas, however, they can
be active or true solvents. Examples are oil-based paints, or alkyds.
Naphtha and mineral spirits are the most commonly used aliphatic
hydrocarbons.

Aromatic hydrocarbons Aromatic hydrocarbons are also derived from

distillation of crude oil, but they are considered stronger solvents for
a wider range of generic paints than the aliphatic hydrocarbons. All
aromatic hydrocarbons contain a benzene-ring molecular structure.
In fact, benzene is a solvent and the base molecule for this family of
solvents. Aromatic hydrocarbons are active solvents for generic paints
including alkyd, oil-based, chlorinated-rubber, certain epoxies, and
a few others. They are also used extensively as diluents. Aromatic
hydrocarbons most widely used in paint formulations are toluene,
xylene, #100 solvent, and #150 solvent.
Ketones This family of organic solvents has very high solvency

power for most generic paints. Acetone, often used for cleanup in or-
ganic chemistry labs, is the base molecule for this family. Ketones most
often used in paint formulas are MEK (methyl ethyl ketone), MIBK
(methyl isobutyl ketone), MNPK (methyl normal propyl ketone), and
MAK (methyl amyl ketone).
Esters Esters, like ketones, have high solvency power for most
generic paints and are used most often in lacquers and furniture fin-
ishes. They have limited use in industrial coatings because of their
high cost and reactivity with certain resins. Ester solvents most often
used in paints are ethyl acetate, butyl acetate, isobutyl acetate, and
amyl acetate.
Alcohols Alcohols are not true (active) solvents for most generic
paints. Exceptions are found in vinyl wash primers and ethyl sili-
cate inorganic zinc–rich primers. Certain water-soluble alcohols are
also used as co-solvents in water-based paints and in water-emulsion
paints. Alcohol solvents found in paints include ethyl alcohol (drink-
ing alcohol), isopropyl alcohol (rubbing alcohol), butyl alcohol, and
amyl alcohol.
Glycol ethers and glycol-ether acetates Glycol ethers are unusual in that
several are water soluble yet also have high solvency power. Because
of their water solubility, they are often used as co-solvents in water-
based paints. Glycol-ether acetates are strong solvents that are often
used in urethane paint formulations.
Pigments and Their Functions

Nearly all paints contain pigments. Exceptions are certain high-gloss
clear coatings. Unlike dyes, pigments used in protective coatings are
essentially insoluble in water and organic solvents. Their chemical
structure can be either organic or inorganic depending on the pigment
type. Most inorganic pigments are derived from minerals that are
mined from the earth. Organic pigments are made synthetically and
are typically much more expensive than inorganic pigments.

94 Chapter Three
Paint manufacturing would be a fast, simple process if it was not

for the need to include one or more pigments in most paint formulas.
Considerable energy and time are required to disperse pigments into
the liquid components of any paint formula. During initial manufac-
turing, the pigment suppliers grind pigments to a very small parti-
cle size. Regrettably, these small pigment particles agglomerate be-
fore they are added to protective coatings. When dispersing pigments
during paint manufacture, the intent is to break apart the pigment
agglomerates into the individual particles originally produced by the
pigment manufacturer. The amount of energy and time required to
accomplish this depends on the type of pigment. In general, organic
pigment agglomerates are much more difficult to disperse than inor-
ganic. Most inorganic pigments can be dispersed using high-speed
dispersion equipment, which is generally the fastest and the most
economical method of pigmented paint manufacture. More time and
energy are required to disperse most organic pigments; other manu-
facturing methods such as sand milling or ball milling are required.
To understand the important functions of pigment in paint, we
need to examine the three major classifications of pigments: prime
pigments, extender pigments, and corrosion-inhibiting pigments.
Prime pigments Prime pigments provide color and hiding power and
can be either organic or inorganic. Red iron oxide, yellow iron oxide,
titanium dioxide (TiO2 ), lead molybdate (toxic), and lead chromate
(toxic) are examples of inorganic prime pigments. Carbon black, ph-
thalocyanine blue, phthalocyanine green, and quinacridone violet are
examples of the more expensive organic prime pigments.
Extender pigments Most extender pigments provide little, if any, color

and hiding. All extender pigments are inorganic minerals that are
mined from the earth. These pigments are used in paint formulas
to provide a variety of desirable properties including gloss reduc-
tion, primer surface roughness for better topcoat adhesion, higher
solids, film reinforcement, lower moisture vapor transmission (MVT),
and thixotropy (sag resistance). Magnesium silicate (talc), barium sul-
fate (barytes), mica, aluminum silicate (clay), calcium silicate (wollas-
tonite), and silica (sand) are examples of extender pigments used in
paint.
Corrosion-inhibiting pigments Also known as active pigments,

corrosion-inhibiting pigments are typically inorganic and have very
low solubility in water. They help control corrosion of steel when used
in certain generic primers (primarily alkyd or oil based). Their low re-
activity with water produces an alkaline condition and/or passivating
ions that interfere with the electrochemical process that causes corro-
sion of steel. Lead tetroxide (red lead), zinc chromate, zinc phosphate,

calcium borosilicate, zinc molybdate, barium metaborate, and cement

are examples of corrosion-inhibiting pigments used in paint. Over the
years, red lead and zinc chromate have served as excellent corrosion-
inhibiting pigments for specific primer formulations, but they have
been virtually eliminated because of their toxicity.
Resins
With only a few exceptions, resins (binders) are organic polymers. De-
pending on molecular weight and resin type, these paint constituents
can be either liquid or solid before they are added to the batch.
The physical properties and performance of a cured paint film are
related more to the types of resins used in the paint than to any of the
other ingredients. This is why most paints are classified according to
the resins (binders) they contain. Single-component paints can some-
times contain more than one type of resin, but these paints are usually
classified by the resin that is present in the formula at the higher per-
centage, with the other resin identified as a modifier. For example, a
paint that contains 80 percent alkyd resin solids and 20 percent acrylic
resin solids would be classified as an acrylic-modified alkyd. Two-
component paints are generally classified using both types of resins.
For example, a two-component paint that consists of epoxy resin in
one component and polyamide resin in the second component would
be identified as a polyamide epoxy.
Curing Mechanisms Versus Resin Types

To gain a basic understanding of the performance properties of the
many generic paints available today, we will discuss the most common
curing mechanisms—that is, the various methods by which a paint
transforms from a liquid to a solid.
Lacquer cure Lacquer cure—the drying of a paint film by solvent

evaporation only—is the simplest curing mechanism. As raw mate-
rials, resins that fall into this category are usually high-molecular-
weight resins. Since no further polymerization takes place when the
dried paint film is formed, the performance properties are already ex-
hibited by the resin as a raw material. In general, the advantages of
paints that exhibit this curing mechanism include fast dry, complete
cure at low temperatures, and fast recoat times. With the exception
of coal tar pitch solutions, the major disadvantage is low solids (high
VOC content), which requires application of multiple coats. In addi-
tion, use of these paints is limited because of air pollution regulations.
Another disadvantage for all of the generic paints that fall into this
category is poor solvent resistance.
Vinyls: Organic solvent-based vinyl paints are made from vinyl-
chloride and vinyl-acetate polymers. They are characterized

96 Chapter Three
by excellent acid resistance, good alkali resistance, and low

MVT. Historically, they have been used on the interior and ex-
terior of steel potable water tanks and in applications where
good chemical resistance is desired. Their use today, however,
has become almost nonexistent because of air pollution regu-
lations.
Chlorinated rubbers: These resins are formed by a reaction of rub-
ber with chlorine. Like the vinyls, they have good chemical
resistance and low MVT. In the past, chlorinated-rubber coat-
ings were used on the interiors of potable water tanks. They
also have excellent adhesion to concrete and have been used
for swimming pools.
Coal tar pitch solution: Coal tar pitch is produced from the destruc-
tive distillation of bituminous coal. A protective coating can be
made by simply dissolving hot pitch in xylene. The cured film
has low MVT and good chemical resistance. Coal tar pitch so-
lutions can be formulated with a higher solids content (lower
VOC) than the other lacquer coatings. These coatings have ex-
cellent adhesion to steel and concrete and are typically used
for sewage and wastewater immersion service. They are only
available in black and have poor resistance to direct sunlight.
Nitrocellulose: This resin is produced by treating cellulose with a
mixture of nitric and sulfuric acids. Nitrocellulose lacquers are
widely used as clear and pigmented furniture finishes. Nitro-
cellulose is also used as a film-forming material in flexographic
and gravure inks. Sunlight resistance is poor, so these coatings
are most often used for interior applications.
Oxidation and polymerization Generic paints that exhibit the oxida-

tion and polymerization curing mechanism include oil-based paints,
alkyds, and epoxy esters. Following application of the wet paint film
and evaporation of the solvent, oxygen from the surrounding atmo-
sphere facilitates the linking together (polymerization) of the resin
molecules. Paints that cure by this process have been available for
many years.
Oil-based paints At the turn of the 20th century, the choice of generic
paints was very limited. Most paints were based on vegetable oils such
as linseed oil. These “oil-based” paints exhibited excellent wetting
and adhesion to marginally prepared steel substrates and provided
long-term weathering and corrosion resistance when exposed to many
atmospheric conditions. The biggest disadvantage was very slow dry
time.
Alkyds Alkyd technology was developed in the late 1920s. It was dis-
covered through laboratory testing that by combining a trifunctional

alcohol with an acid salt and reacting them at high temperature, an

ester-type resin could be formed that had very fast dry time. How-
ever, the film was too brittle to be considered as a paint binder. It was
then discovered that by incorporating a vegetable oil such as linseed
oil in the same reaction, resins could be produced that dried more
quickly than straight linseed oil, yet had good enough film properties
to be considered as paint binders. This development revolutionized
the paint industry at that time.
Further testing revealed that film properties and dry times could
be adjusted depending on the type and amount of oil included in the
reaction. Terminology that came out of this discovery included alkyd,
which stands for an alcohol/acid reaction, and short-oil, medium-oil,
or long-oil alkyd resins. The short–medium–long terminology desig-
nates the amount of oil the alkyd resin contains. Short-oil alkyds dry
more quickly than medium- or long-oil alkyds, but are generally less
flexible and do not produce paints with as good weatherability and
corrosion resistance.
In general, paints that exhibit this curing mechanism produce
paint films with a wide variety of properties depending on the type
and amount of oil in the resin. Overall, alkyd and oil-based paints
are less expensive than other generic types and can provide long-term
corrosion protection of steel as long as the exposure environment is
not too severe. Chemical resistance and color and gloss retention are
considered fair to good, depending on the formulation. Alkyd paint
formulations can be modified with other resins, such as acrylics and
silicones, to upgrade color- and gloss-retention properties.
Epoxy esters Epoxy esters are epoxy resins that have been “esteri-
fied” with fatty acids, resin, and so on. Epoxy-ester paints have better
chemical resistance than oil-based or alkyd paints, but they chalk more
readily when exposed to direct sunlight.
The co-reacting curing mechanism Paints that exhibit the co-reacting

(chemical cross-linking) curing mechanism usually have two or more
components. Examples are two-component epoxies, two-component
aliphatic urethanes, two-component fluorourethanes, and two-com-
ponent polyureas. In this mechanism, a chemical reaction between
two differing resin types is involved. In general, paints that cure by
this mechanism produce cured films with high cross-linked density,
excellent hardness, abrasion resistance, corrosion resistance, low MVT,
and excellent chemical and solvent resistance. One disadvantage is
limited pot life.
Two-component epoxies One component of two-component epoxy

contains an epoxy resin with one or more chemically reactive sites
known as epoxide rings. The second component (a curing agent)

98 Chapter Three
contains an amine or amide-functional resin. When the two paint com-

ponents are combined in the proper ratio and applied to a substrate,
they chemically react with each other to form the cured paint film. Per-
formance properties vary depending primarily on the type of curing
agent used. In general, polyamide-cured epoxies have slightly better
adhesion characteristics, better flexibility, longer pot life, and less acid
resistance than the amine-cured epoxies. It is also more difficult to
develop protective coatings with ultra-high solids (90 to 100 percent)
using polyamide-cured epoxies than it is for those using amine-cured
epoxies. With few exceptions, two-component epoxies chalk readily
when exposed to direct sunlight.
Two-component epoxies can be modified with coal tar pitch, pro-
ducing coal tar epoxy coatings. The coal tar pitch is usually added
to the polyamide or polyamine curing agent. Properly formulated,
these coatings can be applied up to 20 mil (508 m) dry film thickness
(DFT) in a single coat. They typically cost less than unmodified two-
component epoxies and have low MVT and good chemical resistance.
They are most often used for protecting steel and concrete substrates
from immersion in sewage and wastewater.
Two-component polyurethanes Urethane coatings are generally divided

into aromatic and aliphatic types. They are noted for their over-
all balance of high-performance properties. These include excellent
toughness, chemical and solvent resistance, hardness, and abrasion
resistance. One component typically contains a hydroxyl-functional
polyol resin that cross-links with a second component containing an
isocyanate-functional polyisocyanate resin.
Aliphatic polyurethanes usually contain an acrylic or polyester
polyol portion. When cross-linked with an aliphatic isocyanate, they
exhibit excellent color and gloss retention when exposed to direct
sunlight. They are expensive and are most often used as thin-film
topcoats over two-component epoxies.
Aromatic polyurethanes generally have better adhesion to steel
and are less expensive, but they tend to yellow when exposed to direct
sunlight. They are often used as direct-to-steel or direct-to-concrete
stand-alone coating systems. Some formulations can be used for im-
mersion service.
Two-component fluorourethanes Fluoropolymer-based coatings are

known for their outstanding color and gloss retention when exposed
to direct sunlight. Kynarr -based coatings have been around for many
years but require baking at high temperature for proper film forma-
tion.
Two-component fluorourethane coatings are a relatively new tech-
nology. These coatings exhibit the co-reacting curing mechanism at
ambient conditions. One component typically contains hydroxyl- and

carboxyl-functional fluoroethylene vinyl ether (FEVE) resin. Fluo-

rourethane coatings generally exhibit the same performance proper-
ties as two-component aliphatic polyurethanes, but they have much
better long-term color- and gloss-retention properties.
Two-component polyureas These coatings are usually 100 percent

solids, cure extremely quickly (sometimes within seconds of applica-
tion), and have extremely short pot life. Specialized plural-component
spray equipment is required. One component contains either aromatic
or aliphatic isocyanate-functional polyisocyanate resin; the other com-
ponent contains an amine-functional resin that cross-links with the
polyisocyanate resin when the two components are mixed together in
the proper ratio.
Because they are 100 percent solids and exhibit an extremely fast
cure response, these coatings can be spray applied in one coat at very
high film thickness. Depending on formulation, a wide range of per-
formance properties can be achieved using this technology. Rigid or
elastomeric (greater than 100 percent elongation) formulas are possi-
ble. Chemical resistance is generally very good, and some formula-
tions are suitable for immersion service. One hundred percent solids,
two-component polyurethane coatings are also available. Although
their properties are generally similar to those of the polyureas, they
are less flexible and have better chemical resistance.
Baking enamels These are coatings with two co-reacting resins pack-
aged in the same container. One resin is typically a polyester or acrylic
containing one or more hydroxyl groups. The other resin is usually a
melamine or urea formaldehyde. Cross-linking of the resin molecules
only takes place at high temperatures (generally above 250◦ F [121◦ C]).
Home appliances such as stoves, refrigerators, washers, and dryers
have metal substrates that are protected with baking enamels. Com-
ponents are painted in the shop and then sent through ovens, where
the coating cures (usually in 10 to 15 minutes) at high temperature.
Pot life is not a concern, since cure (co-reacting) will not take place at
ambient temperatures.
Moisture cure
Polyurethanes Some aromatic and aliphatic urethane resins are de-
signed to react with atmospheric moisture to form a cured film. These
resins are usually based on toluene diisocyanate (TDI) or dephenyl-
methane diisocyanate (MDI). In the presence of atmospheric mois-
ture (humidity), the resin molecules react with each other to form
cured films. Single-package aromatic urethane primers, intermedi-
ate coats, and topcoats containing aluminum or micaceous iron ox-
ide pigments are available that cure by this mechanism. The aro-
matic urethane types typically have excellent wetting and adhesion

100 Chapter Three
characteristics, even to marginally prepared surfaces, and offer excel-

lent barrier-protection properties. For better color and gloss retention,
single-package aliphatic urethane topcoats are also available that cure
by this mechanism.
Coalescence
Emulsion (latex paints) Water-based paints such as acrylic and
polyvinyl acetate (PVA) emulsions exhibit the coalescence curing
mechanism. Acrylic emulsion resin (acrylic latex), for example, con-
sists of small droplets of acrylic resin emulsified in water. The ap-
plied wet paint film cures through solvent evaporation (mostly water)
and coalescence. As the water evaporates, the emulsified acrylic resin
droplets get closer together. When most of the water is evaporated, an
organic solvent known as a coalescing agent causes the resin droplets
to flow together, forming a smooth, continuous paint film. Water-
based emulsions are often used as interior (PVA) and exterior (acrylic)
house paints. They are also used for other architectural applications.
Substrates that are coated with emulsion paints include masonry (con-
crete, concrete block, plaster, and so on.), wood, and drywall. Specially
formulated acrylic emulsion paints are also available for direct-to-
metal applications.
Free-radical polymerization Polyesters and vinyl esters have been used

extensively for specialty applications since the 1960s. They are based
on unsaturated prepolymer resins that are dissolved in an unsaturated
monomer such as styrene. By addition of peroxide catalyst, carbon-
to-carbon double-bond sites react with each other to form the cured
film. Although the monomer (styrene) is volatile in the liquid state, it
acts as a cross-linking agent and is incorporated into the film.
The polyesters form hard, dense, chemical- and water-resistant
films. They are used primarily as laminating resins and gel coats for
the manufacture of fiberglass boats, shower stalls, bathtubs, bowling
balls, and so on.
The vinyl esters have excellent acid resistance and are used pri-
marily as fiberglass-reinforced and nonreinforced linings for steel and
concrete substrates that come in contact with strong acids. Because of
their reactive nature, vinyl esters have poor package stability, resulting
in short shelf life, especially at high storage temperatures.
Hydrolysis The primary use of ethyl silicates and polysilicates, which

are among the few resins that are not based on organic polymers, is
in the formulation of inorganic zinc-rich primers. The curing mech-
anism is similar to the curing mechanism for concrete (cement). An
inorganic zinc-rich primer contains considerable metallic zinc dust,
typically more than 70 percent by weight in the cured film. When the
liquid paint is applied and the solvent evaporates, moisture from the

surrounding air is absorbed into the film, setting up the hydrolysis pro-
cess. In the case of ethyl silicates, complex chemical reactions occur,
forming a silicon oxide (SiO) matrix in which a small amount of resin
binder holds together a large amount of metallic-zinc pigment. Upon
complete hydrolysis (curing), the applied film is very hard, dense, and
resistant to abrasion and solvents. Inorganic zinc-rich primers protect
steel from corrosion by galvanic action. The zinc becomes the anode
in the corrosion battery and the steel becomes the cathode. The zinc
sacrifices itself to protect the steel in a process that is similar to that
exhibited by galvanized-steel substrates. Inorganic zinc-rich primers
are resistant to high temperatures (up to 750◦ F [399◦ C]), are difficult
to topcoat, and have poor acid resistance.
Additives In addition to the primary paint ingredients (resin, pig-

ment, and solvent), most paint formulations contain one or more ad-
ditives. Although they comprise a minor portion of the liquid paint
(usually less than 2 percent by weight), additives play an essential
role. Most are rheology agents that positively affect the properties of
the paint in the can and during application. Following are a few of the
additives used to enhance paint properties.
Antisettling agents: There is probably more of this additive used
each year than any other paint additive. Since pigment is the
heaviest paint ingredient, gravity causes it to settle to the bot-
tom of the container. Antisettling agents do not actually pre-
vent pigment from settling, but they keep it from hard packing
in the bottom of the container. This makes it much easier to
stir the pigment into a homogeneous mixture with the other
ingredients. The most widely used antisettling agent is hydro-
genated castor oil (trade name: MPA).
Thixotropic agents: These additives are used to incorporate the
property known as false body or high viscosity at rest. This
property is essential to paints that are to be applied without
runs or sags to vertical surfaces at high dry film thickness.
Defoamers: Defoamers are essential additives for water-based
paints and certain solvent-based paints to prevent bubbles dur-
ing manufacture and/or in the applied film. They are normally
silicone based and change the surface tension of water and or-
ganic solvents.
Driers: For driers, metallic soaps are added to oil-based or alkyd
paints to accelerate the oxidation and polymerization process.
Mildewcides: Paint films that contain natural (nonsynthetic) in-
gredients promote mildew growth when they are exposed to
warm, damp conditions. Mildewcides are used as additives in
these types of paints to prevent mildew growth. Most alkyd
and latex paints contain mildewcides.

102 Chapter Three
Anti-skinning agents: These are antioxidants used in oil-based or

alkyd paints to prevent skinning (drying at the surface) in the
can.
Pigment wetting agents: These additives are usually silicone based
and are incorporated into paint formulas to assist the resin in
surrounding and wetting out each of the dispersed pigment
particles.
Coating Calculations
The spreading rate (coverage) of any gallon of paint depends on its
nonvolatile (solids) content. One gallon occupies a volume of 231 in.3
(0.0038 m3 ) or 0.1337 ft3 (0.0038 m3 ). If a gallon or liter of paint con-
tained no volatile (solvent) and it could be applied without any losses,
the spreading rate obtained applied at 1.0 mil (25.4 m) would be
1,604 ft2 (149 m2 ). This figure is expressed as the theoretical spreading
rate per gallon or liter.
If a gallon or liter of paint contains volatile and its percentage
of total volume solids is known, its spreading rate, wet-to-dry film
ratio, and cost per applied mil per square foot (per applied micron
per square meter) can be calculated as follows:
Theoretical spread rate @ 1.0 mil DFT = Percent of volume solids
× 1,604 ft2
Spread rate @ DFTs other than 1.0 mil = Percent of volume solids
× 1,604 ft2 /specified DFT
Wet film thickness = DFT/percent volume solids
Cost per mil per sq ft = Cost per gallon/spread rate @ 1.0 mil DFT
Example
Let us assume we have one gallon of paint that is 50 percent volume
solids. The specified DFT is 5 mil and the cost per gallon is $22. With
this information, we can make the following calculations:
Theoretical spread rate @ 1.0 mil DFT = 0.50 × 1,604 ft2 = 802
ft2 /gal
Spread rate at specified 5.0 mil DFT = 0.50 × 1,604 ft2 /5 = 160
ft2 /gal
Wet film thickness = 5.0/0.50 = 10 mil
Cost per mil per sq ft = $22/gal/802 = 2.74 cents/ft2
Corrosion Protection of Steel Water Tanks with

Liquid-Applied Coatings
A protective coating is a material that, when applied to a structure, iso-
lates the structure from its environment. Properly applied protective

coatings are a cost-effective way to protect both exterior and interior

tank surfaces. A coating applied to the interior wet surfaces of a tank is
also called a lining. Both exterior and interior coating systems must be
carefully selected to provide the best protection value for the money
based on coating life, effectiveness of protection, ease of application,
and ease of adding coats in future years. Many protective coating sys-
tems have become much more complex than the single-component
materials that were prevalent before 1970.
D102 is the AWWA standard for painting the interior and exte-
rior of steel water tanks. The objective of this standard is to provide
information about various coating systems for coating and recoating
the interior and exterior of steel tanks used for potable water storage.
Coating systems for new bolted-steel tanks are not covered by this
standard.
AWWA D102 is not a specification. AWWA standards describe
minimum requirements and do not contain all of the engineering
and administration information normally contained in specifications.
Specifying engineers often reference specific interior and exterior coat-
ing systems contained in D102 when writing detailed specifications
for a steel water tank painting project.
D102—Past, Present, and Future

The first edition of this standard was approved by the AWWA Board
of Directors on February 11, 1964. The second edition was approved
on January 28, 1978, and subsequently withdrawn on June 23, 1991.
The standard was reissued and subsequently approved by the AWWA
Board of Directors on February 2, 1997. The third edition was approved
in 2003.
Inside and outside coating systems contained in D102 have
changed substantially from one edition to the next (Table 3-1).
Following are the primary reasons that substantial changes have
been made to inside and outside coating systems over the years:
r Advances in coatings technology
r State and federal regulations limiting the amount of organic
solvent contained in protective coatings (VOC regulations);
these regulations essentially eliminated the use of chlorinated
rubber and vinyl coating systems
r Restrictions placed on certain coating ingredients such as red
lead and zinc chromate
r Introduction of National Sanitation Foundation (NSF)/ANSI
Standard 61 Drinking Water System Components–Health
Effects in the late 1980s, which virtually eliminated some in-
side coating systems because of potential extraction of high
levels of harmful ingredients into drinking water.

104 Chapter Three
Outside Coating
Standard Systems Inside Coating Systems
AWWA Alkyd Red lead/aluminum phenolic
D102–64 Vinyl Vinyl
Red lead/linseed oil, Zinc/phenolic
alkyd High-solids vinyl
Metallic aluminum Cold-applied wax
Hot-applied wax
Metallic zinc
Hot coal tar enamel
Cold-applied coal tar
Cold taste-and-odor tar
AWWA Alkyd Two-component epoxy
D102–78 Vinyl Vinyl
Alkyd/silicone alkyd Chlorinated rubber
Alkyd (two primer coats) High-solids vinyl
Chlorinated rubber/alkyd Hot-applied coat tar
Cold-applied coal tar
Metallic sprayed zinc
ANSI/AWWA Alkyd Two-component epoxy (two
D102–97 Alkyd (two primer coats) coats)
Alkyd/silicone alkyd Two-component epoxy (three
Vinyl coats)
Epoxy/epoxy/aliphatic Zinc-rich primer/epoxy/epoxy
urethane Vinyl
Zinc-rich primer/epoxy/ Hot-applied coal tar
aliphatic urethane Cold-applied coal tar
ANSI/AWWA Alkyd Two-component epoxy (two
D102–03 Moisture-cured coats)
polyurethane Two-component epoxy (three
Water-based acrylic coats)
emulsion Inorganic zinc/epoxy/epoxy
Zinc-rich primer/epoxy/ 100 percent solids
fluorourethane polyurethane
Epoxy/epoxy/aliphatic Organic zinc/epoxy/epoxy
polyurethane
Zinc-rich primer/epoxy/
aliphatic urethane
Sources: AWWA standards D102–64, D102–78, ANSI/AWWA D102–97, ANSI/

AWWA D102–03.
Note: ANSI = American National Standards Institute.
TABLE 3-1 Changes in Inside and Outside Coating Systems Specified in

Various Editions of AWWA D102

Inside and outside coating systems contained in future editions

of D102 will be further restricted by the continued implementation of
more stringent VOC regulations for both shop- and field-applied pro-
tective coatings. Future editions of D102 will most likely only contain
inside and outside coating systems that are very high solids based or
water based.
Surface Preparation
Before a protective coating system can be applied to a steel or concrete
water storage tank, appropriate surface preparation must be under-
taken. The purpose of surface preparation is twofold: to clean the
substrate of contaminants and to roughen or “profile” smooth sur-
faces to ensure mechanical adhesion of the first (primer) coat. Welds
may be ground, corners and edges may be smoothed, and voids may
be filled so that the applied coating system does not fail prematurely.
The Society for Protective Coatings (SSPC), established in 1950 and
headquartered in Pittsburgh, Pennsylvania, assesses and advances
surface preparation and its understanding by conducting research
and “developing standards, specifications, and guides covering tech-
niques and materials of surface preparation.” NACE International is
a professional technical society that provides education and commu-
nicates information to protect people, assets, and the environment
from the effects of corrosion. It, too, develops surface preparation and
other standards, provides education and certification, and publishes
numerous books and journals. Founded in 1943, NACE is the largest
organization in the world committed to the study of corrosion, with a
membership consisting of 15,000 engineers, scientists, and researchers
in 91 countries.
Together, these two organizations have issued joint standards that
are commonly referenced by those who need to specify proven sur-
face preparation methods. For example, AWWA D102–03 Standard
for Coating Steel Water-Storage Tanks cites four SSPC/NACE sur-
face preparation standards (SP10/NACE 2 Near White Blast Clean-
ing, SP6/NACE 3 Commercial Blast Cleaning, SP7/NACE 4 Brush-
Off Blast Cleaning, and SP11 Power Tool Cleaning to White Metal).
Surface preparation methods vary and may not be appropriate for all
materials of construction. Methods may use abrasive blast cleaning
(SP10/NACE 2, SP6/NACE 3, and SP7/NACE 4, for example), hand
or power tools (SP2, SP3, SP11, and SP15, for example), or water under
pressure (SP12/NACE 5, for example).
The surface preparation standards listed previously are primar-
ily used for steel surfaces. Methods such as SP13/NACE 6 Surface
Preparation of Concrete exist for cementitious substrates. In addi-
tion, SP13/NACE 6 further identifies surface preparation practices

106 Chapter Three
FIGURE 3-9 Surface preparation under controlled conditions.
developed by ASTM International, formerly the American Society

for Testing and Materials. Organized in 1898, ASTM International is
one of the largest voluntary standards development organizations in
the world. ASTM International is a not-for-profit organization that
provides a forum for the development and publication of voluntary
consensus standards for materials, products, systems, and services.
Surface preparation is so important to the successful completion of a
water tank project that members of AWWA, ASTM, and other orga-
nizations donate time to review and update industry standards and
methods and to develop new standards and methods as to surface
preparation equipment improves and changes.
Surface Preparation—Steel
For welded-steel water-storage tanks, surface preparation completed
in a fabrication shop before the first (primer) coats are applied is un-
derstandably faster and easier than surface preparation that must be
carried out after erection. Shop conditions are controlled, in that op-
erations can be continued regardless of outside weather. Lighting and
access to all areas of the structure being fabricated are generally su-
perior to field lighting and access (Fig. 3-9). After cleaning, steel plate
surfaces are abrasive blasted to remove mill scale and/or create a sur-
face profile to which the applied coating will adhere. Abrasive blasting

FIGURE 3-10 Open-nozzle blasting.
operations may be performed within an enclosed chamber or booth or

by an operator using an open nozzle. Speed and efficiency are two of
the advantages of centrifugal blasting machines; open-nozzle blasting
(Fig. 3-10) may allow an operator more time to prepare difficult-to-
access areas.
AWWA D102–03 Section 1.2, Surface Preparation, paragraph
4.5.2.1 states for outside surfaces of new tanks:
Exterior surfaces shall be cleaned in accordance with SSPC SP6/NACE 3.

If specified, tanks located in coastal areas or industrial environments shall
be blast cleaned to SSPC SP10/NACE 2. Blast cleaned surfaces shall have
a surface profile that is appropriate for the specific primer and coating
system as recommended by the manufacturer of the coating.
For outside surfaces of existing tanks, AWWA D102–03 Section

1.2, paragraph 4.5.2.2, states:
When the. . . coating system will adhere to the existing coating, all corro-
sion products and deteriorated coatings shall be removed by spot clean-
ing to SSPC SP11 or SSPC SP6/NACE 3 and the remainder of the exterior
surfaces shall be cleaned by SSPC SP7/NACE 4 or by washing with an al-
kaline cleaner. . . to remove all dirt, dust, coating/paint chalk, and foreign

108 Chapter Three
matter. When the new coating system is not compatible with the exist-
ing coating, all existing coatings shall be removed and the surfaces blast
cleaned to SSPC SP6/NACE 3 or, if specified, to SSPC SP10/NACE 2.
For interior surfaces of new tanks, AWWA D102–03 Section 1.2,

paragraph 4.5.3.1, states:
The interior surfaces of new tanks shall be cleaned in accordance with
SSPC SP10/NACE 2, excluding interior surfaces of dry risers and dry
pedestals. Interior surfaces of dry risers and dry pedestals shall be cleaned
in accordance with SSPC-SP6/NACE 3. Blast-cleaned surfaces shall have
a surface profile that is appropriate for the specific primer and coating
system as recommended by the manufacturer of the coating.
Finally, for interior surfaces of existing tanks, paragraph 4.5.3.2

states:
When existing coatings have not deteriorated extensively and the new
coating system will adhere to and is compatible with the existing coating,
all corrosion products and deteriorated coatings shall be removed by spot
blasting to SSPC SP10/NACE 2 and the remainder of the interior surfaces
shall be cleaned by SSPC SP7/NACE 4. When existing coatings have dete-
riorated extensively or the new coating system is not compatible with the
existing coating, all existing coatings shall be removed and the surfaces
cleaned to SSPC SP10/NACE 2. Blast-cleaned surfaces shall have a sur-
face profile that is appropriate for the specific primer and coating system
in accordance with the coating manufacturer’s recommendations.
Steel plate for bolted tanks may be prepared by first cleaning

and rinsing in a hot alkaline solution followed immediately by hot-
air drying. This process removes not only dirt but also hydrocar-
bon contaminants that will prevent future coating adhesion. There-
after, these plates are sent through a centrifugal blast machine where
an approximately 2-mil-deep (51-m-deep) profile is made on the
surface in accordance with SSPC SP10 Near White Blast Cleaning
procedures.
AWWA D103 Standard for Factory-Coated Bolted Steel Tanks for
Water Storage recommends that tank components to be protected with
glass coatings receive either Near White Blast Cleaning (SSPC SP10)
or, as an alternative, Pickling (SSPC SP8).
Surface Preparation—Concrete
Surface preparation for concrete water-storage tanks is different from
surface preparation for steel. Most surfaces of concrete tanks are al-
ready rough due to the way these vessels are constructed, so there is
little need to add a profile. This coarse surface characteristic allows for
good adhesion of paints and coatings. That does not mean, however,

that water tanks constructed of concrete are without their own surface
preparation requirements.
What needs to be accomplished is the removal of areas of poorly
adhered concrete—flake-shaped fragments that detach from the sur-
face of concrete in a process known as spalling. Laitance, a poorly
adhered layer of concrete, may also be found on concrete surfaces
and must be removed. Although less severe and less damaging than
spalling, laitance comprises cement and tiny particles called fines that
may be caused by improper vibration of concrete within forms. Left
unrepaired, spalling may continue and expose reinforcing bars to cor-
rosion, causing damage to the water tank that is difficult to repair.
One way to remove spalling and laitance is by mechanical means, in
accordance with ASTM D4259.
Roughening of concrete surfaces may be desired on cast-in-place
concrete surfaces, such as composite tank columns. These smooth, as-
cast surfaces may not be of uniform appearance, so abrasive blasting—
again in accordance with ASTM D4259—helps to regulate the appear-
ance of the completed pedestal tower. “Bugholes,” small irregular
cavities uncovered by surface preparation procedures, should be filled
with an appropriate material so they will not trap airborne contami-
nants and mold spores.
Occasionally, bolted tanks without bottoms are constructed and
placed on concrete slabs. These concrete slabs act as the floor and must
be constructed of materials that will not leach out into the potable
water supply and thereby contaminate it. NSF International, founded
in 1944 as the National Sanitation Foundation, created ANSI/NSF
Standard 61, a certification protocol that addresses these concerns.
Regarding “bottomless” tanks resting on a concrete base, NSF may
recommend that these concrete surfaces be constructed of ANSI/NSF
61-certified cements and admixtures, for example, or coated with an
ANSI/NSF 61-certified coating before being placed in service. These
concrete floor surfaces would also require surface preparation prior to
the application of ANSI/NSF 61-approved paints and coatings. Once
again, preparation routines in accordance with ASTM D4259 may be
used.
Even though most concrete tank surfaces may need little to no
roughening, they do require cleaning. The SSPC SP12 standard for
low-pressure water cleaning, WJ4, will accomplish this. This proce-
dure usually removes loose shotcrete clusters and may remove de-
bris left behind by, or concrete escaping through, placement forms.
Placement forms are coated with release agents or compounds to pre-
vent adhesion of concrete to these forms and thereby allow placement
forms to be removed cleanly. These compounds may act as contam-
inants, however, and prevent adhesion of paints and coatings. WJ4
may be specified for this purpose. If non–water soluble or non–water
dispersible form-release compounds are used, low-pressure water

110 Chapter Three
cleaning will not remove them, and abrasive blast cleaning through
the use of ASTM D4259 may be required.
Concrete surfaces, because they are porous, also absorb moisture.
Whether because of rain or the use of water when cleaning equipment
for surface preparation, these surfaces must be allowed to dry before
coating. The length of, and need for, time to dry are influenced both
by the temperature and humidity at the tank site and by the type of
coating to be applied to the tank.
Coating Selection
To the residents of the community or neighborhood closest to a water-
storage tank, how the tank looks—its exterior color and design—may
be the most important characteristic. The city name or the name and
mascot of the local high school may be emblazoned on the elevated
steel tank for all to see. In contrast, residents living near ground con-
crete storage tanks may not want them visible at all, desiring them
to blend into the background landscape. Enhancements are made to
both steel and concrete tanks to increase their visual impact. Such en-
hancements may be from fabrications such as pilasters or simply from
the allure of carefully chosen and illustrated paints and coatings.
Understandably, residents want the exterior surfaces of water stor-
age tanks to remain colorfast and appealing. But the interior surfaces,
which few see, are in fact more important, because it is there where
bacteria can grow and corrosion can occur undetected. Coating sys-
tems for both interior and exterior surfaces can be selected based on
information provided by tank fabricators, by coating manufacturers
and their representatives, by the owner’s or specifier’s preference, or
by reviewing applicable AWWA standards.
Interior Surfaces—Welded-Steel Tanks

An interior coating system in the wetted areas of the tank must with-
stand constant immersion; it must be able to resist alternate wetting
and drying in the upper portion of the operating range and high hu-
midity above the top capacity level; it must be resistant to the actions of
ice abrasion in cold climates; and, in some geographic areas, it must
be able to withstand extreme temperature fluctuations. In addition,
interior coating systems must be able to be both shop-applied and
field-applied, must be cost-effective, and must meet the minimum
requirements of ANSI/NSF 61 Standard for Drinking Water System
Components.
For interior surfaces of tanks, ANSI/AWWA D102 lists five interior
coating systems (Table 3-2). Such coatings “shall have been evaluated
for long-term fresh water resistance and the system shall have demon-
strated satisfactory service in fresh water for at least 18 months. Any
coating that cannot meet these requirements, whether or not included
in this standard, shall not be used.”

ICS Number Description

1 Two-coat, two-component epoxy
2 Three-coat, two-component epoxy
3 Three-coat, zinc/epoxy/epoxy, interior nonimmersed
surfaces above TCL
4 One-coat polyurethane or polyureas
5 Three-coat zinc/epoxy/epoxy, interior surfaces above
and below TCL
Note: ICS = interior coating system; TCL = top capacity level.
TABLE 3-2 Five Interior Coating Systems
Interior Surfaces—Bolted-Steel Tanks

Interior surfaces of bolted-steel tanks are subjected to the same condi-
tions as are welded-steel tanks and must also be protected by specially
designed coating systems. ANSI/AWWA D103 Section 10.3–10.6 al-
lows four protective coating systems to be used: galvanized coatings,
glass coatings, thermoset liquid (epoxy) coatings, and thermoset pow-
der (epoxy) coatings.
Glass coating systems may begin with a primer coat of catalytic
nickel oxide prior to the application of 6 to 19 mil (152 to 482 m) of
the glass coating, firing, and fusion of the glass coating to the substrate
in a furnace operating at temperatures above 1,200◦ F (649◦ C).
Thermoset liquid (epoxy) coating systems are generally applied
in two coats. The first coating is applied and heated to create a tacky,
partially cured first coat followed by additional coating to achieve
a minimum 5-mil (127-m) DFT. It is subsequently baked at 425 to
525◦ F (218 to 274◦ C), to thermally cross-link the complete coating
system.
Thermoset powder (epoxy) coating systems are electrostatically
applied to achieve a minimum 3-mil (76-m) DFT prior to oven curing
and baking.
According to ANSI/AWWA D103, when galvanized coatings
are to be supplied, “zinc metal suitable for immersion in drinking
water shall be applied to the tank parts after fabrication in accor-
dance with the recommended practice of the American Hot Dip Gal-
vanizers Association in compliance with ASTM A123 and ASTM
A153.”
Exterior Surfaces—Welded-Steel Tanks

As mentioned earlier, the exterior coating attracts more attention and
may cause more concern than any other aspect of a tank project be-
cause it is all that the neighboring community sees.

112 Chapter Three
OCS Number Description

1 Three-coat (optional four-coat) alkyd
2 Three-coat moisture-cured urethane
3 Three-coat water-based acrylic
4 Three-coat zinc primer/urethane/fluorourethane
5 Three-coat epoxy primer/epoxy/urethane
6 Three-coat zinc primer/epoxy/urethane
Note: OCS = outside coating system.
TABLE 3-3 Six Outside Coating Systems
For exterior surfaces of tanks, ANSI/AWWA D102–03 lists six

outside coating systems (Table 3-3). According to the standard,
“[p]roprietary formulations will be acceptable provided the coating is
of the same generic type and that the performance of the formulation
offered meets or exceeds the performance of the formulation defined
in the referenced coating standard and also is suitable for the specified
service conditions.”
Exterior Surfaces—Bolted-Steel Tanks

Tanks fabricated in compliance with ANSI/AWWA D103 and pro-
tected with thermoset liquid or thermoset powder coatings are top-
coated with acrylic or urethane baking enamels to yield a minimum
of 3-mil (76-m) DFT.
Regarding the occasional repair and touch-up of welded- and
bolted-steel tank coating systems, it is very important to remember
that such activity does not require the full removal and replacement
of the coating system. Repair and touch-up are only required in areas
that have been damaged, and it is only after many years of service that
welded- and bolted-steel tanks require overcoating and/or complete
repainting. Table 3-4 describes common water tank coating character-
istics.
Application Techniques and Equipment

By definition, a protective coating is a material that is applied to the
exterior of the tank that acts as an insulator between the tank and its
environment. A protective lining has a similar definition, the major
difference being that the protective lining system is applied to the
interior of the tank and acts as a barrier between the tank and its
cargo.

Generic Color and Gloss Corrosion
Type Retention∗ Resistance† Adhesion‡ Long-Term Durability§
Alkyds Fair to good Fair to good Good Good
Acrylics Very good Good to very good Very good Very good
Urethanes Excellent Very good Very good to excellent Excellent
Fluorourethanes Outstanding N/A Very good to excellent Excellent
Polyureas ¶ Very good Excellent Excellent
Epoxies ¶ Very good to excellent Excellent Excellent (topcoat outside)
Zinc-rich primers ¶ Excellent Very good Excellent (as part of system)
Glass coatings Excellent Excellent Excellent Excellent
∗ Some generic classifications are not used as topcoats.

† Some generic classifications do not include primers.
‡ Each generic classification is rated for its adhesion to primer or existing coating.
§ Some generic classifications should not be used as topcoats.
¶ Polyureas, epoxies, and zinc-rich primers are interior coatings or are topcoated. They do not have color and gloss exterior “expectations.”

TABLE 3-4 Common Water Tank Coating Characteristics
113
114 Chapter Three
Many considerations need to be taken into account when selecting

the proper interior and exterior protective systems, including govern-
mental regulations, various environmental considerations, the effec-
tiveness of the protection, the ease (or difficulty) of application, and
the anticipated service life of both the coating and the structure. For
instance, an interior lining must be able to withstand constant im-
mersion in water, varying water temperatures, alternate cycles of wet
and dry periods, ice abrasion, high humidity and heat, and varying
levels of chlorine and mineral contact. The lining materials must not
pose a health risk to the general public and must be approved for
such use by the appropriate state or federal regulatory agency. Alter-
nately, an exterior coating system should take into consideration the
type of atmosphere to which it will be subjected, the expected ambient
temperatures, the areas surrounding the tank, and the desired overall
appearance and aesthetic value of the coating.
In some cases, the entire coating system may be applied in a shop
environment; in other cases, the coating system may be applied en-
tirely in the field. Quite often, though, it is in a combination of these
two settings. Each method will be examined individually.
Shop Application
The two major advantages in applying a protective coating in a shop
environment are control and accessibility. Often the interior and ex-
terior sections receive just a primer coating in anticipation of field
application of the subsequent coatings specified. This is done to al-
low fabricators to quickly clean, prepare, and prime the surface in
accordance with specifications while still allowing them the ability to
continue working the plate and shipping it without compromising the
surface preparation. When this approach is taken, an area is generally
left uncoated around the perimeter of each plate, commonly referred
to as the margin. The margin area can vary by specification but is usu-
ally between 4 and 6 in. (102 and 152 mm) wide. This allows for field
welding to be performed during the erection in the field. Of course,
these areas will not meet the surface cleanliness requirements of the
specification, and they will need to be addressed in the field prior to
the continued application of the coating system. This will be covered
in more detail in the following section on field painting.
In some instances, the entire protective coating system is applied
in the shop—for example, with bolted tanks. The steel panels are gen-
erally coated following roll forming and bolt-hole punching. In this
situation, a thermoset liquid coating may be applied and then baked
at a prescribed temperature, or a thermoset powder coating may be
applied and then baked according to the manufacturer’s instructions.
Depending on the type of interior lining system, the bake tempera-
tures can vary between 425◦ F (218◦ C) and (in the case of a glass lining)

1,600◦ F (870◦ C). Factory-applied lining systems are discussed in more

detail in AWWA D103. The factory-applied liquid baked-on exte-
rior coating systems generally combine epoxy primer with an acrylic
enamel topcoat or an acrylic polyurethane topcoat.
The basic premise of a shop application of the protective coating
system, whether it be the primer or the entire system, is that since
application is all performed on the ground and under controlled shop
conditions, the result will be (and most often is) a very uniformly
applied and fully cured protective coating system. However, the coat-
ing system is often damaged during loading, transporting, unloading,
and erection of the plates. Depending on the extent of the damage, ma-
jor field work may be needed for repair. Anticipating this situation, a
combination of shop priming and field painting is often employed.
Field Painting
Although coating and/or lining systems applied in the field share
many of the considerations we reviewed under shop-applied coat-
ings, other factors specific to field painting need to be evaluated.
Among these items are the type of lining that is environmentally com-
pliant, tank heating and ventilating, dehumidification requirements,
the landscape surrounding the tank, and the type of environment that
the tank is subject to during the preparation, application, and cure
of the lining and coating systems (e.g., chloride sources in a marine
environment).
Generally speaking, the interior of the tank requires the highest
degree of surface cleanliness and preparation. Many of the protective
lining systems require a minimum surface cleanliness equaling an
SSPC SP-10/NACE 2 Near White Metal Blast. In an effort to achieve
this, painting contractors typically blast the bottom of the tank first
and then begin to blast the wall section by section. Each section (called
a drop) is blasted and coated during the work shift unless an environ-
mental control such as dehumidification is needed. The abrasive used
in the cleaning process is allowed to fall to the floor of the tank and ac-
cumulate there. This abrasive provides an insulation of sorts from the
environment so that the initial blast on the floor is held or maintained.
If the blast is lost, the contractor reblasts the floor area and coats it as
he is finishing the interior of the tank. Special care needs to be taken to
ensure that spent blast media is not billowed and deposited into the
freshly coated surface. Although this is a common approach to lining
the inside of storage tanks, it is not the only correct way to perform
this task.
When a primer has been applied in the shop and the contractor is
only applying finish coats to the tank in the field, the surface prepa-
ration specification is usually a bit different than that just described.
Two concerns must be addressed: (1) the condition of the shop primer

116 Chapter Three
and its preparation before receiving any topcoats and (2) the weld
seam preparation. Typically, the shop primer is swept blast according
to SSPC SP7/NACE 4 Brush-Off Blast Cleaning, while the weld seams
are prepared according to SP10/NACE 2 Near White Metal Blast. In
some cases the primer is reapplied, while in others only the finish
coat(s) are applied. Proper ventilation of the tank’s interior is critical
to ensure a thoroughly cured lining.
When coating the outside of the tank, the consequences of over-
spray, dry spray, and ambient conditions must be considered. Adverse
conditions during surface preparation, coating application, and cur-
ing can (and most often do) lead to premature catastrophic failure of
the coating system. Specifiers and contractors should also be aware of
the areas surrounding the tank and the environment to which the coat-
ing system will be subjected. For example, an elevated water tank in a
congested urban area may require coating materials that can only be
roller applied or that tend to “dryfall” if applied by a sprayer. By way
of environmental considerations, coastal regions may require a coat-
ing system that has a higher film build and more barrier protection to
protect the tank from a chloride-rich environment. Depending on the
tank and the contractor, the coating process is completed in different
ways, but completing drops is still the most common way to ensure
that a properly cleaned surface is maintained. Typically, the specified
cleanliness for the exterior of the tank would be an SSPC SP6/NACE 3
Commercial Blast. For a shop-applied primer, the primer is swept blast
according to SP7/NACE 4, while the weld seams require SP6/NACE
3. Again, a primer may be reapplied if specified or the finish coats
may be applied over the prepared existing primer.
Equipment and Techniques

Many shops that use liquid coatings apply them with spray equip-
ment, which has undergone many improvements since the intro-
duction of conventional spray units. This section briefly describes
methods and equipment for applying protective coating and lining
systems in both shop and field. In the field, paint contractors typi-
cally use rollers or airless spray systems. However, because of new
coating technology and new environmental concerns and legislation,
plural-component spray equipment is increasingly showing up on
tank coating projects. The use of brushes and rollers in a shop envi-
ronment is mostly limited. Brushes and rollers are typically used for
touch-up or to coat difficult or complexly designed areas.
Brushes
Brushes are not as high tech as sprayers, and many consider them an
outdated way to apply paint. However, many situations still require
their use. A “stripe coat,” often specified for added protection of edges,

rivets, corners, bolts, and welds, is best applied by a brush, because it

improves the wetting capabilities of the coating by forcing the coating
into areas that would be problematic if sprayed or roller applied. A
worker applying a coating by brush should hold the brush at an an-
gle of approximately 45 degrees and, using the wrist and arm, spread
the coating evenly and quickly onto the substrate. Once the coating is
evenly spread, it should be smoothed by light brush strokes to elimi-
nate any irregularities. When the next brushload of coating is applied,
the final smoothing strokes should extend from the newly applied
coating into the previous brush-applied adjacent area. This spreads
the coating over the overlap between the two areas and provides a
smooth, uniform coating over the whole surface while maintaining
the wet edge (the end of the stroke of the previous applied coating).
As the next brushload of coating is applied, the painter should sweep
the coating from the substrate back into the wet edge of the previous
application to help prevent lap marks. It is also important to brush
out over an edge, not against it. Brushing against an edge creates an
action that pulls the coating away, causing a thin area on the edge.
Rollers
Rollers have earned a bit more respect than brushes, but their produc-
tion results still pale in comparison with results of spray application.
Application by roller is faster than application by brush, but is not
quite as fast as application by spraying. Because the roller cover holds
considerably more coating than a brush, a much larger area can be
covered with one load. Rollers are excellent for large, flat areas—for
example, the tops or sidewalls of tanks. Rollers can be used wherever
the skill of a brush or spray application is not called for. Rollers can
also be used if spray applications are prohibited due to overspray
concerns.
The procedure for using the roller is to immerse it in the coating
tray or bucket and roll it back and forth on either a tray ramp or bucket
grid on the inside of the bucket. This removes the excess coating from
the roller and prevents excessive drip and spatter. Continue spreading
the liquid coating onto the surface in the form of a W or an M over
an area that one roller’s worth of coating will cover. After initially
spreading the coating by this method, fill in the area by rolling the
roller back and forth over the entire surface being covered. Finish
by rolling the coating in one direction. This is called laying off, and
it aids in developing a uniform finished appearance. Spraying and
backrolling is another example where the roller is used to ensure a
uniform application and finish.
Conventional Spray Equipment

Many shops apply the protective coatings with what is termed con-
ventional air spray equipment. This method uses compressed air to

118 Chapter Three
Fluid pressure Regulator and

regulator moisture separator
Regulated air
Fluid
Regulated air
Pressure container
FIGURE 3-11 Components needed to apply paint by air spray. (Courtesy of

Spray-Quip, Inc., Houston, TX.)
atomize the paint as it exits the spray gun. In other words, air is in-
jected into a stream of paint through the air nozzle in front of the gun,
creating the mist that is propelled out. The basic components needed
to apply paint by air spray are an air compressor, paint pot or cup, oil
and moisture separators, air supply hoses, material hose, regulators,
and air spray paint gun (Fig. 3-11).
The most important element in air spray painting, as in other ap-
plication methods, is the person operating the gun. Conventional air
spray equipment affords the applicator a great level of control and
results in a high-quality finish. The applicator is responsible for ap-
plying the paint correctly, using the best technique, and keeping the
equipment in good working order. Generally, coating manufacturers
list the optimum pressures for applying their coatings. They also list
the type of gun and the correct sizes of paint nozzle, air cap, and
needle to produce the best-quality applied film. Typically, external-
mix air caps are used. The space between the fluid nozzle and the air
nozzle is called the annular ring. It provides a column of air around
the fluid stream. As the fluid and air leave the air cap, they begin to
expand and mix. As this mixed stream leaves the center of the noz-
zle, it is further atomized with additional force from the holes on the
horns of the external-mix air cap. The biggest advantage of a con-
ventional air spray system is the control the applicator has over the
finish; relatively easy adjustments to the fluid pressure and air pres-
sure give the applicator tremendous flexibility and versatility. The
biggest drawback is probably low transfer efficiency; conventional
air spray equipment has a transfer efficiency of approximately 25 to
30 percent.

FIGURE 3-12 Standard setup of an airless spray system. (Source: Graco)
Airless Spray Equipment

Because of the low transfer efficiency and slow production time asso-
ciated with conventional spray equipment, airless spray equipment
was developed and introduced (Fig. 3-12). Airless sprayers work by
pressurizing paint fluid. The paint travels through a supply hose to
an airless spray gun tip, and the coating atomizes into fine droplets
as it exits the tip. In airless spray, as the name suggests, air is not used
to atomize the paint. The basic components of an airless spray system
are a power source, a pressure pump, a paint container, high-pressure
fluid hose, an airless spray gun, and an airless spray tip. The paint
is pressurized by the pump and forced through high-pressure hose
to the airless spray gun. When the stream of high-pressure coating

120 Chapter Three
FIGURE 3-13 Standard setup of an air-assisted airless spray system. (Source:

Graco)
reaches the airless tip, it is atomized and shaped by the specific size
of the tip. The only adjustment to an airless spray gun is to change
the tip. Coating manufacturers specify the correct tip size and proper
atomizing pressures needed to produce the best-quality applied film.
Although the airless spray applicator does not have the control
that a conventional spray system affords, the trade-off is that greater
speed and greater transfer efficiency value (35 to 50 percent) are pos-
sible.
Some have found a way to combine the best features of conven-
tional air spray with airless spray equipment and have created a new
spray finishing capability. The process has been termed as air-assisted
airless spray. This equipment (Fig. 3-13) uses a standard airless pump
and an airless spray tip to atomize the coating and shape it into a fan
pattern. However, in contrast to normal airless operations, the fluid
pressure in an air-assisted airless spray system is relatively low. As
expected, a low fluid pressure (usually below 1,000 psi [6,894 kPa])

fails to produce an acceptable spray pattern, and the resulting pattern

has heavy edges called tails. To eliminate the tails and assist in the
atomization process, air is added at a low volume, typically 1 to 3 cfm
(0.47 to 1.41 L/s), and at a low pressure between 10 and 20 psi (69
and 138 kPa). To this coating stream, assisting air is directed into the
airless pattern through horns on a special air cap. This results in a
good appearance, very good transfer efficiency (50 to 65 percent), re-
duced overspray, less tip wear, and longer pump life due to the lower
pressures.
Because of new VOC regulations as well as owners seeking
“solvent-free” coating solutions in order to reduce or eliminate
extractables, manufacturers have begun formulating coatings with
much higher percentages of solids. Fabrication shops and field
painters have had to comply with tougher emission laws and have
begun looking for equipment that provides higher transfer efficiency
and higher solids protective coating material application capabil-
ity. These concerns have led to the development and refining of
equipment known as electrostatic and plural-component spray equip-
ment.
The main advantages of electrostatic spraying are the savings in
materials and labor and the high transfer efficiencies of material to the
surface. In electrostatic spraying, a high-voltage electrical charge is
imparted to the atomized paint particles via an electrode on the gun,
causing the paint particles to be attracted to the substrate, which is
grounded. This virtually eliminates all overspray. Paint particles that
would normally be lost because of overspray are instead attracted
to the edges and even the back of the substrate. Transfer efficiencies
obtained with electrostatic spray painting range between 65 and 98
percent. The equipment typically used for electrostatic spraying is the
same as that used with conventional air spray equipment and with
airless equipment. However, airless and/or conventional electrostatic
systems have electric power supplies to give the paint the negative
charge needed to draw it to all sides of the substrate being painted. An
electrode at the tip of the gun adds a high-voltage electrostatic charge
to the atomized paint particles (Fig. 3-14). This technology lends it-
self to the application of both liquid-applied coatings and powder
coatings.
In plural-component spray painting, two-component (or more)
catalyzed coatings are proportioned, mixed, and applied by the spray
equipment. This method is generally for use with coatings that have
a very short pot life (from a few seconds to a few minutes) and a
very high solids content (typically 100 percent). The base resin and
converter are mixed at the spray gun, or at a mixing manifold pre-
ceding the spray gun. The two components are then immediately
sprayed onto the substrate being coated. There are two basic types of
plural-component systems: fixed ratio and variable ratio. A fixed-ratio

122 Chapter Three
FIGURE 3-14 Standard setup of an electrostatic spray system. (Source: ITW

Ransburg)
system provides only one ratio of volume for the multiple components
of the coating. A variable-ratio system can be adjusted for different
component ratios (e.g., 1:1, 2:1, 3:1). The equipment consists of two or
three airless pumps (feed pumps) attached to an air motor. The pumps
move the coating components individually from their containers in
separate lines to a proportioning pump. The materials are then nor-
mally heated and directed either to the spray gun tip or to a mixing
manifold assembly fitted with one or more in-line static mixers. From
the manifold, the mixed material travels through a whip hose to the
spray gun.
Heaters play an important role with plural-component systems.
They are used to reduce the coating viscosity, improve fluid flow, and
optimize the reactivity of the materials. Heaters are often installed
in-line and are placed on the material containers. The material hoses
are often heat traced and insulated, as well help maintain the desired
temperature. Plural-component systems also use a solvent-fed purge
pump that connects the container of solvent with the back of the mix-
ing manifold. When an applicator shuts down the equipment, the
valve for the purge pump is opened, and a solvent flush is delivered
to flush out any material that could set up in the mixing manifold
(Fig. 3-15).
The mixing manifold, when required, is critical for properly blend-
ing and mixing the materials before they leave the spray gun. The
manifold usually contains a static in-line mixer that works by split-
ting the coating stream and rotating it to 90 degrees. This is done
numerous times so that the components are mixed thoroughly when
they exit the spray gun.

Mixer/manifold
Spray gun
Solvent supply Base supply Proportioning pump Catalyst supply

and pump and pump and pump
FIGURE 3-15 Standard setup of a plural-component spray system. (Source:

JPCL, October 1989)
Inspection of Linings
Independent Inspector’s Duties
Coating contracts usually involve a significant investment of both time
and money. The owner has a well-written job specification completed
and conducts some type of bid process to select the coating contractor.
The inspector provides the owner with written assurance that the
project has met the specifications and that the coating system will
perform for its intended full life. The coating inspector is also viewed
as providing additional assurance that the risk of catastrophic failure
is significantly decreased or altogether eliminated. The inspector most
often becomes the eyes through which the owner observes the finished
work and determines whether the contract has been fully satisfied.
Many tests can be performed after the coating has been applied.
It is often difficult to find deficiencies in the coating system, however.
Once the job is finished, a poor job may look the same as a high-quality
job. Therefore, it is important that inspections occur not only at the
conclusion of a project, but also during coating operations. This will
help determine that the coating specification was met.
During Application
During the application, the inspector may need to conduct a wide
array of key tests and observations:

124 Chapter Three
r Ensure that the proper environmental conditions exist. With

the emergence of new coating technologies, the temperatures
and relative humidity constraints change from product to
product, and it is always best to refer to the manufacturer’s
data pages. A good rule of thumb is to ensure that the steel
surface temperature is at least 5◦ F (3◦ C) above the dew point
and rising.
r Ensure that the surface is free of dust, dirt, or any abrasive
residue from the blast cleaning operation.
r Ensure that coatings have been properly thinned ahead of
time. This includes being certain that the recommended thin-
ner has been used in the recommended amount.
r Ensure that the wet film is of the proper thickness. This can
be determined by conducting mathematical calculations of
the surface area coated and/or by periodically taking direct
readings while the coating is being applied.
r Ensure that the coating is applied evenly, that the passes are
overlapped, and that there are no thin spots, discontinuities,
dry spray, and so on.
The more often problems during application are addressed imme-

diately, the greater the likelihood that runs, sags, and discontinuities
can easily be brushed out and corrected.
After Installation
Once the coating installation is complete, the inspector should check
that the proper curing and drying conditions are being maintained.
The inspector should also make certain that there has been no con-
densation on the surface or that any type of contamination has
been deposited on the coating during the curing process. Over-
spray, pinholes, runs, or any other imperfections not uncovered at
the time of application should be marked now and repaired before
another coat is applied. In many cases, other tests may be required
once the coating application is complete, including the following
points:
r Discontinuity (holiday) testing: low- or high-voltage type

r Dry film thickness measurements: Type I, Type II, or Tooke
Gage method
r Adhesion testing: tape test or tensile adhesion tests
r Degree of cure using durometers or the solvent-resistance
method

Inspection Tools
Individuals responsible for quality control should be familiar with
basic inspection tools including those listed here. This is not to be
construed as an exhaustive list.
WFT gauge: The wet film thickness gauge is used to measure the
thickness of paint being applied at the point of application.
Two common gauges used are the interchemical gauge and a
notch-type gauge.
DFT gauge: The dry film thickness gauge is used to measure the
thickness of paint after it has been applied and, preferably,
cured.
Type 1A magnetic DFT gauge: Commonly called a banana gauge,
this is a single-point lift-off gauge (Fig. 3-16). It measures non-
magnetic coatings over a magnetic surface. It operates by mag-
netic contact and resistance of the magnetic force to the surface
by the coating thickness. Calibration assurance in the field is
strongly recommended.
Type 1B magnetic DFT gauge: Commonly called a pencil pull-off
gauge, this is a single-point lift-off gauge. It measures nonmag-
netic coatings over a magnetic surface. It operates by magnetic
contact and resistance of the magnetic force to the surface by the
coating thickness. Calibration assurance in the field is strongly
recommended.
Type 2 electromagnetic DFT gauge: This gauge measures noncon-
ductive coatings over a ferrous metal surface. It operates by
electromagnetic contact and resistance of the electromagnetic
force to the surface by the coating thickness. Calibration assur-
ance in the field is strongly recommended.
Eddy current gauge: The eddy current gauge (Fig. 3-17) measures
nonconductive coatings over a nonferrous surface. It operates
by emitting an eddy current and measuring the difference be-
tween the emitted signal and the return signal. This differ-
ence in time is affected by the coating thickness. Calibration
FIGURE 3-16 Type

1A magnetic DFT
gauge (banana
gauge). (Source:
KTA-Tator)

126 Chapter Three
FIGURE 3-17 Type 2

electromagnetic
DFT gauge.
(Source: KTA-Tator)
assurance in the field according to the gauge manufacturer’s

instructions is strongly recommended.
Probing knife: This knife is useful in determining adhesion of coat-
ings. It can be used in conjunction with adhesive tape.
Magnifiers/magnified light scopes: These are useful for closer ex-
amination of substrates and evaluation of pits, contamination,
pinholes, etc.
Mirrors and telescoping mirrors: Mirrors are helpful when the appli-
cator needs to check behind hard-to-reach areas such as nuts
and bolts—wherever direct viewing is impossible.
Surface comparators: These are effective in evaluating various
surface profiles obtained using various abrasive materials.
Replica tape and spring micrometers: Replica tape and micrometers
are often used to determine surface profiles.
Sling psychrometer and U.S. Department of Commerce Weather
Bureau Psychrometric Tables: These are used in conjunction to
determine relative humidity values and dew point tempera-
tures.
Surface temperature gauges and infrared noncontact temperature
gauges: These gauges aid in determining when the surface is
approaching the dew point and when surface temperature is

excessive. Either situation may be problematic and can lead to

blistering.
NACE/SP0178 Design, Fabrication, and Surface Finish Practices for
Tanks and Vessels to Be Lined for Immersion Service (also known as
the weld replica standard): This is quite effective in determining if
the welds have been prepared properly and are in a condition
to receive a protective coating or lining.
Holiday detectors: Both wet-sponge low-voltage and high-voltage
DC units are effective in examining a lining to ascertain the
number of discontinuities (holidays) within it. The low-voltage
unit is effective if the lining is less than 20 mil (508 m) DFT.
The delimiters on the high-voltage equipment are the coating’s
thickness and its dielectric strengths; voltages can range from
500 to 200,000 V. Neither unit is recommended for coatings
that contain conductive pigments (e.g., aluminum, zinc, and
graphite).
Adhesion testers: These reveal possible problems with adhesion by
defining a numerical adhesion value in pounds per square inch
or kilopascals and revealing where the break has occurred.
Tooke Gage: Commonly called a paint inspection gauge, the Tooke
Gage is known as “the referee,” so named for its ability to ex-
amine individual coatings within a multicoat system as well
as the system in its entirety. It is highly accurate in determin-
ing film thicknesses up to coating films of 50 mil (1,270 m).
Testing with the Tooke Gage is destructive, so repairs will be
required.
Soluble-salt testing (chloride and ferrous): Wide arrays of tests can be
performed to obtain information determining the presence of
invisible contaminants that will be detrimental to the lining.
Two common extraction methods are the swabbing method
and the Bresle patch method:
r In the swabbing method, the salts are extracted by using
distilled water and medicinal-grade wool or cotton swabs. A
defined surface area is then swabbed with the cotton swabs
that have been saturated with the distilled water solution
and dried with additional dry cotton swabs. The wet and
dry swabs are then placed back into the beaker containing
the distilled water and stirred for several minutes.
r In the Bresle patch method, the salts are extracted by us-
ing distilled water in conjunction with a plastic patch self-
adhesive cell and syringe. Distilled water is injected into the
cell, allowed to dwell for 20 seconds, and then drawn back
into the syringe. The same solution is then reintroduced into
the cell, and the process is repeated. This process is con-
ducted a total of three times.

128 Chapter Three
It is generally accepted by the industry that neither extraction

method removes all of the soluble salts, but only up to a maximum of
approximately 45 to 50 percent. Once you have obtained your solution,
there are several methods for measuring the amount of soluble salts
obtained, including the following:
r Potassium ferricyanide test
r Conductivity test
r Kitagawa tube test
r Quantab strip test
r QuantoFix test
Bibliography
Bauer, M. 1988. Changing Regulations on Coatings for Contact with Potable
Water. Journal of Protective Coatings and Linings (JPCL) 5(12):27–33.
Bauer, M. 1996. Organic Zinc Rich Primer for the Interior and Exterior of
Potable Water Tanks. In Proc. SSPC Expanding Coatings Knowledge World-
wide, Charlotte, N.C. (Nov.). Pittsburgh, PA.: SSPC.
Crist, M. T. 1996. The Cost and Timing of Water Tank Maintenance. In Proc.
SSPC Technologies for a Diverse Industry, Charlotte, N.C. (Nov.). Pittsburgh,
PA.: SSPC.
Dromgool, M. B. 1996. Maximizing the Life of Tank Linings. Journal of Protective
Coatings and Linings (JPCL) 13(3):62–74.
Dubcak, T. O. 1995. Inspecting Water Tank Linings: The Importance of the First
Anniversary. Journal of Protective Coatings and Linings (JPCL) 12(9):60–8.
Finch, D. 1996. Protecting Water Storage Tanks in an Era of Environmental
Compliance. WATER/Engineering & Management, Nov.
Huffman, L. R. 1997. Going with the Flow: A Sampling of Water Tank Mainte-
nance Painting Programs. Journal of Protective Coatings and Linings (JPCL)
14(5):38–46.
Ippoliti, T. 2000. Waterborne Coatings for Water and Wastewater Treatment
Plants. PWC (Nov./Dec.):92–7.
Ippoliti, T. 2002. Polyurea Coatings Win Place in Water, Wastewater Facilities.
WaterWorld (Nov.):16–8.
Kapsanis, K. 1990. A Water Tank Update: Issues and Practices in Removing
Lead-Based Paint. Journal of Protective Coatings and Linings (JPCL) 7(5):
50–6.
Knoy, E. 1992. When to Repair Pits in Steel Water Tanks. Welding Design and
Fabrication (Oct.):51–2.
———1993. Maintaining Aged Steel Water Tanks: What to Look for and Why.
Journal of Protective Coatings and Linings (JPCL) 10(5):61–5.
Maronek, A. H. 1988. Evaluating Acceptability of Potable Water Tank Coatings.
Munger, C. G., and D. V. Louis. Corrosion Protection by Protective Coatings. 2nd
ed. Houston, TX: NACE International.

O’Donoghue, M., R. Garrett, and V. J. Datta. 1998. Optimizing Performance of

Fast-Cure Epoxies for Pipe and Tank Linings: Chemistry, Selection, and
Application. Journal of Protective Coatings and Linings (JPCL) 15(3):36–51.
O’Toole, D. 1997. Overcoating: An Effective and Economical Solution for Wa-
ter Storage Tank Exteriors. In Proc. 1997 SSPC Technologies for a Diverse
Industry, San Diego, Calif.; Pittsburgh, PA.: SSPC.
Preparation of Protective Coating Specifications for Atmospheric Service.
2000. Joint Technology Report SSPC-TR 4/NACE 80200. In Proc. Planning
and Specifying Industrial Protective Coating Projects (2004). Pittsburgh, PA.:
SSPC.
Public Works. January 1995. Specifications for Coating Water Storage Tanks are
Evolving.
Roetter, S. P. 1993. Liability Enormous for Lead-Based Paint Removal. Opflow
(March).
Schubert, R. 1999. Construction and Operation of Water Storage Tanks in
Rural Alaska. In Proc. 1999 AWWA Engineering and Construction Confer-
ence, Orlando, Fla. Denver, Colo: AWWA.
Shannon, G. B. 2003. Selecting Coatings for an Elevated Water Tank in a
Densely Populated Business Park. Journal of Protective Coatings and Lin-
ings (JPCL) 20(7):85–7.
Smith, L. M., ed. Generic Coating Types: An Introduction to Industrial Mainte-
nance Coating Materials. SSPC 95–08. Pittsburgh, PA.: Technology Publish-
ing Company.
SSPC Painting Manual Volume 1: Good Painting Practice. 4th ed. 2004. Pittsburgh,
PA.: Society for Protective Coatings.
SSPC Painting Manual Volume 2: Systems and Specifications. 8th ed. 2000. Pitts-
burgh, PA.: Society for Protective Coatings.
Stein, G. R. 1994. Community Acceptance of Lead Paint Removal Projects.
Zienty, D. 2002. Tanks Pull Double Duty. WATER/Engineering & Management
149(2):9–13.
Zienty, D., and L. Dornbusch. 2002. Painting for Antenna Installations on Water
Storage Facilities. Journal of Protective Coatings and Linings (JPCL) 19(9):
73–9.


CHAPTER 4
Contractual
Considerations
William J. Dixon, P.E.
Dixon Engineering, Inc.
This chapter includes general discussions of the most important topics

to be investigated and detailed in preparing the specifications for a
new steel water-storage tank. Some topics are common to steel water
tank repair and/or repaint. Repair and repaint comments are made
following the new-tank comment in some cases. This chapter describes
the content and purpose of each of the documents involved in bidding,
and it notes the duties of owner, engineer, and constructor with respect
to specifications, bidding, design, construction, and inspection.
Competitive Bidding
The water industry in the United States has been serving municipal
clients for more than 200 years. In that time, the industry has de-
veloped competitive bidding practices, which are required by law in
most states and by governmental subdivisions. Many variations to
the standard construction project (designed and bid by the engineer)
and some old practices such as maintenance contracts have resur-
faced with new twists. Different methods of contract administration
have been developed, the most common of which is usually called
design/build. Newer practices include computer bidding and what
is called a reverse or negative auction. These alternative methods have
different benefits depending on the project and on whether you are a
public owner or a private owner.
It is necessary to fully understand the benefits of the selected
method. The closed competitive bid process was developed to elim-
inate fraud and political influence in the awarding of contracts and
131
Contractual Considerations
132 Chapter Four
generally led to the project being awarded to the lowest responsi-

ble and responsive bidder. Responsible refers to the company’s ability
to successfully complete the project within the required time frame,
and responsive refers to the thoroughness of the bid submittal—that
is, whether the bidder properly completed the bid, included all bid
documents (e.g., noncollusion of contractor with other contractors;
minority and/or women and/or disadvantaged participation forms),
and submitted the required bonding.
Negative Bidding/Reverse Auction

In the negative or reverse auction, the bidder submits the bid on an
open auction Web site. The bid is open for a set time, generally a couple
of days. Other bidders then submit lower bids. The first bidder is free
to resubmit his or her bid, undercutting successive bidders. If bidding
is active, the owner can extend the bid time in his or her interest. In the-
ory, this procedure may work well for industry or private firms—but
a low bidder, in a common industry joke, is the bidder who made the
biggest mistake. It is in everyone’s interest that the constructor make
a fair profit. When a constructor, after submitting what he calculates
is his best competitive price, is squeezed to cut his bid further, he en-
ters the project in makeup mode: During the project, he will either be
looking to cut corners or looking for extras to increase the scope of the
project. Negative bidding, therefore, requires a very thorough set of
contract documents and full-time third-party inspection.
The closed or sealed public bid process was designed to eliminate
the awarding of projects under political influence, and, in general, it
has worked. Remember, the issue is more than a fair and open public
bid process; it is also the appearance of a fair and open public bid
process. With the ability to extend the closing time, it is possible to
close bids after the preferred bidder has submitted his bid. His price
is the lowest, but if a significant change order develops, at least the
appearance of favoritism is there.
Roles of the Owner, Constructor, and Engineer in

Standard Municipal Contracting
There are usually two contracts—the construction project contract and
the contract between the engineer and the owner. Figure 4-1a best
exemplifies the standard roles of the owner, constructor, and engineer
in the competitive bid process.
In the construction project contract, the constructor and the owner
have vertical privity of contract that is extended downward to sub-
constructors, suppliers, and subcontractors’ suppliers (Fig. 4-1b). It
is important to insulate the owner from subcontractors and sup-
pliers through contract protective clauses (safety, indemnification,
insurances, etc.). In the second contract, that between the engineer

Contractual Considerations 133

Horizontal
privity
Owner Engineer
Vertical
privity
(a) Constructor
Owner Engineer
Constructor
Supplier
Subcontractor
Subcontractor supplier
(b) Tier-2 subcontractor
Horizontal
privity
Owner Engineer
Vertical
privity
Lawsuits
Constructor
(c) Lawsuits
FIGURE 4-1 Roles of owner, constructor, and engineer in the competitive bid
process. (a) Privity means a direct relation throughout the contract. (b) In this
situation, there is still no privity between the constructor and the engineer.
There is vertical privity from the second-tier subcontractor all the way up to
the owner. Contract clauses and performance and payment bonds insulate
the owners as much as possible from lawsuits by subcontractors and
suppliers. (c) Because the engineer and the constructor are not third-party
beneficiaries of each other’s contract, neither can sue the other. The owner is
under privity both ways, so if he or she is sued by the constructor, the owner
can bring the engineer in by filing a claim against the engineer.
and the owner, the engineer is hired to prepare specifications and

complete project management and inspection services. This is known
as horizontal privity of contract between the engineer and the owner.
On a repaint or rehabilitation, the engineer is hired to also do a prelim-
inary bid inspection to establish the scope of the project. The owner
does not intend the constructor to be a third-party beneficiary of this
contract. The owner is the sole beneficiary of the services performed.

134 Chapter Four
That is why in Fig. 4-1b there is no connection to the constructor in

the horizontal privity line.
The lack of contractual privity and the owner’s intention that no
third-party beneficiaries are to be part of the owner/engineer contract
are represented in Fig. 4-1c by the jagged lines between the engineer
and the constructor. The arrows indicate that the engineer does not
control the constructor and that the constructor cannot rely on (and,
in most cases, cannot sue) the engineer. The arrows further indicate
that the engineer cannot interfere with the constructor’s “ways and
means” or “means and methods.” If the engineer, who might be more
experienced, starts giving directions or advice to the constructor, he
or she has crossed the line. The engineer assumes the risk for this
advice, which has broken down the wall between vertical and hori-
zontal privity. The engineer’s job is to review, report, and interpret the
contract documents. Sometimes the owner’s staff members provide
their own specifications and inspection services. In this situation, a
definition is needed so that the owner’s inspections do not substitute
for the constructor’s own quality control responsibilities.
Design/Build
The design/build procedure puts the responsibility of engineering
design on the constructor. This procedure works well in new-tank
projects but has conflicts with rehabilitation projects. The benefit of
design/build lies primarily in expediting the project. On new-tank
or new-tower projects, the owner supplies the bidder with soil
investigations, establishes capacity and high water level, and designs
standards and a time schedule to follow. The bidder can properly
prepare costs and submit the bid. If the procedure is followed prop-
erly, there should be no unknowns and no cost increases. Third-party
inspection is still necessary.
On tank repair/repaints, the extent of repairs and the condition of
interior coatings and corrosion may not be known to the tank owner
because of the complexities of removing the tank from service for in-
spection. It is not recommended that the party completing the work
be allowed to establish the scope of the work. New construction with
no unknowns and larger budgets may be competitively bid as de-
sign/build. Design/build is not practical on lower-budget rehabili-
tation projects, because the unknowns cannot be competitively bid.
A pre-bid independent tank inspection can eliminate the unknowns
on repair/repaints. But to ensure competent work and competitive
bidding, a full set of specifications is still necessary, because the con-
structor’s idea of what constitutes a proper repair and coating system
would be different than what the owner expects. Third-party inspec-
tion would be necessary for quality assurance, but without specifica-
tions and a contract, what would the inspector inspect?

Project Administration
For a further discussion of all of the major methods of project ad-
ministration, see American Water Works Association Manual M47,
Construction Contract Administration.
Pre-Bid Inspections—Existing Tanks

Sealed competitive bids are only as accurate as the bidding docu-
ments that all bidders review. For rehabilitation projects, the bidding
documents are only as good as the pre-bid inspection report they are
based on. Pre-bid inspections are discussed in Chap. 10. The pre-bid
inspection report is intended to provide the owner with a thorough
understanding of existing conditions and repaint options. The pre-bid
inspection report is not intended for use as the bid document. If the
owner permits the pre-bid inspection to be performed by a constructor
instead of an inspection firm with no vested interest, that inspection
should be the total scope of the constructor’s contract. The constructor
should never be awarded an emergency repair or repaint change order
based on his or her inspection. With the exception of vent screens and
open cathodic protection holes, most repairs can wait to be completed
through the competitive bidding process.
The practice of some inspection/paint firms to sell emergency re-
pairs has led some state regulatory agencies to develop or expand
permit requirements. Most states require permits for all work on a
tank’s interior; some states also require permits for exterior work to
verify compliance with air quality laws. Permits are issued on the ba-
sis of specifications submitted to the regulatory agency and on the
condition that work is to be completed according to those submitted
specifications. New or relocated tanks require location and elevation
review and may require a permit from the Federal Aviation Adminis-
tration (FAA). If, in the tank owner’s opinion, the emergency repairs
detailed by the constructor’s pre-bid inspection are truly emergencies,
contact the regulatory agencies to determine whether an emergency
permit is needed and whether proposed repairs meet state standards
or codes.
The pre-bid inspection report can be included in the bidding docu-
ments. It is important that it be labeled an “appendix for information
only” so that its language does not conflict with the specifications.
The appendix report should state that it is a service only and that field
verification by site visit is required.
Contract Documents
The terms contract document and bidding document are often incor-
rectly thought to be interchangeable. Bidding documents used to so-
licit project bids traditionally include all the information needed to

136 Chapter Four
prepare and submit a bid, including the advertisement, information

for bidders, general conditions, supplementary general conditions,
addendums, technical specifications, drawings, and proposal pages.
Contract documents include all the bidding documents as well as
signed contract or agreement forms, bonds, insurance certificates, no-
tice of award, notice to proceed, and submittals. As the project pro-
gresses, change orders and field orders are added. Most contracts
permit the inclusion of minutes from preconstruction and progress
meetings.
Design Standards
Incorporation of Standards
Standards prepared by national associations have been incorporated
into technical specifications for decades. Standards are beneficial be-
cause the bidder is using the same terminology as used by the owner/
engineer and for the most part knows what is expected. The Amer-
ican Water Works Association (AWWA) standards incorporate other
nationally recognized standards—American Welding Society (AWS),
Society for Protective Coatings (SSPC), National Association of Cor-
rosion Engineers (NACE), American Petroleum Institute (API), and
concrete standards, to name a few. The benefits are obvious, but there
are pitfalls.
Using standards requires the engineer’s understanding of the in-
cluded standards. He or she must decide whether to include the entire
standard or just portions of it. An example would be the use of AWWA
D100 Standard for Welded Carbon-Steel Tanks for Water Storage.
D-100 defines how many X-rays are to be taken. If you want more
X-rays or a different selection process for X-ray locations, the spec-
ifications must detail the variance. The potential for a conflict may
occur more in the incorporation of other standards within the spec-
ified standards (i.e., when a second tier of standards is incorporated
by reference within the specified standard).
Most standards are submitted to the American National Standards
Institute (ANSI) for certification. One condition for ANSI certification
is that the standard be reviewed and formally updated at least every
5 years. Because of the constant updating and the long bid process,
specifications must identify which standard is being incorporated, ei-
ther by date or by clause in the section that outlines general conditions.
This also applies to standards incorporated within the standard spec-
ified. Industry standards recognize the standard in use when bids are
opened. If a standard is updated during a nonbid situation, a negoti-
ated contract, the negotiating parties should define which standard is
to be used.

Policy Changes of AWWA Standards

After review by AWWA attorneys and officers and by the AWWA Stan-
dards Council, AWWA has made two major changes to the standards.
First, they are to be minimal standards that establish the minimal ac-
ceptable level of performance. Second, all contractual language is to
be removed from all AWWA standards. These significantly change
both the content of the standards and the effect of incorporating them
into the specifications.
Minimal Acceptable Level of Performance

AWWA D100, AWWA D103 Standard for Factory-Coated Bolted-Steel
Tanks for Water Storage, and the proposed standard for composite el-
evated tanks assign specific responsibilities to the owner (or to the
owner’s agent), the engineer, and the constructor for the construc-
tion of steel tanks. AWWA D102 Standard for Coating Steel Water-
Storage Tanks is the painting standard used for both new and ex-
isting tanks. AWWA D104 Standard for Automatically Controlled,
Impressed-Current Cathodic Protection for the Interior of Steel Water
Tanks is the cathodic protection design standard for new and existing
tanks. The design requirements for water, snow, wind, and live loads
in D100, D103, and future composite-tank standards are the minimum
allowable values. The design methodology has changed, permitting
thinner steel.
D100 and the proposed composite-tank standard have changed
the approach to design of allowable loads for the buckling of conical
and dished sections. The owner/engineer must choose one of three
methods for the design. All three are safe, but the owner/engineer
must select the level of conservatism preferred. When specifying ac-
cording to these minimal standards, construction tolerances are more
critical. These standards do not carry a default design for this partic-
ular choice if the specifier fails to select the design methodology. The
result may be a less conservative design than the owner would prefer.
The owner and the engineer are expected to be aware of all local codes
and ordinances and should provide the required loading information,
design information, or both to the constructor if local requirements are
more stringent than those found in the applicable AWWA standard.
Note: Contractual language often shifts the responsibility for knowing
local codes to the constructor.
This change to a minimal standard is not in itself a bad practice.
The new tank still will be designed per the International Building Code
and will perform as needed. The trade-off in connection with the new
minimal standard is the higher degree of attention to maintenance,
which is still the responsibility of the owner. The same level and cost
of maintenance exist, regardless of the change in the standard: The
difference is that there is no time cushion and so no opportunity to
delay that maintenance.

138 Chapter Four
The old tanks were behemoths and used more steel than the tanks
designed today. Water towers were once designed by engineers us-
ing slide rules, and steel thicknesses were rounded up to the next
one-eighth of an inch (0.125 in. [0.3 cm]). To this, the early engineers
added another 0.125 in. (0.3 cm) for corrosion allowance. When weld-
ing replaced riveting, welds had to have only 66 percent penetration.
Whereas earlier, steel was one size going down the stem until the
next fraction was needed, now every stem section could be differ-
ent, because current steel manufacturing allows steel with differences
in size, measured by 10 mil (254 m), to be purchased. Computer-
assisted design, cutting, and rolling permit the new thinner de-
sign methodologies for conical and dished sections. Welding require-
ments have increased to 100 percent penetration welds. Corrosion al-
lowances are still the option of the design engineer, but they are seldom
specified.
The excess steel—not maintenance or the coating systems—is the
reason some of those older towers are still standing. The interior coat-
ing of old towers was two coats of red lead primer protected by a
wax (grease) coating. Lead was good from a durability standpoint but
obviously now is out of favor because of health effects. The grease
coating had a very short life, particularly in cold climates. Pit weld-
ing was standard in maintenance projects, whether it was needed or
not. The modern epoxy, urethanes, and polyureas, as well as cathodic
protection, offer a far superior and cleaner method of protection. Main-
tenance painting and cathodic protection are more critical now that
new designs have no corrosion allowance. There is a safety factor in
design calculation, but that is not the same thing as a cushion factor
for steel loss.
Removal of Contractual Language in the D100 Standard

The second change was the removal of contractual language from
AWWA standards. It used to be that the specifier would incorporate
the D100 standard, which would then speak for itself (“the constructor
shall design . . . , shall fabricate . . . , shall erect . . . , shall test welds,”
and so on). It is still necessary to incorporate the various sections
of the standard or the whole standard, but the contractual language
must now be in the specifications. Failure to use contractual language
could make complying with the incorporated standard optional for
the constructor.
Contractual language has also been removed from the D102 stan-
dard for painting projects. D102, however, always required the se-
lection of an interior or exterior coating system. The new standard
does require input by the specifier (questions found in the appendix)
defining some responsibilities.

Assignment of Responsibilities of the Engineer, Owner,

and Constructor
In standards, a foreword poses a series of questions that need to be
answered by the owner and the engineer. Their answers will define the
responsibilities of the owner, engineer, and constructor on the project.
The constructor needs this information to prepare the bid.
Rebuilt Tanks—Going Where No Standards Have Gone Before

The other pitfall of design by standard is specifying the building of,
or a repair to, a product to which no standard applies. Examples are
specifying structural repairs according to D102 or rebuilding a tank
according to D100. D102 concerns painting only; D100 covers the de-
sign loads of and construction methods for new tanks, not the design
of rebuilt tanks. Rebuilt tanks were built earlier according, it is hoped,
to some standard in existence at that time. It would be the exception,
not the rule, if a rebuilt tank met current D100 requirements for new
tanks. The D100 committee had struggled with the concept of whether
to establish a subcommittee to develop a standard for rebuilt tanks,
but the complexities involved and potential liabilities may prevent
any effort to establish a standard for rebuilt tanks.
If you live in seismic zone 0, for example, it does not matter where
the rebuilt tank came from. If you live in seismic zone 2, you must
know what zone the tank was originally built in, what the code re-
quired when the tank was bid, and preferably the original design
calculations and as-builts (original construction drawing revised to
reflect any construction changes). Without this information, you have
to do reverse engineering—measuring every section of steel, account-
ing for corrosion losses, and then designing according to current stan-
dards and codes to see whether the old tank meets them.
Remember, older tanks may have been built under a standard that
required only 66 percent weld penetration. Old riveted tanks should
never be rebuilt. A rebuilt tank should be built from only one tank,
not from a collection of parts from more than one tank. If you choose
to allow a rebuilt tank, your specifications should be very thorough.
Specify only the pertinent sections of D100 and D102.
Exercise caution in evaluating a used tank that is bid as an alterna-
tive to a new tank. The initial cost should not be the only consideration;
total-life-cycle maintenance costs should also be projected. An owner
considering a used tank should require the following:
r A copy of an up-to-date inspection report of the structure
provided by a qualified, registered professional engineer.
r A signed document from the present owner stating that the
tank is available for sale.

140 Chapter Four
r A detailed proposal of any remedial coating or repair work

that will be done to bring the structure to the equivalent state
of the proposed new tank.
It is also important to know the specifications and standards under

which the tank was originally built, as well as the wind and seismic
loadings for which it was designed and constructed. An additional
caution is the possibility that lead or other hazardous materials are in
the coating of the tank being considered for reuse.
Tanks that were designed and used for fuel oil or a liquid other
than nonpotable water may not have the desired wall thickness for
water storage, having been designed to contain a liquid with a dif-
ferent specific gravity. Additional problems in converting a tank from
hydrocarbon storage to potable water storage have to do with how
clean the steel that contacts the water can be made, and how that
affects adhesion and the purity of interior coatings.
New-Tank Designs
New-tank designs are another example of industry outpacing the
speed at which standards are developed. For example, the compos-
ite tank (concrete pedestal/steel tank) was built without an AWWA
standard for more than 15 years before a standard was developed.
Other smaller associations, including the Steel Plate Fabricators, were
able to develop standards more quickly, permitting the composite-
tank industry to grow until the more comprehensive AWWA stan-
dard could be developed. A hybrid—a glass-lined bolted-steel tank
on a pedestal—was developed and applied for inclusion in AWWA’s
proposed composite-tank standard, but it failed to make the standard
because of timing.
The caution is not to avoid new products but to understand the
standards before you specify something that is not included in them.
Factors in Competitive Bidding

Construction Time Frame
The time frame during which you expect a constructor to build an
elevated tank is inversely proportional to the amount of money you
hope to spend. A tower can be designed, fabricated, and erected in
6 months, but it would be expensive. A year and a half—540 days—is
a reasonable time frame. (In the northern half of the country, because
of the shorter painting season, the tank can be constructed over the
winter, but painting will be delayed for a few months; allowing 540
days enables painting to take place during warm weather.)

The construction process involves design by the constructor and

submittal of design to the engineer for review. Sixty days may elapse
before the engineer turns it around. After the design drawings are
accepted, steel is ordered, cut, rolled, and fabricated in the shop. Af-
ter shop drawings are approved and before foundation work is even
scheduled, the constructor prioritizes the project (based on comple-
tion schedule and the schedule of other projects). The foundation is
constructed and allowed to cure for 28 days. Steel is usually deliv-
ered during this cure time. The start of concrete work is dependent
on timely review of submittals. Actual time to erect steel in the field
ranges from 6 to 20 weeks, depending on weather, site conditions, and
tank size. The painters need 30 to 60 days, again, depending on tank
size, containment requirements, and, particularly, weather. Including
a break for winter, it is not hard to see why 540 days and scheduling
flexibility are needed.
On new-tank projects, the existing tank or operating system is
not removed until the new tank is operational. On repair and repaint
projects, time restrictions are critical as there may be no backup system.
Again, the shorter the project time, the more the project may cost, but
conditions may justify the cost. Weather, summer demands on the
water supply, extra time for containment projects, and even the start
of the owner’s budget year may delay the project start. Money can
be saved by specifying a maximum out-of-service time and the latest
completion date. This way, the constructor knows that it is a 60-day
job and can schedule on the basis of the crew’s availability. The days
can be scheduled anytime, as long as work starts 2 months before the
specified latest completion date.
Pre-Bid Meeting
A pre-bid meeting is beneficial and useful for discussing specific non-
technical portions of the project, the timing of subcontracting require-
ments, and forms in the bid documents (e.g., noncollusion, minority,
affidavits, subcontractor lists, lien waivers, etc.). This meeting can be
mandatory or optional. If it is mandatory, interest in the project can
be gauged by how many constructors are present, which could affect
the contract price. As a minimum, job-site visits should be required.
On repaint projects, if the tank is empty, it should be made available
to the painter for inspection. If the tank is elevated, prospective bid-
ders should provide proof of insurance before they climb it. To limit
time infringements on the owner, limit the days the site is available
for inspection.
Prequalification of Bidders
To shorten the time between bidding and awarding of the contract,
prequalify the constructors. Tank constructors specialize in the design,

142 Chapter Four
fabrication, and erection of welded-steel, bolted-steel, or composite

concrete/steel tanks. Tank painters are also very specialized. Con-
structors and painters rely on extensive safety training, specialized
equipment (some that they have manufactured themselves), and ways
and means procedures developed during years of experience. Tank
constructors and tank painters should be prequalified on the basis
of their experience, ability to supply bonds, financial condition, and
safety and environmental record. For future discussion, see Chap. 4
of AWWA Manual M47, Construction Contract Administration. Cau-
tion: Competitive bidding laws in some states restrict prequalifi-
cation.
Constructor Assistance
Generally, tank constructors and manufacturers are more than willing
to discuss individual tank needs and to assist the owner and the en-
gineer by providing standard design information for a project. For an
elevated tank of a given capacity, each manufacturer has different ge-
ometric parameters. These dimensions normally do not vary enough
to cause difficulty if one constructor is selected for a project on the
basis of the information supplied by another constructor.
To become familiar with current industry standards and practices,
the owner is advised to contact prospective bidders and discuss a
project before issuing an invitation to bid. Most manufacturers are
willing to provide copies of preliminary specifications developed for
tanks of varying styles and capacities. The owner must be careful to
make every effort to write a specification that is open and does not
exclude bidding by any qualified manufacturer or supplier. In par-
ticular, a given manufacturer’s proprietary design details should not
be included in a project’s contract document; this would create an
inequitable bidding situation for other qualified suppliers or manu-
facturers.
Forced Use of Subcontractors

Some communities have requirements to hire a set percentage of mi-
nority or disadvantaged subcontractors. Some communities call for
hiring of local employees or subcontracting of local firms in the bid
package. Although these are laudable goals, care must be taken for
tank constructing and painting on how and by what part of the con-
tract these provisions are enforced. Concrete foundations are a good
example. Many firms are capable of forming and placing the concrete,
but the foundation is just that—a foundation that supports millions
of gallons or liters of water. The AWWA standard requires the con-
structor’s engineer to design the foundation. For liability reasons, the
constructor should design and be responsible for the foundation and
all structural items.

Painting projects consist primarily of abrasive blast cleaning,

painting, and containment construction, some metal repairs, but not
a lot of subcontract work. Also, remember that these constructors and
painters have had extensive safety training and that they are respon-
sible for every worker, including subcontractors and forced hires, on
that tank.
Elements of tank construction, for example, electrical controls, site
restoration, and waste disposal, can be separated and bid locally. In
fact, it is good practice for the local firm or employees responsible
for future maintenance and controls to be familiar with their instal-
lation. Telemetry must interface with the owner’s master system. If
the mechanical system requires piping and several valves inside the
tank, this work can be bid to local mechanical constructors. On paint
projects, it may be possible to tie another project into the paint work
to accommodate the subcontracting.
Bid Security
Bonds are required on most publicly funded projects over a minimum
dollar amount. This dollar amount under the Miller Act is $50,000;
many states have a “Little Miller Act” that may have lower limits.
Some municipalities have lower limits yet. The Miller Act requires
performance and payment bonds. Bid bonds are generally required
by state or local statute.
The bidding process is time-consuming and involves significant
expense for both owner and constructor. In a tank-painting contract in
the northern states, a bidder failing to honor his or her bid could delay
the project into the next season. In that case, the second constructor
would still have to honor his or her bid but, thinking the job had been
lost, may have taken another project. Painting constructors usually bid
and complete all of their contracts within the same weather-restricted
season. They traditionally fill their season and do not leave openings
for jobs for which they came in second. Fairness may require adjusting
the project schedule into the next year if the first bidder cannot meet
the agreement.
The bid bond, intended to cover the increased cost of rebidding
or awarding to the higher bidder, has been traditionally set at 5 per-
cent of the bid. This amount should be sufficient on a $500,000 new-
tank contract ($25,000), but it may not be sufficient for a $50,000 paint
project ($2,500). If the constructor defaults and the bond is collected,
the constructor could lose the ability to purchase bonds. In the coat-
ing industry, with its limited number of qualified constructors, limited
seasons, and wide range of bids, a painter who has received a more
lucrative contract may buy his or her way out of the smaller job. For
this reason and others, many engineers require a bid bond higher than
5 percent or for a set amount.

144 Chapter Four
General Conditions and Supplementary

General Conditions
Common parlance at the beginning of a bid document, referred to
as boilerplate, includes information for bidders concerning general
conditions and supplementary general conditions, or supplementals.
This is standard text that remains virtually unchanged for all projects.
Rewriting the general conditions and the supplementals is not neces-
sary on all projects, but they should be reviewed each time. AWWA
M47 recommends modifying the general conditions or supplementary
general conditions to be more job specific.
The general conditions, if prepared by the engineer, are gener-
ally standard documents prepared by the Engineers Joint Contract
Documents Committee (EJCDC). This is a coalition of engineering as-
sociations with endorsements by the Associated General Constructors
of America and the Construction Specifications Institute. Even more
important is another EJCDC document, “Guide to the Preparation
of Supplementary Conditions.” This guide follows the general condi-
tions, explaining what they really say and listing some of the standard
exceptions. Most bidders do not review the general conditions every
time, but they should read them at least once. The general conditions
are generic to construction work; supplementary general conditions
that are tank specific should be clauses describing the following local
requirements:
r Bonding
r Warranties, guarantees
r Maintenance contracts
r Insurance
r Indemnification
r Prevailing wage and documentation
r Use of local or union labor
r Payment application, change order procedures
r Steel or concrete cost escalation
r Dispute resolution
r Safety
r Meetings—preconstruction, progress, final punch list
r Severability
r Schedule of values
r Termination

Bonding
Bid bonds, discussed in the previous section, are usually free. The
constructor pays roughly 1.5 to 2.5 percent depending on the work-
load and contract default experience for the actual performance and
payment bonds.
A combined performance and payment bond for 100 percent of
the project bid meets the requirements of the Miller Act, but it may
not be sufficient to cover a project gone sour. The performance bond
is used to ensure that all aspects of the project are completed. If the
original constructor is unable to meet the terms of the contract, the
bonding company brings in another constructor to finish the project,
at least to the extent of the bond. A payment bond is used to pay
all legitimate subcontractors and suppliers if the constructor fails to
make payments.
Problems generally start with the lowest bid process. If the project
is awarded to the lowest bidder, who turns out not to be responsible
(is incapable), a 100 percent combined performance payment bond is
inadequate. Remember, the bond is for an amount that all of the other
bidders thought was too low. Out of that amount, you must pay off all
unpaid subcontractors and suppliers and bring in another constructor
to finish.
Separate performance and payment bonds, each at 100 percent of
contract amount, are the current requirements of the EJCDC general
conditions. That and control of partial payments should be sufficient
to fund completion of the project.
Bonding companies have a contract with the constructor, who usu-
ally must personally guarantee the bond. The bond names the owner
as the intended beneficiary. Consider the bond a product supplied as
a condition precedent to award of the contract to the constructor. A
word of caution: There is no contract between the surety and the owner,
but consider the relationship an obligation. The owner is required to
control payments to the constructor and to receive waivers from all
subcontractors and suppliers; this is called a waiver of surety. If the
owner pays the constructor too much and there is insufficient money
left to complete the project, the surety has an affirmative defense to
avoid payment.
Another caution: Verify that the surety company can be served pro-
cess to force payment. Offshore bonding companies have attractive
bonding rates. Offshore companies are offshore for several reasons;
they enjoy tax advantages and are untouchable for claim enforcement.
The surety should be licensed in the state of the project.
Another aspect of the payment bond is the notice requirements
of the Miller Act and the Little Miller Act. The prime contractor’s
subcontractors and suppliers do not require notice of hire. Their con-
tract with the prime contractor, or prime, is evidence of notice. If the

146 Chapter Four
second-tier subcontractors or suppliers and subsequent tiers fail to

give notice to the prime at the start and again within so many days
of the last workday, the surety does not have to pay the claims un-
der the Miller Acts. The Miller Acts and surety also limit coverage:
Third-tier subcontractors and suppliers to second-tier subcontractors
are not protected. Although they have to draw a line somewhere, this
is unfortunate, because those companies are usually the locals. Al-
though local companies are not protected by surety, the owner can
still hold payment until they are paid if the owner has received notice
of nonpayment.
The fourth type of bond is a maintenance bond. Maintenance
bonds are usually free, but a separate maintenance bond is reassur-
ance that any work due under the warranty will be completed. Be-
cause a warranty is a contractual requirement, it technically is already
covered by the performance bond. However, a separate maintenance
bond with the time frame defined is recommended for all warranties
exceeding the standard 1-year/13-month warranty. A maintenance
bond should also be considered for the full term of a multiyear main-
tenance contract (extended maintenance bonds would not be free).
Performance bonds or maintenance bonds expire unless the owner
gives notice to the bonding company that work is needed.
Warranty
The standard construction warranty period is 1 year. The painting war-
ranty is also 1 year, but the D102 standard allows a 13-month period
in which to complete the paint warranty. The time extension recog-
nizes the difficulty some communities have in isolating their tank.
Also, weather may interfere with draining the tank within the speci-
fied time. There is also a trend toward specifying longer warranties, a
practice that theoretically raises project costs. Constructors prefer to
wrap up a project in 1 year. Their bonds are then released; they have
less unknown potential liability and can bid other jobs.
A multiyear warranty on tank construction has little benefit un-
less full use of the tank must be delayed. Most problems with welded
steel are evident within a year. Extended warranties are specified
more often on painting of new tanks or repainting contracts. Con-
tractual problems of long maintenance periods can be resolved by
the use of a maintenance bond. The problem is that the constructor
is giving a warranty on a product that deteriorates with age, weath-
ering, ablation, ultraviolet (UV) degradation, and so on. There is no
standard against which to hold a 2- or 5-year paint project warranty
condition.
Unless specified differently, D102 limits holiday testing (direct-
current voltage testing for coating pinholes) to the high waterline and
down on the wet interior surface. If there are coating breaks in that

area after 1 year, repairs under the warranty are justified. Small coating
breaks, rust staining at the lap seams of the wet interior roof or pinhole
rusting on the dry interior surface or on the exterior, are not warranty
issues. If the specifications did not require a holiday-free coating, a
warranty cannot require a holiday-free system at 1 year. At 2 years
or 5 years, enforcement can only be to whatever expected coating
condition at warranty was set in the original specifications.
Also, a 5-year warranty has to consider normal wear and tear. If
you are looking for annual maintenance, bid the maintenance contract
into the construction bid as a separate cost; do not try to complete the
work under a warranty clause. Trying to have work completed in
5 years for no pay is tricky—it is difficult enough just remembering
that there is a 5-year warranty. The second concern is whether the
constructor is still in business. The owner needs to pay only when
scheduled work is completed.
Maintenance Contracts
Maintenance contracts are for a set period of time. Generally, the ac-
tual painting takes place during the first year, and touch-up and re-
pairs happen in subsequent years. Some maintenance contracts are
attractive because they begin with an enticing finance offer, in which
a company finances the initial high cost of the first painting over the
first couple of years. As always, some good constructors and some bad
ones offer these services. To differentiate, follow the money—or, bet-
ter yet, control the money. Work including maintenance procedures
should be controlled by specifications prepared by your engineer. The
work should be inspected annually by a third-party inspection firm.
The financial portion should be written or at least reviewed by your
attorney. Your attorney should also offer an opinion if competitive bid-
ding of the entire maintenance project is required by local ordinance
or state statute. Some of the cost advantage is in the constructor pro-
viding financial and engineering services, but the owner must decide
whether the savings are in the owner’s interest or are in the vested
interest of the maintenance constructor.
When the dust settles, the good constructors will be there; the
ones with prices too good to be true will be the missing parties. Main-
tenance contracts are not new. In the 1960s and early 1970s, annual
maintenance contracts were a major portion of the painting market.
Most contracts were for 10 or 12 years and had the same annual pay-
ments. Painting on the interior would take place during years 5 and
11, and the exterior would be painted during years 6 and 12. Essen-
tially, the major work was paid in advance. A painter had enough
other contracts at varying stages and so could finance the work when
it was due. When the gas crunch came in 1973 and gas and paint
prices skyrocketed, most constructors merely walked away with the

148 Chapter Four
up-front money or went bankrupt, and the owners were left with paid
but uncompleted projects.
Insurance Requirements
Risk managers or insurance consultants for the owner should establish
limits and types of coverages. Sometimes, consultants are nervous
because the project requires elevated work or may involve exposure
to lead paint. Most tank constructors carry large limits of liability,
and some are self-insured. But smaller paint constructors have lower
limits. The limits and types of insurance should match the owner’s
exposure to risk. Excess coverage sometimes cannot be purchased by
smaller constructors. Purchasing special insurance for one job drives
project costs up, particularly if some constructors are eliminated from
bidding.
Workers’ Compensation insurance levels are standard and protect
the constructor’s workers. General liability insurance protects site vis-
itors other than employees. Auto insurance is recommended even if
all work is on one enclosed site, because the constructor still will run
errands. An umbrella policy would be in addition to these policies.
The owner and the engineer require the constructor to name
them as additional insureds. Any claim against the additional insured
other than gross negligence is covered under the constructor’s policy.
Over the years, attorneys have expanded the gray area between the
constructor’s and the owner’s insurance liability. Some insurance
counselors now require the constructor to furnish a separate owner’s
protective policy.
Professional liability insurance covers errors and omissions and
is associated with the engineer. If the constructor acts as an engineer
in a design/build contract, this insurance may still be needed. Some
states have strict liability laws for accidents involving gravity, falls, or
injuries from dropped objects. Under contract terms, this liability can
be covered under an owner’s protective policy paid by the constructor.
But with strict liability laws, responsibility is automatic and liability
cannot be avoided. The owner should consider his or her own policy
and consider making the constructor’s policy a primary-pay policy.
The EJCDC documents require the owner to supply a builder’s
risk policy. As noted, the EJCDC documents were prepared by profes-
sional engineers’ associations and endorsed by a constructors’ associ-
ation; the owner’s municipal associations were not involved. Owners
prefer that the constructor provide a builder’s risk policy. This policy
provides insurance for the project during construction (e.g., if a tor-
nado blows over an unfinished tank). The policy covers the cost to
replace the tank.
Some owners require submittal of the constructor’s entire policy,
but most owners prefer just a certificate. The certificate warrants that

the constructor will not cancel the policy without giving a specified
number of days’ advance notice. The expiration date, additional in-
sureds, deductibles, and all policy exclusions should be checked.
Indemnification
Indemnification clauses are often points of contention and are some-
times contract killers. The owner’s attorneys want contractual lan-
guage that protects the owner from all claims—suits that might origi-
nate as a result of this project regardless of who is liable. Constructors
have attorneys who are as smart as the owner’s attorneys, and the in-
surance companies, especially, have their share of attorneys. Although
a constructor may agree to any clause to get a contract, it does not mean
his or her insurance policy covers the indemnification. A constructor
who has no insurance policy does little to offer true indemnification.
Rather than proceed under the false assumption of coverage, both
the owner’s and constructor’s policies should carry comparative li-
ability coverage. Each party pays on the basis of each one’s share of
liability.
Both bonds and insurance are conditions precedent to contract
award. We are a litigious society and, as with bonds, insurance is now
a requirement before the owner even signs a contract.
Prevailing Wage, Local/Union Labor, and Local Restrictions

If prevailing wages are required, the required pay scales should be in-
cluded in the supplementary general conditions. Constructors should
be aware that unlike the standards, which are fixed at the date of
bid opening, wage rates can change within the project time frame.
Supplying certified pay records or other pay documentation is an
overhead cost that the constructor should be made aware of through
the supplementary general conditions section when preparing his or
her bid.
The use of local labor or union labor should be detailed, as well
as local time or noise ordinances. Can the constructor work late or on
weekends? Time restrictions to painters are critical because of weather
restrictions.
Payment, Change Order

The general conditions and supplementary conditions detail how to
file the application for payments, how the applications are reviewed
and by whom, denial of partial payment, reasons for denial, deter-
mination of any denied amount, and the corrective actions needed to
recoup lost payments. The change order process and other ways of
changing the contract are included here.

150 Chapter Four
Steel/Concrete Escalation Clause

Most painting projects are completed within 1 year. Project costs are for
fuel, abrasive material, paint, equipment, and labor. Abrasive material
and paint are (or can be) purchased and stockpiled soon after project
award. Equipment is a known cost, and labor cost increases are also
foreseeable, as prevailing labor rates have expiration dates. On new-
tank construction, concrete costs for composite tanks and steel costs
for steel and composite tanks are a major portion of expenses. Gas or
diesel for their equipment is also a major expense. Their costs cannot
be fixed immediately upon award. Steel is not bought until after the
design is approved. Concrete for composite tanks is not purchased
until it is time to start construction. Gas or diesel is purchased on-site.
We have previously discussed the length of time between bid
award and steel purchase. What we did not discuss was the delay
in award. Working from the basic premises that a constructor who
runs a project efficiently is entitled to a fair profit and that material
costs fluctuate, constructors are either covered by a safe cost cush-
ion (overpricing the project) or construction contracts should have a
steel escalator clause and possibly a concrete escalator clause. These
costs are beyond the constructor’s control and can rise rapidly, as was
demonstrated in 2006, when increased demand drove the price of steel
up 30 percent almost overnight; it continues to drive the price upward.
An escalator price should require the constructor to specify quantity
with the bid and to tie that quantity to a certain appropriate index.
Calculation of the index should be by an independent third party. The
owner either pays for the inflated bid price or pays the exact increase,
if any.
Dispute Resolution
The best defense for disputes is to detail in the specifications a clear
procedure for dispute resolution. The two primary alternatives are
the courts and arbitration. Owners, being deep pockets without a face
(usually a municipality), prefer the courts. In fact, a future trend by
municipalities will be to contractually require the constructor to waive
the right to a jury trial. The owner wants a decision based strictly
on the law, with no human element, whereas the constructor prefers
arbitration by certified arbitrators.
The three alternative methods of dispute resolution are negotia-
tion, mediation, and arbitration. There are many variations and hybrid
methods of resolution (minitrials, for example). Negotiation—direct
talking between the contract parties—is the first step in all contract
disputes. Mediation, generally the second step in the process, brings
in a third party. The EJCDC’s general conditions make the engineer
the mediator. He or she both hears arguments and tries to get the
parties to resolve the problem. Because the argument is usually about

specifications that the engineer wrote, he or she, as mediator, is given

the final interpretation of the specifications. The engineer’s decision,
while final, is usually not binding. If either party is still dissatisfied,
it may proceed to court or to arbitration, whichever method is named
in the contract documents.
Arbitration can be binding or nonbinding. Recent legislation per-
mits contracting away the right of appeals to the court system. The
courts are attempting to unload their dockets, and to appeal a binding-
arbitration award to a court has very little chance, barring the allega-
tion of fraud, capriciousness, or a significant procedural error.
Safety
Engineering associations have always maintained that the engineer
is not responsible for job-site safety. If the engineer and the owner
assume such responsibility, they are potentially liable for extensive
damages. If the engineer (and ultimately the owner) was responsible
and thus liable, the employee can sue through Workers’ Compen-
sation and collect a much larger settlement. So far, most cases have
dismissed the engineer because he or she had horizontal privity with
the owner and no contractual relation with the constructor. The ex-
ceptions were when the on-site inspector or project manager took on
the constructor’s role by directing some of the work or giving advice
on ways and means.
No one questions the need for safety, but constructors are contrac-
tually responsible for safety. Constructors have developed their own
construction procedures and ways and means. Constructors have de-
veloped safety programs and trained their personnel accordingly. The
engineer’s personnel should follow the constructor’s safety program
in addition to the engineer’s safety program.
Meetings
The number and types of meetings required, and who must attend,
should be listed as accurately as possible. If home office personnel
are required to attend, it is a cost issue. Most repaint contracts have
a preconstruction meeting and a final punch list meeting with the
job superintendent. Interim progress meetings are usually attended
by the on-site superintendent but not by the project manager unless
there are problems. On new-tank projects with larger budgets, it is
more common to require the project manager to attend.
Severability and Termination

Every contract should have a severability clause. This clause says that
if a court finds one or more clauses to be illegal, they can be severed
from the contract. The rest of the clauses and the contract still remain.

152 Chapter Four
Contract termination clauses describe the procedure for giving

notice to, and terminating the contract by, the owner for convenience
or cause and by the constructors for cause.
Schedule of Values
New-tank projects are usually bid as lump-sum contracts. Extras, such
as complicated logos, are bid as deductible alternates. On a project that
is spread over a year or 18 months, partial payments are in order. To
avoid paying too much money up-front or not enough, a schedule of
values is submitted early. This schedule can be negotiated until it is
acceptable to all parties. The engineer has the final say. Following is a
sample schedule:
Design phase 10 percent

Foundation 10 percent
Fabrication 25 percent
Erection 30 percent
Painting 15 percent
Electrical 5 percent
Site work and punch list 5 percent
Repaint and repair projects can be bid with line items for repair
items and unit prices for painting different areas—wet interior, dry
interior, and exterior. If the project is bid lump sum, a schedule of
values should be included on a bid form that assigns costs to wet
or dry interior and exterior. Payment is then figured on percentage
of work completed (e.g., wet interior abrasive blast cleaning 40 per-
cent, prime coat 20 percent, intermediate coat 20 percent, and topcoat
20 percent).
Technical Specifications—New Tanks

Technical specifications, while ultimately the heart of the entire
project, are only briefly discussed in this chapter. Technical specifica-
tions for almost all new-tank projects are performance based. Discus-
sions about siting, type of tank to select, foundations, appurtenances,
and other topics appear in other chapters.
Tank Water Testing and Disinfection

Tank disinfection procedures are usually performed in accordance
with AWWA C652 Standard for Disinfection of Water-Storage Fa-
cilities or the more stringent requirements of local health agencies.

Specifications should define responsibilities for this work. AWWA

D100 requires that the tank be tested before coating and that the owner
furnish the water to fill the tank and provide a means of disposal fol-
lowing testing. D100 permits the coating to be applied before water
testing if the tank constructor and the owner have specifically agreed
to that. Tank testing and disinfection phases of the project could then
be concurrent, thus saving the owner the cost of producing and dis-
posing of a large quantity of water. However, water testing the welded
tank before applying the coating has the following advantages:
r It allows identification and correction of distortions caused

by anticipated or unanticipated foundation settlements before
coating.
r It allows the identification and correction of any leaks that
might have been temporarily covered by the interior coating
system.
Valves and piping should be tested in accordance with AWWA

C600 Standard for Installation of Ductile-Iron Water Mains and Their
Appurtenances.
The constructor may be required by contract to provide and dis-
pose of the testing and disinfection water. However, the owner will
ultimately bear these costs, plus the constructor’s overhead and profit.
Federal Aviation Administration

The owner or the owner’s engineer should file notification of construc-
tion with the FAA before construction of tall standpipes or elevated
tanks. The FAA will determine whether the site is acceptable and, if
so, the requirements for temporary and permanent tank markings and
lighting.


CHAPTER 5
Foundations
Sayed Stoman, Ph.D., P.E., S.E., M.L.S.E., and Kevin Gallagher, P.E.
Caldwell Tanks, Inc.
The proper design and installation of the foundation is essential to all

welded-steel tanks for water storage. Before any water tank is erected,
it is vital to ascertain the suitability of the site and the feasibility of the
project from engineering, construction, cost, and safety perspectives.
The site soils must be capable of supporting all loads imparted on or
generated by the water tank system with appropriate factors of safety
against bearing, uplift, and lateral stability and with reasonable total
and differential settlements. (The factor of safety [FS] is a function of
the foundation type, the source of loading, and whether the loads are
long-term or transient.) Most importantly, conditions at the site must
constitute a safe working environment for the construction crew.
The ideal sites for erecting water tanks are relatively large and
level, dry, and easily accessible. The ideal bearing soils are sandy soils
that range in relative density from medium to dense, to very dense,
and clayey soils with consistency ranging from stiff to very stiff, to
hard. These characteristics are well suited to shallow foundations,
with a minimal amount of settlement and full-foundation stability. The
ideal site should be large enough to accommodate the construction
equipment, provide ample lay-down area, and be free of obstructions,
especially high-voltage electric power lines.
Appropriate Foundation Type

The foundation type is governed by the soil characteristics in the ef-
fective zone of influence of the bearing soils, the loading, the size of
the property, topography, site location, and the presence or absence
of structures and facilities within the site premises. Future expansion
plans and modifications are also factors in the selection of the appro-
priate foundation type.
155
Foundations
156 Chapter Five
The foundation system should provide for adequate factors of

safety for both strength and stability. Since the water-storage system
is very sensitive to settlement, care must be exercised to limit the total
and differential settlements so that they do not adversely affect the
tank and piping system. As a practical rule, it is preferred to limit the
total and differential settlements to a maximum of 2 in. (5 cm) and
1 in. (2.5 cm), respectively. If high settlements are expected, special
piping or piping joints may be necessary to allow more flexibility in
the system. While settlement is a major design consideration, each
system is unique and must be examined on its own merits.
Generally, for economic reasons, shallow foundations are pre-
ferred. As long as the bearing soils have reasonable capacity, provide
for tolerable settlements, and, when excavated, are properly suited for
backfill around the footings, shallow foundations are possible. A net
allowable bearing capacity of about 3,000 psf (144 kPa) or an ultimate
bearing capacity of 9,000 psf (430 kPa) or better would be most ap-
propriate for shallow foundations. Shallow foundations can be built
where the allowable bearing capacity is lower, but the resulting foot-
ings would be much larger and the settlements possibly higher.
Deep foundations are more appropriate where the bearing soils
are composed of loose sands or soft clays or where tests have detected
the presence of sinkhole cavities, sandy layers that are prone to liq-
uefaction, or silty soils with high moisture content that are likely to
consolidate under loading. The type of deep foundation is influenced
by cost, availability of piles, and local practice.
Regardless of the foundation type used, grade beams intercon-
necting the individual footings may be necessary where the horizontal
shear resulting from either wind loads or seismic loads is too large for
a single footing to resist. These compressive elements are designed as
beam-columns on elastic foundations.
Location/Orientation
Aside from accessibility, site location is crucial for several reasons.
Specific site location and accurate determination of property lines are
vital, as disputes resulting from even minor infringements onto ad-
jacent property can cause major delays in construction and possibly
even cancellation of the project. Precise staking of the foundation foot-
print early in the project can eliminate orientation concerns and facili-
tate establishing the proper grade elevations and boring locations for
geotechnical investigations. Ownership of the site property is essen-
tial, as building a storage tank on land owned by others can be costly.
Orientation of the tank on the site must also be considered with
respect to the piping layout, existing utilities, and other obstacles. To
facilitate the connection to the inlet piping as well as to accommodate

Foundations
Foundations 157
the flow of overflow water away from the footing area, situating the
tank in the proper orientation, is essential. Tank and piping orientation
should be clearly shown on the site plan with reference to reliable
existing benchmarks.
Sites in low-lying areas and floodplains, areas with sinkholes, and
areas with underground shafts, tunnels, or fault lines are not recom-
mended, nor are areas containing substantial fills. Similarly, coastal
regions are not recommended—they are generally subject to signifi-
cantly higher wind loading and may require design for tidal waves.
The loose sandy soils in these areas could require additional stabiliza-
tion measures.
Establishing Existing and Final Grade Elevations

Because system hydraulics depend on the attained potential energy,
it is very important to accurately establish the high water line and
low water line so that proper pressure and flow can be maintained
at all water levels. When the water delivery system has several tanks
along the line, each with precise overflow and low water line eleva-
tions calibrated, it is extremely important that all elevations, including
grade elevations, be accurately established for the system to function
as intended.
Grade elevations must be established with respect to verified
benchmarks, and a clear distinction should be made between existing
and final grade elevations. Grading across the tank footprint is also
critical. The final grade should provide for drainage away from the
tank foundations, but the slope should not be so steep that excessive
footing exposure is required. The footing exposure above grade level
must be properly considered and accounted for, especially when the
site work involves major cuts and fills. These considerations will cir-
cumvent expensive consequences, which can and must be avoided
with proper planning and attention.
Minimum Depth and Projection Above Grade

The proper minimum depth at which a foundation should be placed
depends on several factors. The slope and drainage of the site, the
location of the frost line, the magnitude of the uplift and lateral loads,
the potential for erosion, the presence and location of surface water
or groundwater, and the type of soil all affect the depth at which a
footing is set to bear. Similarly, these factors also influence how far the
footing should project above the final grade level.
Generally, geotechnical reports recommend a minimum depth on
the basis of regional experience and familiarity with local conditions.
However, it is preferred to bear the individual spread footings below

Foundations
158 Chapter Five
the frost line at a minimum depth of about 4.5 to 5 ft (1.4 to 1.5 m).
A minimum projection of 6 in. (15 cm) above the final grade should
be sufficient to protect the column base plates from ground moisture.
On sites where standing water or settlement is a concern, higher pro-
jections would be appropriate to further protect the base plate and
anchor bolts from corrosion.
Excavation Requirements
Before any excavation begins, it is essential that all aspects of job-
site safety are well understood. All excavations, particularly confined
excavations, must be performed in strict compliance with the latest
federal Occupational Safety and Health Administration (OSHA) stan-
dards.
It is recommended that excavated slopes, including those for shal-
low foundations, be laid back at a maximum of 2H:1V (horizontal to
vertical) slope. Permanent slopes of 3:1 may be used on fill slopes that
have been placed on suitable subgrade. Grass should be planted or
other measures taken for erosion control.
Vertical cuts for shallow foundations are not recommended unless
all the requirements for job-site safety and foundation stability can
be assured. Such cuts are not possible in dry, sandy soils, but they
can be made to a critical height in an undrained soil where the pore
pressures are negative. If vertical cuts are used, however, one must
ensure that clear and achievable compaction requirements for the soil
wedge alongside shallow foundations are well defined.
Site Access and Drainage

Water tanks are often erected in remote rural areas where site access
can be difficult. Even in urban areas, with everyone competing for
prime space, site conditions can be challenging and access to the job
site arduous. As noted previously, the job site requires access by large,
heavy construction vehicles and lay-down and work areas for material
and construction equipment. Thus, a suitable access road is vital to
the entire construction effort.
Preparing the site for work often requires tree removal, clearing,
and grubbing. Where rain and mud can prohibit site access, building
some basic roadwork is necessary before all other activity can begin.
Geotechnical investigation and reconnaissance activities also require
full access to the site. Where there is standing or seeping water, wa-
ter removal measures will be required. As the presence of water can
severely complicate construction efforts, site drainage must always be
addressed for all phases of construction, including postconstruction,
to facilitate proper maintenance.

Foundations
Foundations 159
Water Table and Perched Water

The presence of groundwater is always a major construction concern.
A high groundwater table is also a significant design consideration
due to the buoyancy effects, increased settlement and consolidation,
and potential liquefaction associated with ground seismic activities.
Likewise, the presence of perched water can be a challenge during
construction and will require creative dewatering measures.
Where water is present, various dewatering techniques are avail-
able to drain the excavations and to dry the soils before pouring the
foundations. Pumping water out from established well points is a
common method of dewatering. In situations where the bearing soils
are or could be under water, a layer of crushed stone can be placed
on the bearing soils to allow for seepage and flow of water to the well
points for pumping during construction. Alternatively, where possi-
ble, a mud mat 4 to 6 in. (10 to 152 cm) thick of low-strength (2,000 psi
[14 MPa]) concrete can be placed on the bearing soils to contain the
water below or to protect the bearing soils from rain.
As buoyancy reduces the effective unit weight of soil, it may be
necessary to place the footing deeper to maintain lateral and uplift
stability in shallow foundations. Similarly, in pile foundations, pile
uplift and lateral capacity may be affected by the presence of water.
During a seismic event, clean, submerged sands may liquefy, causing
down-drag on piles and loss of lateral support in the liquefied zone.
A high water table also creates design challenges when drilled piers
or auger-cast piles are used.
Soils and Geotechnical Investigations

Once the suitability of the site is established on the basis of a visual
reconnaissance, a subsurface investigation must be performed to de-
termine the geotechnical composition and engineering properties of
the bearing soils and to identify suitable foundation types based on the
prevailing conditions. Although there is some uniformity in geotech-
nical investigations, substantial differences exist in the scope of work
performed and in the extent of data included in the final reports. To
ensure that all relevant information is included in a well-documented
report, a clearly defined scope of work for the investigation is essential.
The number, location, and depth of borings should be speci-
fied before commencing work. Geotechnical professionals who are
thoroughly familiar with their regions can offer specific guidance in
defining the scope on the basis of their field experience and past ex-
plorations. Generally, the shallow-foundation option is pursued for
reasons of economy and ease of construction. When unsuitable soils
are encountered and/or the required bearing depths become exces-
sive, other alternatives should be considered. When there are feasible

Foundations
160 Chapter Five
alternatives, the geotechnical report must include all the engineering

data necessary for each design option.
Exploration, Sampling, and Testing

Explorations for water tank foundations should include a minimum of
three borings that preferably extend to depths of 30 to 50 ft (9 to 15 m).
When details about the average soil properties in the upper 100 ft
(30 m) of the site profile are required by code for site class determina-
tion, one boring may be drilled down to 100 ft (30 m) or to the bedrock,
if it is encountered sooner. To ensure that sound rock has been reached,
it is recommended that rock cores of 5 to 10 ft (1.5 to 3 m) or deeper
be obtained. Additional borings may be required if the subsurface
conditions so dictate. Sinkhole cavities, liquefiable soils, rock lenses,
and boulders are examples of conditions that may warrant further
investigation.
In all cases, boring locations must be precisely staked out and doc-
umented with reliable measurements to well-established benchmarks
or other dependable references. The presence of adjacent structures
and facilities affecting, or affected by, the project must be noted and
realistically evaluated. The borings should be spread out over the foot-
print of the water tank. For leg tanks, it is preferable to perform two
borings on the column circle 180 degrees apart and a third boring at
the center of the tank. Generally, the deepest boring is located at the
center of the tank.
The boring logs should carefully reflect the geological profile en-
countered in the borings as well as the groundwater observations.
Standard-penetration blow counts resulting from split-spoon sam-
pling must also be reflected on the boring logs. Whether the borings
were advanced by dry augering or by rotary-wash drilling should be
noted. Soils encountered during drilling should be examined and clas-
sified in the field. All factual, inferred, and interpretive information
included in the final report must be unambiguously noted.
It is important to recognize that geotechnical investigations per-
formed on the basis of a few borings provide only a limited view of
the overall conditions at the site. Given this large margin of uncer-
tainty, the designer should anticipate and be prepared to accommo-
date changes as they are encountered. Often, to lessen uncertainty,
field testing is performed to verify the proposed soil properties, the
allowable bearing, and other geotechnical concerns.
As borings are performed, soil samples are generally taken at
about 2-ft (0.6-m) intervals in the top layers and at 5-ft (1.5-m) inter-
vals subsequently. These intervals may be shortened where noticeable
variations are observed. Collected samples of disturbed and undis-
turbed soil must be properly handled, correctly labeled, and care-
fully transferred to the soil laboratory for testing. As the boreholes

Foundations
Foundations 161
Specified Test ASTM Designation

Classification of soils D2487
Cone penetration D3441
Consolidation
One dimensional D2435
Undrained triaxial D4767
Direct shear D3080
Modified proctor (compaction) D1557
One-dimensional swell D4546
Seismic refraction method D5777
Standard penetration D1586
Standard proctor (compaction) D698
Triaxial compression D2850
Unconfined compression D2166
Note: ASTM = American Society for Testing and Materials.
TABLE 5-1 ASTM Test Designations
are occasionally left open overnight to permit measurement of

groundwater level the next day, preventive measures should be taken
to ensure safety during the night. In most states, regulations require
that the boreholes be filled with grout or plugged on completion of
the investigation.
The standard penetration test and cone penetration test are the
common tests used to determine properties of soils. The subjective
seismic refraction method is also used in subsurface explorations.
To obtain an estimate of consolidation settlement in saturated clayey
soils, consolidation tests are run in the laboratory. The unconfined
compression test, direct shear test, and triaxial compression test are
alternative test procedures for determining the shear strength of soils.
The unconfined compression test is primarily suited for cohesive soils.
The magnitude of potential swell in clays is determined by the unre-
strained swell test or the swelling pressure test. Table 5-1 describes
some of the American Society for Testing and Materials (ASTM) stan-
dards that can be referred to for further details regarding geotechnical
test procedures.
Similarly, test procedures are available for checking the design
capacity of piles. The axial compression load test (ASTM D1143), the
pullout load test (ASTM D3689), the lateral load test (ASTM D3966),
and the dynamic pile load test (ASTM D4945) are among some of the
common methods used.

Foundations
162 Chapter Five
Engineering Properties
Soils that are well suited for foundation design can be cohesive (pri-
marily clayey) or cohesionless. Cohesion, the characteristic that en-
ables the soils to bind or stick together, gives added shear strength to
the soil. Shear strength defines the suitability of the soil for the type
of foundation work being considered. Cohesionless soils are predom-
inantly sandy soils having particles that lack cohesion. Cohesionless
soils draw their shear strength from sliding friction and interlocking
of grains.
Some soils combine cohesive and cohesionless characteristics and
are classified according to which is dominant. Silty soils have been
classified as cohesionless by some authors and cohesive by others,
depending on the soil’s clay content. Silty soils, however, are not con-
sidered good foundation material due to their compressible nature
when wet. Similarly, topsoil and organic soils are unsuited for foun-
dations. Sandy and clayey soils with high bearing capacities and low
plasticity are best suited for foundations of water-storage tanks.
Engineering properties of soils are defined by the soil grain size
distribution. For cohesionless soils, this distribution is determined by
sieve analysis, which involves sifting the soil through sieves having
openings of different sizes arranged in descending order from coarse
to fine. The amount of soil retained in each sieve after sieve agita-
tion serves as the basis of measurements and plots used in defining
grain-size distribution. Particle-size distribution in fine-grained cohe-
sive soils is determined by hydrometer analysis based on sedimenta-
tion of soil particles in water over a given length of time.
Another measure of cohesive soil consistency is called the
Atterberg limits. The Atterberg limits refer to the moisture content at
which a given volume of a cohesive soil changes consistency from one
state to another. These states are defined as solid state, semisolid state,
plastic state, and liquid state. The moisture content at which the soil
transitions from the solid state to each successive state is referred to
as, respectively, the shrinkage limit (SL), the plastic limit (PL), and the
liquid limit (LL). The difference between the LL and the PL is defined
as the plasticity index (PI). The PI is also a measure of the expansive
potential of the soil. Soils with high PI values (PI > 35) have severe
shrink/swell characteristics and require additional consideration in
design, as is discussed later in the chapter.
Soil Consolidation
All structures are subject to foundation settlement. Given the loading,
footing size, and properties of soils, these settlements can be evaluated
with reasonable accuracy. Consolidation is time-dependent settlement
that can be significant in saturated soils. It occurs when the soil under-
goes compressive deformation under the loading from the structure.

Foundations
Foundations 163
Water is extruded from the voids in the soil as the soil rearranges
its grains to accommodate the increased pressure. The process occurs
rapidly in granular soils because of their permeability, whereas in co-
hesive soils or fine-grained silty soils, consolidation can take a long
time.
Consolidation is also a function of footing size. The larger the foot-
ing, the greater the depth of the soil affected by the loads on the footing.
The affected depth is also referred to as the effective bearing depth of
the foundation. Depending on the thickness of the compressive layer
or layers and the depth(s) at which they occur, the resulting settle-
ments can be substantial and could create serious consequences for
the structure. Therefore, to ensure structural stability, it is important to
include consolidation testing in the work scope where deformation-
prone soil layers are present within the effective bearing depth of the
foundation.
Total and Differential Settlement

All water tank foundations undergo settlement. For reasons of stabil-
ity and serviceability, it is necessary to minimize these settlements to
tolerable limits. The extent or severity of the settlement depends on
the foundation type, the magnitude and direction of loading, and the
properties of the bearing and supporting soils. In addition to the over-
all settlement, a foundation undergoes relative or differential settle-
ment, which, in essence, is the settlement of one part of the foundation
with respect to another. Although all settlements must be evaluated
for their effects on the system, differential settlements must be ex-
amined more closely, as they are critical to foundation strength and
overall system stability.
As noted in the section “Appropriate Foundation Type,” at the
beginning of the chapter, for water tanks, it is preferable to limit the
total and differential settlements to a maximum of 2 in. and 1 in.
(5 cm and 2.5 cm), respectively. If the use of shallow foundations
will cause excessive settlement, deep foundations can be used to
further limit these settlements. Special piping and fittings are avail-
able that can offer flexibility in the system when high settlements are
expected.
Settlement is a major design consideration, and its effects on the
entire water tank system must be carefully evaluated. In addition to
the direct vertical settlements, the foundations are also subject to hori-
zontal displacements under the influence of lateral loading. The extent
of the horizontal displacement depends on the amount of foundation
movement that is necessary to activate the surrounding soils’ passive
resistance against the lateral loads. Therefore, geotechnical reports
must fully address all settlement and lateral displacement considera-
tions.

Foundations
164 Chapter Five
Required Design Information

The subsurface investigation report must be prepared by qualified,
registered geotechnical engineers who are experienced in designing
water tank foundations. The number, location, and depth of the bor-
ings should be provided by the foundation designer and confirmed by
the geotechnical engineer. All data supporting the recommendations
for each feasible foundation type must be included in the geotechnical
report, which should establish the following basic requirements:
For all foundations
r Site classification
r Site topography and site preparation
r Description of the soil and its engineering properties
r Classification of soil strata
r Liquefaction potential and its consequences under dynamic
loads
r Presence of rock, rock lenses, and boulders
r Potential for and consequence of shrink/swell
r Replacement or remediation of shrink/swell soils
r Anticipated total and differential settlements
r Drainage considerations
r Elevation of groundwater and dewatering requirements
r Minimum recommended bearing depth of foundation
r Excavation and backfill requirements
r Suitability of site soils for backfill
r Compaction and compaction testing requirements
r Seismic design parameters for American Water Works Asso-
ciation (AWWA) and/or other applicable codes
For shallow foundations

r Soil ultimate and net allowable (FS = 3.0) bearing capacity
r Soil carrying capacity for lateral load based on soil passive
resistance
r Extent of overexcavation, if necessary, and backfill recom-
mendations
For pile and caisson foundations

r Anticipated pile/caisson type, size, and length
r Required pile/caisson spacing
r Pile/caisson axial load capacity—compressive and pullout
(include values for end bearing and skin friction separately,
as appropriate)

Foundations
Foundations 165
r Minimum required reinforcement for caisson (as per local
practice)
r Pile/caisson allowable lateral load capacity
r Pile/caisson bending-moment diagram for appropriate end
conditions
r Pile or caisson safety factor for long-term and transient load-
ing
r Pile testing required and type of test
r Special installation considerations
r Appropriate uplift connection recommendations
Typically, laboratory analysis of selected samples includes vi-

sual classification, cohesive shear strength tests, determination of
Atterberg limits, grain-size analysis, determination of field moisture
content, and the following additional parameters:
Soil unit weight

Unit cohesion of soil c
Coefficient of soil active pressure Ka
Coefficient of soil passive pressure Kp
Standard penetration resistance values N
Angle of internal friction
Coefficient of friction, if different than tan f
Modulus of subgrade reaction ks
All the required seismic parameters—including the mapped max-

imum considered earthquake acceleration at short period Ss and at
1-second period S1 —should be specified. Where required, site-specific
geotechnical investigation and dynamic site response analysis should
be performed to determine the appropriate values.
Problem Soils
Experience indicates that certain types of soils pose special challenges
in design and require remedial measures before they can support wa-
ter tank foundation loads. Among these are expansive soils, which
are prevalent in many areas of the United States and elsewhere in
the world. The expansive clays with very high plasticity index values
are not suited for shallow foundations unless remedial measures are
taken that include lime mixing, prewatering, use of water barriers, or
soil replacement. All of these measures are costly and require strict
quality control.

Foundations
166 Chapter Five
Foundation considerations in shrink/swell soils depend on the

depth of the active zone and the swell potential of the soil. When
a thin layer of these soils is present near the surface, it can be re-
placed with more suitable low-plasticity soils. When deeper layers
are encountered, the expansive soils under the foundation can simi-
larly be replaced and properly compacted. An approach that is often
recommended is to place the shallow footing below the depth sus-
ceptible to shrink/swell and to replace or remediate the surrounding
soils to make them suitable for backfill. However, the footing can be
placed within the active zone as long as the uplift forces caused by the
swelling of the soils are taken into account in design and as long as
the structure can tolerate the resulting movements in the foundation.
It is important to note, however, that as long as moisture is prevented
from entering the soil, the shrink/swell volumetric changes cannot oc-
cur. Therefore, pouring a concrete slab over the footing area or placing
waterproof barriers around the footing are alternate remedial options.
Further detailed options are discussed later in the chapter.
When materials such as organic soils, fills, or any other type of
loose soil are encountered at the bearing level, they should be undercut
and replaced with suitable soils. The replacement soils may be what
is commonly known as select structural fill, sand, or crushed stone.
Select structural fill consists of uniformly graded sands to silty or
slightly clayey sands, free of organics and other deleterious material,
with less than 30 percent passing through a no. 200 sieve. Select fill
is also recommended for backfill around the footings and pile caps
when unsuitable soils are present.
Structural fills are commonly placed in thin (6- to 8-in. [15- to
20-cm]) lifts and compacted to 95 to 98 percent of the soils’ modified
proctor maximum dry density (ASTM D1557) or other ASTM criteria.
They may require some manipulation of the moisture content (wetting
or drying) to achieve the required compaction. Flowable fill is another
material that can be used for this purpose. Of course, replacing the
undercut soils with low-strength concrete is always an option.
Structural Concrete
Water tank foundations are primarily constructed of reinforced struc-
tural concrete. Concrete is a mixture of hydraulic cement with fine
and coarse aggregates and water in appropriate proportions. Sand,
gravel, crushed stone, and, in some cases, iron blast-furnace slag con-
stitute the aggregates. The governing properties of hardened con-
crete are defined by the quality of the cement paste, ratio of water
to cement, and the properties of the aggregates. Structural concrete
is concrete mixed to a uniform distribution of materials on the basis
of a precise mix/design and satisfactory quality control for required

Foundations
Foundations 167
durability and compressive strength. The design and construction of

water tank concrete foundations follow the Building Code Require-
ments for Structural Concrete (ACI 318-05) and Commentary (ACI
318R-05) (see “Bibliography”).
r Building Code Requirements for Structural Concrete, ACI
318-05.
r Building Code Requirements for Structural Concrete, Com-
mentary, ACI 318R-05.
Concrete is strong in compressive strength, but its tensile strength
is limited to a small fraction of its compressive strength. Therefore, in
flexural design applications such as for foundations, steel reinforce-
ment is provided on the tension side of a member to resist the ten-
sile stresses. Similarly, as shrinkage and temperature reinforcement
normal to flexural reinforcement, or wherever tensile stresses can de-
velop, steel reinforcement is added to provide tensile resistance. Re-
inforcement bar sizes and details as well as criteria for determining
the amount of reinforcement needed are all outlined in ACI 318-05.
Materials
Structural concrete materials include cement, aggregates, water, and
admixtures. The reinforcing steel used in water tank foundations com-
prises deformed bars ranging in diameter from 3/8 in. (9.5 mm) to
about 13/8 in. (3.5 cm). Two larger bar diameters of 13/4 in. (4.5 cm)
and 21/4 in. (5.5 cm) are also available but are seldom used in water
tank foundations. Welded wire fabric is another form of reinforcement
often used in floor slabs.
Cement is a powdered substance produced from a burned mixture
of clay or shale and limestone. Portland cement is the most common
type, grayish in color, consisting chiefly of calcium and aluminum
silicates. Portland cement is manufactured to various designations
on the basis of the physical and chemical requirements as defined
by ASTM C150. Type I designation represents the general-purpose
cement for foundations subject to normal exposure. Where sulfate
attack from soil or water is a concern, if high strengths at an early
period are required, or if hydration heat needs to be minimized,
other ASTM cement types would be better suited and should be
specified.
Aggregates are generally classified into fine and coarse categories
on the basis of their particle size. Fine aggregates consist of sands that
pass through a no. 4 sieve, meaning that their maximum particle size
is less than 1/4 in. (6.4 mm). Some references include a particle size
up to 3/8 in. (9.5 mm) in fine aggregates. Coarse aggregates constitute
any material larger than 3/8 in. (9.5 mm). The most common aggre-
gate size is about 3/4 in. (19 mm). However, the maximum coarse

Foundations
168 Chapter Five
aggregate size is governed by the space limitations between the re-

inforcing bars. As the aggregates constitute up to 75 percent of the
total volume of concrete, it is important not only that the proper size
be selected carefully, but that the aggregates be properly graded and
have the requisite strength, durability, and weather resistance for the
exposure environment.
Uniformity and workability of concrete are affected by the aggre-
gate gradation or particle size distribution within the aggregate. A
properly graded aggregate has a balanced distribution of particle size
that remains consistent from batch to batch. Aggregates with smaller
gradation minimize the number of air voids and result in more dense,
stronger, and better concrete.
Water is a necessary ingredient that initiates the hydration process
of cement. The mixing water should be clean potable water that is
free of oils, acids, alkalis, salts, and other organic materials that are
detrimental to concrete or the reinforcement steel. Likewise, the water
should be free of high concentrations of dissolved solids.
Admixtures should be used only when required by design. They
improve the workability of plastic concrete and enhance the proper-
ties of hardened concrete. The admixtures include air-entraining to
increase resistance to freezing, water-reducing admixtures to reduce
the quantity of water needed to maintain a certain slump, retarding
agents to slow the setting of concrete, accelerators to hasten strength
development at an early age, and fly ash and ground, granulated blast-
furnace slag to improve the plastic or hardened properties of cement
concrete.
Durability
Because water tanks are erected in varied climates and locations, the
environmental effects on their concrete foundations can be harsh and
must be taken into consideration. Conditions that can profoundly af-
fect the service life of the foundations include extreme temperature
fluctuations, freeze/thaw cycles when exposed to water, and exposure
to chemicals, salts, deicers, etc. Durability, in essence, refers to the ca-
pability of concrete to withstand these exposure conditions without
damage, distress, or deterioration.
ACI 318-05 provides detailed durability requirements for im-
proving the performance of concrete. These requirements include
air-entraining recommendations for concrete exposed to freez-
ing/thawing or deicing chemicals, maximum water-to-cement ratios
and minimum strength values for concrete exposed to special condi-
tions, maximum percent of total cementitious material by weight for
exposure to deicing chemicals, and criteria for resistance to sulfate-
containing solutions and soils. This reference also provides require-
ments for corrosion protection of the reinforcing steel.

Foundations
Foundations 169
Quality Control, Mixing, Placing, Finishing, and Curing

For reinforced-concrete foundations to perform and function as in-
tended, it is essential to adhere strictly to specific requirements in
each phase of construction. These requirements govern the strength,
durability, mixing, conveying, depositing, workability, and curing of
concrete. The quality of concrete is mostly defined by the quality of the
cement paste, its proper mixing with the aggregates, and the aesthetic
quality of the finish.
Structural concrete is proportioned to achieve an average con-
crete strength based on a mix design and an anticipated exposure.
The proportions are established to provide workability and consis-
tency, resistance to special exposures, and conformance with strength
test requirements as outlined in ACI 318-05. The mix design must be
followed precisely to produce concrete that satisfies structural perfor-
mance requirements.
Placing and compacting are also important to the quality of con-
crete. Concrete must be placed and vibrated properly to avoid segrega-
tion, honeycombing, settling, and separation of the heavier aggregates
from the rest of the mix. Concrete should be placed continuously in
lifts or layers using multiple discharge locations. This eliminates ag-
gregate separation caused by the horizontal flow of concrete within
the formwork and the need for concrete to be moved into its final
position. During placement, samples of the plastic concrete should be
taken for field testing of unit weight, slump, and air content to en-
sure compliance with mix specification. The tests can be performed
according to the appropriate ASTM specifications.
Unless cured by accelerated curing techniques, poured-concrete
foundations should be maintained at a temperature above 50◦ F (10◦ C)
and in moist condition for at least 7 days after placement. Schedule-
driven activities often dictate backfilling around the footings sooner
than 7 days. In no case should backfilling be started before the con-
crete has gained sufficient strength to withstand the loading induced
on the footing by the pressures resulting from the backfill and the
construction equipment.
Required Strength, fc

Structural concrete can be proportioned to a wide range of design
strengths and characteristics. The design of water tank foundations is
based on a specified design strength for concrete. The design strength
refers to the compressive strength gained by concrete after 28 days of
curing and is referred to as f c . For water tank foundations, a minimum
design strength of 3,000 psi (20.7 MPa) is recommended. The preferred
specified strength, however, is 4,000 psi (27.6 MPa). A common prac-
tice in the industry is that when concrete strength in excess of 3,000 psi
(20.7 MPa) is required by the specifications, a design strength that is

Foundations
170 Chapter Five
500 psi (3.4 MPa) less than the specified strength is used in compu-
tations. This allows some flexibility in situations where strict quality
control measures cannot be maintained due to the long hauls to remote
job sites.
Evaluation and acceptance of concrete for a given design mix can
be made on the basis of actual testing. Adjustments to the mix design
can be made to improve the resulting strength as necessary. To en-
sure that the concrete furnished meets the specified design strength
requirements, fresh concrete specimens can be prepared at the job
site for testing in the laboratory. ACI 318-05 provides criteria for con-
crete sampling and testing and for acceptance of concrete compressive
strength: Concrete strength is considered acceptable when the aver-
age of any three consecutive strength tests equals or exceeds f c and
no individual test (average of two cylinders) falls below f c by more
than 500 psi (3.45 MPa).
For the test results to be meaningful, it is critical that the samples
be taken, handled, and cured in strict compliance with the applicable
ASTM standard. Testing of cylinders that are mishandled or ignored at
the job site may not be truly representative of the concrete furnished.
Also, it is important that qualified personnel test all specimens, as the
outcome of the tests determines the acceptability of the foundations.
If the strength test results fail the acceptability criteria, hardened
concrete can be tested by taking core bores in accordance with ASTM
C42. Experience indicates that unless the requirements of ASTM C42
are strictly adhered to, the core bore test results will underestimate
the true strength of the hardened concrete. ACI 318-05 also provides
specific criteria for core drill testing.
Reinforcing Steel
The reinforcement steel used in water tank foundations is generally
deformed bars conforming to the ASTM A615 specification and hav-
ing a minimum yield strength of 60 ksi (414 MPa). Where weldability
is a requirement, low-alloy steel deformed bars conforming to ASTM
A706 can be used. As noted previously, welded wire fabric reinforce-
ment can be used in floor slabs. These fabrics conform to ASTM A185
for plain wire and ASTM A497 for deformed wire. Epoxy-coated bars
or wires are not necessary for water tank foundations.
ACI 318-05 provides detailed criteria for determining the amount
of reinforcement necessary in design as well as requirements for the
spacing, cover, development length, and splice length. Requirements
for the development length of bars should be carefully reviewed, espe-
cially for the horizontal bars. If horizontal bars are so placed that more
than 12 in. (30.5 cm) of concrete is cast in the member below the bar,
the required development length is 1.3 times the normal development
length.

Foundations
Foundations 171
Nonconforming Concrete
On occasion, it is possible that concrete mixed and furnished under a
specification fails to meet the minimum design strength requirement.
As a result, contractual obligations may require the foundation con-
tractor to remove the nonconforming foundations and reinstall the
foundations according to the stated requirements. This can signifi-
cantly delay project completion and can be very costly. Therefore, it is
important that serious attention be paid to quality control during the
mixing, conveying, placing, and curing phases of the concrete foun-
dation construction.
Contractual obligations notwithstanding, not every nonconform-
ing condition warrants the removal of the footings. If testing confirms
low strength, calculations can be performed to check the adequacy
of the furnished concrete with reduced strength and the actual rein-
forcement provided. The nonconforming condition may be accepted
if the calculations confirm that the load-carrying capacity of the foun-
dation is not significantly reduced and that the design intent has been
met.
Formwork and Removal

Construction of water tank foundations requires the use of formwork.
Proper formwork ensures that the foundations conform to the shape
and dimensions shown on the drawings. The formwork also prevents
moisture loss from concrete, and it facilitates curing, especially when
the top surface of concrete is kept moist. To be effective, the formwork
must be strong enough to withstand the loads and pressures from
pouring concrete and any other loads that are present. Leakage of
plastic concrete must not occur.
Formwork should remain in place for as long as possible, espe-
cially in cold weather. Formwork must not be removed until the con-
crete has gained sufficient strength to withstand its dead load and
any other construction loads. When properly cured, general-purpose
concrete gains about 500 psi (3.45 MPa) strength in 24 hours; within
1 week of placement, it reaches nearly 70 to 75 percent of its maximum
compressive strength. Although many contractors are in a hurry to re-
move the formwork so that they can complete backfilling around the
footings, it is best to keep the formwork in place based on achieving
a defined minimum strength.
To simplify formwork, some contractors take the liberty of adjust-
ing footing thickness or other dimensions. This should not happen
without the explicit consent of the engineer of record. The formwork
required for water tank foundations is relatively simple and should
easily facilitate the required shape and dimensions shown on the en-
gineering drawings.

Foundations
172 Chapter Five
Tank diameter
Overflow elevation
Head range
High water line
Column (typ.)
Low water line
Center riser
Top of
footing Elevation
elevation
Centerline to
centerline of
foundation
at
eter
Diam rline of
e
cent ation
o un d
f
Plan
FIGURE 5-1 Typical leg tank elevation and shallow foundation plan.

Foundations
Foundations 173
Shallow Foundations
The shallow foundation is the most cost-effective foundation for water
storage tanks. Shallow foundations typically include isolated footings
or mat or raft foundations placed just below the columns on the lowest
part of the structure, as applicable. Footings can be placed as shallow
as possible as long as the bottom of the footing is below the frost line,
the resulting bearing pressures are within the allowable limits, the
overall settlements are tolerable, and the stability requirements are
met. However, for water-storage tanks, shallow foundations can be
typically placed as shallow as 4.5 ft (1.4 m) and as deep as 10 ft (3.0 m)
below grade.
Shallow foundations transfer structural loads to the bearing soil
or rock strata occurring below the base of the footing. Shallow foun-
dations for multicolumn elevated storage tanks typically consist of
isolated piers with footings (Fig. 5-1). For ground storage tanks and
elevated single-pedestal tanks, the foundation may take the form of
a ringwall, a ring-tee, or a ring-slab. These ring-type foundations are
further discussed later in the chapter. Ground storage tanks may also
be founded on a slab or a granular berm. Figures 5-2a, 5-2b, 5-2c, and
5-3 show several common types of shallow foundations.
Based on tank geometry, site conditions, and specific environmen-
tal loading effects, various foundation alternatives should be evalu-
ated. Typically a shallow foundation is the preferable option. If poor
soil conditions, high settlement expectations, or low bearing capac-
ities dictate, deep-foundation alternatives must be considered. Low
bearing capacities generally result in large footings, causing the adja-
cent footings to encroach upon each other. As a result, the overlapping
of the pressure bulbs from the individual footings can exacerbate the
bearing stresses and magnify settlements. Therefore, when the net
allowable bearing pressure falls below 2,000 psf (96 kPA), the deep-
foundation alternative should be pursued.
Loads and Load Combinations

Water-storage tanks are subjected to a variety of loads. The gravity
forces consist of the weight of the tank metal, appurtenances, and
the liquid. Common appurtenances include roof-mounted cellular
antenna systems, platforms, floors, walkways, ladders, and piping.
Snow loading consists of the weight of snow on the tank balcony and
the tank roof where the roof slope with the horizontal axis is flat to
moderate. The tank roof may also be subjected to live loading that is in
excess of the snow loads. Lateral forces on the tank and tower consist
of loads resulting from wind pressures or earthquake ground motion.
The AWWA D100-05 Standard for Welded Carbon-Steel Tanks for
Water Storage states that a unit weight of 62.4 lb/ft3 (1,000 kg/m3 ) for
water, 490 lb/ft3 (7,850 kg/m3 ) for steel, and 144 lb/ft3 (2,310 kg/m3 )

Foundations
174 Chapter Five
Radial centerline Centerline of

foundation
Centerline of foundation
Centerline of
anchor bolts
Diameter at centerline
of anchor bolts
Plan
Anchor bolts
Centerline of
Centerline of
anchor bolts
foundation
Top of footing Dowels

elevation
Hoops
Top of grade Exposure
Bars
Bars
(a) Elevation
FIGURE 5-2a Typical shallow foundations: sloped slab

Foundations
Foundations 175
Centerline of
Radial foundation
centerline
Centerline of
anchor bolts
Diameter at
centerline of
anchor bolts
Plan
Anchor bolts
Centerline of
Centerline of
anchor bolts
foundation

elevation
Hoops
Top of grade Exposure
Bars
(b) Elevation
FIGURE 5-2b Flat slab

Foundations
176 Chapter Five
Hole for
inlet/outlet
pipe
A Anchor bolts
on bolt circle
Plan
Centerline of
Anchor bolts
foundation
Top of footing Slab reinforcing

elevation
U-bars
Exposure Dowels
Dowels
Vault for U-bars

piping
Bars
Bars
(c) Elevation
FIGURE 5-2c Riser.

Foundations
Foundations 177
Centerline of
tank and foundation
Top of footing elevation
Exposure
Mat top
reinforcement bars
Mat bottom
(a) reinforcement bars
Extend 3–5 ft
(0.9–1.5 m) Tank plate
beyond tank 2 ft (0.6 m) Centerline of
minimum tank and foundation
Top of berm Compacted crushed stone,
elevation screenings, fine gravel, clean
Slope 1 sand, or similar material
To drain down 1.5
away
1
1
Coarse stone or Coarse stone or
coarse gravel Thoroughly compacted fill of
gravel, coarse sand, or other coarse gravel
(b) stable material
FIGURE 5-3 Examples of shallow mat and berm foundations: (a) typical square
mat foundation and (b) typical granular berm foundation for flat-bottom tanks.
for concrete should be considered in the design of tank structures and

foundations. The standard also recommends consideration of a min-
imum allowance of 25 lb/ft2 (1,205 N/m2 ) for the pressure resulting
from the design snow load on the horizontal projection of the tank
roof surfaces with slopes not exceeding 30 degrees. A reduction of
this allowance is permitted in warmer regions where snow loading
is smaller. However, D100-05 limits the minimum roof design load to
15 lb/ft2 (720 N/m2 ).
AWWA D100-05 has adopted the American Society of Civil Engi-
neers (ASCE) standard 7-05 for wind loading criteria. However, it re-
tains the minimum design pressures to be 30 Cf lb/ft2 (1,436 Cf N/m2 ),
with the force coefficient Cf being 1.0 for flat surfaces, 0.60 for cylin-
drical or conical surfaces with apex angle <15 degrees, and 0.50 for
double-curved or conical surfaces with apex angle ≥15 degrees. For
seismic design, the AWWA standard has essentially adopted ASCE
7-05 criteria with some variation in the minimum design acceleration.
ASCE 7-05 provides detailed criteria for both wind and seismic load-
ing. Proper determination of the period of oscillation of the water tank
system is necessary in all seismic evaluations.
The wind or seismic forces can originate from any direction. Struc-
tural analysis indicates that the greatest uplift force in a multicolumn
elevated water tank occurs in the column that is situated exactly up-
wind of the lateral force. The maximum uplift force generally occurs

Foundations
178 Chapter Five
when the tank is empty. For taller and shorter towers in areas of high
seismic risk, it would not be unusual to find that the maximum uplift
occurs under seismic loading when the tank is full. Similarly, the great-
est downward force occurs in the column situated exactly downwind
of the lateral force. The downward load will be a maximum when the
tank is full. The direction of the lateral force that will cause the greatest
uplift may not be the same as the direction of the lateral force that will
cause the greatest downward force. Hence, all column foundations
are candidates for the worst-case orientation.
Typically, structural loading required for the design of the founda-
tions is determined from the analysis of the elevated tank and tower
or pedestal. The resulting reactions, shears, and overturning moments
due to the gravity loads, wind loads, and seismic loads are all defined.
These loads, in addition to the foundation dead loads and other loads
emanating from soil pressure or swells, constitute the design loads.
Foundations are generally designed according to ACI 318-05 and
its commentary, ACI 318R-05. This building code for structural con-
crete stipulates that the foundations be designed to have design
strengths at all sections at least equal to the required strength based
on factored loads in defined load combinations. Although ACI 318-05
still retains the classical factored load combinations in its Appendix
C as an alternative, in its 2005 edition it has adopted the ASCE 7-05
factored load combinations for design. In the seven load combinations
stated for determining the required strength U, loads not present can
be eliminated from the load combinations:
U = 1.4(D + F) (5-1)
U = 1.2(D + F + T) + 1.6(L + H) + 0.5(L r or S or R) (5-2)
U = 1.2D + 1.6(L r or S or R) + (1.0L or 0.8W) (5-3)
U = 1.2D + 1.6W + 1.0L + 0.5(L r or S or R) (5-4)
U = 1.2D + 1.0E + 1.0L + 0.2S (5-5)
U = 0.9D + 1.6W + 1.6H (5-6)
U = 0.9D + 1.0E + 1.6H (5-7)
where
D = dead loads
F = load due to weight of fluids
Lr = roof live load
L = live load
R = rain load
S = snow load
E = load effects of seismic forces
W = wind load
H = loads due to weight and pressure of soil, water in soils,
or other materials

Foundations
Foundations 179
T = cumulative effect of temperature, creep, shrinkage, differential

settlement, and shrinkage-compensating concrete
It is the authors’ view that the load due to the weight of the flu-
ids F should be included in the strength design load combinations
(Equations [5-4] and [5-5]) with a load factor of 1.2—that is, 1.2F . Oth-
erwise, the case of the full tank under wind or seismic loading may
not be appropriately considered. Interestingly, the ASCE 7-05 basic
load combinations for allowable stress design correctly include the F
loads in load combinations for wind and seismic loading. Likewise,
the weight of fluids should be included in the seismic load combina-
tions (Equations [5–7] and [5–10]), as the seismic uplift can be more
significant in areas of high seismic risk or in the case of tall elevated
tanks when the tank is full. The uplift for an empty tank is generally
governed by the wind load combinations.
Alternatively, the classical ACI 318 Appendix C load combinations
may be used:
U = 1.4(D + F ) + 1.7L (5-8)
U = 0.75(1.4D + 1.4F + 1.7L) + (1.6W or 1.0E) (5-9)
U = 0.9D + (1.6W or 1.0E) (5-10)
U = 1.4D + 1.4F + 1.7L + 1.7H (5-11)
Where structural effects (differential settlement, creep, shrink-

age, expansion of shrinkage-compensating concrete, or temperature
change T) are significant, U should not be less than the larger of the
following equations:
U = 0.75(1.4D + 1.4F + 1.4T + 1.7L) (5-12)
U = 1.4(D + T) (5-13)
Regardless of the load combination selected for design, the load

factor 1.6 on wind load can be reduced to 1.3, where W has not been
reduced by a directionality factor. Also, the factor 1.0 on the seismic
load should be increased to 1.4, where E is based on service-level
seismic forces. Refer to ACI 318-05 for other specifics in using these
load combinations.
It should be noted that in foundations design, AWWA D100-05
Section 12.1.1 requires the water load to be considered as live load,
with appropriate factors for live load. Furthermore, the standard does
not require the inclusion of snow loading in the load combinations
that include wind or seismic loads.
An additional AWWA D100-05 requirement that is associated with
the ductility of the bracing rods states that foundations should be
checked for stability at a lateral seismic force equal to yielding of
bracing rods. For A36 rods, the actual yield stress may be as much
as 1.33 times the minimum published yield. Hence, the anchor bolts
must also be checked to ensure load transfer.

Foundations
180 Chapter Five
Bearing Capacity
Bearing capacity refers to the ability of the soil strata below the footing
to safely resist the structural loading on the foundation with reason-
able safety and tolerable settlements. The loads described earlier must
all be transferred to the bearing strata through the foundation system.
The pressure resulting from the structural loading at the interface
between the foundation and the bearing strata is referred to as the
bearing pressure. The bearing pressure must always remain below
the ultimate bearing capacity of the bearing soils. The ultimate bear-
ing capacity may correspond to a general shear failure or a punching
shear failure in the soils. However, in design, safety factors are applied
to further limit the bearing pressures to levels commonly referred to
as the allowable bearing capacity.
The determination of the ultimate bearing capacity follows
Terzaghi’s equations (Terzaghi and Peck 1967). Based on equilibrium
analysis and experimentation, Terzaghi expressed the ultimate bear-
ing capacity in semiempirical forms that can be expressed as
q ult = c Nc + qNq + BN (5-14)
where
= 1.0 for strip foundation

= 1.3 for square and circular foundations
= (1+ 0.3B/L) for rectangular foundation
= 0.5 for strip foundation
= 0.4 for square foundation
= 0.3 for circular foundation
= 0.5 × (1 − 0.2B/L) for rectangular foundations
= unit weight of soil
q = surcharge, or times the bearing depth of the foundation
B = width or diameter of the footing
L = length of the footing
c = cohesion of soil
Nc , Nq , N = bearing capacity factors determined on the basis of
the angle of internal friction of soil
Refer to a soil mechanics textbook for further details (Terzaghi and

Peck 1967 or Smith and Pole 1981).
Water tank foundations are designed using the net allowable bear-
ing capacity. The net allowable bearing capacity is determined by sub-
tracting the effective surcharge or the overburden pressure from the
ultimate bearing capacity and dividing the result by a factor of safety
(FS). The FS included in the recommendations of the geotechnical
report is critical and should be reviewed carefully. Typically, the FS

Foundations
Foundations 181
ranges from 2.5 to 3.0. However, AWWA D100-05 prescribes specific

values for various bearing conditions:
r A safety factor of 3.0 shall be provided on the basis of calcu-
lated ultimate bearing capacity for gravity loads.
lated ultimate bearing capacity for gravity loads plus wind
load, excluding overturning toe pressure caused by shear at
the top of the footing, unless specified otherwise. The safety
factor may be reduced to 2.25 when specified in the geotech-
nical report.
lated ultimate bearing capacity for gravity loads plus seismic
load, excluding overturning toe pressure caused by shear at
the top of the footing, unless specified otherwise.
lated ultimate bearing capacity for gravity loads plus wind
or seismic load, including overturning toe pressure caused by
shear at the top of the footing.
Therefore, the geotechnical investigation must identify the net al-

lowable bearing capacity using an FS of 3.0 against the ultimate bear-
ing capacity. If a different FS value is specified in the geotechnical
report, the bearing pressures must be corrected to an FS of 3.0, as
required by AWWA D100-05.
Design of Isolated Spread Footing

The isolated spread footing for water-storage tanks consists of a pier
or pedestal sitting on top of a flat or sloping slab. The dimensions of
the pier are generally specified to be compatible with the column ori-
entation and size and the base plate bolting requirements. The height
of the pedestal is a function of the bearing depth recommended by the
geotechnical engineer. The pedestal height is also affected by lateral
stability of the footing.
Determining the footing size is an iterative process. The base area
of the footing is initially determined from unfactored forces and mo-
ments by maintaining the resulting bearing pressures below the al-
lowable bearing pressure. With this determination, the footing is then
subjected to the load combinations defined previously. Appropriate
adjustments to the slab dimensions, thickness, or bearing depth are
made to satisfy both equilibrium and stability (lateral and uplift) re-
quirements. Once the footing is sized, the concrete sections and rein-
forcement requirements are selected in accordance with ACI 318-05
Ultimate Strength Design Method.

Foundations
182 Chapter Five
The footing thickness is based on the concrete flexural and shear-

strength requirements of ACI 318-05. The slab bending moment re-
quirements are checked at the faces of the pedestal. Checks are also
made on strain compatibility to ensure that the failure mode will be
by yielding of the reinforcement, not by the crushing of concrete in
the compression zone. Beam or flexural shear is checked at the critical
distance d from the face of the foundation pier. The punching shear
is checked at the critical distance d/2 from the face of the pier. The di-
mension d represents the distance from the extreme compression fiber
to the centroid of longitudinal tension reinforcement in the slab, at the
face of the pedestal. When checking the punching shear or flexural
shear capacity of sloping slabs, caution is warranted to use the actual
depth at the location under consideration, not the depth at the face of
the pier.
For the multicolumn tank, the tower typically consists of tubular
columns with a base plate at the top of the footing. The column pier
or pedestal is sized to accommodate the column and base plate in
addition to providing adequate embedment depth and edge distance
for the anchor bolts. The pier can be circular, square, or—in the case of
a battered column—rectangular. The rectangular pier allows the line
of action of the column axial force to be centered on the footing, thus
avoiding creation of eccentric moments at the base. The depth of the
foundation below grade, the pedestal projection above grade, and the
thickness of the footing determine the required height of the pier.
The designer must ensure that the pier does not become slender.
Otherwise, it would have to be designed as a column. ACI 318-05 limits
the ratio of the pedestal height to its average least-lateral dimension
to a maximum of 3. By this limitation, the ACI building code provides
specific reinforcing requirements that are lighter than those typically
required for columns.
The minimum reinforcement of flexural members relevant to the
foundation slabs is defined in ACI 318-05 Section 10.5.1. Section 10.5.3
allows this minimum reinforcement requirement to be waived if the
calculated area of the reinforcement is increased by one-third. For foot-
ings of uniform thickness, Section 10.5.4 specifies that the minimum
area of tensile reinforcement in the direction of the span should be the
same as that required for shrinkage and temperature defined in Sec-
tion 7.12. The upper bound on the flexural reinforcement ratio should
satisfy the requirements of Section 10.3.5 or Section B.10.3.3. Shear
friction should be checked at the pedestal/slab interface according to
Section 11.7.4.
For cast-in-place pedestals, ACI 318-05 Section 15.8.2.1 specifies
that an area of reinforcement across the interface should not be less
than 0.005 times the gross area of the pedestal. For shorter pedestals,
this value can be arguably reduced by as much as 50 percent accord-
ing to industry practice. However, the value should be maintained

Foundations
Foundations 183
for taller pedestals. The vertical dowels constituting the pedestal re-
inforcement resist the uplift and lateral forces that exist at the upwind
columns when the tank is empty and the design wind load is fully
active. They also strengthen the pier in resisting compressive loads.
The dowels will require lateral ties as per ACI 318-05 Section 7.10.5.
As was noted earlier, once the foundation system is sized based
on all other requirements, it is necessary to check for vertical (uplift)
and lateral (sliding) stability. When the uplift is severe, it will control
the embedment depth and the size of the footing. To ensure stability
against uplift, the foundation can be placed deeper or made larger
to maximize the dead weight of the soils directly above the footing.
This will be especially necessary when the water table is high and
when buoyant weights are used in design. Similarly, footing depth
or slab dimensions can be adjusted to provide stability against lateral
sliding. When such adjustments are made, it will be necessary to revisit
final reinforcement requirements for code compliance. (Typically, for
buoyant weights, 60 lb/ft3 [961 kg/m3 ] for soil and 82 lb/ft3 [1,314
kg/m3 ] for concrete are used in design.)
AWWA D100-05 requires the weight of the pier (footing) plus the
weight of the soils directly above the pier to be sufficient to resist the
maximum net uplift occurring when the tank is empty. The lateral
stability is provided by the passive resistance, cohesion, and adhesion
provided by the soils. It is recommended that a minimum FS of 1.3 be
maintained against the working load uplift by including the weight
of a 25-degree soil wedge and that a minimum FS of 1.5 be maintained
against lateral sliding.
In elevated tanks, the riser carries a major portion of the water
weight. For torus-bottom tanks, this loading may be equivalent to the
weight of water within half the diameter of the tank times the tank
head range. The loading and the requirement for pipe entry and exit
at the base of the riser footing make the riser foundation unique. The
pipe pit design differs considerably from the column pedestal design
in that it has a top slab that can support a considerable load.
The pipe pit often has an open front so that support for the top slab
is provided by just three walls. The slab, treated as a two-way slab, is
supported on the front edge by a deepened girder or non-deepened
band beam across the open face of the pit and on the other three edges
by the walls of the pit. The load transferred to the top slab by the riser
consists of two parts: one part comprises the direct loads from the tank
transferred by a compressive axial stress in the riser pipe walls, and
the other part is due to the water column that bears on the riser floor
in wet risers. Another design consideration for the top slab is whether
the diameter of the riser pipe is less than or greater than the clear span
in the pipe pit below. If the riser diameter is smaller than the clear
span, the reinforcement in the slab must be attuned to account for the
additional bending moment in the slab.

Foundations
184 Chapter Five
The walls of the pit are reinforced in accordance with ACI 318-05
Sections 14.3.2 and 14.3.3, unless required otherwise. The walls are
typically 12 in. (30.5 cm) thick. They are checked for compressive stress
as well as flexural bending stress caused by soil active pressure, and
their thickness is adjusted as required. Where the water table is high
or when the soils are saturated, soil lateral pressures will be affected,
and their effect must also be considered when specifying design lateral
pressures.
Centerline
of tank and
foundation
ll
wa
r ing
r
ete
iam all
d ringw
s ide e d ia meter
n
I Outsid
Centerline of
tank and foundation
3 in. (76 mm)

at wall Clean dry Crushed
A stone or
6 in. (152 mm) sand
concrete
minimum slab
Plan at center
Compacted
regular or Expansion
gravel fill material
Detail X
Centerline of
tank and foundation See Detail X
Top of footing Sand cushion
elevation Inlet or outlet pipe Exposure
Compacted
backfill 6–12 in. (152–305 mm) Compacted regular
crushed stone Compacted
Hoops Verts. or gravel fill backfill
Section A-A
Concrete
thrust block
FIGURE 5-4 Typical ring-wall foundation plan for a flat-bottom tank. (Verts. =
vertical reinforcement dowels.)

Foundations
Foundations 185
Ringwall, Ring-Tee, and Ring-Slab Footings

Aside from the standard isolated shallow footings, there are other
types of shallow foundations. A ringwall foundation is a cylindrical
wall footing of defined thickness and height. Ringwalls are typically
used for flat-bottom storage tanks, commonly known as reservoirs and
standpipes (Fig. 5-4). Ringwall foundations allow the tank contents to
be supported directly on the soils at grade as long as the allowable
bearing capacity of soil is not exceeded. The ringwall itself supports
the weight of the tank container and significant appurtenances, a small
tributary weight of the water, the roof loads, and pressures resulting
from the effects of wind and seismic loads.
Foundation systems consisting of ring-tees or ring-slabs are typ-
ically used for single-pedestal tanks such as pedespheres, fluted-
column tanks, or composite elevated tanks. A ring-tee foundation is
essentially a ringwall supported on a footing (Fig. 5-5). When the
bearing capacity of the soil is exceeded under a ring-tee foundation, a
ring-slab foundation is used. The ring-slab comprises a ringwall sup-
ported on a slab whose diameter is larger than the diameter of the
ringwall (Fig. 5-6).
Ringwall, ring-tee, and ring-slab footings—as with the single
footing—are sized for the load combinations defined previously.
These footings may be symmetrical about the ringwall or asymmetri-
cal. Asymmetrical footings are used to balance the shear or bending
moments along the two faces of the ringwall. This minimizes tor-
sional moments on the ringwall. Footings containing a ringwall are
also subject to hoop stresses from the soils or the surcharge that must
be considered in design.
As with all foundation systems, it is necessary to check for stability
against overturning as well as lateral sliding.
Backfill and Lateral Stability

Backfill is an essential component of properly designed foundation
systems. The geotechnical investigations generally determine the suit-
ability of the in-situ soils for structural backfill. If the on-site soils are
determined to be unsuitable, recommendations for alternate backfill
material are made in the report. Backfilling in accordance with the
geotechnical recommendations ensures lateral stability of the founda-
tion and stability against uplift.
Backfill soils must be capable of providing the necessary passive
resistance to stabilize the foundation against horizontal sliding and
to eliminate the possibility of water accumulation and buildup. As
noted earlier in the chapter, select structural fills that consist of uni-
formly graded sands to silty or slightly clayey sands are well suited
for backfill. However, they must be free of organics and other dele-
terious material, and it is preferable that less than 30 percent passes

Foundations
186 Chapter Five
Anchor bolts
on bolt circle
Ins
dei
Ou
dia
tsi
O ut de
me
side dia
diam Centerline of
te
m
e
r sl
ete tank and foundation
r rin ter s
gwa la
ab
ll b
all Centerline of
in w
g tank and foundation
m e ter r
e dia
Insid
Ch
or
d
Ch
or
d
Radial bars
Hoops
A
Plan
Anchor bolts Centerline of

Hoops tank and Do not grout
Top of (special top bars) foundation
footing pipe. Bend bars
Gravel floor around opening
elevation
Regular or Bars
Hoops Concrete gravel fill B
Verts. Pipe sleeve
Verts. Thrust block for inlet-
B outlet
Section B-B
Hoops Roughened
Suitable bearing construction joint
Radial bars strata (or a shear key)
Section A-A
FIGURE 5-5 Typical ring-tee foundation plan for a single-pedestal tank. (Verts.
= vertical reinforcement dowels.)
through the no. 200 sieve. Soils with high values of liquid limit and/or
plasticity index should be avoided. A liquid limit in the range of 30 to
35 and a plasticity index of less than 15 are commonly preferred.
Backfill may also be required to replace unsuitable bearing soils.
The backfill material in this case may consist of well-compacted struc-
tural fills as defined above, clean-washed crushed stone (e.g., no. 57
stone), or a lean-concrete mud mat with a compressive strength of

Foundations
Foundations 187
Anchor
bolts on
bolt circle
Centerline
Ou of tank and
Out tsi foundation
side de
diam dia
eter me
ring te
wal r s
l la
b
Centerline Centerline
A of inlet- gw all of tank and
A
r rin
outlet pipe mete foundation
e dia
Insid
spCh
ac ord
Ch
in
g
or
d
sp
ac
in
g
ing
sp oop
ac
Radial bars
H
Hoops
Plan
Centerline of Compacted Do not grout pipe.

Anchor bolts tank and Bend bars around
regular or opening
Top of footing foundation gravel fill
Gravel floor or Compacted
elevation concrete slab
backfill
Bars
Compacted soil
Hoops or ground fill B Pipe sleeve
Concrete for inlet-
Verts. Verts. thrust block Inlet-outlet pipe B outlet
Section B-B
Slab bottom Bearing
Hoops elevation
Outside reinforcement Roughened
diameter Radial bars bars construction joint
Slab top (or a shear key)
reinforcement
bars
Section A-A
FIGURE 5-6 Typical ring-slab foundation plan for a single-pedestal tank. (Verts.
= vertical reinforcement dowels.)
about 2,000 psi (14 MPa). The geotechnical consultant must provide
specific recommendations as to the appropriate backfill material and
required compaction.
Resistance to sliding is generally derived from the passive resis-
tance of the soils acting against the foundation. Cohesive soils also

Foundations
188 Chapter Five
draw passive resistance from the soil cohesion. In addition, the shear-
ing resistance at the base of the footing resulting from internal friction
in the soil may contribute to the lateral resistance. The coefficient of
friction at the base of the footing ranges from about 0.3 for silty soils
to 0.5 for coarse-grained cohesionless soils.
The backfill around the footings is commonly placed in 6- to 8-in.
(15- to 20-cm) lifts and is compacted to 95 to 98 percent of the soil’s
standard proctor maximum dry density (ASTM D698) or modified
proctor maximum dry density (ASTM D1557) criteria. Manipulation
of the moisture content of the backfill material may be necessary to
achieve the required compaction. Flowable fill is an alternative back-
fill material that is simple to place and does not require elaborate
compaction.
Settlement
All structural foundations are subject to settlement. As long as the set-
tlements are reasonably small and uniform, their effect on the structure
is relatively small. However, if the settlements become large and the
differential settlements excessive, there can be serious consequences
that could lead to failure. Therefore, it is absolutely essential that the
settlement of all foundations is estimated and that its effect on the
structure as a whole is examined before construction proceeds.
Geotechnical consultants are expected to provide proper assessment
of the total and differential settlements.
As was stated earlier, for water tanks it is preferable to limit the
total and differential settlements to a maximum of 2 in. and 1 in. (5 cm
and 2.5 cm), respectively. If shallow foundations will cause excessive
settlement, deep foundations can be used to further limit these set-
tlements. The effect of the settlement on the piping should also be
carefully examined. Special piping and pipefittings are available that
should be used when flexibility in the system is required.
Settlement of foundations bearing on rock is not a concern as
long as all individual footings bear on rock. However, the rock layer
must be thick and strong enough to support the loads without be-
ing crushed (as might be the case with a rock lens). The designer
should be very cautious of situations in which the foundation bears
partly on rock and partly on soils, as high differential settlements can
result.
Pile Foundations
When the bearing soils are weak or prone to excessive settlements,
deep foundations—of which one type is the pile foundation—are
necessary. Piles transfer the structural loads deep into the stronger

Foundations
Foundations 189
underlying soil strata or rock formations. They also transfer the lat-
eral loads to the surrounding soils and maintain lateral stability.
Given the complex nature of the resisting soils, competent advice
from a qualified geotechnical engineer should always be sought on
the basis of a thorough subsurface investigation to assess the ap-
propriate pile type, length, and other characteristics necessary for
design.
Pile foundations consist of long, slender structural members that
are either driven into the soil or poured in place after drilling. Whether
a pile develops its capacity from end bearing or side friction depends
on how deep it is embedded and on the properties of the soils sur-
rounding it. Piles driven to and bearing on hard rock or very dense
layers of soil are primarily end bearing, as they axially transmit the
loading to the bearing strata. Piles driven to shallower depths and not
resting on hard and dense layers transmit the loading mainly by skin
friction and, hence, are referred to as friction piles. Generally, piles
develop resistance through a combination of both end bearing and
skin friction. The resistance varies on the basis of the pile length and
the relative density and consistency of the soil layers.
The lateral capacity of the pile is a function of the soil characteris-
tics near the surface. A simulation technique called “beam on elastic
foundation” can be used to assess the lateral resistance capacity of
the pile. The spring constant necessary for the evaluation can be de-
termined from the elastic or shear modulus of the soil. Pile lateral
load is also a function of the flexural capacity of the pile itself. Pro-
fessional advice must be sought in determining pile lateral capacities
and load-displacement characteristics.
Pile Types
Piles can be driven or cast in place. Available driven pile types in-
clude timber, precast, prestressed concrete, steel pipe, and H-piles.
All have certain advantages and disadvantages. Where the resulting
vibrations from pile driving can be a problem, cast-in-place piles may
be more suitable. Auger-cast piles, drilled piers, or caissons are alter-
natives often preferred over pile driving because of their lightweight
equipment.
Although pile selection depends on many factors—among them
cost, availability, and load test requirements—there are advantages to
using a particular pile type for a given job. The common pile types are
listed in Table 5-2 (ASCE 1993b).
The main disadvantage associated with timber piles is the diffi-
culty of achieving a high-strength connection between the pile and
the pile cap. Similarly, prestressed-concrete piles can pose a challenge
in achieving uplift connection. Dowels can be embedded into the
pile head for transfer of tensile load, but because of physical space

Foundations
190 Chapter Five
Type of Pile Description

Timber piles Best suited for short-embedment, low-capacity, and
low-cost applications. They are also appropriate
for application in regions with corrosive
groundwater. Timber piles are difficult to splice
and very susceptible to decay if untreated.
Precast concrete Usually prestressed, these durable and
corrosive-resistant piles are well suited for higher
capacities. Although precast concrete piles are
available in long lengths, splicing them is difficult
and can be a problem because water tank
foundations resist significant tensile (uplift) and
lateral loads.
Steel pipes Open- or closed-ended steel pipes pose some
difficulties in driving but offer high-resistance
capacity, especially when filled with concrete.
Steel pipe piles are well suited for splicing, but
they do have the potential for corrosion.
H-piles Available in a wide variety of sizes and lengths,
H-piles can be driven easily and spliced rather
conveniently. They offer high axial load carrying
capacity and flexural bending resistance.
Corrosion can be a problem, as with all steel,
which can be alleviated with preventive measures.
Source: ASCE 1993b.
TABLE 5-2 Common Pile Types
constraints, developing a connection for moment transfer can be dif-

ficult for smaller piles.
Capacity and Driving Formulas

Pile capacity evaluation has evolved significantly in recent years, as
have requirements for the design of piles. Criteria endorsed by many
recent building codes reflect the National Earthquake Hazards Reduc-
tion Program (NEHRP) Recommended Provisions for Seismic Regu-
lations for New Buildings and Other Structures (NEHRP 2003). These
requirements are indeed very different than the common design prac-
tices employed by the design community thus far. The differences
are more pronounced for regions of high seismic risk. Therefore, the
geotechnical engineer’s role is extremely important in the design of
pile foundations for elevated water tanks.
Geotechnical engineers generally provide the allowable pile load
either as a set of recommendations or in the form of raw soil borings

Foundations
Foundations 191
that must be interpreted. (The latter approach is not preferred.) The

recommendations may also include options for other pile types and
capacities for flexibility in design. A pile’s load-carrying capacity de-
pends on its type, size, and depth of penetration. Geotechnical consult-
ing engineers also provide pile lateral and uplift capacities, a center-
to-center spacing recommendation, and anticipated settlement under
the governing loads. The minimum center-to-center spacing should
be at least three times the diameter or side dimension of the pile for
end-bearing piles but larger for friction piles, especially when many
piles have been driven in a group.
Pile capacity may be limited either by the pile’s internal struc-
tural capacity or by the external capacity offered to it by the resisting
soils. For the pile to be able to furnish the full resistance offered by
the surrounding soils, the internal capacity must exceed the external
capacity. Pile capacity is basically determined on the basis of an ap-
proved pile-driving formula, wave equations, or load test methods.
IBC-2006 limits the allowable compressive load on any pile to 40 tons
(356 kN) when it is determined on the basis of a driving formula alone.
For allowable loads above 40 tons (356 kN), it recommends the use of
the wave equation method of analysis and verification of this allow-
able load by a load test in accordance with ASTM D1143 Test Methods
for Piles Under Static Axial Compressive Load and ASTM D4945 Test
Methods for High-Strain Dynamic Testing of Piles. Similarly, IBC-2006
provides criteria for allowable frictional resistance and uplift capacity.
IBC-2006 and/or ASCE 7-05 provide detailed criteria for
longitudinal-reinforcement and transverse-confinement reinforcing
steel for precast, prestressed piles as a function of site class and seismic
design category. ACI 318-05, in its Chapter 21, also provides design
criteria for piles, pile caps, and foundations that resist earthquake-
induced forces.
The ultimate capacity of a pile that derives its resistance from both
the side friction and the end bearing is given by
QUltimate = QSide Friction + QTip Bearing (5-15)
The evaluation of Q depends on whether the pile is driven in

sand or clay. Refer to any textbook on pile foundations for appropriate
methodologies for determining pile capacity.
AWWA D100 requires a minimum FS of 2.0 for gravity loads and
an FS of 1.5 for gravity loads plus wind or seismic loads. Other refer-
ences define factors of safety on the basis of whether load tests have
been performed.
Various theoretical pile-driving formulas can be used to estimate
pile load-carrying capacity. These formulas do not correlate well with
test results and are historically inaccurate. However, they are help-
ful in establishing when to stop driving a pile to achieve a capacity

Foundations
192 Chapter Five
equivalent to that of the load test. The reader is once again referred to
a textbook on pile foundations (Prakash and Sharma 1990). The equa-
tion that is often used by the industry is known as the Engineering-
News formula (Liu and Evett 1987):
2Wr H
Qa = (5-16)
S+C
or
1000 Wr H
Qa = (in SI units) (5-17)
6(S + C)
where
Qa = allowable pile capacity (lb [kN])
Wr = weight of ram (lb [kN])
H = height of fall of ram (ft [m])
S = amount of pile penetration per blow (in./blow [mm/blow])
C = 1.0 for drop hammer; 0.1 for steam hammer (25 or 2.5)
There is a built-in FS of 6 associated with Equations (5-16) and

(5-17).
Pile Driving and Load Tests

Before any pile driving begins, the geotechnical engineer must ap-
prove the equipment and method of pile driving. This, along with
monitoring the penetration resistance, will ensure safe driving of the
piles. When the engineer authorizes pre-boring to a certain depth, the
diameter of the hole must be smaller than the pile diameter or side di-
mension. A limit of two-thirds times the diameter or side dimension is
recommended. Larger or oversize holes cause loss of the skin friction
and, consequently, a reduction in the pile axial and lateral load capac-
ity. If pre-boring is necessary, then auger-cast piles or drilled shafts
may be a better option.
The geotechnical engineer must provide guidance on pile driving
and criteria for pile length or depth of penetration. Such guidance can
be in the form of limiting the penetration resistance based on a par-
ticular hammer and rate of energy or recommendations for dynamic
testing using a Pile Driving Analyzer (PDA) to ensure pile structural
integrity and adequate load-carrying capacity.
Unless geotechnical recommendations for pile capacity are based
on previous experience in the site vicinity, most recommendations are
theoretical and only an estimate of the carrying capacity of the pile.
Load tests are performed to determine or to confirm those theoretical
capacities. Test piles should be driven where the soil conditions are
known. Test piles must be the same as the actual piles being used

Foundations
Foundations 193
for the job. Also, the pile-driving techniques on test piles must be the
same as those that will be used for the production piles.
For foundations for elevated water tanks, the most common
method of testing piles under static axial compressive loading is that
performed under ASTM 1143. But this method is costly and requires a
minimum waiting period of 7 days for piles in granular material and
14 days for piles in cohesive soils for dissipation of excess pore wa-
ter pressure after test pile installation and before load testing begins.
More recently, the use of PDA is gaining much acceptance over the
static load tests in view of its fast pace and the more quickly avail-
able resulting data. PDA is also used to assess pile capacity and pile
stresses from measurements of the applied force and acceleration at
the head. Refer to ASCE 1993b and ASCE 7-05 for further details.
Pile Spacing and Group Efficiency

As stated previously, the minimum pile center-to-center spacing
should be three times the pile diameter or side dimension for end-
bearing piles. For friction piles, however, a larger spacing (three to
five times the pile diameter or side dimension) may be required, as
determined by the geotechnical consultant. A larger center-to-center
spacing makes the pile cap larger and heavier and can increase the re-
quired number of piles. However, a smaller center-to-center spacing
between piles may reduce the group efficiency of the piles. Some ref-
erences allow smaller spacing, but the geotechnical engineer should
carefully review any reduction in the spacing.
As the spacing between piles decreases, the group capacity of
piles may not equal the sum of the individual pile capacities in the
group. Pile group efficiency must be evaluated on the basis of a rational
evaluation that considers the overlapping effects of individual piles.
The geotechnical engineer should provide the required pile spacing
and group efficiency along with all other pertinent information.
Auger-Cast Piles
Auger-cast piles are piles that are installed by pumping grout under
pressure into holes drilled to required depth by continuous-flight,
hollow-stem augers. The common diameters of these piles range from
12 to >30 in. (30 to >76 cm), with lengths from 50 to 100 ft (15 to 30 m)
and compressive capacities that can exceed 125 tons (1 MN). Auger-
cast piles are reinforced by inserting a single reinforcing bar and/or a
reinforcing cage through the unset grout. The cage extends to a defined
length based on the structural requirements of the pile in resisting
tensile, compressive, and lateral loading. The single reinforcing bar at
the center of the pile typically continues to the bottom end of the pile.
The grout mostly consists of portland cement, sand, and water.

Foundations
194 Chapter Five
Where vibrations due to pile driving can cause damage to other

structures or entities in the near vicinity of the job site, auger-cast piles
offer a better alternative. As with all other piles, pile axial load tests
are performed in accordance with ASTM 1143. Pile installation records
are maintained as required. More information on these piles can be
found in the Augered Cast-in-Place Piles Manual prepared by the Deep
Foundations Institute (DFI 1990).
Lateral Loads and Bending Moments

In addition to the gravity loads, foundations for elevated water tanks
must resist vertical and lateral loads caused by wind or seismic load-
ing. These loads are transferred through the pile cap to the resisting
piles. Depending on pile head fixity and the characteristics of the sur-
rounding soils, the lateral shears can cause significant bending mo-
ments, in addition to the axial loads acting on the piles. Therefore,
both the pile and the surrounding soils should be analyzed and inves-
tigated for strength and stability.
Along with the allowable load capacity for the pile, geotechnical
engineers must provide allowable lateral load capacities for various
pile head fixity conditions. If the pile head extends into the pile cap
and is anchored by uplift connections, there is very little, if any, pile
head rotation. But due to the movement of the pile cap, lateral trans-
lation of the pile head is possible. For design purposes, however, it
is helpful to define the pile head boundary condition to locate where
the maximum bending moment occurs. With the pile head restrained
against rotation, the maximum bending moment in the pile generally
occurs at the restrained end, at the pile cap. Otherwise, it occurs at
some distance below the pile cap.
Geotechnical engineers often provide bending-moment curves as
a function of the lateral loads and an assumed pile head fixity condi-
tion. These curves are very helpful in design and should be included
in all geotechnical reports that recommend pile foundations. They
should be carefully reviewed for assessing pile structural capacity as
well as the effect of pile lateral displacement on the elevated water
tank system.
Pile Caps and Uplift Connections

Pile caps are reinforced-concrete structural elements that resist direct
vertical and lateral forces and transfer them to the supporting piles.
Pile caps are sized based on the number of piles required for a tower
column loading. Depending on their thickness, pile caps can be rigid
or flexible. Typically, pile caps join the pedestal from the top and the
piles from the bottom. Piles usually extend 4 to 6 in. (10 to 15 cm) into
the pile cap.
The size of the pile cap is a function of the number and center-to-
center spacing of piles required for a footing. The larger and thicker
the pile cap, the heavier it is. Thus, the dead weight of the pile cap

Foundations
Foundations 195
S S
S
S
S S S
Four piles Five piles Six piles
S S
S
S
S
S
S
S S
Seven piles Eight piles Nine piles
FIGURE 5-7 Typical pile group patterns for single foundations. (S = spacing
between piles.)
itself can cause an increase in the number of required piles. Fig-

ure 5-7 provides typical pile layouts for four-, five-, six-, and seven-
pile footings. Pile caps must accommodate pile spacing as well as edge
distance requirements. The pile cap edge distance, measured from the
centerline of the outer piles to the edge of the pile cap, is generally a
function of the pile diameter. Typically, for piles with a diameter of
about 1 ft (30 cm), a distance of 1 ft. 3 in. (38 cm) is used.
The thickness of the pile cap should be checked for punching shear
caused by piles exterior to the critical section. The critical section is
taken to be at a distance of d/2 from the pedestal, where d is the current
depth to centroid of tensile steel in the pile cap. The punching shear
should also be checked around the individual piles at a critical section
taken a distance d/2 from the face of the pile, and the pile cap thickness
should be adjusted, if necessary.
The pile cap thickness is also dependent on flexural shear both
tangentially and radially at a distance d from the face of the pedestal.
Flexural reinforcement, in both the tangential and radial directions,
should also be checked at the faces of the pedestal. The flexural rein-
forcement determined should then be compared against the minimum
flexural reinforcement requirements of ACI 318-05 Section 10.5.1 and
adjusted, if necessary.
IBC-2006 provides specific criteria for pile connection to the pile
cap. For prestressed piles, uplift anchorage to pile cap can be achieved

Foundations
196 Chapter Five
by exposing and developing the reinforcing strands at the top of the

pile. This approach, however, is not permitted by other model codes
in areas of high seismic risk. IBC-2006 permits the option, provided
that the reinforcing strand results in a ductile connection. As an alter-
native, reinforcing dowels can be grouted into the top of the pile. In
timber, steel, or pipe piles, a reinforcing bar can be inserted through
the member and bent upward into the pile cap. The pile cap must
be thick enough to accommodate the anchoring mechanism. Refer to
Figures 5-8a and 5-8b for typical pile foundations.
Radial centerline
Prestressed
concrete piles
Pile layout plan view
Anchor bolt Centerline of foundation

elevation
Projection
Hoops
Exposure
Aggregate
Radial bars Tangential bars
Typical uplift Prestressed

anchor concrete piles
(a) Column foundation with piles elevation view
FIGURE 5-8a Typical pile foundations: typical pile foundation for a single
column

Foundations
Foundations 197
Prestressed, precast
Anchor bolts A concrete piles
on bolt circle
Centerline of
tank and foundation
Outs
ide d
iame
ter rin
gwall
ringwall b
iameter la
Inside d rs
ete slab
m
dia
eter
side
diam
In
d
d
or
or
side
Ch
Ch
Out
Radial bars
Hoops
A
Plan
Anchor
Exposure
Top of bolts Centerline of

footing tank and foundation
Expansion
elevation joint Concrete
slab
Subgrade
Hoops Hoops
(special Verts. Verts.
top bars)
Radial bars
Typical uplift Reinforced
anchor concrete
Hoops thrust
Radial bars
Prestressed block
Hoops concrete piles
(b) Section A-A
FIGURE 5-8b Typical ring-tee pile foundation). (Verts. = vertical reinforcement

dowels.)

Foundations
198 Chapter Five
The pedestals for pile foundations are usually compressive ele-

ments with a ratio of unsupported height to least lateral dimension
not exceeding 3.0. As noted in the section on shallow foundations,
although ACI 318-05 Section 15.8.2.1 recommends a minimum rein-
forcing ratio across the interface (between the pedestal and footing) of
0.5 percent of the gross area of supported member (i.e., the pedestal),
a smaller reinforcement ratio can be justified for the tower column
pedestals due to their large size. For taller pedestals, the 0.5 percent
reinforcement ratio should be maintained. However, the reinforce-
ment furnished must be sufficient to meet the requirements for uplift
as well as flexural requirements necessitated by the lateral shears on
the pedestal.
Pile Stability and Settlement

Piles are required to be laterally braced in all directions. Piles intercon-
nected by a rigid pile cap may be considered braced for lateral stabil-
ity provided they are situated as defined in IBC-2006 Section 1808.2.5.
Elsewhere, the surrounding soils furnish lateral stability along pile
length. In regions where the piles extend vertically through voids or
holes, the piles should be analyzed as columns.
Settlement is an important aspect of design for all water tank foun-
dations, including pile foundations. The geotechnical engineer should
evaluate pile settlement as well as potential differential settlements for
full consideration in design. As discussed previously in the section on
soils and geotechnical investigations, settlement not only affects the
structural behavior but also the piping systems and the interconnec-
tions among the various components and appurtenances.
Drilled-Pier (Caisson) Foundations

Drilled piers, caissons, or shafts offer an alternative design option in
deep foundations. Drilled piers are cast-in-place reinforced-concrete
shafts with or without a bell at the bottom. They are installed by
drilling a hole of predefined diameter and depth at the design location
and then filling the excavation with concrete and reinforcement.
Drilled-pier construction is relatively easy and can be accom-
plished with rotary drilling equipment. Depending on the soil con-
ditions, casings or laggings may be needed to prevent the soils from
falling or caving into the hole. Typical diameters of drilled piers for
water tank foundations range from about 3 to 6 ft (0.9 to 1.8 m). Larger
diameters may be needed for higher-capacity tanks, depending on
availability of large-diameter drill bits. Otherwise, several smaller-
diameter piers will be used, which would require a larger pier cap.

Foundations
Foundations 199
There are many advantages to using drilled piers for water

tank foundations. Because of its size and capacity, a drilled pier
can replace a group of piles and eliminate the need for a pile cap.
Installation of drilled piers does not generate much noise or vibration,
as pile driving does, and drilled piers can be set up in hard-to-access
places. A primary advantage of drilled piers is that they can sustain
large axial loads with minimal settlement when bearing on bedrock.
Pier Bearing Capacity

Drilled piers draw their structural capacity from the reinforced-
concrete shaft. They develop their bearing capacity from side resis-
tance, generated by skin friction, and base resistance, generated by
end bearing. In equation form, the ultimate static load capacity of a
pier can be expressed as
Qu = Qbu + Qsu − Wp (5-18)
where
Qu = pier ultimate resistance (kip [kN])

Qbu = pier ultimate end-bearing resistance (kip [kN])
Qsu = pier ultimate side friction resistance (kip [kN])
Wp = pier dead weight (kip [kN])
An FS of 3.0 is applied for allowable-stress design application.

Some references may apply an FS of only 2.0 on the ultimate resistance
due to side friction Qsu in service-load design application. Others may
also apply a load factor on the shaft dead weight Wp . Refer to Fig. 5-9
for a typical belled, drilled pier configuration.
A very useful and relevant reference on drilled piers is a re-
port entitled Drilled Shafts: Construction Procedures and Design Methods,
issued by the US Department of Transportation, Federal Highway
Administration (Reese and O’Neill 1988). This reference provides de-
tailed analysis and design, fabrication, and quality control criteria
for drilled-shafts foundations. The Bearing Capacity of Soils, prepared
by the American Society of Civil Engineers (ASCE 1993a), is another
source for criteria regarding analysis and design of drilled shafts.
Pier Side or Skin Friction

Pier side resistance offered by the skin friction along the pier shaft is
a function of the undrained shear strength of clay soils as determined
by testing. Shear strength varies with depth and soil strata and is
empirically related to the shaft load transfer in side resistance. The

Foundations
200 Chapter Five
Axial load Additional reinforcement,

if required
Lateral
load
Exposure
Hoops
Diameter depends on
loading and depth
of excavation
Depth can vary per design
Qsu side resistance
Reinforcing steel
WP
Bell—may be used
when required.
Size varies—no larger than
θ
three times shaft

diameter at base. Underream
angle θ is 45° or 60° typically.
Q bu base resistance
Base resistance
FIGURE 5-9 Typical drilled shaft. (Wp = pier dead weight; Qsu = pier ultimate
side friction resistance; Qbu = pier ultimate end-bearing resistance) (Source:
Reese and O’Neill 1988.)
shear resistance offered by sands or cohesionless silts, however, is a

function of the soil angle of internal friction.
The resistance capacity offered by side friction can be very sig-
nificant. For piers socketed into bedrock, it is possible that the entire
resistance capacity is emanating from the side resistance furnished by
the socket.
It is a recommended practice within the drilled-pier design com-
munity to ignore the contributions of side friction and passive re-
sistance in the top 5 ft (1.5 m) or 1.5 diameters of the shaft when

Foundations
Foundations 201
evaluating lateral stability. The reason is that lateral movement by

the pier causes a wedge of soil to move up and out, resulting in a
loss of side friction and passive resistance. Similarly, in clay soils, the
side friction in the periphery of the bell or the bottom of the straight
pier within 1 diameter of shaft length is ignored in determining the
resistance capacity of the pier. The reason is that movement of the
base of the pier can result in the development of a tensile crack in
the soil, which in turn can cause a lateral stress at the base of the pier
and, consequently, a reduced load transfer in side friction (Reese and
O’Neill 1988).
Design of Piers
The diameter of the pier is a function of the soil characteristics within
the profile, the location of the water table, and the presence of lateral
loads and/or moments. The design of the concrete mix and its strength
are also of critical importance.
The geotechnical profile of the soil dictates not only how far down
to extend the pier, but also the method of construction to be em-
ployed, the need for casing and/or dewatering, and the need for
under-reaming. Special characteristics of soils—shrinking/swelling
of plastic soils, occurrence of boulders, remains of abandoned foot-
ings, presence of debris or other unsafe materials, and so on—all re-
quire that certain measures be taken into full consideration. Aside
from the basic structural design, the most important consideration is
the amount of the expected settlement of the pier foundation and its
effects on the elevated water tank system.
Under-reaming, where possible, helps increase the pier bearing
surface and consequently the bearing resistance. Under-reaming can
also be used interchangeably with socketing where required. The lon-
gitudinal reinforcement for drilled shafts depends on the many fac-
tors noted previously, but as a minimum, industry practice has been
to provide at least 0.5 percent of the cross-sectional area and at least six
bars, forming a cage of equally spaced bars. This minimum is actually
based on ACI 318 Sections 10.8.4 and 10.9.1. Section 10.9.1 requires a
minimum reinforcement of 1 percent. For regions of low-to-moderate
seismic risk, Section 10.8.4 states that, for compression members with
cross-sectional areas larger than required by consideration of loading,
it should be permissible to base the minimum on a reduced effective
cross-sectional area not less than half the total area.
Additional reinforcement may be required where heavy tensile
loading or bending moments are present. The longitudinal reinforce-
ment cage may extend a partial depth or the full length of the pier
when required. Hoop or spiral reinforcement is also used for drilled
shafts. Hoops are more economical, but from a performance stand-
point spirals are preferred. Figure 5-10 illustrates pier reinforcement.

Foundations
202 Chapter Five
Ties—spacing as per
ACI 318 Sections 7.10
and 7.11. Refer to
ACI 318 Chapter 21 for
special provisions for
seismic design.
Length of cage depends

on lateral design
requirements and may
not require extending
to full length. Extend
cage to full length in
highly plastic (expansive)
soils.
FIGURE 5-10 Drilled shaft reinforcing cage.
In regions of high seismic risk, special reinforcement requirements

may also apply, including confinement steel near the interface regions
with the pedestal or grade beams. Refer to ACI 318-05 Chapter 21 for
special provisions in seismic design.
Where belled bottoms are needed, the base diameter is generally
limited to less than three times the shaft diameter. The under-ream
angle is typically in the range of 45 degrees to 60 degrees, with 60
degrees often used for water tank foundations. Also, a toe height of
about 1 ft (30 cm) is maintained at the base.
Lateral and Uplift Stability

Lateral stability of the drilled pier depends on the shaft length and
flexibility. It also depends on whether the pier is drilled in cohesive or
cohesionless soils. Lateral stability of piers should be carefully exam-
ined, especially where the piers are not socketed into bedrock.
Similarly, uplift stability is a function of pier length, geometry,
and side friction. Where the uplift forces are relatively small, the re-
sistance provided by the side friction along a straight shaft pier may

Foundations
Foundations 203
be adequate for uplift stability. But when the loads are severe, belled,
or under-reamed, piers are necessary. Detailed criteria on the stability
of drilled piers and a discussion on the potential collapse of the bell
in loose soils during construction can be found in the study by Reese
and O’Neill (1988).
Socketing into Bedrock

Piers drilled into rock derive their load-bearing capacity from end
bearing and side friction offered by the length of socket in the rock.
A roughened socket length equal to 1 diameter into rock with higher
modulus than the pier shaft enables the pier to carry 50 percent of the
load by side friction, whereas a 4-diameter roughened socket length
into the bedrock allows the shaft to transfer nearly all the load by
side friction (Wyllie 1992). Hence, the depth of socketing should be a
function of design requirements and not established arbitrarily.
The socket drilled into bedrock also provides end fixity, allow-
ing the pier to develop moment resistance at its base. Geotechnical
engineers generally provide a simplified, uniform, unit side friction
value along with the end bearing for design. This information may be
presented as ultimate capacity or service load capacity. The references
previously noted provide further information on the subject.
Design Considerations in Plastic Soils

Plastic soils can be found in many parts of the world. In the United
States, Texas, Oklahoma, and the upper Missouri Valley area have
highly expansive soils. Increases in moisture cause swelling in these
clayey soils, and, as a result, foundations are subjected to rather large
uplift forces. These forces can be large enough to pull the drilled pier
out of the ground unless it has been properly designed. Similarly,
if the pier shaft is not adequately reinforced, it could break apart from
the base because of the tensile forces caused by swelling soils along
the shaft.
Piers in shrink/swell soils should terminate in bells that bear deep
in soil layers not in the zone of seasonal activity and movement. The
reinforcement cage in these belled piers should extend the full height
to allow the belled segment to anchor the uplift forces in the upper
areas of the shaft. The tensile reinforcement needed is in addition to
the reinforcement needed for normal tensile loading.
Load Testing
A clear way of establishing the structural integrity of a drilled pier is
by load testing. However, due to the high costs and logistical difficul-
ties associated with the arrangement of reaction shafts, such testing
is rarely performed. If it is absolutely necessary that a load test be

Foundations
204 Chapter Five
performed, the pier (as with piles) must be able to sustain without
excessive settlement a load that is at least twice the working load.
The best way to ensure the structural integrity and intended per-
formance of drilled piers is to follow a credible quality control program
of inspection and installation procedures. Recent studies have shown
that minor construction flaws that may not be detectable by common
nondestructive evaluation methods can lead to significant capacity
reduction in drilled piers. Such flaws include the presence of small
voids, soil inclusions, misaligned cage or other reinforcement steel,
weak concrete, or corroded reinforcing bars. Refer to Sarhan et al.
(2004) for further details.
Settlement
Settlement concerns associated with drilled piers are similar to those
defined for piles. Typically, if the drilled pier is bearing on or socketed
into the bedrock, settlement caused by direct loading is negligible.
Under other installation conditions, it is important that a proper set-
tlement analysis be performed by the geotechnical engineer to ensure
that the expected settlements are tolerable from operations and per-
formance perspectives and from the standpoint of structural design.
Reservoir and Standpipe Foundations

Reservoirs and standpipes are considered flat-bottom tanks. The de-
sign of foundations for flat-bottom tanks follows the criteria de-
fined previously for shallow and deep foundations. AWWA D100-05
provides detailed guidance on various foundation types for reservoirs
and standpipes.
The bearing pressure induced by the water at the base of a flat-
bottom tank is equivalent to the height of the high water line H times
the density of water. For a flat-bottom tank to be supported on the
ground, with or without a ringwall, the bearing soils must have an
allowable bearing capacity of at least 62.4 H lb/ft2 (9.81 H kN/m2 ).
Pile foundations may be necessary if the induced bearing is in excess
of the allowable bearing capacity of the resisting soils. Therefore, it is
important to reiterate that a formal geotechnical investigation must
be performed to verify that the bearing soils can carry the resulting
loads.
AWWA D100-05 also provides criteria for grading the interface
between the bottom of the tank and the supporting base and the use
of oiled or clean sand, crushed rock, or asphalt road mix. The standard
also provides information on granular berms, grout, foundation toler-
ances, anchor bolts, etc. Refer to Fig. 5-11 for a typical granular-berm
foundation for a flat-bottom tank (Fig. 5-3b).

Foundations
Foundations 205
Flat-bottom water-storage tank
Extend 3–5 ft (0.9–1.5 m) 2 ft (0.6 m)

beyond tank minimum
Compacted crushed stone,
screenings, fine gravel, clean
sand, or similar material Berm
Slope
To drain down
away
Coarse stone or 1
coarse gravel 1 Coarse stone or
Thoroughly compacted fill of coarse gravel
gravel, coarse sand, or other
stable material
Subgrade
FIGURE 5-11 Typical granular berm foundation.
Slab Foundation
Where the bearing soils are strong or when the water tank capacity is
small, flat-bottom tanks can be supported by a mat or slab foundation.
The slab is uniformly loaded by the pressure head in the tank. The
resulting bearing stress under the slab is the pressure due to the weight
of the tank and its contents added to the uniform pressure caused by
the thickness of the concrete slab. Overturning moment resulting from
wind or seismic loading also contributes to the bearing stress.
The reinforcement requirements of the mat or slab foundation
are based on the loading and deformation characteristics of the
footing. Often the minimum reinforcement requirement defined by
ACI 318-05 will control. Anchorage and stability requirements should
be investigated when the tank is full and when it is empty. Slab exte-
rior edges supporting the tank wall may be thickened, if necessary, to
accommodate the additional bearing stress caused by wind or seismic
overturning moments. Consideration should also be given to the frost
depth in determining slab thickness and bearing elevation.
Ringwall Foundation
As discussed in the section on shallow foundations, ringwall founda-
tions are used when the bearing pressure under the tank shell exceeds
the allowable bearing pressure of the soil near grade. The ringwall
carries the loads deeper and distributes the pressure over a wider
area. When the overturning moments are severe and anchor bolts are
required for stability, a ringwall foundation best accommodates these
anchorages. The design must consider hoop stresses caused by the
internal soil pressure resulting from the weight of the tank and its
contents.

Foundations
206 Chapter Five
Ring-Tee and Ring-Slab Foundations

When the bearing soils are weak or when the tank is large, a ring-tee or
a ring-slab foundation may be required. The foundation system is then
designed for the direct bearing loads and the overturning moments
caused by wind and seismic loading for the load combinations defined
earlier. The ringwall requires vertical and horizontal reinforcement
and full consideration of the hoop stresses. The tee or slab portion
requires radial and tangential reinforcement.
Uplift is a loading condition that must be considered for wind with
the tank empty and for seismic with the tank full. Hence, for uplift
stability, the dowels must be developed within the slab by hooks at the
ends. It is possible that piping has been routed through the ringwall. At
such locations, additional reinforcing must be provided to strengthen
the opening periphery.
Deep Foundations
Where the bearing soils are weak or the settlements are excessive,
flat-bottom tanks require deep foundations. Driven piles and auger-
cast piles are typically used under flat-bottom tanks. In regions of
high seismic risk, special reinforcement requirements also apply, as is
discussed subsequently in this chapter.
Anchor Bolts (Rods)

Load combinations governing the design of foundations for elevated
water tanks were defined in Equations (5-1) through (5-13). The load
combination causing the maximum uplift and shear in the bolt gener-
ally governs the design of the anchor bolts (more recently also referred
to as anchor rods). Because of the significant forces imposed on ele-
vated water tanks, it is critical to properly design all anchor bolts to
safely transmit these forces to the foundation.
Flat-bottom tanks may or may not require anchor bolts. All el-
evated water tanks require anchor bolts. Cast-in-place anchors are
the most common type of anchor bolt for water tank foundations, al-
though post-installed anchors have their uses. For all bolt or anchor
types, the embedment length, center-to-center spacing, edge distance,
and group action should be evaluated for the design loads, with ap-
propriate factors of safety.
Of the many references addressing structural design of anchor
bolts, Appendix D of ACI 318-05 is entirely devoted to anchors in
concrete. This reference requires anchors and anchor groups to be
designed for critical effects of factored loads as determined by elastic
analysis.

Foundations
Foundations 207
Minimum Embedment and Projection

The minimum embedment length should be such that the anchor bolt
is capable of developing the required uplift strength and shear re-
sistance for all loads in the design load combinations. Most manu-
facturers of post-installed anchors recommend an embedment length
for their proprietary anchors. However, cast-in-place anchors require
unique design based on the governing loads or as determined by
testing or past evaluations.
Bolt projection is especially important when settlements are ex-
pected to be large. In such cases, the projection should be long enough
to accommodate shimming, as required. To provide for variations in
foundation elevations, AWWA D100 further requires a projection of
the anchor bolts’ threaded ends an additional 2 in. (5 cm) beyond
the anchor nuts. Typically, a 7-in. (18-cm) projection above the top of
concrete is sufficient.
Allowable Tension and Shear

AWWA D100 recommends sizing the bolts for tension using the root
area and a basic allowable tensile stress of 15 ksi (103 MPa), with a
one-third increase for the wind load combination. The minimum bolt
diameter specified is 1.25 in. (31.8 mm), and the maximum center-to-
center spacing required is 10 ft (3 m).
For the wind load combination, tension–shear (linear) interaction
is checked using a basic allowable tensile stress of 15 ksi (103 MPa)
(as per AWWA D100-05) or 19.1 ksi (as per Table I-B in Allowable
Stress Design, American Institute of Steel Construction [AISC 1989]) for
A36 anchors in tension. For shear, AWWA D100 recommends 7.5 ksi
(51.5 MPa) for unfinished bolts, and AISC (1989) recommends 9.9 ksi
(68 MPa). An interaction value less than or equal to 1.33 renders the
design acceptable.
For the seismic load combination, AWWA D100-05, in Section
3.3.3.2, provides a higher allowable tensile stress for mild steel an-
chors based on the lesser value of 0.8Fy or 0.5Fu , where Fy and Fu
refer to the anchor bolt yield and tensile stresses, respectively. For
A36 steel, this means 28.8 ksi. For concurrent shear, the AWWA D100-
05 allowable stress of 10 ksi (i.e., 1.33 × 7.5 ksi) or the AISC allowable
shear stress of 13.2 ksi (1.33 × 9.9 ksi) is used. An interaction value
of 1.0 renders the design acceptable. Note that in this case, the ten-
sile allowable stress is increased by a different multiplier than 1.33, so
the increases are taken directly in the denominator of the interaction
equation for the seismic load combinations.
For single-pedestal and ground-supported flat-bottom tanks, the
design tensile load in the anchors is calculated from Equations (3-41)
and (3-42) in AWWA D100. For all styles of tanks, when checking
bolt interaction under seismic loads, the resistance offered by friction

Foundations
208 Chapter Five
forces may also be taken into consideration, in the authors’ view. Some
codes specifically disallow this, but for water tanks it is justified since
nearly the entire mass is considered effective in the formation of all
seismic loads, including the seismic overturning moments. This must
be done with attention to signs (load direction), since a column under
uplift cannot generate frictional resistance.
Bolt Interaction
Interaction can be checked by means of a simple equation. For the
combined effects of tension and shear, the following linear interaction
equations may be used in design:
T V
+ ≤ 1.33 (wind) (5-19)
Tallowable Vallowable
T V
+ ≤ 1. (seismic) (5-20)
Tseismic-allowable 1.33Vallowable
where
T = tensile load
V = concurrent shear load on bolt
The allowable tensile and shear stresses are as defined previously

for mild steel. For other types of steel anchors, refer to AWWA D100
or other applicable codes for all allowable stresses.
If high-strength or stainless-steel bolts are required, D100 allow-
able tensile stress for these bolts is based on the lesser of 0.4 times
the minimum published yield stress or 0.25 times the published ten-
sile strength. The calculated bolt size may need to be adjusted when
corrosion allowance is required in design. AWWA D100 discourages
the use of J and L bolts because of their tendency to straighten out, as
observed in pull-out tests.
Quality control in placement of bolts is essential. Given the size
and embedment length required, bolt relocation may not be possible,
and remedial measures can be expensive. Therefore, proper bolt place-
ment, including correct embedment and projected length, is critical to
proper design.
Bearing Stress Under Base Plates

The design bearing strength of concrete is defined in Section 10.17 of
ACI 318-05. Typically, under service load conditions, the allowable
bearing stress Fp is 0.35 fc when the entire area of concrete support is
covered (AISC 1989). Otherwise, when the supporting √ surface is wider
on all sides, the bearing stress is based on 0.35 f c A2/A1 ≤ 0.7 f c . An
additional one-third increase may be taken for wind or seismic load

Foundations
Foundations 209
combinations. Refer to ACI 318-05 for determination of A2 and A1

areas.
Foundations in Regions of High Seismic Risk

Seismic design has undergone a tremendous evolution over the past
decade. For years, seismic design of elevated water tanks was primar-
ily based on the fixed-percentage method, in which the total weight
of water and structure was multiplied by a specified coefficient based
on seismic risk zones. This was subsequently changed to the pseu-
dodynamic approach, but design was still governed by seismic risk
zones. Concurrently, some building codes required seismic design to
be performed using the velocity-based acceleration Av and the effec-
tive peak acceleration Aa .
AWWA D100-05 has entirely eliminated language regarding de-
sign on the basis of seismic zones (0, 1, 2, 3, and 4). Instead, the AWWA
standard has essentially adopted the ASCE 7-05 criteria (based on
NEHRP [2003]) with some variation with respect to the minimum de-
sign acceleration. These requirements are substantially different from
the procedure thus far used by older AWWA standards.
The International Building Code in Section 1613.1 invokes the re-
quirements of ASCE 7-05 for the design and construction of elevated
and flat-bottom water tanks to resist the effects of earthquake motions.
Special Design Provisions

As per ASCE 7-05, seismic design involves a procedure in which spec-
tral response acceleration parameters for the maximum considered
earthquake ground motions are determined from figures and then
modified for local site effects with site coefficients. The resulting ac-
celerations are then scaled down to design values. ASCE 7-05 also
permits the use of site-specific procedures in design and mandates
this procedure where provisions specifically require it.
ASCE 7-05 classifies sites based on shear wave velocity and other
features. Depending on soil consistency ranging from hard rock to stiff
soils, site classifications A, B, C, and D are defined. Site classification
E involves any profile with more than 10 ft (3 m) of soil having high
plasticity index or high moisture content, or low shear strength as
defined in the reference. Site class F involves soils that are vulnerable to
potential failure or collapse, highly organic soils, very high-plasticity
soils, and very thick, soft/medium clays. Site class F soils require site-
specific evaluations.
With the site classifications defined, the 5 percent damped design
spectral acceleration at short period SDS and at 1-second period S D1
are determined. The elevated water tank system is then assigned to
a seismic design category (SDC) based on these accelerations and on

Foundations
210 Chapter Five
the appropriate seismic use group (I, II, or III). All structures having
SDS ≥ 0.5g or SD1 ≥ 0.2g (where g is the acceleration of gravity) are as-
signed an SDC of D. Also, seismic use group III structures with 0.33g
≤ SDS < 0.5g or 0.133g ≤ SD1 < 0.2g are assigned an SDC of D. Seismic
design category E is assigned to seismic use group I and II structures
located on sites with mapped maximum considered earthquake spec-
tral acceleration at 1-second period S1 equal to or greater than 0.75g.
Similarly, seismic use group III structures at these accelerations are
assigned an SDC of F. Structures assigned to categories C, D, E, or
F require special attention to quality assurance during construction.
Structures assigned to category E or F are prohibited from being sited
where there is a known potential for an active fault.
The seismic importance factor IE significantly affects seismic de-
sign. (This factor is defined in ASCE 7-05 Section 11.5.1, Table 11.5-1.)
Values of the importance factor range from 1.0 to 1.5, depending on
the seismic use group category assigned to the elevated water tank
system. AWWA D100-05 assigns a default value of 1.5 to IE unless
otherwise specified by the purchaser, but it allows the use of 1.0 for
systems not supplying water for fire protection.
Reinforcement Criteria
In regions of high seismic risk, ACI 318-05 requires structures to com-
ply with requirements defined in Sections 21.2 through 21.10. These
sections define maximum and minimum flexural and transverse re-
inforcement, maximum spacing for hoops and crossties, bar develop-
ment length, and other requirements. Section 21.10 provides criteria
for the design of foundations. Footings, mats, piles, pile caps, piers,
and caissons are all required to be designed under this section.
ASCE 7-05 refers to ACI 318-05 for design and construction of
concrete foundations assigned to seismic design categories D, E, and
F. ASCE 7-05 requires individual pile caps or drilled piers in these
categories, as well as in category C, to be interconnected by ties. Like-
wise, spread footings founded on site class E and F soils are required
to be interconnected by ties. The design strength for ties in tension or
compression is required to be greater than 10 percent of SDS times the
larger pile cap or column factored-dead plus factored-live load, with
some exceptions. There are also rigorous requirements for the design
of piles in site class E and F soils.
Precast Prestressed and Cast-in-Place Concrete Piles

IBC-2006 provides detailed criteria for the design of foundations, piles
(including precast prestressed piles), and pile cap connections. It spec-
ifies a 28-day compressive strength f c of 5,000 psi (34.5 MPa) and
requires the prestressing strands to conform to ASTM A416. For pre-
stressed piles, IBC-2006 also specifies a minimum volumetric ratio of

Foundations
Foundations 211
spiral reinforcement, defines the ductile region of the pile as a function

of its length, and establishes bounds for the center-to-center spacing
of the spirals or hoop reinforcement. Similarly, IBC-2006 also estab-
lishes criteria for the design and detailing of cast-in-place pile and pier
foundations.
The IBC-2006 requirements are very similar to the ASCE 7-05 re-
quirements. Either reference can be used as required, individually or
in conjunction with ACI 318, in designing elevated water tank foun-
dations.
Foundation Stability
Design for stability is critical in regions of high seismic activity. Foun-
dations must be designed to withstand all design loads with adequate
factors of safety. Foundations must also be stable against all forces
causing uplift, lateral sliding, and overturning. The safety factors re-
quired for stability and strength are defined in various ways by differ-
ent codes. It is important to appreciate the reasoning and philosophy
associated with these factors to ensure structural integrity, safety, and
stability.
Lateral stability in saturated soils, settlement evaluation in satu-
rated or high-moisture-content silty soils, and potential liquefaction
in sandy soils are all conditions that require competent evaluation and
assessment before elevated water tanks are built on sites with these
characteristics. Piles designed for fixity at the pile head must be prop-
erly connected to or embedded deep into the pile cap to develop uplift
and moment capacity.
Backfill around spread footing and pile foundations must be con-
sistent with the geotechnical engineer’s recommendations. Special
recommendations made in terms of moisture content, maximum loose
lifts, or soil remediation measures must be followed. All nonconform-
ing conditions must be brought to the attention of the engineer of
record for evaluation and disposition.
Special Considerations
Design of elevated water tank foundations requires close coordination
with the project geotechnical engineer and with the construction team
at the job site. All parties must clearly understand the design require-
ments and must appreciate what is essential to quality design and
construction. Structural engineers must not assume that all geotech-
nical requirements defined in the subsurface evaluation report will
be routinely implemented during construction. In today’s fast-paced
construction, it is not unusual to see the forms removed the following
day, or long before the concrete has achieved its specified 28-day com-
pressive strength f c . Backfill placement could be started immediately

Foundations
212 Chapter Five
thereafter. Therefore, the engineer of record (EOR) must be specific

in defining any special design and construction, including formwork,
requirements.
Thus, as the concrete strength and backfill procedures are critical
to the various phases of construction, strict quality control measures
must be put in place to ensure that the correct concrete strength is
achieved before any backfill activities commence. Site inspection and
testing will be necessary to ensure proper compaction or soil remedi-
ation.
It is possible that the site soils are unsuited for backfill, and so suit-
able soils must be imported. It is also possible that topsoils containing
organics and other deleterious material could get mixed with other site
soils during backfilling around the footing. Some states have defined
certain soils or soil mixtures as “select fills” that are recommended for
specific structural fills or backfill. All of these issues are important in
foundation design, and the necessary quality control steps should be
taken before starting any construction activity.
Vertical Versus Sloped Excavations

Some foundation contractors prefer making footing excavations no
larger than is required to place the footings. Where possible, these
excavations are vertical unless they are deep enough that OSHA reg-
ulations mandate them to be sloped. Unless backfill compaction is
clearly specified and required by explicit notes on drawings, it is pos-
sible that backfill compaction in these excavations will not occur. If
compaction of soils within the 25-degree to 45-degree wedge was also
included in lateral stability consideration, those soils would not be
compacted if the excavations are vertical.
Therefore, it is crucial to clearly define all requirements for excava-
tion, backfill, backfill compaction, moisture content, and dewatering
where necessary. Curing procedures, minimum concrete strength be-
fore backfill can be placed and/or compacted, and the extent of com-
paction beyond the footings must be defined as well. These require-
ments must be delineated precisely by concise notes on the foundation
drawings.
Backfill Compaction
Geotechnical engineers generally specify compaction in terms of max-
imum thickness of loose lifts and standard proctor maximum dry den-
sity unit weights per ASTM D698 or modified proctor maximum dry
density tests per ASTM D1557 (see Annual Book of ASTM Standards).
These requirements are usually specified to be 95 to 98 percent of
the maximum dry density; even higher percentages are specified for
subgrade compaction. Soil compaction is accomplished by the use of
hand tampers and sheepsfoot or pneumatic rollers.

Foundations
Foundations 213
Backfill compaction not only improves lateral and uplift stabil-

ity by improving soil shear strength, but it also reduces permeability
in cohesive soils. In cohesionless soils, surface drainage must be ac-
complished by proper grading to avoid basin or boat effects around
foundations. A typical compaction note reads:
Backfill material should be placed in 6 to 8 in. maximum loose lifts and

compacted to at least 95% of the Standard Proctor maximum dry density
(ASTM D698).
As a rule, for water tank foundations, a minimum unit weight of

100 pcf (15.7 kN/m3 ) should be achieved. Rock fragments and stones
larger than 3 in. (7.5 cm) in diameter should not be used in the vicin-
ity of the footings. Any soft or loose material should be removed
from the bearing areas before concrete is placed. Thus, it is recom-
mended that a geotechnical representative be present before and dur-
ing the pouring of the foundations as well as during placement of
backfill.
Water- and Moisture-Control Measures

Design in areas of high water table should consider buoyant unit
weights of soil and concrete for lateral and vertical stability. If the
water table is likely to be encountered during excavation or construc-
tion, proper measures must be taken to dewater the bearing areas to
a minimum depth of 3 ft (0.9 m) below the bottom of the founda-
tion. When rainfall is imminent or when the excavation must remain
open overnight, a 4- to 6-in. (10- to 15-cm) mud mat of lean concrete
(2,000 psi [13.8 MPa]) should be poured over the bearing soils, with the
top of concrete being at the required bearing elevation. If water does
enter the excavations or if unsuitable soils are encountered, softened
soils should be completely removed and excavation brought back to
bearing grades with a mud mat or no. 57 stone. The geotechnical en-
gineer should approve this activity.
It is best to maintain soil moisture content as close to (within 2
to 3 percent of) the optimum moisture content. Plastic soils are not
ideal for fill or backfill, but when the geotechnical engineer approves
their use, they should be placed with a higher moisture content of
±5 percent of the optimum. These levels of moisture content facilitate
compaction and help accomplish the desired unit weight.
Shrink/Swell Soils
Expansive soils and challenges associated with shrink/swell soils
were discussed earlier in the chapter in the section on design con-
siderations in plastic soils. Expansive soils are generally plastic clays,
also known as fat clays, that swell with increases in moisture content.

Foundations
214 Chapter Five
U U
FIGURE 5-12 Typical foundation in shrink/swell soils. (Uw = vertical uplift due
to wind; Us/w = vertical uplift due to shrink/swell.)
They are classified as CH clays in the ASTM’s Unified Soils Classi-

fication Chart. The depth of the expansive soil active zone can vary
from a few feet or less than a meter to perhaps more than 15 ft (5 m).
Foundations constructed in these soils can potentially be subjected to
very large uplift and possibly destabilizing forces if they are not de-
signed properly. Refer to Fig. 5-12 for a typical foundation in plastic
zone with an effective active zone of 12 ft (3.7 m).

Foundations
Foundations 215
The plasticity index (PI) is generally used as a measure of swell

potential in plastic soils. Soils with low swell potential are those with
PI values below 25; a PI value of 25 to 35 indicates moderate swell po-
tential; PI values exceeding 35 correspond to soils with very high swell
potential. Foundations for elevated water tanks sited on soils with low
swell potential can be constructed using the standard practices, but
foundations sited on soils with PI values exceeding 25 (see ASCE 7-05
or IBC-2006) require precaution, remedial action, and special design
considerations (Das 2006).
Typically, when placing foundations in plastic soils, geotechnical
engineers recommend bearing the footing deep in nonplastic soils or
below the active zone. To mitigate the effects of the uplifting forces,
the use of select structural fill and backfill, soil stabilization with lime,
and/or inclusion of clearly defined uplift forces in the design have
been recommended. Das (2006) provides recommendations for reme-
dial measures as well as a procedure for estimating the uplift forces
caused by the swelling forces.
Typical remedial options often recommended for shallow footings
include the following:
1. Bear the footing below the active zone and replace the backfill
with select structural fill.
2. Bear the footing in the active zone, replace 3 to 5 ft (0.9 to 1.5 m)
of soil below the footing with select structural fill, and use
select structural fill for backfill.
3. Use site soils for backfill, but ensure that soil moisture content
is greater than the plastic limit and that the moisture content
is 3 to 5 percent above the optimum moisture.
4. Use a polyethylene or bitumen material on the vertical faces
of the footing.
5. Same as point (2), but use site soils for backfill with lime mix-
ing.
6. Same as point (1) or (2), but use site soils for backfill, and
consider the uplifting forces in the design of the footing and in
its stability. Use J-voids where necessary to allow room for the
soil to swell without imposing any forces on the foundation
element.
7. In soils where piers are recommended, geotechnical engineers
should recommend belled piers, with emphasis on the rein-
forcement requirements for resisting the uplift forces at the
junction of the shaft and the bell.
In summary, when site soils involve clays prone to swelling, spe-

cific geotechnical guidance must be sought. The geotechnical report

Foundations
216 Chapter Five
must clearly define the active zone, the potential for swell, uplift,
or adhesion forces that need to be considered in design, the bearing
depth at which the foundations are to be placed, and suitability of
site soils for backfill or recommendations for imported soil. If im-
ported soils are to be used for backfill, clear criteria must be provided
regarding the nature of the soil, its Atterberg limits, compaction re-
quirements, and guidance on local availability of the recommended
soils.
Conclusion
Foundations are critical to the design, construction, operation, and
performance of welded-steel tanks for water storage. Therefore, foun-
dation design and construction require attention to detail and proper
understanding of all criteria and requirements.
Sites that are relatively dry, level, and easily accessible and that
have good soils properties are ideal locations for erecting elevated
water tanks. The suitability of sites must always be established by
a qualified geotechnical engineer. Grade elevations and site bound-
aries must be established carefully to achieve the proper overflow and
foundation elevations.
Geotechnical investigation reports must provide all the neces-
sary information for design. This includes detailed soil properties
and other characteristics defined in this chapter. Certain soils exhibit
shrink/swell or other characteristics that require extra measures and
precautions in design. All of these characteristics must be clearly de-
fined and appropriately addressed in the report. Site classification
and settlement evaluation must also be included in the geotechnical
report.
Generally, isolated spread footings or shallow foundations are
the most economical foundation type when suitable to site condi-
tions. Otherwise, deep foundations are necessary. Detailed criteria
are provided herein to assist the designer in selecting the most suit-
able foundation type and to assist the designer with the design, be it
shallow footings or deep foundations using piles and drilled piers. In
regions of high seismic activity, special design requirements apply that
must be incorporated into the foundation design. Both the logic and
philosophy for these requirements are explained in this chapter, and
further resources are provided in the bibliography at the end of the
chapter.
Requirements for the quality control, mixing, placing, finishing,
and curing of concrete have also been defined here. These require-
ments are critical, as they govern the strength, durability, and work-
ability of concrete foundations. Also, criteria have been introduced

Foundations
Foundations 217
for sizing anchor bolts and for allowable bearing capacity under base
plates.
Requirements for backfill compaction and lateral and uplift sta-
bility are defined. It is further emphasized that to ensure safety, all
excavations must be performed in full compliance with the latest
OSHA construction standards.
Foundation Design Example

Problem Statement
To illustrate the design of a shallow foundation, the loading resulting
from the analysis of a typical 500,000-gal (1,893-m3 ) elevated torus-
bottom water tank will be considered. The tank has a diameter of
50 ft (15.24 m), head range of 37 ft (11.28 m), and a high water line of
116 ft (35.4 m). It is supported by six columns, similar to the tank shown
in Fig. 5-1. The design service loads on the footing are as follows:
Vertical Loads Horizontal Loads

Dead load D = 37.0 kip
Water load F = 520.0 kip
Snow load S = 7.0 kip
Wind load W = ±105.0 kip Wind shear WS = 30 kip
Seismic load E = ±142.0 kip Seismic shear ES = 28 kip
Assume the live load to be zero, and assume that the wind load
has been reduced by a directionality factor so that the 1.6 load factor
applies.
Use a net allowable bearing pressure of 3,000 psf (144 kPA) at a
minimum embedment depth of 5.5 ft (1.68 m) below existing grade and
a concrete compressive strength f c of 4,000 psi (27.58 MPa). Assume
the pedestal to be 4 ft × 4 ft (1.2 m × 1.2 m) with a 1-ft (0.3-m) projection
above grade. Refer to Fig. 5-13.
Footing Design
Step 1: Governing Load Combinations
The load combinations were defined in Equations (5-1) through (5-7).
A quick examination of these equations reveals that only load combi-
nations (5-1), (5-4), (5-5), and (5-6) are governing. After simplification,
these equations are as follows:

Foundations
218 Chapter Five
V
Grade Footing
exposure
Embedment depth
Pedestal width p w
d
Flexural
shear d
Punching
2 shear
t d
Footing width B
FIGURE 5-13 Shallow footing example.
Vertical Loads
U 1 = 1.4 (D + F ) = 780 kip (5-1)
U 4 = 1.2 (D + F ) + 1.6 W + 0.5 S = 840 kip (5-4)
U 5 = 1.2 (D + F ) + (1.4 × 1.0) E + 0.2 S = 869 kip (5-5)
U 6 = 0.9 D + 1.6 W = −135 kip (5-6) (uplift)
Corresponding Horizontal Loads
V1 = 0.0 (5-1)
V 4 = 1.6 W S = 1.6 × 30 = 48 kip (5-4)
V 5 = (1.4 × 1.0) E S = (1.4 × 28) = 39 kip (5-5)
V 6 = 1.6 W S = V 4 = 48 kip (5-6)
Step 2: Size Footing Using Service Loads

Corresponding to the load combinations in Step 1, as per ASCE 7-05
Section 2.4.0, the service (vertical) loads are as follows:
Service U 1 = (D + F ) = 557 kip (5-1)

Service U 4,Wind = (D + F ) + W = 662 kip (5-4) (governs wind)
Service U 5,Seismic = (D + F ) + E Service = 699 kip (5-5) (governs seismic)
Service U 6,Wind = 0.6D + W = −83 kip (5-6) (uplift)

Foundations
Foundations 219
Required footing area:
662 kips
AWind = = 220 ft2 (governs)
3000 psf
699 kips
ASeismic = = 175 ft2
1.33 × 3000 psf
AWind governs and a square footing of 15 ft × 15 ft (4.6 m × 4.6 m)

provides the required bearing area. The net bearing pressure at the
toe of the footing is
662 kips 30 kips × 6.5 ft

f bearing-Wind = 2
+
152
ft (15 × 152 /6) ft3
= 3,290 psf < 4,500 psf (using FS = 2.0)
699 kips 28 kips × 6.5 ft

f bearing-Seismic = 3
+
152
ft (15 × 152 /6) ft3
= 3,430 psf < 4,500 psf (using FS = 2.0)
Although the increase in the net overbearing pressure resulting

from the differential weight of concrete is not yet accounted for, the
initial size selected is reasonable.
Step 3: Check Punching Shear at a Distance d/2 from the Pedestal

Punching shear is checked at a distance d/2 from the face of the
pedestal as shown in Figs. 5-13 and 5-14a . The maximum punching
shear is caused by the load U 5 of 869 kip (3.87 MN) and the factored
weight of the pedestal. Assuming a depth d of 17 in. (43.2 cm) for the
slab, the critical perimeter is given by
b 0 = 4( pw + d) = 4 × [4 ft × 12 (in./ft) + 17 in.] = 260 in.
where p w is the width of the square pedestal.

For a bearing depth of 5.5 ft (1.68 m) and a slab thickness t of ap-
proximately 20 in. (50.8 cm), the pedestal height will be 4.83 ft (1.47 m).
The factored weight will be
Dpedestal = 1.2 × (4 ft × 4 ft × 4.83 ft) × 0.144 (kip / ft3 ) = 13.36 kip,
which results in a new factored U 5 of 882 kip.

Foundations
220 Chapter Five
Pedestal
d width pw1 d
2 2
2
d
width pw2
Pedestal
2
d
Critical perimeter = 4(p w + d )

(square footing)
Footing width B
(a)
Footing width B
Pedestal width pw
d Footing width B
(b)
(B – pw)
–d
2
FIGURE 5-14 Design shear and bending moment evaluations: (a) punching
shear, (b) flexural shear, and (Continued)

Foundations
Foundations 221
Footing width B
Pedestal width pw
Footing width B
(c)
(B – pw)
2
FIGURE 5-14 (Continued) (c) flexural bending.
The nominal punching shear capacity as per ACI 318-05 Section

11.12.2.1 is

Vc = 4 f c .b 0 .d = 4 × 3, 500 × 260 × 17 = 1,046 kip
Here, f c is reduced by 500 psi (3.45 MPa) for reasons described in the
section on structural concrete. Assuming no contribution from slab
reinforcement, using a shear reduction factor of 0.75 as per ACI 318-
05 Section 9.3.2.3, the nominal punching shear capacity is
Vn = 0.75 × 1,046 = 784 kip
The punching shear caused by U 5 is
882 kips
Vu = [152 ft2 − (4 + 1.42)2 ft2 ) = 767 kip
152 ft2

Foundations
222 Chapter Five
Since V n is greater than V u kip, a slab depth d of 17 in. (43.2 cm)

satisfies punching shear requirements.
Step 4: Check Flexural (Beam) Shear at a Distance d

from Pedestal
Flexural shear will be checked at a distance d from the face of the
pedestal, as shown in Figs. 5-14b and 5-15. The bearing pressures for
pw
W3 d
W1
(soil) (Concrete—pedestal)
W2 (Concrete—slab)
3,740 psf (179 kPa)
3,920 psf (188 kPa)
4,370 psf (209 kPa)

4,040 psf (193 kPa)
4,125 psf (198 kPa)
pw
} 450 psf
(22 kPa)
2
pw (B – pw)
d+
2 2
B/2
FIGURE 5-15 Flexural shear and bending moment evaluation. (psf = pounds
per square foot, kPa = kilopascal.)

Foundations
Foundations 223
the factored wind and seismic load combinations are as follows:
(840 + 13.36) kips 48 kips × 6.5 ft

f bearing-Wind = 2
± = 3.79 ± 0.55
152
ft (15 × 152 /6) ft3
= 4,347 (psf) maximum < 9, 000 psf (ultimate bearing)
= 3,238 (psf) minimum
(869 + 13.36) kips

39 kips × 6.5 ft
f bearing-Seismic = + = 3.92 ± 0.45
152 ft3 (15 × 152 /6) ft3
= 4,370 (psf) maximum < 9,000 psf (ultimate bearing)
= 3,470 (psf) minimum
The bearing stress due to seismic loading governs. From Figs.

5-14b and 5-15, the bearing pressure at a distance d from the face of
the pedestal is
(2 + 17/12)
f bd = 3, 920 + (450) = 4, 125 psf
(15/2)
The resulting flexural shear at the same location is

1 (15 − 4) 17
VFlex = (4, 125 + 4, 370)(15) − = 260 kip
2 2 12
The flexural shear capacity of the footing slab as per ACI 318-05
Section 11.3.1.1 is

Vcf = 2 3, 500(15 × 12)(17) = 362 kip
Vnf = 0.75 × 362 = 272 kip
Since Vnf > VFlex , the 17-in. (43.2-cm) depth selected is adequate
for shear.
Step 5: Determine Required Flexural Reinforcement

The bearing pressure at the face of the pedestal is (Figs. 5-14c and 5-15)
(4/2)
f bp = 3, 920 + (450) = 4,040 psf
(15/2)
The resulting ultimate bending moment at this location is

1 15 − 4 2 1
Mu = (4, 040) (15)+ (4, 370 − 4, 040)
2 2 2

2 15 − 4 2
(15)
3 2
Mu = 967 ft · kip

Foundations
224 Chapter Five
Determine the reinforcement needed to carry this bending mo-

ment. The minimum reinforcement required as per ACI 318-05 Section
10.5.1 is
(200)(15 × 12)(17)
A s, min = = 10.20 in.2
(60,000)
Try a steel area of 14.0 in.2 (PCA 1999):

14
= = 4.58 × 10−3
(15 × 12)(17)

fy
Mn = A s f y d 1 − 0.59 ·
fc

60
Mn = (0.9)(14)(60)(17) 1 − 0.59(4.58 × 10−3 )
3.5
Mn = 1,021 ft · kip
Mn > Mu
Therefore, fourteen no. 9 bars each way, 14.5 ft (4.42 m) long, will
suffice.
Step 6: Reinforcement for the Pedestal
ACI 318-05 Section 15.8.2.1 requires a minimum pier reinforcement
of 0.005Ag , where Ag is the gross area of the pier across the interface.
Thus,
Aspedestal = (0.005)(48 × 48) = 11.52 in.2
As noted in the section on design of isolated spread footing, for the

short pedestal, this reinforcement area can be reduced per industry
practice. But this is a matter of decision by the designer. If this rein-
forcement is to be maintained, sixteen no. 8 bars will provide 12.64 in.2
(81.55 cm2 ).
Under the combined action of the 135-kip uplift U 6 and the 48-kip
shear V 6 , the resulting stress in the concrete pedestal will be
135 (48 × 4.83)
ped = + = 210 psi
(4 × 4) (4 × 42 /6)
This stress is less than the concrete modulus of rupture f r where,

as per ACI 318-05 Section 9.5.2.3,
√
f r = 7.5 3500 = 444 psi
As per ACI 318-05 Section 11.5.6.1, assuming a 3-in. (7.62-cm) cover

for the pedestal dowels, since
√
0.5Vc = 0.5(0.75)[2 × 3500(48)(45)] = 96 kip > U6 = 48 kip,

Foundations
Foundations 225
no shear reinforcement is needed. However, since the dowels will be

in compression due to the other load combinations, No. 4 ties at a
spacing of 12 in. (30.5 cm) on center are recommended.
Step 7: Check Uplift Stability

AWWA D100 requires the concrete weight along with the weight of
the soils directly above the footing to be greater than the service load
uplift. The weight of the concrete (W1 + W2 = 65 kip) and the soil (W3
= 80 kip) amounts to 145 kip total, which is in excess of the 105-kip
wind uplift. Refer to Fig. 5-15. Therefore, uplift stability is maintained.
Other Steps
The anchor bolts should be designed for tension and shear interac-
tion. Lateral stability should be checked on the basis of the active and
passive pressures and the cohesion, if any, of the backfill soils. Backfill
compaction is a function of the stability requirements. A compaction
to 95 percent standard proctor maximum dry density, as discussed
in this chapter, may be recommended. The embedment depth can be
adjusted, if necessary, to achieve additional passive resistance.
Refer to Fig. 5-16 for the reinforcing details. Note that, in gen-
eral, strain compatibility must be checked to ensure that a balanced
condition prevails and that the footings are not over-reinforced. Also,
note that a nominal top mat reinforcement can be added as required
for uplift or shrinkage control.
4 ft (1.2 m)
Anchor bolt
1 ft #4 ties @ 12 in. (305 mm)

(0.3 m) centerline to centerline
5.5 ft (1.65 m)
16 #8 dowels
No. 5'S–nominal–each
way (only if required
for uplift or shrinkage
control)
(508 mm)
20 in.
14 #9 each way
15 ft (4.5 m)
FIGURE 5-16 Shallow footing design example.

Foundations
226 Chapter Five
Bibliography
American Institute of Steel Construction (AISC). 1989. Manual of Steel Con-
struction, Allowable Stress Design. 9th ed. Chicago: AISC.
American Petroleum Institute (API). 2008. Design and Construction of Large,
Welded, Low-Pressure Storage Tanks, 11th ed. API Standard 620. Washington,
D.C.: API.
American Society of Civil Engineers (ASCE). 1993a. Bearing Capacity of Soils.
Technical Engineering and Design Guides as Adopted From the U.S. Army Corps
of Engineers, No. 7. New York: ASCE Press.
American Society of Civil Engineers (ASCE). 1993b. Design of Pile Foundations.
Technical Engineering and Design Guides as Adopted From the U.S. Army Corps
of Engineers, No. 1. New York: ASCE Press.
Bowles, J. E. 1995. Foundation Analysis and Design. 5th ed. New York: McGraw-
Hill.
Das, B. M. 2006. Principles of Foundation Engineering. 6th ed. Florence, KY: CL
Engineering.
Deep Foundations Institute (DFI). 1990. Augered Cast-in-Place Piles Manual. 1st
ed. Englewood Cliffs, NJ: DFI.
Kosmatka, S. H., and W. C. Panarese. 1990. Design and Control of Concrete Mix-
tures. 13th ed. Skokie, IL: Portland Cement Association.
Liu, C., and J. B. Evett. 1987. Soils and Foundations. 2nd ed. Englewood Cliffs,
NJ: Prentice-Hall.
National Earthquake Hazards Reduction Program (NEHRP). 2003. NEHRP
Recommended Provisions for Seismic Regulations for New Buildings and Other
Structures (FEMA 450). Part 1: Provisions. Washington, D.C.: NEHRP.
Nilson, A. H., D. Darwin, and C. Dolan. 2004. Design of Concrete Structures.
New York: McGraw-Hill.
Peck, R. B., W. E. Hanson, and T. H. Thornburn. 1974. Foundation Engineering.
2nd ed. New York: John Wiley & Sons.
Portland Cement Association (PCA). 1999. Notes on ACI 318–99 Building Code
Requirements for Structural Concrete, with Design Application. 7th ed. Skokie,
IL: PCA.
Prakash, S., and H. D. Sharma. 1990. Pile Foundations in Engineering Practice.
New York: John Wiley & Sons.
Reese, L. C., and M. W. O’Neill. 1988. Drilled Shafts: Construction Procedures
and Design Methods. US Department of Transportation, Federal Highway
Administration FHWA-HI-88-042, ADSC-TL-4. McLean, VA: US Depart-
ment of Transportation Federal Highway Administration; and Dallas, TX:
ADSC, the International Association of Foundation Drilling.
Sarhan, H. A., M. W. O’Neill, and S. W. Tabsh. 2004. Structural Capacity Reduc-
tion for Drilled Shafts with Minor Flaws. ACI Structural Journal 101(3):291–
297, May/June.
Smith, G. N., and E. L. Pole. 1981. Elements of Foundation Design. New York:
Garland STPM Press.
Terzaghi, K., and R. B. Peck. 1967. Soil Mechanics in Engineering Practice. New
York: John Wiley & Sons.
Woodward, R. J., W. S. Gardner, and D. M. Greer. 1972. Drilled Pier Foundations.
New York: McGraw-Hill.
Wyllie, D. C. 1992. Foundations on Rock. New York: E. and F. N. Spon, an imprint
of Chapman and Hall.

CHAPTER 6
Construction of
Welded-Steel
Water-Storage
Tanks
Jim Noren, P.E
Advance Tank Construction
Donita Fredricks, P.E.

CB&I Constructors
Steel Fabrication
Tank constructors have developed specialized equipment, tools, and
procedures for the construction of ground storage tanks and elevated
water tanks. In the construction of nearly all steel-welded tanks, the
steel components are fabricated in a shop environment and shipped
to the tank site, where the fabricated components are fit and welded
into the finished tank by field construction crews. Steel plate layouts
are developed by the constructor, which minimizes welding and max-
imizes the use of the ordered plate, with consideration to the size and
weight restrictions for shipping. For the composite elevated tank, spe-
cialized forms and equipment have been developed and are used in
the construction of the concrete components.
Material
Plate material may be purchased from a steel warehouse or directly
from a steel mill. Steel warehouses stock plate material in most of the
227
Construction of Welded-Steel Water-Storage Tanks
228 Chapter Six
grades used in welded-steel tanks. Material that is required to con-

form to supplementary requirements (e.g., silicon-killed, fine-grain
practice, normalized, or ultrasonically inspected material) is gener-
ally not available from a warehouse.
Material Purchased from Steel Warehouse

Steel plate from a warehouse is typically available in standard plate
widths ranging from 48 to 96 in. (1.2 to 2.4 m) in 12-in. (0.3-m) in-
crements. Plates are normally stocked in 20-ft (8.8-m) lengths, though
some warehouses have the capability to cut coiled plate to length.
Stock plates commonly used in welded-steel water-storage tanks are
available in the following thicknesses: 3/16 in. (4.8 mm), 1/4 in.
(6.3 mm), 3/8 in. (9.5 mm), and in 1/4-in. (6.3-mm) increments for
thicknesses between 1/2 in. (12.7 mm) and 2 in. (51 mm). Stock mate-
rial grades, sizes, and thicknesses vary from warehouse to warehouse,
so availability of a specific plate size and thickness must be verified in
the design phase of the project. Delivery time for warehouse material
is shorter and minimum tonnages normally do not apply, but the cost
is higher than for material purchased directly from the mill.
Material Purchased from a Steel Mill

Plate material purchased from a mill may be ordered to the customer-
specified width, length, and thickness. Material conforming to specific
supplementary requirements is available from most mills. Steel mills
typically require a minimum order, and delivery times are signifi-
cantly longer than for warehouse-purchased material.
Regardless of whether a plate is purchased from the warehouse or
mill, conformance to the American Society for Testing and Materials
(ASTM) requirements for the ordered plate should be confirmed by
the constructor on receipt of the material. This can be accomplished by
reviewing material test reports or certificates of compliance furnished
by the supplier. If the plate cannot be traced to a material test report
or a certificate of compliance, testing by a qualified testing laboratory
may be used to verify that the plate complies with the chemical and
mechanical requirements of the specified ASTM standard.
On receipt of the plate, measurements should be taken to verify
that its width, length, and thickness are consistent with the ordered
plate size. Permitted variations in dimensions are outlined in ASTM
A6. Visual examination of the plate should be performed to verify that
the material is free from injurious defects and has a workable finish.
Thick plate should also be checked along the edge for lines that would
indicate a possible lamination.
Cutting
Several methods are available for cutting plates to size in the shop,
including thermal cutting by either oxy-fuel gas torches or plasma

C o n s t r u c t i o n o f We l d e d - S t e e l Wa t e r - S t o r a g e Ta n k s 229
FIGURE 6-1 Cutting plate by oxy-fuel torches.
arc. Thermal cutting methods are commonly used in the shop for
irregularly shaped plates (Fig. 6-1). Oxy-fuel gas torches may be au-
tomated by either setting up a track burner or by use in a numerically
controlled burning bed. Plasma arc cutters are used in a numerically
controlled burning bed. Using either method, the plate may be simul-
taneously cut to size and the edges tapered and beveled for welding
by using multiple burning heads. The finished edges should closely
follow the detailed plate dimensions to ensure good fit-up in the field.
The edges should be uniform and smooth and cleaned of slag accu-
mulation when necessary.
Machining and shearing are other methods of cutting plates to
size. Rectangular plates that are ordered with minimal trim allowance
may be trimmed and squared by machining the edges using an edge
planer. Shearing is another option for straight edges that are less than
the width of the constructor’s plate shear. American Water Works As-
sociation (AWWA) Standard D100 limits plate thicknesses for shearing
to 1/2 in. (12.7 mm) or less if the joint is to be butt welded. Edges that
will be lap welded are limited in thickness only by the capacity of
the plate shear. Sheared edges should be square and burrs removed
before welding.
Forming
Single-curvature plates for welded-steel tanks are typically cold rolled
in the shop to the appropriate curvature using a plate roll. Plate widths

230 Chapter Six
Plate Thickness
Minimum Maximum
(in.[mm]) (in.[mm]) Minimum Plate Diameter (ft[m])
>3/8 [>9.5] ≤3/8[≤9.5] 40 [12]
>1/2 [>12.7] 1/2 [12.7] 60 [18.2]
>5/8 [>15.8] 5/8 [15.8] 120 [36.6]
Must be rolled for all diameters
TABLE 6-1 Minimum Diameter for Plates Not Rolled
may be limited by the fabricator’s plate roll capacity. AWWA D100

makes provisions for plates that need not be rolled on the basis of the
minimum diameter and plate thickness as outlined in Table 6-1. Single-
curvature plates are frequently used to construct a double-curvature
surface if the radius is large enough. One example of this is pie-shaped
plates in a dome roof.
Double-curvature plates are cold pressed using repeated blows
with a mortar-and-pestle-shaped die (Fig. 6-2). Typical examples of a
double-curvature plate include the flare and ball of a pedestal tank.
Press breaks are used to form sharp bends in a plate—for example,
the fluted plate in a fluted-column-style tank. Press breaks and presses
can also be used to simulate a rolled plate by repeatedly hitting
the plate with a straight die, allowing for short spaces between hits
(Fig. 6-3). This method can also be used to form cone-shaped plates
and is particularly useful for thick plate.
Angle rolls are commonly used to roll structural angles for welded-
steel tanks. With all forming operations, it is important to have
FIGURE 6-2 Pressing double-curvature plate.

FIGURE 6-3 Forming fluted plates in a press break.
adequate dimensions on the shop drawings to verify the accuracy

of the formed plate.
Shop Subassemblies
Fabricated plate may be subassembled and welded in the shop
(Fig. 6-4). This is done to maximize the welding that can be performed
FIGURE 6-4 Shop assembly of cut and formed plates for a dome roof.

232 Chapter Six
in the controlled environment of the shop using optimal welding pro-

cesses and configurations. Shipping restrictions dictate the extent of
shop subassembly that can be executed.
When plate sections are sized such that the shop-welded sub-
assemblies are not an option, complete or partial shop assembly may
still be warranted to ensure proper fit-up in the field. This is especially
useful for complex geometries with double-curved surfaces.
Blast and Prime

The life of a coating system depends on the surface preparation. A
smooth, regular surface with the proper steel profile will provide a
good basis for the application of a protective coating system. The
fabricator should ensure that weld contours are smooth and that any
unacceptable weld undercutting is eliminated. Weld flux and weld
spatter should be removed and the sharp edges ground smooth.
Most tank constructors recommend that tank components be abra-
sive blast cleaned and primed in the climate-controlled atmosphere of
the shop. Exterior surfaces and interior dry surfaces should be cleaned
to a commercial finish per Society for Protective Coatings (SSPC) SP6
as a minimum. Inside wet surfaces should be cleaned to a Near White
Blast finish per SSPC SP10. Blast-cleaned surfaces should have a sur-
face profile that is appropriate for the specified primer and coating
system per recommendations of the coating manufacturer.
Blasting may reveal small laminations or pitting in the plate sur-
face not previously apparent. If these imperfections are large enough
to produce holidays in the coating system, they should be removed
by grinding. Occasionally, deeper laminations may require welding
or further testing.
The prime coat should be applied immediately after surface prepa-
ration, before the occurrence of any surface rusting or accumulation of
dust or moisture. The type and thickness of primer should be defined
in the customer’s specifications. AWWA D102 Coating Steel Water-
Storage Tanks may be referenced for interior and exterior coating sys-
tems. Prime coats may be applied using any method recommended
by the coating manufacturer, except that rolling should not be used for
the prime coat on interior wetted surfaces unless required for rough
pitted surfaces. An unpainted margin approximately 4 in. (102 mm)
wide should be provided around all plate edges that will be field
welded.
Shipping
Shipping from the shop to the job site is almost exclusively by truck
(Fig. 6-5). The current weight limit is 80,000 lb (36,287 kg) gross for
the truck, trailer, and load, resulting in a net load capacity of approx-
imately 45,000 lb (20,412 kg), depending on the weight of the truck

FIGURE 6-5 Shipping formed plates by truck.
and trailer. Rules for oversized loads vary from state to state. Special
permitting, routing, and escorts may be an option when oversized
loads cannot be avoided or are deemed to be more economical. Plate
layouts are often dictated by shipping limitations. Site access should
also be considered in planning shipping loads.
Material should be sufficiently blocked, braced, and tied down to
secure the components to the trailer and maintain the fabricated shape
during shipping.
Welding
In the 1950s, welding replaced riveting as means of connecting tank
joints. Welding can be performed in all climates and in a variety of po-
sitions. Over time, the technology has improved, leading to increases
in productivity. To convey the correct welding information from the
tank designer, weld symbols in accordance with AWS Standard Sym-
bols for Welding, Brazing, and Nondestructive Examination should
be used on the fabrication and erection drawings.
Welding Processes
The primary welding processes used in the shop and field are shielded
metal arc welding (SMAW), submerged arc welding (SAW), and flux
cored arc welding (FCAW). All are arc welding processes that use
an electric arc generated by an electric current between the tip of the
electrode and the base metal. Heat from the arc melts the electrode

234 Chapter Six
and adjacent base metal which then combines, cools, and solidifies to
form the weld bead.
Welding may be performed manually, semiautomatically, auto-
matically, or by machine welding. Manual welding requires the welder
to manually maintain the proper positioning and arc length and re-
place the electrode as it is consumed. Semiautomatic welding is per-
formed with a handheld gun that continuously feeds the electrode
and flux. Automatic welding is accomplished with equipment that is
capable of performing the welding operation without a welding op-
erator. This type of welding is more commonly used in assembly line
operations. In machine welding, specialized equipment performs the
complete welding operation; however, the welding equipment must
be monitored by a qualified operator who is responsible for position-
ing the steel components, starting and stopping the weld, setting the
speed, and adjusting the controls.
Shielded Metal Arc Welding

SMAW, also referred to as stick welding, utilizes a stick electrode—
typically 9 to 18 in. (229 to 457 mm) long—a solid metal wire core that
conducts electric current and provides filler metal for the joint. The
metal core is coated with a material that provides arc stability and a
shielding gas or a flux coating as the electrode is consumed. Shielding
gases are needed to eliminate oxygen from the molten weld metal.
The fluxing agents allow the molten metal to wet the surfaces of the
base metal and remove impurities from the weld metal.
SMAW is one of the most versatile weld processes and is widely
used in tank construction. The equipment is relatively simple and
portable. SMAW can be used in a wide range of positions and in areas
with limited access. It is also less sensitive to wind than either SAW or
FCAW. Shielded metal arc welding is limited to manual welding and,
consequently, has one of the lowest deposit rates. The electrodes are
relatively short and frequent stops are required to replace them. When
present, slag must be removed before restarting. As a result, SMAW
is the least efficient welding process for long production welds.
Submerged Arc Welding

The electrode for SAW is a continuous bare wire inserted into a wire-
feeding mechanism that automatically feeds the electrode toward the
joint at a controlled rate. The weld is submerged in a blanket of gran-
ular flux that is continuously deposited ahead of and around the elec-
trode. During welding, some of the granular flux is melted and serves
the same purpose as the electrode coating in SMAW welding. This
weld process can be used in a semiautomatic, automatic, or machine
mode.
SAW has one of the highest deposit rates because of the continuous
wire feed. The process is limited by joint position and accessibility.

SAW is typically limited to the flat position for butt welds and the
flat and horizontal positions for fillet welds. With specialized equip-
ment to contain the flux, SAW may be used for lap and butt joints
in the horizontal position. Moving the bulky wire-feeding mecha-
nism for the continuous-feed electrode may make SAW a less desir-
able option for inaccessible areas. Good joint fit-up is also critical for
SAW.
Flux Cored Arc Welding

Like SAW, flux cored arc welding is a continuous-feed wire welding
process. The electrode has a center core of flux encased in a tubular
metal sheath. Two types of FCAW exist: gas shielded and self shielded
(Fig. 6-6). Gas-shielded FCAW uses a gas envelope, usually CO2 or
argon/CO2 , to protect the molten metal from the air. This method is
not suitable for use when the weld cannot be protected from the wind.
For self-shielded FCAW, shielding is provided by gas emitted by the
flux as it vaporizes and by a slag blanket that covers the molten metal.
Self-shielded FCAW is no more sensitive to wind than SMAW, so it is
commonly used in the field.
FCAW is a semiautomatic process in which a handheld weld gun
is used. Deposition rates are typically higher than for SMAW but lower
than for SAW. FCAW is versatile, in that it can be used in all positions
for all the basic joint types. Like SAW, FCAW requires moving the
wire-feeding mechanism; consequently, it may not be the best option
for inaccessible areas in the field.
FIGURE 6-6 Shop flux cored arc welding.

236 Chapter Six
Weld Procedure Specification and Procedure

Qualification Record
Constructors are required to develop weld procedure specifications
(WPS) that define the welding parameters to be used in the fabrication
and erection of the tank. Each WPS must be qualified in accordance
with the rules in American Society of Mechanical Engineers (ASME)
Boiler and Pressure Vessel Code, Section IX, or American National
Standards Institute (ANSI)/AWS B2.1, Standard for Welding Proce-
dure and Performance Qualification. To qualify a WPS, the constructor
welds test coupons and tests the specimens per ASME Section IX or
ANSI/AWS B2.1. The weld parameters and test results are recorded
in a document known as a procedure qualification record (PQR). The
constructor is required to certify that he or she has qualified each WPS
with a PQR. Constructors maintain standard WPSs and PQRs for weld
parameters that are routinely used.
The constructor may elect to use an ANSI/AWS standard welding
procedure to justify a weld in lieu of performing an independent qual-
ification. If this option is selected, it is necessary to comply with all
the rules in AWS B2.1 that govern the use of the ANSI/AWS standard
welding procedure.
Welder Qualification and Identification

Welders in both the shop and the field are required to demonstrate
their ability to perform acceptable welds. Weld testing shall be in ac-
cordance with ASME Boiler and Pressure Vessel Code, Section IX, or
American Welding Society (AWS) B2.1 Standard for Welding Proce-
dure and Performance Qualification. The tank constructor is respon-
sible for testing all welders for the specific weld processes that the
individual welder will use. Records of the testing dates and test re-
sults must be maintained by the tank builder.
Each qualified welder is assigned a number, letter, or symbol that
is stamped on the tank to identify the weld operator employed for each
joint. The stamp is placed adjacent to and at intervals not exceeding
3 ft (0.9 m) along the weld. Alternately, the tank constructor may keep
a written record of the welders employed on each joint and omit the
stamping. This record must be certified by the tank constructor and
included in the inspection report when specified by the purchaser.
Grinding
Some grinding of welds may be required to ensure that the finished
weld contour is suitable for cleaning and painting and will not be detri-
mental to the life of the coating. Grinding should be used to remove
weld slag, weld spatter, burrs, and any sharp surfaces along welds.
If the purchaser requires special grinding, it should be noted in the

contract specifications, and a recognized standard addressing weld

profiles should be referenced to clearly define the extent of grinding
required.
When lapped plates are joined with fillet welds that are less than
the full thickness of the plate, the exposed sharp edge of the plate
should be removed by grinding. This will minimize the potential for
a paint failure at the edge of the plate.
Construction
Scheduling Issues
In addition to the production capacity and workload of the tank con-
structor, construction schedules are affected by the style and size of the
tank, the availability of material, the time of year, and daily weather
conditions. The purchaser may specify either the number of weeks
to completion after award of contract or a set calendar date for com-
pletion, or the purchaser may allow the tank builder to propose a
scheduled completion. If a specified completion date is critical to the
owner, the purchaser may specify liquidated damages to be charged
on a daily basis if the work is not completed on time.
If the purchaser chooses to set a construction schedule, consider-
ation should be given to mill delivery time to allow the constructor to
use material from the mill. In northern climates, a schedule should be
specified that allows painting to be performed at an appropriate time
of year. Tanks with very short construction schedules are typically
purchased at a premium.
Weather can significantly influence the field schedule. Wind, ex-
treme temperatures, rain, snow, and sleet can affect what work can
be performed and how productive the crew is. Even moderate winds
can make it unsafe to pick up and place steel plate.
Weather and temperature conditions must also be appropriate for
welding. Welding is not permitted when the parts to be welded are
wet from rain, snow, or ice, or during periods of high wind, unless the
welder and the work are properly protected. No welding is allowed
when steel is wet. The protection is typically an enclosure to block the
wind. Welding is not allowed if the base metal temperature is lower
than 32◦ F (0◦ C) unless the base metal is preheated to at least 100◦ F
(38◦ C) through the thickness and maintained for a distance along the
weld of four times the thickness of the parts to be welded. If base metal
temperatures fall below 0◦ F (−18◦ C), welding is not recommended.
If welding is performed, low-hydrogen electrodes or low-hydrogen
processes must be used, and the base metal must be preheated to 200◦ F
(93◦ C) in accordance with AWWA D100.

238 Chapter Six
Unless protection is provided, concrete should not be placed in

rain, sleet, snow, or extreme temperatures. American Concrete Insti-
tute (ACI) 306.1 and 305.1 provide guidelines for concrete placed in
cold and hot weather.
Site Issues
The engineer’s drawings typically show the site layout superimposed
on a topographic drawing. This gives some indication of the relative
slope of the site and of potential access problems. If, during the bid
stage, questions arise regarding the site, a site visit may be warranted.
Access
The owner should provide a suitable right-of-way for access from
the nearest public road to the tank site. The access should be able
to handle a semitrailer tractor rig with a trailer that is 53 ft (16 m)
long and that weighs 80,000 lb (36,287 kg) under ordinary weather
conditions. Side clearance needs to be adequate to accommodate the
maximum shipping width for the job. The access road should be free
from underground and overhead obstructions that could be damaged
by the truck traffic. A minimum vertical clearance of 14 ft (4.2 m) is
required.
Site Size
Final property lines should be located sufficiently far from the founda-
tion footprint to permit construction operations. During construction,
additional clearance is required for steel delivery, storage, staging, and
subassembly. If a permanent site of adequate size is not available, the
owner should provide an adequate temporary construction easement.
As a minimum for the construction operations, a site clearance from
the center of an elevated tank to the site limits should be equal to the
height of the tank. For a ground tank, it is preferable to have at least
20 ft (6 m) clear around the entire tank so that a crane can be used
around the full circumference of the tank.
The site should also be big enough to permit abrasive blasting
and painting without impacting neighboring property, both after ini-
tial construction and during future recoating operations. Clearance
requirements between the tank and the neighboring property vary
with the prevailing wind conditions, type of paint application, and
consequence of damage. Sites should be evaluated on a case-by-case
basis, but as a general rule, a clearance of approximately 100 yd
(91.5 m) is suggested. If adequate clearance cannot be provided, it
may be necessary to shroud the tank during initial and future paint-
ing operations. Shrouding the tank is costly and should be avoided if
possible.

Drainage
The tank site should have good drainage during construction. Storage,
staging, and subassembly areas should be free from standing water.
For sites with poorly draining soils, the bearing surface for a shal-
low foundation should be protected from becoming saturated prior
to concrete placement.
Power Lines
Overhead or buried power lines present a significant safety risk for
tank construction. Sites having power lines within 40 ft (12 m) of the
tank or tank foundation are unacceptable.
Security
Access to the tank should be blocked when the tank is left unattended.
If the location is remote or subject to frequent vandalism, additional
measures to ensure site security may be required. The additional mea-
sures may include fencing and full- or part-time security.
Power During Construction

Power Supply
The purchaser should indicate whether electrical power is available
at the site. If power is available, the purchaser should indicate the
voltage and whether it is direct current or alternating current (if al-
ternating current, what cycle and phase). The specifications should
define who will furnish the power to the site and who is responsible
for the associated costs.
Tank constructors frequently provide their own power supply in
the form of generators.
Power Requirements
Power requirements in the field vary depending on the types of tools
that will be used on the job. If the purchaser is furnishing power to
the site, the power requirements should be coordinated with the tank
constructor.
Construction of Welded-Steel Ground Water-Storage Tanks

Anchorage and Grout
If the tank is not anchored, the interface between the tank bottom and
the concrete foundation can be either grouted or filled with the place-
ment of asphalt-impregnated cane fiberboard. If the tank is anchored,
the use of grout is recommended, since the fiberboard may deteriorate
over time. This deterioration would cause vertical movement of the
tank, which would require subsequent tightening of the anchor bolts.

240 Chapter Six
Tank-Bottom Construction
The bottom of a ground-supported reservoir or standpipe is essen-
tially a nonstressed membrane, the purpose of which is to contain the
product inside the tank and transmit the water-bearing load directly
to the foundation. The minimum thickness of the bottom plate is 1/4
in. (6.3 mm); it may be thicker if a corrosion allowance is specified.
The tank bottom should be crowned up from the shell to the cen-
ter with a minimum slope of 1 in. (25.4 mm) vertical to 10 ft (3 m)
horizontal.
Layout The typical plate layout for bottom plates is a “rect-and-

sketch” layout, which refers to rectangular plates with sketch plates at
the outside cut to a radius. The outside radius must be a minimum of
2 in. (50 mm) outside of the shell. Additional projection may be pro-
vided to compensate for shell out-of-roundness and weld shrinkage in
the bottom welds. With a lap-welded bottom, there will be three plate
laps at the corners of the rectangular plates. Three plate laps must be
at least 1 ft (0.3 m) from the shell. Refer to Fig. 6-7 for an example of a
rect-and-sketch bottom layout.
Annular ring An annular ring may be required under two possible

scenarios. First, if the shell uplift due to seismic overturning is large, a
thickened annular ring will increase the uplift resistance of the shell.
This strategy is used to eliminate tank anchorage. For this circum-
stance, butt-welded sketch plates may be substituted for an annular
ius 1 ft. (0.3 m)

rad
Cut minimum
1.5 in. (38 mm) typical

(1 in. [25.4 mm] minimum)
)
.3 m
um
1 ft. (0
minim
Inside shell
FIGURE 6-7 Typical rect-and-sketch layout for bottom plates.

)
6m
. (0.
2 ft nimum ll
mi e she
d
insi
1.5 in. (38 mm) typical

Shell (1 in. [25.4 mm] minimum)
)
m n
1 m ctio
(5 o je
n. pr
2 i ide
ts
ou
Annular plate
FIGURE 6-8 Tank bottom with annular ring.
ring. Second, if the tank is designed in accordance with AWWA D100

Section 14 and is greater than 150 ft (45.7 m) in diameter, an annular
ring is required.
If an annular ring is provided, the minimum inside projection
shall be 2 ft (0.6 m) or the minimum width required for seismic uplift
resistance, whichever is less. The bottom plate is lap welded to the
annular ring and has a rect-and-sketch layout. Refer to Fig. 6-8 for an
example of a tank bottom with an annular ring.
Welding sequence The welding sequence for the bottom plates shall
minimize out-of-plane distortion. A general sequence for bottom
welding is described as follows:
r Weld the sketch plate to sketch plate joints.

r Weld the rectangular short side joints.
r Weld the rectangular long side joints.
r Weld the rectangular plate to sketch plate welds.
Lap welded versus butt welded Bottom plates can be welded by either
lap welds or butt welds. For bottom plate thicknesses up to 3/8 in.
(9.5 mm), the plates are typically lap welded from the top side only.
The weld is a full-thickness fillet weld and, for thicknesses equal to
or greater than 5/16 in. (8 mm), the fillet is typically a two-pass weld.
If it is necessary to seal the underside of the bottom or if the bottom

242 Chapter Six
FIGURE 6-9 Shell plate

Breakdown of
lapped area
beneath shell and
projected outside
of tank.
Bottom plate
plate thickness is greater than 3/8 in. (9.5 mm), butt welding of the
bottom plates is appropriate.
For lap-welded bottoms, the lapped area that is beneath the shell
and projected outside of the tank must be “broken down.” The pur-
pose of the breakdown is to provide a smooth transition at the lap on
the top side so that there will be no gap at the shell-to-bottom connec-
tion in the region of the lap. Refer to Fig. 6-9 for an illustration of a
breakdown.
Annular ring splice welds must be butt welded. The welds may
be either single butt welds with backup bars or double butt welds.
Commonly, the fabrication shop will subassemble annular ring sec-
tions with double butt welds, and the subassemblies will be welded
in the field with single butt welds.
Butt-welded bottom plates can be either a one-sided weld with
a backup bar or a double butt welded. Single butt welding is the
preferred method since the bottom can be laid out and the welding
performed from the top side. Double butt welding is difficult for large
tanks due to the inaccessibility of the underside of the bottom. For
small tanks, however, the initial weld pass can be performed down-
hand and the bottom can be flipped over so that the welding can be
completed down-hand.
Shell-to-bottom junction The connection of the shell to the bottom

plate shall be a continuous fillet weld on both sides of the shell. Table
18 in AWWA D100-05 gives the minimum size of the fillet welds to be
used on the basis of the thickness of the shell plate. If the fillet weld is
5/16 in. (7.9 mm) or larger, the weld shall be two-pass minimum.
The weld should be inspected for watertightness using dye pene-
trant, penetrating oil, or diesel fuel. The inside fillet weld is completed
first, and indicator is sprayed on the weld. If any indicator is visible
outside of the shell after a wait period, a leak is indicated and should
be repaired. Once there are no indications of leakage, the outside weld
can be completed.

Shell Construction
Crane versus jacking Normally, there are two basic methods for erect-
ing the tank shell. The more common method is to start from the
bottom ring and use a crane to place each individual shell plate for
each successive ring until the shell is complete. However, for tall tanks
a method using hydraulic jacks may be more economical. Using this
method, the top two shell rings and roof are erected on temporary jack
stands. Once this is complete, the shell is jacked using hydraulics, and
the next shell ring is placed. This process is repeated until the shell is
complete. This method reduces the crane requirements since there are
no high picks. Also, temporary scaffolding for the shell is not required
since all shell erection and welding activities are performed at ground
level.
The shell is the critical component of a storage tank. It is the pri-
mary stressed membrane that contains the liquid. Therefore, great care
must be exercised in laying out, fitting, and welding the shell.
Listed here are general steps for layout and fit-up of a tank shell.
This procedure varies among contractors; however, the general steps
are the same.
r Check the elevation of the tank bottom.

r Scribe the radius of the inside of the shell on the bottom. This
will be used as a guide for setting the first ring.
r Mark the chord dimensions for the first ring.
r Install erection nuts or lugs on the bottom that follow the
outside radius of the first ring.
r Set the first plate starting at the first chord mark and following
the circular scribe mark.
r Set the second plate and use fit-up gear to align the vertical
seam for welding.
r Set the remainder of the first ring plates and check for level.
r Weld the vertical seams.
r Weld the first ring to the bottom.
r Hang the second and subsequent rings using fit-up gear on
the horizontal and vertical seams.
Shimming and the tub ring After the bottom plate is laid, the first shell
ring, or “tub ring,” is set. Care must be taken in setting the tub ring,
since the roundness of the tub ring is the basis of whether the rest of the
shell will be round. An essential part of making the shell round is that
the tub ring must be level. If the tub ring is not level, tank roundness
is difficult to achieve. The tub ring is leveled by using shims between
the bottom and the foundation.

244 Chapter Six
Wind stability During shell erection, the shell is susceptible to wind

loads. The shell is designed to be stable and resist buckling due to
wind when the tank is complete. However, during construction the
shell can buckle easily, even during moderate winds. Therefore, the
shell must be braced to prevent a “blow-in” incident. One method of
bracing the shell is with the erection scaffolding. The scaffolding is
normally set 3 to 4 ft (0.9 to 1.2 m) below the top of the ring being
erected and consists of brackets and scaffold boards or planks span-
ning between the brackets. If the boards overlap at the brackets and
are tied down securely, the scaffolding itself acts as a ring stiffener
on the shell. Because of this phenomenon, incomplete scaffolding is
normally not allowed to be left overnight.
Partial- versus complete-penetration welding The shell vertical welds

are always complete-penetration welds. The horizontal welds may be
either complete-penetration or partial-penetration welds. When the
shell thickness of the thinner of the two plates being joined is greater
than 3/8 in. (9.5 mm), the horizontal weld can be a partial penetration.
The finished weld must have at least two-thirds the strength of a
complete-penetration weld. Partial-penetration welds are not allowed
in the shell plates for Section 14 designs.
Weld clearances Weld clearances for shell vertical joint offset, perma-
nent attachments, and shell penetrations should meet the require-
ments of AWWA D100 and good industry practice. Section 14 of
AWWA D100 prescribes the requirements for weld clearances. The
base code does not describe any weld clearance requirements; how-
ever, good practice indicates the following weld clearances:
r Vertical shell plate offset = 12 in. (305 mm) minimum

r Permanent attachments = 3 in. (76 mm) (horizontal) and 6 in.
(152 mm) (vertical)
r Shell penetrations = 3 in. (76 mm)
Construction openings A construction opening in the shell is normally

provided to allow easy access to the interior of the tank. This opening
is usually in the form of a short plate that is removed in the first shell
ring. The short plate left out of the shell, called a door sheet, is 6 to
12 ft (1.5 to 3.6 m) wide and has a height equal to the width of the first
shell ring. If a crane must be driven inside of the tank for roof erec-
tion, a taller opening is usually required. This is accomplished with
a first and second ring door sheet, which may or may not include
the full height of the second ring. Temporary stiffening must be pro-
vided around the door sheet to bridge the vertical loads during tank
erection.

Shell penetrations Shell penetrations are required for access to the

interior of the tank and for an overflow to prevent overfilling. The
inlet and outlet piping may go through a shell penetration but more
commonly goes through the tank bottom. A minimum of two man-
holes are required for access. The penetrations have a neck plate that is
welded to the shell, and additional reinforcing in the form of a circular
reinforcement plate may be required.
Post-weld heat treatment Post-weld heat treatment is only required

for shell penetrations that are 12 in. (305 mm) in diameter or greater in
shell plate that is thicker than 1 in. (25.4 mm). The penetrations should
be prefabricated in the shell and stress relieved before shipment. This
requirement applies only to Section 14 designs.
Roof Construction
Roof configuration can be either supported on structural framing or
self-supporting. The self-supporting roofs can be unstiffened, or stiff-
eners can be welded to the roof plate.
Method of roof erection—crane versus air raised For tanks with struc-
turally supported roofs, the typical method of construction is to use
a crane to lift the various components into place. The roof framing is
erected after the bottom and shell are in place, and the roof plate is
placed after the framing is complete.
For tanks with self-supporting roofs, there are more options. The
roof can be built in place using a crane and temporary support for
the roof or the roof can be built on temporary supports outside of the
tank and the entire roof can be lifted into place. The latter method
is advantageous for tall tanks, roofs requiring seal welding, and in
situations where a crane with enough capacity can be used econom-
ically. Another option is to have the roof erected on the floor of the
tank and to lift it into place after the shell erection is complete. This is
accomplished by sealing the outside edge of the roof to the shell and
pressurizing the underside of the roof to lift it to its final position. This
method is economical for large-diameter tanks that are relatively tall.
Surprisingly, the pressure required to air-raise a roof is on the order
of a water column of 3 to 6 in. (76 to 152 mm). The roof can be raised
using high-velocity fans bolted to the shell manholes.
Subassembly For self-supported roofs, to minimize the number of

crane picks and reduce the need for welding in place, the field crew
may elect to subassemble some of the roof sections. This may also
decrease the amount of time the crane needs to be on-site, therefore
reducing costs.
Roof-to-shell junction The roof-to-shell junction can be configured in

several ways. For cone- and dome-type roofs, an angle can be either

246 Chapter Six
butt or lap welded to the top of the shell. The roof plate laps onto
the top of the angle. This arrangement can be advantageous since
fabrication and erection variances can be tolerated. The top angle is
used to aid in keeping the tank shell round. As an alternative to using
an angle, a bar may be used.
A double-curved transition may be used for either a supported or
unsupported roof. This type of transition can be more visually pleasing
for taller tanks and any tank for which aesthetics are important.
Seal welding Seal welding may be specified to reduce rust bleeding

from the inaccessible plate lap areas. If seal welding is required, the
type of roof must be considered. If the roof plate is supported on struc-
tural members, the surfaces mating the tops of the structural mem-
bers to the underside of the roof plate will also be inaccessible. Seal
welding the roof plate lap welds will solve only part of the problem.
However, if the roof framing is to be seal welded to the roof plate,
the tank designer must consider the effects of thermal expansion and
contraction caused by temperature differentials on the roof framing.
Also, the amount of welding required to seal all of the roof framing
might induce additional weld distortion in the roof plate. If the roof
is self-supported, any framing will be welded to the roof plates by
design.
Ponding For supported cone roofs, the minimum roof slope is a 3/4-
in. (19-mm) rise in a 12-in. (305-mm) run. This is a very shallow roof
slope and it may therefore produce ponding if the roof plate is built
with excessive distortion. In many local jurisdictions, ponding is not
allowed by law. The easiest solution to potential ponding is to increase
the roof slope.
Construction of Elevated Steel Water-Storage Tanks

Method of Erection—Crane Versus Derrick
The constructor should select the type of crane to be used to construct
an elevated water tank, basing the decision on the tank geometry,
schedule, equipment availability, and cost.
Elevated tanks are frequently erected using a derrick. A derrick is
a fixed-mast, guyed crane that is positioned at the center of the tank.
Placement of a derrick is optimal for construction of circular tanks,
because the boom is capable of a full 360-degree swing. On the basis
of years of construction experience, tank constructors have developed
guyed derricks specifically designed for tanks.
Mobile or tower cranes are also options for erecting an elevated
tank in the field. Because of rental expenses, mobile or tower cranes
are usually limited to use on a smaller-capacity tank of limited height

with a short field-construction schedule. Use of a mobile crane requires

more site clearance around the tank.
Whether using a derrick, tower crane, or mobile crane, adequate
support is critical. For either a derrick or stationary tower crane, sup-
plemental groundwork or a support pad should be provided if re-
quired before setting the crane. A mobile crane requires a reasonably
level surface around the tank and site conditions capable of support-
ing the loaded crane.
General Requirements
Field subassemblies Shipped plates are frequently subassembled on
the ground in the field. The subassemblies are planned on the basis
of the maximum weight and size feasible to lift and fit into place.
This erection practice allows the welding to be performed close to the
ground in more favorable positions.
Construction aids Specialized erection equipment developed by the

tank constructor is used to aid in lifting, fitting, aligning, and spacing
plates with the appropriate weld gaps. Maintaining the proper gaps,
alignment, and overall dimensional accuracy is critical for subsequent
plate placement. Some construction aids may be permanently left in
place, while others are temporary and are removed after the plate is
secured. Temporary attachments need to be removed without damag-
ing the plates, and the remaining weld should be chipped or ground
smooth before painting. Dimensional accuracy is maintained by con-
sistently checking dimensions. Levels and transit levels may be used
to verify elevations, check angles, and to verify that a component is
plumb.
Access to the tank Safe access to the tank and tower is required for
welders, inspectors, and painters. Temporary scaffolding is commonly
used in conjunction with permanent and temporary ladders for access.
Aerial lifts such as a boom lift or scissor lift may also be used. A work
basket or chair hung from a spider line is another frequently used
option. This option requires a secure anchor point above the area to
be accessed. Regardless of the method of access, fall protection needs
to be considered and special measures taken to ensure the safety of
the workers.
Fluted-Column-Style Tank
The fluted tower rests on butt-welded base plates that are seated on
shims and fixed to the foundation by the anchor bolts. It is critical that
the base plate is level before erecting the fluted plates. After the tank
has been erected, but before it is filled with water, the space between
the base plate and foundation is grouted.

248 Chapter Six
FIGURE 6-10
Erection of tower,
fluted-column-style
tank.
The vertical joints in the fluted tower are lap welded, and the
horizontal joints are butt welded. Water-bearing plates in the tank
are welded with full-penetration butt welds. The roof is typically lap
welded on the top side only. When specified by the purchaser, the
overhead laps in the roof are also seal welded (Figs. 6-10 to 6-12).
Access to the inside of the fluted tower is required at all times and
is typically provided at the opening for the overhead door. A bottom
manhole provides access to the inside of the tank. Tank constructors
have developed specialized equipment to enable safe access to dif-
ficult areas such as the outside of the cone or the underside of the
roof.
The constructor must be cognizant of the stability of the structure
at all times, but especially when the structure is left overnight. Re-
gardless of what component the crew is erecting, the crew should not
leave the tank unattended until all the plates in a given ring are in
place and adequately secured. Provisions should be taken to stiffen
unfinished sections of the tank in case of high winds. This may include
providing stiffening or continuous scaffolding at the upper limits of
construction or guying the structure to the ground.

FIGURE 6-11
Erection of cone
plate, fluted-
column-style tank.
FIGURE 6-12
Erection of
cylindrical shell
plate, fluted-
column-style tank.
249
250 Chapter Six
FIGURE 6-13 Erection of spherical plate, pedestal-style tank.
Pedestal-Style Tank
Complete-penetration butt-welded construction is used for all the
components of the pedestal tank and tower except the roof. The roof
is usually lap welded on the top side. At the request of the purchaser,
the underside may also be seal welded or, alternately, the roof may be
butt welded (Figs. 6-13 and 6-14).
The base cone of the tower of a pedestal-style tank sits on a thick
base plate that is welded with complete-penetration butt welds, set
on shims, and fixed to the foundation by the anchor bolts. As with the
fluted-column-style tank, it is critical that the base plate be level before
erecting and welding the base cone. The base cone is fillet welded to
the base plate. As with the fluted-column-style tank, grout is placed
under the base plate after the tank is completely erected but before it
is filled with water.
Multicolumn-Style Tank
Erection for a multicolumn-style tank typically begins with one bent
in the first panel, consisting of a pair of columns, one bolted strut,
and loosely connected cross-bracing that will either be welded or

FIGURE 6-14 Erection of roof plate, pedestal-style tank.
bolted to the columns. Base plates are welded to the bottoms of the
columns in the first panel. The base plates of the first bent are set
on shims placed on the foundation pedestals and are fixed using the
anchor bolts. Additional bents are erected by sequentially adding a
column, strut, and cross bracing around the tower. After all the bents
are in place in the first panel, the cross braces are adjusted to length,
as required, to ensure that the panel is square and true before pro-
ceeding to the next panel. Subsequent panels are erected in a similar
manner with the columns of the upper panel welded to the lower
columns.
Typically, the portion of the tank that is welded to the upper col-
umn is welded to the column before the tank is erected. After the
tower is complete, the intermediate plates are fit-up and welded with
complete-penetration butt welds. It is critical to maintain dimensional
accuracy of the tower for proper fit-up of these plates. The tank joints
between water-bearing plates are joined with complete-penetration
butt welds. The roof plates may be lap welded with or without seal
welding, or they may be butt welded.
An alternate construction sequence is to construct the upper bents
and tank without the lower columns in place. The advantage of this
construction method is that the tank and upper tower can be con-
structed and painted when closer to the ground. After this portion of
the tank is complete, it is lifted by cranes and the lower columns are

252 Chapter Six
FIGURE 6-15
Multicolumn-style
tank.
set underneath (Fig. 6-15). The structure is then lowered to its proper
height and seated on the columns.
After the tank is complete, but before it is filled, the final tightening
or welding of the cross-braces is done. At this stage, the grout is also
placed under the base plates.
Construction of Composite Elevated Tanks

Concrete Support Structure
The concrete support structure for a composite elevated tank is cast
in place. Jump forms are commonly used. Constructors of this style
of elevated tank have specialized forms conforming to their standard
geometries. Forms have horizontal and vertical rustications built into
the exterior face to provide architectural relief and help mask form
panel joints and construction joints (Fig. 6-16).
Wall reinforcing is placed and tied before the forms are installed.
Special reinforcing is required around the overhead, mandoor, and
other significant openings. After the rebar is placed, the forms are
set and prepared to receive concrete. Large openings in the tower are
blocked out, while smaller openings may be cut or drilled after con-
crete placement. The concrete is delivered to the tops of the forms

FIGURE 6-16 Tower

construction,
composite
elevated tank.
either by pumping or by bucket. Hand compaction and power vi-

bration are performed in accordance with ACI 309 to ensure proper
compaction, minimize segregation, eliminate air voids, and to ensure
close contact with the reinforcement and forms.
After the first ring is poured and has had time to cure, the forms are
removed. The sequence of placing reinforcement, jumping the forms,
and pouring the concrete is repeated until the support tower is com-
plete. Forms are set for the dome and ringbeam or the flat slab. Similar
to the tower forms, the tank constructor will have developed special-
ized forms for their geometry and construction practice. Reinforce-
ment is placed and tied and embedments secured before the concrete
is poured.
Concrete Mix
The concrete mix should be suitable for the method of placement
and the weather conditions. The proportions of the mix should be
adjusted to provide adequate workability and the proper consistency
for placement.
For each tank, the material should be from a consistent source
and the mix design number verified upon delivery. The arrival of
trucks should be sequenced to sustain a pour without long delays.
Retempering of the concrete should be controlled to maintain the mix

254 Chapter Six
FIGURE 6-17 Roof

erection,
composite
elevated tank.
parameters. Concrete testing should be in conformance with ACI 318,

Building Code Requirements for Structural Concrete.
Welded-Steel Container for the Composite Elevated Tank

Welding for the cone, shell, and roof of a composite elevated tank is
similar to that for a fluted-column tank (Fig. 6-17).
Method of Erection—Hoisted Versus Crane or Derrick

Construction methods for the steel container on the composite ele-
vated tank and the fluted-column tank are similar if a crane or derrick
is used. One additional construction technique, hoisting, has been
successfully employed for composite tanks. This method allows the
constructor to erect and weld the cone, shell, and upper cone roof
transition as a complete unit around the concrete shaft at grade. After
the welding is complete, the container is hoisted into position using
a series of cables and hoists. It is welded into place, and the roof is
installed. (Fig. 6-18a and b).
Liner Plate
An interior liner plate is placed over the dome or flat slab. The liner is
lap welded on the top side only. For tanks with a dome, formed liner
plates may be used and constructed so that the liner lies directly on

(a)
(b)
FIGURE 6-18 (a) CET hoisted tank erection as the tank is being raised. (b) CET
hoisted tank erection with the tank in the final position.
255
256 Chapter Six
FIGURE 6-19 CET

liner plate formed
to fit dome with
derrick-erected
cone and shell
plates.
the dome. Alternately, unformed steel liner plates that do not match
the shape of the dome may be used if the space between the plate
and the dome is completely filled with flowable grout after welding
(Fig. 6-19).
Inspection and Testing

Foundation Tolerances
Before construction is started of the tank’s support structure, the foun-
dation should be checked to verify that it is within specified toler-
ances. AWWA D100 provides minimum requirements for foundation
tolerances for multicolumn tanks, single-pedestal tanks, and ground-
supported flat-bottom tanks.
Minimum tolerances for anchor bolts need to be maintained for
installation of the base plates for elevated tanks and anchor chairs
for ground storage tanks. AWWA D100 anchor tolerances for all tank
styles are as follows: Anchors should be within ±1/4 in. (±6.3 mm)
of the theoretical location and plumb within 1/8 in. in 12 in. (19 mm
in 305 mm). The anchor projection above the top of the foundation
should be within ±1/4 in. (±6.3 mm).

Foundations and anchors that are not within tolerances should

be identified and addressed before the tank is erected. A design pro-
fessional representing the tank constructor should evaluate whether
the structural capacity of the foundation has been compromised and
should provide details for remedial action when required.
Tolerances for Concrete Support Structure

Dimensional tolerances for the concrete support structure of a com-
posite tank are outlined in ACI 371R and repeated here. Fabrication
and placement tolerances for rebar should be in compliance with ACI
117.
Dimensional tolerances for the concrete support structure are as
follows:
Variation in thickness
r Wall: −3.0 percent, +5.0 percent
r Dome: −6.0 percent, +10.0 percent
Support wall variation from plumb

r In any 5 ft (1.5 m) of height: 3/8 in. (9.5 mm)
r In any 50 ft (15 m) of height: 1.5 in. (38 mm)
r Maximum for total height: 3.0 in. (76 mm)
Support wall diameter variation

r 0.4 percent (not to exceed 3.0 in. [76 mm])
Dome tank floor radius variation

r 1.0 percent
Level alignment variation

r From specified elevation: 1.0 in. (25.4 mm)
r From horizontal plane: 1/2 in. (12.7 mm)
The offset between adjacent pieces for formwork facing material
should not exceed the following:
r Exterior exposed surfaces: 1/8 in. (3 mm)
r Interior exposed surfaces: 1/4 in. (6.3 mm)
r Unexposed surfaces: 1/4 in. (6.3 mm)
The finish tolerance of troweled surfaces should not exceed the
following when measured with a 10-ft (3-m) straightedge or
sweep board:
r Exposed floor slab: 3/8 in. (9.5 mm)
r Tank floors: 3/4 in. (19 mm)
r Concrete support for suspended steel floor tank: 1/4 in.
(6.3 mm)

258 Chapter Six
Maximum Diameter (ft[m]) Radius Tolerance (in. [mm])

40 [12] ± 0.5 [±12.7]
150 [45.7] ± 0.75 [±19]
<250 [<76.2] ± 1.0 [±25.4]
≥250 [≥76.2] ± 1.25 [±32]
TABLE 6-2 Roundness—Cylindrical Shells
Welded Tank Tolerances

Tank constructors are responsible for maintaining the quality of their
work. AWWA D100 outlines some specific shell tolerances for plumb-
ness, roundness, peaking and banding, and localized flat spots for
ground storage tanks. The maximum out-of-plumbness from top to
bottom should not exceed 1/200th of the total shell height of the shell.
Single shell plates should also meet the flatness requirements of ASTM
A6. To check the roundness of the shell, measurements should be
taken 1 ft (0.3 m) above the shell to bottom seam and not exceed the
tolerances in Table 6-2. Peaking, defined as the out-of plane distor-
tion across a vertical weld seam, should not exceed 1/2 in. using a
36-in.-long sweep board. Banding, the out-of-plane distortion across
a circumferential weld, is also limited to 1/2 in. using a 36-in.-long
sweep board.
Industry-wide tolerances for plumbness, roundness, peaking and
banding, and localized flat spots for elevated tanks do not exist.
AWWA D100 indicates that the tanks are to be “trued up” after erec-
tion and prior to grout placement. To guard against fit-up problems
as field erection progresses, most individual constructors have self-
imposed tank and tower tolerances to ensure that the tank is built
plumb and within fairly tight construction tolerances.
AWWA D100 provides erection tolerances for plates designed for
stability. If the compression allowables of AWWA D100 are used in
the design, the maximum local deviation from the theoretical shape
must be less than
√
ex = 0.04 Rt
√
Lx = 4 Rt
where
Lx = gauge length to measure local imperfection

ex = local deviation from theoretical shape
t = shell thickness
R = radius of exterior surface of the shell, normal to the plate at the
point under consideration and measured from the exterior
surface of the plate to the axis of revolution.

Subject to Subject to
Thickness Primary Stress Secondary Stress
(in. [mm]) (in. [mm]) (in. [mm])
0 < t ≤ 5/8 [15.8] 1/16 [1.6] 1/8 [3.1]
>5/8 [15.8] Lesser of 0.10t or Lesser of 0.20t or
1/4 [6.3] 3/8 [9.5]
t = Nominal thickness of the thinner plate at the joint.
TABLE 6-3 Maximum Allowed Offset for Butt-Welded Plates Subject to Primary
or Secondary Stress
Alternately, if a plate thickness is based on a buckling analysis per-

formed by the tank constructor, measurements must be taken to verify
that the deviations assumed in the analysis have not been exceeded.
AWWA D100 also provides tolerances for plate alignment for lap-
welded joints and butt-welded joints. Lap joints should be held in
as close contact as possible, with no plate separation exceeding 1/16
in. (1.6 mm). When plate separation is present, the weld size should
be increased by the amount of separation. The maximum allowed
offset for butt-welded plates subject to primary or secondary stress is
defined in AWWA D100 and repeated in Table 6-3.
Welded Tank Inspection

Shop welds that will carry stress from the weight or pressure of the
water are typically inspected by the tank constructor before the com-
ponent is shipped. Some purchasers may choose to visit the tank con-
structor’s shop during fabrication to observe the fabricating practices
and operations. Per AWWA D100, either the constructor or a qualified
inspector hired by the owner is required to check the quality of the
field welding.
Visual inspection by an individual who is competent to perform
the inspection is to be performed on all welds. Competency for vi-
sual inspection can be obtained either by training or by experience. A
weld shall be repaired or replaced if any of the following defects are
found: a crack, lack of fusion, unfilled craters, overlap resulting from
the protrusion of weld metal beyond the weld toe or root, undersized
weld, and porosity in butt joint subjected to primary stress. Welds
exhibiting any of the following defects in excess of the AWWA D100–
specified limits also need to be repaired or replaced: excess butt joint
reinforcement, fillet-weld convexity, undercut, porosity, and plate mis-
alignment. All welds should also be visually inspected to ensure the
removal of all weld spatter, sharp surfaces, overlaps, and unacceptable
undercuts that would be detrimental to the coating life.

260 Chapter Six
The most common method of evaluating complete-penetration

butt-welded joints for weld quality is by radiography. Radiographic
inspection is to be performed in accordance with ASME Section V,
Article 2 by Level II radiographers. Section 11 of AWWA D100 pre-
scribes the number and location of radiographs for tank shells, risers,
and single-pedestal columns. The radiographic inspection should be
performed as the work progresses. The purchaser may participate
in the selection of the specific radiograph locations. AWWA D100 has
radiograph inspection standards that must be used for evaluating dis-
continuities and defects present in the radiograph film. A weld shall
be repaired or replaced if cracks, incomplete fusion, or inadequate
penetration are noted. In addition, welds with inclusions or rounded
indications that exceed the limits specified in AWWA D100 need to be
repaired or replaced.
The welds in bottom plates of a ground storage tank are required
to be tested for watertightness by magnetic-particle testing or by air-
pressure or vacuum testing.
There are additional inspection requirements for ground storage
tanks built to AWWA D100 Section 14. The weld between the shell
and the bottom should be inspected for watertightness using dye pen-
etrant, penetrating oil, or diesel fuel. The inside fillet weld is completed
first and indicator sprayed on the weld. If any indicator is visible out-
side of the shell after a wait period, a leak is indicated and it should be
repaired. Once there are no indications of leakage, the outside weld
can be completed. Section 14 requires more extensive radiographic
examination of butt welds in the shell and annular plate. It also re-
quires that all welds attaching manholes, nozzles, and other pene-
tration be inspected for cracks by either the magnetic-particle or the
dye-penetrant method by a qualified inspector.
Hydrotest
Water testing is typically performed on the completed tank after it is
painted and disinfected. The purchaser is responsible for furnishing
the water to the site with sufficient pressure to fill the tank. Water
should be filled to the top capacity level, and weld seams should be
inspected for any signs of leakage. If leaks are found, the water must
be lowered at least 2 ft (0.6 m) below the point of repair, and the defect
must be repaired and rewelded. If no leaks are found, the tank can be
put directly into service, which eliminates the need to dispose of the
test water.

CHAPTER 7
Construction of
Bolted-Steel
Water-Storage
Tanks
Keith McGuire, P.E.
Columbian TecTank
Richard Field, P.E., S.E.

Engineered Storage Products Company (retired)
Bolted-steel tanks are factory coated and conform to American Water

Works Association (AWWA) D103, Factory-Coated Bolted Carbon
Steel Tanks for Water Storage.
Erection of the Tank

Equipment required to erect a bolted-steel tank varies among tank
manufacturers. When jacks are used, a complete ring of shell sheets
is assembled and raised by the jacks so that the subsequent ring of
sheets may be assembled under them. The tank is assembled from the
top down while the construction crew operates from a location at or
near ground level.
Tank walls consist of shell sheets or panels carefully packed at
the factory to prevent damage to the factory-applied coating during
shipment. Shell sheets and panels typically have varying bolt-hole
patterns designed to resist the increased hoop loading as tank height
increases (Fig. 7-1). The roof may consist of factory-coated roof panels,
deck plates supported by structural members, or a self-supported alu-
minum dome.
261
Construction of Bolted-Steel Water-Storage Tanks
262 Chapter Seven
FIGURE 7-1 Bolt-hole patterns vary. (Photo courtesy of CST Industries.)
Various lengths and strengths of structure bolts, washers, and nuts

are used to assemble the tank and appurtenances. Each manufacturer
specifies the degree to which bolts must be tightened to properly as-
semble the tank and its components. Typical instructions indicate a
specified torque or a visual observation of gasket deformation that
must be achieved.
Coated areas that are damaged in the erection process must be
repaired in accordance with the tank manufacturer’s written proce-
dures. In addition to wrenches and drift pins, proprietary equipment
may be used to ensure that the sheets are properly joined and aligned.
Sealant and Gaskets

To ensure watertightness, gaskets and/or sealer are used to seal all
joints. Sealer is dispensed from tubes or sausage packs via hand-held
caulking guns. The most common sealant materials used are urethane
and silicone.
Gaskets to seal along sheet seams are supplied in strips. Strip
gasket is pre-punched and furnished in rolls, and it must be cut to the
required length during erection. When splices are necessary, lap joints
are required, and a bead of sealant is applied along the joint. Circular
gaskets for items such as nozzles, manways, and collars are furnished
in one piece or are in several pieces that are joined to form rings.
Special gaskets are sometimes required to seal certain areas of the
bolted tank panels, depending on the type of construction. Lap gas-
kets, radius fillet gaskets, and other types of gaskets may be furnished
by the tank manufacturer for use at specific tank locations during as-
sembly.

C o n s t r u c t i o n o f B o l t e d - S t e e l Wa t e r - S t o r a g e Ta n k s 263
Floor
If the tank will have a concrete floor, specially sized sheets and panels
will be embedded into the floor, forming a ring of foundation sheets
on which the remainder of the tank will be erected. If the tank will
have a steel floor, factory-coated floor segments/panels will be used.
Unloading and Storage

Tank components and materials are usually shipped to the site on
skids or special racks on open trailers. Before unloading, the materi-
als should be checked for shipping damage. During unloading, it is
recommended that a forklift truck or crane be used to avoid damage to
tank components and boxed items. An inventory should be performed
to ensure that all items shipped agree with the bill of lading.
The materials should be stored near the tank location on dry
ground, out of the way of other construction activities, and secure
from pilferage and theft. All materials should be covered to prevent
damage from the weather.
Concrete Floor Construction

Footing and floor concrete placement are usually performed in two
separate pours (Fig. 7-2). Leveling plate assemblies are installed in
FIGURE 7-2 Footing and floor concrete placement.

264 Chapter Seven
FIGURE 7-3 Leveling plate assemblies.
the top of the footing pour. These assemblies are used to position the
embedded tank foundation sheet (Fig. 7-3). The foundation ring is set
on the leveling plate assemblies, then leveled and rounded to specified
tolerances (Fig. 7-4). Floor sumps are then installed, and other piping
is stubbed off above the floor line (Fig. 7-5).
FIGURE 7-4 Foundation ring set on leveling plate assemblies.

FIGURE 7-5 After floor sumps are installed, other piping is stubbed off.
Reinforcement is placed in the floor area and around the curb to

the outside of the foundation ring in accordance with the engineer-
ing drawings (Fig. 7-6). Forms are usually used for creation of the
foundation floor curb around the entire perimeter of the foundation
FIGURE 7-6 Reinforcement placed in floor area and around curb.

266 Chapter Seven
FIGURE 7-7 Floor and curb concrete placed and finished.
assembly. Forms may be installed during or after placement of the

floor and curb reinforcement steel. A water-stop seal strip is installed
onto the foundation sheet prior to concrete placement. The floor and
curb concrete is placed, finished, and allowed to cure before further
tank construction takes place (Figs. 7-7 and 7-8).
FIGURE 7-8 Concrete allowed to cure.

Steel Floor Construction

For a tank with a steel floor, a concrete ring footing should be installed
that contains properly prepared soil and provides support for the tank
shell, steel floor, and the stored liquid load. Special soil conditions and
other requirements may dictate the use of a concrete slab to properly
support the tank. Embedded anchor bolts, if required, are installed to
the specified projection above the top of the footing (Fig. 7-9).
If the tank does not require anchor bolts, and depending on the
site conditions and tank loadings, then tanks with steel floors may
be supported on a prepared granular base or berm in lieu of a concrete
foundation. A steel perimeter ring at least 1 ft (0.3 m) larger than the
tank radius may be used to contain the compacted base.
The steel floor is constructed in accordance with the tank man-
ufacturer’s instructions. Assembly of pie-shaped segments begins at
a circular plate at the center of the tank, and segments are assem-
bled from the center outward until the outside of the tank is reached
(Fig. 7-10). For floors using rectangular segments (Fig. 7-11), workers
start at a designated point along the tank perimeter and work their
way across the tank to the other side.
FIGURE 7-9
Embedded anchor
bolts are installed
to specified
projection.

268 Chapter Seven
FIGURE 7-10 Assembly of pie-shaped segments.
FIGURE 7-11 Floors using rectangular segments.
Tank Construction
Jacking Method
Specially designed jacks are used to build the remainder of the tank.
The jack assemblies are anchored to the tank floor, one at each sheet

FIGURE 7-12 Jack assemblies are anchored to tank floor.
location around the tank perimeter (Fig. 7-12). After all jacks are in-
stalled, the faceplates to which the wall sheets will attach are leveled
with each other.
As the first ring of sheets is placed on the jacks, sealer is placed
in the overlapping vertical joints, and the sheets are bolted together.
Before tightening the bolts, a special tool is used to spread the joint
apart, simulating the loading that will be placed on the vertical joint
from the stored liquid load. As the joint is held in its spread condition,
the bolts are tightened to their specified torque.
At this point, the roof segments are bolted and attached to the top
of the first ring of sheets (Fig. 7-13). The center of the roof is temporarily
supported while the roof segments are bolted into place. When the
roof is complete, the temporary support is removed and the entire
structure is jacked up to the next level.
After the first ring of sheets is completed and tightened, the jacks
are energized and the structure is raised to a height that allows the
next ring of sheets to be installed in the same manner as the first
ring (Fig. 7-14). The second ring of sheets rests in supports attached
to the foundation sheet while being bolted to the ring above it. On
completion of the second ring of sheets, the jacks are disconnected
from the first ring, lowered to a specified location in the ring just
completed, and connected to that ring. The jacks are energized, and
the structure is raised to the next level. Tank erection continues in
this manner until the last ring of sheets is installed and tied into the
foundation sheets. Tanks taller than 120 ft (36.5 m) have successfully
been erected in this manner.
The last sheet of the bottom ring is typically left out to provide
an easy means of access for other work inside the tank and for the
removal of the jacks.

270 Chapter Seven
FIGURE 7-13 Roof segments are bolted and attached.
Wind stiffeners, ladders, ladder cages, and platforms, as applica-

ble, are attached to the side of the tank as successive rings are assem-
bled and jacked to the next level (Fig. 7-15).
Scaffold Method
Exterior scaffolding similar to that shown in Fig. 7-16 is required. The
quantity and length of scaffold planks required are determined by
FIGURE 7-14 The jacks raise the structure.

FIGURE 7-15 Wind stiffeners, ladders, ladder cages, and platforms are
attached.
FIGURE 7-16 Exterior scaffolding.

272 Chapter Seven
FIGURE 7-17 Scaffolding bracket.
the size of the tank and the width of the tank panels. Scaffolding is
moved from ring to ring as work progresses, meaning that the min-
imum quantity of planks is enough to encircle the tank at one level,
plus a few extra. Some crews prefer to leave all scaffolding in place un-
til the shell and roof are complete, thus requiring two planks for every
shell segment in the tank. Scaffold planks are supported by scaffold-
ing brackets (Fig. 7-17). Normally, the brackets consist of steel angles
2.5 in. × 2.5 in. × 0.25 in. (63.5 mm × 63.5 mm × 6.35 mm). Ac-
commodation for safety line uprights must be provided. The bracket
shown in Fig. 7-17 will receive a tubular upright that fits over the 6-in.-
(152-mm)-long plank-retaining rod at its end. Any plank-retaining de-
vice should have a minimum height of twice the scaffold plank thick-
ness. The safety line upright must be 42 in. (1 m) in height, measured
from the top of the scaffolding board. Common practice is to leave all
scaffolding brackets in place until the shell and the deck are complete.
This requires one bracket for each shell segment in the tank. All boards
must be secured.
If interior scaffolding is not used, a hooked drive-out ladder sim-
ilar to the one shown in Fig. 7-18 is required during erection. Points
that bear on the shell should be padded to protect the interior finish of
the tank. Hoisting equipment capable of lifting components weighing
as much as 1,000 lb (453.6 kg) is required. A variety of devices can
be used, but gin poles similar to the one shown in Fig. 7-18 are the

FIGURE 7-18 Drive-out ladder and gin pole.
most common. Made of steel pipe or tubing, they are supported from
scaffolding brackets and extend well above the top of the ring under
construction. The tank shell must be protected from gin-pole bearing
points.
Roof Installation
At least three types of roofs are commonly installed on factory-coated
bolted-steel tanks. These types are self-supported (Fig. 7-19), center-
supported (Fig. 7-20), and self-supported aluminum domes (Fig. 7-21).
Self-Supported Roofs
Self-supported roofs consist of one-piece pie-shaped panels that are
temporarily supported in the center while the panels are lapped onto
each other and bolted together. The roof slope is typically 20 degrees,
and the outer end of the panel is formed in a rounded shape, creating a
knuckle, which adds stiffness to the panel. This outer edge bolts to the
top of the tank wall (Fig. 7-13). After all roof panels are assembled and
attached to the tank wall, the temporary center support is removed,
allowing the roof to support itself.
Low-profile self-supported roofs are typically sloped at a 1:12 pitch
to allow for rain runoff. One or two horizontal support members span
the tank diameter, and an elevated collar is located at the mid-span to
set the roof pitch and accept the radial rafters or stiffened roof sheets.

274 Chapter Seven
FIGURE 7-19 Self-supported roof.
Center-Supported Roofs
Center-supported roofs are supported by a center pole that extends
from the tank floor. A common system consists of a center support
column of prefabricated pipe with a base plate at the bottom, a rafter-
bearing plate at the top, and radial rafters (see Fig. 7-22 for proper
arrangement of parts).
FIGURE 7-20 Center-supported roof.

FIGURE 7-21 Self-supported aluminum dome roof.
Self-Supported Aluminum Domes

Self-supported aluminum domes are spherical structures conform-
ing to the dimension of the tank. It is a clear span structure with a
fully triangulated space frame complete with noncorrugated closure
panels. These domes can be constructed in place on the top ring of the
FIGURE 7-22 Arrangement of parts, center-supported roof.

276 Chapter Seven
FIGURE 7-23 Self-supported aluminum domes can be constructed in place.
tank shell sheets (Fig. 7-23) or constructed on the ground and lifted
onto the top ring of shell sheets as a completed assembly. In either
case, the mounting shoes of the dome attach to a formed angle that
has been attached to the upper horizontal bolt line of the first ring
of shell sheets. The dome consists of an aluminum I-beam structure,
lightweight aluminum roof panels, and flashing. Although the dome
is attached to the tank shell at the perimeter roof angle, the attachment
allows necessary movement between the roof and the tank.
Tank Appurtenance and Accessory Installation

Ventilator
Normally, the roof ventilator is located in the center of the roof at
the roof cap (Fig. 7-24). The ventilator capacity needs to be sufficient
to pass air so that the maximum possible rate of water entering or
leaving the tank will not cause excessive pressure or cause a vacuum
to be developed. For potable water storage, screening to prevent birds
and bugs is required, and the ventilator must be capable of relieving
pressure and vacuum if the screens become clogged.
Roof Accessories
A hinged, lockable roof access door is normally provided near the
outside ladder to allow liquid samples to be withdrawn from the top

FIGURE 7-24 Roof ventilator usually located in center of roof at roof cap.
of the stored liquid (Fig. 7-25). The door is bolted onto an opening
provided in the roof panels for this purpose.
A roof railing assembly is installed after the roof installation is
complete. Depending on the pitch and the type of roof, a walk-
way may be provided (Fig. 7-26). The assemblies are bolted together
FIGURE 7-25 Hinged, lockable roof access door near outside ladder.

278 Chapter Seven
FIGURE 7-26 A walkway may be provided.
using appropriate fasteners; sealer and neoprene pads are used where
support brackets attach directly to the roof panels.
A caged ladder with a roof manway landing platform that meets
requirements of the Occupational Safety and Health Administration
(OSHA) is supplied with most tanks (Fig. 7-27). Depending on tank
height, some ladder/cage assemblies also include one or more step-
off platforms. Individual sections of the ladder/cage assemblies are
constructed on the ground. The sections are then attached to the rings
of shell sheets during the tank erection process.
Shell Penetrations
Tank manufacturers’ policies vary concerning penetrations through
the tank sidewall for piping and other instrumentation. Some manu-
facturers supply the openings in the panels shipped from the factory;
others provide detailed instructions for locating and cutting the open-
ings in the field during or after tank erection. Depending on the open-
ing size, the tank manufacturer may require and provide the means
to reinforce the area around the opening. Depending on the tank coat-
ing, the reinforcement can be a plate welded around the opening in
the factory or bolted on in the field (Fig. 7-28).

FIGURE 7-27 Caged

ladder is supplied
with most tanks.
FIGURE 7-28 Depending on coating, tank can be reinforced in factory or field.

280 Chapter Seven
FIGURE 7-29
Brackets holding
overflow pipe
should be correctly
located.
After the tank sidewall has been tied into the foundation sheet, one
or more access door assemblies (manways) are mounted to the tank
sidewall within the lowermost full-height ring of shell sheets (refer to
Fig. 7-26).
A variety of overflow piping designs can be installed. Care should
be taken to ensure the brackets holding the overflow pipe are correctly
located and cushioned against the tank to prevent damage to the coat-
ing (Fig. 7-29).
When level indicators and other control devices are being in-
stalled, workers need to heed the tank manufacturer’s instructions
regarding penetrations of the sidewall and prevention of coating
damage.
Completion
After all tank sections have been erected, all appurtenances have been
installed, and piping is complete, the interior is cleaned of all con-
struction equipment and debris. Any damaged coating areas on the
tank interior or exterior are repaired in accordance with the coating
manufacturer’s instructions. The tank exterior is examined to ensure
that all safety decals are in place, if applicable.

FIGURE 7-30 Completed reservoir tanks.
The tank is tested for leaks by filling it to the overflow level. Any
leaks found are repaired in accordance with the tank manufacturer’s
recommendations. The test liquid is usually disposed of using the
tank’s drain system.
FIGURE 7-31
Completed
standpipe.

282 Chapter Seven
Once testing is complete, the tank should be disinfected. AWWA

Standard C652 and the tank manufacturer’s recommendations should
be followed to achieve proper tank disinfection.
All construction equipment is removed and the tank site is cleaned
in accordance with the project specifications. Figures 7-30 and 7-31
show the completed tanks.

CHAPTER 8
Inspecting New-Tank
Construction
Steven P. Roetter, P.E.
Tank Industry Consultants
The purchaser should inspect the tank as construction proceeds to en-

sure that the structure complies with specifications, both in form and
in quality. This chapter discusses the specific items to be checked by
the purchaser and those that are the responsibility of the constructor.
On the job, the purchaser and the constructor should keep in mind
that a quality structure is the result of the cooperation and affirma-
tive efforts of all parties. If the relationship between the purchaser
and the constructor is combative or antagonistic, the project is likely
to be completed late and quality will probably be adequate at best.
On the other hand, a purchaser with a sound knowledge of specifica-
tions, standards, and trade practices can work with the constructor to
complete a high-quality project on schedule.
Responsibility for Quality

In general, tank constructors are responsible for maintaining the qual-
ity of their work. The American Water Works Association (AWWA)
D100 Standard for Welded Carbon-Steel Tanks for Water Storage and
its D103 Standard for Factory-Coated Bolted-Steel Tanks for Water
Storage give the constructor the responsibility for designing and con-
ducting a welding or bolting quality assurance program. According to
the provisions of AWWA D103, the constructor is responsible for se-
lecting a factory-coated bolted tank (manufactured in accordance with
that same standard) that meets the capacity and height or diameter re-
quirements of the owner. This is a logical assignment of responsibility,
because the designing and building of tanks is a specialized field, and
283
Inspecting New-Tank Construction
284 Chapter Eight
specialized equipment and knowledge are required to conduct a total

quality control program. Nonetheless, the owner or the owner’s engi-
neer, as well as third-party consultants where needed, should monitor
the work for inevitable human errors and misunderstandings.
Even before construction begins, the owner or engineer should
check shop drawings to ensure that the special requirements of the job
are met. The constructor is usually responsible for the structural design
of the tank, but the interplay of the legally defined responsibilities of
the engineering and construction phases is not firmly established. The
engineer reviewing the drawings should clearly specify the purpose
and the extent of the review to avoid being held legally liable for a
more detailed examination than was actually performed.
A quality assurance program for the tank project should take into
account each of the functions involved in building a tank. In gen-
eral, these functions fall under the following categories: foundation,
fabrication, steel delivery, tank erection, field-applied coatings, shop-
applied coatings, and appurtenances.
The Foundation and Composite-Tank Pedestal

The foundation for a tank and the concrete pedestal of a composite
tank are the most significant sources of potential failure, and so most
owners and engineers apply greater expertise to inspecting the foun-
dation than to any other function of tank construction. If neither the
owner nor the owner’s engineer has expertise in this area, other ex-
perts should be contacted. The reinforced-concrete foundation and
the soil on which it bears are made up of nonhomogeneous materials
over which there is little control. In addition, the foundation is fre-
quently installed by a specialty constructor or under another division
of the water project, giving the tank constructor little control over this
function. Figure 8-1 shows foundation construction in progress.
Soil Investigation
The purchaser should be at the site when the borings are being taken
as part of the soil investigation. This gives the purchaser better insight
into problems that might be encountered during construction of the
foundation.
Activities Before Concrete Is Placed

The excavation should be properly shored to prevent cave-in. The soil
conditions at the bottom of each excavation should be evaluated to
confirm that they are the same as those used when the foundation
design was developed. The bottom of the excavation can be sealed
with a mud mat to prevent water from changing the soil character-
istics and to provide a solid surface from which to work. If piles are

I n s p e c t i n g N e w - Ta n k C o n s t r u c t i o n 285
FIGURE 8-1 Foundation construction in progress.
being driven, the purchaser must determine if the piles are driving
as predicted. The purchaser should be on the job to verify that the
pile-driving log is properly completed. Site and final concrete eleva-
tions should be confirmed. Placement of forms, reinforcing steel, and
anchor bolts should be verified, and photographs should be taken to
document these steps before the constructor allows the concrete to be
placed.
Concrete
If the owner, constructor, or engineer does not have extensive previ-
ous experience with the materials supplied by the concrete ready-mix
plant, a design mix should be developed and tested. The consistency
of the concrete should be evaluated as it comes out of the chute, and
concrete test cylinders should be taken. The storage location for the
test cylinders should provide satisfactory moisture conditions and
controlled temperature. The cylinders should not be transported dur-
ing initial cure. Test cylinders that are taken after approximately one-
eighth of the truck’s load has been discharged give a more represen-
tative sample of the concrete than the initial material that is placed.
Slump tests should also be performed. For large pours, it is necessary
to sequence the pouring operations so that the concrete does not set
up before placing fresh concrete next to it, creating “cold joints.” The
concrete should not be dropped into the forms from excessive heights.
After the concrete is placed, it should be vibrated with a mechanical

286 Chapter Eight
vibrator to ensure proper distribution of the material. Mechanical vi-

brators should not be used to move the concrete from one location
to another. Proper removal of the forms and of all form ties should
be verified. Tolerances on concrete foundations are given in AWWA
D100 and AWWA D103. The purchaser’s representative should be fa-
miliar with American Concrete Institute (ACI) Standards 301 (Specifi-
cations for Structural Concrete for Buildings) and 318 (Building Code
Requirements for Structural Concrete).
With a good cure, quality concrete develops the necessary com-
pressive strength and is less likely to experience surface deterioration.
If the surface of the foundation is not cured properly to protect it from
the elements, chipping or spalling failure of the concrete can occur
long before the steel tank deteriorates.
Backfilling
If backfilling is done improperly, the concrete is overstressed. Proper
backfilling techniques are necessary to provide a structurally stable
foundation, to prevent water from ponding on moisture-weakened
soils, and to make the site more solid for the tank-erection crew.
Proper backfilling operations require the specialized services of
a qualified soil-testing laboratory to determine the optimal moisture
content and maximum density of the backfill material, check the mois-
ture content of the material being placed, and conduct relative den-
sity tests in the field after the backfill material has been placed and
compacted. This last test ensures that the specified degree of soil com-
paction has been obtained. Adequate soil compaction is particularly
important for the foundations of ground storage tanks. Although the
bottoms of these tanks are usually quite flexible, particular care is nec-
essary for backfilling pipe trenches beneath the tank and the soil or fill
material adjacent to concrete ringwalls. Severe differential settlement
in these places can rupture the underlying pipes and cause possible
failure of the tank bottom.
The contract documents may assign the responsibility for provid-
ing necessary soil-testing services to either the tank constructor or the
owner. In either case, copies of all soil test reports should be furnished
promptly to all interested parties.
Fabrication
It is recommended that the owner and the engineer visit the construc-
tor’s facility while the tank is being fabricated. Fabricators approach
the process differently in terms of flow of materials and the sequence
of operations, which eventually influences how the structure is evalu-
ated. The owner should inspect the quality of shop fabrication, weld-
ing, and fit-up; the type of surface preparation; and the shop coating

if applied. The owner should compare the mill test certifications to

the heat numbers stamped on the steel plates. These mill test certifi-
cations should be filed for future reference. The material should also
be examined for excessive corrosion or pitting, signs that the material
has been stored outdoors for too long.
The amount of fabrication performed in the shop varies depend-
ing on the contractor. One function that will be performed in the shop
is rolling of the plates. Table 19 in the D100 standard outlines the con-
ditions under which the shell plates must be rolled. If D100 requires
that the shell plates be rolled, the owner should verify that they are
rolled to the correct radius. The conditions for welding and the avail-
able equipment make welding in the shop much easier than welding
in the field. The welds should be visually evaluated in the shop to de-
termine if they meet the requirements of D100 and can be adequately
painted. Welding repairs can be more effectively performed in the
shop than in the field.
This visit also serves to open communications for the balance of
the project.
Steel Delivery
The owner should be on hand when the steel is delivered to the job
site. The owner can help resolve conflicts with neighboring property
owners, document any damage occurring in the unloading process,
and protect underground utilities on the site or under the access road.
Tank Erection
Erecting and welding or bolting the steel are tasks for which the exper-
tise of the constructor is vital to the success of the project. Erecting steel
is a dangerous operation, requiring skills acquired only through ex-
perience. During this phase, the purchaser’s representative may need
assistance. Independent testing laboratories are usually equipped to
take radiographs of welded seams, but they know little about steel
erection and fit-up and are not willing to climb to the heights usually
associated with water-storage facilities.
Using someone from another tank constructor’s organization as
the purchaser’s representative can lead to conflicts of interest and
other problems. It is very difficult for a competitor to be unbiased in
the evaluation of another constructor’s work. Even if this competitor
is fair, it is difficult for the tank constructor to accept the opinion as
an unbiased one. Therefore, it is usually best to secure the services of
a consultant who specializes in this type of inspection work and has
the expertise and climbing ability to accomplish the job.

288 Chapter Eight
Fit-Up Quality
Tank Bottom
The levelness of the tank’s base plate(s) is critical if the rest of the tank
is to be erected properly. The constructor’s steel-erection supervisor
should check the foundation(s) for differences in elevation. Any such
differences should be compensated for by shims underneath the base
plate(s). If this task is not done properly, or if there were fabrication
problems with the steel, the purchaser may see slivers of steel that need
to be cut from seams, frequent use of a large hammer to form the steel
variations in the seam gaps, or plates not aligned in accordance with
the tolerances required in AWWA D100 Section 10.6.3. These prob-
lems usually produce a tank of unacceptable aesthetic or structural
quality.
Tank Shell
AWWA D100 has several fit-up requirements. Most are related to ap-
pearance of the structure, but improper fit-ups can be structurally
significant if bad enough. AWWA D100 addresses plumbness of the
shells of ground storage tanks. The shell’s deviation from vertical
should be measured as the shell is erected, and variations from verti-
cal should be corrected when they approach the limits set in AWWA
D100. The standard also establishes the roundness of the shell. The
tank diameter should be measured in several locations 1 ft (0.3048 m)
above the tank bottom corner weld. The measurements should not
exceed the tolerances established in AWWA D100. AWWA D100 also
establishes tolerances for peaking and banding of the shell of ground
storage tanks. Peaking is the out-of-plane distortion across a vertical
weld seam, and banding is the out-of-plane distortion across a hori-
zontal weld seam. To measure peaking and banding, a sweep board
is useful. A sweep board can be constructed from a piece of plywood
36 in. (0.9144 m) long. One side of the plywood board should be curved
to the radius of the tank, and the other side should be flat. The curved
side should be used to measure peaking, and the flat side should be
used to measure banding. The offset of aligned shell courses is gov-
erned by AWWA D100. During fit-up, it should be verified that the
plates are aligned within the tolerances established in D100, and these
tolerances should be maintained throughout the welding process.
Double-Curved, Axisymmetrical, Conical, and Cylindrical Section

In D100, tolerances for these sections are specifically established for
stability and are structurally significant. The owner should verify
through the formulas given in D100 that the double-curved, axisym-
metrical, conical, and cylindrical sections are within the tolerances
established.

Welding Quality
According to AWWA D100, the constructor is required to check the
quality of the welding. It is the job of the owner to monitor the con-
structor’s quality control program. First, the owner should collect the
certification papers of all welders on the site. These papers detail the
types of welding and the steel thicknesses for which the welder is
certified. The most common method of evaluating weld quality is
by means of radiography. The purchaser should participate in the
selection of radiograph locations, watch for documentation of the ra-
diographs, and review the radiographs with the constructor’s quality
assurance expert. The areas selected should be in strict accordance
with the AWWA D100 standard, which requires that the radiographs
reflect the general quality of welding on the tank. Equal proportions
of shop, ground, and air welds should be reflected in the radiographs.
The contractor should document these radiograph locations on a roll-
out. AWWA D100 has radiograph inspection standards that must be
used for evaluating discontinuities and defects present in radiograph
film. The penetrameter, which ensures the reviewer that the radio-
graph was sensitive enough to identify the smallest defect addressed
by the standard, must be visible in each radiograph. It is vital that
the radiographs be evaluated in accordance with this criterion and
that any repairs and follow-up radiographs be conducted in accor-
dance with D100. It is also important for the tank constructor and
purchaser to visually inspect all welds to ensure the removal of all
weld splatter, sharp surfaces, overlaps, and unacceptable undercuts
that will be detrimental to the coating life. Welds do not need to be
perfectly smooth, but sharp edges must be removed. Ground storage
tanks erected under AWWA D100 Section 14 Alternative Design Basis
require many more radiographs than standard tanks.
Bolting Assembly
Bolted-steel tanks require the proper placing of steel sheets, gaskets,
and sealants. Some erection methods may also require pre-tensioning
the sheets and tightening the bolts to a prescribed torque. These details
are covered by the manufacturer’s erection instructions and drawings.
The engineer or purchaser may require that a set of these instructions
be included with the shop drawing package that is submitted.
Tank Appearance
Tank appearance is of great importance to many owners. The final
appearance is known only after the tank is coated, when dents and
buckles become apparent. It is then that the owner expresses dissat-
isfaction. Determining how well the tank complies with the speci-
fications and applicable codes and negotiating a settlement for poor
appearance are time-consuming and stressful. Usually these problems

290 Chapter Eight
can be avoided if the constructor checks to see that the tank is level,
round, and plumb as it is being built. Incorporating dimensional tol-
erances into the contract will also minimize disputes. AWWA D100
specifies some of these tolerances.
Surface Regularity
A smooth, regular surface provides a good base for the application of a
protective coating system, thus helping reduce maintenance costs. To
this end, the constructor and the purchaser should ensure that the weld
contour is smooth, that unacceptable weld undercutting is eliminated,
that weld spatter is ground off, that remains of welds used to attach
erection and fit-up equipment are chipped and ground smooth, and
that unacceptable gouged-out places in the steel are filled in.
Representatives of the tank constructor and the purchaser should
be alert for sharp edges or areas that would cause premature coat-
ing failure so that corrective action may be taken as the work pro-
gresses.
Factory-Coated Bolted Tanks

A bolted-steel tank is delivered to the location with a factory-applied
coating. If the steel has not been damaged in transit, the surfaces will
be smooth. Each panel should be carefully inspected before erection.
If the panel is damaged in transit, in handling, or during erection of
the tank, it may be necessary to repair or replace it.
Water Testing
When welded tanks are water tested before they are coated, any leaks
that are found can be repaired without requiring any coating to be
redone. If the tank is not filled until after it has been coated, small pin-
holes in the welds may be plugged temporarily with coating; these
will cause leaks later if the coating breaks loose. The owner should
ensure that water for the test is available at the time and pressure nec-
essary to coincide with the constructor’s schedule. The owner should
also ensure that provisions are made for draining and disposing of
the test water. If leaks are found in factory-coated bolted tanks, the
constructor should make repairs according to the manufacturer’s rec-
ommendations.
Field Cleaning and Coating

This section discusses the cleaning and coating of welded-steel tanks
after they have been erected and before they are placed in service.

Erection Scar Removal

By the time the tank is erected, irregularities in the surfaces of the
tank should have been eliminated. The erection crew has the equip-
ment and scaffolding to smooth out these defects economically and
effectively. On the other hand, the coating crew is often subcontracted
or made up of members of another department of the tank company.
Their rigging is sometimes not as well suited for this cleanup work as
the scaffolding of the erection crew.
Occasionally, abrasive blasting reveals small laminations in the
steel plate. If these are not removed completely by blasting and remain
large enough to produce holidays in the coating system, they should be
removed by grinding. Occasionally, deeper laminations may require
welding or further testing.
Steel Cleanliness
The first requirement for a good coating is a clean surface. The steel
should be free from dirt and oil, both of which may accumulate during
construction. All weld seams, abraded areas, scratches, shop or field
markings, or poorly adhering shop primer should be removed by
abrasive blast cleaning. The areas cleaned by abrasive blasting should
blend well into the adjoining undisturbed shop primer. Some shop-
applied primers must be scarified or otherwise prepared before ensu-
ing coats are applied.
The purchaser should also be aware that welding or cutting activ-
ity on one side of a plate is likely to damage the coating bonding on the
opposite side of the plate. This is especially important if shop priming
is used. The areas opposite welding or cutting operations should be
examined for coating damage resulting from the heat induced by the
cutting or welding process.
The manuals Good Painting Practice and Systems and Specifications
visual standards and an inspection manual available from the Society
for Protective Coatings give good guidelines for inspecting coatings.
Inspection Instruments
Instruments needed to inspect coating include at a minimum a wet-
film thickness gauge, a calibrated dry-film thickness gauge, equip-
ment for measuring air temperature and humidity, a steel-temperature
thermometer, a surface-profile measuring device, and a wet-sponge-
type holiday detector. The holiday detector is used to inspect the coat-
ing for voids that will cause premature coating failure. If full-time
inspection is not conducted, destructive testing involving the use of
a Tooke Gage and/or other instruments will be necessary to evaluate
the thickness of each coat and to obtain an indication of the cleanliness
of the substrate.

292 Chapter Eight
Inspection Planning
The purchaser should plan work to aid in the timely completion of the
tank field coating. This will require open lines of communication with
the coating company and an understanding of the effects of weather
on coating progress. The constructor will also need to work efficiently
in good weather.
The purchaser should state requirements for the number of lo-
cations to be tested according to the total surface area of the plate.
Minimizing testing is unwise, but an excessive number of testing lo-
cations places an unreasonable burden on the constructor and can
substantially delay the progress of the tank coating. SSPC PA2 delin-
eates procedures for measuring coating dry-film thicknesses. To avoid
excessively delaying the coating progress, large tanks may require
more than one purchaser’s representative to conduct the required field
tests.
Technical Aspects of Coatings

Today’s coatings require exactness in measuring and mixing com-
ponents and thinners. The appropriate application equipment must
be used, and the proper combination of humidity and dew point,
air, and steel temperatures is critical during both application and
curing.
The tank interior must be ventilated to ensure the safety of work-
ers and to allow for the proper curing of the field-applied coatings.
Fans or air horns are usually required to move air through the tank.
Even with forced-air ventilation, proper breathing equipment is nec-
essary for the safety of the workers and the purchaser’s represen-
tative(s).
Shop-Applied Coatings
Bolted tank panels are coated at the factory under controlled condi-
tions. AWWA D103 requires that the panels be grit-blasted to near-
white metal (SSPC SP10) and coated within 15 minutes of cleaning
to prevent rust from starting. The coating is then either baked on or
fused on. If desired, the purchaser may observe these operations dur-
ing shop inspection.
If specified, a preconstruction primer may be shop applied to new
welded-steel tanks. Observation of the shop painting and fabrication
of the steel components of welded-steel tanks is necessary to evaluate
proper fabrication techniques, shop surface preparation, and shop
primer application.

Mechanical and Electrical Appurtenances

Mechanical and electrical items that need to be checked vary depend-
ing on the equipment specified for a particular project. The following
should be checked on every project:
r Electrical wiring should meet applicable codes.
r Conduits, fixtures, pipes, valves, or other items should not
interfere with the safety of the ladders or platforms.
r Cathodic protection anodes and hand-hole covers should be
properly placed, and the purchaser’s representative should
witness the energizing of the cathodic protection system and
the initial potential profile being conducted as per the project
specifications.
r All hatches should be locked.
r Safety belts and sleeves should be furnished for the ladder
safety devices. Safety sleeves should be checked for proper
operation along the full height of the rails or cables. Any coat-
ing, deviations, or obstructions that prevent the free operation
of the sleeve should be removed.


CHAPTER 9
Operation
José N. Hernández, P.E.
City of Cleveland
Sami F. Sarrouh, P.E.

Brown and Caldwell
There are an estimated 400,000 potable water storage tanks in the

United States. Worldwide, the total number of tanks is undoubtedly
staggering. Water-storage facilities have played and continue to play
a pivotal role in the growth and success of water distribution systems.
Existing and new tanks come in a variety of styles, construction
materials, configurations, and functions, and they have a wide range
of capacities. Tanks have been constructed from materials as diverse
as steel, wood, and concrete; they can be at ground level, elevated, or
buried. They can have many different inlet and outlet configurations
and can hold as little as 10,000 gal (37,854 L) or as much as 140 mil gal
(529,957.6 kL) or more (Fig. 9-1).
These and other factors influence tank performance and capabil-
ities. Such characteristics also play a role in how utility personnel
operate the tanks, and how they affect the important tank functions
such as emergency storage capacity and the preservation of water
quality. Concerns about the effects of tank characteristics and opera-
tion on water quality present new challenges for both operators and
tank design engineers.
It is the intent of this chapter to provide the reader with practical
information, in addition to technical and scientific data, regarding the
design and operation of potable water storage tanks.
Although tanks are often the most visible part of a water distribu-
tion system, they are just one among several major components in a
system designed to bring potable water to utility customers. Potable
or drinking water can be defined as the water delivered to the con-
sumer that can be safely used for drinking, cooking, and washing
(De Zuane 1997).
295
Operation
296 Chapter Nine
FIGURE 9-1 Masonry and concrete reservoir, capacity 23 mil gal (87,064 kL),
in Parma Heights, Ohio.
A utility’s water may begin its journey to its customers from a

nonpotable (raw) source or from another utility acting as a wholesaler.
Raw water can be groundwater, such as in the case of wells. It can also
be surface water from lakes, rivers, or dams (Fig. 9-2). It may also be
desalinated brine or seawater.
The raw water is treated according to rules and regulations
established by government regulatory agencies. Myriad treatment
strategies are available. The applicability of each treatment strategy
depends on the characteristics of the source water and the technology
FIGURE 9-2 Water intake crib in Lake Erie; Cleveland, Ohio, is in the
foreground.

Operation
Operation 297
Ground
Elevated Surge tank
tank tank
Treatment
plant
Trunk and
distribution
water mains
Clearwell Pumping
Raw water station
source
FIGURE 9-3 Major components of typical water distribution system.
required to make it potable. There is an extensive list of references

on potable water treatment. AWWA’s Water Quality and Treatment,
A Handbook on Drinking Water, published by McGraw-Hill (AWWA
2010), is an excellent general reference on this subject. Once treated,
water is usually stored in a large tank or reservoir at or near the
treatment plant. Finished water is then pumped into a water dis-
tribution system serving a community or service area, as demand
requires.
Water distribution systems comprise several major components
(Fig. 9-3). Pipes and piping systems include trunk and distribution
mains, valves, and hydrants. Trunk mains are large-diameter pipes
that carry water away from treatment plants. Trunk mains branch off
into distribution mains. Distribution mains are smaller pipes that carry
the water through individual streets or zones. Each water customer
has at least one service tap into a distribution main. In a well-designed
distribution system, trunk and distribution mains act as a network, or
web, with many connecting nodes. The web connections are redun-
dant. If a section of pipe inside the web should break, valves could
be closed to isolate a small affected area and conduct repairs while
maintaining service to the remaining customers in the system. The
pipe network’s design serves another function. A completely closed
pipe network has no dead ends, thus eliminating the possibility of
water stagnation.
When a distribution system’s area is very large or encompasses
substantial differences in elevation, it is often necessary to divide the
system into two or more pressure zones, or “highs.” Booster pumping
stations are used to raise the pressure of the water as it is pumped from
a low-pressure zone to a high-pressure zone.
Storage tanks are connected to the trunk-main web. Under opti-
mal conditions, a distribution system is designed so that one or more
storage tanks are located at the opposite end of the system’s grid from
the pumping station.
The main function of a water distribution system is to deliver suf-
ficient quantity of potable water at a minimum established pressure

Operation
298 Chapter Nine
to utility customers. This role has been expanded in recent years to

that of preserving the quality of the water in the system. More detailed
information regarding the design and operation of potable water dis-
tribution systems is available from many sources, including the Water
Distribution Systems Handbook (Mays 2000).
Modeling of Tanks in Water Distribution Systems

Tanks are an integral element of water distribution systems. Water
distribution system modeling has become an essential tool for distri-
bution system operators, scientists, and engineers. Models represent a
water distribution system as a web or network in which different hy-
draulic scenarios can be simulated. The network model is represented
by a collection of pipe lengths interconnected in a specific topological
configuration by node points where water can enter or exit the sys-
tem (Fig. 9-4). Tanks represent boundary conditions in these models.
Traditionally, tanks in water distribution system models define the
hydraulic grade line limits at its boundaries.
Modeling is concerned with water behavior when the water is
both moving and stationary. An important parameter in a water dis-
tribution system is pressure. For instance, pressure is a function of
depth within a tank. The pressure is the same at two points that are
at the same fluid depth regardless of the shape or volume of the tank.
Pascal’s law states that the pressure at any point inside a tank has
the same magnitude in all directions. In other words, pressure is a
scalar quantity, not a vector. Pressure is always perpendicular to a
submerged surface regardless of the surface’s shape or orientation.
Tank 1
Treatment
plant
Tank 2
Tank 3
FIGURE 9-4 Node representation of a distribution system network; arrows

indicate average pipe segment flow.

Operation
Operation 299
Hydraulic grade line

Energy grade line
V 2/2g
Total
dynamic
head Ground
H static tank
Elevated
tank
Trunk and
distribution
water mains
Datum
Pump
H atmospheric
Absolute zero pressure
FIGURE 9-5 Distribution system energy lines. (Note: V 2/2g = velocity head.)
The energy head at any point in the distribution system is calcu-

lated using the Bernoulli equation:
P V2
H= + +Z
2gc
Each point in the system has an energy level, H. The energy grade
line (EGL) is a graph of the total energy versus position in a pipeline
from a reference line or datum. Friction losses in the pipe cause the
EGL to slope downward in the direction of flow. Since friction in
the pipe is a function of fluid velocity, the faster the flow, the steeper
the EGL. The hydraulic grade line (HGL) is a graph of the sum of the
pressure and potential energies versus position in a distribution sys-
tem from the reference line or datum. The HGL level at any point in
the distribution system is equal to the height a water column would
rise inside a vertical conduit open to the atmosphere. It can also rep-
resent the height of the water in a tank connected to the system at that
point. The kinetic energy component is the difference between the EGL
and the HGL. Figure 9-5 shows distribution system energy lines.
In most water distribution applications, the elevation and pressure
head terms are much greater than the velocity head term. For this
reason, velocity head is often ignored, and modelers work in terms of
hydraulic grades rather than energy grades. Therefore, given a datum
elevation and an HGL, the pressure can be determined as (Walski,
Chase, and Savic 2001):
P = (HGL − Z)
Both the EGL and the HGL increase at the location of a pump
within the pipeline by an amount equal to the head added by pump-
ing. The head added by pumping (h A ) may be calculated if the flow

Operation
300 Chapter Nine
velocity and gauge pressure are known for the suction and discharge
pump nozzles.

P V2 P V2
hA = + +Z − + +Z
2gc discharge 2gc suction
The head added by the pump is a function of the flow rate through
the pump. A graph of pump head versus flow rate is called the pump
curve. The plotted head is the head difference across the pump and is
given the name total dynamic head (TDH).
Traditional hydraulic modeling provided an insight into how the
water level in tanks changed under simulated demand conditions.
However, the water in the tank was considered of invariable quality.
Hence, there was no difference in quality whether the water was en-
tering the tank, inside it, or leaving it. In the past decade, increased
emphasis on water quality in storage facilities prompted more detailed
investigations of water dynamics in tanks.
Modeling of water flow regimes and water quality in tanks has
been done using multiple methods:
r Scaled laboratory models
r Tracer studies
r Semiempirical mathematical models
r Computational fluid mechanics using finite-element or finite-
volume models
Scaled Laboratory Models

For more than a century, scaled models have been used to validate
hydraulic structures, vessels, and tank designs. Geometrically similar
scaled-down models are tested under laboratory conditions to sim-
ulate actual operating conditions involving fluid flow phenomena,
since governing equations regarding the latter do not have exact math-
ematical solutions. The laws governing tank scale-model construction
are known as principles of similitude. These principles make it possi-
ble to construct models that accurately represent actual tank perfor-
mance.
The main dimensionless parameters used in scale models are
Froude number
U
Fr = √
Lg
Reynolds number
UL
Re =


Operation
Operation 301
Weber number
U 2 L
We =

where U is the free stream velocity of the system; L is the characteristic

length, diameters, or depths; g is the gravitational acceleration; is
water density; is the viscosity of water; and is the surface tension
of water.
A tank and a geometrically similar model are in similitude if Re ,Fr ,
and We are the same. The scaling of the model, however, is based on the
predominant parameter for each particular application. The Froude
number is generally used in the scaling of water-storage facilities. A
scale model should be as large as possible to minimize scaling errors
on tank performance. Many sources quote typical scale ratios for tank
models in the range of 1:30 to 1:50.
Chemical tracers or dyes are used to visualize streamlines in the
model. However, most sophisticated technology, such as optical tech-
nology and lasers, has been used to quantitatively measure scalar
values. The objective of these tests is to trace water streamlines from
the inlet as they mix with ambient water in the vessel and exit the tank.
The tracer is monitored at the outlet and compared with a constant
influent concentration over time to determine how much time water
spends in the tank (residence time).
The effects of temperature on the buoyancy of influent water are
of particular interest for tank modeling. Temperature distribution in
tanks can be measured at different locations and depths in the model
with temperature probes.
Tracer Studies
Tracer studies are conducted in water storage tanks to determine res-
idence time and/or water quality distribution (Fig. 9-6). Tracers may
be chemicals or dyes that can be tracked or measured by ion-specific
electrodes, conductivity probes, colorimeters, or visual/camera ob-
servation. A tracer can be introduced at a known concentration by a
dosing pump at the tank’s inlet piping. Typical tracers include chloride
ions (such as sodium chloride, calcium chloride, and lithium chloride),
fluoride ions, and the fluorescent dye Rhodamine WT. Local regula-
tory agencies must approve the use of a particular tracing chemical
for studies of actual potable water tanks. The tracer should not be
consumed or removed during treatment. Fluoride ions are not typ-
ically present naturally in water. Therefore, fluoride can be used in
lower concentrations than chloride tracers. Rhodamine WT must be
used following certain guidelines found in Appendix D of the August
1999 US Environmental Protection Agency (USEPA) Disinfection Pro-
filing and Benchmarking Guidance Manual (USEPA 1999). Selection of a

Operation
302 Chapter Nine
24
22
Chloride ion concentration (mg/L)
20
18
16
14
12
10
0 1 2 3 4 5
Time (h)
FIGURE 9-6 Tracer study results of live pass-through distribution system tank
with plug flow to determine contact time.
particular chemical tracer may depend on the salt concentrations

present in the water. Specific instructions on chemical tracers and the
conditions under which they are most effective are found in Appendix
D of the Guidance Manual. If a tracer study is needed to find the con-
tact time (CT), a water system should consult the latest tracer study
guidance from its regulatory agency.
The tracer chemical should be added at the same injection points
as the disinfectant to be used in the CT calculations. Tracers are com-
monly added in two ways: the step-dose method and the slug-dose
method. In the step-dose method, the tracer is injected at a con-
stant dosage, and the endpoint concentration is monitored. To acquire
90 percent recovery for the tracer, endpoint sampling should continue
until the tracer concentration reaches steady-state level. In the slug-
dose method, however, a large dose of tracer chemical is instanta-
neously injected. The tracer concentration is then monitored at the
endpoint until the entire dose (slug) has passed through the tank. Fig-
ure 9-7 shows apparatus to measure CaCl tracer concentration. A mass
balance is required for this method to determine whether the entire
tracer dose was recovered. Contact time is then determined mathe-
matically from the concentration versus time profile. A tracer study is
generally done as follows:
r Determine the flow rate or rates to be used in the study.
r Select the tracer chemical and determine the raw water back-
ground concentration of the tracer chemical. This is needed

Operation
Operation 303
FIGURE 9-7 Apparatus to measure CaCl tracer concentration at tank inlet and
outlet. Ion-specific probes can be seen at left. Data were recorded on a PC.
to determine the dosage of chemical to feed and to properly

evaluate data.
r Determine the appropriate tracer dosage.
r Determine tracer addition locations, sample collection logis-
tics, and sampling frequency. Sampling frequency depends on
the size of the tank—the larger the tank, the less frequent sam-
pling is needed, but the longer the duration of the sampling
event.
r Conduct the tracer test using either the step-dose or slug-dose
method.
r Compile and analyze the data.
r Calculate CT.
Semiempirical Mathematical Models

These models interpolate or extrapolate experimental data using sim-
plistic mathematical relationships to predict tank performance on
the basis of inlet and outlet operational conditions and simplified

Operation
304 Chapter Nine
geometric parameters. Some of these models have an extensive

database of experimental data so that they can produce highly accu-
rate results if the operational conditions are within the data’s range.
These models have been adopted by the USEPA for outfall modeling
of discharges into bodies of water.
Computational Fluid Mechanics Using Finite-Element or

Finite-Volume Models
A multitude of computational fluid dynamics (CFD) software pro-
grams are commercially available. Although the mathematical algo-
rithms might differ slightly, they all predict fluid behavior in vessels
and around structures by solving the Navier–Stokes equations. These
algorithms use a multitude of turbulence models to predict time-
averaged turbulent flow characteristics. Often, the software packages
offer a choice of turbulence models to the modelers. The model cho-
sen affects the accuracy of the solution depending on the phenomena
involved in each particular study.
Grid generation is another critical software component that can
drastically affect accuracy and convergence. The modeler must de-
termine the regions where finer grids are required and track conver-
gence at critical locations. Models differ in their ability to handle com-
plex geometries, curved surfaces, inlets/outlets, buoyancy effects, and
boundary conditions such as free boundaries (e.g., changing level at
the water-to-air boundary) (Fig. 9-8). Transient flows, separation and
FIGURE 9-8 Computational fluid dynamics model of reservoir with two inlets
and two outlets showing velocity contours. Light gray is highest velocity and
dark gray is stagnant.

Operation
Operation 305
reattachment, vortices, and swirl are not accurately predicted by cur-

rent mathematical procedures. Specialized application software pack-
ages are available to predict such phenomena, but they are limited in
application and cannot easily be generalized due to their reliance on
numerous empirical and semiempirical equations. These limitations
allow general-purpose CFD modeling to be more useful for baffling
and plug flow optimization than mixing, for example. Some packages
include dispersion models or the ability to incorporate transport equa-
tions programmed by the modeler. Nevertheless, the capability of the
packages to produce accurate predictions is limited by the mathemat-
ical constraints of modeling curvilinear flow patterns and fluctuating
or time-dependent phenomena. Their capability is also limited by the
modeler’s grasp of advanced fluid mechanics principles and his or
her ability to detect an improbable or inaccurate result. Otherwise,
the phrase “garbage in, garbage out” may apply. Many CFD users
have learned that validating models against real-world testing or ex-
perimental data is needed to prevent faulty or inaccurate solutions.
CFD is a step in the design process that minimizes trial and error, but
it is not in itself a validation of design.
Water Quality Issues for Water Storage Tanks in

Distribution Systems
The quality of potable water in the United States is defined by regula-
tions that govern all facets of water treatment and distribution. These
include, among others, means of filtration and disinfection, reduc-
tion or elimination of contaminants, control of taste and odor, back-
flow prevention, and distribution system minimum pressures. The
USEPA’s Division of Drinking Water has the primary charge of es-
tablishing and enforcing these regulations. Regulations set forth by
the Safe Drinking Water Act (SDWA) directly affect the operation of
water distribution systems. “The intent of the SDWA is for each state
to accept primary enforcement responsibility (primacy) for the oper-
ation of the state’s drinking water program. Under the provisions of
the delegation, the state must establish requirements for public wa-
ter systems that are at least as stringent as those set by the USEPA.
In some states, the primacy agency is the health department, and
in others it is the state’s environmental protection agency, depart-
ment of natural resources, or pollution control agency” (Von Huben
1999).
Amendments to the SDWA made in 1996 included the requirement
that water utilities create and distribute consumer confidence reports
(CCRs). The yearly CCR includes information regarding compliance
with maximum contaminant levels (MCLs) and treatment techniques
for drinking water. Utilities serving more than 10,000 people were

Operation
306 Chapter Nine
FIGURE 9-9 Chlorine residual analyzers monitor inlet and outlet disinfectant
residual at ground storage tank.
mandated to mail the CCR directly to their customers. The report

discloses contaminant levels in a community’s drinking water supply
alongside MCLs and maximum contaminant level goals (MCLGs). It
also tells what type and concentration of disinfectant, such as chlorine
or chloramines, is used by the utility.
Whenever a disinfectant molecule reacts with organic or inorganic
molecules in the water, it forms what are commonly known as disinfec-
tion by-products (DBPs). DBPs are a collection of chemicals, including
trihalomethanes (THMs) and haloacetic acids (HAA5), that are highly
soluble in water. Animal testing has demonstrated that some of these
compounds have carcinogenic properties. Other disinfectants, such as
chloramines, chlorine dioxide, and ozone, may yield DBPs upon react-
ing with organic or inorganic molecules. The reaction products from
such interactions are similar to those produced by chlorine, but they
tend to occur in smaller concentrations. Figure 9-9 illustrates a chlorine
residual analyzer. The USEPA announced in December 1998 the Dis-
infectant/Disinfection By-Product Rule, Stage 1 (D/DBP). This rule
created the following MCLs: total THMs, 80 g/L; HAAs, 60 g/L;
and bromate, 10 g/L. The rule also established MCLs for disinfec-
tants: chlorine, 4 mg/L; chloramines, 4 mg/L; and chlorine dioxide,
0.8 mg/L (AWWA 2003 Principles and Practices).
The USEPA published the Stage 2 Disinfectant/Disinfection By-
Product Rule (Stage 2 DBPR) on January 4, 2006. The Stage 2 DBPR

Operation
Operation 307
builds on existing regulations by requiring water systems to meet

DBP MCLs at each monitoring site in the distribution system to better
protect public health.
The Stage 2 DBPR includes a provision requiring all community
water systems (CWSs) and only nontransient noncommunity wa-
ter systems (NTNCWS) serving more than 10,000 people to conduct
an initial distribution system evaluation (IDSE). NTNCWSs serving
fewer than 10,000 are exempted from IDSE requirements, but will
need to comply with the Stage 2 DBPR compliance monitoring re-
quirements. The goal of the IDSE is to characterize the distribution
system and identify monitoring sites where customers may be ex-
posed to high levels of total trihalomethanes (TTHM) and haloacetic
acids.
Compliance with the IDSE requirements includes the modeling
or monitoring of distribution system site on the basis of criteria out-
lined by the EPA that include all water-storage facilities. Utilities are
required to report trended averages per site and address MCL viola-
tions.
The Total Coliform Rule mandates monthly monitoring of distri-
bution system samples for coliform bacteria. A fecal or Escherichia coli
bacteria test must be performed on any positive sample. Additional
samples and analyses are required at the positive-sample point as well
as in its vicinity within 24 hours. Utilities must submit a sampling plan,
and samples must be taken at the customers’ taps or from taps that
are representative of the distribution system. The Total Coliform Rule
does not require the testing of water from storage facilities or their
outlet pipes.
The Lead and Copper Rule (LCR) mandated desktop studies to
determine the need for corrective action regarding lead and copper
concentrations. Many utilities discovered that they required the ad-
dition of corrosion-inhibiting agents to arrest the release of copper
or lead ions from piping materials into the water. Action levels for
copper and lead were set at 1.3 mg/L and 0.015 mg/L, respectively.
Pilot studies were conducted by utilities to identify the most suitable
inhibitor for their system. Orthophosphate and phosphoric acid are
popular choices. Stagnant areas of distribution systems, such as dead
ends or poorly mixed tanks, readily lose their residual concentration
over time. Residual phosphate compounds used as inhibitors do not
readily dissipate like disinfectants do. In the absence of sufficient dis-
infectant, these compounds can foster the development of biological
agents such as algae and bacteria. Regular flushing of dead ends in
distribution systems and appropriate tank design and operation are
needed to eliminate this potential problem.
The water quality in distribution systems’ water storage tanks and
reservoirs is affected by several other factors. Some likely issues, their
cause, and suggested solution are listed in Table 9-1.

308
Issue Possible Cause(s) Suggested Solution(s)
Low disinfectant residual r Excessive water transport time to r Investigate low flow or demand for size of mains
in water entering the tank r Boost disinfectant in large/multizoned distribution systems
storage facility r High biological or chemical r Find valve issues in trunk mains
demand in distribution system r Biofilms on interior pipe surface are difficult to eliminate,
r Distribution system backflow or prevention is best choice: adequate pipe velocity,
pipe integrity issues disinfection, and flushing
r Number and size of tanks in r Improve disinfection and flushing after main repair or
series upstream of facility and installation
their water quality issues r Eliminate dead ends in distribution system
r Find and repair main breaks, leaks, and backflow issues
r Solve issues at upstream tanks, if any
r Boost disinfectant
Operation
Nitrification r Chloramines r Ensure adequate chloramines residual throughout the system

(2.0 to 2.5 mg/L)
r Reduce water age by avoiding excessive storage residence
time
r Use chlorine:ammonia:nitrogen ratio of 4.5:1
Inadequate tank turnover r Low inflow and/or outflow r Hydraulic analysis to investigate proper distribution piping,
r Low tank volume change over pumping, or storage requirements

time r Revisit tank location and elevation

r Low normal demand versus tank r Excessive pump operation
size
r Improper tank location, elevation,
or size

Stratification r Tank geometry and location of r Increase tank turnover
inlets and outlets r Improve tank mixing
r Buoyancy caused by temperature r Properly locate inlets and outlets
difference between tank ambient
and influent water
Loss of disinfectant r Accumulation of sediments: r Periodic cleaning of tank sediments
residual in distribution chemical and biological demand r Eliminate tank elements that allow sediments to accumulate
system tanks r Exposure to ultraviolet (UV) light or provide habitat for biofilms
(open tanks) r Install appropriate tank cover
r Diffusion due to ambient air r At tank design, decrease ratio between water surface area
temperature and tank volume
r Corrosion of tank metals r Design venting to prevent excessive air drafts above water
r Excessive water tank residence surface
time r Conduct periodic tank maintenance and rehabilitation
r See Stratification
Operation
Biofilms r Availability of nutrients in water or r Perform periodic tank cleanings

tank surfaces r Monitor condition of vents and hatches to prevent biological
r Low disinfectant residuals pollutants from entering tank
r Nitrification r See Loss of Disinfectant Residual
r Warm temperatures r See Nitrification
r Corrosion inhibitors r Improve tank turnover and mixing
r Tank coatings r Replace with NSF61-approved coatings that do not promote
biogrowth; minimize crevices, gaps, and substrate

imperfections before applying coatings to minimize biofilm
habitat
TABLE 9-1 Water Quality: Issues, Causes, and Suggested Solutions (Continued)
309
310
Issue Possible Cause(s) Suggested Solution(s)
Short-circuiting r Tank geometry and location of r Provide tank mixing
inlets and outlets r Provide baffling for pass-through tanks (if inlet and outlet are
connected to different pressure zones with little or no
backflow)
Regions of stagnant r Tank geometry and location of r At design, optimize ratios of tank surface area to volume and
water in tank inlets and outlets of height to diameter
r Short-circuiting r Locate inlets and outlets such that the largest volume of
r Stratification water in the tank is contained between them
r See Short-Circuiting
r See Stratification
Coating leachates r Insufficient coating cure r Follow manufacturer’s recommendations for cure time versus
r Improper coating component substrate surface temperature
Operation
mixing or mixing ratios r Follow manufacturer’s recommendations for component

r Not NSF61-approved coatings mixing and use factory-measured quantities
r Use NSF61-approved coatings
Effects of metal r Improper coating or coating r Consult coating specialist for proper coating selection and
corrosion in contact application surface preparation
with the water, such as r Ice damage r See Effects of Ice
red water, iron-reducing r Presence of sharp edges, r Ensure proper surface preparation during maintenance or
bacteria, and high cavities, or gaps that prevent rehab

turbidity proper coating or cause coating r Follow periodic and comprehensive tank inspection and

failure maintenance program
r Inadequate tank inspection and
maintenance program

Contamination, r Improper vent and overflow r Provide and maintain vent and overflow screening
intentional or screening r Provide secure hatches and proper tank covers to prevent
unintentional, through r Improper hatches animals, insects, vandals, or rain from entering tank
tank openings such as r Tank cover r Provide proper air gaps
vents and hatches r Improper tank overflow and drain
Chemical composition of r Metals come out of solution and r Periodically clean tank sediments
precipitates in the settle inside tanks, affecting
storage facility, such water quality in high
as iron and concentrations
manganese
Effects of ice including r Winter conditions in colder r Provide insulation and heating of a tank for surge tanks
Operation
tank damage; climates r Provide tank cover

encapsulation and r Thermal stratification r At design, bury or partly bury tanks
later release of r See Stratification
contaminants, debris,
and coating particles
upon meltdown
Note: NSF = National Sanitation Foundation.

TABLE 9-1 Water Quality: Issues, Causes, and Suggested Solutions (Continued)
311
Operation
312 Chapter Nine
Water Quality Monitoring in Distribution System Tanks

The importance of monitoring the quality of water in storage facili-
ties is fast becoming a focus of interest. Although no US regulations
require the monitoring of water quality parameters in storage tanks,
there are several important reasons why utilities should maintain a
monitoring program for tanks. Other countries, such as the United
Kingdom, mandate frequent monitoring of tank water quality.
Tanks are an intermediate stage between the water treatment
plants and customers. Depending on the size of a distribution system,
demand, and other factors, water in tanks could have left the treatment
process from several minutes to several days, or even longer, before it
entered the tank. Testing of water quality parameters in storage pro-
vides a utility with valuable information regarding what happens to
the water as it moves through the distribution system en route to the
customer. Testing can also indicate whether detrimental occurrences
are taking place in the storage facility itself. Sampling data can also
be used in the determination of tank mixing efficiency and turnover.
Security Concerns
The events of September 11, 2001, raised questions about the safety
of the nation’s water supply. In general, water tanks are accessible
through access hatches, air vents, overflows, and access ladders. These
entry points can be tempting targets for vandals or criminals attempt-
ing to contaminate the water supply. They are also potential sources
of accidental contamination from rainwater, birds, insects, and other
natural sources. Continuous online monitoring of water quality pa-
rameters can provide an indication or alarm should contamination
occur. The affected tank can be isolated from the system, and author-
ities and the public can be notified. Figure 9-10 illustrates protective
measures for water tanks.
Tank-Monitoring Program
The tanks themselves can be a source of concern. Leaching from coat-
ings, red water and bacteriological concerns from rusting wet surfaces
or overhead structural elements, accumulation of sediments that may
contain biofilms, settled metal particles, or other detriments to water
quality may be present in storage. Some may engender complaints
about the water’s taste, odor, or appearance.
A tank-monitoring program provides a utility with background
data of tank parameters for different seasons and weather conditions;
it helps establish schedules for tank inspection and maintenance; and it
can serve as a guide to water quality managers for planning treatment
strategies for changing distribution system conditions.
Although no regulatory requirements in the United States man-
date tank monitoring, there are regulations that require periodic water

Operation
Operation 313
FIGURE 9-10 Tank overflow and catch basin are protected from insects or
vandals by combination of stainless-steel cage and insect screen; flow switch
(not shown) triggers overflow alarm.
testing in distribution systems. The National Primary Drinking Water

Regulations (CFR 141.23), the Surface Water Treatment Rule (SWTR),
the Lead and Copper Rule (LCR), the Total Coliform Rule, and Stage
1 of the D/DBP all establish contaminant restrictions and monitor-
ing requirements for distribution systems. Table 9-2 provides water
quality parameters and related regulatory information.
Regulations cover such important contaminants as total and fecal
coliform bacteria and heterotrophic bacteria, disinfection by-products
such as trihalomethanes (THMs) and haloacetic acids (HAAs), nitrite,
nitrate, lead, and copper. They also limit the concentration in the sys-
tem of residual disinfectants such as chlorine or chloramines.
The USEPA has identified several secondary contaminants
(nonenforceable) under the National Secondary Drinking Water Regu-
lations (NSDWR). The enforceability of this rule is up to the individual
states’ regulatory agencies. Table 9-3 lists the secondary nonenforce-
able contaminants.
At a minimum, utilities should attempt to monitor disinfectant
residual, heterotrophic plate count (HPC), and coliform bacteria in
water-storage facilities. Suspected biofilms on tank surfaces should
be sampled and speciated to determine source or cause. An ideal way
to test a tank for biofilms is to prepare a metal or concrete coupon with
a coating and finish similar to those of the tank. The coupon can be

Parameter Sample Location Regulatory Limit Reference Comments
Disinfectant residual Entry point to 0.2 mg/L on a continuous basis Surface Water Treatment Rule Only applies to systems using
distribution system surface-water supplies
314
Disinfectant residual Distribution system Maximum residual disinfectant levels D/DBP Rule, Stage 1 Surface-water systems serving
(MRDLs): chlorine, 4.0 mg/L; more than 10,000 people
chloramines, 4.0 mg/L; running
annual average
Disinfectant residual Throughout Detectable levels of disinfectant Surface Water Treatment Rule Only applies to systems using
or heterotrophic distribution system residual or HPC bacteria count of 500 surface-water supplies; in the
plate count (HPC) or fewer colony-forming units (cfu) per United States, Legionella is
bacteria count milliliter in 95 percent of samples also regulated by a treatment
collected each month for any 2 technique
consecutive months
Nitrite Entry point to 1.0 mg/L as N National Primary Drinking Water Maximum contaminant level
distribution system Regulations (NPDWR), Code of
Federal Regulations (CFR) 141.23
Nitrate Entry point to 10.0 mg/L as N NPDWR, CFR 141.23 Maximum contaminant level
distribution system
Operation
Total Throughout 80 g/L, running annual average based D/DBP Rule, Stage 1 Systems serving more than
trihalomethanes distribution system on quarterly samples 10,000 people
Haloacetic acids Throughout 60 g/L, running annual average based D/DBP Rule, Stage 1 Systems serving more than
(sum of five) distribution system on quarterly samples 10,000 people
Total coliform Throughout 0 cfu in 95 percent of samples Total Coliform Rule Number of samples determined
bacteria distribution system by population served
Lead Customer’s tap 0.015 mg/L Lead and Copper Rule Action level at 90th percentile
Copper Customer’s tap 1.3 mg/L Lead and Copper Rule Action level at 90th percentile

pH Throughout the 7.0 minimum pH units Lead and Copper Rule Unless state determines
distribution system otherwise
Source: Adapted from Kirmeyer et al. (1999; p. 37, Table 2.1).

TABLE 9-2 Contaminant Restrictions and Monitoring Requirements for Distribution Systems
Operation
Operation 315
Secondary Maximum
Parameter Conditions Contaminant Level
Aluminum Colored water 0.05–0.2 mg/L
Chloride Salty taste 250 mg/L
Color Visible tint 15 color units
Copper Metallic taste, blue-green 1.0 mg/L
stain
Corrosivity Metallic taste, corrosion, Noncorrosive
fixture staining
Fluoride Tooth discoloration 2 mg/L
Foaming Frothy, cloudy, bitter taste, 0.5 mg/L
agents odor
Iron Rusty color, sediment, 0.3 mg/L
metallic taste, reddish or
orange staining
Manganese Black to brown color, black 0.05 mg/L
staining, bitter metallic
taste
Odor Rotten-egg, musty, or 3 TON (threshold
chemical smell odor number)
pH Low pH: bitter metallic 6.5–8.5
taste, corrosion
High pH: slippery feel, soda
taste, deposits
Silver Skin discoloration, graying 0.10 mg/L
of the white part of the
eye
Sulfate Salty taste 250 mg/L
Total Hardness, deposits, 500 mg/L
dissolved colored water, staining,
solids (TDS) salty taste
Zinc Metallic taste 5 mg/L
TABLE 9-3 Secondary Nonenforceable Contaminants in Water Distribution

Systems

Operation
316 Chapter Nine
Parameter Indication Parameter Is Present

Iron oxide Distribution system corrosion
Aluminum hydroxides Excess aluminum after flocculation
Calcium carbonates Supersaturation of minerals in hard waters
Manganese Source-water problem
Heterotrophic plate Taste and odor problems; recurring bacterial
count (HPC) counts
Depth of sediment Rate of accumulation; disinfectant residual
loss due to resuspension
Gross microbial System cross-connection, poor hydraulic
examination circulation, or failed facility vent screening
Source: Adapted from Kirmeyer et al. (1999).
TABLE 9-4 Sediment Monitoring Parameters
fastened to a wet tank surface so that its conditions mimic those of the
tank walls. In this manner, the coupon can be retrieved from the tank
for analysis without interfering with tank function.
Samples from tank bottom sediment may be collected to determine
chemical composition or biological activity. Kirmeyer et al. (1999) rec-
ommend the sediment monitoring parameters outlined in Table 9-4.
The need to monitor other bulk water parameters varies depend-
ing on the type of tank, source water, environmental conditions, and
so on. Monitoring the amount of sediment accumulated in the tank
bottom can give a utility an indication when the next cleaning cycle
should take place. Table 9-5 lists water quality monitoring parameters
recommended for storage facilities.
Several nitrification tests should be performed where chloramines
are used for secondary disinfection. As a minimum, tests for hetero-
trophic plate counts (HPC), nitrite, nitrate, ammonia, and chlorine
species can help to ensure that optimal conditions are maintained in
storage facilities when chloramination is used.
Sampling Methods and Equipment

The main issue regarding monitoring of water quality parameters is
access to adequate sampling points. The tools used depend on the
type of sample being retrieved, available access to the tank, and the
utility’s budget. Regardless of method, technicians must be conscious
of water quality concerns regarding accidental contamination. The
materials for the sampling equipment must be compatible with use in
potable water, if applicable.

Operation
Operation 317
Sampling Procedure
Parameter Purpose Used
Alkalinity Indicates potential buffering On-line ion-selective
capacity (pH stability) electrode or grab
sample and
laboratory analysis
Aluminum Indicates potential On-line ion-selective
coagulant overfeeding electrode or grab
sample and
laboratory analysis
Ammonia, Indicates potential for On-line ion-selective
total and/or nitrification electrode or grab
free sample and
laboratory analysis
Chlorine Indicates protection from On-line chlorine
residual, bacterial growth and residual analyzer
total and/or provides early warning or grab sample and
free sign of water quality amperometric
deterioration; monitored titration laboratory
at inlet and outlet to analysis
control rechlorination
when practiced
Coliform, total Indicates the presence of Grab sample and
and/or indicator bacteria laboratory analysis
fecal
Conductivity, Can quickly indicate relative On-line ion-selective
specific changes in total electrode or grab
dissolved solids (e.g., sample and
alkalinity) laboratory analysis
Disinfection Represents potential for Grab sample and
by-products ongoing chemical laboratory analysis
reactions and DBP
formation
Heterotrophic Indicates conformance to Grab sample and
bacteria MCL; provides early laboratory analysis
warning sign of water
quality deterioration
Iron Indicates potential On-line ion-selective
corrosion reactions electrode or grab
sample and
laboratory analysis
TABLE 9-5 Water Quality Monitoring Parameters for Storage Facilities

(Continued)

Operation
318 Chapter Nine
Sampling Procedure
Parameter Purpose Used
Nitrate Indicates possibility of On-line ion-selective
nitrification electrode or grab
sample and
laboratory analysis
Nitrite Indicates possibility of On-line ion-selective
nitrification electrode or grab
sample and
laboratory analysis
pH Indicates changes from the On-line ion-selective
water source; indication electrode or grab
of corrosion of concrete sample and
or an unlined new facility laboratory analysis
Taste and Evidence of water quality Grab sample and
odor problem in progress laboratory analysis
Temperature/ Differences in storage On-line sensor
temperature facility indicate possible
profile stratification and
stagnant zones—early
warning sign of potential
microbial problems
Turbidity Provides early warning sign On-line turbidimeter
of water quality sensor and
deterioration analyzer
TABLE 9-5 Water Quality Monitoring Parameters for Storage Facilities

(Continued)
Sampling is divided into two categories: grab samples and contin-

uous sampling. Grab samples are small volumes of tank water man-
ually collected by a technician to be analyzed either in the field or
under laboratory conditions. Continuous sampling is accomplished
by means of electronic equipment such as ion-specific probes, temper-
ature probes, or on-line colorimetric chemical analyzers. The sensors
are in the tank water, or tank water is continuously or periodically
piped to the sensor. The data are recorded locally either on paper or
electronically, or they are transmitted via telemetry to the utility’s data
storage and monitoring facility.
Kirmeyer et al. state that
a [monitoring] program that takes all necessary parameters into account
and schedules sampling when needed to provide adequate information
is a conceptual starting point. A recommended approach. . .is to further

Operation
Operation 319
break down sampling needs into two categories of routine sampling and
diagnostic sampling. Routine sampling is defined as monitoring of param-
eters on a regular (continuous, weekly, monthly) basis and may include
regulatory and operating parameters. Diagnostic sampling is defined as
special purpose monitoring to document condition or to determine the
cause of a problem. Routine monitoring is used to document general water
quality conditions, whereas diagnostic monitoring is often more facility
specific. Diagnostic monitoring can first provide a baseline for a storage
facility and identify problems. After the problems are corrected, routine
monitoring can then be used to track tank conditions to detect or prevent
the recurrence of the problems or the beginning of new ones.
The types of data being collected or the purpose for the data col-
lection dictates whether grab or continuous sampling is required for
a particular parameter.
Grab Samples
In theory, technicians using grab sampling techniques can collect ev-
ery type of parameter. Certain types of tank-water tests such as het-
erotrophic plate count, total coliform bacteria, chemical composition
of tank sediments, biofilm analysis, and disinfection by-products can
only be performed on manually collected tank water, as they require
analysis under laboratory conditions. A rare exception is when a tank
is located next to the utility’s laboratory and sampling lines exist di-
rectly between the tank and the laboratory. The laboratory technician
or chemist is then able to fill the test vial directly from the sampling tap.
In most cases, technicians must travel to several tanks in remote lo-
cations as often as the monitoring program dictates the taking of grab
samples. Under the best of circumstances, the sampling program in-
cludes the installation of permanently and properly located sampling
taps on the tanks. The taps can be connected through small-diameter
piping to several sampling locations inside the tank. This scheme can
provide a more or less three-dimensional view of the tank’s water
quality parameters. The more sampling taps are installed, the more
complete the water quality picture will be. The ideal locations for sam-
ple points are the inlet(s) and outlet(s), if any; locations that may be
subject to short-circuiting or stagnation, such as regions not in the
general path between the tank’s inlet and the outlet; and at several
depths to test for the effects or the presence of stratification.
The sampling lines can be designed such that the water will flow
out of the tank by gravity or be pumped out with the use of a jet (vac-
uum) or metering pump. Technicians must flush the sampling lines
for sufficient time to provide a proper sample from the tank water
at the sampling point. Twenty minutes of flushing is usually suffi-
cient. Some systems allow for continuous flow through the sampling
lines, which facilitates this step. Others are designed so that the sam-
pling lines can be periodically disinfected and flushed. Permanently

Operation
320 Chapter Nine
installed sampling lines are a security concern, as they may provide

easy access for intentional contamination of the tank water. Sanitation
of the sampling collection area and frost proofing are also concerns.
Sampling lines create an environment in which a relatively small vol-
ume of tank water is enclosed by a relatively large surface area. This
condition is favorable to the formation of biofilms inside the sam-
pling line. Technicians should monitor test results such as elevated
bacterial counts or turbidity that may show such a problem exists.
In general, sampling taps should be enclosed and protected from the
elements. Pump station buildings, insulated enclosures, and heated
underground vaults are all potential locations for sampling taps. (The
vaults are not so convenient if confined space is a concern.)
Most tanks, however, are not fitted with sampling taps, and tech-
nicians must use roof hatches to gain access to the tank. This implies
that the technician must be fit, trained, and equipped to safely climb
the tank and open the hatch and that the tank meets all Occupational
Safety and Health Administration (OSHA) requirements for climb-
ing safety. Several depth sample collection probes and tubes, such as
weighted collection bottles with string-operated trap doors or rods
with check valves, can be used to collect samples at different depths
(Fig. 9-11). Although samples can be taken at various water depths
using these devices, the number of sampling locations is limited by
the location and number of roof hatches.
The sampling equipment, including tethers, must be adequately
disinfected before it is introduced into the tank. Care must be observed
FIGURE 9-11 Grab sample retrieved from ground tank using calibrated depth
sampling tube with check valves.

Operation
Operation 321
so that the equipment disinfection process is compatible with the pa-

rameter being tested. For instance, disinfecting a sample bottle and
tether with 200 ppm of chlorine and then lowering the bottle into the
tank to retrieve a chlorine residual sample may not yield a satisfactory
result. The water must be collected in a properly prepared and labeled
bottle suitable for the type of sample being prepared, and it must be
transported to the laboratory in a timely and sanitary manner to pre-
serve the validity of the test results. Some tests, such as colorimetric
chlorine residual, may be performed by the technician on site, as there
is a chance that the residual concentration may decrease in transit.
More than one technician may perform grab samples. Proper train-
ing in sample collection and the use of calibrated tethers or rods helps
maintain consistency among technicians and samples. Both on-site
and laboratory analyses should be performed by a certified water
quality laboratory using the methods specified in Standard Methods
for the Examination of Water and Wastewater and the USEPA’s Manual of
Methods for Chemical Analysis of Water and Wastes.
Continuous Sampling
Climbing tanks to take grab samples can be dangerous or impossible,
especially during rainy or cold weather. Continuous online sampling,
which relies on technology rather than technicians to collect, analyze,
record, and communicate the data, does not require that operators
climb the tank to retrieve water or that they even be present during
sampling, and its performance is independent from weather condi-
tions. Online sensing probes may be located at various locations inside
the tank, or water from various locations in the tank may be brought
to an analyzer through sampling lines. The analyzer, data recorders,
telemetry, and other equipment should be located in a secure, heated,
and sanitary location to preserve the integrity of the readings. Figure
9-12 shows a pressure transmitter, analyzer, and corrater.
Online sampling technology offers the advantage of a continu-
ous data stream or data collected at relatively brief time intervals,
depending on the type of test, type of instrument, and the utility’s
desired sampling rate for particular parameters. Grab samples reflect
only a momentary condition of the water in the tank, often with no
accurate time reference.
The analyzed data can be stored on site using a personal computer,
data logger, printout, or pen-chart. Operators can retrieve the data in
either electronic or paper form at scheduled intervals that can vary
from daily to yearly, depending on the purpose of the testing. Data can
also be sent to a central data collection and monitoring facility using a
telephone modem connection, remote terminal unit (RTU) with radio
frequency communication capability, cellular technology, broadband
(T lines), satellite, and so on. These methods of communication are

Operation
322 Chapter Nine
FIGURE 9-12 Left to right: tank water elevation (pressure transmitter),

disinfectant residual (analyzer), and water corrosivity (corrater) are monitored
for ground water storage tank.
discussed in the section “Telemetry” of this chapter. The amount of

monitoring data storage is limited by several factors:
r Number of parameters
r Total sampling rate for all the parameters
r Capacity of the data storage device
r Frequency of data retrieval by the utility
r Size of the data packages being stored for each parameter
Online monitoring can also serve a local control function. This

is particularly useful in cases where rechlorination is desirable
(Fig. 9-13). Online chlorine residual analyzers can serve both as mon-
itoring devices for chlorine residual and as controllers for chlorine-
boosting feed systems. For example, when an analyzer at the inlet of
a tank detects that the chlorine residual is below a level accepted by
the utility’s water quality manager for the tank or is in violation of
minimum disinfectant concentration regulations, it can start or adjust
the on-site chlorine feed to match a target concentration. It is ideal
for this process to take place at the inlet of the tank to provide some
contact time for the disinfectant and as a backup safety mechanism for
potential malfunctions of unmanned chlorine-feed systems. Ideally, a
controller for such system should be able to determine whether the

Operation
Operation 323
FIGURE 9-13 Hypochlorite generation system makes disinfectant from brine

for rechlorination at this water-storage facility.
water is inflowing or outflowing from a single inlet/outlet tank; and

the boosting should only take place into the inflow. A second chlorine
residual analyzer monitoring the outflow from a tank with an inlet
and an outlet can be integrated as part of a feedback loop control for
the boosting system and/or can serve as an alarm trigger in case of
overfeeding.
It is important to note that the effectiveness and accuracy of an on-
line monitoring system is only as good as the maintenance it receives.
A calibration schedule must be established for each type of analy-
sis equipment being used. Most ion-specific probes have a life-span
and must be replaced as recommended by the manufacturer. Some
analyzers have moving parts that require cleaning and maintenance.
Chemical solutions used by analyzers should be checked and replen-
ished as needed. Sampling hoses must be kept clean and should be
replaced when fouling is suspected or on a regular schedule. Mainte-
nance technicians first introduced to a monitoring program should be
made aware of these issues and should be appropriately trained.
Water Storage Tank Applications and Their Operation

Water storage tanks are categorized by function. The main categories
are clearwells, distribution system storage tanks, surge relief tanks,
and hydropneumatic tanks.

Operation
324 Chapter Nine
Clearwells
Treatment plants require tanks in which to store treated water be-
fore it is pumped into the distribution system. These tanks are com-
monly known as clearwells. Clearwell sizing is critical to reducing
or eliminating fluctuations in use of filtered water. These tanks also
provide the utility with disinfectant contact time (CT) credit. The
SWTR requires that all surface water treatment facilities provide fil-
tration and disinfection that achieves at least a 99.9 percent (3-log)
removal-inactivation of Giardia lamblia cysts and a 99.99 percent (4-
log) removal-inactivation of enteric viruses. The SWTR assumes that
for effective filtration, a conventional treatment plant achieves 2.5-
log removal of Giardia and a 2-log removal of viruses. Disinfection is
required for the remainder of the removal-deactivation. The amount
of disinfection credit to be awarded is determined with the CT con-
cept, CT being defined as the residual disinfectant concentration (C,
in mg/L) multiplied by the contact time (T10 , in minutes) between
the point of disinfectant application and the point of residual mea-
surement. The SWTR Guidance Manual (USEPA 1991) provides tables
of CT values for several disinfectants that indicate the specific disin-
fection or CT credit awarded for a calculated value of CT (AWWA
1990).
Clearwell design must take into account CT requirements for the
particular treatment process served by the tank. It is a recommenda-
tion of the Ten States Standards (1992) that intermittently operated
filtration plants with automatic high-service pumping from the clear-
well during nontreatment hours provide enough extra clearwell vol-
ume to compensate for depletion of storage during nighttime hours to
ensure adequate disinfectant CT. It is commonplace to design clear-
wells with two or more compartments. One compartment may be
removed from service for maintenance or during times of low de-
mand. Methods for optimizing CT in clearwells will be discussed in
the section on plug flow in this chapter. Figure 9-14 is an aerial view
of a baffled tank optimized for contact time.
Clearwell operation must follow preset parameters for flow and
elevation. It is desired that a clearwell should operate in a condition
as close to steady state as possible. The hydraulic control of raw water
pumping, filter effluent flow, and pumping out into the distribution
system should be coordinated and interlocked, if possible, to prevent
pulsating flows and other transients. Interlocks should also maintain
elevations and should properly shut down during power failure. This
allows for fast recovery when power is restored. To further prevent
flow fluctuations, since most raw water pumps are vertical pumps,
caution should be exercised so that pumps are not operated in unstable
regions of their curves.

Operation
Operation 325
FIGURE 9-14 Aerial view of baffled tank optimized for contact time.
Distribution System Tanks

Distribution system tanks are by far the most common category of
tanks in service. This type of tank serves multiple purposes for water
utilities, including the following:
r Reduce the need for pumping during high water demand
periods (peak shaving)
r Act as pressure relief should a pressure transient occur in a
distribution system
r Provide extra water storage for emergencies such as fires or
power outages
r Help optimize water treatment by allowing treatment plants
to maintain relatively constant treatment rates
r Help reduce the size of trunk and distribution mains
r Act as a sediment trap to settle solids from distribution system
water
r Serve as a landmark, even a source of identity and pride, for
a community
The most important parameters in the design of distribution sys-
tem water storage tanks are location and size. A tank serves a specific
area of a distribution system. Hydraulic engineers, designers, and
modelers must determine how much water must be stored in the tank

Operation
326 Chapter Nine
to serve both typical and emergency capacity demand. The water must
be stored at sufficient elevation to meet the pressure requirements of
the service area. It is important to keep in mind the preservation of
water quality when sizing the tank and determining its operating
characteristics.
Distribution system tank operation is inherently intertwined with
the pumping facilities that feed it, other parallel tanks and their eleva-
tions and capacities, the size of upstream piping, and valve operation.
Each of these is discussed hereafter.
Isolation valves in a distribution system may be inadvertently
closed after repair of a water main break or after a new installation,
for example. This may cause a lower-than-expected pressure at the end
of a system, where tanks are likely located. Such a condition may pre-
vent adequate filling of a tank, increase water transport time between
the pump station and the tank (system residence time), and create
customer dissatisfaction from low pressures. Periodic isolation valve
surveys and valve exercising should be carried out to ensure proper
valve operation and position to prevent isolation valve mishaps. Hy-
draulic modeling is a tool that can be used to assist in the location
of improperly positioned isolation valves, pressure zone boundary
valves, and faulty pressure regulators. All of these may reduce sup-
ply pressure to tanks.
Altitude valves (Fig. 9-15) are installed at the system connection
to tanks. There are two types of altitude valves: single acting and
FIGURE 9-15 Piston-style altitude valves shown here control flow and water
elevation for two adjoining storage tanks.

Operation
Operation 327
double acting. Single-acting altitude valves allow water to enter the

tank when it is less than full. They do not reopen if the system pres-
sure drops; they remain closed until some other connection or a check
valve bypass causes the tank to draft, dropping the water level be-
low the valve’s set point. Double-acting altitude valves are used for
single-inlet/outlet tanks, since they eliminate the need for a check
valve bypass. This reduces installation cost and possible transients
caused by check valve slam. Double-acting valves close when the wa-
ter elevation in the tank reaches a set point, and they open when the
system pressure drops below another set point. Altitude valves can
have hydraulic or electric solenoid pilot controls that allow for throt-
tling inflow and outflow, further controlling fill and draft cycle times.
Altitude valves, in general, have a considerable pressure drop. This
is why many are installed with a bypass containing a remotely con-
trolled valve, such as a motor-operated butterfly valve. Some utility
operators opt to lock the altitude valve in the full open position and
use a motor-operated valve to isolate the tank if the water level gets
too high. Some operators attempt to control the tank level by turning
pumps on and off, throttling pump discharge check valves, or reduc-
ing a variable-speed pump’s revolutions per minute (rpm), essentially
negating the reason for having an altitude valve.
In most water distribution systems, a tank experiences two cycles
of fill and draft in a 24-hour period. The fill cycles usually take place
at noon or midnight, ±3 hours. To optimize power consumption, the
fill and draft periods should be long enough that the fewest number
and the smallest sizes of pumps are needed to satisfy demand. This
also ensures that tank turnover is optimal.
The number and size of pumps operating upstream of a tank can
affect the available pressure at the base of the tank. In most cases,
a pump is run continuously, and a second pump is added to help
satisfy demand during peak hours if the tank cannot supply enough
peak shaving. This number of pumps varies seasonally. Hydraulic
modeling is useful not only during system design, but it can also be
valuable for system operation because it can predict the amount of
pumping required under various system conditions.
If enough pumps are in operation, the tank will be filled with
leftover water after system demand is satisfied. If upstream pumping
is reduced before the tank is full, several things may occur; for instance:
r The tank may not fill at a rate fast enough to provide sufficient
turnover and/or mixing.
r If the fill rate is low enough, the tank may not fill high enough
to help peak shave pumping demand during the next high-
demand period.
r If pumping is reduced excessively, the tank may begin drafting
even though it has not been filled.

Operation
328 Chapter Nine
If the upstream pumping is larger than demand so that the tank

fills at a fast rate, it may cause a different set of events:
r The altitude valve will close to prevent overflow, preventing
the tank from floating on the system. This creates a “hard”
system condition. In other words, the tank is no longer avail-
able to dissipate pressure transients in the system, and the
pressure in the system could change drastically, causing the
pumps to operate at higher pressures that may result in pump
shutdowns. Pump shutdowns cause system pressure surges,
which are further exaggerated by the subsequent opening of
the altitude valve when the system pressure drops below the
valve’s set point. This operation mode is undesirable.
r The high inflow into the tank may fluidize sediments at the
bottom of the tank, and they may reenter the distribution sys-
tem at the next draft, creating water quality concerns.
r High flows and flow reversals may scour water mains; this
also creates water quality issues, such as red water and re-
duction in disinfectant residuals.
Low pressures are seen at a tank when a system outgrows the

capacity of its water mains. Some operators address this issue by in-
creasing pumping at the upstream pump stations, or they add booster
pumps between stations. But this is only a temporary fix that may re-
sult in large pressure variants between the pump station and the tank.
Additionally, the water mains may not provide sufficient flow capac-
ity to fill the tank within the required time.
Oversized water mains do not have a low-pressure problem, but
excessively low flow rates can cause increased distribution system res-
idence time, biofilms, and sediment accumulation, resulting in areas
of high disinfectant decay. The overall effect is a reduction of disinfec-
tant residual available at the tank.
Tanks installed in parallel to serve the same pressure system are
rarely of the same dimensions. Some even have different overflow and
bottom elevations. Further, the ability of the utility to supply water
to each tank varies because of its location, adjacent demands, and
water main size. This results in different fill rates. Hence, the tanks
may not readily cycle in unison. Hydraulic models may help properly
size the mains and valves to new tanks, providing more uniformity
with less throttling. Modeling can also be used to devise elaborate
operational schemes, resulting in similar uniformity in tank use. In
the absence of a hydraulic model, operators often resort to throttling
of inlet and outlet valves to equalize the fill or draft time such that
fill and draft cycles occur simultaneously, preventing the need to shut
off one tank before the others. Although hydraulically it is not critical

Operation
Operation 329
for tanks to cycle in unison, water tank residence time will vary, and
water quality may become inconsistent between facilities in the same
system. However, the effect of differing tank shutoff times in the same
system may render transient travel times not constant. Hence, pump
check valves at pump stations may require a much longer closing or
opening time to prevent pressure surges.
Distribution system water storage tanks should be periodically
taken out of service for cleaning and inspection. Most of the water
in the tank should be allowed to flow back into the distribution sys-
tem. A few feet or meters of water, typically between 3 and 5 ft (0.9
and 1.5 m), are left in the tank to prevent collected sediments from
reentering the system. Debris and sediments are brought to the tank
from the system’s piping and are collected at the bottom of tanks.
This is a desirable means of removing contaminants and particulates
from the water. However, as discussed in the water quality section of
this chapter, several potential biological and chemical water quality
issues are associated with tank sediment. Removal of this sediment is
the only way to eliminate these potential problems. NSF International–
approved chemicals are available that remove persistent, troublesome
stains , such as those caused by manganese and iron deposits. Tanks’
inlets and outlets should be designed such that settled solids are not
fluidized by the inflow and do not return to the distribution system
through the outlet.
In the authors’ experience, a good practice is to allow the wa-
ter main supplying the tank to flush through the tank, using it as a
sediment trap. This can be achieved by emptying the tank and then
opening the tank shutoff valve while the drain is open. This creates
high-velocity scouring of the mains in the vicinity of the tank. A con-
siderable amount of sediment has been removed from tank supply
mains using this scheme. The flushing can be accomplished in as little
as a couple of hours. The drained water must be treated according
to local regulations (dechlorination and solids removal, for example)
prior to disposal.
The tank should be isolated with at least two valves in series with
the tank. All hatches should be open, and the tank drain should be
kept open. After all water and sediment are drained from the tank,
the tank is typically powerwashed prior to inspection. Regulations
require that the tank be disinfected before it is returned to service. The
tank should be disinfected according to the latest version of AWWA
C652, Standard for Disinfection of Water-Storage Facilities. The stan-
dard describes the materials needed, facility preparation, application
of disinfectant to interior surfaces of the tank, and sampling and test-
ing procedures for the detection of coliform bacteria. It also contains
instructions for disinfecting equipment used in on-line underwater
inspections.

Operation
330 Chapter Nine
Three methods of tank disinfection are discussed in the standard.

Only one method is required to be used, although it is possible to
combine two or all three methods, if necessary. The three methods
are
r Full tank chlorination so that, at the end of the retention pe-
riod, a free chlorine residual equal to or greater than 10 mg/L
is achieved.
r Spraying the interior tank surfaces with a solution of 200
mg/L of available chlorine.
r Chlorinating the bottom portion of the tank with 50 mg/L
of available chlorine, then filling the tank to overflow and
maintaining a free chlorine residual of at least 2 mg/L for
24 hours.
The standard should be consulted for more detailed instructions.

According to the Ten States Standards, two or more sets of sam-
ples taken from the tank at 24-hour intervals shall indicate that wa-
ter is microbiologically satisfactory before the facility is placed in
operation.
One of the main functions a tank serves is to dissipate pressure
transients that may take place in the distribution system. When a
tank is taken offline, a distribution system may become vulnerable to
water hammer or other damage stemming from a potential pressure
surge. It is a good practice to plan the installation of temporary pres-
sure relief valves at critical locations in the distribution system for the
duration of the tank work. Insulated enclosures should be provided
along with the surge relief valves in locations subject to freezing con-
ditions. Large customers in the tank’s service area should be notified
and encouraged, if at all possible, not to run processes that demand
a great amount of water while the tank is offline. This is important
because emergency water storage is lost when a tank is taken out
of service. The risk of losing water pressure during a fire is reduced
when large customers temporarily abstain from using large volumes
of water during tank work.
The Federal Aviation Administration (FAA) determines and reg-
ulates structures that may pose an obstruction hazard to air traffic.
Elevated tanks, tanks located at high elevations, and those within an
FAA-designated three-dimensional controlled volume of airspace at
the end of airport runways are required to carry aviation obstruction
markings and/or equipment. The equipment may be an air traffic ob-
struction light at the highest point on the tank. Utility operators and
maintenance crews must monitor the light’s operation and replace
the light when it burns out. Failure to comply with FAA requirements
may lead to fines in the best of cases, or an airplane’s collision with
the tank in the worst case.

Operation
Operation 331
Cold-Weather Operation of Distribution System Tanks

Table 9-6 describes the influence of atmospheric temperature on steel
tanks. Tanks in areas where the lowest one-day mean temperature
(LODMT) is less than 5◦ F (−15◦ C) are likely to experience tank-
freezing conditions. Other factors may contribute to tank freezing even
in areas where the LODMT is above 5◦ F (−15◦ C):
r Elevated tanks with small-diameter (less than 36 in. [0.9 mm])

uninsulated risers
r Tanks with uninsulated and unheated appendages that expe-
rience no flow (e.g., a drain valve at the end of a nipple)
r Tanks experiencing overflow problems because of frozen or
otherwise malfunctioning controls for valves or pumps
r Tanks with abnormally static water conditions
r Tanks with inadequate vents or overflows
The National Bureau of Fire Underwriters publishes tables giv-

ing the heat loss per hour from various types of tanks. A 250,000-gal
(1-ML) elevated tank located in a −10◦ F (−23◦ C) area losses 1 million
Btu/h (292 kJ/s) in a 12-mph (19-km/h) wind. Move the tank to a
−40◦ F (−40◦ C) location, such as International Falls, Minnesota, and
the heat loss almost doubles (Knoy 1991a, 1991b). See the isothermal
map in Fig. 9-16.
Although it is possible to replace most of the water in a tank with
warmer water, that is seldom the way the tank is operated. Water stor-
age tanks (particularly elevated tanks and standpipes), because they
are connected to the main only by the riser pipe, usually float on the
system. As a result, it is possible that a tank can serve a system that
uses several times more water per day than the tank capacity yet still
receive only a small percentage of fresh water daily. The operating pro-
cedures discussed in the following paragraphs allow a tank to make
optimal use of the heat available in incoming water and highlight
limitations and effects of water quality.
Most utilities in cold-weather regions experience a drop in de-
mand during the winter months. If the upstream pump station to a
tank that is part of such a utility’s distribution system does not allow
for adequate reduction in pumping so that the tank is allowed to cycle,
tank turnover may be jeopardized, water quality may suffer, and, in
extreme weather, ice may form. Drafting the tank and filling it may
slow the formation of ice caps inside tanks. This is the only opera-
tional tool available to slow icing. The more a utility is able to cycle
a tank, the less ice buildup will occur. However, tanks should not be
kept at low levels for long periods, because the amount of heat energy
contained in the tank is a function of the volume of tank’s water. The

Tank Capacities—Thousands of US Gallons (Thousands of Liters) Add Btu
332
Heat
per Lineal
(Btu) 25 30 40 50 75 100 150 200 250 Ft (kW-hr/m)
Loss (94.6) (113.6) (151.4) (189.3) (283.9) (378.5) (567.8) (757.1) (946.4)
2 Uninsulated
per Ft
2 Square Feet (Square Meters) of Tank Surface∗ Steel Riser
(kW-hr/m )
Atmospheric Tank 1,210 1,325 1,550 1,800 2,370 2,845 3,705 4,470 5,240 3 ft 4 ft
Temperature Radiating (112.4) (123.1) (144.0) (167.2) (220.2) (264.3) (344.2) (415.3) (486.8) (0.914 m) (1.220 m)
(◦ F[◦ C]) Surface Btu Lost per Hour, Thousands (kW) Dia. Dia.
35 [1.7] 32.3 40 43 51 59 77 92 120 145 168 50 69

(0.102) (0.012) (0.013) (0.015) (0.017) (0.023) (0.027) (0.035) (0.042) (0.049) (0.048) (0.066)
30 [−1.1] 46.1 56 62 72 83 110 132 171 207 242 144 192
(0.145) (0.016) (0.018) (0.021) (0.024) (0.032) (0.039) (0.050) (0.061) (0.071) (0.138) (0.184)
25 [−3.9] 61.5 75 82 96 111 146 175 223 275 323 255 340
(0.183) (0.022) (0.024) (0.028) (0.033) (0.043) (0.051) (0.065) (0.081) (0.095) (0.245) (0.327)
Operation
20 [−6.7] 77.2 94 103 120 139 183 220 287 346 405 380 506
(0.243) (0.028) (0.030) (0.035) (0.041) (0.053) (0.064) (0.084) (0.101) (0.119) (10.365) (0.486)
15 [−9.4] 93.6 114 125 146 169 222 267 347 419 491 519 692
(0.295) (0.033) (0.037) (0.043) (0.049) (0.065) (0.078) (0.102) (0.123) (0.144) (0.499) (0.665)
10 [−12.2] 110.9 135 147 172 200 263 316 411 496 582 670 893
(0.349) (0.039) (0.043) (0.050) (0.059) (0.077) (0.093) (0.120) (0.145) (0.171) (0.644) (0.859)
5 [−15.0] 128.9 156 171 200 233 306 367 478 577 676 820 1,092

(0.406) (0.046) (0.050) (0.059) (0.068) (0.089) (0.108) (0.140) (0.169) (0.198) (0.788) (1.05)
0 [−17.8] 148.5 180 197 231 268 352 423 551 664 779 982 1,309
(0.468) (0.053) (0.058) (0.068) (0.079) (0.103) (0.124) (0.161) (0.195) (0.228) (0.944) (1.26)
−5 [−20.6] 168.7 205 224 262 304 400 480 626 755 884 1,152 1,536
(0.532) (0.060) (0.065) (0.077) (0.089) (0.177) (0.141) (0.183) (0.221) (0.259) (1.11) (1.47)

−10 [−23.3] 190.7 231 253 296 344 452 543 707 853 1,000 1,329 1,771
(0.601) (0.068) (0.074) (0.087) (0.101) (0.132) (0.159) (0.207) (0.250) (0.293) (1.28) (1.70)
−15 [−26.1] 213.2 258 283 331 384 506 607 790 954 1,118 1,515 2,020
(0.672) (0.076) (0.083) (0.097) (0.112) (0.148) (0.178) (0.231) (0.279) (0.327) (1.46) (1.94)
−20 [−28.9] 236.8 287 314 368 427 562 674 878 1,059 1,241 1,718 2,291
(0.747) (0.084) (0.092) (0.108) (0.125) (0.165) (0.197) (0.257) (0.310) (0.363) (1.65) (2.20)
−25 [−31.7] 262.3 318 348 407 473 622 747 972 1,173 1,375 1,926 2,563
(0.827) (0.093) (0.101) (0.119) (0.138) (0.182) (0.219) (0.285) (0.343) (0.403) (1.85) (2.46)
−30 [−34.4] 288.1 349 382 447 519 683 820 1,068 1,288 1,510 2,145 2,860
(0.909) (0.102) (0.112) (0.131) (0.152) (0.199) (0.240) (0.313) (0.377) (0.442) (2.06) (2.75)
−35 [−37.2] 316.0 383 419 490 569 749 900 1,171 1,413 1,656 2,381 3,174
(29.3) (0.112) (0.122) (0.143) (0.067) (0.219) (0.264) (0.343) (0.413) (0.485) (2.29) (3.05)
−40 [−40.0] 344.0 417 456 534 620 816 979 1,275 1,538 1,805 2,620 3,494
(0.997) (0.122) (0.134) (0.156) (0.182) (0.239) (0.287) (0.373) (0.450) (0.529) (252) (3.36)
Operation
−50 [−45.5] 405.6 491 538 629 731 962 1,154 1,503 1,814 2,126 3,139 4,186
(1.28) (0.144) (0.158) (0.184) (0.214) (0.282) (0.338) (0.440) (0.531) (1.623) (3.02) (4.02)
−60 [−51.1] 470.8 570 624 730 848 1.116 1,340 1,745 2,105 2,467 3,702 4,936
(1.48) (0.167) (0.183) (0.214) (0.248) (0.327) (0.392) (0.511) (0.616) (1.722) (3.56) (4.75)
Notes: To determine heat loss per hour, find the minimum mean atmospheric temperature for 1 day from the isothermal map (Fig. 9-16) and note the
corresponding heat loss. Heat loss for a given capacity with a different tank radiating surface than shown here shall be obtained by multiplying the radiating
surface by the tabulated heat loss per square foot (square meter) for the atmospheric temperature involved. The minimum radiation surface area shall be the

wetted tank steel surface area plus the top water surface area. For tanks with large steel plate risers, the heat loss from the riser shall be added to that from

the tank. The riser loss per linear foot (linear meter) shall be as tabulated above.
∗ These numbers are square feet (square meters) of tank radiating surfaces used for each capacity to compute the tabulated heat loss values and are typical for
tanks with D/4 ellipsoidal roofs and bottoms.
Source: AWWA Manual M42.
333
TABLE 9-6 Thousands of British Thermal Units (Btu) Lost per Hour from Elevated Steel Tanks Based on Minimum Water Temperature of 42◦ F (5◦ C)

and Wind Velocity of 12 mph (5 m/s)
Operation
334 Chapter Nine
FIGURE 9-16 Isothermal lines for lowest one-day mean temperatures and
normal daily minimum 30◦ F (−1◦ C) temperature line for January, United
States and Southern Canada.

Operation
Operation 335
FIGURE 9-16 (Continued)

lower the volume of water in a tank, the less heat energy is present
in the tank. It is this heat energy that minimizes ice formation. For
example, it takes longer for a pitcher of water than a glass of water to
freeze. The authors have been successful in eliminating icing by using
tank mixing systems in combination with normal tank operation.

Operation
336 Chapter Nine
FIGURE 9-17 More than 100 tons (90.72 metric tons) of “dirty” ice was
discovered inside this 3-mil-gal (11.35-ML) elevated tank in the spring of
2004. It took weeks for repair crews to remove the ice before work on the
tank could begin.
Elevated tanks are more susceptible than ground tanks to icing

conditions (Fig. 9-17). Ground tanks are more likely to experience
icing than partly or fully buried reservoirs. Insulating the tanks is not
a viable option; insulation only slows heat loss but does not stop it.
Hence, prolonged extreme weather eventually causes icing despite the
insulation. The reason partly or fully buried tanks have less propensity
to freeze is that the ground temperature is much higher than the air
temperature below freeze depth. Freeze depth is the depth of ground
from surface (grade) below 32◦ F (0◦ C). Ice formation on the interior
surfaces of the tank acts as insulation and retards further icing of the
water. However, formation of an ice cap on the surface of the water is
the most damaging result of cold weather.
Ice can have several detrimental effects on tank longevity and
water quality. The physical effects of ice have been well documented.
Floating and falling ice can scrape coatings and damage or destroy
structural elements of the tank such as tie rods, ladders, overflow
weirs and piping, painter’s rings, and so on. The effect is cumulative
and progressive; more and more damage is done in each cold season
if the ice problem is not addressed.
Although it is often desirable to take tanks down for inspection
or rehabilitation work during low-demand winter months, caution
must be taken to ensure that the tank is drained before ice forms.
Otherwise, substantial amounts of ice may remain in the tank after it is
drained. In some cases, so much ice may be present that it would need

Operation
Operation 337
to be physically removed before work can proceed. This can cause

considerable unexpected delay and added expense to tank inspection
or rehabilitation.
Thawing a frozen tank is a difficult task. There are, however,
some alternatives for dealing with such situations. A common tech-
nique is to lift a heavy-duty hose 3/4 in. (19 mm) in diameter from
the ground up over the top of the tank. The end of the hose termi-
nates with a 1/2-in.- (12-mm)-diameter pipe about 10 ft (3 m) long.
This pipe is dropped through the tank vent or manhole located di-
rectly over the inlet or riser pipe. It must be kept slightly off the ice
to keep it from sticking to it. Warm water from a fire department tank
truck is pumped through the hose, parting the ice as the probe drops
down. The water must be used judiciously. Stopping in the middle of
this process could allow the thawing pipe to freeze in the tank, which
may mean the equipment must be left in the tank for the rest of the
winter.
When the tank truck is empty, it must be refilled from utility mains.
However, the refill water will not be as warm as the original water in
the truck, which was taken from the heated fire station. Steam gen-
erators have been attached to this type of probe, but the warm-water
method seems to work best, is cheaper, and creates fewer safety haz-
ards. The same equipment can be used to thaw a frozen riser. How-
ever, it may be difficult to thread the probe into the riser pipe. The
tank drawings or a recent tank inspection report should be reviewed
before this type of riser thawing is attempted, unless the position-
ing of the piping arrangement is known from experience. Artificially
thawing tanks is expensive and dangerous. In addition, a warm front
may move through the day after the crew has thawed the tank, which
makes a high thawing bill hard to justify.
Rust, coating debris, and damaged structural elements are nor-
mally encapsulated in tank ice. Because the damage occurs as the ice
develops, the substances—which may include biofilms, metals, de-
bris, and assorted particulates such as coating chips—become trapped
in the ice for the duration of the cold season. Most of the accumula-
tion is in the tank’s ice cap. These substances are released when the ice
thaws. Since they are released near the surface, they may be suspended
long enough to be drafted out of the tank with the outflow.
Another way ice formation can affect water quality in tanks is that
if the ice cap becomes thick enough (as it often does), the water level
underneath the ice may drop while the ice cap remains stationary.
This creates a small vacuum above the surface of the water that can
cause air and chlorine (if used as a secondary disinfectant) to leave the
water. This wet chlorine gas and air can cause accelerated corrosion
in the areas damaged by the ice.
Several devices on the market claim to prevent ice formation inside
tanks. Some involve the use of mechanical mixers such as pumps

Operation
338 Chapter Nine
or propellers, some use air bubblers (bubbles remove heat from the
water, and there must be an influx of warmer water for the system to
operate properly), and others use turbulent jets. They are all essentially
mixing devices with varying degrees of mixing efficiency and energy
consumption. A small tank may be fitted with a recirculation pump
and a water heater. This energy-intensive option is limited by tank
size and equipment expense.
There are reasons beyond tank operation that may make a tank
susceptible to freezing problems. Isolation or altitude valves may be
improperly applied or malfunctioning, which creates conditions of
overflow, low flow, or no flow that may cause a tank to freeze. Valves’
hydraulic control piping should be protected against freezing by pro-
viding heated enclosures, insulation, or electric heat tracing.
Some traditionally designed air vents for tanks can collect conden-
sation, freezing rain, or snow that freezes on vent parts and hinders
operation. A common problem in such cases is the insect screen serv-
ing as a substrate for ice formation. Water outflow from a tank with a
frozen vent creates a vacuum condition that may result in progressive
collapse of the tank shell until structural failure allows venting or the
water outflow stops. Several manufacturers supply frostproof tank
vents designed either to prevent the formation of ice on their parts or
to dislodge any ice formed by using a tank vacuum condition. Utility
crews should inspect tank vents periodically to ensure their proper
functioning.
A common cause of tank freezing is inadequate cover over the pipe
leading to the tank. Sometimes soil conditions preclude installing the
tank foundation deep enough to provide adequate frost cover; in such
cases, fill should be brought in for cover, or other means should be
devised to insulate the pipe.
Site grade In some tower-type elevated tanks, the riser foundation is

built higher than the column foundations. The site should be graded
higher in the center to prevent the inlet/outlet piping from being ex-
posed to the atmosphere between the concrete riser foundation tunnel
and the earth.
Compacting fill If the earth over the connecting piping is not com-
pacted properly, it will settle during the first few years of operation.
The ensuing lack of adequate cover, combined with moisture satu-
ration of the depression, creates a potential trouble area that may
freeze.
Supervising the covering The base of the tank is usually the location
of the interface between constructors, and it may become a no-man’s-
land. Unless the construction work is properly supervised and in-
spected, the piping may not even get backfilled before the first winter
operation (AWWA 2003).

Operation
Operation 339
Tank overflows are subject to icing when the tank overflows at a

slower rate than the freezing rate. This may occur for any of several
reasons: a utility may want to misguidedly trickle-overflow a tank to
prevent it from freezing; the hydraulic controls for the altitude valve
may be set above the overflow elevation or may be damaged by ice or
frozen; an altitude or isolation valve may be leaking; telemetry may
be faulty due to weather conditions or damage; or the altitude valve
may have been locked in the open position so that the tank can float
on the system, and the system pressure may have increased beyond
the tank capacity.
When tanks overflow in freezing weather, several problems can
develop. An overflow-to-grade may freeze solid, especially where
there is a trickling overflow, where screens are plugged, or where
flap or rubber valves are stuck on the discharge. If water continues
to be pumped into the tank after the overflow pipe is frozen solid,
the tank may overflow through the roof hatch for a while and then
through the vent. The tank will then freeze solid, build up pressure,
and burst.
When water overflows through the hatch and vent, it invariably
forms a large icicle, weighing tons, on the tank exterior. The same
problem can be caused by normal overflow through a stub overflow
(one extending only a few feet or meters from the shell of the tank).
Either situation places a large eccentric load on the structure which,
in the case of water towers, can exceed the structure’s design stress.
The icicle usually forms on the side of the tower that is away from
the prevailing wind, and the wind and icicle together create additive
loadings. Even if the structure is not damaged by the ice load, it may
be damaged when the ice thaws or breaks off and falls. Eccentric ice
loadings or tower members damaged by falling ice have caused water
towers to fall (AWWA 2003).
Proper tank design and operation will prevent many of the freez-
ing problems discussed previously and will allow operators to deal
with the problems that do occur. Special consideration should be
given to tank inside appurtenances. Tanks located in areas where the
LODMT is −5◦ F (−15◦ C) or colder should not be equipped with inside
ladders or overflow pipes. As ice forms and moves up and down, it
can exert tons of force on ladders and pipes, tearing them loose from
their supports and possibly ripping or punching holes in the container.
The resulting leak will occur at a very inopportune time. If an inside
overflow pipe is broken, the tank will rapidly lose all water down to
the break, creating a large icy area on the ground below. If the vent
is plugged with ice or snow, the tank roof may collapse when water
evacuates the tank rapidly.
It is acceptable to equip a tank with inside ladders and overflow
pipes if the tank is known to have a high turnover rate of warm water.
A ladder and overflow can also be installed at the center of the tank

Operation
340 Chapter Nine
and supported by the access tube, as in single-pedestal tanks and

extremely large column-type tanks.
The use of interior girders, roof bracing, painters’ rails, or virtually
any other protrusion below the high water line or within an area af-
fected by floating or suspended ice is a poor design practice for areas
with an LODMT of −20◦ F (−29◦ C) or colder. Certain local conditions
or tank use patterns such as those discussed earlier may cause equally
severe icing problems in warmer areas.
Surge Relief Tanks

Surge tanks are connected to trunk mains at locations where there is
a change in elevation, such as a peak or a knee. They are designed to
hold a specific volume of potable water at a specific head or pressure.
Special valve arrangements maintain the water level until a down-
surge or column separation occurs in the nearby main. Within a few
seconds, a large amount of water is released to fill the void created.
This prevents the destructive vacuum from damaging the pipe and
minimizes flow reversal, which can create extremely high pressure in
the pipe. The size of the surge tank and its associated piping and the
speed of water delivery are based on transient surge analysis, which
should be conducted by experienced professionals. Misapplied surge
tanks may cause more damage than if there was none. Figure 9-18
shows an insulated surge tank.
FIGURE 9-18 Insulated surge tank in metropolitan area provides protection

from water column separation to large pump station located approximately
200 ft (60.961 m) lower than tank.

Operation
Operation 341
Since these tanks do not float on the distribution system, special

attention must be given to the quality of water, disinfection residual,
and turnover in the tanks. In colder climates, special consideration has
to be given to eliminate freezing, which can affect the available volume
and speed of delivery at the desired head. A circulation pump changes
the water in the tank at a sufficient rate to maintain water quality. Tank
water is returned to the system, and fresh water is supplied from the
system upstream of the return location. In cold climates, surge tanks
are typically insulated, and a recirculation pump may be used to cycle
the water in the tank through heaters. The chlorine residual should
be monitored in tanks whose recirculation systems do not exchange
water with the main. If needed, chlorine should be boosted using a
rechlorination system.
Hydropneumatic Tanks
Hydropneumatic pressure tanks are used in very small systems to re-
duce the amount of pumping required to provide water at pressure.
They can also serve as temporary replacements to elevated storage
in small systems during prolonged rehabilitation work. Hydropneu-
matic tanks are pressure vessels typically made of steel. A portion of
the tank is kept filled with compressed air. Once full of water, the tank
provides water in excess of the pump capacity as required. This keeps
the pump from short-cycling and provides pressured water for short
periods during power outages. These vessels can also act as pressure
surge relief tanks.
Fluid Dynamics in Tanks

The water entering the tank may be turbulent (as in the case of a pipe
inlet) or laminar (as in the case of a tank with a large-diameter wet
riser). The inlet geometry and the flow characteristics affect how the
water interacts with the tank’s structure and the ambient water within
it. During design, depending on the tank’s operational requirements,
it is typically desired to have one of two flow modes, plug or mixed.
Plug Flow
Short-circuiting, in general, occurs when influent water bypasses most
of the tank volume, having only minimal interaction with ambient
water, and flows directly to the outlet. Clarifiers, clearwells, or tanks
where contact time is needed should not be designed to have any form
of short-circuiting. It is ideal if the water enters the tank and leaves
the tank in an orderly fashion like a train, with the oldest water in the
tank leaving first. This is referred to as the first-in, first-out flow mode
or the regime generally known as plug flow.
In such a flow mode, a decrease in tank elevation decreases the
flow area. This increases the speed by which the influent water reaches

Operation
342 Chapter Nine
the outlet. Hence, the detention time of water in the tank decreases.
Also, depending on geometry, some areas inside the tank may expe-
rience a change in the nature of the flow. For example, a separation
area near where water changes direction may trap more water and
increase velocities in the rest of the cross-sectional area, which may
cause some mixing and/or short-circuiting.
The Disinfection Profiling and Benchmarking Guidance Manual of the
USEPA (1999) has procedures for CT calculations that list baffling
factors for various baffling conditions. These are approximate, and the
regulatory agency usually approves the baffling factor during design
reviews. Computational fluid dynamic analysis can accurately predict
baffling factors by modeling a tracer flow through the tank. However,
the use of CFD is limited to design validation and should not be a tool
for determining baffling factors because drastic difference in results
are possible depending on the model and modelers’ limitations. Tracer
studies are the most accurate means of determining baffling factors.
A tank’s theoretical detention time (TDT) is computed by dividing
the volume (V) of the tank by the peak hourly flow rate (Q): (TDT
= V/Q). The baffling factor, T10 /T, is multiplied by the TDT to yield
an estimate of the contact time (also known as the effective detention
time), T10 , as follows:
T10 = Contact Time = V/Q × T10 /T
Baffling factors are a function of tank design. T10 /T equal to 1.0
represents pure plug flow characteristics where TDT is equal to the
contact time, T10 . Figure 9-19 illustrates a CFD model of a baffled
clearwell showing velocity contours.
FIGURE 9-19 Computational fluid dynamics model of baffled clearwell showing

velocity contours.

Operation
Operation 343
Baffling Condition T10 /T Baffling Description

Unbaffled (mixed 0.1 None, agitated basin, very low
flow) length-to-width ratio, high inlet and
outlet flow velocities
Poor 0.3 Single or multiple unbaffled inlets and
outlets, no intrabasin baffles
Average 0.5 Baffled inlet or outlet with some
intrabasin baffles
Superior 0.7 Perforated inlet baffle, serpentine or
perforated intrabasin baffles, outlet
weir or perforated launders
Perfect (plug flow) 1.0 Very high length-to-width ratio
(pipeline flow), perforated inlet,
outlet, and intrabasin baffles
Source: USEPA (1999).
TABLE 9-7 Baffling Classifications and Factors
Baffling classifications and factors as given in the USEPA Guidance

Manual are shown in Table 9-7.
A discussion of CT calculations is outside the scope of this text.
However, we shall discuss the effect of baffling geometries on deten-
tion times. Baffles are obstructions in a tank that can be made from
a variety of materials. These obstructions contain and direct the flow
in tanks to create plug flow conditions. Ideally, the objective is to
transform the tank into a wound pipe, creating a first-in, first-out con-
dition. Baffle configurations can be numerous; however, some baffle
arrangements produce better plug flow than others. Design modifica-
tions that can increase T10 may allow the same inactivation (CT) with a
decreased disinfectant residual or a decrease in tank size. The follow-
ing summarizes some of the design features that optimize clearwell
plug flow design and operation.
Inlet Treatment
The inlet to a tank can be modified—for example, by adding a perfo-
rated wall—so that the flow entering the tank does not create an un-
even stream but is distributed evenly along the flow’s cross-sectional
area. In addition, the flow should not be directed such that it impinges
on and attaches itself to a boundary; this creates a skewed velocity
profile.
Flow Area Length Versus Width

To optimize the total water volume, proper baffling must be designed
to eliminate short-circuiting and dead zones (stagnant areas). The flow

Operation
344 Chapter Nine
area should be long and narrow, eliminating secondary flows and

allowing a pressure gradient along the length of the flow channel
only. Ideally, the entire volume of water should be in motion in one
direction with no recirculation, stagnation, or mixing.
Number of Turns and Angle of Flow Direction Change

The number of direction changes in a baffled tank should be mini-
mized. Turns skew the velocity profile, causing the water at the outer
(concave) edge of the turn to move faster than at the near (convex)
end. The increased turbulence creates some localized mixing. The an-
gle of flow direction affects the amount of skew and the size of the
separation region in which recirculation occurs. Keeping the turns as
mild as possible improves the plug flow. This means that curvatures
are better than corners, especially if the radius of curvature is larger
than the width of the flow area. The tank designer should take the
following into consideration:
r Avoid any dead zone by eliminating (filling in) corners be-
tween horizontal and vertical surfaces.
r Add perforated walls at each change of flux direction. To ob-
tain good water distribution on the whole cross section, a cer-
tain head loss must be created through holes, given by the
following formula:
P = k (V 2 /2g)
where
k = 0.62
v = velocity through hole, in feet or meters per second
g = acceleration of gravity = 32.2 ft/s2 (9.80665 m/s2 )
A velocity of 2 ft/s (0.6 m/s) creates sufficient head loss to ensure
good distribution.
Laminar Versus Turbulent Velocity Profile

The perforated walls just discussed can be used at various intervals
along the flow path to modify the shape of the velocity profile from a
laminar bullet shape to a more blunt (turbulent) shape, thus producing
a more uniform velocity distribution across the flow cross section.
Outlet Treatment
To maintain water level, an outflow weir may be provided at the tank’s
outlet. Water drop should be minimal to prevent aeration of water and
disinfectant loss. However, a bypass through the weir should be added
to allow drafting of the clearwell in an emergency. A perforated wall in
the vicinity of the outlet may help in eliminating preferential currents
and help ensure an even distribution into the outflow from all parts
of the baffled channel’s cross-sectional area.

Operation
Operation 345
FIGURE 9-20
Computational fluid
dynamics model of
0.8 baffling factor
tank.
Inflow and Outflow Control

To prevent the creation of transients and encroachment of high-
velocity influent water into the lower-velocity ambient water, the flow
into and out of the tank should be controlled to maintain steady-state
conditions.
Baffling factors are directly proportional to water depth. As the
flow depth decreases, the baffling factor decreases. Figure 9-20 shows
a CFD model of a 0.8 baffling factor tank. In addition, the decrease in
water depth increases the average velocity across the baffled channel.
This decreases the theoretical detention time, resulting in a consid-
erable decrease in contact time. A way to ensure a constant water
depth is to use variable-speed pumping with a proportional integral
derivative (PID) controller using tank elevation as a feedback and the
desired water depth as a set point. Valve throttling can achieve similar
conditions but is not as accurate or energy efficient.
Mixed Flow
Mixed flow occurs when the influent water impinges on the ambient
water, resulting in a diluted volume representative of their propor-
tions. The water leaving the tank is no longer the oldest water in the
tank but is of an averaged age based on the tank’s turnover ratio and
mixing efficiency. Just as the baffling factor represents the scale of plug
flow present in a pass-through tank, mixing efficiency or effectiveness
represents the amount of mixing achieved. Just as baffling schemes
are not equal in performance, performances of mixing schemes differ.
Mixing in tanks can be achieved in many ways; we shall attempt to
discuss most of them, concentrating on turbulent jet mixing due to its
applicability in water storage tanks as discussed hereafter.

Operation
346 Chapter Nine
Mixing Times and Dynamics

Mixing is a function of time. For mixing to be complete, it must be
efficient and sufficient. Whenever an attempt is made to mix multiple
ingredients, the amount of time during which the ingredients are sub-
jected to the mixing process dictates the degree of mixing. The nature
of a particular mixing process determines its mixing effectiveness or
efficiency. Therefore, the mixing time needed to provide a homoge-
nous tank is indirectly proportional to its mixing efficiency. Inefficient
mixing may cause water in some areas of the tank not to be mixed.
Water in a tank without an effective mixing system may have stratifi-
cation, short-circuiting, or stagnant areas. Even a significant amount
of filling and drafting would not result in mixing in the tank. Calcu-
lation of a minimum theoretical mixing time is insufficient to make a
decision regarding the mixing condition of a tank. Consider, for ex-
ample, a standpipe that experiences stagnation due to a stratification
problem, or a tank with stagnation issues due to short-circuiting. If
sufficient mixing time is available, as calculated by the method dis-
cussed later in this section, the stagnant water will remain there and
will not mix.
Basic Discharge Flux Quantities
The basic discharge (from diffuser) flux quantities are used to deter-
mine mixing characteristics.
r Discharge volume flux: Qj = uj Aj
r Discharge momentum flux: Mj = uj Qj
r Discharge buoyancy flux: J j = g Qj
j
where
j = parameters pertaining to the discharge jet

A = cross-sectional area
u = velocity along main flow axis
gj = reduced gravitational acceleration of the discharge jet due to
the density difference between the effluent and ambient
environment (AWWA 2003)
The momentum flux affects the relatively near-field mixing re-
gion and has minimal affect on the far field region. Momentum flux
is directly proportional to the micromixing; however, it is not an in-
dication of mixing in the entire tank. Available research specifically
dictates that macromixing or large-scale folding of the interfaces is
what affects mixing efficiency, not micromixing or small-scale wrin-
kling. There is ample research on turbulent mixing, and the more ap-
propriate measures to use would be the coefficient of variance, decay
function, and/or mixing efficiency calculations based on fluid inter-
faces (Hjertager et al. 2008; Nathman, Aguirre, and Catrakis 2004).

Operation
Operation 347
Mixing Times and Efficiency

The mixing time equation currently in use for water storage tanks
has been derived from experimental data on 4- and 6-ft- (1.2-
and 1.8-m)-diameter cylindrical tanks under fill-and-draw operation
(Grayman et al. 1999).
Mixing time (in seconds) = 10.2 V 2/3 /M1/2 = 9 V 2/3 (d/Q)
where
M = momentum (inflow velocity times inflow rate, UQ)

V = volume of water in the tank at the start of the fill cycle in cubic
feet (cubic meters)
Q = inflow rate in cubic feet per second (cubic meters per second)
U = inflow velocity in feet per second (meters per second)
d = tank inlet diameter in feet (meters)
This calculated mixing time must exceed the tank’s fill time for
sufficient mixing (Grayman et al. 1999). In small tanks, such as models,
the effects of mixing by molecular diffusion affect a larger proportion
of the tanks’ volume within a given time than is the case with large live
tanks, in which the concentration gradients may be lower or dissipate
more quickly, causing slower diffusion. Hence, the previous equation
may overestimate the mixing capabilities of a tank. Nevertheless, this
approach is a good tool, since a tank that does not meet the criteria of
the previous mixing time equation most probably has water mixing
problems.
On the other hand, if a tank meets the criteria, there is still a
good chance that it has a problem regardless of the apparent sufficient
exchange of water volume. The most definitive and accurate means of
checking a tank for water quality is to take multiple samples at differ-
ent areas and elevations and calculate the tank’s coefficient of variance
(CoV), as discussed later. More research should be conducted to de-
termine satisfactory CoV or mixing efficiency values. Some authors
set a target CoV of 0.05 (5 percent) for well-mixed tanks. Recent pub-
lications point to a CoV of 0.1 (10 percent) as a more attainable goal.
Although we agree with the higher CoV requirement change (less mix-
ing), we still believe it may not always need to be this stringent. Many
tanks that meet the criteria of the previous mixing time equation do
not have a CoV less than 0.1, which further reinforces the statements
made about its limitations.
Some references state that a CoV between 0.05 and 0.1 is compara-
ble to achieving complete mixing. A CoV of 0.05 is typically considered
in the industry to constitute an excellent mixing condition for a wide
variety of applications (such as paints), all of which are required to be
more homogeneous than ambient potable water in tanks. A CoV value
close to zero reflects optimum homogeneity, while a value close to 1.0

Operation
348 Chapter Nine
implies the solution is heterogeneous. High levels of homogeneity are

not needed in tanks to maintain water quality. In addition, it may be
very expensive to achieve CoV levels less than 0.1 (10 percent). A CoV
of 0.15 (15 percent) is considered good for industrial applications if
mixing time is increased by approximately 25 percent. A higher CoV
may then be acceptable if we are able to increase mixing time, slightly
boost incoming disinfectant residual, or make other less expensive
tank modifications that can slow disinfectant deterioration—or if it is
determined that the higher variation’s effect on water quality is ac-
ceptable. Hence, adequate mixing is subject to many interpretations
from many perspectives. It should be based on achieving water quality
objectives with an efficient use of resources. Nevertheless, one cannot
go wrong by targeting a CoV lower than or equal to 10 percent if the
ability to do tank analysis is hindered.
The coefficient of variance is defined as
CoV = /xave
where
n = number of samples
= standard deviation of the measured readings corrected to the
true value of the population by using n−1 in the denominator
x ave = average of the measured readings
Determination of CoV is the most definitive method of establish-

ing tank mixing performance. Several samples are collected at var-
ious elevations and areas within a tank. A good rule of thumb is
to collect at least six samples; more samples should be collected for
tanks larger than 1 mil gal (3.78 ML). Any number of parameters can
be studied—disinfectant residual, disinfection by-products, turbidity,
temperature, and so on. A CoV of 0.2 or less is an indication that the
tank is sufficiently mixed and does not require mixing enhancements.
Another widely used parameter for quantifying mixing effective-
ness is the range mixing effectiveness (ERange ). The ERange is defined as
E Range = 1 −RangeOut /RangeIn.
where
RangeOut = range of concentration readings leaving the tank
RangeIn = range of concentration readings entering the tank
An E Range value of 1.0 implies excellent or homogeneous mixing,

whereas a value of zero implies a heterogeneous solution or poor
mixing. This measure may yield misleading effectiveness values if
short-circuiting is present in a tank. However, this anomaly will be
revealed when the tank is drafted and stagnant water leaves the tank.
Stagnant water may cause an E Range value below 1.0.

Operation
Operation 349
The advantage of this method of mixing evaluation is that it lends

itself easily to online monitoring, since only the inlet and outlet re-
quire continuous sampling and testing for water quality parameters
(disinfectant residual, for example).
Mixing time calculations involving inflow and outflow rates
should be compared with tank ambient volume to determine whether
or not a mixing system would work. If tank turnover rate is insuf-
ficient—that is, if the volume of water entering and leaving does not
change a certain volume in the tank—turbulent jet mixing systems
may not result in a homogenous tank, regardless of the mixing effi-
ciency. The required percentage of tank volume to be exchanged is
a function of the scalar concentrations of the ambient water and the
incoming water. These scalars include disinfectant residual, contam-
inants, disinfection by-products, and so on. Other influential param-
eters may be the temperatures of the incoming and ambient water,
chemical compositions and biological contents, atmospheric condi-
tions, and the tank’s physical condition.
Mixing Theory
Of the many mixing technologies, no single one can be used for all mix-
ing duties. This makes choosing mixers a complex task that requires
adequate understanding of mixing processes—their application and
limitations. Many texts have been written on this subject. Hereafter
we shall attempt to discuss the subject in general and then focus on
its applications in tanks.
The mixing of one or more components or materials in a fluid
system can be described in terms of two separate but interlinked pro-
cesses, macromixing and micromixing:
r In macromixing, there is no mixing on the molecular scale, but

fluid elements or different components are well distributed or
blended to create uniformity throughout the mixture.
r Micromixing is complete mixing of species on the molecular
level. Droplet or particle size reduction (dispersion) of one or
more components of the mixture produces increased homo-
geneity of the system.
Most fluid-mixing problems may be analyzed in terms of the mis-

cibility or ease of mixing of the components. The ease of distribution
affects the mixing approach to be adopted. For instance, where the
rate of reaction between the components is to be improved (such as
combining water with a high disinfectant concentration with water
that contains bio-matter water or with low-residual water), the mix-
ing approach is focused on maximizing distribution (macromixing).

Operation
350 Chapter Nine
For mixing immiscible fluids such as chlorine gas or solids with water,
the mixing approach is focused on reducing droplet or particle size to
maximize the area of contact (micromixing). Hence, micromixing has
a higher tendency to resuspend settled solids or diffuse gases.
Another point of consideration is the mode of operation involved,
of which the fluid mixing is normally only a part. The most important
distinction that affects the mixing operation is whether the tank op-
eration is a batch process (distinct fill or draft cycles) or a continuous
process (pass-through where fill and draft occur simultaneously). In
a batch process, a discrete volume of material is mixed in the tank,
whereas in a continuous process, a stream of material is mixed.
The Mixing Process

Fluid mixing is the process by which a nonuniform system is made
uniform. The degree of mixing or uniformity can be analyzed by eval-
uating how well the flow is macromixed or micromixed. A measure
of uniformity is the coefficient of variance, which is the ratio of the
standard deviation in concentration and the average concentration.
The CoV may be viewed as a measure of macromixing. Therefore, if
the CoV is close to 1.0, the fluid is not mixed; if it is close to zero, it is
homogenous.
To analyze the degree of micromixing, another quantity, the decay
function, is evaluated. The decay function (d) can be expressed as the
ratio between the cross-sectional average of concentration fluctuations
and the cross-sectional average of the concentration (Nathman et al.
2004).
A recirculation zone in the area near the inlet increases the mixing
effect. Experimental results show a uniform concentration distribu-
tion at that point. However, farther along the path, the concentration
becomes less uniform as the flow becomes better macromixed than
micromixed (Nathman et al. 2004). In other words, the turbulence in
the region very near the inlet can result in micromixing, but micromix-
ing cannot be maintained as the flow gets farther from the source. This
has been confirmed by several researchers, and the mechanisms that
cause each type of mixing are known.
Before discussing mixing mechanisms, it must be understood that
the aforementioned research implies that even high-jet-velocity inlet
sources will always revert to macromixing as the flow gets farther
from the source, whereas micromixing is limited to areas close to the
source. The energy cost required to expand this effect to cover an en-
tire tank or reservoir would be excessive. In addition, such a high level
of mixing is not necessary to achieve high mixing efficiency, as will be
shown later. Micromixing is also not maintainable, because as soon as
the inflow to the tank stops, density and temperature gradients will
increase the coefficient of variance with time. Biochemical reactions

Operation
Operation 351
taking place over time also contribute to the increase in the coeffi-
cient of variance. In addition, the higher velocities reduce the total
amount of tank influent water because of the effect of the resulting
higher head loss on pump performance. Reducing the amount of wa-
ter with high disinfectant residual entering the tank results in a lower
overall increase in tank disinfectant residual over time. Therefore, cre-
ating high-velocity jets in an effort to increase micromixing is not an
optimum way of mixing tanks.
Tank Mixing Systems

Various tank mixing systems employ several energy transfer and con-
version mechanisms to create various effects for particular mixing
results. In the case of distributive or macromixing, swirl created by
rotating parts or directed flow causes laminar thinning of the material
interfaces, thereby increasing volumetric combination of the materi-
als. A repeated cutting and folding action of the mixture also increases
the distribution of various material components. The effectiveness and
efficiency of a mixer in distributive mixing is therefore a function of
how the machine interacts with the fluid in a geometric sense. That
is, the volume enclosed by the outer interfaces (i.e., the interfaces be-
tween the pure fluid and the mixed fluid) rather than the interfacial
surface area is what determines mixing efficiency.
Conversely, the effectiveness and efficiency of a mixer in disper-
sive mixing (micromixing) is dependent on the means of the system’s
shearing interaction with the fluid. The higher the shear stress, the
smaller the resulting particles or droplets are in the mixture. The uni-
formity of the stress distribution determines the uniformity of the
mixing. Without uniform distribution of the shear stresses, it is im-
possible to guarantee that the same level of mixing is applied to all
parts of the water in the tank.
Mixing technologies are often available in either batch or contin-
uous form, but rarely both. In situations where both are offered, there
are typically some performance trade-offs. Care must be taken to se-
lect a system that performs effectively with the tank’s operational
scheme. In addition, energy consumption and availability must be
considered. Since the amount of water to be mixed is huge, the en-
ergy needed to mix the tank can add up to a considerable amount
over time. Also, many tanks are constructed in remote areas where
electrical power may not be readily available. Some mixing systems,
due to the requirement for high inlet pressure, may require additional
pumps, pump replacement, or a change in operation points of existing
pumps. This not only changes the distribution system’s performance
and efficiency during filling, but it also may affect the drafting of the
tank, which could become critical in achieving fire flows from the tank.
We shall next discuss some of the various mixer types on the market.

Operation
352 Chapter Nine
Mixer Types
The many types of mixers applicable to water mixing in tanks may be
categorized as follows:
Impellers
These have specially shaped blades on a rotating shaft driven by a gear
motor and/or variable-speed drive. They are used almost exclusively
for batch processes. Applying them in water tanks or towers may be
prohibitive because of structural support requirements and energy
consumption issues. Also, these systems may not work in changing
water levels.
Static Mixers
These devices require continuous fluid motion to work, so they are
used in continuous processes. They comprise a set of nonmoving ob-
structions or orifices in a pipeline. The obstructions or orifices are
shaped and positioned in such a way as to create cutting and fold-
ing effects and/or turbulence for mixing of piped fluid streams. Static
mixers are a reliable and low-cost alternative. Nevertheless, any high-
pressure drop across the mixer may require larger and more expensive
pumps, increase energy consumption, or alter pump operating points.
Dispersers
Dispersers comprise a range of complex machines and systems that
deliver relatively uniform dispersions in particular fluid applications.
A valve homogenizer comprises a very high-pressure pump and a
controlled valve nozzle through which the fluid is forced at very high
velocity to rupture the droplets through extensional stressing. The jet-
impinging mixer, another type of disperser, uses high-velocity fluid
streams—except that in this case, the fluid is jetted against a plate or
contra-jet to rupture the droplets or particles using impact stressing.
The high level of mixing that these systems provide is localized; use
of this technology in water tanks would require many nozzles or jets,
making it high in installed cost and energy consumption.
Pump Mixers
Available in both batch and continuous forms, pump mixers use in-
ternally generated energy to force fluid through small nozzles at very
high velocities while extending and shearing it. The fluid flowing
through the nozzles at high velocity then impinges on an internal
wall of the mixer. A dynamic cutting and folding action added to
vigorous turbulent flow provides distributive mixing. Pump mixers
are suited to a wide variety of applications because they can handle a
wide range of materials and viscosities with high mixing performance.
In addition to energy consumption and availability issues, these sys-
tems may not be as effective in large tanks. Tests conducted for the US

Operation
Operation 353
Department of Energy have shown that a considerable volume of wa-

ter in large tanks remains unmixed (Lee and Dimenna 2001).
Turbulent Jet Mixing

Turbulent mixing is involved in most mixing processes due to its abil-
ity to mix and transport species, momentum, and energy much faster
than can be done by molecular diffusion (Hjertager et al. 2008). De-
tailed understanding of the mixing process and validation of proto-
types are even more critical in mixing systems used in tanks that store
potable water for human consumption, because they can affect water
quality.
Turbulent jet discharge into a crossflow (i.e., a transverse jet) is
a turbulent free-shear flow (Shan and Dimotakis 2001). Crossflow in
a tank can be created by change in water elevation and/or outflow
streamlines (as in a pass-through tank). The discharge, in the form
of a round buoyant jet into a nonturbulent stratified crossflow, may
contain five asymptotic regimes (AWWA 1986) in which self-similar
flow conditions exist:
1. Pure jet
2. Pure plume
3. Pure wake
4. Advected line puff
5. Advected line thermal
Of these regimes, the first three are free turbulent flows dominated
by transverse shear (shear normal to jet axis); the latter two are dom-
inated by azimuthal shear (parallel to jet axis). In the latter case, an
internal double-vortex structure is generated within the jet. No self-
similar regime is possible in the presence of density stratification (Jirka
1999). In that case, axial pressure forces influence and finally destroy
the boundary-layer evolution of the flow and lead to strong horizontal
spreading—the so-called collapse motion during the terminal-layer
phase of a buoyant jet in stratified surroundings. The flow is no longer
jet like. In actual discharge situations, one or more of the five regimes
can occur as asymptotic regimes. Regime (1) is often the initial regime
(whenever the jet velocity U j > ambient velocity ua ). Regime (5) is
usually—but not always—the final regime (Jirka 1999). Figure 9-21
shows a model of a positively buoyant turbulent jet.
The buoyant jet is a flow phenomenon with free turbulence. It
represents a gradually evolving flow along its axis and thus exhibits
boundary-layer characteristics with its possibilities for mathematical
simplification including self-similarity techniques (Schlichting 1968).
However, because of the variety of forcing elements, buoyant jet mo-
tions are in general not self-similar (Jirka 1999). They are self-similar

Operation
354 Chapter Nine
FIGURE 9-21 Model of positively buoyant turbulent jet.
only in five possible asymptotic regimes in which they have an in-

variant internal force balance, invariant turbulence, and entrainment
properties. In between these regimes, the buoyant jet properties are
variable and cannot be scaled uniquely by local jet parameters (Jirka
1999).
The local Reynolds number (Re) in the case of a turbulent jet in
a quiescent reservoir is approximately equal to the jet exit Reynolds
number Rej = d j U j / (d j is the jet diameter, U j is the jet velocity, and
is the kinematic viscosity of water). If the jet momentum flux Mj U j =
j (d j /2)2 U j 2 ( j is water density and Mj is the mass of water in the jet)
is held constant as the jet diameter decreases and its velocity increases,
the jet approaches a point source of normal momentum, which gener-
ates a counterrotating vortex pair (Shan and Dimotakis 2001). At low
Reynolds numbers, in particular, tertiary and sometimes quaternary
vortices are formed such that the vortices no longer have equal circu-
lation, and each vortex could have a different induced velocity. In that
case, the induced vertical velocities would be substantially smaller,
and the overall mean trajectory would be shallower. At high-Re trans-
verse jets, the counterrotating vortex pair is the dominant structure
and the primary mechanism for entrainment of free stream fluid. The
vortex circulation is a decreasing function of downstream distance de-
caying through viscous diffusion (Shan and Dimotakis 2001). In other
words, high-Re (more turbulent) inlet sources would result in a flow
with higher vertical velocities and a vortex pair as the main source
of water entrainment. As will be explained later, entrainment is the
major mixing mechanism. The higher vertical velocity means that the
influent would reach the water surface faster, resulting in even less
entrainment of the ambient water in the tank. Hence, the high level

Operation
Operation 355
of micromixing caused by the high Reynolds number results in more

homogenous mixing in the vicinity of the inlet, but less water is mixed.
Experimental data indicate that the scalar gradients are steeper in
the horizontal directions than in the vertical directions. Anisotropy of
the transverse jet’s scalar field is in contrast to the far field isotropy
for axisymmetric jets discharging into a tank (Shan and Dimotakis
2001). Scalar mixing in the transverse jet is enhanced by increasing
the Reynolds number. In the case of turbulent jets discharging into a
tank, the concentration’s probability distribution functions lose their
peak at the highest Reynolds number (Shan and Dimotakis 2001). This
means that when discharging into a reservoir or tank in which there
isn’t a crossflow, mixing is decreased in the far field at high Reynolds
numbers (or influent velocities). The minimum Reynolds number for
which turbulent mixing can be considered as fully developed is ap-
proximately 10,000 (Nathman et al. 2004).
On the basis of the previous analysis, tank mixing systems need
to reduce the Reynolds number to slightly above 10,000 by reducing
the velocity through each inlet. An efficient way to accomplish this
is to have more inlets to spread the flow. Having more inlets results
in a larger volume of water involved in near-field micromixing, and
it introduces the influent at various locations in the tank, further en-
hancing the distribution of the micromixing effect.
Flow-dependent mixing is explained by noting that turbulent mix-
ing is essentially a three-stage process (Shan and Dimotakis 2001):
Entrainment: Engulfment of irrotational (ambient) flow into the tur-

bulent flow region, or macromixing.
Stirring: Kinematic motion responsible for creating interfacial area
between species.
Molecular mixing: Diffusive mixing on the molecular scale, or mi-
cromixing.
The balance among these three stages determines the probabil-

ity distribution function of the mixed water. Nevertheless, the mean
concentrations are a measure of entrainment rather than of molecu-
lar mixing. Hence, the transverse jet entrains less ambient fluid than
the ordinary turbulent jet. Transverse jets homogenize the entrained
fluid more thoroughly. This indicates that for transverse jets, there is
more stirring and molecular mixing; for ordinary jets, there is more
entrainment (Shan and Dimotakis 2001).
Increases in water elevation in tanks with mixing systems result
in little transverse motion. Transverse motion is especially negligi-
ble in a large tank unless it is a pass-through tank with considerable
and independent inflow and outflow rates. Therefore, the influent jets
need to entrain larger volumes to mix the whole tank. To maximize
entrainment, the inlets should be located such that the inflow engulfs

Operation
356 Chapter Nine
the maximum amount of volume possible. This is done by creating

sheet flows at different elevations that intersect and or interact in the
far field. This is further enhanced if the influent is buoyant, as will
be shown later, whether positively or negatively due to temperature
difference between influent and ambient water.
Turbulent Jet Mixing Efficiency

Turbulent jet mixing efficiency is determined by the behavior of the
mixed influent/ambient water interfaces. Knowledge of the dynam-
ics of the interfaces is crucial for physical descriptions, predictions,
and control of the mixing efficiency (Nathman et al. 2004). Although
it has long been recognized that large-scale entrainment is important
for mixing, one must understand that entrainment alone sets an up-
per bound only on mixing efficiency. In other words, knowledge of
the growth rate of a turbulent shear flow is not sufficient to deduce
mixing efficiency (Nathman et al. 2004). The crucial point is that the
volume enclosed by the outer interfaces (i.e., the interfaces between
the influent water and the ambient tank water)—rather than the in-
terfacial surface area—determines mixing efficiency. Hence, it is the
large-scale dynamics of the outer interface that provide the dominant
contributions to the mixing efficiency (Nathman et al. 2004).
On the basis of high-resolution measurements in the far field of
fully developed round jets, it was found that large-scale folding of the
interfaces, as opposed to the small-scale wrinkling of the interfaces,
provides the dominant contribution to mixing efficiency (Nathman
et al. 2004).
A discharge with no buoyancy is referred to as a nonbuoyant jet or
pure jet. A release of buoyancy only (no initial momentum) is called
a pure plume. A discharge with both momentum and buoyancy is
called a buoyant jet or forced plume. Positively buoyant flows are
defined where the buoyancy force acts vertically upward against the
gravity force; negative buoyancy is defined as acting downward in
the direction of the gravity force (MixZone 2005).
Dilution in turbulent buoyant jets is caused by entrainment of
surrounding ambient water into the influent water jet. Entrainment is
a turbulent process caused by shear stress between the discharge flow
and the surrounding ambient water (MixZone 2005).
As briefly stated earlier, to improve mixing efficiency, a tank mix-
ing system would have to engulf or entrain as large a volume of water
as possible during inflow. To take maximum advantage of this mixing
phenomenon, a tank mixing system must have inlets creating sheets
of flow at different elevations. The interfaces should then fold over
each other for optimum mixing efficiency. The directed momentum of
the created sheets of flow would create more folding if the inlet ori-
entations and design allow for the energy to be converted to directed
momentum rather than excessive stirring turbulence. This process

Operation
Operation 357
requires less throttling at each inlet and multiple inlets at the

proper elevations. The locations of the inlets should be such that
boundary-layer attachments occur only as part of the entrainment
scheme and do not disrupt it.
Mixing Regions
The hydrodynamics of an influent continuously discharging can be
conceptualized as a mixing process occurring in two separate areas.
In the first area, the initial jet characteristics of momentum flux,
buoyancy flux, and outfall geometry influence the jet trajectory and
mixing. This area, called the near field, encompasses the jet subsurface
flow and any surface or bottom interaction (or, in the case of a stratified
ambient, terminal–layer interaction). The mixing zone is the part of
the near-field area in which the initial dilution of a discharge occurs.
Many hydrodynamic definitions of mixing zones include both near-
field mixing and boundary-interaction processes (MixZone 2005). In
this area, mixing system design can usually affect the initial mixing
characteristics through appropriate manipulation of design variables.
In particular, designs with dynamic bottom attachments should be
avoided (MixZone 2005). Dynamic plume attachments occur when
the discharge flow interacts strongly with a boundary in the near
field. Such near-field boundary interactions present the possibility
of high influent concentrations near the discharge (MixZone 2005).
Often, near-field attachments are avoidable with proper design of the
mixing system. This flow also exhibits a subsequent buoyant liftoff
and an unstable near field (MixZone 2005).
Two types of attachment are typically found: wake attachment
forced by the crossflow and Coanda attachment forced by the entrain-
ment demand of the influent jet itself. A physical description of these
processes is given below (MixZone 2005).
In wake attachment, the presence of the discharge structure and
the jet influx interrupts the ambient velocity field and causes a recir-
culation area in the wake downstream from the discharge (MixZone
2005).
A Coanda attachment is a rapid dynamic attachment that occurs
when a jet discharges close to a nearby parallel boundary. This process
is referred to as a Coanda effect. It occurs because of the entrainment
demand of the jet flow at its periphery. If a boundary limits the ap-
proach flow of ambient water, then low-pressure effects cause the jet
to be deflected toward that boundary, thereby forming a wall jet. Thus,
the mixing process of Coanda-attached flow is governed by the dy-
namics of the wall jet (MixZone 2005). Figure 9-22 shows a negatively
buoyant turbulent jet flow with wall attachment.
This implies that a tank mixing system discharging very close to
the bottom or surface may cause flow attachment to the bottom or a
buoyant film at the top with reduced mixing.

Operation
358 Chapter Nine
FIGURE 9-22 Negatively buoyant turbulent jet flow showing wall attachment.
(Source: MixZone.)
Boundary interactions occur when the flow contacts the surface,

bottom, or sides or forms a terminal layer in a density-stratified
ambient environment (Fig. 9-23). Boundary interactions also deter-
mine whether mixing is controlled by stable or unstable conditions at
the discharge source (MixZone 2005). Boundary interaction generally
FIGURE 9-23 Turbulent jet flow into density-stratified tank model. (Source:
MixZone.)

Operation
Operation 359
provides the transition from near-field (discharge source–controlled,

or micromixing region) to far-field (ambient environment–controlled,
or macromixing region) mixing processes. However, boundary inter-
actions in the form of dynamic plume attachments to the bottom are
considered near-field mixing processes (MixZone 2005).
As the turbulent buoyant jet travels farther away from the source,
the source characteristics become less important. Conditions existing
in the ambient environment control trajectory and dilution of the flow
through the spreading of buoyant density-current motions and pas-
sive diffusion due to ambient turbulence. This region is referred to as
the far field (MixZone 2005). A counterrotating vortex pair has been
noted in the far field of a transverse jet; however, the mean flow state
is not necessarily a symmetric vortex pair but can be unsteady and
asymmetric under certain conditions (Shan and Dimotakis 2001).
The rotational flow created by the vortex is what helps to make
the flow regime more homogenous. Hence, a properly designed tank
mixing system would attempt to encourage or extend the creation of
lateral vortices that do not reach the surface so that it enhances the
mixing of the largest volume of water.
The assessment of near-field stability (i.e., distinguishing stable
from unstable conditions) is a key aspect of analyzing influent di-
lution and modeling the mixing zone. It is especially important for
understanding the behavior of the two-dimensional plumes resulting
from multiport diffusers (MixZone 2005). Discharge plumes may be
classified as having the following characteristics:
r Stable discharge conditions usually occur for a combination
of strong buoyancy and weak momentum (MixZone 2005).
r Unstable discharge conditions occur when a recirculation phe-
nomenon appears in the discharge vicinity. This local recircu-
lation leads to re-entrainment of already mixed water back
into the buoyant jet region (MixZone 2005).
The previous discussion of stability emphasizes the importance of

eliminating recirculation at the discharge vicinity. Recirculation at the
inlets can be minimized by (1) reducing influent velocity to minimize
the recirculation region size or (2) having a boundary surface very
close to the point of discharge to prevent or minimize rotational flow
in the direction opposite that of the discharge. In the second case, this
means not using discharge nozzles.
Buoyant Discharges
Information about the density distribution in the ambient water body
is very important for correctly predicting influent discharge plume
behavior. Density currents are buoyancy driven far-field flows that
are defined by transverse horizontal spreading while being advected

Operation
360 Chapter Nine
downstream by an ambient current. These spreading processes can

intrude into the ambient flow, forming a buoyant upstream wedge
and stagnation point. These flows are caused by the density difference
of the mixed flow relative to the ambient density. Density currents are
preceded by turbulent jet mixing in the near field and are followed
by passive diffusion in the far field. Density currents may or may not
form upstream intrusions, depending on the crossflow magnitude and
internal buoyancy at boundary interaction (MixZone 2005).
Buoyant jets discharged horizontally along the water surface from
a laterally entering channel or pipe bear some similarities to the more
classical submerged buoyant jet. For a relatively short initial distance,
the effluent behaves like a momentum jet, spreading both laterally and
vertically due to turbulent mixing (MixZone 2005). After this stage,
vertical entrainment becomes inhibited due to buoyant damping of
the turbulent motions, and the jet experiences strong lateral spread-
ing. During stagnant ambient conditions, ultimately a reasonably thin
layer may be formed at the surface of the receiving (ambient) water;
that layer can undergo transient density-current buoyant spreading
motions (MixZone 2005).
In the presence of ambient crossflow, buoyant surface jets may
exhibit any one of following three types of flow features (MixZone
2005):
r They may form a weakly deflected jet that does not interact
with the bottom or surface.
r When the crossflow is strong, they may attach to the down-
stream boundary, forming a bottom-hugging plume.
r When a high discharge buoyancy flux combines with a weak
crossflow, the buoyant spreading effects can be so strong that
an upstream intruding plume is formed that also stays close
to the surface near the inlet.
Density currents are effective transport mechanisms that can
quickly spread a mixed effluent laterally over large distances in the
transverse direction, particularly in cases of strong ambient stratifi-
cation. In this case, influent of considerable vertical thickness at the
terminal level can collapse into a thin but very wide layer, unless this
is prevented by lateral boundaries (MixZone 2005).
If the influent water is nonbuoyant or weakly buoyant, there is
no buoyant spreading area in the far field—only a passive diffusion
area. Depending on the type of near-field flow, ambient density strat-
ification, and boundary interaction process, several types of density
current buoyant spreading may occur:
r Spreading at the water surface
r Spreading at the bottom
r Spreading at a sharp internal interface with a density jump

Operation
Operation 361
FIGURE 9-24 Buoyant discharge from single port inlet at 45-degree angle into
stagnant tank.
r Spreading at the terminal level in a continuously (e.g., lin-

early) stratified ambient.
Turbulence in the ambient environment becomes the dominating

mixing mechanism in the far field at sufficiently large distances from
the discharge point. In general, the passively diffusing flow grows
wider and thicker until it interacts with the vessel bottom and/or
sides (MixZone 2005). The strength of the ambient diffusion mecha-
nism depends on several factors that relate mainly to the geometry
of the ambient shear flow and the amount of ambient stratification.
From classical diffusion theory, gradient diffusion processes in the
bounded flows can be described by constant diffusivities in the ver-
tical and horizontal direction that depend on turbulent intensity and
on channel depth or width. In the presence of a stable ambient strat-
ification, the vertical diffusive mixing is generally strongly damped
(MixZone 2005).
In the surface approach condition, the weakly bent flow impinges
on the surface at a near-vertical angle (>45 degrees) (Fig. 9-24). After
impingement, the flow spreads more or less radially along the wa-
ter surface as a density current. In particular, the flow spreads some
distance upstream against the ambient flow and laterally across the
ambient flow. The strong buoyancy of the discharge dominates this
spreading. The lateral spreading of the flow in the surface impinge-
ment area is driven by both the flow momentum and the buoyancy
force (MixZone 2005).

Operation
362 Chapter Nine
Since influent water is rarely at the same temperature as ambient

water in the tank, buoyant flow can be used as a free source of mixing
energy. Substantial additional mixing can be created and further op-
timized if the tank mixing system is designed to use density streams
as part of the entrainment scheme. This would further negate the use
of inlets at the bottom or near the surface in a tank mixing apparatus.
Such inlets prevent the density streams’ interfaces from folding due
to stratification and prevent the engulfment or entrainment of large
volumes of water needed for mixing efficiency.
Flow Diffusers
Influent
A multiport diffuser is a linear structure consisting of many more or
less closely spaced ports or nozzles that inject a series of turbulent jets
into the ambient receiving water body. These ports or nozzles may
be connected to vertical risers attached to an underground pipe or
tunnel or they may simply be openings in a pipe lying on the bottom
(MixZone 2005).
Flow diffusers in water tanks’ mixing systems should be designed
to use and accommodate all the physical phenomena associated with
turbulent jet mixing. Such systems should optimize mixing efficiency
using as little energy as possible, as follows:
r Use reasonable inlet velocities. Higher velocities are not only
energy consuming, but are also detrimental to proper mixing
as shown earlier. This requires a multitude of inlets to suffi-
ciently divide the flow so that velocities are lower.
r Do not use nozzles—not only because of higher velocities and
head loss, but also because more recirculation is associated
with nozzles. Orifices are closer to the conduit/pipe surface
and minimize such recirculation.
r Entrain larger volumes of ambient water by having flow
streams at different levels and sides, creating three-dimensio-
nal mixing effects. A three-dimensional arrangement would
be required for the inlets such that the influent creates undis-
turbed streams or currents, which engulf the majority of am-
bient volume.
r Enhance folding of the interfaces in the far field (Fig. 9-25).
The inlets must be positioned so that inflow streams impinge
on each other and interact with boundary surfaces in such a
manner that they create large-scale folding and lateral vor-
tices.
r Use density currents.

Operation
Operation 363
FIGURE 9-25 Sheet flow from multiport diffuser into tank model showing
formation of vortices and folding of interfaces in far field. Note: better mixing
distribution in far field.
Effluent
The flow diffuser discussed so far is an influent structure. However,
it may also be a tank effluent structure. The filling cycle typically uses
a fraction of the total time of a tank’s operating cycle. The tank may
remain idle for some time or may draft for a long time, feeding back
into the system. Reliance on influent mixing alone is not optimum, be-
cause throughout the longest part of the operational cycle, no mixing
is taking place. To optimize mixing, it is ideal to also mix during the
draft cycle. After the fill cycle, ambient water in the tank will stratify,
lose disinfectant, or be rendered otherwise nonhomogeneous because
of some physical or biochemical activity. As a result, water quality
may progressively decrease.
More importantly, impurities or disinfection by-products may set-
tle, stratify, or accumulate unevenly because of temperature gradients
and removal of fluid solely at one or two locations on the tank bot-
tom. To prevent the possibility of high concentrations of accumulated
impurities (such as some disinfection by-products or solids being fed
back into the distribution system by excessive drafting), it is prudent
to mix or blend the effluent water from various areas and elevations
of the tank as it is drafted. At a minimum, effluent mixing will accom-
plish the following:
r Prevent the sequential removal of stratified or accumulated
components

Operation
364 Chapter Nine
r Minimize the number of dead spots due to drafting from the

regions close to the outlet
r Decrease reliance on passive diffusion which may be dis-
rupted by many environmental factors
r Decrease the possibility of icing in colder climates due to dis-
tributed motion of the fluid and removal of fluids near the
top
r Remove water close to the top, where biological activity is
higher and disinfectant residual is lower
r Make effluent more uniform and provide a more homo-
geneous starting point for the next fill cycle, making turbulent
jet mixing more effective
r Because of uniformity of mixing, achieve better and more
meaningful tracking of water age
r Eliminate mixing cost, since water from the tank or reservoir is
mixed by gravity and forced back into the distribution system.
When water is combined from different areas and elevations

through the diffusers, it creates large interface areas internally which
are then stirred by the turbulence in the diffusers and tanks’ piping,
valves, and fittings. This further mixes the flow to create a more ho-
mogeneous effluent and consistent water quality.
Tank Venting
Most water storage tanks are nonpressurized tanks that require ad-
equate venting. By allowing the removal or replenishment of air as
water enters or exits the tank, venting prevents both pressurized and
vacuum conditions. Atmospheric tanks are not designed to handle
pressurization; the absence of sufficient venting to handle the air out-
flow generated as water enters the tank would cause the air in the
tank to compress and exert pressure on the tank walls that may ex-
ceed design stress limits. Likewise, tanks are not designed to handle
the vacuum conditions created when water is drafted from a tank
without adequate venting. Buckling of tank walls takes place even
when differential pressure is small.
The styles of air vents most commonly found in water tanks are
the mushroom, pan, and 180-degree types. AWWA D100, Standard
for Welded Steel Tanks for Water Storage, requires that one tank vent,
even if more than one is required, always be located near the center of
the roof. A reasonable offset is allowed for tanks designed with center
dry-access tubes. Vent designs, examples of which are given in Figs.
9-26 to 9-28, should meet the following requirements:

Operation
Operation 365
Screen
24 in. (588 mm)
Tank roof assembly
FIGURE 9-26 Double 90-degree elbow roof vent detail. (Source: AWWA Manual
M42, Steel Water Storage Tanks.)
r Prevent insects and animals from entering the tank (a non-

corrodible mesh is recommended)
r Prevent rainwater or surface water from entering the tank
r Prevent air drafts from entering the tank
(Outside diameter)
3/16
C
ov
er
A A
.
am
di
nt
Ve
Tank ro
of
di
am
(Inside diameter)
et
er
(Outside diameter)
(Hole in roof)
Plan view Section A-A
FIGURE 9-27 Pan deck vent detail. (Source: AWWA Manual M42, Steel Water
Storage Tanks.)

Operation
366 Chapter Nine
re
su
pres 1/2- no. 13 × 15 flattened
Air expanded metal bird screen
Pres
PTFE gaskets (typical) sure
palle
t
Vacuum
pallet
Screen
(brass material is normal)
Support bars
Air vacuum
Carbon-steel body
Roof
Install vent vertical 5 +
FIGURE 9-28 Typical clog-resistant vent detail. (Note: PTFE = polytetrafluoroe-

thylene. Pallets should be removed during coating to prevent clogging of the
screens. Periodic inspection and maintenance are required to keep in proper
working condition.) (Source: AWWA Manual M42, Steel Water Storage Tanks.)
r Exclude dust and debris, as much as possible, from entering

the tank
r Provide some level of security against accidental or inten-
tional contamination
r Prevent direct sunlight from entering the tank
r Be frostproof in cold-weather areas
r Be tall enough, or installed high enough, not to be blocked by
drifting snow or debris
It is a requirement of the Ten States Standards that overflows not

be considered as vents. Obviously, a tank using its overflow as a vent
would be left without venting during overflow conditions. It also
points out that vents on ground-level tanks should terminate in an
inverted U shape with its opening 24 to 36 in. (609.6 to 914 mm) above
the tank’s roof or ground. The U-shaped overflow should be covered
with 24-mesh noncorrodible screen installed within the pipe at the
location least susceptible to vandalism. AWWA manual M42, Steel
Water Storage Tanks, recommends clog-resistant vents with pressure-
and vacuum-releasing pallets.
Large tanks should be provided with more than one vent. One
should be installed near the center of the roof, and the other(s) closer

Operation
Operation 367
to the tank’s walls. This facilitates crossflow ventilation through the

tank.
Vent sizing is of special concern in the case of tanks in systems
that have experienced demand growth. Increased tank inflow and
outflow rates must be handled by tank venting. Undersized vents
must be replaced with properly sized ones to prevent problems related
to increased maximum flow rates.
Telemetry
Most storage facilities for potable water are located in unmanned sites.
Some tanks are located in sites manned by a handful of operators
whose main responsibilities are to monitor a water treatment process.
In either case, it is desirable to have automated systems that moni-
tor hydraulic and water quality parameters of tanks. These devices
can store data in electronic form or on paper. They can also transmit
information collected to a central location or manned facility where
an operator can keep track of and control multiple facilities through-
out a plant and/or a distribution system. Telemetry is the science and
technology of automatic measurement and transmission of data by
wire, radio, or other means from remote sources, pumping stations,
distribution system tanks, or other facilities or processes to receiv-
ing stations for recording and analysis. Most telemetry systems used
by water utilities are commonly known as SCADA (for supervisory
control and data acquisition) systems.
Tank Water Elevation

Water utilities that operate a SCADA system have a central monitor-
ing facility where one or more operators are able to remotely control
the fill or draft of tanks, the opening and closing of motor-operated
valves, and chemical feed processes such as disinfectant boosting. In
addition, the SCADA system monitors, records, analyzes, and identi-
fies trends regarding myriad parameters from online sensors, analyz-
ers, and transmitters at each facility in the communication network.
Although some of the systems monitored by SCADA may be auto-
matic (e.g., the closing of an altitude valve to prevent tank overflow,
the sounding of alarms, and so on), some may be entirely controlled
by the SCADA operator (e.g., starting and stopping pumps). Tank ele-
vation information lets SCADA operators know when pumps should
be turned on or off as part of normal distribution system operation.
In many cases, tank elevations are the only source of information to
operators regarding distribution system pressures.
Trended elevation data over time paints a picture of a tank’s daily
fill and draw cycles. Parameters such as rates of inflow and outflow
throughout the fill and draw cycles can be indirectly determined from

Operation
368 Chapter Nine
elevation data if the tank geometry is known, and changes in tank vol-
ume can be calculated over time. This trended flow and volume data
can help water quality personnel monitor mixing patterns in tanks
over the course of the year and provide operators with strategies to
enhance tank mixing for each tank’s operating characteristics. A bal-
ance must be maintained between the need for optimum water quality
and emergency storage. Allowable low levels in tanks should leave
sufficient water in storage to satisfy potential emergency demands
from fires, power outages, water main breaks, and so on.
Excessively high water elevation can trigger an overflow alarm
to alert the SCADA operator that an altitude valve may be malfunc-
tioning and that tank overflow is probable. The operator can respond
by turning pumps off at upstream pumping stations and/or closing a
remotely controlled motor-operated isolation valve, if one is present
at the tank. Depending on the particular circumstances, the operator
can dispatch a road crew to the tank. Skillful operation and knowl-
edge of the distribution system are required to appropriately address
high-pressure conditions.
Isolating a tank from the distribution system without taking other
measures, such as shutting pumps off, may create abnormally high
pressure and leave a system vulnerable to water main breaks and
other catastrophic failures. Tanks that are left open to a distribution
system provide surge relief should pressure transients be generated.
Such systems are deemed soft systems. A system operated at high
pressures with its tank(s) offline loses this surge protection and is
said to be a hard system. It should be noted that some utilities have
chosen to forego altitude valves and rely entirely on telemetry and
motor-operated isolation valves to control water level in the tank.
Many strategies are available for sensing water level. A few of
the most common technologies are listed here, divided into two cate-
gories: contact sensors and noncontact sensors.
Contact-Level Sensing Technology

Bubbler systems use a source of compressed air to push bubbles out of
a conduit at the bottom of the tank. The higher the pressure required to
push the bubbles, the higher the water level. Bubblers provide contin-
uous level sensing relatively accurately, but they require an external
source of compressed air. The air pressure is transmitted as an analog
voltage or current signal.
Radio-frequency (RF) capacitance sensors, tuning-fork sensors,
and floats are switches that are engaged when submerged in water
and disengaged when water levels drop below them. Several of these
switches can be installed on a track or some other means of support at
several tank depths. An elevation signal is generated for each particu-
lar depth where switches are located. A transmitter unit is commonly

Operation
Operation 369
used to monitor each switch in the tank and transmit digital outputs
corresponding to tank level. Icing is a concern with this technology.
Noncontact Level Sensing Technologies

Ultrasonic level sensors and radar level sensors are noncontact de-
vices that are suspended above the water surface. They measure the
rate of travel of energy through air to measure the distance between
the sensor and the surface of the water. Although some sensors are
susceptible to condensation, many styles are available with features
that prevent condensation from affecting the measurement. Icing may
or may not be a concern, depending on how the sensor has been in-
stalled. Another concern is the proximity to the sensor’s energy beam
of tank walls or other tank appurtenances. Some sensors are able to
“calibrate out” such obstacles. Under certain conditions, a stilling well
can be used to focus the sensor’s beam through a pipe dipped into the
water. This method has the added advantage of reducing water sur-
face turbulence that may affect accuracy, but it increases the chance
for icing damage.
Pressure transmitters measure the pressure, or head, of the tank at
some point along the tank’s piping or wet riser. These devices use the
compression of a pressure-sensing element, typically a strain gauge or
a capacitor, to generate a continuous-voltage or current-analog signal
corresponding to tank elevation. A tap is required in the tank’s piping
or riser where small-diameter piping (copper that is 0.25 to 0.75 in. [6.3
to 19.0 mm] in diameter is common) connects to the transmitter. This
small-diameter pipe is susceptible to freezing and should be installed
in a heated or insulated enclosure in cold-weather regions. Pressure
transmitters are the most common type of level sensor used in SCADA
systems for distribution system storage tanks.
Level-sensing methods require accurate information regarding
tank elevation and dimensions. A maintenance and calibration sched-
ule should be followed, and good records should be kept.
Street Pressure
Street pressure is measured on the street side of the altitude or tank iso-
lation valve. The most common technology used is the pressure trans-
mitter. A continuous-voltage or current-analog signal corresponding
to the pipe pressure at the sensor elevation can be continuously sent
to the SCADA operator. (The pressure transmitter need not be at
the same elevation as the pipe centerline, but this discrepancy must
be accounted for in the determination of street pressure.) If the alti-
tude valve is not locally controlled, a decrease in street pressure signals
the operator that the altitude valve or tank isolation valve should be
opened and the tank drafted to meet demand. Data on street pressure
enable the operator to monitor distribution system pressures even if

Operation
370 Chapter Nine
the tank has been isolated. For instance, a sharp decrease in street
pressure may be an indication of system failure such as a water main
break.
Disinfectant Residual
Water quality managers are installing monitors for disinfectant residu-
als at representative distribution system locations. Some installations
are made because of regulatory requirements, others as a voluntary
measure. Tanks are ideal for such systems since they are usually in
secured locations, electrical power is available, and utility personnel
periodically visit the tank site.
Various technologies are commercially available to measure and
transmit concentrations of disinfectant residual. The most common
secondary disinfection chemical in the United States is chlorine. The
following methods are described for measuring concentrations of
chlorine residual at water storage facilities.
N,N-Diethyl-p-phenylenediamine (DPD) Colorimetric Method

DPD is oxidized by chlorine in solution. This results in two oxidation
products. The contrast between the colors of these two compounds, as
measured by a colorimeter or a spectrophotometer, reveals the amount
of free or total chlorine in the water. A voltage or current-analog sig-
nal corresponding to the calibrated concentration of residual is then
transmitted to SCADA.
Iodometric Method
Potassium iodide reacts with free chlorine in the sample water to pro-
duce iodide. The iodide concentration is measured by the instrument
to yield total chlorine. Free chlorine is not measurable by this method.
Polarographic Membrane Sensor Technology

A pair of electrodes is immersed in a conductive electrolyte and sepa-
rated from the sample water by a chlorine-permeable membrane. Free
chlorine travels through the membrane and is reduced to chloride at
the electrode’s surface. The reduction of free chlorine generates an
electric current between the electrodes that is proportional to the free
chlorine concentration.
Amperometric Electrodes
Combinations of probes consisting of a silver anode and a platinum
cathode measure free chlorine concentration, pH, and temperature.
A current proportional to the free chlorine concentration is produced
within the electrodes. The amperometric electrodes require replace-
ment after a manufacturer-specified lifetime (Pollack et al. 1999). Al-
though automated, these systems may require chemical replenish-
ment and periodic maintenance and calibration to sustain accuracy.

Operation
Operation 371
Disinfectant-residual sensors can also act as feedback for booster

disinfection at storage facilities. The SCADA operator can remotely
control the feeding system. System parameters such as start, stop, feed
rate, remaining disinfectant, and set point may be controlled and/or
monitored.
Temperature
Temperature transmitters and thermocouples can be used to measure
and transmit temperatures. Low air temperature inside equipment en-
closures can alert the SCADA operator to potential freezing conditions
that are detrimental to tank-monitoring devices. A water temperature
probe inserted either in the tank or in the pipeline leading to the tank,
or in both, can provide the utility with information regarding tank
stratification conditions. A complete temperature profile of a tank can
be obtained by using a weighted line of thermocouples designed to
measure temperature at various water depths. The information may
be transmitted to SCADA or stored locally and downloaded manu-
ally. Trending of such temperature profiles over time can help water
quality managers and operators to determine operational parameters
for seasonal or changing conditions.
Flow
Many different types of meters can be used to measure water flow.
SCADA monitoring of flow into and out of a tank can indicate prob-
lems in the distribution system, assist water quality managers in de-
termining optimum tank operation, determine water depletion time
during emergencies, and so on. Some of the systems used to mea-
sure flow at tanks are differential-type flowmeters such as venturis,
insertion meters (V-Cone, Annubar, etc.), and orifice plates and
electronic-type meters such as ultrasonic, temperature, and magnetic
flowmeters. Detection of flow direction is inherent to the operation of
some of the meters, such as magmeters. Other meters, such as ven-
turis, require additional devices to determine flow direction. Each
meter named, whether as a primary or a secondary device, makes
use of a transmitter to calculate and convert the flow into an analog
current or voltage signal.
Security
As discussed earlier in this chapter, security at water storage facilities
is a concern to utility officials and law enforcement. SCADA systems
can also transmit data from security sensors and video from cam-
eras either to the SCADA operator or directly to a separate security
SCADA monitoring center or to law enforcement monitoring officials.
A variety of sensors are available to detect intrusion to a tank site or

Operation
372 Chapter Nine
tampering with tank appurtenances such as hatches and vents. Each of

these sensors is capable of sending a digital signal to SCADA that in-
dicates the appropriate security-related reaction. Pan/tilt/zoom cam-
eras can be operated remotely to track an intruder’s activities. Avail-
able camera technology is able to register images even at very low light
levels. Video recording devices at each site may be accessed remotely
through SCADA to download video images of the intrusion.
Local Monitoring and Control

The section of this chapter on water quality monitoring in distribution
system tanks outlined several parameters that a utility may choose
to monitor. Water corrosivity, pH, conductivity, total organic carbon
(TOC), and turbidity are among the most commonly encountered. All
and any of these parameters, as well as the ones described in detail
previously, can be relayed by telemetry to the utility’s SCADA center.
Data about all parameters measured at each tank are routed
through a common device for processing and communication. There
are essentially two devices found throughout the industry—the
remote terminal unit (RTU) and the programmable logic controller
(PLC). RTUs are generally associated with remote monitoring of field
devices. PLCs have been traditionally used for the automatic and/or
remote control of processes. These differences have become cloudy in
recent years; each device can now serve both monitoring and control
duties. Debate rages on over the reliability of each system, their con-
trol and sampling rates, and their capability or lack thereof to handle
large numbers of data points, store data during power outages, and
so on.
RTUs and PLCs require a protective splash-proof/weatherproof
enclosure. Each data point is wired to the device’s input/output cards.
The data are analyzed according to the terminal or the controller’s pro-
gramming and are stored or transmitted. It is possible for either device
to be connected to a local personal computer. Operators and mainte-
nance personnel can use these local computers to monitor or trou-
bleshoot data and device performance locally without the assistance
of, or feedback from, the SCADA center operator. The local computer,
which is called a human/machine interface (HMI), can be a laptop
computer brought from site to site or a desktop computer stationed
at the site.
Technology options available for remote telemetry communica-
tion can be categorized as follows: telephone, cellular, radio frequency,
fiber, and satellite. Telephone communication technology requires a
hard line be installed. The information is transmitted and received
in analog form by means of modems at the site and at the SCADA
center. Several communication rates are available depending on the
utility’s budget and the need for fast transmission of large volumes

Operation
Operation 373
of information. Available services range from basic telephone modem

communication (2,400 bits per second [bps]) to leased duplex lines
with high broadband capability (>2 million bits per second [Mbps]).
Because telephone lines are already available near most water stor-
age facilities and the needed equipment is relative inexpensive and
commonly available, telephone communication is the most popular
technology used for SCADA systems. Telephone lines are suscepti-
ble to weather such as lightning or windstorms, which may disable
SCADA access to the facility. Careful consideration should be given
during design of any SCADA system to default operation of controlled
devices if there is a communication failure.
Cellular communication technology can be used if the water stor-
age facility and the SCADA center lie within the coverage range of
a cellular communication company. A cellular modem is installed at
the tank and at the SCADA center, but no hard wire exists. Instead,
data are transmitted via a wireless cellular network of communica-
tion towers and cellular antennas between the site and the control
center. Cellular communication is a good choice for locations where
hard phone lines do not reach, as the cost is relatively higher than for a
hard telephone line and fewer broadband options are available. Many
electronic devices, such as pressure transmitters, small process con-
trollers, security sensors, and so on are now available with integrated
cellular modems capable of sending a small number of monitoring
and/or control signals to a cellular modem at the SCADA center.
Radio-frequency communication systems use a radio modem and
a low-powered transceiver at the tank location, and a transceiver is
connected to an RF base station at the SCADA center. Several tank
sites or other remote facilities can be polled over a single ultrahigh-
frequency (UHF) or very-high-frequency (VHF) system. Any station
can serve as a repeater to extend the line-of-sight transmission of
the SCADA center (Pollack et al. 1999). In a typical application, the
SCADA base requests data from a remote location, such as a tank, by
transmitting a wake-up signal to send data. When the remote begins
transmitting, the base reverts to the receive mode and collects the data
package. After transmitting the data, the remote goes back to the re-
ceive mode and awaits instructions from the base. The output of the
sensors at the remote site has usually been converted to digital data by
the RTU or PLC. This signal (typically in the range of 300 to 3,000 Hz)
is delivered to a modem that converts it to an analog form that can
be frequency-modulated to the RF carrier. When the base receives
the analog data, the base modem converts it back to digital data. The
Federal Communications Commission (FCC) has allocated certain fre-
quencies that can be used for fixed operation. Certain frequencies are
available for RF transmission in the low band (25 to 50 MHz), mid-
band (72 to 76 MHz), VHF band (150 to 174 MHz), UHF band (450 to
512 MHz), and 900 MHz (928 to 960 MHz). The low band provides the

Operation
374 Chapter Nine
best communication range because the path loss is less than at higher
frequencies. However, it is susceptible to interference from electri-
cal noise. The UHF band is the most frequently used because of the
large number of available channels and its relative freedom from elec-
trical noise.
Line-of-sight studies are always required to test the feasibility for
RF systems. Obtaining a license from the FCC for the exclusive use
of a frequency can be expensive and time-consuming. Start-up cost
is high, and although maintenance cost is low, the utility becomes
the sole owner and operator of the RF communication system and is
responsible for its maintenance and upgrades (Pollack et al. 1999).
Fiber SCADA communication systems require the installation of
expensive fiber-optic cable (often several miles or kilometers) to a fiber
utility cable. Fiber provides the best broadband of any communication
method, often surpassing 100 Mbps. This expensive option should
only be considered when a utility requires the fast transmission of
very large volumes of data, including real-time video. The fiber lines
can be leased or owned, depending on availability and agreement
with the provider. When the water utility owns the line, it becomes
responsible for its maintenance and upgrades.
When a tank is located where telephone lines, cellular commu-
nications, or RF systems are impractical, satellite communications
is an option if a satellite covering the distribution system area is in
space. In this case, the satellite acts as a relay station between the tank
and the SCADA center. Transmitters and receivers are required at
both ends to communicate through the satellite. This option is more
expensive than hard phone lines or cellular technology, but it may
be well worth the cost when no other communication alternative is
viable.
SCADA Systems
The sophistication of the SCADA system depends on the utility’s bud-
get, the equipment supplier, and the programmer/system integrator.
Often a single operator is in charge of remotely controlling and mon-
itoring thousands of data points throughout the distribution system.
A master station at the SCADA center is usually a single device (cen-
tralized system), a master with submasters (hierarchical system), or a
parallel group of processors (distributed system). For the purpose of
this discussion, each will be referred to as the master station.
The functions of the master station include scanning PLCs and
RTUs throughout the distribution system. This is accomplished by
monitoring the proper operation of remote control devices, ensur-
ing that messages from these devices are error free, retrying when
messages are incorrect, and reporting PLC or RTU failures. A mas-
ter station also processes data received from RTUs and PLCs. The

Operation
Operation 375
master station does this by checking for alarm conditions, averaging

and trending data, storing event changes, and entering data into a
database.
The transmittal of operator commands is another important func-
tion of the master station. The transmittal of commands involves
several steps: interrupting the scan and arming the proper remote,
encoding and transmitting the command, verifying that the proper
command has been received, permitting execution of the command,
and verifying command execution.
Master stations must also maintain a database for historical data.
To accomplish this, data received from the remote location are typi-
cally condensed into hourly and daily averages, peak values are de-
lineated, and various data-compression techniques are used to min-
imize storage. Additionally, the historical data might include status
information such as valve positions, water level elevations, and sim-
ilar items to allow later correlation with flows and pressures. The
historical database must also provide very flexible data retrieval
capabilities.
The master station is also responsible for driving the human/
machine interface. This is done by presenting data on video display
screens, map boards, printers, or similar mediums; providing the abil-
ity to define screen formats, including graphics; and providing the
ability to define report formats. Master stations provide the impor-
tant function of providing failover to a backup when necessary. This
involves maintenance of duplicate data files in a backup processor
and monitoring of the primary processor (by the backup) and switch
to the backup (i.e., failover) on detection of a stall or error. Master sta-
tions may also perform advanced functions such as supply prediction,
demand prediction, optimal pumping, and leak detection.
The human/machine interface is the point at which the operator
interacts with the SCADA system. Current SCADA systems offer in-
teractive HMI modules. These allow building of display screens by
those with no programming knowledge. This permits operations per-
sonnel who will be using the system to design and build graphic and
tabular displays that precisely meet their needs. These displays may
be interactive—that is, the symbol for a pump may change color de-
pending on pump status, or a reservoir icon may “fill” as the reservoir
level increases.
Inputs to a SCADA system occur either as real-time events au-
tomatically sensed and reported by the remote control device or as
manual inputs through an HMI. Inputs from RTUs or PLCs include
status, flow, pressure, and level. Inputs from HMIs include commands
for open/close, run/stop, and set point.
SCADA outputs are either for driving the HMI or for executing
the commands at the remote location. HMI outputs include periodic
reports, alarms, alarm summary reports, graphic pictures, displays of

Operation
376 Chapter Nine
real-time data, displays of averaged or trended data, and historical

reports. Control outputs include set point and on/off or start/stop.
SCADA technology should permit building of reports by persons
with no programming skills. An operator or engineer, for instance,
can readily define a report format. The resulting report can then be
produced for a special study or scheduled to be automated output as
a routine operational report (Gotoh et al. 1993).
HMI screens should be organized and labeled in a way that they
are easily identified and uploaded to the screen. The interface with the
operator should be clear and simple (feedback, on-screen help, and
option menus are useful tools). A sufficient number of control and
monitoring screens should be provided to keep screen cluttering to a
minimum. Alarms should be in a scroll window that does not overtake
the main window. Symbols, colors, and terms should be consistent and
self-descriptive. The overall screen environment should be intuitively
simple and allow operation with minimal computer programming
skills.
Energy Conservation in the Distribution System

In many communities, water utilities are the largest consumers of
electric power. Pumping is the largest consumer of electric power for
utilities and is their largest operational cost. Pumping is a continuous
process that is typically rarely interrupted. It is used to meet high
demand in the early morning and late afternoon (and/or evenings)
and to fill the tanks between those times. Utilities can take advantage
of elevated water storage to reduce these power expenses.
Electrical energy is converted to pressure and velocity head
through the pumping process. When the water reaches an elevated
storage facility, the water rises to an elevation equal to the remain-
ing energy in the pipe. Hence, energy is stored in tanks in the form
of potential energy, or head. Tanks can be filled during low-demand
hours to take advantage of reduced power rates. The volume of stored
water and its specific head reduces the need for pumping (additional
electric energy) during peak-demand hours. This process is known as
peak shaving. Peak shaving not only reduces power consumption, it
reduces the size of pump stations and trunk mains to satisfy the same
demand. This capital cost savings is in addition to the operational en-
ergy cost reduction. However, this must be weighed against the cost of
additional storage to satisfy demand. An optimum design is one that
achieves the lowest life-cycle cost including maintenance and cost of
demolition and replacement.
Tanks help save energy in other ways. Drafting of tanks in con-
junction with pumping during hours of peak demand reduces water
flow in trunk mains. Lower pipe velocities result in less head loss

Operation
Operation 377
(energy loss) due to friction. Maintaining water elevation in tanks

within a specific range can help set limits on the minimum and max-
imum system curves. Coordinating that with the pump curves forces
pumps to operate at or close to their highest possible efficiencies. The
size and number of pumps installed must be optimized in keeping
with the demand conditions and storage available. Pumps operating
efficiently can reduce electric consumption drastically. In many cases,
pumps with best efficiency points (the combination of flow and pres-
sure at which a pump operates most efficiently) that are greater than
90 percent provided wire-to-water efficiencies of less than 25 percent
due to inappropriate operation. Pump wire-to-water efficiencies dur-
ing operation are commonly less than 50 percent. To improve pump
efficiency, elevated tanks should be used to satisfy the marginal de-
mand, and pumps should be used only when there is enough demand
in the system for them to operate near their optimum efficiency. In
addition, allowing tanks to float on the system eliminates the need
to start pumps to meet marginal demand increases, which eliminates
maximum electric demand surcharges.
Most water utility power rates are based on demand or capacity
surcharge. That is, the water utility pays based on its peak power con-
sumption during the billing period, for the entire billing period. Rates
are also higher for power consumed during peak hours. Therefore,
water administrators and operators should look for ways to decrease
overall and peak-hour power consumption.
The action of starting pumps, in particular, draws large and in-
stantaneous amounts of inrush current from power grids. Electrical
distribution systems may experience serious problems if a power com-
pany does not have enough standby power to meet this instantaneous
demand. In fact, utilities that require large pumps for distribution
pumping often must get clearance from the power company before
starting a pump. Power companies may also require water utilities to
install soft-starter technology to reduce starting motor current.
Some small utilities with sufficient storage are able to shift pump-
ing to periods of low electrical demand and pay a reduced rate for
power. This strategy, however, is difficult to implement in large sys-
tems because of the excessive volume of storage that would be re-
quired.
Many utilities have resorted to variable-speed pumping to meet
variable demand. Variable-speed drives allow pumps to operate be-
low normal speeds to reduce flow and pressure output. Although
wire-to-water efficiencies may be low at lower speeds, the amount
of energy used is less, reducing energy consumption. Hence, peak
power consumption is reduced, because only a fraction of the total
potential pump power is used. The efficiencies achieved by variable-
speed pumping can be exceeded by properly designed and operated
constant-speed pumping systems and tanks.

Operation
378 Chapter Nine
Ideally, the most efficient means of conveying water is to have

no more than two pumps operating continuously at constant speed
near or at optimum efficiency. The number of pumps may increase
or decrease on a seasonal basis if demand patterns change, but it re-
mains constant otherwise. Tanks are filled when daily demand drops
and drafted when it increases; the draft and fill limits are set by the
boundaries of the pumps’ high-efficiency region. Of course, not all
systems have pump stations or tanks that are set up to do that. In
addition to the peak-shaving volume for energy considerations, tank
size should provide sufficient reserve storage for fire flow and emer-
gencies while maintaining tank turnover for water quality. It is not
easy to optimize the tank’s diameter versus height for the required
storage volume to satisfy all or most of these considerations.
Bibliography
American Water Works Association (AWWA). 2003. Principles and Practices of
Water Supply Operations—Water Treatment. 3rd ed., p. 210. Denver, CO:
AWWA.
AWWA. 1986. Maintaining Distribution-System Water Quality. Denver, CO:
AWWA.
AWWA. 1990. Water Quality and Treatment, A Handbook of Community Water
Supplies. 4th ed., p. 14.4. New York: McGraw-Hill.
AWWA. 2010. Water Quality and Treatment, A Handbook on Drinking Water.
6th ed. New York: McGraw-Hill.
Clark, R. M., and W. M. Grayman. 1998. Modeling Water Quality in Drinking
Water Distribution Systems. Denver, CO: AWWA.
Clesceri, L. S. (ed.), A. E. Greenberg, and A. D. Eaton. 1998. Standard Methods
for the Examination of Water and Wastewater. 20th ed. Washington, D.C.:
American Public Health Association, AWWA, and Water Environment
Federation.
Code of Federal Regulations. 2004. Title 14—Aeronautics and Space, Chap-
ter 1, Subchapter E Airspace, Part 77. Objects Affecting Navigable
Airspace. Washington, D.C.: Federal Aviation Administration, Depart-
ment of Transportation.
Code of Federal Regulations. 2004. Title 40—Protection of Environment, Chap-
ter 1, Part 141. National Primary Drinking Water Regulations. Washington,
D.C.: US Environmental Protection Agency.
Connell, G. F. 1996. The Chlorination/Chloramination Handbook. Denver, CO:
AWWA.
Crozes, G. F., et al. 1999. Improving Clearwell Design for CT Compliance. Denver,
CO: American Water Works Association Research Foundation (Awwarf).
De Zuane, J. 1997. Handbook of Drinking Water Quality. 2nd ed. New York: John
Wiley & Sons.
Gotoh, K. (ed.), J. K. Jacobs, S. Hosoda, and R. L. Gerstberger. 1993. Instru-
mentation and Computer Integration of Water Utility Operations, pp. 113–4.
Denver, CO: Awwarf and Japan Water Works Association.

Operation
Operation 379
Grayman, W. M., L. A. Rossman, C. Arnold, R. A. Deininger, C. Smith, J. F.

Smith, and R. Schnipke. 1999. Water Quality Modeling of Distribution System
Storage Facilities. Denver, CO: Awwarf and AWWA.
Guidance Manual for Maintaining Distribution System Water Quality. 2002.
Denver, CO: Awwarf and AWWA.
Hjertager L. K., B. H. Hjertager, N. G. Deen, and T. Solberg. 2008. Ex-
perimental and Computational Studies of Turbulent Mass Transfer in
a Mixing Channel. International Journal of Chemical Reactor Engineering
6:A105.
Jirka, G. H. 1999. Five Asymptotic Regimes of a Round Buoyant Jet in Stratified
Crossflow. 28th International Association of Hydraulic Engineering and
Research (IAHR) Biennial Congress, Graz (Austria).
Kirmeyer, G. J., L. Kirby, B. M. Murphy, P. F. Noran, K. Martel, T. W. Lund,
J. L. Anderson, and R. Medhurst. 1999. Maintaining Water Quality in Fin-
ished Water Storage Facilities. Denver, CO: Awwarf and AWWA.
Knoy, E. C. 1991a. Good Design Eliminates Frozen Storage Tanks. Opflow
17(2):1.
Knoy, E. C. 1991b. Solving Cold Weather Problems for Storage Tanks. Opflow
17(1):2.
Lee, S. Y., and R. A. Dimenna. 2001. Performance Analysis for Mixing Pumps in
Tank 18. Report WSRC-TR-2001-00391 prepared for US Department of En-
ergy, contract DE-AC09-96SR18500. Aiken, SC: Westinghouse Savannah
River Co.
Lindeburg, M. R. 2001. Mechanical Engineering Reference Manual. Belmont, CA:
Professional Publications.
Mays, L. W., ed. 1999. Hydraulic Design Handbook. American Water Works
Association. New York: McGraw-Hill.
Mays, L. W., ed. 2000. Water Distribution Systems Handbook. American Water
Works Association. New York: McGraw-Hill.
MixZone, Inc. 2005. www.cormix.com.
Moegling, S. D. 1992. Modeling the Effects of Reservoir Mixing on Water Quality in
Water Distribution Systems. Graduate thesis, University of Akron, Akron,
OH.
Nathman, J. C., R. C. Aguirre, and H. J. Catrakis. 2004. Far-Field Turbulent Mix-
ing Efficiency and Large-Scale Outer-Fluid-Interface Dynamics. 42nd Ameri-
can Institute of Aeronautics and Astronautics (AIAA) Conference, Sacra-
mento, CA.
Pollack, A., A. S. C. Chen, R. C. Haught, and J. A. Goodrich. 1999. Options for
Remote Monitoring and Control of Small Drinking Water Facilities, pp. 52–120.
Columbus, OH: Battelle Press.
Recommended Standards for Water Works [Ten States Standards]. 1992. Great
Lakes–Upper Mississippi River Board of State Public Health and Envi-
ronmental Managers. Albany, NY: Health Research.
Roberts, P. J. W., X. Tian, S. Lee, F. Sotirepoulos, and M. Duer. 2004.
Physical and Numerical Modeling of Mixing in Water Storage Tanks:
Progress Report. Denver, CO: Georgia Institute of Technology and
Awwarf.
Sanks, R. L., ed. 1989. Pumping Station Design. Stoneham, MA: Butterworth-
Heinemann.
Schlichting, H. 1968. Boundary Layer Theory. 6th ed. New York: McGraw-Hill.

Operation
380 Chapter Nine
Shan, J. W., and P. E. Dimotakis. 2001. Turbulent Mixing in Transverse Jets.

Report CaltechGalcitFM:2001.006. Pasadena, CA: Graduate Aeronautical
Laboratories, California Institute of Technology.
Ten States Standards. 1992. See Recommended Standards for Water Works. Great
Lakes–Upper Mississippi River Board of State Public Health and Environ-
mental Managers. Albany, NY: Health Research.
US Environmental Protection Agency (USEPA). 1974. Manual of Methods for
Chemical Analysis of Water and Wastes. Washington, D.C.: USEPA.
US Environmental Protection Agency (USEPA). 1991. Guidance Manual for Com-
pliance with the Filtration and Disinfection Requirements for Public Water Sys-
tems Using Surface Water Sources [Surface Water Treatment Rule Guidance
Manual]. EPA 570391001. Washington, D.C.: USEPA.
US Environmental Protection Agency (USEPA). 1999. Disinfection Profiling and
Benchmarking Guidance Manual. Appendix D, Determination of Contact
Time. EPA-815-R-99-013. Washington, D.C.: USEPA.
US Environmental Protection Agency (USEPA). 2006. Stage 2 Disinfectant and
Disinfection Byproducts Rule (Stage 2 DBP rule). EPA 815-F-05-003. Wash-
ington, D.C.: USEPA. Compliance Help: http://www.epa.gov/safewater/
disinfection/stage2/compliance.html#quickguides. Accessed February
2008.
von Huben, H. 1999. Water Distribution Operator Training Handbook. 2nd ed.,
p. 16. Denver, CO: AWWA.
Walski, T. M., D. V. Chase, and D. A. Savic. 2001. Water Distribution Modeling,
p. 31. Waterbury, CT: Haestad Press.
Walski, T. M., J. Gessler, and J. W. Sjostrom. 1990. Water Distribution Systems:
Simulation and Sizing. Chelsea, MI: Lewis Publishers.

CHAPTER 10
Maintenance,
Inspection, and
Repair
Jennifer Coon, C.H.M.M., C.E.T.
Why have a maintenance program? The answer is simple: Preventive

maintenance has been, and always will be, less expensive than cri-
sis maintenance. Preventive maintenance allows owners to identify
potential problems and develop solutions before the problems reach
crisis proportion. For example, it can be much cheaper to identify
and arrest coating failure and corrosion before they turn into metal
loss requiring more extensive repair. Additionally, tank painting, if
done properly, is typically required at intervals of 15 to 20 years. If the
coating adhesion is monitored regularly during inspections, topcoats
can be applied to the exterior to restore the aesthetics and extend life
of the original or underlying coating system beyond the anticipated
15 to 20 years. Topcoating can cost only a fraction of the cost of full
repainting.
Tank Evaluations and Resources

Three types of evaluations are recommended during the life of the
tank: (1) initial or baseline tank evaluations, (2) update evaluations,
and (3) operator evaluations. Several organizations have established
standards by which water storage tanks are evaluated and maintained.
AWWA publishes standards dealing with specific aspects of tanks.
Additionally, all water storage tanks should be compliant with any
local building codes.
381
Maintenance, Inspection, and Repair
382 C h a p t e r Te n
Initial Evaluation
An initial evaluation is a thorough evaluation performed to deter-
mine the tank’s structural, sanitary, safety, and coating condition. The
AWWA manual M42, Steel Water Storage Tanks, recommends that a pro-
fessional engineer familiar with the design and maintenance of water
storage tanks perform this type of evaluation. The initial evaluation
is the most detailed and intensive evaluation.
Update Evaluation
An update evaluation should be performed approximately every 3 to
5 years following the initial evaluation. The update evaluation is per-
formed to monitor changes in the coating condition and rate of corro-
sion and to verify that tank conditions have not changed significantly
since the previous evaluation or rehabilitation. The same person or
firm that performed the original initial evaluation should perform the
update evaluation. The update evaluation is not as time-consuming
and detailed as the initial evaluation.
The advantages of having an initial evaluation and subsequent
update evaluations performed by a professional are that these evalu-
ations will identify the optimum time for tank repainting and repairs,
and the owner can better plan for and budget for proper tank mainte-
nance or rehabilitation. The evaluations will identify all of the work
that is required to properly maintain the tank. This eliminates sur-
prises and change orders during a repair or repainting project.
Operator Evaluation
Tank owners should perform a cursory evaluation of the tank’s con-
dition at least annually. The purpose of this evaluation is to identify
items that can be easily remedied by maintenance personnel and to
indicate if any issues exist that require professional evaluation. If any
significant deterioration is found, or if the tank has been damaged
in some way, a professional should be called in to evaluate the prob-
lem. Items requiring basic maintenance can be remedied by operat-
ing personnel at this time. If the owner’s personnel are not properly
equipped or qualified to climb the tank, the professional person or firm
that performs the initial and update evaluations could perform this
function.
The advantages of operator evaluations are that any significant or
serious changes that may require further evaluation by a professional,
such as a potential leak or metal loss on the anchor bolts, can be iden-
tified by the operator. Also, routine maintenance can be performed by
water department personnel, thus saving the costs of replacing items
or repairing items at the next rehabilitation.

Maintenance, Inspection, and Repair 383
Resources for Tank Owners

r The AWWA Standard D100 for steel water storage tanks was
originally published in 1935 and has undergone continual up-
grading and modification.
r The National Fire Protection Association (NFPA) initially
adopted NFPA 22 Standard for Water Tanks for Private Fire Pro-
tection in 1914.
r AWWA M42 Steel Water Storage Tanks Manual. This manual
incorporates and updates much of the information contained
in AWWA D101 Standard for Inspecting and Repairing Steel
Water Tanks, Standpipes, Reservoirs, and Elevated Tanks for
Water Storage. The publication of D101 has been discontin-
ued.
Additional Steel Tank References
r AWWA D100-05—AWWA Standard for Welded Carbon Steel
Tanks for Water Storage
r AWWA D102-06—AWWA Standard for Coating Steel Water-
Storage Tanks
r AWWA D103-97—AWWA Standard for Factory-Coated
Bolted Steel Tanks for Water Storage
r AWWA D104-04—AWWA Standard for Automatically Con-
trolled, Impressed-Current Cathodic Protection for the Inte-
rior of Steel Water Tanks
r API Standard, ANSI/API 65-1992—Welded Steel Tanks for
Oil Storage
r API Standard, ANSI/API 653-1995—Tank Inspection, Repair,
Alteration, and Reconstruction
r API Standard, ANSI/API 620-1992—Design and Construc-
tion of Large, Welded, Low-Pressure Storage Tanks
r NFPA 22—Standard for Water Tanks for Private Fire Protec-
tion
Composite-Tank References
r ACI 371R-98—Guide for the Analysis, Design, and Construc-
tion of Concrete-Pedestal Water Towers
r AWWA D107—AWWA Standard for Composite Elevated
Tanks for Water Storage
Inspection and Repair by Operator

The first step of any preventive maintenance program is inspection
to identify the items requiring maintenance. Following are items that

should be inspected periodically by the operator and instructions re-

garding repair.
Site Access
Inspection
The tank and the site should be monitored for signs of unautho-
rized access and vandalism, which are a potential liability for the tank
owner. Signs of unauthorized access include damage to the tank or
site, graffiti on the tank or site appurtenances, paint chipping caused
by rocks being thrown at the tank, and bullet holes or indentions in the
steel caused by from guns being shot at the tank. Personnel should
look for damage or loose wiring in the site fence and barbed-wire
strands as well as gaps between the fence and the ground. The proper
operation of the gate locking mechanism, site motion detectors, site
lighting, and surveillance cameras should also be verified. If the tank
is equipped with an exterior ladder, the proper operation of its vandal
deterrent and locking mechanism should also be confirmed.
Repair
If the site is not already enclosed by a fence, a fence at least 6 ft
(1.8 m) tall and topped with barbed wire should be installed around
the entire tank site. Barbed-wire strands should be added to the top
of the fence if they are not already present. Personnel should then
regularly maintain the fence and barbed wire so that they are in good
condition. Any holes, broken wire, or bent sections should be repaired.
The fence should be close enough to grade to prevent intruder passage
under it. All barbed-wire strands should be taut. The fence should be
equipped with a gate or gates that can be locked whenever the site
is unattended. Vegetation should be regularly trimmed back from the
fence so that it does not damage or restrict view of the fence. All
light fixtures, motion detectors, and surveillance cameras should be
regularly maintained. If the tank has an exterior ladder, it should be
equipped with a locked vandal deterrent.
Site Maintenance
Inspection
The operator’s personnel should evaluate the condition of the tank
site not just for appearance purposes, but also to help protect the
tank from damage and corrosion. The presence of any trees, bushes,
or other vegetation touching the foundation, bottom plate, or tank
should be noted. Foliage traps moisture against the steel and creates
a damp atmosphere that can accelerate corrosion. These areas should
also be checked to see if grass clippings or other debris have accumu-
lated there. If saturated or eroded soil not caused by precipitation or

overflow effluent is noted around the base of the tank, a professional

engineer familiar with water storage tank issues should be contacted,
as this may signify a structural problem.
Repair
Trees and bushes should be trimmed back to prevent the limbs and fo-
liage from touching the tank. Vegetation should be trimmed so it does
not grow up on the foundation, base plate(s), and tank. Personnel
should remove any debris found on the foundation and base plate(s).
When the tank site is mowed or other maintenance operations are
performed with similar equipment, the discharge should be directed
away from the base of the tank to prevent any rocks discharged from
hitting the tank and damaging its coating. This will also prevent grass
clippings from accumulating on the foundations and base plates and
accelerating corrosion. Care should be taken that maintenance equip-
ment, trucks, and so on do not come into contact with the tank or
foundation.
If necessary, personnel should regrade the site so that the founda-
tion projects 6 to 12 in. (152 to 304 mm) above grade and adequate
drainage away from the foundation occurs. Rainwater runoff and
overflow discharge should be directed away from the foundation.
Foundation(s)
Inspection
The foundation should be checked to verify that it projects between
6 and 12 in. (152 and 304 mm) above grade and that there is proper
drainage away from the foundation. The foundation(s) should be ex-
amined for signs of settlement and the concrete inspected for evidence
of cracking, spalling, or exposed reinforcing steel. Deep cracks or ex-
tensive crumbling of the foundation signal, a potentially serious issue,
and a professional evaluation should be conducted. Also, if the foun-
dation tops are not approximately level with each other, this may be
evidence of differential settlement of the tank foundation, and a pro-
fessional evaluation should be conducted.
Repair
Personnel can apply a bonding agent and vinyl emollient concrete-
patching mortar to any deteriorated areas or voids found in the con-
crete foundation to build up the surface to its original contour. The
condition of this repair should then be monitored.
Grout, Fiberboard, and Sealant

Inspection
The condition of any grout, fiberboard, or sealant located at the in-
terface of the foundation and the bottom plate should be evaluated

for any cracks, voids, or deterioration. These can allow moisture to

build up between the tank foundation and bottom plate and cause
underbottom corrosion.
Repair
A flexible sealant can be applied to any deteriorated areas or voids
found in these materials to restrict the ingress of moisture through the
voids and under the tank bottom plate. The condition of these repairs
should then be monitored.
Anchor Bolts and Chairs

Inspection
The anchor bolts and chairs should be inspected regularly by the op-
erator’s personnel for any signs of corrosion and metal loss. Metal loss
is most critical at or below the nut. If metal loss is observed, the area
of metal loss should be measured and compared with the diameter of
the bolt where no corrosion has occurred. Typically, if the anchor bolt
has deteriorated by more than the thread depth (to the root diameter
or less), if any of the anchor bolts are bent or otherwise damaged, or if
any of the nuts are not completely threaded, a structural engineer fa-
miliar with the design and maintenance of water storage tanks should
be contacted to evaluate the anchor bolts and chairs further.
Repair
Personnel should keep the chairs free of debris, vegetation, and grass
clippings.
Manholes and Access Doors

Inspection
Manholes and doors should be checked periodically to confirm that
they are secured against unauthorized entrance. Unlocked manholes
and doors are a potential liability for the tank owner.
Repair
Personnel should install locks on manholes and access doors. They
should also replace any manhole gaskets that do not create a positive
seal. If any of the bolts have corroded, they should be replaced with
stainless-steel bolts to prevent rust staining from streaking onto the
tank surfaces.
Exterior Overflow Pipes

Inspection
The overflow pipe should be checked to verify that no potential exists
for cross connection between the potable water stored in the tank and

the water in the storm or sanitary sewer. The proper operation of a

flap gate or elastomeric check valve should be confirmed, and the
inspector should verify that no gaps exist between the flap gate and
the pipe. The condition of the screen should be assessed to verify that
it is adequate to prevent the ingress of insects and small animals. The
brackets and associated attachments should be evaluated for corrosion
and metal loss.
Repair
If gaps are noted, personnel should reposition the flap gate or elas-
tomeric check valve on the overflow pipe to eliminate them. Any dam-
aged screening on the overflow pipe discharge should be replaced to
prevent the ingress of insects and small animals.
Venting
Inspection
The proper operation of the clog-resistant vent and its pallets needs to
be checked before and after freezing weather. The condition of the vent
screening needs to be assessed to verify it is adequate to prevent the
ingress of insects into the tank. Shielding over any vertically oriented
screening also needs to be assessed so that wind-driven dust and
debris do not enter the tank.
Repair
Personnel should replace any damaged vent screens to prevent the
ingress of insects into the tank. Any damaged shielding over the
screening should also be replaced so wind-driven dust and debris
do not enter the tank.
Ladders
Inspection
All ladders should be carefully checked for deteriorated members
that might pose a danger to climbers. The ladder brackets and their
attachments to the tank and the ladder itself should be checked for
missing or deteriorated bolts and/or cracked welds. The rungs should
be inspected for metal loss, especially where they join the side rails.
Repair
If deteriorated members are noted on a ladder or its associated brack-
ets, repairs should be made. If the ladder is equipped with a safe-
climbing device, the device should be shielded from any paint or sol-
vent being used to ensure its continued proper operation.

Balcony and/or Platform

Inspection
Any balcony and/or platform should be evaluated to verify that the
surface does not retain water. If the surfaces allow water to pond, a pro-
fessional engineer should be contacted to discuss options for drainage.
The floor or safety railing access openings should be assessed. Clos-
able covers should be located over all floor openings, and protective
chains or bars should be located at all railing access openings.
Repair
Personnel should replace any missing covers from floor openings
and/or protective chains from safety railing.
Interior Lighting
Inspection
Personnel should check to make sure all interior lighting fixtures op-
erate properly. The condition and presence of the protective cages and
globes on the fixtures should be verified. The conduits associated with
the lighting should be assessed to confirm that they enclose all wiring
and are adequately supported. If they do not appear to be adequately
supported, a professional engineer should be contacted.
Repair
Personnel should maintain any interior light fixtures so that they oper-
ate properly. Burned-out bulbs should be replaced, as should damaged
protective cages and globes. If the fixtures or associated conduits ex-
pose wiring, it should be covered in accordance with National Electric
Code (NEC) guidelines.
Obstruction Lighting
Inspection
If obstruction lighting is required on a tank, personnel should make
sure it is operational and lit. The condition of any globes and bulbs
should be verified. The lighting should be evaluated to confirm that
it is adequately braced and that it and the associated conduits do not
have exposed wiring. The condition of the photoelectric cell should
be checked. The proper operation of the lighting should be verified
both at night and during the day to ensure that the lights are on when
required and off during daylight hours (unless otherwise required
by the Federal Aviation Administration) to reduce electricity use and
the frequency of replacing bulbs. If the fixture and its conduits do not
appear to be adequately supported, a professional engineer should be
contacted.

Repair
Personnel should replace damaged bulbs or globes. If wiring is ex-
posed, it should be covered in accordance with NEC guidelines.
Inspection of the Tank Exterior

Exterior Steel Welded Tanks and Leaks
The general condition of the exterior coating and any evidence of cor-
rosion should be monitored. If the exterior of the tank is in poor con-
dition, the condition of the interior coating could be as poor or worse.
The tank should be observed for signs of leakage or rust streaking that
a leak could have caused. Leaks can develop in flat steel plates, but
the most common sites are at seams and joints. Dark rust stains are
usually evidence of leakage. Small leaks in the seams may rust closed
over time, so water may not actually be running down the tank at the
time of the inspection. Also, if the tank is empty, a leak may not be
visible. If a leak or excessive corrosion is noted, a professional engi-
neer should be consulted regarding repair, as the leak may indicate a
more serious issue.
Exterior Steel Riveted Tanks and Leaks

The tank should be observed for signs of leakage or rust streaking that
a leak may have caused. Leaks can develop in flat steel plates, but the
most common sites for leaks are seams and joints. Dark rust stains are
usually evidence of leakage. Small leaks in the seams may rust closed
over time, so no water will actually be running down the tank at the
time of the inspection. Also, if the tank is empty, a leak may not be
visible. Rivet heads should be closely evaluated as extreme metal loss
on these items may indicate a structural issue. If a leak or excessive
corrosion is noted or severe corrosion observed on rivet heads, a pro-
fessional engineer should be consulted regarding repair, as the leak
may indicate a more serious issue.
Exterior Steel Bolted Tanks and Leaks

The tank should be observed for signs of leakage or rust streaking
that a leak may have caused. Leaks can develop in flat steel plates,
but the most common sites for leaks are seams and joints. Dark rust
stains are usually evidence of leakage. Small leaks in the seams may
rust closed over time, so water may not actually be running down the
tank at the time of the inspection. Also, if the tank is empty, a leak

may not be visible. Additionally, the gaskets or sealants between the

bolted joints should be evaluated to confirm they create a watertight
seal. Glass-lined coatings should also be checked for damage around
the bolts from over-tightening. If a leak or excessive corrosion is noted,
a professional engineer should be consulted regarding repair, as the
leak may indicate a more serious issue.
Professional Evaluation
AWWA recommends that water storage tanks be professionally evalu-
ated at least every 5 years and otherwise whenever conditions warrant
evaluation. A thorough professional evaluation will enable the tank
owner to accurately schedule required maintenance, prolonging the
structure’s useful life.
A professional evaluation should consist of a careful study of the
tank’s interior, exterior, foundation, and accessories. All necessary sur-
faces on the tank should be accessed by rigging and rappelling the
interior and exterior as required by the condition and design of each
tank.
Selecting a Professional Inspection Company

When retaining a firm or person to perform a professional tank evalu-
ation, the owner should refer to the AWWA manual M42 Steel Water-
Storage Tanks, which states: “The tank maintenance engineer should
have knowledge of the traditional engineering disciplines and have
specialized training and practical experience in the design, fabrica-
tion, erection, inspection, sanitary integrity, coating, and maintenance
of steel water-storage tanks.”
A tank owner who invests in a professional tank evaluation should
expect the evaluation to be thorough, professional, and complete. In
addition to supplying the usual components of a professional evalua-
tion, the evaluation team should identify any peculiarities associated
with the tank.
The Evaluation Report

A certified engineering report should be issued concerning the con-
dition of the tank. The evaluation report should describe the obser-
vations of the technicians and engineers and their recommendations
for optimum rehabilitation. Color photographs of the tank interior
and exterior provide aid to the tank owner in analyzing the data
presented. The report should address the condition of the tank—
structural, coating (including heavy-metal content analysis), corro-
sion control (including cathodic protection), safety (OSHA compli-
ance), operational and sanitary conditions, and compliance with other
applicable standards—and environmental considerations such as

containment and proper disposal of abrasive blast residue. Those

items not in compliance with current OSHA regulations concerning
safety, sanitation, and operation should be identified so that the tank
owner can make informed decisions regarding compliance with these
important issues.
A registered professional engineer familiar with the design, con-
struction, and maintenance of water storage tanks should certify the
report, which should serve as a decision-making document for
the tank owner. The report should also include budget estimates for
the recommended work, anticipated life of the coatings and the struc-
ture, and estimated replacement cost of the tank. Recommendations
should address what rehabilitation work needs to be performed to
meet the short-term and long-term needs of the water system.
Inspection of the Tank Interior

Underwater Evaluation Using a Diver
Although this method does not require the tank to be drained, it should
be taken off line and isolated from the system the entire time the diver
is in the tank. When performing an underwater evaluation, the diver
must wear a full dry suit and full-face diving mask to prevent contact
between the diver’s body and the potable water. Before entering the
tank, the diver and all related equipment must be thoroughly disin-
fected in accordance with the latest revision of AWWA Standard for
Disinfection of Water-Storage Facilities C652.
Underwater Evaluations Using a Remotely Operated Vehicle (ROV)

Remotely operated vehicles can be used to evaluate the interiors of
water tanks without interrupting service and isolating the tank from
the system by operating valves. These ROVs provide closed-circuit
video to an on-site technician who operates the unit. As with divers,
the ROV must always be disinfected before use in potable water tanks
in accordance with AWWA C652. The vehicles are typically made of
nonporous materials, and the bearing seals must be filled with a food-
grade glycerin.
Float-Down Evaluation
The interior of the tank is full of water as a float-down evaluation
begins. A field technician in a small raft evaluates the interior surfaces
as the owner drains the tank. The interior wet riser of elevated tanks
is typically evaluated by rigging after the float-down evaluation has
concluded. The duration of this evaluation is determined by the rate
at which the tank is drained.
Drained (Dry) Evaluation

During a dry evaluation, the tank is drained before the evaluation
and dewatered. The remaining water and sediment are removed from

the tank to access the bottom plates. Technicians get to the necessary
surfaces of the tank by rigging and rappelling the interior (including
the interior of riser pipes 36 in. [0.9 mm] in diameter and larger) as
required by the condition and design of each tank.
Both dry and wet evaluations have limitations. First, with a diving
evaluation, the diver is able to access all of the steel surfaces. During
a dry evaluation, the tank evaluation crew can only access, via simple
rigging from roof manholes and vents, surfaces that are adjacent to
ladders. However, the diver must evaluate the steel surfaces up close,
as the limited light does not allow him or her to move away from the
tank shell and evaluate the overall corrosion patterns inside the tank.
Also, because there is usually silt in the bottom of the tank, the tank
bottom cannot be as thoroughly evaluated by diving. When a diver
stirs up this sediment, visibility is impaired, diminishing the quality
of the evaluation of the tank bottom and the lower portion of the tank
shell. In addition, the diver is working with a limited supply of air,
which typically causes him or her to accelerate the evaluation. Some
of the physical tests that are normally performed, including adhesion
tests and dry film thickness readings, cannot be accomplished on the
submerged surfaces. Perhaps most importantly, only one diver sees
the tank interior, whereas it is ideal for several members of the tank
evaluation crew to visually assess the tank interior and evaluate the
problems found. This provides for greater accuracy in the evaluation.
Structural Evaluation
Structural evaluations are normally only performed if the owner or the
engineer believes a tank does not meet current structural standards or
if the structural integrity of the tank is suspect. Structural evaluations
should evaluate metal loss compared to the apparent or observed orig-
inal metal loss obtained by ultrasonic testing. A structural analysis is
not commonly required for properly maintained existing tanks unless
the tank has been modified (if, for example, equipment or antennas
have been added to the structure) or has experienced an extreme event
such as high winds or an earthquake.
The first step of a structural analysis is an engineering evaluation
of the tank to determine its condition. A structural engineer should
review deterioration of the foundation to determine its effects on the
tank’s structural integrity. A level should be used to determine if dif-
ferential settlement has occurred since construction of the tank.
The original design drawings should be reviewed for compliance.
Measurements should be taken in the field to analyze the tank and
anchorage for compliance with current structural codes and require-
ments that may have changed or come into effect since the tank was
originally designed and constructed. The latest AWWA standards and
local building codes should be used. Careful attention should be given

to the tank’s compliance with current design requirements for wind,

snow, and seismic loadings.
The original weld quality on welded steel tanks should be ver-
ified by ultrasonic or radiographic testing. The testing is normally
performed at locations as required by current AWWA standards and
evaluated in accordance with American Welding Society (AWS) stan-
dards for weld quality. If the owner has the original radiographs used
to verify the original weld quality and has confidence in their accuracy,
this additional weld testing may be redundant.
Specialized Inspections
Ultrasonic Thickness Measurements
Ultrasonic thickness measurements of the steel should be taken, and
areas of metal loss and deterioration should be analyzed for structural
deficiencies.
Coating Evaluation
The coating survey should include laboratory analysis of coating sam-
ples to determine the total lead content by weight. Similar tests should
be conducted for other regulated heavy metals such as arsenic, barium,
cadmium, chromium, mercury, selenium, and silver. Additionally, the
coating type, thickness, condition, and adhesion should be tested to
assess the ease of applying a topcoat to the existing coating. If such
tests indicate that topcoating is an option, recoating costs could be
significantly reduced.
Cathodic Protection Evaluation

Annual inspection of the cathodic protection system by the manufac-
turer or other qualified person is recommended. At a minimum, this
should include overall inspection of the entire cathodic protection sys-
tem (including removal of expended or damaged anodes, if required),
replacement of all defective parts, complete potential profile survey, a
physical check of the anode placement and wiring continuity, obser-
vation for corrosion at areas of exposed steel, and a written report.
Inspections Following Extreme Events

In the 1970s, national design standards first began to include proce-
dures for designing liquid storage tanks that resist earthquakes. The
basic design standards and codes for tanks, which focused on provid-
ing better details and structural resistance, were based on observed
behavior and problems. Design standards have evolved since that
time with regard not just to seismic design, but to wind-load design.
Older tanks may not meet these current standards. Therefore, inspec-
tion and upgrading, and then maintaining retrofits, may reduce the
effect of a seismic or other natural event on the tank. Additionally,

inspections can identify problems that potentially could arise during

freezing weather. Many of these problems could be easily repaired
and maintained before they cause hazardous conditions or before the
tank fails during freezing weather.
Owner-Performed Inspection
Immediately following a tornado, hurricane, major windstorm, or
earthquake, or during freezing weather, tanks should be evaluated for
any possible damage. If damage is evident, a professional structural
engineer familiar with water tank design and maintenance should be
contacted as quickly as possible to evaluate the structural condition
of the tank.
Professional Seismic Evaluations

In-depth seismic evaluations (as with structural evaluations) are typ-
ically above and beyond the standard initial evaluation. Seismic eval-
uations are typically performed only if the owner or the engineer
believes a tank does not meet current seismic requirements. Because
of the ever-changing seismic regulations and prescribed design loads,
some owners are compelled to do a seismic evaluation of their struc-
tures, especially those built before the advent of seismic design.
In addition to a professional engineer–performed initial field eval-
uation of the tank, the engineer should obtain and research all avail-
able original tank erection drawings, design calculations, specifica-
tions, as-built drawings, and other historical data. Based on the field
evaluation and the historical documentation review, a certified en-
gineering report should be submitted outlining the observations and
recommendations for replacement or retrofit and maintenance to meet
the objectives of the owner. The analysis should determine not only
whether the original design complies with the current seismic stan-
dards, but also whether the tank in its current condition complies with
these standards.
Because there have been significant changes in the AWWA design
and construction standards (especially in the approaches to design for
seismic loadings) and design philosophies, the owner may desire a
more complete structural evaluation. Accordingly, the entire tank and
anchorage system should be analyzed for compliance with present
structural codes and requirements, which may have changed or come
into effect since this tank was originally designed and constructed.
Careful attention should be given to each tank’s compliance with the
present design requirements for wind and seismic loadings. Addition-
ally, the original foundation design drawings and soil report should
be reviewed for compliance. This design review of the foundations
and the amount of reinforcing steel actually inside the concrete will
be based solely on the drawings, as there is no economical method
of verifying the amount and the location of reinforcing steel and/or

concrete construction practices. Without the original foundation de-

sign drawings, this review could not be reasonably performed, and
several assumptions would have to be made. The engineer should al-
ways provide a recommendation after reviewing the foundation de-
sign regarding the necessity of additional physical foundation evalu-
ation. If original drawings of the foundation(s) are not available, then
portions of the existing foundation may have to be excavated.
During the seismic evaluation, the original weld quality on welded
steel tanks should be verified by ultrasonic or radiographic testing.
The testing is typically performed at locations as required by present
AWWA standards and evaluated in accordance with AWS standards
for weld quality. If the owner has the original radiographs used to
verify the original weld quality and has confidence in the accuracy of
their radiographs, this additional weld testing may be redundant.
If one has not been recently performed, a soil investigation may be
part of the seismic evaluation of the tank. The additional information
gained from a site-specific soil investigation is important in assessing
the overall tank behavior. The soil information may also be useful in
determining the design load to comply with the building code require-
ments and to identify potential soil abnormalities that may affect the
performance of the tank. However, a soil investigation is not always a
necessary expense that adds value. It is typically in the owner’s best
interest to require soil investigations only when the information gath-
ered may have a significant influence on the outcome of the seismic
evaluation. It is recommended that a soil investigation be conducted
only for the following conditions:
r Tanks where the foundation is extremely sensitive to the in-
tegrity of the tank
r Sites believed to have a potential or known soil problems, or
r Sites where the potential properties suggest that the default
building code soil factors used in determining the seismic load
are not credible.
For example, when tanks are located in areas subject to soil lique-
faction or gross slip failures, additional site investigation and remedi-
ation may be required. The size, location, and type of tank influence
the relative value of a soil investigation. A tank of low height with
an unanchored flat bottom typically imposes less load on the soil and
may not be susceptible to soil and foundation problems during an
earthquake. Conversely, a large standpipe or elevated tank with sub-
stantial anchorage requirements may be significantly affected by the
soil behavior.
The existing building codes and national tank design standards
all specify factors to adjust the seismic design load for the site soil
classification. All of these documents also have a default value when

sufficient detail is not available. For many sites, the default site clas-
sifications are conservative for determining the design load.
Another important factor in assessing the need for a site-specific
soil investigation is the history of tank failures during seismic events.
The types of tank failures most often encountered are related to pip-
ing flexibility, damage to the shell anchorage, shell buckling, or slosh-
ing damage to the roof and roof support structure. Few foundation
problems resulting from earthquakes are reported. When foundation
problems are reported, they are often related to gross soil failures (e.g.,
a tank sliding down the hill) that may not be addressed by the typical
soil report or may be a consequence of inadequate anchorage design.
Tank Inspection Issues

Confined Space and Other Safety Issues
Personnel accessing the interior of a tank should be trained in proper
procedures for entering confined spaces. This includes training in
measuring atmospheric conditions for oxygen levels and lower explo-
sive limits, emergency response procedures, and roles of each of the
workers on the site crew. Worker training programs are available from
a variety of sources including OSHA, which offers outreach training
courses through its Outreach Training Program.
Before climbing a tank, the operator’s personnel should be trained
to work at heights and should be comfortable doing so. The worker
should use all appropriate safety equipment and follow all safety pro-
cedures. Whenever someone enters the tank, at least one additional
person should act as a ground person who is available to get help, if
needed. Emergency response procedures should be established and
reviewed with all crew members at the start of the tank access. Train-
ing may be available through an OSHA Outreach Training Program,
through your local fire department, or from recommendations by dis-
tributors of fall-protection equipment.
Tank Cleaning/Washouts
As water is held in the tank, suspended solids begin to settle out of
the water and onto the tank bottom. Without regular washouts, large
amounts of sediment may accumulate in the tanks. In addition, proper
evaluation of the interior surfaces of the tank cannot be conducted
with sediment covering the bottom of the tank.
Draining the Tank

Before scheduling work crews to wash out a tank, it is a good idea
to determine if the tank is equipped with a drain. If so, its location
should be noted.

Washing Out the Tank

The tank can be washed out with low-volume, moderate-pressure
(2,500 psi [17,237 kPa]) pumps, firefighting equipment, or other
means. Water should be sprayed on all interior surfaces to remove
as much residue as possible. In areas where sedimentation is a prob-
lem or where strict local environmental regulations apply, it may be
necessary to separate the sediment from the washout water and prop-
erly dispose of it by some means other than allowing it to enter storm
sewers or nearby streams. Also, care should be taken so that large
amounts of sediment are not allowed to enter the tank piping; this
could clog pipes or damage valves.
If the tank has been equipped with aluminum cathodic protection
anodes, many of them may have fallen since the previous washout.
Because these anodes may damage the interior coating, they should
be removed from the tank during the washout.
Operating Without the Tank

Operating without a tank may require notification of local businesses
and residents so that temporary large uses of water such as lawn
watering or equipment washing can be scheduled for other days,
thereby leaving the operator with adequate fire protection capacity.
It may be necessary to provide pressure relief valves for the one-tank
pressure planes when the single tank is out of service.
Refilling the Tank and Disinfection

The disinfection of water storage facilities should be done in accor-
dance with the latest revision of AWWA C652. This standard offers
three chlorination methods by which disinfection can be accom-
plished.
r Method 1. This method requires that the tank be filled with

chlorinated water (10 mg/L chlorine) for the sole purpose
of disinfecting the tank. After the required retention period,
the disinfection water is drained and the tank is filled with
potable water. After the potable water has satisfied bacte-
riological tests and is determined to be of acceptable aes-
thetic quality, the water may be delivered to the distribution
system.
r Method 2. This method requires that a chlorine solution
(200 mg/L) be applied with brush or spray equipment to all
parts of the tank that would be in contact with water when
the tank is full to the overflow elevation. After rinsing, the
tank is then filled with potable water. After the potable water
has satisfied bacteriological tests and is determined to be of
acceptable aesthetic quality, the water may be delivered to the
distribution system.

r Method 3. This method requires that the tank be filled to

approximately 5 percent of the total storage volume with a
mixture of potable water and chlorine containing 50 mg/L
of available chlorine. After a retention time of not less than
6 hours, the tank is filled to the overflow level with potable
water. After a 24-hour retention period, the water should be
tested. Once the water has been tested for bacteria and aes-
thetic quality, the water may be delivered to the distribution
system.
Of the three disinfection methods listed in AWWA C652-02,

Method 1 is the least popular because an entire tank full of water
must be wasted to accomplish disinfection. In addition to wasting the
water, discharging large volumes of highly chlorinated water is not
environmentally acceptable. The primary drawback to Method 2 is
that personnel disinfecting a tank must be equipped with proper res-
pirators and protective clothing to help protect them from the vapors
released into the air when chlorine is applied.

CHAPTER 11
Potable Water
Security
John McLaughlin, P.E.
Jordan, Jones and Goulding
The use of water as a leveraging tool in conflict is not new, or is the con-
cept of water security. Besides the air we breathe, water is the single
most critical element to human survival. In Water Conflict Chronology
(Gleick 2008), more than 100 incidents are documented in which water
was the cause of, or was integral to, a major conflict or event. These
events or types of conflicts are grouped into one or more of the fol-
lowing categories: control of water resources, water as a political or
military tool, terrorism, water as a military target, and disputes re-
lated to development of water resources. As early as 2,500 bc, water
was used as a military tool to help defeat an enemy. With humans’
reliance on safe and sustainable potable water, its use as a tool of war
and conflict should be no surprise.
Threats to Water Systems

Any lack of attention to or understanding of the critical importance
of potable water changed dramatically after September 11, 2001. Sud-
denly, the concepts of unconventional threats and how they could use
critical infrastructures against a population became real. An imme-
diate reaction to the events of September 11 was the introduction of
federal legislation to require US water systems to complete vulnera-
bility assessments (VAs) and emergency response plans (ERPs). This
federal legislation became Public Law 107-188, and it required every
public water system in the United States serving more than 3,300 peo-
ple to complete a VA and an ERP, on a staggered schedule, before
399
Potable Water Security
400 Chapter Eleven
December 31, 2004. The actual schedule for completion of the VAs
was the following:
r March 31, 2003, for systems serving a population of 100,000
or more
r December 31, 2003, for systems serving a population of 50,000
or more but less than 100,000
r June 30, 2004, for systems serving a population greater than
3,300 but less than 50,000
In each case, the system was to complete an ERP as soon as pos-
sible, but no later than 6 months after completion of the VA. The ERP
was to incorporate the results of the VA.
Water systems were not fundamentally insecure before Septem-
ber 11; most had fences, locks, and other systems to detect and delay
intruders. Larger water systems sometimes had guards and more in-
tricate electronic security systems. What fundamentally shifted was
the focus—away from protecting water systems against natural dis-
ruption and contamination and toward reducing the risk from an
intentional malevolent human attack (and not necessarily from an
international terrorist organization). Some of the most prevalent, best
documented, and least appreciated threats to water systems come
from disgruntled current or former employees, a lone vandal or a
group of vandals, and common criminals.
Almost monthly since September 11, news stories have docu-
mented break-ins at water facilities. These types of events almost
certainly occurred as often before 9/11, but they received little pub-
licity or attention. The main difference between the pre- and post-
9/11 incidents is that the Federal Bureau of Investigation (FBI) and
other law enforcement agencies paid the former—usually unorga-
nized attempts at vandalism—little attention. Nevertheless, the in-
cidents directly pointed to the need for better risk reduction at water
systems, which quickly began to improve formerly minimal security
practices.
Definitions
People tend to think of “providing security” at water systems, and this
chapter uses that terminology, but the real goal is to reduce risk by
eliminating vulnerabilities. This process of risk reduction ultimately
leads to the security that water system operators and the public seek. In
that connection, the following definitions are provided (Sandia 2002).
r Risk—Measure of the potential damage to or loss of an asset
based on the probability of an undesirable occurrence.
r Risk assessment—Process of analyzing threats to and vul-
nerability of a facility, determining the potential for losses,

Potable Water Security 401
and identifying cost-effective corrective measures and resid-

ual risk.
r Vulnerability—An exploitable security weakness or defi-
ciency at a facility.
r Physical protection system—Integration of people, proce-
dures, and equipment for the protection of assets or facilities
against theft, sabotage, or other malevolent human attacks.
The goal of any security system is to prevent an attack if possible. It

is generally not cost-effective, though, to stop any and all attacks. Can
a water system afford the level of physical protection used at a nuclear
facility or strategic military facility (armed and trained guards; “kill
zones,” clear areas outside the perimeter where deadly force is autho-
rized; and so on)? The answer to this question is almost always no,
and so water system management must be willing to develop strate-
gies to mitigate the consequences of an attack. This ensures that even
though a water system may not be able to stop the attack from happen-
ing, it can still cost-effectively reduce the overall level of risk. Conse-
quence mitigation, in addition to reducing the risk from an intentional
malevolent human act, also helps reduce the risk to a system from a
natural disaster. By providing a double benefit, consequence mitiga-
tion measures may be the most cost-effective risk reduction measures
of all.
Certain aspects of risk reduction at a potable water storage system
also benefit the system during natural disasters. The focus of this chap-
ter, however, is still on reducing risk from an intentional, malevolent
human act.
Types of Threats
A water-storage facility can be intentionally attacked in three ba-
sic ways: physical disruption, contamination (radiological, chemi-
cal, or biological), and interference with supervisory control and
data acquisition (SCADA), computer, and information technology (IT)
systems.
Physical Disruption
Much has been written about contamination being the worst-case sce-
nario for a water-storage facility. This is valid and worthy of discus-
sion, but perhaps the simplest and the most effective way of having
an impact on potable water storage is through physical disruption.
The amount of water that humans actually consume is only a frac-
tion of a percentage of the total potable water produced. In Milwaukee,
Wisconsin, and Albuquerque, New Mexico, for example, the percent-
age of potable water actually consumed is one-half to one-quarter of
1 percent of the total produced (Danneels 2001). Having storage, and

402 Chapter Eleven
therefore supply, of nonpotable water still allows fires to be fought,

industry to operate, contamination to be contained, and basic sani-
tation to continue. In cases where potable water storage is compro-
mised, potable water can be temporarily provided using bottled water,
mobile treatment systems, and bulk water that has been hauled in.
The purpose of this chapter is not to identify specific vulnerabili-
ties of a water-storage facility or to provide direction for adversaries;
therefore, the discussion will remain general. Physical disruption of
storage facilities generally requires some knowledge of the specific
water system to be truly effective. However, almost every water sys-
tem relies on critical storage facilities that, if eliminated, would criti-
cally disrupt its ability to supply water to the distribution system or
to critically important customers.
It is easy to disable or eliminate a storage facility without sophisti-
cated chemical or biological knowledge and equipment. Imagine the
damage that can be done to electrical systems with basic tools. Sugar in
the fuel tanks of emergency generators can create substantial damage.
Valves can be broken and extensively damaged without explosives.
Any simple Internet search reveals recipes for various homemade ex-
plosives capable of doing substantial damage.
Contamination
Three types of contaminants are of concern in water systems. These
are, in order of concern, biological, chemical, and radiological. Tra-
ditional water treatment has focused on removal or inactivation of
naturally occurring contaminants and contaminants unintentionally
introduced by humans. Each case of intentional or malevolent contam-
ination can cause unique problems. Besides the obvious—customers
getting sick or dying—one of the most likely overall problems is the
widespread public perception and panic that water is not safe to drink
(Burrows, Valcik, and Seitzinger 1997). Additionally, there is the prob-
lem of timely determination of what agent (or agents) has been intro-
duced.
Charlotte-Mecklenburg Utilities (in North Carolina) has dealt with
this issue twice. The first event was unintentional and involved Foam-
gate (Krouse 2001); the other occurred after 9/11 and was intentional.
In each case, even with rapid detection of the contamination, the test-
ing necessary to determine its exact nature and potential harmful ef-
fects was one of the most difficult parts of the entire response effort.
A chemical agent might be easily detected through the taste, odor,
or appearance of the water, especially if enough of the agent is present
to do physical harm to a person. The problem with radiological or
biological agents is that they are much more difficult to detect and
deal with. The first means of detecting these agents in water, even
large quantities of agents, might be through symptoms that do not
appear in an affected population until days or weeks later. Moreover,

symptoms could still be difficult to trace back to the water system

without good coordination, cooperation, and relationships between
the water system and public health personnel. Add to these difficulties
the fact that most potable water storage facilities are still not well
protected and thus contamination is relatively easy to accomplish. If
the threat has knowledge of the system and chooses a storage facility
that serves a critical part of the distribution system, the situation can
easily be made worse.
It would be difficult for a terrorist or other threat to have a broad,
long-term impact on a water system through use of a contaminant. At
a minimum, an adversary would need each of the following to create
widespread consequences: (1) specific knowledge of which storage fa-
cilities are in the most critical parts of the distribution system, (2) access
to agents and knowledge of which agent(s) might be most effective
and difficult to detect or inactivate, and (3) access to the equipment
to distribute the agent. Many agents can be introduced into a water
supply system. Any one of them can cause panic among the public
(Deininger 2000). This means that the contamination threat, though
difficult to carry out, cannot be minimized.
SCADA/IT Interference
A third method of disabling a water-storage facility is through cyber
attacks against a SCADA system. Fortunately, many water systems
still practice manual operation and allow their SCADA systems to
perform very little, if any, control. Those that do not practice manual
operation or that allow maximum control by their SCADA systems
run the very real risk of losing control through hackers entering their
system. These hackers can be current insiders or employees, disgrun-
tled former employees, lone thrill-seeking hackers, or a group of orga-
nized and highly capable hackers bent on significant and coordinated
destruction.
General Site Considerations

Location
Possible locations for existing storage facilities are as varied as each
facility’s vulnerabilities. It is difficult to conclude what would be an
ideal location from a security standpoint. A facility in a heavily popu-
lated area might be less vulnerable because it would be harder to attack
with so many people potentially watching; it is more critical, though,
because it serves more customers. In a remote setting, there are fewer
people to observe and possibly detect an intrusion, but the criticality
of the service area is probably lower. This section will only review the
more common security issues for remote and urban locations.

404 Chapter Eleven
Remote Location
When a storage facility is remotely located, its primary vulnerabil-
ity is that few people are around to detect an intrusion. Unless it is
a manned facility, such as a clearwell at a water treatment plant, the
only reliable means of detecting an intrusion would be through an
accurate, automated detection system, which many remote facilities
do not have. Even when a system has the capability of accurately de-
tecting an intrusion attempt at a remote site, response would normally
take too long because of the distance from a regular patrol area.
Urban Location
An urban or heavily populated location does not have the same vul-
nerabilities as a remote location, but several inherent vulnerabilities
still exist. Location in a congested area means that many more people
have close access to the site and are potentially aware of the facil-
ity’s importance. In general, in many urban areas, a lot of criminal
activity goes unnoticed and unreported. One thing common to virtu-
ally all water system facilities is the presence of graffiti, especially on
tanks. Most water systems have not worried about this in the past,
but the presence of graffiti points to the ease of access by and the
poor detection of intruders. In addition, because these storage facil-
ities are so close to large population centers and because they tend
to serve more critical customers, they are usually much more vital
assets.
In both remote and urban settings, the key is good detection. Ob-
viously, until a system accurately detects an intrusion attempt in the
first place, delay of the intruder will not be possible. A response force,
no matter how close or aware, will have not have any impact, and no
facility location will be safer than any other.
Accessibility
Accessibility, as discussed here, has to do with the number of peo-
ple allowed to access the facility. Almost all potable water storage
systems allow nonutility personnel to have unmonitored access to
storage tanks. These are most often employees of telecommunication
companies, electrical utilities, and other city departments. An unsci-
entific survey of results of many vulnerability assessments shows that
almost all facilities allow this access without maintaining any direct
control over who accessed the facility or when.
An equally critical vulnerability is the common practice by many
water systems of allowing too many of their own personnel to
have keys to facilities. Maintaining access control over the water
department’s own personnel is a more difficult problem to solve than

controlling access by employees of other agencies. Many tank person-

nel legitimately need access to the site, but many others have no real
need for keys or key cards.
Both cases require policies that call for monitoring of all personnel,
utility or nonutility, who might access a site. Access should be limited
to those who legitimately require it. Background checks should be
conducted on anyone who has access privileges.
Visibility, Perimeter, and Size

The visibility, perimeter, and size of any site are difficult to control.
Most sites are selected on the basis of hydraulic considerations and
the ability to acquire suitable property without any forethought to
vulnerabilities and security, a practice that must change. Even then,
cost will be a primary concern, and creative means of eliminating
vulnerabilities will be required.
A tank’s visibility is a given. Large ground storage tanks and el-
evated tanks of any size are obvious. What often works in the tank
owner’s favor is that most people take water tanks for granted and
forget they are there. As long as police, fire, and the tank owner’s
personnel do not do this, the visibility issue can be minimized.
What should always be avoided is taking a potentially bad visibil-
ity issue and making it worse. Neighborhood aesthetics may dictate
some screening, but hiding a tank too well makes it more difficult
to detect intrusions. Site perimeters should not be camouflaged or
screened unnecessarily. A tank owner should enlist the public rela-
tions staff to help explain this to the community. Whether the site is
large or small, the tank and related critical facilities should not be
placed near the perimeter. A small site may be dictated by economics
or location (tight, congested area), but as long as good detection of
potential intruders is maintained, additional layers of delay can be
added without huge cost, especially at a new site.
General Tank Considerations

Water storage tanks tend to be fairly standard in how they are designed
and accessed. The biggest differences are elevated versus ground stor-
age tanks and, in the realm of elevated tanks, leg supports versus
enclosed pedestal supports. There are differences in construction ma-
terial (steel, concrete, or a steel/concrete composite) and variations
within each category of tank (standpipes, clearwells). These specific
differences tend to have less impact on tank security. For this chapter,
only design elements that are pertinent to security of storage facilities
will be discussed.

406 Chapter Eleven
Elevated Storage Tanks

Elevated tanks generally offer a security advantage over ground stor-
age tanks in that they do not usually require integral, on-site booster
pumping. That is not to say that pumping is not part of the design of
an elevated tank system, but maintaining pumps on site is usually un-
necessary. Where pumping is not integral to the tank site, the number
of vulnerabilities is reduced accordingly.
Enclosed-Base Elevated Tanks

The two main types of enclosed-base elevated tanks are the fluted-
column tank and the pedestal/spheroid tank. Figure 11-1 shows a
typical fluted-column type of enclosed-base elevated tank; Fig. 11-2
shows a typical pedestal/spheroid type of elevated tank. Both types
usually contain a single pedestrian access door with an integral lock.
The fluted-column tank, with its (usually) larger-diameter base, can
often accommodate a vehicle protected by a lockable door similar to
an automatic garage door. Both styles of tanks almost always contain
in their bases tank-specific piping (Fig. 11-3), including the supply and
FIGURE 11-1 Typical fluted-column type of enclosed-base elevated tank.

FIGURE 11-2 Typical pedestal spheroid type of elevated tank.
FIGURE 11-3 Tank-specific piping for enclosed-base and pedestal spheroid

elevated tanks.

408 Chapter Eleven
FIGURE 11-4 SCADA components stored in enclosed tank base.
discharge piping, sampling ports, overflow piping, shutoff valve(s),

and an altitude control valve (if used).
An integral pumping system is commonly provided in the base
of a fluted-column tank, but usually not in a pedestal/spheroid tank.
In one sense, the large size of the enclosed base is one of the pluses of
a fluted-column tank, because more assets can be stored in the space.
The other side of the coin is that storing all critical assets in one place
creates potential vulnerabilities, because only a single door serves to
delay an intruder.
Within the base of each style are usually found SCADA compo-
nents such as tank pressure gauges, residual chlorine analyzers, re-
mote terminal units (RTUs), and radio/dialer equipment (Fig. 11-4).
Internal ladders providing access to the top of the tank bowl are
almost always located in the base of both styles of tanks. These ladders
allow direct access to the water storage portion of the tank by way of
direct hatch access or through the water-storage vent.
Multicolumn Tanks
Multicolumn tanks have many of the same features as an enclosed-
base tank, but without the same level of protection. Figure 11-5 shows
the base of a typical multicolumn elevated tank with a ladder guard.
Usually, multicolumn tanks have detached underground vaults to

FIGURE 11-5 Typical base of multicolumn style tank with ladder guard.
house critical piping and shutoff and altitude control valves (see
Fig. 11-6 for an example of this arrangement). SCADA components
and other related instrumentation are sometimes housed in the same
vault, but more often they are located in the open on the tank leg or
possibly in an unprotected shed detached from the tank.
FIGURE 11-6 Detached underground vault for piping and valves, multicolumn
tank.

410 Chapter Eleven
In the past, ladder access was available at ground level in many

systems, but that practice began to change even before September 11.
The practice of cutting off ladders 20 ft (6 m) or so above ground
level and adding locked access gates began as a way of controlling
vandalism, and it has now become an even more accepted means of
limiting access.
Ground Storage Tanks/Standpipes

In both ground storage tanks and multicolumn tanks, the piping,
valves, SCADA, and so on. are usually located in underground vaults
or separate sheds, or they are mounted outside on the tank itself. Lad-
ders are now being cut off above ground level, and lockable access
gates are being installed. Figures 11-5 and 11-7 show examples of how
this is accomplished on both multicolumn and ground storage facili-
ties.
As noted previously, ground tanks often differ from elevated tanks
because a booster pump station is often integral to ground tanks’
operation. Usually, both the tank and the pump station are located
on the perimeter of the same site. Often the pump station is a more
critical and easily accessed asset and becomes more of an issue to
FIGURE 11-7 Ladder cutoff and guard on ground storage tank.

secure. The principles of detection, delay, and response (Sandia 2002),

along with consequence mitigation, apply to the pump station if it is
located on site.
Construction Materials
Materials of construction play only a minor part in the security
of a tank. Almost all tanks are constructed of concrete, steel, or a
steel/concrete composite. A study of explosives, tank characteristics,
and materials of construction would be needed to determine which
of the three would be most susceptible to destruction. Suffice it to
say that a steel, concrete, or composite tank of proper structural de-
sign will withstand about the same level of explosive force, all other
factors being equal.
Water-Storage Vulnerabilities
This section is general and avoids describing specific methods and
means of contaminating or disrupting a water system through inten-
tional acts at a potable water–storage facility.
Most key elements of water system vulnerability have been cov-
ered previously. Specific locations exist on most storage facilities that
are the most vulnerable points. These include vents, sampling ports,
fiberglass hatches, and local chemical feed stations. Many utilities
have hatches that are lightly screened or not screened at all because
of wear and tear. Fiberglass hatches are common on ground storage
tanks and present a minimal barrier to a determined adversary. The
locks usually provided for metal hatch covers are of the type found
at the hardware store and are easily cut with large bolt cutters. Read-
ily accessible sampling ports, fire-hose connections, or local chemical
feed systems (for maintaining residual chlorine levels, for example)
are simple points of access for possible contamination.
Disruption of a water system through physical destruction at a
water-storage facility is a bit more difficult, but it is possible just the
same. It would take a large amount of explosive placed strategically
close to a storage facility to ensure complete destruction. Because of
this, we tend to focus on the possibility that an adversary would at-
tempt the same level of disruption through focused destruction of
critical piping, valves, booster pumping, or other on-site components.
As with a tank’s access hatches and vents, most enclosed tank base
doors or exterior vaults are only secured with a minimal hasp-and-
lock system.
SCADA/IT vulnerabilities are not currently severe or common,
because not many water systems rely on SCADA/IT to control func-
tions. Many utilities use SCADA only to monitor a few key parameters
and are alerted either when the signal is lost or when values are out

412 Chapter Eleven
of range. This does not mean, though, that these vulnerabilities can
be ignored. Reliance on SCADA signals without verification can be
dangerous, and often SCADA systems alarm over many minor occur-
rences, leaving operators to filter these alarms and potentially miss
something of real importance.
SCADA-related vulnerabilities will probably increase as security
systems (closed-circuit television [CCTV], perimeter alarms, and so
on) begin running signals through the same SCADA system used for
operational data. This also opens up a new avenue to be concerned
about: A disgruntled employee who controls not only the operation of
a system but the security system as well is known as a super insider.
Effective Security/Risk-Reduction Practices

All security or risk-reduction measures can be placed in one of sev-
eral categories. The major categories are physical protection systems
(PPS), operational security (OS), and consequence mitigation (CM).
Within the PPS are three subcategories: detection, delay, and response
(Sandia 2002). The basic concept is to try to prevent an attack from
occurring through PPS and OS (and, as a result, through effective
detection, delay, and response). The CM piece of risk reduction auto-
matically presumes that the attack has occurred and was successful.
Through good CM, a system can effectively respond to an event and
minimize the damage. Water systems have an inherent ability to mit-
igate consequences, because they face similar issues every day when
lines break, power goes out, spills occur, and storms move in. In some
instances, it is probably more cost-effective for the same risk reduction
to focus energy not on preventing the attack but on mitigating its con-
sequences. (The cost and physical difficulties of protecting every part
of a water distribution system, or even the most critical parts, would
be extreme. However, most systems incorporate beneficial elements
such as redundant facilities, system loops, and interconnects. A rapid
response by personnel trained in these matters will almost certainly
reduce the attack’s effectiveness.) This does not mean to ignore the
effort to prevent an attack; it just acknowledges that no water system
can truly afford to prevent every attack from all possible threats.
Physical Security
Physical protection systems are security measures such as CCTV (cam-
era) systems, motion sensors, alarms, fences, locks, and guards. The
basic concept of PPS is to detect an adversary as early as possible.
Detection means not just having a camera system record an intruder,
but having a person assess the alarm or image and react quickly and
effectively to alert whatever response mechanism is planned. Delay is
the combination of measures that will slow an adversary who is on the

path to the water-storage facility. As noted, early detection followed

by effective delay is the ideal sequence for PPS. Response comprises
the time and process involved to intervene with the adversary. If the
delay is inadequate, the adversary will succeed in carrying out the
malevolent act before the response arrives, and therefore the response
is ineffective. A response that arrives on scene and reaches the adver-
sary in time, but fails to intercept the adversary, is equally ineffective.
An example of this would be having an unarmed guard trying to stop
a group of heavily armed adversaries. Although the unarmed guard
may arrive in time, he or she can do little to stop the adversary.
Detection, delay, and response comprise a three-legged stool.
Without all three legs in place and of equal strength, the stool will
not stand.
In addition to the information provided here, the reader should
review information provided in “Guidelines for the Physical Secu-
rity of Water Utilities,” a Water Infrastructure Security Enhance-
ment guidance document produced by the US Environmental Pro-
tection Agency (USEPA) and funded by the American Society of Civil
Engineers (ASCE), American Water Works Association (AWWA), and
the Water Environment Federation (WEF). Also see the USEPA’s Water
and Wastewater Security Product Guide at http://cfpub.epa.gov/
safewater/watersecurity/guide/tableofcontents.cfm.
Detection Practices
Digital CCTV
Many utilities installed CCTV capability before September 11. Some
of these provided digital image storage. The majority used tape and
relied on an operator to see an event in real time or to forensically
view what happened. After 9/11, digital CCTV systems became more
prevalent. These systems store images in digital format and provide
an alarm if the viewed image deviates from a stored baseline image.
In such a case, in addition to providing the alarm, they pull up the
correct segment of video image, including the moments immediately
preceding and following the event.
With any camera system, lighting conditions and clear lines of
sight are critical. An uninterrupted fence line and clear areas at least
15 ft (4.5 m) outside the fence line are essential to successful early de-
tection. Adequate lighting, properly designed with the camera system
to provide optimum contrast, is also essential. Lights should be the
quick-strike type so that after a power outage has been resolved, it
does not take several minutes for the lights to warm up. (Quick-strike
lights come up to full candlepower almost instantly after power is
restored. They do not operate without power. The best means of pow-
ering lights and other critical functions during a power outage is to
provide a generator.)

414 Chapter Eleven
Hand in hand with these systems is the ongoing maintenance of

each part of the tank and the tank site. Areas inside and outside of
fences must be kept clear, lights that burn out must be immediately
replaced, and any camera system must be designed to work with the
light level available. These practices fall somewhat into the operation
systems category because they are policy-level practices that humans
take care of operationally. They are listed here, though, because they
are also integral to physical security, and they clearly demonstrate the
need for all protection systems to be just that—systems. Table 11-1
shows some basic comparisons of CCTV technologies, with pros and
cons for each.
Perimeter Detection Systems

Numerous types of perimeter detection systems are available. The
highest levels of these involve multiple integrated systems including
combinations of microwave, infrared, capacitance, taut wire, and fiber
optic. Fiber-optic technology can be cost-effective and can be adjusted
or tuned to minimize nuisance alarms. All physical alarm systems
must still rely on a human to assess the alarm and react properly. Table
11-2 shows some basic comparisons of detection technologies, with
pros and cons for each. As with the CCTV systems, the training and
policies necessary for this are discussed in the section “Operational
Security.”
Guard Dogs or Geese

Depending on the criticality of the facility and whether it is operator
attended or not, trained guard dogs may be an option. This option
obviously carries certain maintenance and liability issues, but it may
be a valid option where human monitoring is difficult, requires aug-
menting, or is impractical. Similar to guard dogs, but less of a liability
concern, are geese. The mess and maintenance for geese may be a prob-
lem, but they are very good at sounding an alarm. Once the alarm is
sounded, a human must intervene effectively, or the alarm has not
been fully assessed. An alarm without human assessment is not an
alarm at all.
Access Control
Controlling access is another key component of both detection and
delay. Access control can be as simple as basic door and window
locks or it can comprise state-of-the-art biometrics. Basic lock-and-key
systems can be effective against many adversaries, but they require
strict key-control policies that are practiced and enforced. If everyone
has a key to all facilities and assets, locks cease to be effective. Good
key control can detect and delay both insider and outsider adver-
saries. If padlocks are used at remote storage facilities to which other

Photographic
Technology Comments
Night-vision camera r Good for day and night viewing
r Will not have to redo or add lights
r Expensive
Black and white (B&W) r Good for day and lower-light vision
camera r Inexpensive
(recommended) r Not good for dark conditions
r Not as easy to distinguish during the day
r Will have to redo site lighting to have
effective monitoring
Color camera r Good for day viewing
(not recommended) r Not good for low-light or dark conditions
r Expensive
Day/night (color/B&W) r Color is good for day viewing
r B&W is better for night viewing
r More expensive than B&W or color
Recording Technology
No recording r Must monitor at all times to be functional
r Nothing is available that can be used for
prosecution
Tape recording r Used for backup validation of alarms
r Hard to find previously recorded moments
r Cannot record while viewing a previously
recorded moment
Digital recording r Used for backup validation of alarms
r Begins recording based on motion in the
field of view
r All recordings are date/time stamped for
ease in finding a particular moment when
viewing
r Accessible from a remote location
r Images are in PC-friendly format and can
be stored electronically indefinitely
TABLE 11-1 CCTV Summary

416
Types of
Detection Technology Pros Cons
Fence sensors r Vibration r Most economical and easiest to r Must have properly installed and maintained fence lines
install of the fence sensors r Prone to all types of vibrations, which can be minimized
r High probability of detection with enhancements:
r Weather sensor station—feeds weather information to
field processor, which then adjusts its vibration alarm
sensitivity
r Pulse count accumulator—sensitivity is determined by
choosing number of pulses needed to create an alarm
r Nuisance alarms can be caused by shrubbery, trees,
animals, and severe weather that causes fence to
vibrate
r Taut wire r Not overly sensitive to wind r Regular tensioning maintenance is required
r Very reliable r One of the most expensive fence sensor systems

r Low false-alarm rate and low because of laborious installation and maintenance time
nuisance-alarm rate
r Fiber optic r Immune to electrical or r The more activity there is at fence, the lower the
electromagnetic interference sensitivity setting
(EMI) disruption r Sensitive to extreme temperature changes and blowing

r Intrinsically safe and uses very debris

stable equipment, resulting in r Could be sensitive to large-animal activity
high reliability r The fence must be stable, free of vibration, and in good
r Adjustable sensitivity. condition

r Strain-sensitive r Capable of “hearing” what may r Very sensitive to high-EMI sources (for example,
cables be causing an alarm (similar to substations) and radio frequency interference
pressing ear against the wall) r Sensitive to poor fence construction or maintenance
r E-field r Self-adjusting circuit rejects wind r Adverse weather such as rain, snow, and lightning can
and ambient noise create problems
r Extremely low nuisance-alarm r Vegetation and animal movement can cause sensors to
rate react
r Capacitance r Weather and EMI/radio frequency r Generally mounted on top of fence, so use in conjunction
interference (RFI) have no effect with another type of sensor on lower part of fence fabric
on sensors’ ability. r Anything making physical contact that changes fence
characteristics may cause an alarm
In-ground r Balanced r Mostly immune to weather and r Should use additional surveillance/detection when
sensors pressure line environmental noise working with large expanses of concrete
r Tree roots may cause problems when tree blows in wind
r Sensitive when in close proximity to roads/rails due to
machinery.
r Ported coaxial r Mostly immune to weather and r Avoid installing under chain-link fences; install at least
buried line environmental noise 3 ft (0.9 m) above buried metallic pipes
r Susceptible to buried metal
r Affected by high-EMI sources such as large electrical
equipment or substations (should not be used in close

proximity to these areas)

TABLE 11-2 Perimeter Detection Technologies (Continued)
417
418
Types of
Detection Technology Pros Cons
r Buried fiber r Mostly immune to weather and r Must be installed away from poles and trees at a
optic environmental noise distance equal to at least the height of the pole or tree)
r Immune to electrical or EMI r Should not be installed in or under concrete or asphalt
disruption r Susceptible to erosion where either more exposure or
r Adjustable sensitivity deeper burial affects the sensitivities
r Sensitive to tree roots as the tree blows in wind
r Buried r Mostly immune to weather and r Sensitive to medium in which geophones are buried
geophone environmental noise r Sensitive to trees, fences, light poles, and telephone
poles, which can trigger the alarms when blowing in wind
Volumetric r Active infrared r Send multiple-beam pattern, r Precise alignment of sensors is critical
sensors increasing coverage r Not good with hilly terrain
r Good probability of detection r Sensitive to snow and grass around the sensors
r Available in portable versions r Sensitive to fog, heavy rain, and dust
r Narrow detection zone good for r Sensitive to vegetation overgrowth
monitoring perimeter sectors
r Microwave r Can be used to monitor an area r Sensitive to high-frequency spectrum
or a definitive perimeter line r Sensitive to areas that contain strong emitters of
r Use monostatic sensors where electric fields (radio transmitters) or magnetic fields

well-defined area of coverage is (large electric motors or generators)

needed (400 ft [122 m] coverage) r Can interpret ionization cycle created by fluorescent
r Bistatic sensors can be used up bulbs as motion
to 1,500 ft (457 m) r Potential health hazards

r Passive r Send multiple-beam pattern, r As ambient temperature approaches temperature of
infrared increasing coverage intruder, sensor is less likely to respond
r Good probability of detection r Sensitive to all heat sources (heaters, animals, and
so on)
r Precise alignment of sensors is critical
r Not good with hilly terrain
r Sensitive to snow and grass around the sensors
r Sensitive to fog, heavy rain, and dust
r Sensitive to vegetation overgrowth
r Passive r Greatly reduces false-alarm rate r Reduces probability of detection since both sensors
infrared/ if used in predictable and/or must positively detect before sending an alarm
microwave controlled environment r Has all the cons of each technology
r Cost effective (cheaper than r Potential microwave health hazards
purchasing two individual
sensors)
r Radar r Good for detecting helicopter or r Susceptible to uneven terrain
plane intrusions r High maintenance
r Potential radio-frequency health hazards

Video sensors r Motion r Can help to limit false alarms r Needs lighting
detection r Provides record of events during r Needs unobstructed viewing
an intrusion
r Monitoring field can be
manipulated

r Typically used in conjunction with
other monitoring technologies
TABLE 11-2 Perimeter Detection Technologies (Continued)
419
420 Chapter Eleven
utilities and/or agencies may need access, the tank owner should avoid
daisy-chain systems (several interlocking padlocks); all that is neces-
sary for an intruder to do is to break the weakest lock.
Swipe cards and/or personal identification number (PIN) access
control can be more secure and allow easier “key” control. The person
must remember his or her code and remember to carry his or her key.
An advantage of these systems is that they allow logging of who enters
the facility—or at least of whose card and PIN were used to enter. This
may not stop the adversary (the wrong person with the right key or
access code can enter), but it will dissuade those who want to escape
undetected.
Biometric systems control access by using characteristics and traits
that are unique to an individual. Among the most common are finger-
print and retina/eye scanners. These systems are virtually impossible
to trick, and they do not involve having to carry a key. Their cost may
prevent widespread use, but they can be especially effective against
an insider or as a second layer to a perimeter detection system for an
especially critical facility. See Table 11-3 for a further breakdown of
various access control systems.
Glass-Break Sensors
Delay and response are most effective when there is early detection. If
an adversary gets through a fence or other outer perimeter undetected,
the time available to a response force for intervention is greatly dimin-
ished. However, using glass-break sensors on building windows may
be necessary if perimeter detection at a fence line is not available or
practical. Certainly, it is preferable to have the extra distance and de-
lay, but short of moving entire facilities, that may not be possible. This
type of sensor may also be considered a layer in a detection system
for a highly critical storage facility or where threat by an insider is the
main concern.
Door Alarms
Door alarms, too, are more appropriate when the adversary is an in-
sider or as an extra layer in a detection system. The use of alarms for
storage facilities within the property’s perimeter can detect an insider
who, although legitimately within the perimeter of the facility as a
whole, may need to be restricted from entering key buildings that
house specific assets.
Contaminant Detection Technology

Contamination is less likely to occur than physical disruption and
may not have the same impact. Contamination may be more difficult
for an adversary to accomplish, and detecting such an attack is also
much more difficult. Current technologies generally detect contam-
ination by looking at the effect the contaminant has on certain key

Technology Entry Method Comments
Photo ID r Verification of a personal picture
r Procedural—relies on identification of person by guard
on a badge to actual person r Relies on guard for all access control
wearing badge
Stored-image badge r Verification of picture on badge to r Procedural—relies on identification of person by guard
stored picture of same person to r Relies on guard for all access control
actual person wearing badge
Personal identification r Correct combination of numbers r PIN entry can be coded per entry point, but not per person
number (PIN) entered on keypad for entry into r Automated process
restricted area
Key card entry r Automated verification of card r Key cards can be coded per entry point and per card
with predetermined criteria for r Access can be denied to some cardholders and allowed to
access to restricted area others, or access can be only during certain times of day
r Access privileges can be modified
r Automated process
Key card/PIN entry r Verification by matching PIN to r Must have both the card and PIN for access
badge number for entry r Access can be denied to some cardholders and allowed to
others, or access can be only during certain times of day

r Automated process
Biometrics r Verification of a personal r Characteristics include fingerprints, retina or voice
characteristic to authorize access recognition, or face scanning
to a restricted area r Each characteristic is unique to the individual

r Access can be denied to some individuals and allowed to

others, or access can be only during certain times of day
r Automated process
421
TABLE 11-3 Entry Control Summary

422 Chapter Eleven
indicators—among the most common residual chlorine and oxida-

tion–reduction potential (ORP). The theory behind residual chlorine
analysis is that a biological contaminant exerts a chlorine demand and
therefore creates a drop in the residual. This unusual or unexpected
drop would raise an alarm, but there would be no specific information
about what caused the drop and whether it was intentional or natu-
ral. The same is true for ORP detection. This indicator may react to
more contaminants, including chemical and biological contaminants,
but there is no way of identifying a specific agent or of determining
whether contamination occurred naturally or intentionally.
Several criteria should be considered when deciding whether to
implement an early warning system for water system contamination.
In no particular order, they are:
r Provides warning sufficiently ahead of time to allow for
proper action
r Is economically affordable
r Requires little skill or training
r Is flexible enough to cover all possible threats
r Is able to identify the source
r Is sensitive to changes at regulatory levels
r Provides minimal false-positive and/or negative results
r Is durable and robust
r Provides results that are reproducible and verifiable
r Can be operated remotely
r Has year-round all-climate functionality
Any decision to choose an early warning system must be made

locally, and the relative costs (monetary, physical, social, and organi-
zational) must be weighed against the relative benefits. It is also im-
portant to keep in mind the relative infancy of this technology. Little
is known about which contaminants the technology might most
accurately detect, and any early warning system currently consid-
ered would not likely be able to score high on all the criteria just
listed.
The following sections detail types of systems and tools that will
likely be used when early warning systems become more effective and
prevalent. For much more specific information on planning, design-
ing, implementing, and operating an early warning system, please
refer to “Early Warning Monitoring to Detect Hazardous Events in
Water Supplies,” from which much of this contaminant monitoring
information is taken (Brosnan 1999).

Contaminant Analyzers
Current technologies for detecting a contaminant look at its effect on
certain key indicators. Among the most common are residual chlorine
and ORP, described previously. The very nature of these types of de-
tection means the contaminant is already present in the system and
its consequences must be mitigated.
Another technique is to use biological analyzers—organisms that
react in certain ways to any of several toxic agents. Their reaction is
tied to an electronic signal that creates the alarm. The problems here
are the lack of any specificity as to the cause of the alarm and the
potential for false positives or negatives. A few examples follow:
r In the dynamic fish test, golden ides are exposed to an artifi-

cial water flow/current, which they normally swim against. If
they detect an upset condition, they turn to avoid it, and this
action would be detected and registered. Similar techniques
have been used in Europe since the 1970s.
r In the dynamic daphnia test, water fleas are placed in a con-
trolled column of raw water and exposed to several infrared
light beams, which they regularly interrupt and which indi-
cates a regular level of activity. If a contaminant is introduced
into the water, the activity level initially increases and then
sharply declines because of the death or incapacitation of the
daphnia.
r Recently, some locations have used the mussel as a monitoring
indicator. The theory is that when mussels are subjected to a
contaminant, their shells close at low contaminant levels and
then open wide at severe levels. The monitor takes several
mussels and glues one half of the shell to a wall. The other
half of the shell has a magnet attached that contacts a reed
switch to indicate an open or closed position. Electromagnetic
sensing between the two shell halves can indicate their interim
positions between fully opened and fully closed.
r Delayed algal fluorescence and luminescent bacteria monitors
use the principle that the presence of a contaminant dimin-
ishes the luminescent/fluorescent level of either the algae or
the bacteria.
While these methods may not be desired for use at this stage, they
give an indication of the body of knowledge available to enhance
security at all your facilities.
Technology to monitor and analyze contaminants is constantly
being developed and perfected with a goal of providing accurate,
real-time capability. Already the Sandia National Laboratories staff

424 Chapter Eleven
have developed the ChemLab (microChemLab), a palm-sized an-

alytical laboratory that can virtually instantly detect any number of
chemical or biological contaminants (Sandia 2002). This type of tech-
nology, which can accurately detect and identify an agent in real time,
appears to represent the future of contaminant monitoring.
Placement of Detection Devices

Accurately detecting a contaminant is only the first step. Where do you
put these analytical devices to do the most good? How do you know
where the contaminant originated and where it may be headed in the
distribution system? Several hydraulic models on the market provide
a level of water quality prediction. The most widely used for con-
taminant transport are Haestad’s WaterCad, MWSoft, and MIKENET
from the Danish Hydraulic Institute. MIKENET is already being used
in Europe. The model takes analytical measurements from a series of
parameters to detect a contaminant and then applies its algorithm to
predict the fate and transport of the contaminant from start to fin-
ish. Critical to this or any other model is its calibration to real-world
conditions, the number and locations of analytical devices, and the
operator’s knowledge of the system. As is the case for any detection
effort, a contamination event cannot be considered truly detected until
the alarm has been accurately assessed.
For water-storage facilities identified as critical through pairwise
comparisons, fault tree analysis, or accurate hydraulic modeling, real-
time contaminant analyzers should probably be located on site.
Delay Practices
Delay measures generally are the most cost-effective part of a risk
reduction system that comprises detection, delay, response, and con-
sequence mitigation. There is a multitude of number and types of
measures; the only limits are the constraints of the particular site.
Whether it is an operator-staffed facility such as a water treatment
plant clearwell or a remote, unmanned facility such as an elevated
water storage tank, the most common delay features are fencing and
gates. As with any protection system, fencing and gates are useless
without proper maintenance and training of the staff on how to max-
imize their effectiveness.
A simple way of making a regular chain-link fence more secure
is to use razor wire at the top of the fence instead of three strands of
barbed wire. Traditional fences consist of 6-ft to 8-ft (1.8-m to 2.4-m)
chain link with three strands of barbed wire on outward-facing out-
riggers. Where necessary and practical, the fence can be made more
secure by replacing the three-strand barbed wire with at least one coil
of concertina or razor wire (Fig. 11-8). Even more delay can be built
in by using two layers of fencing. This system is prevalent at critical

FIGURE 11-8 Chain link fence with razor wire at top.
military or nuclear sites and may be appropriate for certain water-

storage facilities.
If the facility is highly visible and is located in a neighborhood,
ornamental or architectural-type security fences may be necessary.
Figure 11-9 shows an example of an ornamental fence that can also
provide security benefits. There are numerous varieties of this type
of fence (the example in Fig. 11-9 is from Delgard) that can help with
both security and public acceptance.
Regardless of the type of fencing used, both fence and clear areas
must be properly maintained. This can be assured by establishing and
complying with a policy to regularly check the entire fence line—that
is, to perform a touch test on the entire perimeter.
Finally, tamperproof nuts and bolts for gates and fences should
always be used. When reviewing the effectiveness of a perimeter fence,
one of the first checks is to see if the nuts and bolts can be loosened
by finger pressure only. This is frequently the case, and it negates the
effectiveness of hardened locks, razor wire, and the like.
At perimeter entrance points for personnel, gates with effective
locks, swipe card, or biometric access control are effective. Because ve-
hicle access is commonly needed at water-storage facilities, the same
locking systems as used for personnel access should be used. Figures
11-10 and 11-11 show examples of vehicle gate entrances at remote
sites. Usually, the gates are only of the vehicle-access type; because

426 Chapter Eleven
FIGURE 11-9 Architectural security fence.
the sites are unmanned, they are almost always accessed by vehicle. If
gates exist but vehicle access is no longer allowed, Jersey-type barri-
ers are very effective. These come in various forms, including plastic
barriers that can be filled with liquid to add weight. If they are to be
effective, they must remain filled with liquid.
FIGURE 11-10 Typical chain link entrance gate.

FIGURE 11-11 Typical architectural gate and multilock system.
A special caution needs to be sounded against using daisy-chain

locks if at all possible. These systems are allowed to exist because of
the normal practice of permitting those other than the tank owner and
personnel (e.g., people from the phone company, emergency services,
police, and so on) to have unrestricted access to a site because they
have equipment there. As noted elsewhere in this chapter, this practice
should be stopped or at least severely restricted.
In addition to preventing unknown personnel from having unre-
stricted access to a site, it is important to have an effective lock. Figure
11-12 illustrates the concept of the weak link in the chain, in which a
simple lock of the type available at a hardware store is all that stands
in the way of an adversary.
r An array of delay features can be placed between the site
perimeter and the storage tank itself. Vehicle barriers in zigzag
patterns are very effective. Additional layers of fencing will
delay an adversary who is on foot. Use reinforced glass for all
exterior windows.
r Depending on the nature of the operations at the site, cer-
tain assets may be contained within a building on site. The
building presents several opportunities to delay an adver-
sary. Heavy exterior metal doors should be installed. Locks
and hinges on all exterior doors should be covered with steel
plates using tamperproof screws. Reinforced glass should be

428 Chapter Eleven
FIGURE 11-12 Daisy-chained locks with weakest link.
used for all windows (if applicable). Figure 11-13 shows a typ-
ical entrance configuration for a fluted-column enclosed-base
elevated tank. If this configuration is not practical, bars, cages,
fence, or mesh can be installed on the inside of the window
frame. Of course, these measures are worth nothing if doors
are not kept locked and if there is no key control.
Presuming the adversary gets to the asset before a response ar-

rives, further delay measures can still be used. The most common and
simplest is to build a steel cage around the actual asset. The design of
this barrier must allow for adequate normal maintenance, but it can
be very cost-effective.
Response Practices
The response component of physical security should focus on provid-
ing the water tank owner and staff with backup communications sys-
tems for all possibilities. This should include landline phones, cellular-
phone backup, and even radio systems tied into emergency frequen-
cies. Local law enforcement should be made fully aware of all facility
locations and should train on these sites. Water system staff should
also be familiar with local law enforcement agencies and should have
all of their emergency phone numbers up-to-date and readily avail-
able at all times.

FIGURE 11-13 Pedestrian and vehicle access to fluted-column base.
In conjunction with good detection system practice, nuisance

alarms must be minimized to prevent the “cry wolf” problem, in which
alarms are ignored because there are too many false alarms.
If private guards are to be used, many issues must be addressed.
Will they be employed 24/ 7/365? Will they be armed? What level of
authority will they have? Will they be used at all facilities? Do they
regularly train with local law enforcement? Generally, the cost of hir-
ing private guards is prohibitive. An adequate response can be made
by local law enforcement if you have worked to improve relationships
with those agencies.
Operational Security
Operational security (OS) can also provide security, detect and delay
an adversary, and enhance response capabilities.
Detection, Delay, and Response Practices for Operational Security

The categories of detection, delay, and response still apply in OS, but
they are different from the physical security or PPS functions. With
OS, policies, procedures, and training—not physical features—have a
much greater role:
r From an operational perspective, one of the best ways to im-

prove detection is to have a well-trained and aware staff.
Preparing and fully implementing a set of security policies

430 Chapter Eleven
and procedures along with emergency operations and re-

sponse plans is absolutely the most important element of any
risk reduction program.
r If guards are to be provided, they should be on site 24/7/365
and should be well trained.
r Complete background checks should be conducted for all em-
ployees. Focus more thorough and more frequent checks on
employees who have critical access.
r Limit access and key availability to only those employees who
need them.
r Plan in advance what deliveries are expected, and record what
company and driver are expected when. Allow no deviations
from this schedule. Require that all delivery personnel be es-
corted at all times on site. Institute a policy to perform basic
assay tests for all chemicals arriving on site.
r Establish Water Watch neighborhoods throughout your sys-
tem, but begin by focusing on areas near your critical facilities.
Train people in these groups in the basics of water system op-
eration and, especially, what security problems to look for and
whom to call. Whoever is tasked with receiving the calls must
be prepared to handle the situation and initiate a response.
r Have local law enforcement stop and check identification on
any person working in or around water system facilities (such
as hydrants, valve boxes, tanks, booster pumps, and so on).
Do not assume that the worker is an authorized employee of
the water company.
r Consider splitting SCADA system monitoring into two
categories—normal monitoring and monitoring for intrusion
detection. Establish two-person control over SCADA and se-
curity access to critical assets.
r Strengthen existing backflow prevention policy or establish a
new policy. Begin requiring backflow prevention on all con-
nections and change-outs.
r As already stated, probably the most effective way to delay
an adversary is by establishing and implementing security-
and emergency-related policies and procedures. These should
specifically address such areas as key control for all facility
locks. If biometrics or other types of access control are used,
proper policies and procedures still must be followed.
The same type of key control policy should apply to vehicles and
at all other points where access needs to be restricted. All employ-
ees should be subject to strict sign-in/sign-out procedures around all
critical facilities and when using any water system vehicle.

If an employee is alone at a facility, a regular passive and active

call-in procedure should be strictly observed. The passive program
means that a call is regularly placed to the lone operator, either from
a base facility or from local law enforcement. The calls can be placed
randomly or at regular intervals. The active method requires the lone
operator to place the regular or random calls. (Having at least two
operators on duty at all times is the ideal situation, thus eliminating or
minimizing the need for active or passive call-in procedures. However,
if having two people available at all times is impractical, the regular
call-in procedure is the next best thing.)
These types of policies are especially effective at reducing the risk
from an insider adversary, but they are also applicable in defending
against the outsider. They are generally very cost-effective to imple-
ment. The biggest obstacle is to change the ways in which a tank
owner’s staff thinks and functions. Even now, not all operators and
staff members consider security to be an important part of utility op-
erations. Thus, it can be difficult to achieve full acceptance of policies
such as these.
Consequence Mitigation
If detection, delay, and response have failed and a successful attack
has occurred, you are left to mitigate the consequences of that attack.
For water systems, conducting mitigation may be one of the most cost-
effective means of reducing the risk of future attacks and ultimately
improving the level of security for the system. Consequence mitigation
provides benefits after an intentional human act and after natural
disasters. As with some high-level adversaries (e.g., international or
domestic terrorists, organized criminal enterprises, and saboteurs), a
natural disaster cannot be prevented from “attacking” a water system;
in either case you must be able to mitigate the consequences. These
general mitigation techniques are applicable to all sites:
r Provide and maintain an inventory of replacement equip-
ment, focused on the most critical assets as determined from a
Sandia-based RAM-WTM (Sandia 2002) or other vulnerability
assessment.
r Do not store replacement or redundant components in the
same location or structure as the primary item.
r Provide generators or other backup power at all critical facil-
ities. They should be capable of powering the critical assets,
at a minimum.
r For utilities that use gaseous chlorine, store less total chlorine
on site, assuming delivery is on time and reliable.
r If you must store large quantities of gaseous chlorine on site,
store it in two or more geographically distant locations to
lessen the amount available at any single place.

432 Chapter Eleven
r Provide additional gas venting and storage locations for all

gaseous chlorine, and regularly provide emergency training.
r In lieu of gaseous chlorine, switch to safer means of disinfec-
tion such as hypochlorite delivered in bulk form or created
through on-site generation.
r Provide tamperproof, lockable fire hydrants. Schedule regular
training and communication with fire department personnel.
r Work toward establishing system interconnects with neigh-
boring utilities whenever possible. Modeling and testing of
the feasibility of an interconnect must happen before any
physical connection is made.
Bibliography
Brosnan, T. M., ed. 1999. Early Warning Monitoring to Detect Hazardous Events
in Water Supplies. International Life Sciences Institute (ILSI) Risk Science
Institute Workshop Report. Washington, D.C.: ILSI.
Burrows, W. D., J. A. Valcik, and A. Seitzinger. 1997. Natural and Terrorist
Threats to Drinking Water Supplies. US Army Center for Health Promo-
tion and Preventive Medicine. In Proc. 23rd Environmental Symposium and
Exhibition, American Defense Preparedness Association, Arlington, VA.
Danneels, J. J. 2001. Department Manager, Sandia National Laboratories. State-
ment to US House of Representatives Committee on Science, hearing on
H.R. 3178 and the Development of Anti-Terrorism Tools for Water Infras-
tructure, Nov. 14, 2001.
Deininger, R. 2000. The Threat of Chemical and Biological Agents to Public Wa-
ter Supply Systems. Water Pipeline Database, Science Application Interna-
tional Corporation (SAIC), Hazard Assessment and Simulation Division.
McLean, VA.: SAIC.
Gleick, P. H. 2008. Water Conflict Chronology (revised). Oakland, Calif.: Pacific
Institute for Studies in Development, Environment, and Security.
Krouse, M. 2001. Backflow Incident Sparks Improvements. Opflow 27:2.
Public Health Security and Bioterrorism Preparedness and Response Act of
2002. Public Law 107-188, 42 U.S.C. Washington, D.C.: 2002.
Security Systems and Technology Center, Systems Analysis and Development
Department, Sandia National Laboratories. May 2002. Risk Assessment
Methodology for Water (RAM-WSM ). Notebook Volume I. Copyright 2002
Sandia Corporation. Contract DE-AC04-94AL85000. Export Control Clas-
sification Number (ECCN) EAR99.
US Environmental Protection Agency. 2006. Guidelines for the Physical Secu-
rity of Water Utilities. ASCE/AWWA Draft American National Standard
for Trial Use. American Society of Civil Engineers (ASCE), American Water
Works Association (AWWA), and Water Environment Federation (WEF).
Washington, D.C.: USEPA.

CHAPTER 12
Tank Rehabilitation
Gregory R. “Chip” Stein, P.E.
Maintaining water-storage facilities is becoming increasingly impor-

tant because of rising replacement costs and the difficulty of obtaining
rate increases and funding for large-scale construction operations. Al-
though the cost of maintenance is also increasing, these smaller out-
lays can substantially delay or even eliminate the need to replace a
utility’s large capital investment in tanks. This chapter is a guide to
the proactive rehabilitation of existing water tanks as well as a guide
to planning short- and long-range maintenance operations on a new
tank.
When renovation of an existing tank is being considered, an eval-
uation must be made to determine the scope of work to be included.
The costs of renovation versus replacement must be compared and
amortized over the life of a new tank to determine if repair is econom-
ically justifiable.
Developing Specifications
If an evaluation of the tank’s condition, components, and appurte-
nances has determined that repair is required—and if repair is eco-
nomically feasible—it is necessary to generate a set of detailed tech-
nical specifications and bonding requirements.
The scope of work must be determined by evaluating the rec-
ommendations and cost estimates from the inspection report and
comparing these to the availability of funds and to the tank owner’s
long- and short-term plans for the tank. Often, there are multiple po-
tential solutions to an observed deficiency. To determine the repair that
best fits the utility’s needs, these solutions and their associated costs
should be evaluated in terms of the level of risk the utility is willing to
accept.
433
Tank Rehabilitation
434 C h a p t e r Tw e l v e
Standards Referenced
For the specification writer to be effective, he or she must have a
working knowledge of and have access to the following material:
r Applicable American Water Works Association (AWWA)

standards
r National Sanitation Foundation (NSF) standards
r SSPC Painting Manual Volume 2: Systems and Specifications.
Society for Protective Coatings
r Local regulations regarding volatile organic compounds
(VOCs)
r National Fire Protection Association regulations
r All pertinent regulations from the Occupational Safety and
Health Administration (OSHA) and American National Stan-
dards Institute (ANSI)
In addition to these, the specification writer must have a working

knowledge of any state or local regulations that apply to water tank
rehabilitation. The writer should also be familiar with the capabilities
and availability of qualified contractors to perform work of the nature
and magnitude required.
Seismic Design Standards

AWWA D100-05 has changed the way tanks in seismic zones are de-
signed. This latest AWWA D100 revision eliminates seismic “zones”
altogether—instead, the coordinates of the tank are entered into a
computer program, and site-specific seismic design criteria are de-
termined. The change in seismic design standards resulted from a
dramatic change in the way engineers view the risk of a seismic occur-
rence, its potential magnitude, and its effect on a structure. Existing
tanks in high-risk areas should be evaluated to determine whether
they meet the current seismic criteria. It may be prudent to reeval-
uate the seismic criteria and the original tank design criteria when
planning future structural upgrades or modifications.
Owner’s Standard Requirements

Also included in the project specifications should be any special re-
quirements the tank owner or local regulatory agencies might have.
Potential contractors must be alerted to special bonding, wage rate
scales, taxes, and licenses that may be required. Local ordinances
may have stipulations concerning hours of work, acceptable noise lev-
els, requirements for air monitoring, and other construction activities.

Tank Rehabilitation
Ta n k R e h a b i l i t a t i o n 435
Open communication among the specification writer, the tank owner,

and local officials is imperative when preparing specifications.
Environmental/Worker Safety
Lead Regulations
Regulations regarding removal of paint that contains lead and other
regulated heavy metals were changed in the early 1990s. Methods of
compliance and the interpretation and enforcement of these regula-
tions to protect the environment and workers have changed dramat-
ically. Many areas of the United States now enforce a policy of no
emissions into the atmosphere or past the property line. Add to this
the concern for the safety of workers while they are removing the
coatings, and it is obvious why the cost of water tank rehabilitation
has doubled.
The largest problem has been collection of the dust and debris gen-
erated by the removal of the tank’s coating while keeping workers’ ex-
posure levels to heavy metals within the permissible range prescribed
by OSHA in its regulation 1926.62. One solution is to shroud the entire
structure with impervious tarps and conduct open blasting within this
containment system. Dust collectors are then used to negate the pres-
surization effect of the compressed-air abrasive blasting, producing
a negative air pressure in the containment enclosure. Workers must
be adequately equipped with respiratory protection while they are in
this hazardous environment. This relatively expensive method of con-
tainment has been very successful and widely used. Other methods of
surface preparation include the use of vacuum shrouding around blast
nozzles and power tools and the use of ultra-high-pressure (35,000+
psi [241.32 MPa]) water jetting.
The most promising technology currently in use and undergo-
ing further development is the robotic blasting system. This system
includes a self-contained centrifugal blasting apparatus that seals
against the tank surface. The unit is raised and lowered by a winch and
cable. There is no compressed air, so there is no pressure to disperse
the debris that is generated. The abrasive media is typically recyclable,
so the amount of debris is minimized. Additionally, because workers
are outside the blasting assembly, they are not exposed to the concen-
trated dust.
VOC Regulations
Volatile organic compounds, the solvents that traditionally have given
coatings their liquidity and workability, are being heavily regulated
nationwide. To complicate matters, different areas of the country are

Tank Rehabilitation
adapting different acceptable levels of VOCs in industrial coatings.

As the solvents in the coatings are released, the coating dries. To re-
duce VOCs in their products, coating manufacturers are producing
more high-solids coatings and more water-based coatings. These new
coatings will greatly affect coating selection for topcoating operations,
recoating, and construction of new tanks.
Water Circulation
Short-circuiting and stagnation of water in tanks is a concern for tank
owners, who have installed baffle walls and piping systems to force
circulation and water turnover. Baffle walls should be carefully de-
signed to account for their effect on the tank structure. Additionally,
these walls present challenges to future tank maintenance. Piping sys-
tems should be evaluated for use of dissimilar metals, increased cost
of interior repainting, and degree of head range loss required to run
the system.
Description of Repair Work

The accurate and thorough description of needed repair work is one
of the most important roles of the specifying engineer. The engineer’s
goal should be to adequately describe the work so that change orders
are minimized or eliminated and bidders are all on the same page with
respect to what is required. Sometimes it is easy to accurately estimate
and describe the work (e.g, installation of a safe-climbing device on a
ladder). However, in other situations, the repair work is not as easily
estimated (e.g., the amount of pit welding required or the length of
a crack in the concrete). For these situations, the specification should
stipulate the method of repair while allowing bidders to submit a
unit price for it (e.g., the price per foot to repair the concrete crack).
This allows the specifier to minimize the possibility of a change order
and solicit prices for the unknown quantity of work in a competitive
bidding atmosphere.
Surface Preparation
A successful coating application depends largely on the quality of sur-
face preparation. Regardless of the substrate (be it steel, concrete, or
a coated surface), the area to be coated must be clean, relatively free
of contaminants, and properly abraded to receive a coating. Surface
preparation should be specified to conform to the applicable SSPC
standards for cleanliness and the coating manufacturer’s surface pro-
file requirements. Depending on the location of the water tank (in
coastal or heavy-industry areas, for example), specific requirements
regarding the degree of cleanliness and additional testing require-
ments may be required for surface contaminants.

Tank Rehabilitation
Coating Systems
There is no longer any such thing as a standard coating system. Tech-
nology in the painting industry, especially in the water storage tank
industry, is in a period of rapid change. There is no longer a “standard
spec”—not if you want a coating system that will truly protect your
tank.
Gone are the days when conventional paints were applied over
minimally cleaned surfaces by everyday laborers. Now, in a period of
increasingly stringent environmental regulations, highly skilled tech-
nicians apply sophisticated coatings onto surfaces cleaned by ever-
evolving surface preparation methods. We must now “design” a coat-
ing system for each tank, taking into consideration all the specific
conditions that may affect the system’s performance.
Coating System Selection
First, we need to realize that in the past, common industry shortcom-
ings caused specifiers to use improper or inadequate coating systems
for water tanks. Engineering education was lacking with regard to con-
trolling corrosion by using coatings, and so specifiers relied mainly on
coating suppliers for guidance. Thus, a trend developed among speci-
fying engineers of using suppliers’ “canned” specifications rather than
developing a specification and system that fit the exact needs of the
tank owner. Additionally, the welded-steel tank specifiers, designers,
and fabricators frequently failed to recognize the need to incorporate
proper design details that extend the coating life. Just because “it’s
by the specification” does not mean it is the best coating design for
long-term corrosion protection.
The first step in designing a proper coating system is to deter-
mine the owner’s needs and research specific operating conditions by
asking questions about the tank itself.
r In what environment is this tank located?
r What are the constraints of the tank site?
r What is the design of this tank?
r What is the current condition of its coating?
r What are the types of coating failures observed on this tank?
r Why did these coating failures occur?
r What can be done to correct these coating failures?
r Where are the existing corrosion problems on this tank?
r What time of year and for how long can the tank be taken out
of service for painting?
r What is the level of community acceptance of this tank?
r What are the owner’s short- and long-term plans for this tank?

Tank Rehabilitation
After these questions have been answered, review possible coating

alternatives. Weigh the advantages and disadvantages of each system
so that the best system can be provided for a specific tank and owner.
Research of coating alternatives should begin with a thorough review
of applicable standards. This review should include AWWA D102
Standard for Painting Steel Water-Storage Tanks and the applicable
standards of the SSPC. The coatings must comply with and be tested
in accordance with the requirements of the NSF for coatings in contact
with drinking water. Finally, the coating manufacturer’s performance
test data and real-life case histories should be investigated, as well as
any independent laboratory testing or documented service history.
Coating System Objectives

The specifier should set objectives that will be compatible with the
needs of the tank painter, engineers, and, most importantly, the tank
owner, as follows:
r Reduce initial cost.
r Provide the optimum coating life for the tank environment.
r Minimize release of VOCs or other harmful materials into the
atmosphere.
r Provide a coating system that will be easily maintained by
touch-up and maintenance topcoating, thus minimizing the
need for abrasive blasting to bare steel until the tank has been
topcoated several times.
r Eliminate unsealed or uncoated interfaces of steel surfaces.
r Provide excellent resistance to abrasion and be self-healing
when subjected to minor abrasions and scratching.
r Provide excellent resistance to ultraviolet (UV) light, moisture,
oil, soil, and chemicals.
r Provide a recoat window varying from a few days to as long
as years for new tank projects.
r Meet all NSF standards and US Environmental Protection
Agency (USEPA) regulations.
r Describe the system in generic or performance terms that do
not rule out qualified coating manufacturers yet that uphold
the standards of quality and performance necessary to pro-
vide the tank owner with the best possible system.
Interior Coating Systems

Interior coating systems should offer long life; ease of application;
abrasion resistance; and (in the case of open-top tanks) resistance to
UV light, oil, dirt, chemicals (chlorine), and other contaminants.

Tank Rehabilitation
Prior to the effective date of NSF Standard 61, Drinking Water

System Components–Health Effects addressing direct and indirect
water additives, there were many types of coatings:
r Vinyls
r Zinc dust/zinc oxide
r Chlorinated rubber
r Bituminous
r Coal tar
r Red lead
r Wax grease
r Phenolic aluminum
In the past, these coatings have worked with varying degrees of
success. However, because strict environmental guidelines for toxins,
heavy metals, VOC emissions, and other health threats have since
been established, significantly fewer of these coatings are likely to
meet new criteria set forth for use on tank interiors.
The success of the two-component catalyzed epoxy appears to
make it the frontrunner at this time. Epoxies can be and have been
formulated with very high solids (low VOCs) and with many chem-
ical varieties available. Two-component catalyzed epoxy is a highly
versatile tank lining and coating.
Another product for consideration is the solventless 100 percent
solids polyurethanes. These products are not mixed, as conventional
epoxies and urethanes are; they are sprayed with a dual pump ar-
rangement that mixes the polyurethane at the spray gun tip. The ad-
vantage is less waste and a coating that cures for immersion within 48
hours. Manufacturers claim that these coatings offer from 20 percent to
38 percent longer life than epoxies, but in this author’s opinion, exten-
sive field testing and evaluation are required to substantiate this claim.
One last coating for consideration for water immersion is not really
a coating at all. Spray metalizing using zinc, aluminum, or a combina-
tion of both has been used successfully for many years. Only recently
has technology made this a viable option when considering costs.
Zinc coatings can also be used for direct application to the water
tank. While there is a lack of extensive service history in our industry,
inorganic zinc coatings could be used on surfaces intended for immer-
sion in potable water if NSF certified. However, these coatings should
not be topcoated unless they are fully cured and hydrolyzed.
Exterior Coating Systems

Like the interior coating systems, the exterior systems should also offer
long life; abrasion resistance; ease of recoating; ease of application;
and resistance to ultraviolet light, oil, dirt, salts, chemicals, and other

Tank Rehabilitation
contaminants. Because of potential community resistance, coatings

requiring minimal to no abrasive blasting are attractive alternatives
for future repainting needs.
With ever-tightening regulations to control VOC emissions, the
alternatives for exterior systems have been reduced significantly.
Solvent-based aluminums, acrylics, vinyls, and chlorinated-rubber
paints do not meet most VOC restrictions and have lately been used
very little. Although we are never quite sure what direction the regula-
tory bodies will take when it comes to environmental issues, the trend
appears to be to reduce emissions even further, resulting in lowering
the VOCs of all coatings. Generic coating systems that currently meet
most areas’ restrictions are high-solids alkyds, water-based acrylics,
epoxies, polyurethanes, and inorganic and organic zinc-rich coatings.
More stringent regulations will probably eliminate alkyds and all but
the high-solids epoxies and polyurethanes in the future.
The most widely used exterior system today is the epoxy–urethane
system, sometimes with a zinc-rich primer and sometimes with an
additional clear urethane topcoat. Water-based acrylics are becoming
more popular—especially for overcoating, due to the minimal stress
they have on existing coatings during cure and for congested sites
where their “dry fall” characteristics are important. (Dry fall coating
overspray releases all of its solvents as it falls through the air. The
overspray is dry when it contacts the surface below.)
A more recent technology is the use of solvent or new water-based
fluorourethanes. Previously, these coatings were only available as a
baked-coil coating material from which more than 25 years of color
and gloss retention was normally expected. Time will tell if the newer
air-dried versions will perform as well. Another category of exterior
coatings comprises inorganic-based siloxane hybrids that claim to
weather as well as, if not better than, conventional polyurethanes,
but apply like high-build epoxies.
Finally, there are varieties of coatings of several generic types
that are formulated so that they can be applied over very minimally
cleaned surfaces. They can be applied over rust, rust stain, old chalked
and cracked paint, and other existing coating defects.
Environmental issues, especially for lead paint removal and abra-
sive dust generation, have caused specifiers to strongly consider alter-
natives to conventional cleaning methods. To avoid open-air blasting,
the coating industry has developed various methods to achieve the
same degree of cleaning. Among these methods are containment of the
structure with tarp material and the use of dust collectors to produce
negative air pressure inside this containment.
Risks and Benefits of Repainting

After considering coating systems and environmental issues, the spec-
ifier and the owner must decide the risks and benefits for the var-
ious options of repainting. Table 12-1 is a way of reviewing this

Risks
Early Coating Environmental Worker Adverse
Alternative Benefit Cost Failure Release Contamination Publicity
Do nothing None/negative 0 N/A N/A N/A High
Spot repair/spot repaint Very low Low High Low Low High
Zone painting Low Low Moderate Low Low Moderate
Spot repair/overcoat Moderate Moderate High Low Low Low
Complete removal/recoat High High Low High High High
Tank Rehabilitation
Replace tank High High Low Low Low Low
Note: N/A = not applicable.
TABLE 12-1 Comparison of Relative Risks, Benefits, and Costs

441
Tank Rehabilitation
information. Costs may be inserted using an engineering estimate or

even from the actual bids.
Another consideration is when the tank can be drained and avail-
able for painting. If this cannot be done in normal warm-weather
conditions, the coating system design must specify either coatings
that can be applied and cured at colder temperatures or coatings that
are applied while using heaters and dehumidification to control the
environment. If the schedule is tight and downtime must be limited,
multiple shifts using environmental control might overcome schedule
constraints.
Also, the coating system must meet the owner’s aesthetic require-
ments. Color availability and color and gloss retention are important
aspects of aesthetic appeal. To the public, aesthetic appeal is often the
most important aspect of the coating system.
One final consideration when designing coating systems—the
specifier must keep in mind the knowledge and abilities of the poten-
tial low bidder. The specifier must realize that if no independent field
inspection of the work will occur, there is greater risk in specifying a
coating system that is very sensitive to the quality of workmanship—
as are nearly all of the new long-life systems. So, if little or no inspec-
tion is to be performed or if prequalification of the bidders is impos-
sible, it is not recommended to specify coatings that are difficult to
properly apply.
Likewise, the specifier must be prepared to address problems that
will undoubtedly occur in the field. This may even include demon-
strating to the contractor how the specified coatings are to be mixed,
applied, and cured. As new systems are developed, adequate prod-
uct knowledge and practical field experience with these new coatings
are lacking. Beware of being the guinea pig for new coating systems.
Know what you are specifying! If you lack previous experience with
the new product, observe application procedures and gain informa-
tion from knowledgeable colleagues. It is important that specifiers
continually learn about new coatings, equipment, procedures, regu-
lations, and other important aspects of the coating industry through
pertinent professional organizations and societies so they can provide
the best service possible.
Overseeing Painting and Maintenance

Many water tank owners seek autonomous verification that recoating
and repairs are being performed in accordance with project specifi-
cations and generally accepted industry practice. By having a quali-
fied and experienced professional be the on-site project representative
while the work is being performed, the owner has independent assur-
ance that the coatings will remain in good condition for their intended
service life.

Tank Rehabilitation
Role and Qualifications of the Project Representative

Long before any abrasive blasting and painting are performed on the
water tank, the tank owner needs to determine what the role of the
project representative will be and what qualifications that person must
possess. Many water utilities require registration and licensing from
one or more of the various industry associations (e.g., SSPC, NACE).
However, possession of a specific industry’s license should not take
the place of numerous years of experience in administering water tank
rehabilitation projects.
The role of the project representative must be determined and
agreed on before the project begins. It is critical, then, that the de-
scription of the project representative’s duties, including the limita-
tion of authority and responsibility, be communicated and made clear
to all parties, including the painting contractor. For example, when
an independent project representative is on-site, his or her responsi-
bility (or lack thereof) for the safe work practices of the contractor’s
personnel should be understood. Generally this is the case, since the
project representative does not have direct control or supervision of
the means, methods, techniques, sequences, or procedures of the con-
tractor’s personnel, nor would the project representative generally be
asked to issue direction regarding or assume control over the contrac-
tor’s compliance with environmental regulations.
In most cases, the role of the project representative is to conduct
on-site observation of the work in process and help the owner deter-
mine whether the work is in compliance with the specifications and
with generally accepted industry practice. The project representative
should also be expected to document and report to the tank owner any
work that appears unsatisfactory or defective and advise the owner
when additional testing appears necessary.
The project representative should document his or her observa-
tions daily on an observation form. Topics might include number of
contractor’s personnel on-site, surface profile measurements, paint
batch numbers, area(s) of tank worked on, and ambient weather con-
ditions. The written narrative should be supplemented with photo-
graphic documentation as determined necessary by the project rep-
resentative. It is critical that this documentation be distributed to the
owner and to the contractor’s foreman on a regular (daily) basis so
that all parties are on the same page.
Role of the Water Tank Owner

As previously discussed, the water tank owner is responsible for
selecting the project representative and communicating the author-
ity and limitations of that person’s duties to all parties. Next, the
owner should designate someone else to act as the utility’s represen-
tative, a person who has the authority to transmit instructions, receive

Tank Rehabilitation
information, and interpret and define the owner’s decisions. The util-
ity’s representative should also be responsible for arranging for access
onto both public and private properties as necessary and for reviewing
and providing input on all documentation submitted by the contractor
and the engineer.
Observation of the Work in Process

As mentioned previously, many utility owners view independent
evaluation of the coating application and repairs as an essential part
of a successful tank rehabilitation project. The capability of a coating
to achieve its anticipated service life is directly related to the quality
of workmanship during application. Verification of the workmanship
assures the owner that the money spent on tank maintenance will be
maximized.
As applicable, the following should be verified daily by the on-site
project representative:
r Temperature of steel
r Weather conditions (temperature, wind velocity and direc-
tion, relative humidity, and dew point)
r Paint batch numbers used on the day of the observation
r Location of work performed
r Quality of work being performed and compliance with the
project documents
r Wet and dry film thickness readings
r Calibration record of dry mil thickness gauge
r Measure of the paint cure
r Number of workers on the job
r Equipment on the job
r Recommendations made
r Estimated completion date
r Photographs of significant details
r Other pertinent data as required or requested
It is necessary that the on-site project representative, in addi-

tion to having the previously discussed qualifications, be trained and
qualified to competently use the equipment necessary to verify the
quality of the work. The project representative generally uses the fol-
lowing equipment and resources to observe the work when required
to do so:
r SSPC-Vis 1 visual blasting standards
r NACE Visual Standard TM-01-70/75 (available)

Tank Rehabilitation
r AWWA standards D100-96 and D102-97
r Testex Press-O-Film Profile Measurement System
r Surface contamination detection device (soluble salts)
r Wet film thickness gauge
r Dry film thickness gauge
r Certified thickness calibration standards
r Steel temperature gauges
r Sling psychrometer and psychrometric tables
r Wet sponge holiday detector (low voltage)
r Tooke Gage (if required)
r Adhesive force measurement device (if required)
Contract Document and Specification Options

It is important that the project specifications include specific tank up-
grades as well as the tank repainting design. Repair specifications can
be worded according to the precise methods to be used or the desired
end result. The specification writer should review the strengths and
weaknesses of each approach.
Precise Methods of Repair

For a specification that outlines precise methods of repair, the spec-
ification writer (and therefore, ultimately, the tank owner) exercises
a good deal of control over the contractor’s activities. This type of
specification can result in fewer bidders who are willing to modify
their standard procedures to comply with the specifications, and bid
prices may be higher. Strictly adhered to, this method of specification
preparation is a bit of overkill and places more liability on the specifier.
End Result Only

This type of specification defines the repairs to be made but leaves it to
the contractor to determine how to accomplish the repairs. The result
is more bidders and possibly lower costs. It allows the contractor to use
his or her standard methods of repair and can lead to the development
of innovative procedures. However, the specifier and the owner have
little or no control over methods used.
A prudent specifying engineer uses the best of each method and
writes a repair specification that results in the best bidders offering
the most competitive bids that result in a long-lasting, high-quality
repair. Quantity does not necessarily mean quality.
The contract documents used in a tank rehabilitation project gen-
erally spell out such necessary requirements as insurance limits,

Tank Rehabilitation
bonding, length of contract time, and bidding information. Sometimes

a utility has a complete set of contract documents; sometimes it just
provides certain documents and/or input on minimum requirement
levels. If the utility lacks a complete set of contract documents, there
are organizations that provide boilerplate documents (e.g., the En-
gineering Joint Contract Documents Committee) that can easily be
modified for tank rehabilitation. Regardless of how the contract doc-
uments are assembled, the utility must provide precisely exact input
in numerous areas, including minimum insurance level requirements,
liquidated damages amounts, on-site availability of water and elec-
tricity, and bid opening dates.
Contract Administration
The specifier and the tank owner should collaborate to administer the
project to make sure that the owner’s needs are being satisfied. The
owner may prefer that some activities be performed on-site during a
rehabilitation project; some activities are best overseen by either the
specifying engineer or the on-site project representative. The project
engineer should verify compliance with the project specifications and
contract documents to ensure that both the letter and the intent of
the documents are being followed. The go-ahead for work to pro-
ceed should not be given until all submittals have been reviewed and
accepted.
After all of the submittals have been reviewed and accepted, nu-
merous other administration activities need to be done, including
these:
r Pre-job conference attendance

r Consultation on adequacy of and compliance with the project
specifications
r Specification interpretation
r Attendance at the preconstruction meeting and all subsequent
meetings, and provision of meeting notes
r Review of all contractors’ submittals and shop drawings
r Review of construction schedule
r Review and approval of materials
r Preparation for negotiations of change orders and assistance
with supplemental agreements
r Review and approval of payment requests
r Dispute settlement
r Public relations

Tank Rehabilitation
First Anniversary Evaluation

A first anniversary evaluation, as recommended in AWWA D102,
should be called for in the project specifications and should be sched-
uled prior to the end of the one-year bonded guarantee. The purpose
of this evaluation is to identify and repair defective work before the
bonding period ends. The water tank owner should be responsible
for making the tank available and coordinating the date of the evalua-
tion with the contractor and the project representative. The contractor
should be required to complete the tank washout the day before the
evaluation and should also be responsible for meeting the following
requirements:
r Have an experienced foreman present.

r Be prepared to perform minor touch-up work.
r Bring all rigging necessary to performance of the touch-up
work.
r Bring at least 1 gal (3.79 L) each of the exterior primer, inter-
mediate coating, and finish coating.
r Bring at least 1 gal (3.79 L) each of an interior coating that can
be placed in immersion service immediately for minor spot
repairs.
r Bring Scotch-BriteTM abrasive disks with power tools and
sandpaper to clean the steel surface.
r Supply equipment with which to apply coating repairs.
r Supply equipment with which to wash out the tank and chlo-
rine to disinfect it following the evaluation and any required
touch-up work.
The project representative should prepare and submit to the water

tank owner a brief report with color photographs of the conditions
found during the first anniversary evaluation and of the touch-up
work.

Tank Rehabilitation


Steel Water Storage Tanks Design Construction Maintenance and Repair PDF

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Steel Water Storage Tanks Design Construction Maintenance and Repair PDF

Transféré par

Droits d'auteur :

Formats disponibles

Source: Steel Water Storage Tanks: Design, Construction, Maintenance, and Repair

Richard A. Horn, P.E.

The rapid development and expansion of public water supply systems

FIGURE 1-1 Geodesic dome on bolted-steel reservoir.

commonly found on wastewater plant storage tanks as well. Figure

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

FIGURE 1-2 Witch’s

FIGURE 1-3 Legged tank with radial arm design.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

FIGURE 1-4 Welded-steel reservoir. (Photo: Gay Porter DeNileon, AWWA)

Approved ladder, cage,

Crushed rock or gravel Concrete foundation

FIGURE 1-5 Cross-sectional view of welded-steel reservoir. (Source: AWWA

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

FIGURE 1-6 Bolted-steel reservoir, glass fused to steel.

Approved ladder, cage, and platform

Roof manway Gravity

24-in. (0.6-m) round Grade

FIGURE 1-7 Cross-sectional view of bolted-steel reservoir. (Source: AWWA

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Capacity Range of Sizes Available

Source: AWWA Manual M42, Steel Water-Storage Tanks.

TABLE 1-1 Capacities and Sizes of Typical Welded-Steel Water-Storage

Any use is subject to the Terms of Use as given at the website.

Copyright © 2010 The McGraw-Hill Companies. All rights reserved.

TABLE 1-2 Capacities of Glass-Coated, Bolted-Steel Reservoirs and Standpipes

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Roof Designs for Reservoirs and Standpipes

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Capacity level Weir box

Painter’s trolley rail

Approved ladder, cage,

FIGURE 1-9 Cross-sectional view of typical welded-steel standpipe.

coatings should be applied to all surfaces deemed necessary from a

Column- and Rafter-Supported Cone Roofs

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

FIGURE 1-10 Bolted-steel standpipe.

Approved ladder, cage, and platform

Floor sloped toward

24-in. (0.6-m) round Grade

FIGURE 1-11 Cross-sectional view of bolted-steel standpipe. (Source: AWWA

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Capacity Range of Sizes Available

Source: AWWA Manual M42, Steel Water-Storage Tanks.

TABLE 1-3 Capacities and Sizes of Typical Welded-Steel Standpipes

Self-Supporting Dome Roof and Umbrella Roof

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

12 in. (0.3 m) Vent 3 ⁄16-in. (4.7-mm)

Girders are Capacity

should be avoided if possible, because they may create corrosion

Knuckle plate 3/16 in. (4.7-mm)

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

3 ⁄16-in. (4.7-mm) minimum thickness

FIGURE 1-14 Self-supporting dome roof or umbrella roof. (Source: AWWA

of triangulated space truss (geodesic) panels. The dead weight of these

Self-Supporting Ellipsoidal Roof

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

1⁄4-in. (6.4-mm) Area above Vent

FIGURE 1-15 Self-supporting ellipsoidal roof. (Source: AWWA Manual M42,

Multiple-Column Elevated Tanks

Downloaded from Digital Engineering Library @ McGraw-Hill (www.accessengineeringlibrary.com)

Medium-Capacity Elevated Tanks