Issue 11, 2013 Journal of Civil Engineering and Architecture

Journal of Civil Engineering
and Architecture
Volume 7, Number 11, November 2013 (Serial Number 72)
David Publishing Company
www.davidpublishing.com
Publishing David

Publication Information:
Journal of Civil Engineering and Architecture is published monthly in hard copy (ISSN 1934-7359) and online (ISSN
1934-7367) by David Publishing Company located at 3592 Rosemead Blvd #220, Rosemead, CA 91770, USA.

Aims and Scope:
Journal of Civil Engineering and Architecture, a monthly professional academic journal, covers all sorts of researches
on structure engineering, geotechnical engineering, underground engineering, engineering management, etc. as well
as other issues.

Editorial Board Members:
Dr. Tamer A. El Maaddawy (Canada), Prof. San-Shyan Lin (China Taiwan), Dr. Songbai Cai (China), Dr. Xiaoyan Lei
(China), Prof. Vladimir Patrcevic (Croatia), Dr. Sherif Ahmed Ali Sheta (Egypt), Prof. Nasamat Abdel Kader (Egypt),
Prof. Mohamed Al-Gharieb Sakr (Egypt), Prof. Olga Popovic Larsen (Denmark), Prof. George C. Manos (Greece), Dr.
Konstantinos Giannakos (Greece), Pakwai Chan (Hong Kong), Dr. K. Muthukkumaran (India), Chiara Vernizzi (Italy),
Prof. Michele Maugeri (Italy), Dr. Giovanna Vessia (Italy), Prof. Valentina Zileska-Pancovska (Macedonia), Dr. J.
Jayaprakash (Malaysia), Mr. Fathollah Sajedi (Malaysia), Prof. Nathaniel Anny Aniekwu (Nigeria), Dr. Marta Sowik
(Poland), Dr. Rafael Aguilar (Portugal), Dr. Moataz A. S. Badawi (Saudi Arabia), Prof. David Chua Kim Huat
(Singapore), Dr. Vail Karakale (Waiel Mowrtage) (Turky), Dr. A.Senem Deviren (Turkey), Dr. Yasemin Afacan
(Turkey), Dr. Ming An (UK), Prof. Ahmed Elseragy (UK), Prof. Jamal Khatib (UK), Dr. John Kinuthia (UK), Dr.
Johnnie Ben-Edigbe (UK), Dr. Yail Jimmy Kim (USA), Dr. Muang Seniwongse (USA), Prof. Xiaoduan Sun (USA), Dr.
Zihan Yan (USA), Dr. Tadeh Zirakian (USA).

Manuscripts can be submitted via Web Submission, or E-mail to civil@davidpublishing.com or
civil@davidpublishing.org. Submission guidelines and Web Submission system are available at
http://www.davidpublishing.com, www.davidpublishing.org.

Editorial Office:
3592 Rosemead Blvd #220, Rosemead, CA 91770, USA
Tel: 1-323-984-7526, 323-410-1082 Fax: 1-323-984-7374, 323-908-0457
E-mail: civil@davidpublishing.com; civil@davidpublishing.org; shelly@davidpublishing.com.

Copyright2013 by David Publishing Company and individual contributors. All rights reserved. David Publishing
Company holds the exclusive copyright of all the contents of this journal. In accordance with the international
convention, no part of this journal may be reproduced or transmitted by any media or publishing organs (including
various websites) without the written permission of the copyright holder. Otherwise, any conduct would be
considered as the violation of the copyright. The contents of this journal are available for any citation. However, all
the citations should be clearly indicated with the title of this journal, serial number and the name of the author.

Abstracted / Indexed in:
Database of EBSCO, Massachusetts, USA
Chinese Database of CEPS, Airiti Inc. & OCLC
Cambridge Science Abstracts (CSA)
Ulrichs Periodicals Directory
Summon Serials Solutions
Google Scholar
ProQuest

Subscription Information:
$520/year (print) $360/year (online) $680/year (print and online)

3592 Rosemead Blvd #220, Rosemead, CA 91770, USA
Tel: 1-323-984-7526, 323-410-1082 Fax: 1-323-984-7374, 323-908-0457
E-mail: civil@davidpublishing.com; civil@davidpublishing.org; shelly@davidpublishing.com.

www.davidpublishing.com
DAVID PUBLISHING
D

Jour nal of Ci vi l Engi neeri ng
and Archi t ect ure

Volume 7, Number 11, November 2013 (Serial Number 72)

Contents
Engineering Structures
1329 Bending Stresses of Steel Web Tapered Tee Section Cantilevers
Boksun Kim, Andrew Oliver and Joshua Vyse
1343 Unanticipated Bond Failure over Supporting Band Beams in Grouted Post-Tensioned Slab
Tendons with Little or No Initial Prestress
Raymond Ian Gilbert
1353 A Computer Modeling of the China Central Television Headquarters in Beijing
Valentina Forcella
1369 Infilled Reinforced Concrete Beams for Sustainable Construction
Patel Rakesh, Dubey Sheo Kumar and Pathak Krishna Kant
Transport Design
1375 Flexible Pavement Design Simulation Using Mechanistic-Empirical Pavement Design Guide
Daba S. Gedafa, Mustaque Hossain, Stefan Romanoschi and Andrew J. Gisi
1385 Critical Factors in Financing and Management of Light Rail Transit Systems: Delphi
Methodology Approach
Julin Sastre Gonzlez and Clara Zamorano Martn
Geotechnical and Geophysical Techniques
1398 Determination of the Scale of Damage in the Wastewater Tunnels Due to Earthquake: A Case
Study
Turgay ogun, aatay Turgut, Baris Sayin and Ahmet ogun
1405 Probabilistic Seismic Hazard Assessment for Bridgetown-Barbados, Employing Subduction
Interface Characteristic Earthquakes
Walter Salazar, Lyndon Brown and Garth Mannette

1423 Use of Geophysical Techniques for the Localization of the Restricting Zones of Permeability in the
Bottom of Basin: Forecasting the Zones of Clogging
Ibrahim Alimi Ichola, Salima Bouchemella and Claudio Fernando Mahler
Urban and Housing Design
1433 Sustainable Methods and Techniques in Old and New Public Constructions
Mauricio Noya, Ana Lucia Torres Seroa da Motta and Wagner Abreu
1445 Social Participation in the Context of the Urban Public Space Renewal: The Case of Lagarteiro
Neighborhood in Oporto
Paulo Tormenta Pinto
1458 Analyzing Methods of Using Sustainable Development Indicators
Fani Samara, Olga Christopoulou and Athanasios Sfougaris
1464 Economy of RHA (Rice Husk Ash) in Concrete for Low-Cost Housing Delivery in Nigeria
Olonade Kolawole Adisa

Nov. 2013, Volume 7, No. 11 (Serial No. 72), pp. 1329-1342
J ournal of Civil Engineering and Architecture, ISSN 1934-7359, USA

Bending Stresses of Steel Web Tapered Tee Section
Cantilevers
Boksun Kim
1
, Andrew Oliver
2
and J oshua Vyse
3
1. School of Marine Science and Engineering, Plymouth University, Plymouth PL4 8AA, UK
2. Structures Team, Gary Gabriel Associates, Taunton TA1 2PX, UK
3. Design and Engineering, Atkins Ltd., Colchester CO1 1TG, UK

Abstract: Although commonly used, no design method is available for steel web tapered tee section cantilevers. This paper
investigates the bending stresses of such beams. Relationships between the maximum compressive stress and the degree of taper were
investigated. An analytical model is presented to determine the location of the maximum stress when subjected to a uniformly
distributed load or a point load at the free end and was validated using finite element analysis and physical tests. It was found that the
maximum stress always occurs at the support when subjected to a uniformly distributed load. When subjected to a point load at the free
end and the degree of taper is up to seven, it was found that Millers equation could be used to determine the location of the maximum
stress. However, it is shown that when the degree of taper is greater than seven, Millers equation does not accurately predict the
location and the analytical model should be used. It was also found that the location of the maximum stress was solely dependent on the
degree of taper, while a geometric ratio, was required to determine the magnitude of the maximum stress. A simple method that
predicts the magnitude of the maximum stress is proposed. The average error in the prediction of the magnitude of the maximum stress
is found to be less than 1.0%.

Key words: Web tapered tee sections, tapered cantilevers, bending stress patterns, maximum stress shift, finite element analysis.

1. Introduction

Steel web tapered cantilevers are used because of
their aesthetic features and lightweight. They are
structurally efficient because the web is tapered along
the beam to closely match the variation of the bending
moment of the beam. The depth of the beam is largest
at the fixed support, where its bending moment is
greatest and gradually decreases towards the free end.
Although steel web tapered tee cantilevers are
commonly used, neither Eurocode 3 nor BS 5950
provides a design method for such beams. In BS 5950
[1], some design rules exist for untapered tees and
tapered beams whose cross-sections are other than tee
sections. The majority of the existing literature deals
with the lateral torsional buckling capacities of tapered
I-beams [2-6] or tapered channels [3]. Tapered tee

Corresponding author: Boksun Kim, Ph.D., research fields:
steel construction, steel buckling and finite element analysis.
E-mail: boksun.kim@plymouth.ac.uk.
cantilevers were studied by Fischer and Smida [7] and
Yuan et al. [8]. However, their study dealt with the
instabilities of the beams.
The design of tapered rectangular levers with a point
load at the tip was studied by Miller [9]. Miller pointed
out that the location of the maximum bending stress of
the tapered levers moved away from the support as the
degree of taper was more than two and determined the
location of the maximum stress as follows:

1
'
x' , 2
'
for

h
h
L
h
h
(1)

where, h and h are the depths of the beam at the
support and the tip, respectively, and x is the distance
from the tip and L is the cantilever length.
Eq. (1) can be rearranged using the distance from the
support, x (x =L x) as follows:

1
'
2
'
, 2
'
for

h
h
h
h
L
x
h
h

(2)

DAVID PUBLISHING
D
Bending Stresses of Steel Web Tapered Tee Section Cantilevers

1330
However, Millers study was limited to rectangular
cross-sections and the stress patterns for other
cross-sections may differ. In a study with tapered
I-beams [2], the maximum stress, hence the plastic
hinge, of the tapered I-beam cantilevers occurred away
from the support when the degree of taper was high.
However, the location of the maximum stress was not
quantified.
This paper investigates the bending stresses of steel
web tapered tee section cantilevers. Cantilevers with
various degree of taper were analyzed. The stress
pattern along the beam is identified and the location
and magnitude of the maximum bending stress are
discussed and simple methods are suggested to predict
the latter two.
2. Web Tapered Tee Cantilevers
In the absence of design guidance, a conventional,
yet onerous, analysis may be used for the design of web
tapered tee cantilevers. The resistance and stiffness of
such beams may then be checked along the tapered
beam using the section properties at each point [10].
The bending stress of a tapered beam may be calculated
using the following well known equation:
I
My
(3)
where, is the bending stress on any fibre at a distance
y from the neutral axis, M is the bending moment and I
is the second moment of area about the neutral axis.
2.1 Bending Stresses
As the cross-section of a tee is mono-symmetric and
the neutral axis lies towards the flange, the
compressive stress is more critical than the tensile
stress. Fig. 1 shows the theoretical (using Eq. (3))
compressive stress along the beam relative to that at the
support. A 152 kg/m 229 kg/m 34 kg/m tee section
cut from a 457 152 67 UKB was used. The
dimensions of the cantilever beams are given in Table 1.
Five cases distinguished by the degree of taper (h/h)
were analyzed. The degree of taper, h/h is defined by
the ratio of the depth at the support (h) to that at the tip
(h), as shown in Fig. 2.
When subjected to a uniformly distributed load the
maximum compressive stress always occurs at the
support regardless of the degree of taper, as shown in
Fig. 1a. While the location of the maximum stress
changes, depending on the degree of taper when
subjected to a point load at the tip, as shown in Fig. 1b.
For a cantilever with a tip point load the stresses in
the untapered section decrease linearly towards the tip,
with the maximum at the support, due to the constant
value of the section modulus. On the other hand, the
stresses in the tapered sections do not decrease linearly
towards the tip because of the different values of the
section modulus along the beam due to the taper. It
should be noted that as the degree of taper increases,
the location of the maximum compressive stress moves
towards the tip, as shown in Fig. 1b. This indicates that
the maximum stress does not always occur at the
support in tapered tees, even if the maximum bending
moment occurs at the support, and therefore the plastic
hinge is likely to be formed away from the support.
This point was also noted in tapered I-beams [2] and
tapered levers with rectangular cross-sections [9]. The
maximum compressive stresses of the 3:1, 4:1 and 5:1
tapered beams are 1.1, 1.3 and 1.5 times more than that
of the untapered one, even though the maximum
bending moment is the same for all cases. This should
be considered in the design of tapered tee cantilevers.
2.2 The Location of the Theoretical Maximum Bending
Stress
It would be beneficial for designers to determine the
location and magnitude of the maximum compressive
stress of tapered cantilevers. The latter is discussed in
the following section.
Consider a web tapered tee-section cantilever
subject to a uniformly distributed load and a
concentrated load at its free end, as shown in Fig. 3. Let
x be the longitudinal axis of the beam, y and z be the
cross-sectional axes parallel to the web and flange,
respectively. For convenience, the origin of coordinates
was chosen to be the centroid of the section. Due to the

1331

(a) Under a uniformly distributed load

(b) Under a point load at the tip
Fig. 1 The predicted compressive stresses along the beam.

Table 1 Dimensions of the cantilevers.
Beam designation h/h h (mm) h (mm) Flange width (mm) Flange thickness (mm) Web thickness (mm)
1A and 1B 1.0 228.9 228.9 153.8 15 9
2A and 2B 2.0 228.9 114.6 153.8 15 9
3A and 3B 3.0 228.9 76.0 153.8 15 9
4 4.0 228.9 57.2 153.8 15 9
5 5.0 228.9 45.8 153.8 15 9
0.0
0.4
0.8
1.2
1.6
0 0.2 0.4 0.6 0.8 1
x
/
s
u
p
p
o
r
t
Support Tip
Untapered
2:1Taper
3:1Taper
4:1Taper
5:1Taper

1332

Fig. 2 Tee section cantilever showing strain gauges.

tapering of the web, the section properties of the beam
are a function of the coordinate x and can be expressed
as follows:
) tan (
2
tan
) tan (
2
2
o
o
o
x b t t b
x b
t x b t
t b
y
wo w f f
wo
f wo w
f f
+
|
.
|
\
|
+ +
=
(4)
=
2
12
) tan (
) tan (
wo
wo w z
x b
x b t I
o
o

+
(
(
(
|
.
|
\
|
+
2
2
tan
f
wo
y t
x b o

(
(
|
|
.
|
\
|
+
2
2
2 12
f f
f f
t
y
t
t b

(5)
where, y

is the distance from the top of the section to
the neutral axis, I
z
is the second moment of the
cross-sectional area about the z-axes, b
f
is the flange
width, t
f
is the flange thickness, b
wo
is the web depth at
the support (x = 0), t
w
is the web thickness and is the
tapering angle.
For a beam subject to a uniformly distributed load
and a concentrated load at its free end M
z
can be
expressed as follows:
2
) (
2
) ( ) ( x l
q
x l P x M
z
+ = (6)
where, P and q are the concentrated and uniformly
distributed loads, respectively, x is the distance from
the fixed support and l is the cantilever length, as
shown in Fig. 3.
The compressive stress,
zC
of the beam then can be
determined as follows:
) (
) (
) (
) ( y b t
x I
x M
x
w f
z
z
zC
+ = o
(7)
Differentiating Eq. (7) with respect x and equating it
to zero yields a value for the location of the maximum
stress.
) (
) (
) (
) (
2
c
c
+
=
c
c
x I
x M
x
M
x I
y b t
x
z
z z
z
w f
zC
o

0
) (
) (
) ( =
c
c

c
c
+
x
y
x I
x M
x
I
y b t
z
z z
w f
(8)
However, it is difficult to drive an exact formula in
Eq. (8). It only gives an approximation. The exact
formula for predicting the location of the maximum
stress is only valid for rectangular cross-section. The
locations of the maximum stress using Eq. (8) are
plotted in Fig. 4 for various degrees of taper and
compared with those using Millers equation (Eq. (2))
and the theoretical stress equation, i.e., Eq. (3).
Fig. 4 shows that when subjected to a point load at
the tip, the maximum compressive stress will always be
at the support for a tapered tee section cantilever whose
tapering ratio (h/h) is less than or equal to two. This
agrees with Millers study [9] even though the two
cross-sections are different. Millers equation can be
used to determine the location of the maximum stress
of a tapered tee-section cantilever whose tapering ratio
is greater than two and up to seven. When the degree of
taper is greater than seven, Millers equation diverges
from the rest of the two data sets, as shown in Fig. 4.
When the degree of taper is greater than seven, Eq. (8)
should be used to determine the location of the
maximum stress.
2.3 The Magnitude of the Theoretical Maximum
Bending Stress
Knowledge of the location of the theoretical
maximum bending stress is very useful. However, the
Cantilever support
Cantilever tip
h
h
Strain gauge

1333

Fig. 3 A web tapered tee-section cantilever beam subject to a uniformly distributed load and a concentrated load at its free
end.

Fig. 4 Location of the maximum compressive stress under a point load at the tip.

magnitude of this stress still needs to be calculated.
Achieving this requires the onerous recalculation of the
location of the neutral axis and the second moment of
area. Table 2 shows the relationship between the ratios
of the maximum bending stress to the stress at the
support, and the degree of taper. 15 UK tee sections
cut from UK Universal Beam sections were used.
Table 2 is plotted in Fig. 5 (using a line of best fit for
each tapering ratio). It was found that the stress ratio,
is determined by the geometric ratio, at the support:
f f
w
t b
ht

(9)
where, h is the height of the section and b
f
is the width
of the flange, and t
w
and t
f
are the thickness of the web
and flange, respectively.
The higher value of a section has, the lower value
of . The lowest and highest values hence represent
the upper and lower curves respectively in Fig. 5. A
high value of means that the web is dominant and a
low value of means that the flange is dominant. Fig. 6
0
2
4
6
8
10
12
14
0 0.2 0.4 0.6 0.8 1
D
e
g
r
e
e
o
f
t
a
p
e
r
,
h
/
h
'
UsingEq.(8)
MillersEq.(2)
TheoryusingEq.(3)
Support x/L
Tip

1334
shows the effect of tapering the web on the component
of the stress equation. As the web is tapered, the second
moment (I) of area decreases rapidly due to the d
3
term
(where I =bd
3
/12, and b and d are the breath and height
of an individual section.), while the distance to the
neutral axis, y
bar
decreases less rapidly.
3. Finite Element Analyses
Prior to the experimental work a FEA (finite element
analysis) was carried out to predict the stress patterns
and stiffness of a web tapered tee cantilever. The same
sections used in the experiments and also in Section 2.1
were analyzed. A 152 229 34 tee section cut from a
457 152 UKB67 was modeled in three
dimensions using a finite element analysis software
program, LUSAS [11]. An eight node thin shell
element was used. Five cantilevers whose support to tip
depth ratios of 1:1, 2:1, 3:1, 4:1 and 5:1 were
analyzed. Fig. 7 shows the bending stresses along the
beam when a point load of 15 kN was applied at
the tip.
It should be noted that as the degree of taper
increases, so does the area of compression in the web,
as shown in Fig. 7. As expected, the location of the
maximum stress is away from the support for the 3:1,
4:1 and 5:1 taper cantilevers. The results from the FEA
are compared with the theory and the experiment in the
Section of Results and Discussion.
4. Experimental Work
A total of six cantilever beams were tested in the
Heavy Structures Laboratory at Plymouth University.
Three sets with support to tip depth ratios of 1:1, 2:1
and 3:1 were prepared with two identical beams in each
set. To simulate a cantilever, two tee beams were
connected to each side of a column using extended end
plates and an equal load was applied to each cantilever
simultaneously, as illustrated in Fig. 8. Fig. 9 shows
one side of this test arrangement. The beams were
inverted, i.e., the top flange was down, so that loading
could be applied from the floor using a hydraulic jack,
as shown in Fig. 9.
The bending strains along the beam were measured
using 6 uni-directional strain gauges. The locations of
the gauges are shown in Fig. 2. Three more strain
gauges were installed at the mid-span, and 100 mm
before and after the mid-span of the 3:1 tapered beams
in order to determine the location and the magnitude of
the maximum stress. The deflections of the beams at
the tip were also measured.

Table 2 Ratios of the maximum bending stress to the stress at the support ().
Section
(From top to bottom of Fig. 5)

(ht
w
/b
f
t
f
)
Degree of taper (h/h)
2 2.25 2.5 2.75 3 3.5 4 4.5 5 5.5
146 127 22 0.50 1 1.012 1.042 1.082 1.128 1.231 1.339 1.447 1.552 1.651
305 305 90 0.60 1 1.009 1.036 1.074 1.119 1.220 1.331 1.446 1.561 1.675
165 152 20 0.54 1 1.009 1.036 1.074 1.118 1.219 1.329 1.444 1.560 1.675
127 152 24 0.80 1 1.009 1.035 1.072 1.117 1.218 1.329 1.444 1.559 1.672
178 203 37 0.68 1 1.009 1.035 1.072 1.116 1.217 1.328 1.443 1.558 1.673
191 229 45 0.72 1 1.008 1.034 1.071 1.115 1.215 1.325 1.440 1.556 1.672
229 305 70 0.79 1 1.007 1.032 1.068 1.110 1.209 1.318 1.433 1.549 1.667
152 229 34 0.89 1 1.006 1.029 1.064 1.105 1.201 1.308 1.421 1.538 1.656
292 413 113 0.87 1 1.006 1.029 1.063 1.105 1.201 1.307 1.420 1.536 1.655
210 267 46 0.82 1 1.006 1.029 1.063 1.104 1.200 1.306 1.419 1.535 1.653
229 305 57 0.85 1 1.006 1.029 1.063 1.104 1.197 1.306 1.419 1.535 1.653
140 203 23 0.86 1 1.005 1.028 1.062 1.102 1.196 1.301 1.413 1.529 1.646
267 381 87 0.95 1 1.005 1.027 1.060 1.100 1.193 1.297 1.408 1.523 1.641
305 457 127 0.93 1 1.005 1.028 1.061 1.102 1.196 1.302 1.414 1.529 1.648
102 152 13 1.24 1 1.003 1.022 1.052 1.088 1.175 1.272 1.377 1.486 1.599


1335

Fig. 5 Relationship between the stress ratio () and the degree of taper (h/h).
1
1.1
1.2
1.3
1.4
1.5
1.6
2 2.5 3 3.5 4 4.5 5
R
a
t
i
o
o
f
t
h
e
m
a
x
i
m
u
m
b
e
n
d
i
n
g
s
t
r
e
s
s
t
o
t
h
e
s
t
r
e
s
s
a
t
t
h
e
s
u
p
p
o
r
t
)
Degreeoftaper(h/h`)
|=0.5 |=0.6 |=0.7 |=0.8 |=0.9 |=1 |=1.1 |=1.2

= 0.5 = 0.6 = 0.7 = 0.8 = 0.9 = 1 = 1.1 = 1.2

1336

Fig. 6 The effect of web tapering on the magnitude of the stress.

(a) Untapered beam

(b) 2:1 taper

(c) 3:1 taper

1337

(d) 4:1 taper

(e) 5:1 taper
Fig. 7 The bending stresses along the beam when a point load of 15 kN was applied at the tip (N/mm
2
).

Fig. 8 Test arrangement.

P P

1338

Fig. 9 Test specimen.

5. Results and Discussion
5.1 The Location of the Maximum Bending Stress
It is often assumed that the maximum bending stress
of a tapered cantilever lies at the support where the
maximum moment occurs. However, due to varying
geometric section properties along the length of the
beam, the maximum bending stress may not occur at
the support and is in fact dependent upon the degree of
taper and the applied loading type. When a UDL is
applied the maximum stress always occurs at the
support, as shown in Fig. 1a, while a point load at the
tip is applied, the location of the maximum stress
depends on the degree of taper, as shown in Fig. 1b.
It was found that when subjected to a point load at
the tip and the degree of taper is greater than two and
up to seven, Millers equation could be used to
determine the location of the maximum stress. This
could be because a tee section has no bottom flange and
only its web is affected by taper. Hence, the
cross-section of the tee can be regarded as being
rectangular. When the degree of taper is greater than
seven, Millers equation does not accurately predict the
location of the maximum stress, as shown in Fig. 4. The
difference is as much as 5% when the degree of taper is
12. 5% could be crucial for long beams. The difference
could be due to the fact that Millers equation is based
on rectangular cross-sections, while the other two on
tee sections. As the degree of taper increases, the
cross-sectional area of the web towards to the tip
becomes relatively small, compared with that of the
flange. Millers equation ignores the latter. Eq. (8)
should be used to determine the location of the
maximum stress.
It should be noted that the location of the maximum
stress of a tapered tee-section cantilever subject to a
point load at the tip is governed by the degree of taper,
as shown in Fig. 4. Neither the length of the cantilever
nor the magnitude of the point load has an effect upon
the location of the maximum compressive stress.
Section designations have little influence on the
location, as shown in Fig. 10. It should also be noted
that Eq. (2) and Fig. 4 are only valid for web tapered
tee cantilevers with a point load at the tip and when the
support depth is bigger than the tip depth.
The bending stresses of all six beams from the
experiment are presented in Fig. 11. As expected, the
stress patterns resemble those in Fig. 1b. The
maximum stress occurs at the support for the untapered
and 2:1 tapered beams, while it moves away from the
support for the 3:1 tapered beams. Using Eq. (2) the
predicted position, x/L of the maximum stress of the


1339

Fig. 10 The locations of the theoretical maximum compressive stress with various tee section designations.

(a) Untapered beam

0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 0.2 0.4 0.6 0.8 1
o
x
/

o
s
u
p
p
o
r
t
x/L
3:1 Taper 152 x 229 x 30 4:1 Taper 152 x 229 30
5:1 Taper 152 x 229 x 30 3:1 Taper 210 x 267 x 41
4:1 Taper 210 x 267 x 41 5:1 Taper 210 x 267 x 41
3:1 Taper 305 x 305 x 119 4:1 Taper 305 x 305 x 119
Support Tip
300
250
200
150
100
50
0
50
100
150
S
t
r
e
s
s
(
N
/
m
m
2
)
Distancefromthesupport(mm)
UntaperedBeam1A:14.43kN UntaperedBeam1B:15.15kN
Theory:14.43kN LUSASOutput:14.43kN
1A and1Btensilemeasurementsareverycloseandare
thereforeclusteredonthisgraph
0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000
x/L

1340

(b) 2:1 taper

(c) 3:1 taper
Fig. 11 Comparisons of predicted and observed bending stresses along the beam.

300
250
200
150
100
50
0
50
100
150
S
t
r
e
s
s
(
N
/
m
m
2
)
2:1TaperedBeam2A:14.39kN 2:1TaperedBeam2B:14.81kN
300
250
200
150
100
50
0
50
100
S
t
r
e
s
s
(
N
/
m
m
2
)
3:1TaperedBeam3A:14.29kN 3:1TaperedBeam3B:14.48kN
0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000
0 200 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000

1341
Table 3 Comparison between the proposed method with the theory.
Section 254 343 76 171 178 26 127 178 17 102 127 13 133 102 15
h (mm) 343.7 177.4 174.4 128.5 103.3
b
f
(mm) 254.5 171.5 125.4 101.9 133.9
t
f
(mm) 21.0 11.5 8.5 8.4 9.6
t
w
(mm) 13.2 7.4 6.0 6.0 6.4
Geometric ratio, 0.849 0.666 0.982 0.901 0.514
Stress ratio (beam theory) 1.105 1.112 1.096 1.105 1.126
Stress ratio, using Fig. 5 1.104 1.115 1.097 1.100 1.122
Percentage difference (%) 0.049 -0.229 -0.088 0.414 0.386

3:1 tapered beam was at the mid-span (x/L = 0.5),
which is 1 m away from the support. Eq. (2) gives a
good agreement with this position, as shown in Fig. 11.
More tests with cantilevers with higher degrees of taper
should be carried out in order to test Eq. (2).
5.2 The Comparison of the Experimental Bending
Stresses with the Theory and the FEA
Fig. 11 compares the experimental bending stresses
along the beam with the theory and the FEA. The
stresses in the web are approximately up to 20 N/mm
2

less than predicted. They agree within a maximum
difference of 10%. Part of the difference could be
because the beam was inverted and the self-weight of
the beam cancelled approximately 4.4% of the applied
load for the untapered beam. This could have caused a
reduction of 2.2% in the bending moment. Further
discrepancies between results could be due to
experimental uncertainties in strain gauge
measurements and the values of Youngs modulus
used.
5.3 The Validation of the Method Determining the
Magnitude of the Maximum Stress
The proposed method for determining the magnitude
of the maximum stress was validated. 5 UK tee sections
were randomly chosen. These are not included in Table
2. Table 3 compares the values of using Fig. 5 with
stress ratios using the theory for 3:1 tapered beams.
The errors of the proposed method are less
than 1%.

6. Conclusions
The bending stresses of steel web tapered tee-section
cantilevers have been investigated. The cross-sectional
properties of such beams vary along the length and
therefore the highest stress may not always occur at the
support. The bending stresses of such beams depend on
the degree of taper and the moment gradient. In design,
the bending stresses should be checked not only at the
support, but also along the beam. The location of the
maximum compressive stress for various degrees of
tapered cantilevers was identified. It has been found
that when a UDL was applied, the maximum stress
always occurs at the support.
When a point load was applied at the tip, it was
found that the maximum stress is always at the support
for tapered tee cantilevers, whose tapering ratio (h/h)
was less than or equal to two. It was also shown that as
the degree of taper increased, the location of the
maximum stress moved towards the cantilever tip and
the magnitude of the stress also increased. It has also
been found that Millers equation (Eq. (2)) can be used
to determine the location of the maximum stress of a
tapered tee section cantilever whose tapering ratio is
greater than two and up to seven. Here, it should be
noted that the location of the maximum stress is solely
influenced by the degree of taper. When the degree of
taper is greater than seven, Millers equation is not
suitable and Eq. (8) should be used to determine the
location of the maximum stress. The determination for
the location of the maximum stress can be summarized
in Eqs. (10)-(13).


1342
When subjected to a UDL:
For all degrees of taper
0
L
x
(10)
When subjected to a point load at the tip,
0 2
'

L
x
h
h
(11)

1
'
2
'
7
'
2

h
h
h
h
L
x
h
h
(12)
7
'
h
h

(13)
for Eq. (13),
L
x
can be determined using Eq. (8).
A method for determining the magnitude of the
maximum bending stress was proposed. It was found
that the ratio of the maximum stress to the stress at the
support, is determined by the geometric ratio at the
support, , Eq. (9). The proposed chart, Fig. 5 can
predict the stress ratio within 1% error.
Acknowledgments
The authors would like to thank Doug Wharf, Mike
McCulloch and Terry Richards for their help and
advice on this project. The authors would also like to
thank the technical staff in the Heavy Structures
Laboratory at Plymouth University. Their technical
support is gratefully acknowledged.

References
[1] BS 5950-1:2000 Structural Use of Steelwork in
BuildingsPart 1: Code of Practice for DesignRolled
and Welded Sections, BSI, London, 2006.
[2] P. Buffel, G. Lagae, R. Van Impe, W. Vanlaere, J. Belis,
Design curve to use for lateral torsional buckling of
tapered cantilever beams, Key Engineering Materials
274-276 (2004) 981-986.
[3] W. Krefeld, D. Butler, G. Anderson, Welded cantilever
wedge beams, Welding Research 38 (1959) 97-112.
[4] S. Kitipornchai, N. Trahair, Elastic stability of tapered
I-beams, Journal of the Structural Division 98 (1972)
713-728.
[5] D. Nethercot, Lateral buckling of tapered beams,
Proceedings of The International Association for Bridge
and Structural Engineering 33 (1973) 173-192.
[6] A. Andrade, D. Camotim, P.B. Dinis, Lateral-torsional
buckling of singly symmetric web-tapered thin-walled
I-beams: 1D model vs. Shell FEA, Computers and
Structures 85 (2007) 1343-1359.
[7] M. Fischer, M. Smida, Coupled instabilities of tapered
cantilevers with a T-shaped cross-section, in: Proceedings
of the Third International Conference on Coupled
Instabilities in Metal Structures, 2000, pp. 379-388.
[8] W. Yuan, B. Kim, C. Chen, Lateral-torsional buckling of
steel web tapered tee-section cantilevers, Journal of
Constructional Steel Research 87 (2013) 31-37.
[9] G. Miller, The design of tapered levers, International
Journal of Mechanical Engineering Education 29 (1997)
95-130.
[10] T. Cosgrove, AD311: T-sections in bendingStem in
compression, New Steel Construction 15 (2007) 38.
[11] LUSAS Modeller on-Line Help, Finite Element Analysis
Ltd., Surrey, 2010.


Unanticipated Bond Failure over Supporting Band
Beams in Grouted Post-Tensioned Slab Tendons with
Little or No Initial Prestress
Raymond Ian Gilbert
Centre for Infrastructure Engineering and Safety, School of Civil and Environmental Engineering, The University of New South
Wales, Sydney 2052, Australia

Abstract: After grouting the ducts in a large band beam and one-way slab post-tensioned floor system in a large shopping mall (in
excess of 1.5 km
2
in plan), doubts arose concerning the level of initial prestress and the strength and serviceability of the slabs was
questioned. To assess the strength of the slabs, it was assumed, as a worst case scenario, that some spans may have been constructed
with zero initial prestress. Load tests on prototype specimens, with similar dimensions and reinforcement quantities as the floor slabs,
were performed to assess the stress development in initially unstressed, but fully grouted, tendons. Unexpected bond failures occurred
at the junctions of slab and band beam where the abrupt change in section depth caused a sudden change in tendon stress. The maximum
stress that was developed in the initially unstressed tendons at these locations was about 60% of the strength of the tendons.

Key words: Anchorage, development length, post-tensioned slabs, stress development, tendons, unbonded construction.

1. Introduction

Problems were encountered recently with the
post-tensioning installation in the floor slabs of a large
shopping mall constructed in the Middle East. The
floors consisted of over 1.5 million square meters of
continuous, post-tensioned, one-way slabs supported
on transversely spanning band beams. Soon after
construction, doubts arose concerning the level of
prestress in the slabs and questions were asked
concerning the strength and serviceability of the slabs
in regions where the level of initial prestress may have
been less than that specified. For the purpose of
assessing the strength of the slabs, it was assumed, as a
worst case scenario, that some spans may have been
constructed with zero prestress in the tendons.
Two series of load tests were carried out on site.
Eight full-scale concrete slab specimens (2.0 m wide
and 6.0 m long) reinforced with a combination of

Corresponding author: Raymond Ian Gilbert, Ph.D.,
research fields: reinforced and prestressed concrete structures.
E-mail: i.gilbert@unsw.edu.au.
conventional deformed reinforcing bars and grouted
strands were tested. The specimen dimensions, concrete
properties, reinforcement types, reinforcement
quantities and locations, duct type, tendon type and
tendon profile were similar to (or where possible,
identical to) the as-built floor slabs.
In each specimen, the tendons consisted of two
post-tensioning ducts, each containing four 12.7 mm
diameter strands. The ducts were fully grouted 28
days prior to testing. The initial prestressing force in
each strand was zero, i.e., the bonded strand was
initially unstressed and thus acted as non-prestressed
reinforcement throughout the test.
Each specimen was subjected to increasing
deformation to failure and the load-deflection
response was measured electronically from first
loading through to the peak load and then beyond into
the post-peak unloading region. Slip of the strand at
the end of each specimen was also measured. Tests
were also conducted on companion specimens to
determine relevant material properties.
DAVID PUBLISHING
D
Unanticipated Bond Failure over Supporting Band Beams in Grouted
Post-Tensioned Slab Tendons with Little or No Initial Prestress

1344
The purpose of the tests was to assess the tensile
stress that develops in the initially unstressed
post-tensioning strand at the critical section at peak
loads. As the strands were initially unstressed, the
stress that developed in the strand under increasing
load was entirely due to bond via the concrete-duct,
duct-grout and grout-strand interfaces.
2. Background
In a situation where the initial prestressing force is
zero, the maximum stress that can be developed in a
prestressing strand grouted inside a post-tensioning
duct depends on:
(1) the development length available for the strand
at the section under consideration;
(2) the quality of bond between the strand and the
grout;
(3) the slip that might take place between the duct
and either the grout inside the duct or the concrete
outside the duct;
(4) the deformation that develops at the level of the
strand due to the imposed actions.
The deformation referred to in four point can be
determined from first principles based on strain
compatibility and the requirements of equilibrium.
The Australian Standard for Concrete Structures [1],
specifies that the minimum development length for
untensioned 12.7 mm diameter strand is
d
=150 d
b
=
1,900 mm. Based on tests results [2-4], ACI 318-08 [5]
specifies that the minimum embedment length
required to develop a stress of f
ps
on each side of the
critical section in a pretensioned member may be
taken as:
b
se ps
b
se
d
7
) (
21
d
f f
d
f
|
|
.
|
\
|
+
|
.
|
\
|
=
(1)
where, f
se
is the effective prestress in the strand (in
MPa), f
ps
is the stress in the strand when the nominal
strength is reached (MPa), and d
b
is the nominal
diameter of the strand. In the case of strand with zero
initial prestress (f
se
= 0), the first term in Eq. (1)
disappears and the development length is the flexural
bond length given by
d
= (f
ps
/7) d
b
. Clearly, Eq. (1)
was never intended to be applied to a situation where
the effective prestress is zero [6, 7]. It does not
account for many of the factors that affect the
development length, including the concrete quality,
the cover, the spacing between wires, the concrete
confinement and so on. Little guidance is available for
estimating or checking the development length of a
strand in a grouted duct. For unstressed, deformed
wire in normal weight concrete, ACI 318-08 [5]
suggests that the development length may be taken as:
b
c
y
d
4 . 1
d
f
f
|
|
.
|
\
|
'
=
(2)
Although not strictly applicable for unstressed 12.7
mm diameter strand with f
y
=1,800 MPa in concrete
with
c
f ' =60 MPa, Eq. (2) gives
d
=2,100 mm.
If the specified tensile yield strength f
py
is to be
developed in an initially untensioned strand at any
cross-section, the imposed actions must cause a
change in strain in the steel sufficient to generate f
py

and the steel strand must be anchored on either side of
the cross-section by a length in excess of the
development length. The embedment lengths of the
strands in the actual floor slabs in the Middle East
shopping mall on either side of the critical positive
moment region at mid-span and the critical negative
moment regions at the interior supports were in excess
of the requirements of both AS3600-2009 and
ACI-318-08. However, because of the uncertain
nature of the bond conditions in the ducts in these
slabs, it was decided to explore this aspect more
thoroughly by testing.
3. Cross-Sectional Analysis
3.1 Series 1
In the first test series, three one-way slab specimens
(S1 to S3) were constructed and tested. The specimens
contained the same non-prestressed reinforcement and
the same prestressing steel and ducting as was used in
the as-built floors and the same concrete was specified.

1345
The grout materials and mix proportions were also
similar to that used in the slabs in the shopping mall.
Each specimen was subjected to the same curing
regime as was applied to the actual floors and all were
tested on the same day28 days after grouting of the
ducts.
Specimens S1 and S2 were tested in positive
bending over a simply-supported span of 5.6 m, as
shown in Fig. 1a. The specimens were prismatic and
of rectangular section.
S1 was 2,000 mm wide and, at 250 mm deep, it had
the same thickness as a typical exterior span of the
actual slabs in the shopping mall. It contained the
same conventional reinforcement as the actual slabs
and also contained a similar prestressing strand layout.
Specimen S2 was 2,000 mm wide, 200 mm thick, and it
was the same thickness and contained the same
conventional reinforcement and prestressing strand
layout as the positive moment region of a typical
interior span. Each specimen contained two fully-grouted

1.2 m 1.2 m 0.8 m 1.2 m 1.2 m

. . . . . . . . . .
A
p
= 784 mm
2

d
p
= 145 mm 200
A
st
= 864 mm
2

d
st
= 170 mm
6mm diam mesh
2.0 m
P/4 P/4 P/4 P/4

. . . . . . . . . .
A
p
= 784 mm
2

d
p
= 193 mm 250
A
st
= 1469 mm
2

d
st
= 219 mm
6mm diam mesh
2.0 m
transverse bottom bars (T10 - 200)
longitudinal bottom bars (11 - T10)

Two ducts with
4 12.7 mm
strands in each
duct (zero
prestress)
(c) Section of S2

D
Longitudinal bottom bars - 13 - T12 (S1)
2 Ducts + strand (zero prestress) 11 - T10 (S2)
5.6 m
Note: Cover to longitudinal non-prestressed bars is 25 mm and to the
duct at mid-span is 45 mm (same as cover in UAE slabs).
(a) Elevation of S1 and S2
transverse bottom bars (T10 - 150)
longitudinal bottom bars (13 - T12)
Two ducts with
4 12.7 mm
strands in each
duct (zero
prestress)
(b) Section of S1

Fig. 1 Elevation and cross-section of specimens S1 and S2.
ducts, with four 12.7 mm diameter prestressing
strands in each duct all with zero initial prestress. An
elevation of slabs S1 and S2 and the loading
arrangement are shown in Fig. 1, together with the
cross-section of each specimen at mid-span.
To test behavior of the slabs in the negative
moment regions over the band beams, an additional
specimen S3 was tested in Series 1. Specimen S3 was
6 m long, 2,000 mm wide and had a depth of 200 mm
at each end of the specimen and 450 mm in the central
2.1 m of the specimen. S3 contained the same
conventional reinforcement and prestressing strand
layout as the negative moment region of a typical
interior support region of the shopping mall slabs.
Details of specimen S3 and the loading arrangement
are shown in Fig. 2. The specimen contained two
grouted ducts, with each duct containing four 12.7
mm diameter strands. The level of prestress in all
strands was zero.
3.2 Series 2
The results for specimen S3 (in Series 1) indicated
that local bond failure occurred at the junction of the
slab and the band beam where there is a relatively
sudden change in the section depth and hence a sudden

Two ducts each containing 4 12.7 mm diam. strands
(zero prestress)

11 T10 (top)
+ 6mm diam square mesh
A
200
250
A 11 T10 (bottom)
11 1.8m

0.2m 1.75m 2.1m 1.75m 0.2m
Note: Top cover to the longitudinal conventional reinforcing bars is 50 mm
and top cover to the duct over the 2.1 m central region is 40 mm.

(a) Elevation S4

11 T10 longitudinal bars
2.0m + 6 mm wire mesh

D = 200 Two ducts each with
4 12.7 mm strands
(zero prestress)
(b) Cross-Section A-A
P/2

200
P/2
200
(a) Elevation S3

Fig. 2 Elevation and cross-section of specimens S3.
(a) Elevation of S1 and S2
(a) Elevation S3
(b) Cross-section A-A

1346
change in the tendon stress. At the critical section A-A
in Fig. 2a, slip occurred locally between the strands
and the grout and the strain in the tendon was less
than the strain in the adjacent concrete.
As a consequence of this observation, a second series
of tests on specimens similar to S3 was conducted in
order to provide confidence in the level of stress that
may reasonably be assumed to develop in the strand at
the critical negative moment region adjacent to the
band. Five additional specimens were constructed,
three with identical dimensions and reinforcement to
S3 (Fig. 2) and two with identical dimensions to S3, but
with heavier amounts of conventional reinforcement.
These additional specimens were labeled S3A, S3B,
S3C, S4A and S4B. The loading arrangement for each
specimen was as shown in Fig. 2. The conventional
reinforcement details for all specimens (Series 1 and 2)
are given in Table 1.
3.3 Test Set-Up and Loading Frame
The specimens were tested on-site in a specially
constructed testing frame (shown in Fig. 3). The load
was applied by a hydraulic jack of capacity 1,000 kN
and distributed to the specimen using an appropriate
spreader beam arrangement. The loads were applied to
the top surface of each specimen at each loading point
as a line load across the full 2 m width of each specimen.
Throughout the tests the loading frame was
adequately braced and the specimen and loading
elements were adequately supported to ensure stability.
Each test specimen and each spreader beam were
simply-supported such that rotation was permitted at
both ends and longitudinal (horizontal) movement
was permitted at one end only.
The loading frame and spreader beams were stiff
enough to ensure movements of the frame and loading
elements were insignificant compared to the
deformation of the test specimen. Frame deformations
were monitored throughout the test to ensure
movements were not excessive.
4. Test Results
4.1 Measurements Recorded
Each slab specimen was installed and appropriately
supported in the test frame, and then subjected to
increasing load P applied using a hydraulic jack. The
weight of the jack and spreader beam arrangements
was measured, so that the initial loads acting on the
top surface of the specimen were known prior to
commencement of the test.
Initially each specimen was subjected to a small
applied load and then unloaded, to ensure that all
measurement transducers and gauges were working,
and to ensure full contact at all loading and support

Table 1 Conventional reinforcement details.
Specimen
Top reinforcement Bottom reinforcement
Longitudinal Transverse Longitudinal Transverse
S1
6 mm square mesh (200 ctrs) 13 T12 T10@150
25 mm 25 mm 37 mm
Cover Cover Cover
S2
6 mm square mesh (200 ctrs) 11 T10 T10@200
25 mm 25 mm 35 mm
Cover Cover Cover
S3 (S3A, S3B, S3C)
11 T10 + 6 mm mesh T10@200 + 6 mm mesh 11 T10 T10@200
50 mm 60 mm 25 mm 35 mm
Cover Cover Cover Cover
S4A (S4B)
20 T12 + 6 mm mesh T12@150 + 6 mm mesh 11 T10 T10@150
50 mm 62 mm 25 mm 35 mm
Cover Cover Cover Cover
Notes: T10 (T12) are deformed bars of 10 mm (12 mm) diam, bar spacings are in millimeters.

1347

Fig. 3 Specimen S2 under load in testing frame.

points. Each test specimen was then subjected to
increasing load to failure. The rate of deformation
applied to each specimen was controlled so that an
accurate picture of the load-displacement relationship
was obtained from first loading to the peak load and
beyond. After the peak load was reached, each
specimen was then unloaded.
The load applied by the hydraulic jack was
measured using an accurately calibrated load cell and
the deflection of each specimen was continuously
recorded at six locations along the span using LVDTs
(linear variable displacement transducers). In addition,
the slip of the tendons at the free ends of each
specimen (relative to the concrete end face) was also
measured using LVDTs.
The deflection of the cross-head or reaction beam at
the top of the loading frame at the point where the
hydraulic jack reacts against the cross-head were
recorded for each test and proved to be insignificant.
Deflection readings along the span of each specimen
were taken relative to the specimen supports (i.e., the
reaction slab) and not relative to the loading frame.
The reaction slab on which the loading frame was
mounted can be seen in Fig. 3.
As the load on each specimen was gradually
increased, the position of each crack was recorded.
From time to time, each test was paused and the
height of each crack was recorded, together with the
crack width at the soffit of each specimen.
For each slab specimen, companion specimens in
the form of standard 150 mm cubes and 150 mm
diameter concrete cylinders were cast and
appropriately cured in accordance with ASTM C
31M-98 and ASTM C 192M-98, as appropriate. At
the time of testing each slab specimen, the companion
specimens were tested to determine the concrete
compressive strength and the concrete elastic modulus
in accordance with ASTM C 39M-99 and ASTM C
469-94.
Tests were also conducted on samples of the
conventional reinforcing bars and the strand to
determine the complete stress-strain curve of the steel
samples, from which the elastic modulus, the yield
stress (0.2% proof stress), the tensile strength, the
corresponding strain at peak stress and the fracture
strain were determined.
4.2 Method of Evaluating Test Results
Each specimen was relatively lightly reinforced and
the conventional reinforcement yielded well before
the peak load was reached. From the measured peak
loads for each specimen and the known strengths and
stiffness characteristics of both the concrete and the
steel, the stress in the 12.7 mm diameter strands at the
critical section at peak loads (just prior to failure of
the specimen) was readily determined using only the
principles of mechanics and satisfaction of the
requirements for equilibrium and compatibility of
deformations.
If the calculated stress in the strand exceeded the
0.2% proof stress for the material, the full design
strength of the specimen was deemed to have been
reached and the strand had developed its full strength
via bond between the strand and the grout inside the
prestressing duct. If slip occurred between the strand
and the surrounding concrete, the full strength of the
strand may not have been developed.
4.3 Specimen S1
At the time of testing, the measured material
properties obtained from the companion specimens

1348
were:
Mean concrete compressive strength, f
cm
= 60 MPa;
Elastic modulus of concrete, E
c
= 30 GPa;
Yield stress of the conventional steel, f
y
= 575 MPa;
Elastic modulus of conventional steel, E
s
= 200 GPa;
Tensile strength of strand, f
pu
= 2,060 MPa;
Elastic modulus of strand, E
p
= 198 GPa.
The measured mid-span moment versus mid-span
and quarter-span deflection curves are shown in Fig. 4.
The specimen reached its peak load capacity when the
mid-span moment reached 456 kNm (which included
the moment caused by the self-weight of the slab) at a
peak applied load P
max
= 454 kN (consisting of the
maximum load applied via the hydraulic jack of 419
kN, plus the self-weight of the loading rig = 35 kN).
The mid-span deflection at the peak load was 128.5
mm. No draw-in of the tendons at the ends of the
specimen was recorded at any stage during the test.
An ultimate strength analysis was undertaken on
the cross-section at mid-span, using the measured
material properties, in order to estimate the maximum
strength in bending M
max
. Perfect bond was assumed
between the strands and the surrounding concrete (i.e.,
it was assumed that no slip occurred on the
strand-grout interface, the grout-duct interface and the
duct-concrete interface).

0
100
200
300
400
500
0 50 100 150 200
Deflection (mm)
M
i
d
-
s
p
a
n

M
o
m
e
n
t

(
k
N
.
m
)
Mid-span
Quarter-span
Calculated M
max
= 437 kNm
1820
1715

1370
1025
850
680
490
Stress in Strand
(MPa)
M
max
= 456 kNm at
max
= 128.5 mm.
M
i
d
-
s
p
a
n

m
o
m
e
n
t

(
k
N
.
m
)

Measured - span
Measured mid-span
Calculated mid-span
Deflection (mm)

M
i
d
-
s
p
a
n

M
o
m
e
n
t

(
k
N
m
)

Fig. 4 Measured vs calculated moment-deflection curves, S1.
Using the measured material properties and the
requirements of strain compatibility and equilibrium,
the calculated maximum moment at mid-span was 437
kNm (compared with the measured maximum
moment of 456 kNm). At the calculated peak moment,
the calculated stress in the tendons was
pu
= 1,820
MPa and, with the strain in the conventional tensile
reinforcement well past the measured yield strain,
st

= f
y
= 575 MPa.
The mid-span moment versus mid-span deflection
curve was also calculated, using E
c
= 30,000 MPa and
the mean flexural tensile strength of f
ctm
= 5.8 MPa.
Yielding of the conventional reinforcement took place
at a moment of 239 kNm when the stress in the
tendons was just 487 MPa. The post cracking
deflection was calculated using the approach for
calculating short-term deflection specified in
Eurocode 2 [8]. This approach is far more accurate for
non-prestressed slabs than the approach in ACI
318-08 (Bischoff [9] and Gilbert [10, 11]). The
calculated moment deflection curve at mid-span is
compared with the measured curve in Fig. 4.
4.4 Specimen S2
The material properties for S2 are identical to those
reported earlier for S1. The measured mid-span
moment versus mid-span deflection curve is shown in
Fig. 5, together with the calculated response of the
specimen. The maximum mid-span moment reached
during the test was 330 kNm at a peak load of P
max
=
325 kN (consisting of the maximum load applied via
the hydraulic jack, 290 kN, plus the self-weight of the
loading rig, 35 kN). The measured mid-span
deflection under the peak loads was 182 mm and no
draw-in of the tendons at the ends of the specimen was
recorded at any stage of the test.
Using the measured material properties and
enforcing the requirements of strain compatibility and
equilibrium, the calculated maximum moment at
mid-span was 298 kNm (compared to the measured
330 kNm). At this calculated peak moment, the stress
1,820
1,715
1,370
1,025
500
400
300
200
100
0

1349
0
50
100
150
200
250
300
350
400
0 50 100 150 200 250
Mid-span deflection (mm)
M
i
d
-
s
p
a
n

m
o
m
e
n
t

(
k
N
.
m
)
Experimental
Calculated

pu
= 2050 MPa
M
max
= 330 kNm at
max
= 182 mm.
Calculated M
max
= 298 kNm
Experimental
Calculated

M
i
d
-
s
p
a
n

m
o
m
e
n
t

(
k
N
m
)

Mid-span deflection (mm)
0 50 100 150 200 250
400

300

200

100

0

Fig. 5 Measured vs calculated moment-deflection curves, S2.

in the tendons was
p
= 2,050 MPa and the
conventional tensile reinforcement was well past
yield with
st
= f
y
= 575 MPa. Clearly, the measure
peak load exceeded the calculated peak load and
the strand must have developed its full tensile
strength.
The mid-span moment versus mid-span deflection
curve is also calculated, using E
c
= 30,000 MPa and the
mean flexural tensile strength of f
ctm
= 5.8 MPa, and the
calculated curve is also shown in Fig. 5. Yielding of
the conventional reinforcement took place at a
moment of 131 kNm when the stress in the tendons
was just 484 MPa. As for specimen S1, the post
cracking deflection was calculated using the approach
specified in Eurocode 2 [8].
4.5 Specimen S3
Specimen S3 was tested in an inverted position to
that shown in Fig. 2, with self-weight contributing to
the peak moment at the critical section (A-A in Fig. 2).
The measured moment versus deflection curves for S3
at Section A-A is shown in Fig. 6. The specimen
reached its peak load capacity when failure occurred
at Section A-A at the junction of the 200 mm thick
slab and the 450 mm deep interior band. The moment
at failure at Section A-A was 195 kNm when the
deflection at this point reached 80 mm and the load P
was 163 kN, which included the self-weight of the
loading rig (16 kN).
0
50
100
150
200
250
0 20 40 60 80 100
M
o
m
e
n
t

(
k
N
.
m
).
D fl ti ( )

M
o
m
e
n
t

(
k
N
m
)

pu
= 950 MPa
0 20 40 60 80 100
250

200

150

100

50

0
Deflection (mm)

M
o
m
e
n
t

(
k
N
m
)

Fig. 6 Measured moment-deflection curve for S3 (at A-A).

No slip was recorded at the ends of the specimen
during loading. In contrast to specimens S1 and S2,
the strands did not develop their full strength during
the load tests, with the tendon stress at peak load
reaching just 950 MPa at Section A-A.
4.6 Specimen S3A, S3B and S3C
were:
cm
= 53.3 MPa;
c
= 46.9 GPa;
y
= 575 MPa;
s
= 200 GPa;
Ultimate tensile strength of strand, f
pu
= 2,060 MPa;
p
= 198 GPa.
As for S3, each specimen was tested in an inverted
position to that shown in Fig. 2, with self-weight
contributing to the peak moment at the critical section
(A-A in Fig. 2). Each specimen reached its peak load
capacity when failure occurred at Section A-A at the
junction of the 200 mm thick slab and the 450 mm
deep interior band. For S3A, the moment at failure at
Section A-A was 219 kNm when the deflection at this
point reached 83 mm and P = 190 kN. For S3B, the
moment at failure at Section A-A was 207 kNm when
the deflection at this point reached 78 mm and P =
177 kN. For S3C, the moment at failure at A-A was
197 kNm when the deflection at this point reached 54
2,050 MPa

M
max
= 330 kNm at
max
= 182 mm

1350
mm and P = 165 kN. These values of P include the
self-weight of the loading rig (16 kN).
The measured moment versus mid-span deflection
curves for S3A, S3B and S3C at Section A-A are
shown in Fig. 7, together with the corresponding curve
for specimen S3. No slip was recorded at the ends of
either specimen during loading.
As observed in the series 1 tests, the tendons in
S3A, S3B and S3C did not develop their full strength
at Section A-A (Fig. 2) during these load tests. The
stress at failure in each specimen was 1,150 MPa
(S3A), 1,050 MPa (S3B) and 960 MPa (S3C).
4.7 Specimen S4A and S4B
were:
cm
= 59.6 MPa;
c
= 47.5 GPa;
y
= 575 MPa;
s
= 200 GPa;
Ultimate tensile strength of strand, f
pu
= 2,060 MPa;
p
= 198 GPa.
As for S3, each specimen was tested in an inverted
position with self-weight contributing to the peak
moment at the critical section (A-A in Fig. 2). Each
specimen reached its peak load capacity when failure
occurred at Section A-A at the junction of the 200 mm
thick slab and the 450 mm deep interior band. For
S4A, the moment at failure at Section A-A was 328
kNm when the deflection at this point reached 77 mm
and P = 315 kN. For S4B, the moment at failure at
Section A-A was 312 kNm when the deflection at this
point reached 64.5 mm and P = 296 kN. These values
of P include the self-weight of the loading rig (16 kN).
The measured moment versus deflection curves for
S4A and S4B at Section A-A are shown in Fig. 8.
Also shown on these curves are the calculated tendon
stresses at the peak load, based on the requirements of
equilibrium. No slip was recorded at the ends of either
specimen during loading.
0
50
100
150
200
250
0 20 40 60 80 100
Deflection (mm)
M
o
m
e
n
t

(
k
N
.
m
).
Slab 4 (10-12-2006)
Slab 4A (17-02-2007)
Slab 4B (18-02-2007)
Slab 4C (20-02-2007)
Stress in tendon at peak
moment load

pu
= 1150 MPa (S3A)

pu
= 1050 MPa (S3B)

pu
= 960 MPa (S3C)

pu
= 950 MPa (S3)
Slab S3
Slab S3A
Slab S3B
Slab S3C
Deflection (mm)

M
o
m
e
n
t

(
k
N
m
)

Fig. 7 Measured moment-mid-span deflection curve for
S3, S3A, S3B and S3C (at A-A).

0
50
100
150
200
250
300
350
0 20 40 60 80 100
Deflection (mm)
M
o
m
e
n
t

(
k
N
.
m
)
Slab 3 (10-12-2006)
Slab 3A (18-02-2007)
Slab 3B (19-02-2007)
pu
= 1330 MPa (S4A)

pu
= 1170 MPa (S4B)
Deflection (mm)

M
o
m
e
n
t

(
k
N
m
)

Slab S4A
Slab S4B

Fig. 8 Measured moment-mid-span deflection curve for
S4A and S4B (at A-A).

As observed in the S3 specimens, the tendons in
S4A and S4B did not develop their full strength at
Section A-A during these load tests. The stress at
failure in each specimen was calculated to be 1,330
MPa (S4A) and 1,170 MPa (S4B).
5. Discussion of Results
For the simply-supported supported specimens S1
and S2, of uniform thickness and spanning 5.6 m, no
slip occurred between the strands and the surrounding
concrete at any stage of loading and the full strength
of the tendons was developed through bond. The
1,150
1,050
960
950
1,330 MPa
1,170 MPa

1351
stress in the strand at the ultimate limit state can be
calculated by enforcing equilibrium and assuming full
strain compatibility, i.e., by assuming the strain in the
tendons is equal to the strain in the concrete at the
tendon level at all stages of loading. For the strand in
these specimens, the full strength of the tendons was
developed over a development length of 2,600 mm
from the end of the specimen to the maximum moment
location under the interior loading point (Fig. 1).
For specimens S3, S3A, S3B, S3C4, S4A and S4B,
local bond failure occurred at the junction of the slab
and the band beam where the sudden change in the
section depth resulted in a rapid change in the tendon
stress and high local bond stresses. The development
length of the strand from the end of each specimen to
the point of maximum moment in each specimen was
in excess of 2,100 mm and, according to AS
3600-2009 and ACI 318-08, this should have been
sufficient to develop a yield stress of about 1,800 MPa.
At the critical section, slip occurred locally between
the strands and the grout and the strain in the steel was
less than the strain in the adjacent concrete. In all of
these specimens, the maximum stress that was
developed in the strand at the peak load was between
46% and 65% of the ultimate strength of the steel.
To provide confidence in the level of stress that
may be assumed to develop in the strand at the critical
negative moment region adjacent to the band, a
simple statistical analysis was undertaken on the
results of the six relevant specimens. The results are
summarized in Table 2.

Table 2 Strand stresses at Section AA in Fig. 2.
Specimen
Stress in tendon at peak load at
Section AA,
pu
(MPa)
S3 950
S3A 1,150
S3B 1,050
S3C 960
S4A 1,330
S4B 1,170
Mean,
pu
(MPa) 1,102
Standard Dev., s

(MPa) 145

If the sample of measured tendon stresses is
assumed to be normally distributed, the lower
characteristic value of the tendon stress at the ultimate
load is
pu.0.05
=790 MPa. This is the value of tendon
stress at peak loads that is smaller than 95% of all
samples. This measure is consistent with the factors of
safety imposed on material strengths in many
international standards, including ACI319-08 [5].
It is suggested, therefore, that the ultimate strength
of the negative moment region in the slabs at the edge
of the band beam (where the section depth suddenly
increases and the tendons are located in the top of the
slab) may be calculated based on the requirements of
equilibrium and strain compatibility, assuming slip
does occur between the strand and the concrete and
the maximum tendon stress does not exceed 790 MPa.
6. Conclusions
Eight full-scale concrete slab specimens reinforced
with a combination of conventional deformed
reinforcing bar and grouted strands (with zero initial
prestress) were tested to assess the stress that could
develop in the slab tendons in the positive and
negative moment regions of a one-way slab supported
on band-beams. In each specimen, the tendons
consisted of two post-tensioning ducts, each
containing four 12.7 mm diameter strands. The ducts
were fully grouted 28 days prior to testing. The initial
prestressing force in each strand was zero, i.e., the
bonded strand was initially unstressed and thus acted
as non-prestressed reinforcement throughout the test.
In each specimen, the strand was full-anchored in
accordance with AS 3600-2009 [1] and ACI
318.08 [5].
For the prismatic specimens of constant depth, the
strand at the critical section developed its full design
strength, with the stress reaching the specified yield
stress of the strand as the peak load was approached.
For the non-prismatic specimens where the slab depth
suddenly changed at the slab and band-beam
connection (from 200 mm to 450 mm), bond failure

1352
occurred and the maximum stress that could be
developed in the strand was little more than 50% of
its design yield strength.
Where a post-tensioned concrete slab suddenly
increases in depth, such as at the edge of a drop panel
or band beam, there is a relatively sudden change in
the stress in the tendon resulting in high local bond
stresses. In situations where the initial prestressing
force in the tendon is small, the assumption that the
full strength of the tendon can be achieved at the
ultimate limit state may not be reasonable.
Acknowledgments
This work was undertaken as part of a consultancy
and a broader research program on development and
lap lengths of reinforcement in concrete with the
financial support of the Australian Research Council.
This support is gratefully acknowledged.
References
[1] AS 3600-2009, Australian Standard for Concrete
Structures, Standards Australia, Sydney, 2009.
[2] N.W. Hanson, P.H. Kaar, Flexural bond tests
pretensioned beams, American Concrete Institute Journal
55 (7) (1959) 783-802.
[3] P.H. Kaar, R.W. la Fraugh, M.A. Mass, Influence of
concrete strength on strand transfer length, Journal of the
Prestressed Concrete Institute 8 (5) (1963) 47-67.
[4] P. Kaar, D. Magura, Effect of strand blanketing on
performance of pretensioned girders, Journal of the
Prestressed Concrete Institute 10 (6) (1965) 20-34.
[5] ACI 318M-08, Building Code Requirements for
Structural Concrete and Commentary, American
Concrete Institute, Detroit, 2008.
[6] L. Martin, W. Korkosz, Strength of prestressed members
at sections where strands are not fully developed,
Prestressed Concrete Institute Journal 40 (5) (1995)
58-66.
[7] PCI Design Handbook, Precast and Prestressed Concrete,
6th ed., MNL-120-4, Precast/Prestressed Concrete
Institure, Chicago, 2004.
[8] Eurocode 2, Design of Concrete Structures Part 1-1:
General Rules for Buildings, DD ENV 1992-1-1, European
Committee for Standardisation (CEN), Brussels, 1992.
[9] P.H. Bischoff, Reevaluation of deflection prediction for
concrete beams reinforced with steel and FRP bars,
Journal of Structural Engineering, American Society of
Civil Engineers 131 (5) (2005) 752-767.
[10] R.I. Gilbert, Reevaluation of deflection prediction for
concrete beams reinforced with steel and fiber reinforced
polymer bars, discussion, Journal of Structural
Engineering, American Society of Civil Engineers 132 (8)
(2006) 1328-1330.
[11] R.I. Gilbert, Tension stiffening in lightly reinforced
concrete slabs, Journal of Structural Engineering,
American Society of Civil Engineers 133 (6) (2007)
899-903.

Journal of Civil Engineering and Architecture, ISSN 1934-7359, USA

A Computer Modeling of the China Central Television
Headquarters in Beijing
Valentina Forcella
Politecnico di Milano, Milano D.I.C.A. 20131, Italy

Abstract: This paper deals with the computer modeling of structures starting from a point cloud. The CCTV (China Central
Television) tower headquarters is the case for study because the shape of this building is non-stellar, concave and multi-connected. It
is composed of sowns and chains. The sown is the representation of a horizontal plane formed by dense points. The chain is a planar
path modeled by rare points. The CCTV structure is defined only by the three orthogonal Cartesian coordinates of the points. The
proposed computer modeling uses a sequence of procedures and the desired outputted 3D model is consistent. The first procedure is
devoted to attributing points to their voxel and to estimating three values needed afterwards. The second procedure is devoted to
analyzing clusters vertically and horizontally, to preliminarily distinguishing chains from sowns and to generating relational
matching. The third procedure is devoted to building closed paths between all chains and all their projections on sowns. The fourth
procedure is devoted to connecting points with triangles. The fifth procedure, still being implemented, is devoted to interpolating
triangles with triangular splines. The results show it is possible to achieve the 3D model using the above mentioned procedures.
These procedures are written, implemented and tested and they form a library of peoples own software. The code is written using
Matlab. It is not possible to obtain the required 3D model if the procedures are applied in the wrong order or one step is skipped. To
conclude, it is possible to obtain the computer model of the CCTV using the provided sequence of procedures.

Key words: CCTV tower, cluster analysis, Delaunay triangulation, Bzier spline.

1. Introduction

This paper deals with the computer modeling of a
building, starting from a point cloud. The test example
is the China Central Television Tower Headquarters,
located in Beijing.
The construction of the building began in
September 2004 on the 20 hectare site of an
abandoned motorcycle factory in Beijings new
Central Business District and was completed by the
OMA (Office for Metropolitan Architecture) in
August 2008 [1].
There are two modeling problems. The first one
relates to the fact that the structure is modeled by
points, acquired from pictures available on the web
and in architectural literature, and defined only by
their three orthogonal Cartesian coordinates, without

Corresponding author: Valentina Forcella, M.Sc., Ph.D.,
engineer, research fields: environmental and infrastructure
engineering, territorial safety and control engineering. E-mail:
forcella.valentina@gmail.com.
any other distinction, as shown in Table 1, extracted
by the input data set.
The second one relates to the fact that the shape of
this building is:
(1) non-stellar: starting from one point, to reach all
the other points, it may be necessary to go outside the
structure;
(2) concave;
(3) multi-connected: this means that there is a hole.
The structure is composed of sowns and vertical or
almost vertical walls (treated as chains).
To solve these problems, the solution strategy is
composed of procedures. The first procedure checks
the input data set [2], identifying the minimums and
the maximums of X, Y and Z. The voxel procedure,
a method based on octree encoding, is used also to
estimate three values needed afterwards. The second
procedure divides the point cloud into clusters, first
vertically and then horizontally [3]. This second
procedure also classifies the data into sets of chains
DAVID PUBLISHING
D
A Computer Modeling of the China Central Television Headquarters in Beijing

1354
Table 1 Extract of input data set.
ID X Y Z
253 0.000 120.000 -20.000
273 200.000 120.000 -20.000
274 0.000 130.000 -20.000
294 200.000 130.000 -20.000
295 0.000 140.000 -20.000

and sowns and relates points belonging to different
clusters. The third procedure generates closed paths
between points classified as belonging to the same
horizontal cluster. These closed paths are built
differently depending on the fact the cluster is
classified as a chain or as a sown. The fourth
procedure connects points with a Delaunay
triangulation. The last procedure, still being
implemented, will create a triangular interpolation
using triangular splines.
The results show it is possible to obtain the 3D
model using the above mentioned solution strategy.
The strategy is composed of coded using Matlab [4].
It is not possible to obtain the required 3D model if
the procedures are applied in the wrong order or one
step is skipped.
To conclude, it is possible to obtain the computer
model of a structure if its shape is non-stellar, concave
and multi-connected. The case study was the CCTV
and the provided sequence of procedures outputted the
desire consistent 3D model.
The imposition of an arbitrary reference system is
needed to remove the uncertainty of the origin,
orientation and scales. The origin is placed on the left
hand side, the axes as right-hand coordinate system,
and 1 m in reality is 1 unit in the data set.
The results will be given in the same reference
system in which the coordinates are provided as
shown in Fig. 1.
The first 3D plot shows the structure represented by
the group of points in Fig. 2.

Fig. 1 Reference system adopted.

Fig. 2 3D plot of point cloud.
Z
X
Y

1355

As mentioned, a sequence of procedures have to be
adopted to obtain three dimensional model. The
procedures used by the author are given
below.
2. Voxel
A voxel is an element of the total volume,
representing a value on a regular grid in 3 dimensional
space. The voxel information is organized in matrix
and sparse data structure must be handled [5]. The
position of a voxel is inferred based upon its position
relative to the other voxels.
The first procedure is devoted to checking the data
set, to detecting the minimums and the maximums of
X, Y and Z, to attributing points to their voxel and to
estimating the three values needed afterwards.
The input file is CCTV.txt containing a list with
the:
(1) point ID;
(2) X-coordinate;
(3) Y-coordinate;
(4) Z-coordinate.
The data set was checked to ensure that both the ID
point and its coordinates are correct, consistent and
without any duplication. If one of these two cases
occurs, a list with the number ID, coordinate X, Y and
Z is provided. If there are no double points, it is
shown. The service file CCTV_RID.txt contains the
checked data which will be used subsequently.
In this specific case, there are neither double points
nor points out of height so CCTV.txt is equal to
CCTV_RID.txt.
The voxel procedure is also used due to the
estimation of the three values needed afterwards:
(1) The first parameter defines vertical scanning
and divides horizontal surfaces (which model sowns)
from chains;
(2) The second parameter defines horizontal
clustering and generates finite point sets at the same
level, if necessary;
(3) The third parameter is used to match vertically
between points belonging to consecutive levels; this
parameter is also used to distinguish sowns from
chains.
In CCTV modeling, the produced parameters are:
(1) height step: 7.969 m;
(2) planimetric distance: 17.678 m;
(3) planimetric tolerance between points at different
levels: 4.000 m.
It is also necessary to find minimums and
maximums for each coordinate, in order to set the
voxel parameters.
In this specific case, the values given are:
(1) X minimum: 0.000 m;
(2) X maximum: 20.000 m;
(3) Y minimum: 0.000 m;
(4) Y maximum: 20.000 m;
(5) Z minimum: -20.000 m;
(6) Z maximum: 235.000 m.
For each order of voxel, a list provides:
(1) the total number of points;
(2) the voxel ID;
(3) the number of points that belong to that voxel.
Another list contains only the full voxels, their
full-voxel ID, their voxel ID, and the number of points
that belong to that full voxel.
The voxel procedure continues up to the order in
which the voxels are all empty.
In this specific case, the results are:
(1) voxel 1
o
has no point;
(2) voxel 2
o
has no point;
(3) voxel 3
o
has 1 point (in yellow in Fig. 3);
(4) voxel 4
o
has 167 points (in blue in Fig. 3);
(5) voxel 5
o
has 1077 points (in red in Fig. 3);
(6) voxel 6
o
has 1802 points (in green in Fig. 3).
At the end of this procedure, the optimal order
voxel is found with its amplitude in X, Y and Z.
In this specific case, the parameters produced are:
(1) optimal voxel: 6
o
;
(2) X amplitude: 3.125 m;
(3) Y amplitude: 3.125 m;
(4) Z amplitude: 3.984 m.

1356

Fig. 3 Voxel distribution.

3. Contours
The second procedure is devoted to checking the
data set, to analyzing clusters vertically and
horizontally, to preliminarily distinguishing chains
from sowns, and to generating relational matching.
The input file could be CCTV.txt or CCTV_rid.txt.
In this case, CCTV.txt is used because in the initial
data set there are neither double points nor points out
of height. If there were outliers, the used file would be
CCTV_RID.txt and not the initial data set.
The input data set is listed with the point ID and X,
Y and Z coordinates for each points.
The point ID can be the name of the point or the
position in the list: both are used and listed.
The output files are:
(1) CCTV_ord.txt;
(2) CCTV_intermediate.txt;
(3) CCTV_assemby.txt.
The service file is CCTV_connections.txt, the data
set can be processed in two different ways:
(1) by checking the input;
(2) by checking the data set and analyzing it in
terms of clusters and relational matching.
The data set is checked in terms of planimetric
distance. If points are too close together, only one of
them is kept. After this checking procedure, the data
set is analyzed.
Input coordinates could be supplied in all possible
combinations (X, Y and Z; X, Z and Y; ), the order
chosen is 123 (X, Y and Z).
Input data set can also be scaled and rotated, the
adopted scales are:
(1) +1.000 m in X axis;
(2) +1.000 m in Y axis;
(3) +1.000 m in Z axis.
and the adopted rotations are:
(1) 0.000 m in X axis;
(2) 0.000 m in Y axis;
(3) 0.000 m in Z axis.
The estimation of the three values needed (the
height step between points at consecutive levels, the
planimetrical distance between points at the same
level and the planimetrical tolerance between points at
consecutive levels) is done using voxel procedures.
The parameters adopted in contours procedure are:
levels: 4.000 m.

1357
The height step parameter, estimated using the
voxel procedure, is necessary to do vertical
clustering.
For each level, the information is listed in this
way:
(1) vertical cluster ID;
(2) number of points belonging to that vertical
cluster;
(3) height;
(4) point IDs at that level.
A typical extract of the vertical clusters is shown in
Table 2.
In this case, fifty two 2-dimensional plots were
obtained. The 3-dimesional vertical clusters is shown
in Fig. 4. Two 2-dimensional plots of the fifty-two
plots are shown in Fig. 5.
The planimetric distance parameter, estimated using
the voxel procedure, is necessary to do horizontal
clustering. In order to generate contours, finite point
sets at each level must be constructed.
These sets could involve all points belonging to a
vertical cluster, or there could be different horizontal
clusters at the same level, if necessary. This second
case is displayed in Fig. 6.
As shown in Table 3, each horizontal cluster is
described by:
(1) its horizontal cluster ID;
(2) the number of points that belong to the clustered
level;
(3) its level;
(4) the point ID that belongs to the cluster.
There are two horizontal clusters at level 55.000 m:
this is the case in which two different sets are done at
the same level.
There are 83 horizontal clusters.
The preliminary distinction between sowns and
chains is made considering firstly the number of
points and secondly, and in the case of ambiguity, the
local density.
Table 2 Vertical clusters.
Vertical ID # points Height (m) Point IDs
1 80 -20.000 1 441
16 38 55.000 2793 2830
52 153 235.000 4338 4496

Fig. 4 Vertical clusters.

1358

Fig. 5 Two of the 52 vertical clusters.

Fig. 6 Two horizontal clusters are needed.

1359

Fig. 7 Preliminary distinction between sowns (in green) and chains (in blue).

Table 3 Procedure output.
Horizontal ID # points Height (m) Point IDs
1 80 -20.000 1 441
17 18 55.000 2793 2810
18 20 55.000 2811 2827
83 153 235.000 4338 4496

Generally, points that belong to sowns are denser
than the ones that belong to chains. If the number of
points is the same, it is necessary to check the local
density.
The 3D plot is shown in Fig. 7. Although it is a
preliminary distinction, all clustered levels are
correctly classified, except height 45.000 m which is
classified as a chain instead of a sown. The
information is listed in this way:
(1) horizontal cluster ID;
(2) type of horizontal cluster;
(3) vertical cluster ID.
At this point, the data set is analyzed in terms of
relational matching and the planimetric tolerance
parameter, estimated with the voxel procedure, is
used.
The top picture in Fig. 8 shows an incorrect model
due to the incorrect matching, while the figure at the

Fig. 8 Incorrect and correct matching.

P
13
P
14

P
9
P
8

P
7

P
10
P
6

P
5

P
4
P
3
P
2
P
1
P
12
P
11
P
21
P
20
P
18
P
17
P
15
P
16
P
1
P
2
P
3 P
4

P
5

P
15
P
16
P
17
P
18

P
20
P
21
P
11
P
14
P
13
P
12
P
10
P
9 P
8

P
7

P
6


1360
bottom shows a consistent relational matching.
The connection must satisfy the following
properties:
(1) the points in question must be as close as
possible while still belonging to consecutive
levels;
(2) if points belong to the same vertical, they are
easily matched;
(3) if not belonging to the same vertical, the
strategy is to create connections along the same
direction which resulted in the right picture in Fig. 8
above.
The relational matching is done separately for
chains and for sowns. The former is classified
by:
(1) relational matching ID (between one chain and
another);
(2) horizontal cluster ID matched to the other
horizontal cluster ID;
(3) level matching the other level;
(4) the number of points belonging to that matching;
(5) the point ID matched to the other or others
point(s) ID.
Table 4 shows the output provided. The latter is
classified by:
(1) relational matching ID (between the sown and a
chain);
(2) horizontal cluster ID matched to the other
horizontal cluster ID;
(3) level matching the other level;
(4) the number of points belonging to that matching;
(5) the point ID matched to the other or others
point(s) ID;
(6) the number of sides matched.
In this specific case, there is a perfect
correspondence between each point so the matching
type is a one to one and the ID point is related to a
single point ID.
The total number of sides is also available. In the
modeling of CCTV, the matched chain sides are 148,
the matched sown sides are 18 and the total number
is 166. The specific results are omitted for
brevity.
4. Closed Paths
The third procedure is devoted to building closed
paths between all chains and all their projections on
sowns.
The input files needed are the ones generated in
contour procedures and are:
(1) CCTV_ord.txt;
(2) CCTV_connections.txt;
(3) CCTV_assembly.txt.
The output files are:
(1) CCTV_closedpath.txt;
(2) CCTV_connections_cla.txt;
(3) CCTV_contours.txt.
The four values needed are the three generated with
the voxel procedure and the fourth is a switch
parameter in order to choose what type of
interpolation should be used.
The switch parameter can assume two values:
(1) 0, for classic paths;
(2) 1, for Catmull-Rom paths.
The approach is to use the classic paths to model the
closed paths of points representing vertices of the
object, and to use Catmull-Rom lines for the smooth
structure.
The values of the first three parameters adopted in
the closed-paths procedure are:
levels: 4.000 m.
For CCTV modeling, this parameter is set at 0
because of the shape of the building.
In the first step, the data set is horizontally clustered,

Table 4 Relational matching for chains.
Hor_clu
ID
Height
(m)
#of
points
Point ID (first Hor_clu and
second Hor_clu)
1 -20 80 2 3 22 43
2 -15 80 443 444 463 484 445

1361

in the second, the data are connected one other, and
each horizontal cluster is finally classified as a chain
or as a sown. At this point, it is possible to generate
the closed paths, one for each horizontal clusters as
previously done. All the points must be taken only
once and all sides, if taken, only once too.
The generation of closed paths, simple in
elementary cases, becomes complex as the shape of
the path becomes more complicated. The shapes of
four letters F, G, R and W were created starting from
singular points in order to test this procedure. The
results are shown in Fig. 9.
To find a closed path, first, the author connect the
first point to its closest point. Then this second point
is connected to its closest point, excluding the first in
order to keep a unique sense of advancement, and so
on. To obtain the desired closed path, the sides do not
have to cross each other. If there are points not
connected to other points, they must be connected. To
do this, different hierarchic ways are used and
topological information is analyzed. In fact, topology
and geometry, in dimensions up through 3, are very
intricately related [6].
Typically, these isolated points belong to levels,
separated by a unique level, quite often consisting in
one or a few points. Moreover, further specific
analyses are developed, using to the same
mechanism.
Only at this step, it is possible to follow the path of
the path. Indeed following a level structure, the path
between its root and the farthest leaf identifies the first
half of the path. The second part is done attempting to
understand if starting from the farthest leaf the root is
reached.
In some cases, it is really impossible to reach the
root. Therefore, the whole path is formed by adding
single steps, starting from the point where the
previous step finishes. After all points have been
taken into account, all the sides inside the closed path
will be deleted.

Fig. 9 The four letters analyzed.

1362
4.1 Piecewise Catmull-Roms Lines
It is possible to continuously model every path with
piecewise Catmull-Roms lines. Starting from three
points, every straight line passes through the second
point and it is parallel to the line joining the first and
the third points.
Fig. 10 shows the Catmull-Rom closed path
between five generic points.
The advantage of a continuous interpolation is the
possibility to extrapolate other points, e.g., the
intersections between consecutive straight lines.
The output given is divided into two groups: the
first one for chains and the second for sowns
contours.
The former are organized as follows:
(1) chain perimeter ID;
(3) side ID;
(4) point ID with the ID of the next point on the
perimeter;
(5) the total number of points in the perimeter that
belong to chains.
The latter (closed paths for the sowns contour) is
organized as follows:
(1) sowns perimeter ID;
(2) horizontal cluster (sown) ID and horizontal
cluster (chain) ID;
(3) side ID;
(4) point ID with other point ID consecutively in the
perimetration;

Fig. 10 Perimeter with Catmull-Rom.
(5) the total number of perimetred points that
belong to chains;
(6) the total number of the points on the perimeter.
Interpolation using Catmull-Rom resulted in the
contours of chains and contours of sowns. The reports
are organized like so:
(1) interpolation ID;
(3) side ID;
(4) ID of the three consecutive points on the
perimeter;
(5) type of straight line:
y = ax + by
x = cy + dx
(6) in the first case, a y is displayed while in the
second, an x;
(7) the slope of the straight line;
(8) y-intercept or x-intercept.
Table 5 shows an extract of the interpolation results
for the chains.
If Catmull-Rom interpolation is set, the straight
lines intersect each other so the closed-path
procedure lists some additional information, divided
for chains and sowns:
(1) ID of the three consecutive points (belonging to
chains or to sowns) on the perimeter;
(2) X, Y and Z coordinates of the intersection
point;
(3) number of intersected points for chains;
(4) number of intersected points for sowns;
(5) total number of intersected points.
4.2 Classic Way
It is also possible to model every path in the classic
way, in order to make a comparison between the two
modes (Catmull-Rom and the classic way).

Table 5 Chains interpolation.
Side ID Point IDs Type Slope Intercept
81 442 443 444 y 0.000 0.000
82 443 444 445 y 0.000 0.000

P
1

P
3
P
5

P
4
P
2


1363
Fig. 11 shows the classic contour between the same
five points displayed in Fig. 10. Fig. 12 shows the
image of the contour of the chains, and Fig. 13 that of
the sowns. Fig. 14 shows the perimeter done in the
classic way:
The final distinction between sowns and chains is
done considering separate points that belong to
connections classified as sowns-chains or
sowns-sowns and points that belong to connections
classified as chains-chains.
For the former, the numbers of points is compared:
if the first group is composed of a greater number than
the other one, the first is classified as a sown while the
second one as a chain. On the contrary, if the first
group is composed of a lower number than the other
one, the first is classified as a chain while the second
one as a sown.
For the latter, first the existence of all sowns or all
chains is checked. If it is not so, a loop begins in order
to fix all chains making a comparison between the
number of points that belong to subsequently
horizontal clusters.
At the end, if there are some non-classified
connections, the type is set as a sown.
As shown in Fig. 15, all the clusters are correctly
classified.
The levels 0.000, 45.000, 205.000 and 235.000 are
correctly classified as sowns, all the others as
chains.

Fig. 11 Classic perimeter.

Fig. 12 Classic perimeter for a chain (10.000 m).

P
2

P
1
P
5
P
3
P
4

1364

Fig. 13 Classic perimeter for the contour of a sown (235.000 m).

Fig. 14 Classic perimeter.

1365

Fig. 15 Distinction between sowns (in green) and chains (in blue).

5. Triangles
The fourth procedure is devoted to connecting
points in triangles. These points belong to subsequent
chains or subsequent sowns or chains and their
projections upon the nearest sowns. There are many
possible triangulations, the adopted one is Delaunay
triangulation. Triangles are as equilateral as possible
and each triangle must not contain any other points.
This last fact implies that a circumcircle passing
through three points, should not contain any other
point. Delaunay triangulations maximize the smallest
of all the angles of the triangles in the triangulation.
This triangulation was invented by Boris Delaunay in
1934 [7].
Delaunay triangulation of a discrete point set P in a
general position corresponds to the dual graph of the
Voronoi tessellation for P.
Fig. 16 shows a random triangulation but it is
wrong because:
(1) The triangle should be as equilateral as possible;
(2) A circumcircle should not contain any of the
other points (it is shown in Fig. 17);
(3) The smallest inner angle must be at its
maximum breadth.
If the structure in concave, Delaunay triangulation
does not work so the following procedure is adopted:
If the middle point of the side of the triangle is inside
the closed path or on the perimeter or outside the
closed path. If it is inside or on the perimeter, the side
is kept. If the line is outside the closed path, it is
deleted.
For example, the middle points P
11
and P
12
in
Fig. 18, are outside the perimeter so sides P
7
-P
6
and
P
7
-P
2
are deleted.
If the body has a hole, the 3D model must have the
same hole. So if a side is found within the hole, it will
be deleted.
For example, the middle points P
9
in Fig. 19, is
inside the hole so the side P
1
-P
5
is deleted.
The input files needed in the triangles procedure
are:
(1) CCTV_ord.txt;
(2) CCTV_connections_cla.txt;
(3) CCTV_assembly.txt;
(4) CCTV_closedpaths.txt;
(5) CCTV_contours.txt.

1366

Fig. 16 Random triangulation.

Fig. 18 Wrong Delaunay triangulation for a concave
body.

Fig. 17 Delaunay triangulation (in blue), Voronoi
tessellation (in orange) and circumcircle (in yellow) [8].

Fig. 19 Wrong Delaunay triangulation for a
multi-connected body.

Fig. 20 Delaunay triangulation (level -5.000).
P
4

P
5

P
6

P
1

P
3

P
2
P
4
P
5
P
9
P
2
P
6
P
3
P
7

P
1
P
8


1367

The Delaunay triangulation is also applied to points
that belong to different but consecutive levels. This is
done in order to create a rough 3D surface. Fig. 21
provides the 3D plot.
The output files given are:
(1) CCTV_triangles.txt;
(2) CCTV_newpoints.txt.
The three values needed are estimated using
voxel procedures and are:
levels: 4.000 m.
The triangulation can connect points belonging to
the horizontal cluster, at the same level, as shown in
Fig. 20.
Each triangulation is associated with a Voronoi
tessellation.
6. Bzier Spline
The fifth procedure, still being implemented, is
carried out using one of peoples own software called
tri_spline. It is devoted to interpolating triangles
with triangular splines. For this task, the Bzier spline
is used, because it forms continuous surfaces with one
continuous derivative surfaces.
Fig. 22 shows 7 points (in red) with their point IDs
(in black) and two connected Bzier splines (in blue).
The endpoint of the first Bzier spline is the start point
of the second Bzier spline.
The computer modeling becomes smoother, even if
surface breaklines are accepted. It is noticed that
discontinuities of second derivatives are geometrically
a bit less evident and therefore neglected.
It is necessary to point out that it is difficult to
identify breaklines not highlighted in the input.
Another theoretical problem is the following: a pair of
coordinates should be used instead of the one
dimensional curvilinear coordinate. This leads to
analytical complexity of the solution because this pair
of coordinates is not yet globally defined.

Fig. 21 Delaunay triangulation.

Fig. 22 Two splines connected in the middle point.

At the moment, some theoretical aspects should be
studied in detail, recognizing the elegance of the
Bzier spline, even in one-dimensional space.
7. Future Works
It is evident that future work has to be focus on
solving existing problems. First of all, the evaluation
of the first three parameters done using the voxel
procedure are an underestimation of the real
values.
The second problem to be analyzed is the
possibility to model almost-horizontal sowns correctly.
In this specific case, the top of CCTV is not horizontal,
while in the 3D model it is analyzed and treated as a
horizontal plan.
The third problem is the fact that the Bzier spline
is well known, but there is no definite coordinate pair
so some extra studies are needed in order to find a
solution.

1368
8. Conclusions
The authors first comment regards the results as a
whole: they try to take a step forward compared to a
wide range of previously collected samples. In any
case, all programs (except for computer graphics
applications) are coded, implemented, tested and used
forming a library of own software.
The second comment relates to the approach used:
the user is provided a graphic result and *.txt files
where all the results are analyzed.
Not only convex structures, but also concave,
non-stellar and multi-connected can be processed.
The geometric characteristics of the model are put
first than the graphic outputted model. It implies the
attention is focused on how to obtain the computer 3D
model automatically and not on the graphical
restitution. For this reason, the final model do not
contain shadows, light effects and so on.
References
[1] P. Jodidio, Architecture Now!, Taschen, Germany,
2010.
[2] D.M. Hawkins, Identification of Outliers, Chapman and
Hall, London, 1980.
[3] L. Kaufman, P.J. Rousseeuw, Finding Groups in Data,
Prentice All, Engliwood Clifts, Wiley & Sons, New York,
1990.
[4] G. Naldi, L. Pareschi, Matlab, Concepts and Projects,
Apogeo, Lavis (TN), 2007.
[5] A. George, J. Liu, Computer Solution of Large, Sparse,
Positive and Definite System, Prentic-Hall, Engliwood
Clifts, New Jersey, 1981.
[6] W. Thurston, Three-Dimensional Geometry and
Topology, Princeton University Press, Princeton,
1997.
[7] B. Delaunay, Sur la sphre vide, Izvestia Akademii Nauk
SSSR, Otdelenie Matematicheskikh i Estestvennykh
Nauk 7 (1934) 793-800
[8] F.P. Preparata, M.I. Shamos, Computational Geometry:
An Introduction, Springer, New York, 1985.


Infilled Reinforced Concrete Beams for Sustainable
Construction
Patel Rakesh
1
, Dubey Sheo Kumar
1
and Pathak Krishna Kant
2
1. Department of Civil Engineering, Maulana Azad National Institute of Technology, Bhopal 462051, India
2. Department of Civil and Environmental Engineering, National Institute of Technical Teachers Training and Research, Bhopal
462002, India

Abstract: This paper presents a study carried out on infilled reinforced concrete beams for sustainable construction. In reinforced
concrete beams, less stressed concrete below neutral axis can be replaced by some light weight material to reduce the weight of the
structure and also achieve the economy. The used infilled material is brick. Sustainability can be achieved by replacing the partially
used concrete. By saving concrete, the authors save cement, which reduces the green house gases emissions. So it is considered as
environment friendly. Since infilled beam acts like a layered member, there needs a theory to analyze it. Method of initial functions is
used for the analysis of the infilled RC (reinforced concrete) beams. This method is successfully applied on infilled beam. Results show
that physical conditions are verified for infilled beam.

Key words: Initial functions, infilled beams, brick masonry, displacements, stresses.

1. Introduction

Sustainability means meeting the needs of the
present generation without compromising the ability
of future generations to meet their needs.
Sustainability can be achieved by reducing, reusing
and recycling the waste. It also means moving
towards green technologies and reduces emission of
CO
2
. Production of cement is also one of the main
causes of man-made carbon dioxide emissions.
Reduction of greenhouse gases and environmental
pollution is great concerned for all of us. So the
authors have to reduce the use of cement from
concrete structures to make concrete more sustainable
and environment friendly. This can be done by
replacing cement in concrete with larger amounts of
supplementary cementing materials, or by replacing
less stressed or unutilized concrete by other low cost
and environment friendly materials like alternative

Corresponding author: Patel Rakesh, research scholar,
research fields: the development of concepts and
methodologies for the concrete beams, composite material and
sustainable construction. E-mail: rakeshasct@gmail.com.
aggregates, bricks etc..
The main concept of the sustainable construction is
to save the natural resources for future generation, by
conserving virgin materials and protecting the
environment by reducing, re-using and recycling the
waste and reducing CO
2
emissions [1]. An attempt has
been made to design and produce sustainable concrete
by using micro silica. It is sustainable concrete
because waste material was used to produce high
strength, workability and durability [2].
These days attempts are being made for sustainable
construction work and economise the costly
reinforced concrete work. Like concrete, brick
masonry was also reinforced to make it ductile and
enable it to resist tensile forces. Reinforced brick work
may be designed either on the basis of linear elasticity
or the limit state philosophy of ultimate load design. It
also follows the assumptions of bending theory that
the plane sections remains plane after bending and
that the tensile strength of brick work may be
neglected.
In RC (reinforced concrete) beams strength of
DAVID PUBLISHING
D
Infilled Reinforced Concrete Beams for Sustainable Construction

1370
concrete lying near the neutral axis is not fully utilized.
The concrete just above the neutral axis less stressed
where as the concrete below the neutral axis serves as a
shear transmitting media. Partially utilized concrete of
RC beam has bean replaced by bricks. In this way, the
economy of reinforced brick beams and strength of RC
beams are tried to be combined in brick filled beams.
The bond between brick and concrete layers at the
brick concrete interface should essentially be very
good. It should be ensured that no slip will occur
between the two layers.
Experimental work is carried on the brick filled
reinforced concrete beams, with the view that the
stresses in the beams are maximum at the top and
bottom and zero at the neutral axis. So a cheap and
light material can be used near the neutral zone for
light weight and economy [3]. Behavior of brick
infilled reinforced concrete beams under cyclic
loading is find out experimentally and compared the
results with the conventional reinforced concrete
beam. It is observed that the load carrying capacity of
an infilled beam was about 80% of the conventional
reinforced concrete beams [4]. Experimental study of
brick filled reinforced concrete beam is done and it is
observed that saving of about 30% concrete is
achieved [5]. The analytical and experimental
investigations on infilled frames under static and
lateral loads are conducted and suggested that the load
carrying capacity and stiffness of the infill frame is
more than the plain frame [6]. The experimental
programme is conducted on the beams with light
weight brick core with ferrocement form. It is
observed that beams achieved higher ultimate load
when welded wire mesh is used in ferrocement forms
[7].
The experimental work was conducted to obtain a
stress-strain curve for the masonry, and determines the
splitting tensile strength of the bricks and Poissons
ratio of the masonry. It was observed that the average
splitting tensile strength of brick masonry was found
to be 6.8% of the average compressive strength of
bricks [8].
Method of initial functions is an analytical method
of elasticity theory. It can be used for layered
members. The method makes it possible to obtain
exact solutions of different types of problems, i.e.,
solutions without the use of hypotheses about the
character of stress and strain. It is used for the analysis
of beams under symmetric central loading and uniform
loading. Results are obtained for different end
conditions [9]. The method of initial functions is used
for the analysis of free vibration of rectangular beams
of arbitrary depth. And present the analysis for
different values of Poissons ratio [10]. MIF (method
of initial functions) has been applied for deriving
higher order theories for laminated composite thick
rectangular plates [11]. Developed governing
equations for composite laminated deep beams by
using method of initial functions. The developed beam
theory can be used for beam sections of small depth
and large depth [12]. Applied MIF (method of initial
functions) for the analysis of orthotropic deep beams
and compared the results with the available theory
based on Bernoullis hypothesis [13].
2. Brick Replacement Zone
The replacement zone in reinforced concrete beam,
where concrete below neutral axis may be replaced by
bricks was obtained with the help of stress block, used
for limit state design of RC beams. Fig. 1 shows the
stress block parameters for RC beams.
By equating total tension to total compression,
X
u
= (0.87 f
y
A
st
)/ (0.36 f
ck
b) as per IS 456 2000
[14].
0.466 fck
0.87 fy/Es +0.002 0.87 fy
D
T
C
4/7 Xu
3/7 Xu 0.002
0.0035
d
b
Xu

Fig. 1 Stress block parameters for RC beams.

1371
d
=(2 cover +diameter of bar)

d is the thickness of concrete layer available to
maintain bond between steel and concrete. It is
assumed the provision of thickness of concrete not
less than the cover provided on each side of the
tension reinforcement will satisfy the bond
requirement.
d
min
=3 Xu/7 +Xb
Brick replacement zone =d
max
d
min
This brick replacement zone is shown in Fig. 2.
3. Formulation Using MIF
The brick filled reinforced concrete beams are
analyzed by using MIF (method of initial functions).
According to this method, the basic desired functions
are the displacements and stresses, the system of
differential equations which are obtained from
equations of Hookes law and equilibrium equations
by replacing stresses by the displacements according
to elasticity relations. The order of the derived
equations depends on the stage at which the series
representing the stresses and displacements are
truncated.
In the case of a layered composite beam loaded at
the top surface, the state of stresses and displacements
at the free bottom surface of the beam is given by:
{ }
1
S = | |
1 1
, , 0, 0
T
u v (1)
Let
{ }
T
S = | |
, , ,
T
T T T T
u v Y X (2)

where, u
T
, v
T
, Y
T
and X
T
are the values of stresses and
displacements at the top surface of the layered beam.

Xu
N.A.
Xb
4/7 Xu
3/7 Xu
0.466fck
dmin
D
b
d'

Fig. 2 Brick replacement zone in RC beams.
Relating the stresses and displacements at the top
surface of the layer to those at the bottom surface by
successive application of the transfer matrix [L
i
]
across each layer, one obtains:
{ } | | | |
1
S A S
T
=

(3)

where,
| | | | | | | |
1 1 2 2
1 2
.......
h y h y h y N
L L L A
N N
= = =
=

(4)

The terms of the matrix [A] are evaluated after
expanding the exponential in the form of a series.
The matrix has a form:
[ ]
uu uv uY uX
vu vv vY vX
Yu Yv YY YX
Xu Xv XY XX
A A A A
A A A A
A
A A A A
A A A A
(
(
(
=
(
(

(5)

The Eq. (3) relates the boundary conditions at the
top surface to those at the bottom surface and is useful
for deriving governing differential equations for a
layered beam having a particular number of layers.
The method adopted for analyzing layered beams
involves the determination of initial functions at the
bottom surface of the beam by relating them through
the matrix [A] to the stresses at the top surface.
4. Application to the Infilled Beam
Brick filled reinforced concrete beam consists of
the three layers as shown in Fig. 3. Top and bottom
layers are of concrete and middle one is of brick. The
area of reinforcement below the brick replacement
zone is also converted in the equivalent area of
concrete by using modular ratio.

Fig. 3 Brick filled RC composite beam.

1372
Therefore, the matrix [A] becomes
| | | | | | | |
3 3 2 2 1 1
1 2 1
.
h y h y h y
L L L A
= = =
= (6)
where, h
1
, h
2
and h
3
are the thickness of three layers,
h
1
and h
3
are the thickness of concrete layers, and h
2
is
the thickness of brick layer.
The conditions at top are given by:
{ } | | , , , 0
T
T T T
S u v p =
(7)
Substituting the expressions (1) and (5) in the Eq. (3),
it can be got:
1 1
0
Xu Xv
A u A v + = (8)
1 1 Yu Yv
A u A v p + =
(9)
These equations are exactly satisfied by:
|
Xv
A u =
1
(10)
|
Xu
A v =
1
(11)

where, | is an unknown auxiliary function substituting
the value of u
1
and v
1
from the Eqs. (10) and (11) in
the Eq. (9), the differential equation governing the
problem of brick filled reinforced concrete beam is
obtained:
p A A A A
Xu Yv Xv Yu
= | ) ( ) (

(12)

The order of the governing differential Eq. (12)
depends on the order of the terms in the
matrix [A].
The auxiliary function | is chosen such that it
satisfies the governing differential Eq. (12), as well as
the boundary conditions at the edges of the beam.
Initial functions are obtained from Eqs. (10) and (11).
By operating on the initial functions by the transfer
matrix [L
i
] successively across each layer, the authors
can determine the stresses and displacements, within
the entire beam.
5. Analysis of Brick Filled Beam
The following values of beam dimensions are
chosen for the particular problem. The replacement
zone is calculated using above relations and thickness
of three layers are decided. Reinforced concrete brick
filled composite beam consists of the three layers as
shown in Fig. 3. Where, h
1
, h
2
and h
3
are the thickness
of three layers, H is the total thickness and b is the
width of beam section. h
1
and h
3
are the thickness of
concrete layers and h
2
is the thickness of brick layer.
H = 800 mm, b =300 mm and l = 4,000 mm;
h
1
= 210 mm, h
2
= 510 mm and h
3
=80 mm.
The following material properties are taken:
For concrete layers:
E = 22,000

N/mm
2
= 0.15, G = 9,200

N/mm
2
For brick layer:
E = 5,000

N/mm
2
= 0.10, G = 2,200

N/mm
2
A uniformly distributed load is assumed, on the top
surface of the beam:
Taking P

= 25 N/mm.
The boundary conditions of the simply supported
edges are:
X = Y = v = 0, at x = 0 and x = l
The boundary conditions are exactly satisfied by
the auxiliary function |.
| = A
1
sin (x/l)
Analysis is done for x = l/2 (mid span of beam)
The value of auxiliary function | is obtained from
Eq. (12), using this value of auxiliary function the
values of initial functions u
1
and v
1
are obtained from
Eqs. (10) and (11).
These values of u
1
and v
1
are substituted in Eq. (3)
for obtaining the values of stresses and displacements.
These values are shown in Table 1.
From Fig. 4, it is observed that the value of
displacement u is more at the top surface and less at
the bottom surface.
It can be seen from Fig. 5 that displacement v is
almost uniform throughout the depth.
From Fig. 6, it is observed that the value of normal
stress (Y) is zero at the bottom and maximum at the
top of beam. The physical condition of normal stress
equal to the applied normal load at the top fiber is
satisfied. This figure is drawn to verify the physical
condition of beam. It is seen from Fig. 7 that the shear

1373
Table 1 Values of stress and displacements at different
depths of brick filled beam.

S. No. y/H u/H v/H Y/P X/P
1 0.0 0.0594 0.1948 0.0000 0.0000
2 0.1 0.0474 0.1953 0.0456 1.7835
3 0.2 0.0373 0.1956 0.1419 2.1064
4 0.3 0.0275 0.1960 0.2521 2.3531
5 0.4 0.0180 0.1963 0.3724 2.5265
6 0.5 0.0085 0.1967 0.4989 2.6294
7 0.6 -0.0008 0.1971 0.6277 2.6646
8 0.7 -0.0102 0.1976 0.7548 2.6349
9 0.8 -0.0211 0.1977 0.8763 2.2355
10 0.9 -0.0331 0.1976 0.9657 1.3180
11 1.0 -0.0453 0.1974 1.0002 0.0010

Fig. 4 Variation of u through the depth of beam.

Fig. 5 Variation of v through the depth of beam.

stress (X) is maximum below the mid depth. It is
calculated at the mid of beam.
From this method results are obtained in terms of
two displacements u, v and stress components Y, X.
These terms for infilled RC beam are shown from

Fig. 6 Variation of normal stress (Y) through the depth of
beam.

Fig. 7 Variation of shear stress (X) through the depth of
beam.

Fig. 4 to Fig. 7. This theory gives better results for

layered members in comparison to elementary beam
theory. The eighth order MIF theory is used for
analysis.
7. Conclusions
The normal stress equal to the intensity of loading
and shear stress equal to zero at the top of beam are
obtain, this shows that MIF is successfully applied for
the analysis of infilled beams. The beam theories
based on MIF have advantage over other theories
because of their capabilities to converge to an exact
linear elasticity solution and so provide a governing
equation of desired order according to the
requirements of a beam problem.
Presence of bricks in the low stressed zone has not

1374
caused significant reduction in strength of reinforced
concrete beams.
Economy and reduction of weight in beams
depends on the percentage replacement of concrete by
bricks. Brick-filled reinforced concrete beams can be
used for sustainable and environment friendly
construction work as it saves concrete.
References
[1] J.P. Desai, New trends in green technologies for
sustainable construction, in: Proceedings of International
Seminar on Green Technologies for Sustainable Concrete
Construction, New Delhi, 2012.
[2] A.K. Al-Tamimi, N. Iqbal, Design and production of
sustainable structural concrete, in: Proceedings of
International Seminar on Green Technologies for
Sustainable Concrete Construction, New Delhi, 2012.
[3] P. Govindan, A.R. Santhakumar, Composite action of
reinforced cement concrete beams with plain masonry
infills, in: Proceedings of International Symposium on
Reinforced and Pre-stressed Masonry, Edinburgh, 1984.
[4] P. Govindan, M. Lakshmipathy, A.R. Santhakumar,
Behaviour of infilled beams under cyclic loading, Indian
Concrete Journal 61 (5) (1987) 124-128.
[5] U.B. Choubey, U. Gupta, Structural response of
brick-filled reinforced concrete beams, in: Proceedings of
4th International Seminar on Structural Masonry for
Developing Countries, Madras, 1992, pp. 226-233.
[6] S. Arulselvan, K. Subramanian, P.E.B. Pillai, A.R.
Santhakumar, RC Infilled frameRC plane frame
interactions for seismic resistance, Journal of Applied
sciences 7 (7) (2007) 942-959.
[7] H.F. Ezzat, N.A. Mohamed, B.S. Yousry, A.A. Ahmed,
Permanent ferrocement forms: A viable alternative for
construction of concrete beams, in: 30th Conference on
Our World in Concrete & Structures, Singapore, 2005.
[8] M. Gupta, A.K. Pandey, Experimental studies on brick
masonry in compression, Indian Concrete Journal 86 (1)
(2012) 43-50.
[9] K.T.S. Iyengar, K. Chandrashekhara, V.K. Sebastian,
Thick rectangular beams, Journal of the Engineering
Mechanics Division 100 (6) (1974) 1277-1282.
[10] K. Iyengar, T.S. Raja, P.V. Raman, Free vibration of
rectangular beams of arbitrary depth, Acta Mechanica 32
(1) (1979) 249-259.
[11] K.T.S. Iyengar, S.K. Pandya, Application of the method
of initial functions for the analysis of composite
laminated plates, Archive of Applied Mechanics 56 (6)
(1986) 407-416.
[12] S.K. Dubey, Analysis of composite laminated deep
beams, in: Proceedings of the Third International
Conference on Advances in Composites, Bangalore,
2000, pp. 30-39.
[13] S.K. Dubey, Analysis of homogeneous orthotropic deep
beams, Journal of Structural Engineering 32 (2) (2005)
109-166.
[14] Plain and Reinforced ConcreteCode of Practice, IS:
456-2000, Indian Standard, 2000.


Flexible Pavement Design Simulation Using
Mechanistic-Empirical Pavement Design Guide
Daba S. Gedafa
1
, Mustaque Hossain
2
, Stefan Romanoschi
3
and Andrew J. Gisi
4

1. Department of Civil Engineering, University of North Dakota, Grand Forks ND 58202-8115, USA
2. Department of Civil Engineering, Kansas State University, Manhattan KS 66506, USA
3. Department of Civil Engineering, University of Texas-Arlington, Arlington TX 76019-0308, USA
4. Kansas Department of Transportation, Materials and Research Center, Topeka KS 66611, USA

Abstract: The KDOT (Kansas Department of Transportation) is currently adopting MEPDG (mechanistic-empirical pavement design
guide) to replace the 1993 AASHTO (American Association of State Highway and Transportation Officials) design method. The main
objective of this study was to compare flexible pavement design using 1993 AASHTO design guide and MEPDG. Five newly built
Superior PERforming Asphalt PAVEments (Superpave), designed using the 1993 AASHTO Design Guide, were selected as test
sections for the design simulation study. Deflection data were collected approximately 8 to 10 weeks after construction using FWD
(falling weight deflectometer). The FWD deflection data were used to back-calculate the pavement layer moduli using three
different back-calculation programs. The existing pavement structures were analyzed for a 10-year analysis period. The maximum
numbers of years the existing pavement structures will be in a serviceable condition as well as the minimum thicknesses of different
layers to serve for 10-years were also determined. Effects of changing subgrade modulus, target distress, and reliability were also
investigated. The MEPDG design analysis shows that the 1993 AASHTO Guide-designed flexible pavements do not show the
distresses currently observed in Kansas for the 10-year design period. The MEPDG design simulation shows that the thinner the
pavement sections, the higher the permanent deformation. The existing pavement structures can serve for more than 20 years as per the
MEPDG design analysis if the default failure criteria and nationally-calibrated models are used.

Key words: Pavement design, MEPDG, 1993 AASHTO design guide, FWD deflection, back-calculated modulus.

1. Introduction

The most widely used procedure for the design of
flexible pavements is specified in the Guide for
Design of Pavement Structures, published in 1986 and
1993, by the AASHTO (American Association of
State Highway and Transportation Officials) [1, 2]. A
few states use the 1972 AASHTO Interim Guide
procedure, their own empirical or
mechanistic-empirical procedures, or a design catalog
[3]. The design methodologies in all those versions of
the AASHTO Guide are based on the empirical
performance equations developed using the AASHO

Corresponding author: Daba S. Gedafa, Ph.D., P.E.,
assistant professor, research fields: mechanistic-empirical
design, pavement management system, non-destructive testing,
materials engineering and traffic engineering. E-mail:
daba.gedafa@engr.und.edu.
Road Test data from the late 1950s.
Due to the limitations of earlier guides, a design
guide, based as fully as possible on mechanistic
principles, was developed under the NCHRP
(National Cooperative Highway Research Program)
[4]. The procedure is capable of developing
mechanistic-empirical design while accounting for
local environmental conditions, local materials, and
actual highway traffic distribution by means of axle
load spectra. Since the resulting procedure is very
sound and flexible and it considerably surpasses the
capabilities of any currently available pavement
design and analysis tools, it has been adopted by
AASHTO as the new AASHTO design method for
pavement structures.
The design method adopted in the Guide for
DAVID PUBLISHING
D
Flexible Pavement Design Simulation Using Mechanistic-Empirical Pavement Design Guide

1376
Mechanistic-Empirical Design of New and
Rehabilitated Pavement Structures is known as
MEPDG (Mechanistic-Empirical Pavement Design
Guide). In MEPDG, prediction of pavement response
and performance must take into account fundamental
properties of layer materials. Among fundamental
layer properties, the most important property of HMA
(hot-mix asphalt) is the dynamic modulus of asphalt
concrete. This property represents the temperature-
and time-dependent stiffness characteristics of the
HMA material. Significant amount of effort has been
devoted to develop a test protocol to determine the
dynamic modulus of HMA [5]. This effort has
resulted in a standard test protocol that can be used for
the Simple Performance Test for Superpave
Mix
Design [6]. This test protocol calls for the use of axial
compression testing for measuring the dynamic
modulus. One of the issues related to the dynamic
modulus is its use in forensic studies and pavement
rehabilitation design.
In the hierarchical design approach proposed in
MEPDG for new HMA pavements, dynamic moduli
measured in laboratory are required for the highest
design reliability (Level 1), which is intended for
pavements with very high traffic volumes. However,
dynamic modulus is used as the primary stiffness
property for HMA at all three levels of hierarchical
inputs in MEPDG. The dynamic modulus test is
relatively difficult and expensive to perform.
Therefore, numerous attempts have been made to
develop regression equations to estimate the dynamic
modulus from mixture volumetric properties. The
predictive equation developed by Witczak et al. [5] is
one of the most comprehensive mixture dynamic
modulus models available today that can predict the
dynamic modulus of dense-graded HMA mixtures
over a range of temperatures, rates of loading and
aging conditions. These inputs are available from
conventional binder tests and the volumetric
properties of the HMA mixture. A revised version of
this model has been recommended in the design of
intermediate- and low-volume roadways (design
Levels 2 and 3) in MEPDG [7].
The KDOT (Kansas Department of Transportation)
is currently considering adopting MEPDG to replace
the 1993 AASHTO design method that is in use now.
However, work is needed to determine whether
MEPDG gives results similar to that of the AASHTO
design method and/or predict the distresses that match
the measured/observed distresses.
2. Objectives
The main objectives of this study are to investigate:
(1) The AASHTO flexible pavement design method
in Kansas using MEPDG;
(2) The effect of subgrade modulus on the predicted
distresses using MEPDG;
(3) The effect of failure criteria and reliability while
using MEPDG.
3. Test Sections
Five newly built superpave pavements, designed
using the 1993 AASHTO Design Guide were selected
in this study. Each test section was 305 m long. Table 1

Table 1 Layer type and thickness.
Layer No. Layer type Material type
Thickness (cm)
US-54
Butler
US-77
Butler
US-283
Grah.
K-7
Don.
K-99
Elk
1 Surface SM-9.5A (PG 64-28) 3.8
a
3.8 3.8 3.8 3.8
2 Binder SM-19A (PG 64-28) 6.3 6.3 6.3 6.3 6.3
3 Base SM-19A (PG 64-22) 21.3 20.0 17.5 12.5 17.5
4 Aggregate base AB-3 N/A N/A N/A 27.5 N/A
5 Subgrade Modified subgrade 15.0
b
15.0
b
15.0
b
15.0
c
15.0
b

a
SM-9.5T PG 64-28;
b
Lime Treated Subgrade (LTSG);
c
fly ash modified subgrade (FASG).


1377

shows the layer type and thicknesses of these sections.
All pavement sections have Superpave 9.5 mm
nominal maximum aggregate size mixture (known as
SM-9.5A and SM-9.5T in Kansas) with PG 64-28
binder in the surface course of 3.8 cm thickness.
Layers 2 and 3 consist of fine graded Superpave 19
mm nominal maximum aggregate size mixture,
SM-19A with PG 64-28 and PG 64-22 binders,
respectively. The base layer thickness varies from
12.5 cm to 21.3 cm. The K-7 project in Doniphan
County has the thinnest asphalt base (12.5 cm) since it
also has 27.5 cm crushed stone base, designated as
AB-3 in Kansas. All projects have lime-treated
subgrade except K-7 in Doniphan County where
subgrade was modified with a Class C fly ash.
4. Data Collection
4.1 Deflection Data
Deflection data were collected approximately 8 to
10 weeks after construction. Multiple target loads
were used on most test sections. The target loads used
in the FWD (falling weight deflectometer) testing
were 40, 53 and/or 67 kN for all sections. Deflection
measurements were made in the outside wheel path of
the travel lane at 11 stations at 30.5 m intervals. The
geophone spacing was 0, 20, 30, 45, 60, 90 and 120
cm for US-54, US-77 and US-283. The last sensor
was located at 150 cm for the K-7 and K-99 projects.
4.2 Volumetric Properties
Most of the mixture data required have been
obtained from the mix design database of KDOT.
Information includes gradation of aggregates
(cumulative percent retained on 19 mm, 9.5 mm and
4.75 mm sieves and percent passing 0.075 mm sieve),
physical properties of the aggregates (bulk and
effective specific gravities), asphalt content and
asphalt specific gravity and theoretical maximum
specific gravity of the mixture.
Bulk specific gravities of the compacted samples
and the cores were determined in the laboratory
following the Kansas Standard Test Method KT-15,
Procedure III. KT-15 closely follows AASHTO T 166.
From these pieces of information, the air void (%),
effective binder content (% by volume), voids in the
mineral aggregates (VMA, %), and percent of VMA
filled with binder (%) were calculated. The original,
mix/lay-down, surface aging and aging at different
viscosities have been determined at different
temperature and frequency. Temperature data for
different locations have been obtained from KSU
(Kansas State University) weather data library.
5. Analysis Methodology
5.1 Back-Calculation of Modulus
FWD deflection data have been normalized to 40
kN load. These normalized deflection data were used
to back-calculate the pavement layer moduli based on
the multilayered elastic theory. The moduli of thin
surface layers or sandwiched layers are usually
difficult to obtain, because surface deflections are
often insensitive to changes in the moduli of these
layers. Changes in the moduli of subgrade or other
thick layers may mask changes in thin layers [8].
Flexible pavements are usually analyzed as
three-layered systems having an asphalt concrete
surface layer, a mechanically or chemically stabilized
base layer and a subgrade [9]. In this study, all
pavement sections were modeled as three layer
systems by combining all asphalt concrete layers into
one layer. Comparison of solutions from different
programs gives an idea of the range of solutions that
can be expected [8]. Thus, three back-calculation
computer programs, EVERCALC, MODCOMP 5,
and MODULUS, were used in this study. It is to be
noted that only back-calculated subgrade moduli have
been used in this study.
In the back-calculation of pavement layer moduli,
the objective is to identify a set of pavement layer
moduli that would produce a deflection basin
matching the measured deflection basin. Since only a
finite number of sensor data points are available from

1378
the deflection measurements, the objective function in
the back-calculation analysis typically involves the
minimization of the root-mean-square difference (D
rms
)
of the measured and computed deflections. A solution
that has a smaller D
rms
value is considered to be a
better fit, and thus a better solution [10].
=
|
|
.
|
\
|
=
m
i i
i i
rms
D
D d
m
D Minimize
1
1

(1)

where, m = number of deflection-measurement points,
d
i
= back-calculated deflection at point i, and D
i
=
measured deflection at point i.
5.2 Analysis Using MEPDG
MEPDG software was used to do the design
analysis. Four cases have been considered at level 3
using default distress targets. The existing pavement
structure was analyzed for a 10-year analysis period as
Case 1. Case 2 considered the maximum number of
years the existing pavement structure will be in a
serviceable condition. Minimum thickness of different
layers to serve for 10-years has been found by
ignoring and considering longitudinal cracking as
Cases 3 and 4, respectively.
Input data used for the MEPDG analysis of flexible
pavements are categorized as: general information,
site/projection identification, analysis parameters,
traffic, climate, pavement structures and
miscellaneous. The main ones are discussed below:
Analysis parameters: Flexible pavement design is
based on the surface-down and bottom-up fatigue
cracking of the asphalt surface, HMA thermal
cracking, fatigue cracking in chemically stabilized
layers, permanent deformation for both asphalt layers
and the whole pavement and roughness. Since there
are no stabilized layers in this study, fatigue cracking
in chemically stabilized layers is not applicable.
Default and modified criteria have been used in this
study. Distress targets for 10-year design life have
been changed for some of the distresses as shown in
Table 2. Reliability was also changed to 50% keeping
the default distress targets (not included in the table).
Traffic data: are one of the key elements required
for the design and analysis of pavement structures.
The basic required information are AADTT (annual
average daily truck traffic) for the base year, percent
trucks in the design direction, percent trucks in the
design lane, and operational speed of the vehicles.
Three functions are available to estimate future truck
traffic volumes: no growth, linear growth and
compound growth. Linear growth rate was used in this
study.
Project-specific linear traffic growth rates varied
from 0.9% to 1.7%. Directional and lane distribution
factor for trucks was taken as 60% and 100%,
respectively. Percent of trucks varies from 13% to
26% as indicated in Table 3. For this study, some
other required traffic inputs (hourly distribution factor,
monthly adjustment factor, axle load distribution
factor, number of axles per truck, axle spacing, etc.)
were derived from the MEPDG level 3 or default
values.
Climate: Environmental conditions have significant
effects on the performance of flexible pavements. The
seasonal damage and distress accumulation algorithms

Table 2 Performance criteria for the study.
Distress type
Criteria 1/default Criteria 2
Distress target Reliability level (%) Distress target Reliability level (%)
Terminal IRI (m/km) 2.6 90.0 2.6 90.0
Long. cracking (m/km) 189.4 90.0 94.7 90.0
Alligator (fatigue) cracking (%) 25.0 90.0 10.0 90.0
Transverse cracking (m/km) 189.4 90.0 94.7 90.0
AC permanent deformation (cm) 0.64 90.0 0.64 90.0
Total permanent deformation (cm) 1.91 90.0 1.91 90.0


1379
Table 3 Summary of traffic data.
Routes US-54 US-77 US-283 K-7 K-99
Initial two-way AADT 3,959 1,217 1,046 1,251 1,862
Percent of trucks 13 26 20 14 16
Linear growth rate (%) 1.5 1.4 0.9 1.4 1.7
Operational speed (km/hr) 110 95 105 90 90
No. of lanes in each direction 1 1 1 1 1
Directional distribution (%) 60 60 60 60 60
Lane distribution (%) 100 100 100 100 100

in the MEPDG design methodology require hourly
data for six weather parameters such as air
temperature, precipitation, wind speed, percentage
sunshine, relative humidity and seasonal or constant
water table depth at the project site [4]. The design
guide recommends that the weather inputs be obtained
from weather stations located near the project site. At
least 24 months of actual weather station data are
required for the computations. In this study, project
specific virtual weather stations were created by
interpolation of climatic data from the selected
physical weather stations.
Miscellaneous: The thermo-hydraulic properties
required as inputs into MEPDG software are
groundwater depth, infiltration and drainage
properties, physical/index properties, hydraulic
conductivity, thermal conductivity, heat capacity, etc..
The recommended calibrated values of 2.72 W/(m.K)
and 1.17 kJ/(kg.K) were used for thermal conductivity
and heat capacity, respectively. Physical and index
properties were derived based on the gradation of the
unbound materials. Surface shortwave absorptivity
and drainage path length were chosen based on the
default inputs, and was 0.85 m and 3.7 m, respectively.
All performance models used in this study are the
nationally calibrated default ones. Only the inputs
represent local conditions or projects in Kansas.
The existing pavement structures were analyzed for
a 10-year analysis period as Case 1. Case 2 considered
the maximum number of years the existing pavement
structure will be in a serviceable condition. Minimum
thickness of different layers to serve for 10-year has
been found by ignoring and considering longitudinal
cracking as Case 3 and 4, respectively. Default criteria
have been used for these cases. Effects of changing
subgrade modulus, target distress, and reliability have
also been presented
6.1 Case 1: 10-Year Analysis Period
The predicted distresses for the existing pavement
structures are shown in Table 4(a). These predicted
distresses are far less than the target distress limits.
The lowest and highest IRI (International Roughness
Index) is 1.36 m/km and 1.41 m/km, respectively.
Longitudinal and transverse crack is zero and 0.19
m/km, respectively, for all projects. The thinner the
pavement sections, the higher the AC and total
permanent deformation.
6.2 Maximum Design Life
Table 4(b) shows the maximum number of years
the existing pavement structure will be in a
serviceable condition. K-7 has the thinnest total AC
thickness and as a result, it has the least service period.
US-77 has the highest service period. The lowest and
highest IRI is observed on US-283 and US-77,
respectively. Insignificant amount of longitudinal
cracking is observed on only K-99. The smallest and
largest AC and total permanent deformation is
observed on US-77 and K-99, respectively.
6.3 Case 3: Minimum Thickness Ignoring
Longitudinal Cracking
The causes and effects of AC surface down cracking


1380

Table 4 Distress predicted for 10-year and maximum design life.

Predicted distress
Distress target
US-54 US-77 US-283 K-7 K-99
(a) 10-year design life
IRI (m/km) 1.37 1.36 1.38 1.41 1.40 2.59
Long. cracking (m/km) 0.00 0.00 0.00 0.00 0.00 189.39
Alligator (fatigue) cracking (%) 0.00 0.00 0.00 0.10 0.10 25.00
Transverse cracking (m/km) 0.19 0.19 0.19 0.19 0.19 189.39
AC permanent deformation (cm) 0.20 0.13 0.18 0.23 0.23 0.64
Total permanent deformation (cm) 0.73 0.70 0.78 0.83 0.85 1.91
(b) Maximum design life
Maximum design life (year) 24 25 23 22 23
IRI (m/km) 1.90 1.92 1.87 1.89 1.90 2.60
Long. cracking (m/km) 0.00 0.00 0.00 0.00 0.02 189.40
Alligator (fatigue) cracking (%) 0.10 0.00 0.10 0.10 0.40 25.00
Transverse cracking (m/km) 0.19 0.19 0.19 0.00 0.19 189.40
AC permanent deformation (cm) 0.30 0.20 0.25 0.33 0.35 0.64
Total permanent deformation (cm) 0.93 0.88 0.95 1.03 1.10 1.91

Table 5 Minimum layer thicknesses ignoring and considering longitudinal cracking.

Predicted distress
US-54 US-77 US-283 K-7 K-99
(a) Ignoring longitudinal cracking
Layer thickness
Surface (cm) 3.75 3.75 3.75 2.50 3.75
Binder (cm) 5.00 5.00 5.00 2.50 6.25
Base (cm) 5.00 2.50 2.50 2.50 5.00
AB3 (cm) - - - 27.50 -
Distresses
IRI (m/km) 1.58 1.61 1.59 1.61 1.56
Long. cracking (m/km) 158.90 191.29 181.63 11.95 806.80
Alligator (fatigue) cracking (%) 6.10 4.70 3.80 1.40 5.70
Transverse cracking (m/km) 0.19 0.19 0.19 0.19 0.19
AC permanent deformation (cm) 0.38 0.35 0.28 0.33 0.33
Total permanent deformation (cm) 1.35 1.50 1.45 1.48 1.30
(b) Considering longitudinal cracking
Layer thickness
Surface (cm) 3.75 3.75 3.75 3.75 3.75
Binder (cm) 5.00 5.00 6.25 6.25 6.25
Base (cm) 10.00 7.50 8.75 5.00 12.50
Total AC thickness (cm) 18.75 16.25 18.75 15.00 22.50
AB3 (cm) - - - 27.50 -
Distresses
IRI (m/km) 1.46 1.45 1.44 1.47 1.44
Alligator (fatigue) cracking (%) 0.70 0.40 0.20 0.40 0.50

(longitudinal cracking) are not well understood yet
and it was ignored (there has been failure in
longitudinal cracking) in this case to determine the
minimum thicknesses of different layers to serve for
10-year as shown in Table 5(a). The lowest to highest
total AC thickness to serve for 10-year is 7.5, 11.25,

1381
11.25, 13.75 and 15 cm for K-7, US-283, US-77,
US-54 and K-99, respectively. K-7 has the lowest
total AC thickness since it has 27.5 cm of AB-3. K-7
and US-54 has the lowest and highest alligator
cracking, respectively. US-283 and US-54 has the
lowest and highest AC permanent deformation
whereas K-99 and US-54 has the lowest and the
highest total permanent deformation, respectively.
6.4 Case 4: Minimum Thickness Considering
Longitudinal Cracking
Minimum layer thicknesses of different layers to
serve for a 10-year period considering the longitudinal
cracking are shown in Table 5(b). The lowest to
highest total AC thickness is 15, 16.25, 18.75, 18.75
and 22.5 cm for K-7, US-77, US-54, US-283 and
K-99, respectively. The lowest total AC thickness is
observed for K-7, which has 27.5 cm of AB-3. The
lowest IRI is observed on a pavement which has the
highest total AC thickness (K-99) and vice versa (K-7)
even though the difference is not significant from a
practical point of view. This shows that longitudinal
cracking does not depend on the thickness of AC
layers. Transverse cracking is constant for all projects.
US-283 has the lowest AC and total permanent
deformation.
6.5 Effect of Subgrade Modulus on Predicted Distress
Back-calculation programs, EVERCALC,
MODCOMP 5, and MODULUS, have been used to
back-calculate the subgrade modulus. One-third of the
back-calculated subgrade modulus was used as the
subgrade design modulus to run the MEPDG software
for a 10-year design analysis period. Default failure
criteria have been used for this analysis. The results
for different projects have been presented and
discussed below:
US-54: The design subgrade modulus varies from
14.63 MPa to 188.68 MPa for US-54 as indicated in
Table 6(a). Transverse and longitudinal cracking
remain the same at all subgrade moduli. IRI, alligator
cracking, and total permanent deformation decrease
with an increase in subgrade modulus whereas the AC
permanent deformation increases/remains constant
with an increase in subgrade modulus. No failure has
been observed at any subgrade modulus level;
US-77: Table 6(b) shows that the design subgrade
modulus for US-77 varies from 7.90 MPa to 66.89
MPa. Transverse and longitudinal cracking remain the
same at all subgrade moduli. IRI, alligator cracking
and total permanent deformation decrease with an
increase in subgrade modulus whereas AC permanent
deformation increases/remains constant with an
increase in subgrade modulus. There is a failure in
total pavement deformation when subgrade modulus
is 7.90 MPa. The result shows that the stronger the
subgrade, the higher the AC permanent deformation;
US-283: The design subgrade modulus varies from
7.76 MPa to 59.80 MPa for US-283 as indicated in
Table 6(c). Transverse and longitudinal cracking
remain the same at all subgrade modulus levels. IRI,
alligator cracking and total permanent deformation
decrease with an increase in subgrade modulus
whereas the AC permanent deformation
increases/remains constant with an increase in
subgrade modulus. There is failure due to total
pavement deformation when the subgrade modulus is
7.76 MPa;
K-7: The design subgrade modulus varies from
43.75 MPa to 69.37 MPa for K-7 as indicated in Table
6(d). Transverse cracking, longitudinal cracking and
AC permanent deformation remain the same at all
subgrade modulus. IRI, alligator cracking and total
permanent deformation decrease with an increase in
subgrade modulus. There is no failure at any subgrade
modulus level;
K-99: For K-99, the design subgrade modulus
varies from 45.29 MPa to 69.92 MPa as indicated in
Table 6(e). Transverse cracking remains constant at
all subgrade modulus. IRI, alligator cracking and AC
permanent deformation decrease with an increase in
subgrade modulus whereas longitudinal cracking and


1382

Table 6 Effect of subgrade modulus on predicted distresses.
Predicted distress
(a) US-54
Design subgrade modulus (MPa) 14.63 25.56 28.34 35.59 188.68
IRI (m/km) 1.56 1.46 1.44 1.41 1.30
Alligator cracking (%) 0.20 0.10 0.10 0.10 0.00
(b) US-77
Design subgrade modulus (MPa) 7.90 14.19 21.18 66.89
IRI (m/km) 1.72 1.56 1.48 1.33
Alligator Cracking (%) 0.10 0.10 0.00 0.00
Transverse Cracking (m/km) 0.19 0.19 0.19 0.19
Total permanent deformation (cm) 2.10* 1.48 1.15 0.60
(c) US-283
Design subgrade modulus (MPa) 7.76 18.73 45.35 53.74 59.80
IRI (m/km) 1.79 1.54 1.39 1.38 1.40
Alligator cracking (%) 0.10 0.10 0.00 0.00 0.00
Total permanent deformation (cm) 2.38* 1.38 0.83 0.75 0.80
(d) K-7
Design subgrade modulus (MPa) 43.75 52.32 64.50 69.37
IRI (m/km) 1.43 1.42 1.40 1.39
Alligator cracking (%) 0.10 0.10 0.00 0.00
Transverse cracking (m/km) 0.19 0.19 0.19 0.19
Total permanent deformation (cm) 0.93 0.85 0.78 0.75
(e) K-99
Design subgrade modulus (MPa) 45.29 46.29 69.92
IRI (m/km) 1.42 1.42 1.38
Long. cracking (m/km) 0.00 0.00 0.02
Alligator Cracking (%) 0.20 0.20 0.10
Transverse Cracking (m/km) 0.19 0.19 0.19
AC permanent deformation (cm) 0.23 0.23 0.25
Total permanent deformation (cm) 0.93 0.93 0.75
*indicates failure.

total permanent deformation decrease with an increase in the subgrade modulus. There is no failure at any

1383
subgrade modulus level.
6.6 Effect of Reliability
The effect of reliability has been investigated by
keeping the distress target at default values and
changing the reliability to 50%. All projects failed in
total permanent deformation and there has not been
any increase in maximum service life due to the
change in reliability under existing traffic condition.
Higher traffic values have been used to investigate the
effect of reliability and maximum service life is higher
using 50% reliability than that of 90%. The results
have not been included since the results are the same
under the existing conditions.
6.7 Effect of Distress Target
The effect of distress targets on the maximum
service of projects in this study has been investigated
using Criteria 2 in Table 2. There has not been
a change in the maximum design life due to the
changes in distress targets under the existing
conditions.
7. Conclusions
Based on this study, the following conclusions have
been made:
MEPDG design analysis of flexible pavements
designed using 1993 AASHTO design guide does not
show the default distresses currently observed in
Kansas in the 10-year design period. By the end of
10-year design period, over 50% of pavements have
had some form of structural rehabilitation from
KDOTs experience.
The MEPDG analysis results also show that the
thinner the pavement sections, the higher the asphalt
concrete layer and total permanent deformation.
The existing pavement structures can serve for
more than 20 years as per the MEPDG design analysis
if the nationally calibrated models and default failure
criteria are used. This is contrary to the current Kansas
experience that shows that by year 20 nearly all
pavements have had some kind of structural
rehabilitation.
Total AC thickness varies from 7.5 cm to15 cm for
a 10-year design period if the effect of AC surface
down cracking (longitudinal cracking) is ignored. The
lowest thickness is observed on K-7, which has 27.5
cm of AB-3.
The minimum total AC thickness to serve for
10-year period considering the longitudinal cracking
varies from 15 cm to 22.5 cm. The lowest IRI is
observed on a pavement that has the highest total AC
thickness and vice versa. Longitudinal cracking does
not depend on the thickness of AC layers.
Back-calculated subgrade moduli obtained from
various back-calculation programs result in variable
predicted distresses for different projects.
Distress target and reliability have not shown
significant effects on the maximum service life of the
existing pavements. This needs to be verified by
performance observation of in-service pavements.
References
[1] Guide for Design of Pavement Structures, AASHTO
(American Association of State Highway and
Transportation Officials), Washington D.C., 1986.
[2] Guide for Design of Pavement Structures, AASHTO
(American Association of State Highway and
Transportation Officials), Washington D.C., 1993.
[3] K.T. Hall, State of the Art and Practice in Rigid
Pavement Design, Transportation in the New Millennium,
Transportation Research Board, Washington, D.C., 2000.
[4] Guide for Mechanistic-Empirical Design of New and
Rehabilitated Pavement Structures, Final NCHRP report
for Project 1-37A, Transportation Research Board,
Washington, D.C., 2004.
[5] M.W. Witczak, T.K. Pellinen, M.M. El-Basyouny,
Pursuit of the simple performance test for asphalt
concrete fracture/cracking, Journal of the Association of
Asphalt Paving Technologists 71 (2002) 767-778.
[6] Simple Performance Test for Superpave Mix Design,
NCHRP report 465, Transportation Research Board,
[7] Interim User Manual for the M-E Pavement Design
Guide, NCHRP, Transportation Research Board,
[8] Y.J. Chou, R.L. Lytton, Accuracy and consistency of

1384
back-calculated pavement layer moduli, Transportation
Research Record: Journal of the TRB 1022 (1991) 1-7.
[9] R.W. Meier, D.R. Alexander, R.B. Freeman, Using
artificial neural networks as a forward approach to
back-calculation, Transportation Research Record:
Journal of the TRB 1570 (1997) 126-133.
[10] T.F. Fwa, C.Y. Tan, W.T. Chan, Back-calculation
analysis of pavement-layer moduli using genetic
algorithms, Transportation Research Record: Journal of
the TRB 1570 (1997) 134-142.


Critical Factors in Financing and Management of Light
Rail Transit Systems: Delphi Methodology Approach
J ulin Sastre Gonzlez
1
and Clara Zamorano Martn
2
1. Royal School of Engineering, University of Seville, Sevilla 41007, Spain
2. Civil Engineering Department, Foundation Caminos de Hierro, Madrid 28002, Spain

Abstracts: In recent years, a large number of LRT (light rail transit) networks and tramways have been built in Spain using the
tender system. The rhythm of construction has been such that very few analyses have been carried out and those that have were only
partially done because the main participants involved had ulterior motives which impeded their having a critical and independent
opinion of the projects carried out. Within this framework, the Delphi methodology presents itself as a very useful instrument, as the
intrinsic advantages that it has added to the characteristic that the experts who participate in the project can remain anonymous
without detracting value from the conclusions obtained as these are grouped together according to their worth.

Key words: Public-private partnership financing, Delphi, public transport, LRT, management, concessions.

1. Introduction

Congestion in cities and new ways of tackling
mobility issues, principally on the part of the citizens
themselves, are making different public bodies aware
that it is worth favoring the development of urban rail
network systems and metros, specifically LRT (light
rail transit)
1
and reserved platform systems. Since the
construction of the first modern tramway, put into
service in Spain in 1994, ten more tramway systems or
light rail networks have been built, and at the current
time, there are more than 20 under study or in
construction. These figures can be extrapolated to
Europe as a whole, where there are currently the region
of 200 tramway networks of which the majority are
located in the 15 member states which made up the
European Union on the May 1, 2004. The development
expectations of this system and secular financing on the
part of the Public Administrations make us think of

Corresponding author: J ulian Sastre Gonzlez, Dr., M.Sc.,
Ph.D., research fields: public-private partnership, railways,
transport planning and sustainable mobility. E-mail:
juliansastre@juliansastre.com.
1
In this paper, LRT or LR and tramways (we refer to modern
tramways networks) are used in the same sense as in Spain that
has been decided for each case for political/marketing reasons
[1].
major development of these systems using the BOT
(built, operate and transfer) method. For these systems
to run in more efficiently, it would appear obvious to
take into account critical factors regarding their
financing. The pace of development with which this
system has expanded, as regards light rail networks,
and the lack of, in some cases voluntary, spreading of
information about these, means that the same mistakes
are being repeated systematically from one project to
another. A recent study carried out in Spain [2],
allowed for the compilation of extensive information
on 22 Spanish case-studies [3, 4], as well as others
from different parts of Europe, analyzing variables of
supply, demand and cost to which was added a detailed
study on nine projects following a laborious
interviewing process. As can be seen in Table 1, 90%
of these had problems during the construction and
operational phases, amongst which unforeseen events
during construction stands out, There are also other
factors to be taken into account, such as an increase in
investment cost, as can be observed in Fig. 1, which
affected a large number of the projects. Over-costing in
the operational phase is the problem which has the least
effect on infrastructures as a whole, perhaps in part

DAVID PUBLISHING
D
Critical Factors in Financing and Management of Light Rail
Transit Systems: Delphi Methodology Approach

1386

Table 1 Comparative analysis of the solutions of underground or overland metro.
Comparative analysis of solutions for underground or overland metro systems
Underground metro system
Advantages Other remarks
Does not interfere with the road network nor reduce its capacity
Does not affect traffic
Greater commercial speed
Less impact on surface building work except when problems involving
subsidence or archaeological difficulties occur
Total times for each mode of transport can be
greater or lesser depending on each case and
require, just as demand assessment does, specific
studies Increased capacity
Increased road safety
Improved initial image which helps social support
Overland metro
Advantages Other remarks
Increased accessibility
Possibility of more stops
Notably cheaper to build and operate
Greater acceptance on the part of users in terms of comfort sensation as they
can enjoy the local environment whilst travelling without producing the feeling
of claustrophobia
On visual or aesthetic impact is also debatable.
Overland systems, at the moment, use catenaries
(although it must be mentioned that the recent
Bordeaux overland did not use overhead cabling)
though often are carried out providing a complete
urbanisation of the environment, from faade to
faade and giving back the city or improving
some areas of this
Facilitates the implantation of restrictive measures for private traffic to relieve
congestion in the city centre
Shorter building work duration in theory (although obtaining residence permits
in public spaces may cause troublesome delays)
Easier to spread
Greater security against terrorist attacks and emergencies (security)
Source: In-house based on diverse interviews.

because they had only recently been put into
service.
A highly advisable way of using past experiences to
the benefit of the construction of new projects is via the
use of perspective, specifically by the application of the
Delphi method [5].
The Delphi method is included in those techniques
known as expert methods which, unlike the
extrapolative or correlation methods, require specific
data on past experiences, based on consulting people
who have vast knowledge on the subject in hand. This
method is usually used when one of the following
conditions occurs:
(1) There is no historical data with which to work. A
typical case in this situation is the provision of new
technologies. One must take into account that the
projects in place on the Iberian Peninsula have all been
recently completed;
(2) The impact of external factors has a greater
influence in the evolution than the internal ones. In this
way, the appearance of favourable regulatory
legislation (this is the case in Spain with the
concessions on Infrastructures Law of 2003 and the
Regulations of Public Service Contracting of
2007), and the support on the part of some agents
toward certain technologies or systems can cause
major development in these areas when, without
which, it would have occurred at a much slower
pace;
(3) The ethical, moral, social and political conditions
dominate over the economical and technological ones
in an evolutionary process. In this case, one type of
technology can see its development hindered if the
technology itself is rejected by society, or stimulated
by political acceptance as in seen in the case of the light
rail networks. They are, also, urban projects of major
political impact.
The aim that has been achieved by the application
of the Delphi method is to establish the key
factors for financing a new tramway or light-rail
network based on the opinion of qualified
specialists [6].

1387

Fig. 1 Main problems encountered.

2. Methodology
The methodology has been carried out in two stages.
The first, qualitative, in which experts were requested
to provide their reasoning with respect to the opinions
presented and later quantitative analysis was carried
out in which the experts were asked to answer survey
about the main parameters.
The first stage was performed by the use of personal
interview which lasted for around 2 h and were held
over 22 months depending on when the experts were
available.
The second stage was done by sending a survey
containing the parameters by e-mail, and through this
system the results were returned in a period of about
two weeks.
For preparation of the first stage, a basic
questionnaire was designed based around the sectors
main issues and added to the results obtained from the
analysis of twenty-two case studies from Sastres 2009
study [7], as mentioned beforehand.
The second step was the selection of the expert panel.
The basis of the idea was to get opinions from a wide
spectrum of specialists, who responded to the group of
agents involved in this type of project. A basic
requirement was that they had had a high level of
authentic responsibility in several projects of the type
under study. Fifteen specialists from Spain and
Portugal have taken part, and these can be grouped into
four categories:
7 heads of public departments;
3 from the financial sector;
3 concessionary partners;
2 sector researchers.
After carrying out the interview analysis, the a priori
conclusions were established and the questionnaire was
designed which would be the basis for the second phase
of the study. The results of both phases allow to extract
the researchs findings.
The practical way of applying the method has been
set out in Fig. 2.
3. Phase I: Qualitative
ConsiderationsLessons Learnt
The points of interest which are obtained from the
first phase analysis of the study are the
following:

0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
C
o
n
s
t
r
u
c
t
i
o
n

d
e
l
a
y
s

P
r
o
b
l
e
m
s

i
n

t
h
e

o
p
e
r
a
t
o
r

-

c
o
n
s
t
r
u
c
t
o
r

r
e
l
a
t
i
o
n
s
h
i
p

U
n
a
c
c
p
e
t
a
b
l
e

q
u
a
l
i
t
y

I
n
c
o
m
e

l
e
v
e
l

b
e
l
o
w

f
o
r
e
c
a
s
t

r
a
t
e

O
v
e
r
-
c
o
s
t
i
n
g

i
n

t
h
e

O
&
M

p
h
a
s
e

D
e
m
a
n
d

l
e
v
e
l

b
e
l
o
w

f
o
r
e
c
a
s
t

r
a
t
e

I
n
c
r
e
a
s
e
d

i
n
v
e
s
t
m
e
n
t

c
o
s
t

E
x
t
r
a

c
h
a
r
g
e
s

o
f

o
&
M
M


1388

Fig. 2 Applied Delphi method.
Source: In-house.

3.1 Interest Caused by Light Rail Networks
From the point of view of its contribution as a
transport system, the utility of light rail networks are
centred on two scenarios:
As a main means of transport in medium-sized cities
of 300,000 to 500,000 inhabitants where the urban
infrastructure allows for this and transport is carried out
largely overland. The minimum suggested level is
2,000 passengers per kilometre of track and a
maximum of 5,000 passengers per hour in each
direction.
In larger cities, acting as a complement to the
conventional metropolitan system.
The light rail network is characterised by greater
accessibility [7] than a conventional underground
metro, this means greater acceptance on the part of
senior citizens and less time between line transfers. Yet
its most outstanding plus, which makes it especially
interesting, is its use as a means of renovation in the
city or area. In short, it is a project which impulses
urban development in all senses of the word.
Nonetheless, it would seem clear the importance of
acting on the existing bus network:
Beforehand, in order to take advantage of the
capacity that this modes of transport has as well as its
efficiency;
Once the light rail network is put into service,
integrating both networks and remodelling the bus
network so that this becomes a complement on the
route and not competition.
Finally, the pros and cons of this system versus a
conventional underground metro network are
described in Table 1.
3.2 Cost ranges for a Spanish Project
The main findings from the first phase regarding cost
issues are:
(1) investment: 10-15 M/km in comparison with
30-35 M/km of conventional underground metro;
(2) development: 8 /km-9 /km (for bus routes
1.3-1.7 /km) on out of town routes 3.5-4 /km on city
routes). Rolling stock: 2.5 M/unit;
(3) distribution between operation and maintenance:
55-60/45-40%;
(4) start-up investment, the proportion with respect
to the total is situated around:
civil 55%-65%;
Analysis
Design of a guide for the initial interviews
Selection of an expert panel
Carrying out interviews
Carrying out
contacts
Research of contact
data
Interview analysis
Initial conclusions and definition of critical issues
Questionnaire design based on preliminary issues
Development
Carrying out the survey
Contact and sending of
questionnaire by e-mail
Writing of conclusions

1389
systems 15%-20%;
rolling stock: 20%-25%.
3.3 Barriers, Critical Aspects and Keys for Success
About these three issues, with the interviews, very
interesting lessons learnt can be highlight as follows.
A transport authority is an important element for
agreements between Public Administrations and
allows for questions to be answered such as who will
pay? How will the financial responsibilities be divided
both for investment and well as operation? It also
facilitates the conciliation between technical criteria
with the political ones.
A guaranteed minimum demand must exist from the
beginning irrespective of what can be expected with
future developments so that a concession can be put
into service with private initiative.
As has already been mentioned, the LRT can be seen
as a regeneration plan for the city so this can boost it
towards success. There must be ambition implicit in the
project (not only in terms of transport) coupled with
financial realism.
Legal and financial problems have been easily
resolved in Spain in recent years, though these factors
could become critical within the framework of the
current economic downturn.
Light rail projects are more complex, due to their
immersion into the urban environment, than
conventional underground metros and less flexible than
the bus network (with this modes of transport, if a
mistake is made in the design phase it is possible to
remodel this as thesupport, the road has more
branches).
They must be planned and co-ordinated with
sufficient time so that they become the project and
does not end up being a project with clearly-defined
aims. It is not good when we are dealing with a project
merely linked to an electoral campaign, or a legislative
period as this is called in Spain.
A problem that light rail networks face is the
construction phase, as this is truly conflictive due to the
inconvenience caused to residents, and, above all, the
businesses that see their level of trade notably reduced
during some time.
It is very useful for the Public Administrations that is
boosting the project to assume the leadership role in it,
although always with wide-ranging agreements
between all the competent bodies. This means, a Public
Administration, or responsible body which is capable
of pushing the project on without forgetting that this
must be co-ordinated with the parties involved.
The implication of local companies, without this
being compulsory, it usually of benefit as they wish to
remain and they want to fit in with the environment.
It is important to include bonus as well as
penalizations in the contracts.
The composition of the partners in the tender group
or consortium is debatable:
The constructors tend to disappear during the
operational phases (when the structure or the
partnership is flexible) yet from then on they do not
assume the same level of responsibility for faults which
can occur;
The operators are often absent in the structure
during construction, yet, these are the ones who will
occupy the space in the long term;
Other agents are unsure if they should form part of
the company structure or not: financers, consultancies,
rolling stock manufacturers.
Signalling priority and a good commercial speed are
key factors for the success of the projects as are the
quality of service and on-board cancellation. Another
critical issue is the decision to include underground
sections or not.
In terms of design, an issue which is becoming of
increasing importance is the catenary, due to the
demands that these have a reduced visual impact, and
in general, as a result of concern for urban aesthetics, as
was mentioned at the beginning.
Communication and public engagement in all phases
with the social stakeholders involved is another essential
task to be included in the project development.

1390
3.4 Financing and Management System.
Regarding financing and management issues, it can
be remarked as follows.
It is not clear whether the construction should be
included in the concession bids, experts generally
concur that, at the end of the day, this does not result in
an earlier setup of the projects, and investment costs
could go over-budget. This can only be justified if:
There is a high opportunity cost, either because of
public deficit or a lack of capacity to manage the
project. On the other hand, as the social benefits are
greater in relation to the investment, then the launch of
the project must be brought forward as much as
possible;
The contract and legal framework allow for the
adequate control of the concession and its costs.
In a period of between 20 and 25 years of concession
is insufficient to cover investment, a residual value can
be established for a transfer under the proper conditions
though it is not advisable to lengthen the concession
periods.
Under an operational tender, the timeframe must be
of between 6 and 10 years. An intermediate alternative
is the Porto formula which has a much reduced
deadline towards being put into service in less time, to
build capacity with and the lessons learnt by the Public
Administration.
The DBOT (BOT including design) tenders (as
established under Article 22 of the Spanish Tenders
Law) can provide a solution worthy of contemplation
in order to take advantage of the capacity of private
initiative. Yet this opinion is not fully shared by many
experts who tend to believe that the citizens needs are
better understood by the Public Administration. This is
a system which was implanted by the Highways
Agency in the United Kingdom, tackling aims such as:
Minimising the contribution required and
optimising the level of riskv [8] assumed by the public
sector;
Ensuring that the projected routes were designed,
maintained and operated satisfactorily to maximise any
adverse impact on the environment thus maximising
the benefits felt by the users;
Promoting innovation, not only in terms of
technical and operational aspects, but also in terms of
financial and commercial organisation;
Fulfilling the government aim of encouraging the
development of a private-sector for United Kingdom
motorways.
All groups agree that the importance of effective
operation for the success of the project, as the service is
the final aim and its effect is that which will be felt for
the entire life-cycle of the project.
Private operation can have three advantages,
although they do not always give good results:
Avoiding union interference;
Optimising the experience of the private company
when it is not present in the Public Administration;
Reducing activities to those which are strictly
necessary in order to avoid inefficient or excessive
expenses, limiting political interference.
An integrator is necessary in the projects, the experts
recommend an engineering firm or even a rolling stock
manufacturer.
The Public Administration must be responsible at all
times for the concession, supervising it and checking
the contract is fulfilled.
The importance of financial advisors seems clear as
they provide a good source of information on demand
and costs.
The BOT Model has the advantage of simplicity and
financial postponement, but the reality in Spain means
that this is complex due to lack of transport bidders
with clear interests in the business, coupled with the
financial capacity to see this through. On the other hand,
although it seems that these are going to change, the
traditional contracting mechanisms do not work
properly, they are rigid, enclosed and do not allow for
fair renegotiation based on incomplete contracts with
gross costs or similar formulae.
For example, the new Spanish concessions law
foresees competitive dialogue, via the method of

1391
incomplete contracts, though one must taken into
account that this causes for time to be spent in contract
negotiations. Some experts contend that there must be
the possibility of provisional pricing and a system of
cost-plus which prevents wear and tear and
over-costing during constructions as changes always
occur. The presence of mediators, for example, would
be a possible solution. An example of current trends in
Spain can be seen with the Granada Metro, for which
the Andalusian Railways Agency will invite tenders for
and contract the project (up to five or six
sub-contractors) and operation will be hired via another
tender, perhaps using the gross-cost with incentives
method. The rolling stock depends on what decisions
are made regarding:
transfer system;
financial equity;
contract length.
Though in the case of the company responsible for
the Tenerife Metropolitan Railway, just 20% was
sufficient because the ticketing income is greater than
50% (commercial income) of the production costs,
which is a necessary condition (not enough), to follow
EU (European Union) rules for funding. The EU
requires that the majority share of a mixed company
(public and private sector) be private as to avoid being
considered part of public deficit. That said, though in
the case of the company responsible for the Tenerife
Metropolitan Railway [9], just 20% was sufficient
because the ticketing income is greater than 50%
(commercial income) of the production costs, which is
a necessary condition (not enough), to follow EU rules
for funding.
An interesting example is that of the last section of
Line 8 of the Madrid Barajas Airport Metro.
Investment and infrastructure management were
private, but the transport service operator is from the
public-sector. This means, investment does not figure
in the public deficit, yet the control and quality
standards of the operation are maintained.
The assessment of the British experience is
interesting in so far as it considers what is known as
value for money, put more simply, profit returns on
public money. When a PPP is undertaken in England, a
Public Sector Comparator is carried out: How much
would it cost if it were done publically or privately? To
do this, an important database is required. A
private-sector company, in theory, provides fewer
delays, as it would be a turnkey project and delays
mean reductions in income. This has been closely
studied in the United Kingdom. The new Spanish Law
for Public Service Contracts of 2007 has already
contemplated this. Though one must take into account
that the United Kingdom is not subject to SEC-95, (EU
rule for public accountings). As well, in Spain, the aim
is that the public accounts are not affected.
The opportunity cost of a BOT is based on the loss of
social benefits due to the delay which could be
produced in a project due to lack of public funds to
finish it and because the repayment of the Public
Administrations contributions are made via company
tax.
3.5 The Process of Tender and Awarding
Tender arrangements are considered by experts as a
critical subject.
Unfeasible requirements must not be (excessively
strict punctuality indicators for surface transport
systems, for example).
Operational proposal must be properly evaluated.
Good base documentation is necessary, above all, in
the planning stages.
Documents must be clear and detailed according to
the majority of the experts, and the quality of the
contracts is a key issue for the proper functioning of the
project as mentioned in other studies, such as Roy
Yvrande-Billion [10].
The progress clause considered in the Spanish legal
framework (that requires that the concessionaire must
adapt the system to the advances over the time period
of the concessions is important although its application
causes uncertainty.

1392
Bid evaluations may and must be carried out by
specialists, who should even be independent. It is
important to analyse every aspect of these bids to
understand every minor detail.
It must be demanded that the offers contain the key
aspects referred to in the awarding criteria in a clear
and concise way in the conditions document in order to
achieve an easier comparison of proposals. Aspects
which would be considered of interest to include in this
documentation are:
Cost per km and cost per passenger valued at
current prices and within different timeframes;
Service offer (timetables, headways, etc.).
Owner recommendations from the expert to enrich
future tenders are as following: Though the Public
Administration must have an important negotiation
capacity and this demands that the Public
Administrations must be highly competent. If not, it is
necessary to seek out an external consultant as is
recommended in different multilateral guides for
tender. This could be a freelancer or company with
experience in the sector. This option could cost
relatively little comparing to the investment. The need
is for someone to identify where the potential problems
are as following:
In terms of the awareness barometers, these must
be done in such a way that, for example, on a scale of 1
to 10, all these items score at least 7, as, for example,
the compensatory aspects of the practices in force at the
moment mean that an unbalanced proposal with good
and bad aspects can win. This should not be the case;
The financial element is not only the economic
value of the construction and/or the operation of the
line, rather the delivery conditions and risks. This is
more difficult to measure though very important.
Therefore, the financial aspect is of great importance
yet, all of its facets must be taken into account.
Looking at the Seville Metro [11] case where the
winning bidder was the one which offered a lower
investment cost, around 30% less, and the true cost has
practically doubled, this shows that giving the contract
to the cheapest brings with it an important risk as the
project was not as well defined as with other bids.
4. Phase IIResults Tabulation
With regards to the results obtained from the
previous phase, seven essential points were established
for which a system of parameters depending on fixed
limits. The issues approached were as follows.
4.1 What Is the Minimum Daily Demand Level Which
Would Justify a LRT?
As can be seen in Fig. 3, the majority of the experts
consider that a daily demand level of 20,000
passengers is sufficient, and this coincides with the
idea expressed in the interviews that light rail networks

Fig. 3 Minimum daily demand level.
Source: In-house based on Delphi survey.
Minimum daily demand

are largely a
tackling a m
could be han
platforms [1
is an oppor
significant p
network.
4.2 Which T
BOT?
In most ca
are key an
contrasts w
Consultancie
have as an
engineering
analysed ten
percentages,
4.3 Which F
Recommend
Rail Networ
The exper
two main an
DBOT,

Fig. 4 Types
Source: In-hou

a city project
mobility issue
ndled by the
12], however,
rtunity to act
project such
Type of Partn
ases, the expe
nd essential
with the cu
es, as can be
important r
agencies a
ndency for th
, as can be ap
Financing and
d in General f
rk?
rts, according
nd extreme op
, in the opinio
s of partners in
use.
C
R
e
s
p
o
n
s
e

p
e
r
c
e
n
t
a
g
e

Critical Fa
Transit
and not excl
e. These leve
use of bus ro
, many expert
t in the city
as a tramw
ners Are Mor
erts consider t
partners, so
urrent situa
seen in Fig. 4
role in the c
as the partn
heir participat
ppreciated in T
d Managemen
for the Develo
g to Fig. 5, are
ptions:
on of some e
n a BOT.
No
Concesionar
Consult
Financ
Operato
Construct
actors in Fina
t Systems: De
lusively a wa
els of passen
outes on reser
ts consider th
and carry o
way or light
re Important
that the opera
omething w
ation in Sp
4, do not seem
concessions f
ers, despite
tion with redu
Table 2.
nt Model Do
opment of a L
e divided betw
experts allows
0
one
rio
tor
cial
ors
tor
T
ancing and Ma
elphi Method
ay of
ngers
rved
hat it
out a
rail
in a
ators
which
pain.
m to
from
the
uced
You
Light
ween
s for
the
from
leve
ope
mos
proj
4.4
Priv
Proj
O
dive
que
Fig
of
priv
Ref
indu
[5]
4.5
a LR
E
ther
10
Typ
Types of BOT
anagement o
ology Appro
private-secto
m the initial
el of efficienc
Public cons
erational conc
st efficient in
jects in Spain
What Moti
vate-Sector P
oject?
Once again t
erse possibil
estionnaire ha
. 6. This mea
consideration
vate-sector p
f. [13]. On the
ustry which c
is considered
Who Should
RT Project?
Expert opinio
re is a divisio
20 3
pe of patner
T patners
of Light Rail
ach
or company
stages, as w
cy whilst wor
struction via
cession, which
n financial term
n.
ives Do You
Participation
the experts a
ities, all the
ave been eva
ans that there
n which pu
participation,
e other hand,
can compete
d interesting b
Carry out th
ons indicate t
on between t
30 40
rs
to be able t
well as ensuri
rk is being ca
independent
h is the soluti
ms for recent
u Consider
in a Light R
are divided
factors sugg
aluated, as ca
are many rea
ut forward
as were sin
the creation o
in the world
but secondary
he Constructio
that, if we lo
those who bel

0
1393
to contribute
ing a certain
arried out;
contract and
ion seen to be
tly completed
Can Justify
Rail Network
between the
gested in the
an be seen in
asons worthy
the idea of
ngled out in
of a powerful
dwide market
y.
on Project in
ook at Fig. 7,
lieve that the
3
e
n
d
e
d
y
k
e
e
n
y
f
n
l
t
n
,

1394
Table 2 Fun
Functions of t
transport syste
Participants
Constructor gr
Rolling stock
manufacturers
Operators
Financers and
Consultancies
Source: In-hou

Fig. 5 Recom
Source: In-hou

construction
Administrati
bidder shoul
nctions of the d
the different age
em
Role
roups
Partn
Cons
s
If the
Supp
Their
Cons
They
d banks
They
They
They
s
These
Winn
Build
The P
They
Bid;
Proje
Proje
Oper
Syste
Their
their
reduc
use based on De
mmendable fin
use based on De
n project shou
ion, and those
ld do this so
M
o
d
e
l
s
Critical Fa
Transit
different agent
ents who norma
ner in the conce
struction and ev
ey are partners,
pliers, being abl
r presence as a p
structor groups p
y represent the f
y provide capita
y often have littl
y can participate
e can work for
ning bidder;
ders Consortium
Public Adminis
y can participate
ect;
ect Managemen
ation assessmen
ems integrator.
r role is still sub
continuity in th
ce the cost of fin
elphi survey.
nancing and m
elphi survey.
uld be carried
e who think th
that it is they
PublicSector
Recomm
actors in Fina
t Systems: De
s.
ally form part o
ssionare genera
ven installations
supply of the ro
e to form part o
partner is not co
power make the
final aim of the
al and financial
le involvement
e just as advisor
the:
m;
stration.
e in the:
nt: quality contro
nt;
bject to great de
he process and s
nancing.
management mo
out by the Pu
hat the succes
y who assum
constructor,priv
Mixed
mendedman
ligh
ancing and Ma
elphi Method
of a bidding com
ally with leader
s
olling stock and
of the supply gr
ommon in Spai
em seem like su
process, the ful
guarantees
in the creative
rs for the financ
ol;
ebate, and in an
shares out the f
odels for a ligh
ublic
ssful
me all
the
answ
ade
0
BOT
DBOT
Public.
vatesector
dCompany
Others
Any
nagement
htrailnetwo
anagement o
ology Appro
mpany for the c
rship through a
d responsibility
roup of the conc
in, though guara
ubcontractors in
lfilment of a se
process and bu
cing design
ny case, it is ofte
financial burden
ht rail network
construction
wer given f
equate model.
10 2
Respons
Fundingm
orks
of Light Rail
ach
construction and
bidding compa
y for its implem
cessionare with
antees experien
n Spain
rvice
usiness design
en with a reduc
n between the v
k.
n risks. This
for question
.
20 30
epercentage
modelfor
d development
any of the Const
mentation
h the constructor
nce and long-ter
ced percentage w
various parties w

vision conc
n C regardin
40
of an urban
tructor Group
r
rm implication
which ensures
which can
curs with the
ng the most
e
t

1395

Fig. 6 Private-sector participation motives.

Fig. 7 Construction project in a BOT case.

Fig. 8 Concession period.

0 10 20 30 40 50
PORCENTAGE RESPONSE
Less than 5 years
From 5 to 10 years
10 to 30 years
30 to 40 years
Others
Variable
CONCESSION PERIOD
0.0 10.0 20.0 30.0 40.0 50.0
Response percentage
Administration
Constructor
Concesionaire
Operators
Any
Construction project in a BOT case
Response percentage

1396

Fig. 9 Economic offer score.

Fig. 10 Bidding evaluation.

4.6 What Tender Timeframe Do You Consider
Generally Appropriate?
The majority, as can be seen in Fig. 8, choose a
longer timeframe, like those used in Spanish projects,
of between 30 and 40 years, which use up the
maximum amount of time permitted by the 2003
Tenders Law. Shorter timeframes and variable
timeframes are not highly regarded mainly when
construction investment is included in the tender.
4.7 What Percentage Score Should the Economic
Criteria Must Have in a Tender?
In their majority, the experts, as we can observe
in Fig. 9, consider that this must not exceed
50% though that this must have a notable
importance, around 30%-50% generally. The
current terms of reference fulfil this criteria
suggested by the expert panel
(Fig. 10).
0 20 40 60 80
Others
Less than10%
10 - 30%
30 -50 %
More than 50%
% Economic offer score
Technical-execution
evaluation
Technical-development
evaluation
Funding-economic
offer evaluation
Economic-financial
solvency evaluation

1397
5. Conclusions
In short, the analysis of the answers provided by the
expert panel allows us to affirm that:
The light rail networks can have a major impact if
they are developed as city projects;
The involvement of private-sector companies can
provide expertise to the projects and allow for an
earlier opening of the service, yet it is necessary to
prepare this carefully and foresee adequate monitoring
in all observational phases of the project;
The development of light rail network projects
based on construction via public budgets and
operational tenders is an option which could integrate
the advantages of the different models and allow for the
versatility necessary in many cities, as long as the
ruling Public Administration has the funds required to
do this.
Acknowledgments
The authors are grateful for the funding provided by
the Madrid Regional Transport Consortium (PTA of
Madrid) to carry out this wide-ranging research project,
focusing on the analysis of the relationship between
private financing and type of project. We would
especially like to express our appreciation to Jess
Rodrguez Molina and Jos Manuel Pradillo, Technical
Director and General Manager of the Madrid Regional
Transport Consortium.
References
[1] Manual for Planning, Financing and Implementation of
Urban and Metropolitan Public Transport Systems,
Madrid Transport Consortium, June 2004.
[2] J. Sastre Gonzlez, Light rail network concessions: Public
and private-sector participation, comparison of
management and financing models, Doctoral Thesis,
Universidad Politcnica de Madrid, 2009.
[3] Conditions Documents for the Parla Tranway Tender,
Madrid Regional Transport Consortium, 2005.
[4] Conditions Documentation for the Contract Reduction of
the Project, Construction and Development of the New
Rail Line between Mostoles Central and Navalcarnero,
Ente Pblico de Derecho Pblico, Infraestructuras del
Transporte, Madrid, 2007.
[5] A. Lpez del Corral, A. Snchez Solio, Economic
analysis of the reform of the tender system for public
works, Revista de Obras Pblicas 149 (3425) (2002)
45-51.
[6] K.C. Green, J.S. Armstrong, A. Graefe, Methods to elicit
forecasts from groups Delphi and prediction markets
compared, Foresight: The International Journal of Applied
Forecasting 8 (2007) 17-20.
[7] Manual for Tramways, Light Railways and Reserved
Platforms: Design, Project, Financing and Implementation,
Madrid Transport Consortium, 2006, p. 279.
[8] J.E. Schaufelberger, I. Wipadapisut, Alternate financing
strategies for build-operate-transfer projects, Journal of
Construction Engineering and Management 129 (2003) (2)
205-213.
[9] Conditions Documents Which Must Be Adhered to during
Tenders for the Selection of Light Rail Network Operators
Participating in the Equity of Tenerife Metro Public
Company Ltd., Cabildo de Tenerife, 2002.
[10] W. Roy, A. Yvrande-Billion, Ownership, contractual
practices and technical efficiency: The case of urban
public transport in France, Journal of Transport
Economics and Policy 41 (2) (2007) 257-282.
[11] Documentation for the Tender Conditions for Bidders on
the Construction and Development of the Inter-Urban Line
1 Seville Metro in Which Services beyond the Scope of the
Project and the Acquisition of Rolling Stock Are Included,
Ministry of Public Works and Transport, Andalusian
Government, 2001.
[12] Documentation for the Tender Conditions for Bidders on
the Construction and Development of the Inter-Urban
Lines 1 and 2 of the Malaga Metro in Which Services
beyond the Scope of the Project and the Acquisition of
Rolling Stock Are Included, Ministry of Public Works and
Transport, Andalusian Government, 2003.
[13] C. Clements, M. OMahony, Traditional or public private
partnerships for procurement of light rail transit?, in: TRB
Annual Meeting, Washington, DC, 2004.


Determination of the Scale of Damage in the Wastewater
Tunnels Due to Earthquake: A Case Study
Turgay ogun
1
, aatay Turgut
2
, Baris Sayin
3
and Ahmet ogun
4

1. Department of Civil Engineering, Istanbul University, Istanbul 34320, Turkey
2. Institute of Science, Istanbul University, Istanbul 34452, Turkey
3. Department of Construction and Technical Affairs, Istanbul University, Istanbul 34116, Turkey
4. Vocational School of Technical Sciences, Akdeniz University, Antalya 07058, Turkey

Abstract: The pre-determination of the effect of earthquake in subsurface structures is gaining importance increasingly. One of the
main factors in determination of the damages due to earthquakes in subsurface structures, especially tunnels are horizontal
acceleration value. The aim of the study is to put forward the scale of damage due to earthquake in a wastewater tunnel in Istanbul,
the most populated city of Turkey, under construction. Possible damage caused by earthquake will be determined by utilizing the
information about the route of the wastewater tunnel analyzed in the study.

Key words: Subsurface structures, tunnels, earthquake damage scale.

1. Introduction

All the structures used to maintain the life of
humankind are defined as infrastructures vital for
places where all necessities for societies such as water,
energy, fuel and communication are produced,
transported and utilized [1]. Damages occurring in
infrastructures during an earthquake extremely affect
the daily life. Apart from the significant economic
losses, such cases may put the public health at risk.
For this reason, in underground structures,
predetermination of the possible damages of the
earthquakes becomes important increasingly.
Horizontal ground acceleration value is one of the
main factors in determination of the damages in the
underground structures especially damages in tunnels
caused by earthquakes.
When the damages occurring in the underground
structures are examined, it can be seen that they are
mainly caused by landslides and faulting [1]. The
earthquake, by itself, rarely causes damages in these

Corresponding author: Baris Sayin, Ph.D., research fields:
earthquake and reinforced concrete structures. E-mail:
barsayin@istanbul.edu.tr.
structures. Tunnel can be blocked by damages, which
are formed in the frames of the tunnel as a result of
the land sliding. The main points in the tunnels at
which the damages are concentrated are crossroads,
parts at which direction and shape changes exist and
parts at which the construction material or the type of
the ground varies. In tunnels with inner lining, the
damages are typically limited as cracks on the lining.
In the study conducted by Owen and Scholl for
tunnels [2], if the ground acceleration is lower than
0.4 g, then the damage in the hard rock tunnels is very
limited. Up to 0.19 g, damage did not occur in any
lined or unlined tunnel. Between 0.19 g and 0.4 g,
small cracks appear and some bricks or stones fell off
from the lining. Collapse of the tunnel was not
observed in the tunnels up to 0.5 g values. It was
stated that serious damages in the inner lining and
tunnel area are the results of the low quality inner
lining and inferior ground conditions. The seismic
dangers that could be observed in the wastewater
tunnels can be grouped as temporary and permanent
landslides that may occur in the ground. While the
temporary ones through seismic waves, permanent
DAVID PUBLISHING
D
Determination of the Scale of Damage in the Wastewater Tunnels Due to Earthquake: A Case Study

1399
landslides can be formed due to different reasons such
as ground faulting, liquefaction, landslide and
settlements [3].
Under the seismic motions, underground structures
demonstrate three types of deformation, namely axial
tension and compression, longitudinal bending and
rounding/bending [4]. The axial deformations in the
underground structures form as a result of the
compressive and tensile stresses, which are
originating from parallel motions in the axis of the
structure, caused by seismic waves. Longitudinal
bending, on the other hand, forms as a result of the
perpendicular particle motions, originated by seismic
waves, along the axis of the structure. It was depicted
that in rounding/bending, shear waves propagate
perpendicular or near perpendicular to the axis of the
tunnel and this yields bending of the cross section of
the tunnel [5].
In the study, the degree of the damage, that could
occur in a wastewater tunnel, which is being built in
Istanbul, the most populated city of Turkey, is
presented. Within this scope, by using the information
about the route at which the tunnel will pass through
(geotechnical report, waste water application plan,
lithological cross-section, cross-section of the RC and
its properties), it was examined if a damage would
occur as a result of an earthquake.
2. Damages Occur in the Infrastructures as a
Result of Earthquakes
Yzgll and Uurlu [6] conducted a study to
determine the damage levels of the wastewater and
water supply networks present in Istanbul
Metropolitan district. The results obtained in the study
are given in Table 1. The total numbers of pipe break
in the wastewater network, including all other
counties, were given as 1959. It was stated that in 940
of them damages occurred due to liquefaction.
3. The Geology of the Study Area
In the study area including close neighborhood,
Basic Mass formations belonging to Paleozoic Era,
i.e., 500 to 300 million years ago, are present. This
mass consists of different formations from
Ordovician, Silurian, Devonian and Lower
Carboniferous Periods as well as from two
granodioritic massive grown into these formations. It
covers the large areas on the coastal region of
Marmara Sea, both sides of the Bosphorus, the
Islands of Istanbul and Gebze in Kocaeli Provience
[7]. Generally in the research area, the regions called
as floor (natural, alluvium, marine sedimentary) or
main rock with high and low amount of cracks,
fractures, faults and folds cut by limestone, shale,
mudstone-limestone, andesite and diabase dikes were
passed on the valley-bay passages close to the sea.
The geomechanical properties of the rock used in
constructed tunnel and the tunnel route are given in
Table 2 and Fig. 1, respectively.
The final cross and longitudinal sections of the
wastewater tunnel, which will be opened through the
studied route is presented in Figs. 2 and 3.
3.1 Underground Water Observations
Table 3 presents the depths of underground water
according to the underground water observations

Table 1 The total numbers of pipe break in the wastewater network [6].
County Location of damage Number of damage Cause of damage Level of damage
Bakrky Wastewater outline 216 Pipe fracture Heavy
Pendik Wastewater outline 123 Pipe fracture Normal
Kadky Wastewater outline 178 Pipe fracture Normal
Kartal Wastewater outline 163 Pipe fracture Normal
Tuzla Wastewater outline 148 Pipe fracture Normal
Adalar Wastewater outline - - Normal
Avclar-Beykoz Wastewater outline - - Heavy


1400
Table 2 Geomechanic properties of the rock units [7].
Units Density (kN/m
3
) Porosity
Non-confined compression
test (MPa)
Youngs modulus
(GPa)
Poisson ratio
Tensile stress
(MPa)
Limestone 25.91-26.45 0.89-2.73 28-40 42-65 0.30-0.31 2.6-5.20
Andesite 4.60 20 0.22 1.32-5.39

Fig. 1 The wastewater tunnel route [7].

Fig. 2 Longitudinal section view of the wastewater tunnel.
Scale 1:5,000

1401

Fig. 3 Cross section view of the wastewater tunnel.

Table 3 The depths of the underground water.
Boring No.
Depth of ground
water (m)
Ground water
grade (m)
S1 6.50 +2.00
S2 3.50 +3.00
S3 6.40 +30.10
S4 7.00 +0.50

Table 4 Technical properties of EPB/slurry TBM AVN
2200 AB [8].
Diameter of the cutter head 2.775 m
Power on the cutter head 160 kW (217.6 HP)
Mechanical efficiency 85%
Pipe-Jacking main station and Interjack
Force (Max.) 500 Bar
13,600 kN
Outer diameter of the pipe installed 2.7 m
Cross-section of tunnel face 6.048 m
2

carried out in the research area [7].
3.2 The Properties of TBM (Tunnel Boring Machine)
The technical information of the TBM and pipe
jacking used in the underground water tunnel under
investigation is given in Table 4.
4. Determination of the Level of the Damage
Originating from an Earthquake for the
Wastewater Tunnel in Question
In the underground structures, especially in the
tunnels, one of the main factors used in
determination of the damages originating from an
earthquake is the maximum horizontal ground value
[8]. Firstly, this value would be determined by using
different acceleration reduction equations. The level
of the damage, on the other hand, would be identified
by using the curves on horizontal round
acceleration-tunnel damages given in the literature.
According to the ground investigation report [7],
prepared for the tunnel in question, 4,300 m of the
ground of the tunnel was determined as basic rock.
Only 250 m of the ground is natural ground and/or
artificial ground. The investigations point out that the

1402
ground base of the ground is A type and the class of
the local ground is Z1. It is advised that periods of
the spectrum characteristics should be taken as T
A
=
0.10 and T
B
= 0.30. For the valley passage, the base
of the ground belongs to group D and the class of the
local ground is Z4. For this, it is considered that
spectrum characteristics values of T
A
= 0.20 and T
B
=
0.90 are suitable. For the weakest part of the tunnel,
calculations in determination of the level of the
damage are carried out as follows. The vertical
distance of the wastewater tunnel to the fault, D is
equal to 48 km. The depth of the tunnel axis,
H was taken as 30 m for the horizontal
distance of 4,190 m. By considering these
assumptions:
(1) If the maximum horizontal ground acceleration
value is calculated by using the equation given in Ref.
[9]:

) (log 2 1
2 2
2
331 . 0 141 . 0
) 82 . 14 ( log 1177 . 1
) 6 ( 079 . 0 ) 6 ( 503 . 0 287 . 3 log
y
a
w w y
P Z Z
D
M M a
o + + +
+ +
+ =
(1)
Considering the conditions of the ground, Z1 and
Z2

were taken as 1 and 0, respectively. Meanwhile, for
84% safety factor, P was taken as 1. If moment
magnitude, M
w
is taken 7.4.
By using these assumptions;
2
1
2 2
2
/ 16 . 119 076 . 2 log
1 141 . 0 ) 82 . 14 ( 48 log 1177 . 1
) 6 4 . 7 ( 079 . 0 ) 6 4 . 7 ( 503 . 0 287 . 3 log
sn cm a a
a
y y
y
= =
+ +
+ =

a
y1
= 119 / 981 = 0.122 g
(2) If maximum horizontal ground acceleration
value is calculated by using the equations given by
Arolu and Ylmaz [10];
o + + +
+ + =
P Z C Z C
R M a
Y Y S S
s y
log 049 . 1 214 . 0 080 . 2 log
(2)
km R h D R
f
547 . 48 27 . 7 48
2 2 2 2
= + = + =

Surface-wave magnitude, M
s
= 7.4, V
s30
360 m/s
in soft grounds and if ground factor is taken as Z
y
= 1,
C
y
= 0.085;
( ) 1 085 . 0 547 . 48 log 049 . 1 4 . 7 214 . 0 080 . 2 log + + =
y
a

It is found that log a
y2
= 1.9798 a
y2
= 95.46
cm/s
2
= 0.097 g.
(3) When another equation given by Arolu and
Ylmaz [10] is used:
s
w y
V D
M a
log 371 . 0 ) 57 . 5 log( 779 . 0
229 . 0 312 . 0 log
5 . 0 2 2
+
+ =
(3)
including V
s
= wave propagation velocity;
= =
36 . 0 36 . 0
6 6 . 107 6 . 107 N V
s

V
s
= 205.1 m/s is found.
If M
w
= 7.4, D = 48 km, V
s
=

205.1 m/s values are
put into Eq. (3), maximum horizontal acceleration
value,
1 . 205 log 371 . 0 ) 57 . 5 48 log( 779 . 0
4 . 7 229 . 0 312 . 0 log
5 . 0 2 2
+
+ =
y
a

8577 . 0 3119 . 1 4 . 7 229 . 0 312 . 0 log + =
y
a
If log a
y
=

0.7870, then it was calculated that a
y3
=
0.163 g.
(4) If the same calculation was performed by using
Eq. (4):
) (ln ) 1112 / ln( 200 . 0 ln 899 . 0
) 6 ( 107 . 0 ) 6 ( 576 . 0 393 . 0 ln
30
2
ay P V R
M M a
s
w w y
o +
+ =

(4)
If D = 48 km and h
f
= 6.91 km are taken,
km R h D R
f
5 . 48 91 . 6 48
2 2 2 2
= + = + = .

) 1112 / 1 . 205 ln( 200 . 0 ) 5 . 48 ln( 899 . 0
) 6 4 . 7 ( 107 . 0 ) 6 4 . 7 ( 576 . 0 393 . 0 ln
2

+ =
y
a

338 . 0 4895 . 3 20972 . 0 8064 . 0 393 . 0 ln + =
y
a

It is found that ln a
y
=

2.1617 a
y4
= 0.115 g.
(5) If the arithmetic average value of this maximum
horizontal acceleration is calculated,
( )
y y1 y2 y3 y4
a a a a a = + + +
= 0.122 g + 0.0970 g + 0.163 g + 0.115 g) / 4 = 0.124 g
By considering a
y
= 0.124 g, H = 30 m and 50%
safety factor, examination of the maximum horizontal
ground acceleration-depth relation given in Fig. 4
yields that light damages can occur in the wastewater
tunnel, in question.


1403

Fig. 4 Maximum ground horizontal acceleration (measured above the ground/calculated)depth of tunnel damage relation
[11].

If the Fig. 4 is examined, it can be clearly seen that
the level of damage caused by the earthquake
decreases with increasing depth of the tunnel.
5. Conclusions
The knowledge of the behavior of the existing
underground structures is highly important to take
precautions for the possible damages. It is vital to
know the response of these structures against
earthquake effects so that the possible damage levels
can be predicted in advance. After an earthquake, not
only damages affecting the daily life but also damages
resulting in significant economic losses and conditions
that can threat the public health can be minimized.
The depth at which the tunnel construction is
carried out, the properties of the ground and the
method of tunnel construction are factors affective to
the level of the damage. For this reason, in the project
design and application stages of such constructions,
these points must be taken in consideration. In order
to reduce the seismic damages in wastewater tunnels,
it is necessary to take precautions such as the use of
mobile junctions, isolation of the tunnel line from the
ground motions etc..
References
[1] . Yzgll, G. Birgren, The behavior of the
infrastructure in earthquakes and damages formed, in:
Fifth National Conference on Earthquake Engineering,
Istanbul, Turkey, May 26-30, 2003.
[2] G.N. Owen, R.E. Scholl, Earthquake Engineering of
Large Underground Structures, Report no.
FHWA/RD-80/195, U.S. Department of Transportation,
Federal Highway Administration and National Science
Foundation, 1981.
[3] T. aatay, The structural behavior of the wastewater
structures under the effect of earthquake and an example
study, M.Sc. Thesis, Istanbul University, Institute of
Science, Istanbul, 2010.
[4] E.C. Cceoglu, The seismic design of the underground
structures, M.Sc. Thesis, Institute of Science and
Technology, Istanbul, 2006.
[5] C.M. John, T.F. Zahrah, A seismic design of underground
structures, Tunnelling and Underground Space
Technology 2 (2) (1987) 165-197.
[6] . Yzgll, K. Uurlu, The seismic performance of the
Istanbuls wastewater and water network, in: Earthquake
Symposium, Kocaeli, Turkey, Mar. 23-25, 2003, pp.
1376-1383.
[7] Genson Ltd. Co., ISKI Additional Geotechnical report,
Report No. 2003/196, IST, 2003.
[8] A. Kahriman, T. ogun, M. Akgl, Experts report:
2006-30, Istanbul, 2006.
[9] C. zbey, A. Sar, A. Manuel, M. Erdik, Y. Fahjan,
Empirical strong ground motion attenuation relations for
0.1 1 10 100 1,000 10,000
P
r
e
d
i
c
t
e
d

m
a
x
i
m
u
m

g
r
o
u
n
d

a
c
c
e
l
e
r
a
t
i
o
n
,

g

(
%
)


1404
northwest Turkey, in: Fifth National Conference on
Earthquake Engineering, Istanbul, Turkey, May 26-30,
2003.
[10] E. Arolu, A.O. Ylmaz, Seismic analysis of the tunnel
gallery with solved problems, Technical Thesis, Istanbul
Branch of Turkish Chamber of Mining Engineers,
2006.
[11] S. Sharma, W.R. Judd, Underground opening damage
from earthquakes, Engineering Geology 30 (3-4) (1991)
263-276.


Probabilistic Seismic Hazard Assessment for
Bridgetown-Barbados, Employing Subduction Interface
Characteristic Earthquakes
Walter Salazar
1
, Lyndon Brown
2
and Garth Mannette
1

1. Seismic Research Centre, The University of the West Indies (UWI), St. Augustine, Trinidad and Tobago
2. Earthquake Unit, The University of the West Indies (UWI), Mona, Kingston, Jamaica

Abstract: A new probabilistic seismic hazard analysis was performed for the city of Bridgetown, Barbados, West Indies. Hazard
computations have been performed using the standard Cornell-McGuire approach based on the definition of appropriate seismogenic
sources and expected maximum magnitudes, the authors take into consideration the possibility of large subduction interface
earthquakes of magnitude 8.0-9.0 beneath the Barbados accretionary prism via application of a characteristic model and slip rates.
The analysis has been conducted using a standard logic-tree approach. Uniform hazard spectra have been calculated for the 5% of
critical damping and the horizontal component of ground motion for rock site conditions setting 5 return periods (95, 475, 975, 2,475
and 4,975 years) and spectral accelerations for 34 structural periods ranging from 0 to 3 s. The disaggregation results suggest that the
magnitude-distance pair that dominates the hazard yields M 7.4 and 8.6 and a distance of 42.5 km in the Interface Subduction Zone
beneath Barbados for the 475 and 975 years RP (return period), respectively. An event with an M 8.0 at a distance of 107.5 km in the
Intraplate Subduction Zone is the second scenario that dominates the hazard for both 475 and 975 years RP.

Key words: Tectonics, seismogenic sources, characteristic earthquakes, recurrence interval, disaggregation.

1. Introduction

The present work aims to assess the PSHA
(probabilistic seismic hazard) [1, 2] at Bridgetown,
Barbados (Fig. 1) in terms of PGA (peak ground
acceleration) and response spectra for 5% of critical
damping in conjunction with a disaggregation scheme
to obtain the magnitude-distance pair that contributes
most to the hazard. Recent studies have computed the
seismic hazard for the whole Eastern Caribbean
including the Island of Barbados [3].
This work attempts to overcome the shortcomings
encountered in previous works: The island of Barbados
differs geologically from other islands in the Eastern
Caribbean for its evolution in an accretionary prism
rather than a magmatic origin arc. Barbados lies just

Corresponding author: Walter Salazar, doctor of
engineering, research fellow, research fields: earthquake
engineering and engineering seismology. E-mail:
walter.salazar@uwiseismic.com.
above the inclined interface subduction zone between
the descending Atlantic oceanic lithosphere of the
American Plate and the overriding Caribbean Plate
(Fig. 2). Despite large interface thrust earthquakes have
been relatively sparse since historic times in the

Fig. 1 Bridgetown, Barbados location, the seismic hazard
is calculated for the coordinates 13.099
o
N, 59.613
o
W (see
red star).
DAVID PUBLISHING
D
Probabilistic Seismic Hazard Assessment for Bridgetown-Barbados, Employing
Subduction Interface Characteristic Earthquakes

1406

Fig. 2 The Lesser Antilles region, (a) Bathymetric map. 1Volcanic Caribbees; 2Limestone Caribbees; 3axis of the
inner arc; 4axis of the outer arc; 5deformation front. Isobaths in m; after Ref. [4]. The authors model the subduction
interface zone SZ3 beneath Barbados as an inclined plane dipping to the west with an angle of 10
o
delimited by the
coordinates 10.71
o
-14.8
o
N and 57.529
o
-60
o
W, with the trench located at 57.529 W and a maximum depth of 50 km at 60
o
W.
The authors also model the subduction interface zone SZ2 in a similar way for the northern part of the Eastern Caribbean.
Reproduction of the figure with the permission of Geoff Wadge and Trevor Jackson.

northern and southern part of the Eastern Caribbean,
the worldwide seismic activity trend suggests that large
megathrust events of magnitude 8.0-9.0 are possible in
regions of long periods of quiescence, clear examples
are the recent Sumatra and Japan Earthquakes in 2004
and 2011, respectively. In this regard, the authors
employed a characteristic model for the Eastern
Caribbean subduction zone and introduced in our
tectonic model and hazard calculations the inclined
geometry of the slab beneath Barbados based on
detailed tectonic and geological information.
The first section of the work for this paper is devoted
SZ3
SZ2

1407
to explaining in detail the tectonic environment and the
seismicity evaluation for the Eastern Caribbean,
including a subsection of the chosen recurrence time
for the characteristic earthquakes for the interface
subduction zones in accordance with the observed slip
rates in the region. Secondly, the authors present the
GMPEs (Ground Motion Prediction Equations)
adopted to compute the hazard those that are
compatible with the tectonics of the region. Thirdly, the
PSHA is computed via application of a logic tree
formulation in terms of PGA and elastic response
spectra, a subsection of the PSHA is presented in terms
of the disaggregation process to find out the
magnitude-distance pair yielding the largest
contribution to the hazard at Bridgetown.
2. Seismotectonic Setting and Geometrical
Delimitations
The authors employed the earthquake catalogue and
the Gutenberg Richter relationships (Table 1)
developed by Bozzoni et al. [3]. Fig. 3 illustrates the
geometrical configuration for both, the shallow and
deep zones covering the subduction, upper-crustal
volcanic island-arc, transform and intra-plate faulting
and transitions zones used for the hazard calculation.
This configuration comprises the subduction trench to
the east, and the deepest part of the Atlantic Plate to the
west. The geometrical delimitation for shallow
seismicity in the arc includes the islands, related
epicenters, and main geological structures such as
volcanoes and seismic faults.
In order to evaluate the seismicity in the region, we
established 15 SZs (seismogenic zones) that affect the
study area based on detailed tectonics and geological
features that already been studied by different authors
(Figs. 2 and 3), as follows.
2.1 Zone 1: Volcanic Island-Arc
The upper-crustal seismicity is concentrated within
the upper 35 km of the Caribbean continental plate in
the Lesser Antilles Arc [5], with epicenters plotting
from the islands of Grenada to Anguilla within a nearly
continuous belt of 100 km width along both, the axis of
the principal active volcanoes and the inland and

Table 1 Main features of the SZs (seismogenic sources) and Gutenberg-Richter Parameters (log N = a bM) for the Eastern
Caribbean [3]. It is noted that the depth set for the seismogenic zones SZ2 and SZ3 are variable in order to model the inclined
slab beneath Barbados. M
max
corresponds to the maximum observed magnitude in the catalogue + increment of 0.5 units for
SZ4-SZ11, SZ13-SZ15, an increment of 0.3 has been adopted for SZ1. For SZ12 M
max
is based on new paleoseismological and
slip rates data [6, 7]. * M
max
used as a characteristic value for SZ2 and SZ3 (see Section 3 and the forth figure).
Seismogenic
zone
a b M
max
Depth (km) Type Main focal mechanism
SZ1 4.794 -1.012 6.9 19.1 Volcanic Normal and strike-slip
SZ2 4.614 -0.893
8.0
>8.0-9.0*
0-50 Interface Thrust (inverse)
SZ3 3.216 -0.725
8.0
>8.0-9.0*
0-50 Interface Thrust (inverse)
SZ4 4.164 -0.821 8.5 86.0 Intraplate Normal
SZ6 4.724 -0.941 8 32.3 Interface Thrust and strike-slip
SZ7 3.043 -0.705 8 28.4 Shallow Normal
SZ9 2.961 -0.727 6.9 24.4 Transition Normal and strike-slip
SZ10 2.127 -0.531 7.2 43.9 Transition/Intraplate Normal and strike-slip
SZ12 4.100 -0.949 7.7 32.5 Crustal Normal and strike-slip
SZ13 3.392 -0.747 7.2 23.3 Crustal Strike slip and thrust

1408
Fig. 3 Seism
amongst the z
hazard calcul

offshore sh
Subduction
are moderate
through hist
north of Ma
Islands, and
activity than
of the Less
Lucia). Volc
300 years oc
the Soufriere
Mt. Pel
Guadeloupe
Montserrat v
earthquakes
with volcan
clusters with
Bernard
evaluation
account thes
the ones that
(5.5-6.0 M
w
Probabi
mic source delim
zones. The alte
lations of this w
hallow faults
Trench. With
e, reaching a
toric times. Z
artinique to A
d we characte
n the Windwa
ser Antilles A
canic disaster
ccurred during
e in St. Vince
in Martini
(1976 to 197
volcanoes. Ho
do not nece
nic eruptions
h no discernib
and Lambe
of seismic h
se shallow to
t occurred on
w
), March 16
listic Seismic
Subdu
mitation for th
ernate geomet
work.
which run
hin this zone
a maximum v
one 1 compri
Anguilla cover
erize it with
ard Islands at
Arc (from G
rs in the regi
g major explo
ent (1718, 18
ique (1902)
77) and from 1
owever, the m
essarily occu
s and frequ
ble mainshock
ert [8] sugg
hazard must
moderate ear
1851 and 189
6, 1985 (6.4
c Hazard Ass
ction Interfac
he Eastern Car
trical delimitat
parallel to
e, the magnitu
value of about
ises an area f
ring the Leew
a higher seis
the southern
Grenada to S
ion over the
osive eruption
812, 1902, 19
), Soufriere
1995 to presen
moderate shal
ur in conjunc
uently appear
k (swarms).
gested that
t also take
rthquakes suc
97 in Guadelo
M
w
) at south
sessment for
ce Characteri
ribbean [3]. T
tions for SZ2 a
the
udes
t 6.6
from
ward
smic
part
Saint
past
ns in
979),
in
nt in
llow
ction
r in
the
into
ch as
oupe
h of
Nev
21,
the
zon
mec
obs
the
cata
20 k
197
and
cha
2.2
T
Eas
plat
Plat
gen
(acc
incl
eart
Bridgetown-B
stic Earthqua
he 3D view of
and SZ3 are p
vis, and the e
2004 (6.3 M
w
Les Saintes I
ne yield bo
chanisms. A
erved in the
other zones
alogue lists o
km) in the W
72 (4.5 M
w
) an
d Saint Lucia
aracteristic of
Zone 2-5: Su
The volcanic
stern Caribbe
te begins to
te reaching d
nerating earth
cording to th
lude in Zon
thquakes (de
Barbados, Em
akes
f the SZs is sho
presented in Fi
earthquake th
w
) in the north
Islands. The f
oth, normal
marked low
Windward I
of the Easter
only two uppe
indward Islan
nd May 19, 1
respectively,
f the Lesser A
ubduction in t
island-arc lie
an Trench, w
submerge u
depths of 20
hquakes as la
he historical
nes 2 and 3
epth 50
mploying

own to illustra
ig. 2 and used
hat occurred o
h-west of Do
fault plane sol
l and strik
wer level of
Islands as co
rn Caribbean
er-crustal eve
nds dated on
1990 (4.7 M
w
, confirming t
Antilles Arc so
the Lesser-An
es about 300
where the Nor
underneath th
00 km below
arge as magni
catalogue).
3 all the sh
km) along
ate the overlap
in the seismic
on November
minique near
lutions in this
ke-slip focal
seismicity is
ompared with
n, the seismic
ents (depth <
September 8,
w
) on Grenada
the quiescent
outh region.
ntilles
km from the
rth American
he Caribbean
w the islands
itude 8.0 M
w
The authors
hallow focus
the inclined
p
c
r
r
s
l
s
h
c
<
,
a
t
e
n
n
s
w

s
s
d

1409
inter-face seismic zone that yields underthrust focal
mechanisms [9]. Convergence between the Caribbean
and North American plates occurs at a rate of about 2
cm/yr [10].
As an alternative geometrical delimitation for the
interface subduction zones presented in Fig. 3, the
subduction zone SZ3 is modeled as an inclined plane
dipping to the west with an angle of 10
o
delimited by
the coordinates 10.71
o
-14.8
o
N and 57.529
o
-60
o
W,
with the trench located at 57.529
o
W with a maximum
depth of 50 km at 60
o
W. Similarly, the authors also
model the SZ2 interface subduction zone as an inclined
plane dipping to the west (Fig. 2). It is noted that the
previous tectonic model assumed a fixed depth for the
interface subduction zone. Due to the proximity of
Barbados to the interface subduction zone, we consider
it is very important to improve the tectonic model by
including a more realistic geometry of the inclined slab
beneath the island in the hazard calculations.
The focal mechanisms of deeper intra-plate events (>
50 km) indicate that there is normal faulting resulting
from initial flexure of the down going Atlantic slab
(Zone 4 and 5) with an average westward dipping angle
of 50 [11]. Zones 2 and 4 cover the latitudes from
14.8 N to 20.0 N and they are characterized with a
higher seismic activity than Zone 3 and 5 (from 11.0 N
to 14.8 N latitude). Bengoubou-Valeruis et al. [11] and
Russo et al. [12] attribute the differences in the seismic
activity to the following reasons: (1) changes in the
tectonic structures mapped by Feuillet et al. [13]; (2)
there being enough sediments to lubricate or decouple
the two plates in the subduction zone; (3) strengthening
caused by thick accretionary prism overburden which
lies above the shallow reach of the subduction zone.
The quiescent area coincides with the deepest part of
the Barbados accretionary wedge. The upper-crustal
seismic activity level observed along the
volcanic-island arc also reflects also the differences
observed in the seismic activity in the subduction zone.
The largest interface event listed in the catalogue
dates October 10, 1974 (7.3 M
w
) with its epicenter
located between Antigua and Barbuda (Zone 2). The
largest intra-plate earthquakes within Zone 4 occurred
on February 8, 1843 and on April 5, 1690, with a
magnitude of 8.0 and 7.5 (M
w
), respectively, with both
epicenters located to the west of Antigua and Barbuda
islands. Other intraplate big events occurred in Zone 5
around Martinique on January 11, 1839, December 3,
1906, and November 29, 2007 with magnitudes 7.3-7.4
(M
w
).
2.3 Zone 6-8: Puerto Rico and Virgin Islands
The Puerto Rico and the Virgin Islands region is
considered as a microplate that is surrounded by the
obliquely subducting North American plate, the
Caribbean Plate and several major faults such as the
Mona Canyon to the east and Abnegada Passage to the
west [14, 15] and the Muertos Trough to the south.
Puerto Rico accommodates approximately 16.9 mm/yr
of deformation relative to North America and 2.4
mm/yr relative to the Caribbean Plate [16]. The area is
divided into three seismogenic sources as follows:
Zone 6: This zone includes the Puerto Rico Trench
area with a depth less than 50 km including the
megathurst faulting along the plate interface of the
subducting North American Plate southward
deepening. This zone also comprises the left lateral
strike slip faulting that is subparallel to the Puerto Rico
trench north and north-west of Puerto Rico including
the Septentrional fault. On July 29, 1943 an earthquake
ruptured the Puerto Trench with a magnitude of 7.5
(M
w
). This seismogenic zone covers the north of Puerto
Rico and the Virgin Islands.
Zone 7: This zone comprises the shallow faults (less
than 50 km depth) inland Puerto Rico and offshore
namely, Mona Canyon, South Lajas Fault, Great
Northern and Southern Puerto Rico Fault Zone, the
Anegada Trough and Sombrero Seismic Zone [16].
This seismogenic source has produced earthquakes of
magnitude 7.5 and 7.3 (M
w
) in the Anegada and Mona
Passage in 1867 and 1918, respectively, yielding
normal faulting in a broad zone of active crustal

1410
extension and accompanied by destructive tsunamis
[17]. The absence of volcanism in Puerto Rico and the
Virgin Islands suggests that this zone is not an
extension of the island-arc Lesser Antilles structure
[18].
Zone 8: This zone includes the intra-plate
subduction seismicity generated by the bending of the
North-American slab with depths greater than 50 km.
Moreover, recent research suggests the existence of the
subducted Caribbean slab confirmed by low velocity
anomalies beneath the island [19]. This seismogenic
area comprises the subduction intra-plates slabs of
North America southward dipping and the Caribbean
northward dipping beneath the microplate. The largest
earthquake listed in the catalogue within this zone is
dated on March 24, 1916 (7.4 M
w
).
The Muertos trough offshore southern Puerto Rico
constitutes the thrust-trench locus convergence
between the microplate and the Caribbean Plate
northward deepening [19]. This seismic zone is
excluded from our analysis since the rates of activity in
this zone are poorly known [17] and it seems that based
on the knowledge of the seismic history, the motion
along the Muertos Trough appears to be a small
fraction that of the Puerto Rico Trench [14]. The
authors confirm this suggestion based on the slow slip
rate in this boundary form historical literature
consulted to compile the catalogue.
2.4 Transition Zone 9 and 10A
These seismogenic zones are defined as the
intersections between the transform faults and
subduction zones with the Lesser Antilles Arc located
at the north and the south of the Eastern Caribbean.
Zone 10A includes the shallow seismic activity in the
southern part of the island of Tobago which is
considered within the Caribbean-South American Plate
boundary [20-22]. Russo et al. [12] suggest also that
the northern boundary of the Eastern Caribbean-South
America plate may lie as far north as the southern end
of the Grenada basin. Moderate but shallow
earthquakes occurred south of Tobago on 1982 (4.8
Mw) and 1997 (6.7 Mw) with right-lateral strike slip
and normal faulting mechanisms, respectively [20, 23].
The transition Zone 9 is characterized by a
low-seismicity level yielding mainly normal focal
mechanisms in the boundary zone between the Lesser
Antilles arc and the Puerto Rico Trench.
2.5 Zone 10B: East of Trinidad
Russo and Speed [24] suggested that the earthquakes
located in this zone are consistent with the detachment
and bending-flexure of the South American slab
moving toward the collision zone. The zone covers
mainly normal faulting mechanisms with ENE-WSW
striking planes and strike slip faults with an average
depth of 45 km. The maximum magnitude reported in
the catalogue for this zone is 6.7 M
w
(March 10, 1988).
2.6 Zone 11: North of Paria Peninsula
This zone constitutes a subducting detached oceanic
lithosphere with depth ranging from 50 km to 300 km
and represents one of the most active seismogenic
sources in the Eastern Caribbean [12]. The largest
reported earthquakes occurred in October 21, 1766 and
January 10, 1888 with magnitudes of 7.5 M
w
and 7.0
M
w
, respectively. The focal mechanisms indicate that
there is a normal faulting resulting from the initial
flexure of the down going slab steeply dipping to the
NW at 60. However, mixed-motion earthquakes with
thrust and strike slip indicated bending of the
subducting slab at deeper depths.
2.7 Zone 12: Trinidad Faults
This zone includes the faults mapped in Trinidad
namely, the Northern-Range and Central Range, and
Darien Ridge and Arima and Los Bajos Fault,
characterized by earthquake with depth less than 50 km.
Weber [21] employed far and near field geodesy and
palaeoseismology to search fossil earthquakes on the
Central Range Fault, the principal active dextral
strike-slip in Trinidad. He concluded that the central

1411
range fault was locked, stored and released significant
elastic motion in the recent past. He suggests a slip rate
of 12 mm/year and that several more meters of motion
could be stored in the fault. In the south, Weber [21]
suggests a slip rate of 6 mm/year of the Los Bajos Fault
with a dextral motion. Regional tectonic-geological
studies conclude that El Pilar Fault might right-step
into Central Trinidad [25], however, Weber [21, 26]
affirms that the N68E oblique trending in the central
range fault is not associated with El Pilar Fault 90
trending of pure wrenching. More recently, Prentice
et al. [6] suggested prehistoric earthquakes of M > 7 in
this fault based on paleoseismological investigations.
A detail PSHA focused only for Trinidad has been
done recently by Ref. [7] incorporating larger
maximum magnitudes (M 7.7) on these faults in a logic
tree formulation, slip rates and deterministically
bounded ground motions for the Bridges
Reconstruction Program and Sir Solomon Hochoy
Highway Extension to Point Fortin. This information
has been taken into account for the new seismicity
parameters presented in Table 1. The northern range
and the Arima Fault comprise a complex fault system
with lateral strike-slip, thrust and normal faulting. On
December 2 and 3, 2004 events with a magnitude of 5.8
and 5.4 (M
w
) occurred in the central north-east of
Trinidad, fault plane solutions suggest mainly a normal
motion with a component of right-lateral strike slip.
The location of these earthquakes and the correspondent
focal mechanisms coincide with the Northern Range
normal fault dipping southward mapped by Algar and
Pindell [27] beneath the Caroni Swamp area.
2.8 Zone 13 and 14: El Pilar Fault
These zones comprise the boundary between the
Caribbean and the South American plate. The events
that have their origin in the fault are shallowless that
50 km depthand they are characterized mainly by
right lateral strike slip mechanism in the northern coast
of South America. The Caribbean Plate is moving
about 20 mm/yr in an easterly direction relative to
South America [28]. However, thrust focal mechanism
also takes place in this region reflecting the oblique
collision at crustal levels between the Caribbean and
the South American Plates. The authors observed a
high level seismic output in Zone 13 that extends from
63.5 W to 62.3 W longitude covering the Araya-Paria
Isthmus, and a moderate seismicity level in Zone 14
that extends from 67.0 W to 63.5 W longitude
covering the vicinity of Caracas to the Araya region.
The maximum magnitude listed in our catalogue
occurred on October 4, 1957 (6.4 M
w
) in Zone 13 and
on September 1, 1530 (8.0 M
w
) in Zone 14.
2.9 Zone 15: South of Trinidad
Russo et al. [12] defined this zone as a passive
margin edge in the foreland basin in north of south
American continent, covering events with strike slip,
mixed thrust and strike slip, and thrust mechanism
around the Orinoco Delta region in Venezuela, with an
average depth of 50 km and a maximum magnitude of
6.6 (M
w
).
Figs. 2 and 3 illustrate the geometrical configuration
for both, shallow and deep zones covering the
subduction, upper-crustal volcanic island-arc,
transform and intra-plate faulting and transitions zones
that affect Barbados.
3. Characteristic Magnitude Distribution for
the Interface Subduction Zones SZ2 and SZ3
Many fault that exhibit seismicity and crustal strain,
indicate that an exponential distribution encompassing
all magnitudes is inappropriate. The continuous
exponential distribution may be adequate for events up
to 8.0 M
W
in the subduction zones, however, larger
earthquakes may occur with a characteristic magnitude
namely M = 8.0 to 9.0 whose frequency of occurrence
is higher than obtained by extrapolating the
exponential part of small earthquakes (Fig. 4). The
authors employ truncated exponential distributions that
could more accurately model the understanding of
future events in the interface subduction zone beneath


1412

Fig. 4 Seismicity model for the Subduction Interface Zone (SZ3) beneath Barbados. Characteristic earthquakes are
considered with a recurrence interval of 600 years for characteristic magnitudes 8.30-8.65. An additional option is a recurrence
interval of 750 years for characteristic magnitudes 8.65-9.00 (Table 2).

Barbados, namely: (1) a distribution between M
min
=
4.5 and M
max
= 8.0 with a standard b value (Table 1);
(2) a distribution between M
min
= 8.0 and M
max
= 9.0
with a b = 0.0 representing the equal likelihood of a
characteristic event in that magnitude range.
The activity rate of the minimum magnitude M
min
for
the exponential part is computed from the classical
Gutenberg Richter (G-R) relationship (Table 1). The
activity rate for the characteristic part can be estimated
via paleoseismological investigations, crustal strain
data (geodesy) or the recurrence interval between large
events. The recurrence interval of large earthquakes in
the Eastern Caribbean is very uncertain since there is
too little activity to estimate what the largest possible
seismic moment could be to constrain the frequency of
occurrence. The Headquarters for Earthquake Research
and Promotion in Japan had determined the recurrence
interval of an M 9.0 earthquake to be 600 years based
on historical records of earthquakes in 869, 1611 and
2011. Suppasri et al. [29] concluded that the recurrence
interval for an event comparable to the Sumatra 2004
tsunami is approximately 520 years.
Since there are no historical accounts of such big
megathrust events in the trench of the southern part of
the Eastern Caribbean near Barbados in the past five
centuries, our analysis is based on slip rates in the
subduction interface zone. Recently, DeMets et al. [10]
affirmed by employing GPS stations located in the
Caribbean and Central American region, that the
Caribbean Plate is in fact moving 2 cm/yr relative to
the North American Plate, indicating the possibility of
large strain accumulation that could trigger large
earthquakes as the afore mentioned examples.
The authors incorporated into the logic tree, two
different characteristic magnitudes with a width of 0.35
units with equal weights, namely: (1) 8.3-8.65; (2)
8.65-9.0. To calculate the associated slip rates for the
exponential and characteristic part, the procedure
below is followed:
The classical Gutenberg and Richter relationship
(G-R) is written as follows:
logN = o - bH (1)

1413
where, N is the number of earthquakes per year with
magnitude equal to or above M, a, b: constants for the
Gutenberg and Richter relationship (Table 1).
If one wants to relate a slip rate s to an activity rate
(number of earthquakes per year) for the seismicity
represented by the exponential form in Eq. (1), the
following formulas yields [30]:
v
M
min
= s po
1
y(1-[y)
k[cxp|[(M
min+
dc)]
[Ho
mux
1-[y
- Ho
mn
1-[y
-1
(2)
and
logHo = cH + J (3)
where,
min
: activity rate for the minimum magnitude (M
min
)
using the Gutenberg and Richter relationship in Eq. (1);
: rigidity of the earth crust equal to 3.0 10
11
dyne/cm
2
;
a
T
: rupture area of the fault;
: b(ln10), where b is from the Gutenberg and
Richter relationship log N = a-bM;
c, d: constants c = 1.5 and d = 16.05 in Eq. (3) to
obtain the seismic moment Mo giving a moment
magnitude M;
: c(ln10);
Mo
max
: moment magnitude in dyne-cm for the
maximum magnitude M
max
employing Eq. (3);
Mo
min
: moment magnitude in dyne-cm for the
minimum magnitude M
min
employing Eq. (3).
In the case of the characteristic part, the slip rate can
be found by inverting Eq. (5) after obtaining the
seismic moment rate (dyne-cm/year) Ho
and
employing the following expressions:
Ho
=
v
chcr
y
Mo
mcx
-Mo
min
M
mcx
-M
min
(4)
Ho
= s po
1
(5)
where,
char
is the activity rate for the characteristic
earthquakes. When evaluating the seismicity for the
interface subduction zone SZ3, the authors employed
in the logic tree two characteristic models with M
8.3-8.65 and an M 8.65-9.0 and an exponential part
with a G-R relationship log N = 3.216 0.725 M
(Table 1) for M 4.5-8.0.
The authors model the characteristic magnitudes of
8.30-8.65 with a
char
= 0.00167 earthquakes per year
which is equivalent to a 600 year return period (Table 2),
a rupture area of 450 km 90 km [31] and using Eq. (4)
and inverting Eq. (5) indicates a slip rate of 0.85 cm/yr.
When analyzing the characteristic magnitudes of
8.65-9.0 with a return period of 750 years (
char
=
0.00133 earthquakes per year) a rupture area of 450 km
220 km yields a slip rate of 0.92 cm/ yr. Historical
seismicity indicates a
4.5
= 0.9 earthquakes/year and b
= 0.725 for the magnitude ranges 4.5 M 8.0;
inverting Eq. (2) yields a slip rate of 0.22 cm/yr. The
total slip rate resulting from the exponential and the
characteristic part for the SZ3 yields 1.99 cm/yr (= 0.22
+ 0.85 + 0.92 cm/yr).
The same analysis is performed for the interface
subduction zone SZ2 employing the exponential part
(Table 2) and the same two characteristic model with
M 8.3-8.65 and a M 8.65-9.0 setting 650 years and 850
years recurrence interval, the total slip rates yields 1.98
cm/yr, respectively (Table 3).
The average slip rate of the interface subduction
zone in the Eastern Caribbean is suggested to be 2
cm/yr on the basis of geodetic data. The previous
example demonstrates that for the two characteristic
models and the exponential part, the slip rate is nearly 2
cm/yr for the region validating the recurrence interval
between large subduction earthquakes for the Eastern

Table 2 Seismicity models for the interface subduction zone SZ3 beneath Barbados, the activity rate is presented for the
minimum magnitude M
min
for the exponential part being 4.5 and for the characteristic earthquakes.
Magnitudes b value M range
Recurrence
interval (yr)
Activity rate
(earthquakes/yr)
s (slip rate)
Exponential 0.725 4.5-8.0 1.1 0.9 0.22 cm/yr
Characteristic 0 8.3-8.65 600 0.00167 0.85 cm/yr
All - 4.5-8.65 - - 1.99 cm/yr


1414
Table 3 Seismicity models for the interface subduction zone SZ2. The activity rate is presented for the minimum magnitude
M
min
for the exponential part being 4.5 and for the characteristic earthquakes.
Magnitudes b value M range
Recurrence
interval (yr)
Activity rate
(earthquakes/yr)
s (slip rate)
Exponential 0.893 4.5-8.0 0.25 3.94 0.38 cm/yr
All - 4.5-8.65 - - 1.98 cm/yr

Caribbean and their use in the Seismic Hazard
computations.
4. Ground Motion Prediction Equations
The GMPEs are regionally dependent and the choice
of appropriate relations depends on the tectonic
environment around the site of interest. No specific
GMPEs have been developed for Barbados and the
Eastern Caribbean. In order to be compatible with the
tectonic environment of the region, the following
GMPEs are employed to compute the seismic hazard:
(a) For subduction zones:
Youngs et al. [32]; Atkinson and Boore [33]; Zhao et
al. [34]; Kanno et al. [35]; Lin and Lee [36].
(b) For crustal zones:
Kanno et al. [35]; Zhao et al. [34]; Abrahamson and
Silva [37]; Boore and Atkinson [38]; Campbel and
Bozogornia [39];
(c) For the volcanic zone:
Sadigh et al. [40]; Zhao et al. [34]; Kanno et al. [35];
Abrahamson and Silva [37]; Chiou and Youngs [41].
To incorporate the GMPEs in the PSHA
computation, the correspondent distance definition of
each GMPEs has been used, incorporating suitable
empirical relations between rupture length and
magnitude for crustal and subduction earthquakes
[42, 43]. Details of the formulation of each GMPEs can
be found in the original articles cited in the reference
section of this report.
5. Logic Tree Formulation and Hazard
Computation
Epistemic uncertainties in the hazard assessment
have been addressed in a logic-tree framework by
considering the following parameters: (1) the
maximum cutoff magnitudes; (2) the GMPEs.
The logic tree for the horizontal component consists
of a total of 15 branches for the interface subduction
zones (SZ2-SZ3), see Fig. 5: Five branches are
associated to the normal exponential G-R relationships
with five GMPEs while the remaining 10 branches
refer to the two characteristic magnitude models and
the five GMPEs; five branches are associated for the
rest of seismogenic sources employing the five GMPEs.
Concerning GMPEs, equal weights have been
associated with GMPEs since no strong motion data is
available for Barbados. Each seismogenic zone in the
Cornell-McGuire method has been treated separately:
All the branches for each seismogenic zone have been
weighted averaged to obtain a mean hazard curve for
each seismogenic zone. The 15 mean hazard curves (1
15 seismogenic zones) have then been summed at the
site of interest to produce a single hazard curve
representing the Cornell-McGuire final computation.
Seismic hazard results are presented in terms of
uniform hazard spectral accelerations for the horizontal
component and 5% damping, calculated for:
(1) stiff ground conditions (NEHRP site
classification B) and flat topographic surface
representing outcropping ground conditions;
(2) 5 percent structural damping of the critical;
(3) 5 return periods RP = 95 475 975 2,475
4,975 years;
(4) 34 spectral periods (from 0 to 3 s);
(5) A truncation value of 3 (sigma) in the GMPEs.
The PSHA results are presented in terms of Hazard
Curves (Figs. 6-8) for the PGA and the acceleration
spectral ordinates for 0.2 s and 1.0 s, and the elastic

1415
response spectra for 5% damping is presented in Fig. 9
for 95, 475, 975, 2,475 and 4,975 years RP. The
spectrum for 2/3 of 2,475 years RP is also presented.
Note the similarity with the spectrum for 975 years RP.
6. Disaggregation of Seismic Hazard
The probablisitic seismic hazard assessment
combines the contributions from all the considered
sources to provide an estimate at the site of a ground
motion parameter of interest with a certain probability
of exceedance during a specified lifetime of the
structure. Therefore, the physical image of an
earthquake in terms of magnitude and source-to-site
distance is not clearly visualized in the PSHA.
However, through a disaggregation analysis it is
possible to identify one or more earthquakes and find
the largest contribution to the hazard [44, 45].
In fact, the disaggregation process separates the
contributions to the

mean annual rate of exceedance of
a specific ground-motion value at a site due to a
magnitude (M) and a source to site distance (R)
providing the earthquake scenarios that dominate the
hazard at the site.
Disaggregation in terms of MR pairs has been
computed for one site located in Barbados (Fig. 10), the
PGA and 475 and 975 years return period. The
magnitude-distance pair that dominates the hazard
indicates an M 7.4 and 8.6 and a rupture distance R of
42.5 km in the interface subduction zone (SZ3) for the
475 and 975 years RP (return period), respectively. The

Fig. 5 Logic tree adopted for the horizontal component of ground motion. M
char
= magnitude for the characterisitc
earthquakes in the interface subduction zones. The weights used in hazard calculations correspond to the numbers written to
the right of each branch of the tree.

1416
1E-005 0.0001 0.001 0.01 0.1 1 10
Peak Ground Acceleration (g)
1E-006
1E-005
0.0001
0.001
0.01
0.1
1
10
100
A
n
n
u
a
l

f
r
e
q
u
e
n
c
y

o
f

e
x
c
e
e
d
a
n
c
e

Fig. 6 Peak ground acceleration hazard curve for Bridgetown, Barbados location.

1E-005 0.0001 0.001 0.01 0.1 1 10
Spectral Acceleration (g)
1E-005
0.0001
0.001
0.01
0.1
1
10
100
A
n
n
u
a
l

f
r
e
q
u
e
n
c
y

o
f

e
x
c
e
e
d
a
n
c
e

Fig. 7 Spectral acceleration hazard curve at 0.2 s for Bridgetown, Barbados location.

1417
1E-005 0.0001 0.001 0.01 0.1 1 10
Spectral Acceleration (g)
1E-005
0.0001
0.001
0.01
0.1
1
10
100
A
n
n
u
a
l

f
r
e
q
u
e
n
c
y

o
f

e
x
c
e
e
d
a
n
c
e

Fig. 8 Spectral acceleration hazard curve at 1.0 s for Bridgetown, Barbados location.

0 1 2 3
0.5 1.5 2.5
Period (s)
0
0.5
1
1.5
2
2.5
S
p
e
c
t
r
a
l

A
c
c
e
l
e
r
a
t
i
o
n

(
g
)
Graph 1
4975 years
2475 years
975 years
475 years
95 years
2/3 of 2475 years

Fig. 9 Elastic design response spectra for Bridgetown, Barbados setting 95, 475, 975, 2,475 and 4,975 years RP (return period).
Units of acceleration are in g. The spectra for 2/3 of 2,475 year RP is also presented. It is noted the similarity with the
spectrum for 975 year RP. Peak ground acceleration values can be taken at the 0 period.
4,975 year
2,475 year
975 year
475 year
95 year
2/3 of 2,475 year
Graph 1

1418

Fig. 10 Disaggregation plots for the peak ground acceleration corresponding to 475 years (above) and 975 years return period
(bottom). The magnitude-distance pair that dominates the hazard yields a M 7.4 and M 8.6 with a rupture distance of 42.5 km
in the interface subduction zone (SZ3) for the 475 and 975 years RP, respectively (indicated by the white arrows). An event
with a M 8.0 at a rupture distance of 107.5 km in the intraplate subduction zone (SZ5) is the second scenario that dominates the
hazard for both 475 and 975 years RP.

1419

distance to the source of 42.5 km represents the closest
distance from the site to the inclined interface zone
Beneath the island. An M 8.0 at a rupture distance of
107.5 km in the intraplate subduction zone (SZ5) is the
second scenario that dominates the hazard for both, the
475 and 975 years RP (Fig. 10).
7. Conclusions
The scope of the work presented here was to carry
out a seismic hazard assessment for Bridgetown,
Barbados based on a state of-the-art PSHA study.
PSHA has been performed using the classical
Cornell-McGuire approach. Design response spectra
have been developed for 5% of critical damping and
the horizontal ground motion on outcropping bedrock
conditions, thereby neglecting local site amplification
effects at this stage.
The present study has shown that Barbados is
characterized by a high seismic hazard, the horizontal
peak ground acceleration expected on stiff ground
yields 0.36 g, setting a 10% probability of exceedance
in 50 years lifetime for the structure, which
corresponds to 475 years RP. The level of hazard
prescribed by this research is substantially larger than
previous works (i.e., Bozzoni et al. [3] suggest 0.22 g
for 475 years RP), this increment is attributed to two
reasons: (1) Former works did not consider the
possibility of megathrust earthquakes beneath
Barbados, in fact the maximum magnitudes considered
in previous works used for this zone yields 7.0-7.5
based solely on 500 year of earthquake history, much
lower than the maximum magnitudes employed in this
work: a M 8.0 within the exponential G-R relationship
and the characteristic earthquakes within the
magnitude range of M 8.3 to 9.0 based on slip rates; (2)
only the classical method of Cornell-McGuire was
taken into account in the computation of the
disaggregation process for this work; the authors did
not introduce in the logic tree formulation the free-zone
method proposed by Woo [46] and used by Bozzoni
et al. [3] as well, it has been observed that the free zone
methods yield lower hazards results than the classical
Cornell method in areas of sparse seismicity in terms of
short period components of ground motion [47, 48].
The authors recommend the use of the 2/3 of 2,475
years RP seismic loads (which is equivalent to the 975
years RP) prescribed by the IBC (International
Building Code) [49, 50] and make use of the 0.46 g for
the PGA in the future design of short period structures
in Barbados (Fig. 9), indeed, the Magnitude-Distance
Pair (M-R) dominating the hazard at this level indicates
an interface earthquake of M 8.6 at a distance R of 42.5
km beneath the island of Barbados and with important
contribution of an M 8.0 at a distance R of 107.5 km in
the intraplate subduction zone for the 975 years RP. It
is noted that the definition of distance employed here is
the closest distance to the rupture fault. The authors
have performed the PSHA employing available
GMPEs for interface subduction zones which maximum
magnitudes in the range of 8.0-8.5, better results would
be available employing GMPEs incorporating the new
strong motion data of recent interface subduction
earthquakes in Chile and and Japan (M 9.0).
If a mega earthquake has not occurred in a region
within the past 500 yearsfor which historical data is
availableit does not mean that an earthquake will not
occur in the future (as the cited cases in Japan and
Sumatra). Barbados is just above the inclined
subduction interface zone having slow slip rates
yielding long recurrence intervals in this region. The
validation of this preliminary characteristic model
provided in this work for mega interface subduction
events in the Caribbean could be done focusing on
seismic plates coupling research and tsunami
geological sedimentation studies in the Islands,
including Barbados.
For 95 years, RP the magnitude distance pair that
contributes the most to the hazard gives an M 6.8 with a
distance of 42.5 km (subduction interface event) and an
M 7.6 at a distance of 107.5 km (subduction intraplate
event) corresponding to a PGA of 0.17 g. This seismic
load can be used to design provisional structures during

1420
the construction phase of buildings in Bridgetown. For
the 2,475 years RP, the magnitude-pair yields an M 8.7
and a distance of 42.5 km (subduction interface event)
and M 8.1 and a distance of 102.5 km (subduction
intraplate event) for a 0.68 g of PGA.
Acknowledgments
The computer program EZ-FRISK 7.62, a
software for earthquake ground-motion estimation,
developed by Risk Engineering Inc., has been used for
the Cornell-McGuire seismic hazard computation.
Maps have been prepared using ESRI Arc Map 10.1
(Arc View) Geographic Information System. The
authors thank to Geoff Wadge and Trevor Jackson to
allow us reproducing Fig. 2. This study has been
funded initially by the World Bank as a part of the Risk
Atlas Project for the Caribbean under the supervision
of the DRRC (Disaster Risk Reduction Centre) at the
University of the West Indies, Mona, Jamaica. The
article was finalized during the project Strategies and
Tools for Real Time Earthquake Risk Reduction
(REAKT) founded by the European Community Work
Package 7.
References
[1] C.A. Cornell, Engineering seismic risk analysis, Bull.
Seismol. Soc. Am. 58 (1968) 1583-1606.
[2] R.K. McGuire, FORTRAN Computer Program for
Seismic Risk Analysis, U.S. Geological Survey, Open-file
report, 1976, pp. 76-67.
[3] F. Bozzoni, M. Corigliano, C. Lai, W. Salazar, L.
Scandella, E. Zuccolo, et al., Probabilistic seismic hazard
assessment at the Eastern Caribbean Islands, Bull. Seismol.
Soc. Am. 101 (5) (2011) 2499-2521.
[4] G. Wadge, The lesser Antilles, Chapter 9, Caribbean
Geology, S. Donovan, T. Jackson (Eds.), The University
of West Indies Publishing Association, Mona, Jamaica,
1994.
[5] C.H. Boynton, G.K. Westbrook, M.H.P. Bott, R.E. Long,
A seismic refraction investigation of crustal structure
beneath the Lesser Antilles island arc, Geophysical
Journal of the Royal Astronomical Society 58 (1979)
371-393.
[6] C. Prentice, J. Weber, C. Crosby, D. Ragona, Prehistoric
earthquakes on the Caribbean-South American plate
boundary, Central Range Fault, Trinidad, Geology 38 (8)
(2010) 675-678.
[7] Halcrow, Bridges Reconstruction Program and Sir
Solomon Hochoy Highway Extension to Point Fortin,
Probabilistic Seismic Hazard Assessment, 2012, p. 26.
[8] P. Bernard, J. Lambert, Subduction and seismic hazard in
the Northern Lesser Antilles: Revision of the historical
seismicity, Bull. Seismol. Soc. Am. 78 (6) (1988)
1965-1983.
[9] D.E. Byrne, D.M. Davis, L.R. Sykes, Loci and maximum
size of thrust earthquakes and the mechanics of the
shallow region of subduction zones, Tectonics 7 (4) (1988)
833-857.
[10] C. DeMets, G. Mattioli, P. Jansma, R. Rogers, C. Tenorio,
H. Turne, Present motion and deformation of the
Caribbean plate: Constrains from new GPS geodectic
measurements from Honduran and Nicaragua, The
Geological Society of America Special Paper 428 (2007)
21-36.
[11] M. Bengoubou-Valerius, S. Bazin, D. Bertil, F. Beauducel,
A. Bosson, CDSA: A new seismological data center for
the French lesser Antilles, Seismological Research Letters
79 (1) (2008) 90-102.
[12] R.M. Russo, R.C. Speed, E.A. Okal, J.B. Shepherd, K.C.
Rowley, Seismicity and tectonics of the Southeastern
Caribbean, J. Geophys. Res. B 98 (8) (1993) 299-319.
[13] N. Feuillet, I. Manighetti, P. Tapponnier, Arc parallel
extension localization of volcanic complexes in
Guadeloupe, Lesser Antilles J. Geophys. Res. B 107 (12)
(2002) 23-31.
[14] W.R. McCann, On the earthquake hazards of Puerto Rico
and the Virgin Islands, Bull. Seism. Soc. Am. 75 (1985)
251-262.
[15] P. Jansma, A. Lopez, G. Mattioli, C. DeMets, T. Dixon, P.
Mann, et al., Neotectonics of Puerto Rico and the Virgin
Islands, northeastern Caribbean, from GPS geodesy,
Tectonics 19 (2000) 1021-1037.
[16] J. Clinton, G. Cua, V. Hurfano, C.G. von
Hillebrandt-Andrade, J. Cruzado, The current state of
seismic monitoring in Puerto Rico, Seismological
Research Letter 77 (5) (2006) 532-543.
[17] C. Mueller, A. Frankel, M. Petersen, E. Leyendecker,
Documentation for 2003 USGS Seismic Hazard Maps for
Puerto Rico and the U.S. Virgin Islands, USGS Open File
report 03-379, 2004.
[18] P. Molnar, L.R. Sykes, Tectonics of the Caribbean and
Middle America regions from local mechanisms and
seismicity, Geol. Soc. Am. Bull. 80 (1969) 1639-1684.
[19] C. Mendoza, W. McCann, Improving the Seismic Hazard
Model for Puerto Rico through Seismic Tomography and a
Reliable Microearthquake Catalogue with Recalculated
Magnitudes and Calibrated Hypoentral Error Estimates,
Research Supported by USGS No. 05HQGR0012, 2009.

1421
[20] J. Latchman, Tobago and earthquakes, Ph.D. Thesis, The
University of the West Indies, 2009.
[21] J.C. Weber, Neotectonics in the Trinidad and Tobago,
West Indies segment of the Caribbean-South American
plate boundary, Occasional Papers of the Geological
Institute of Hungary 204 (2009) 21-29.
[22] R. Burmester, M. Beck, R. Speed, A. Snoke, A
preliminary paleomagnetic pole for mid-Cretaceous rock
from Tobago: Further evidence for large clockwise
rotations in the Caribbean-South American plate boundary
zone, Earth and Planetary Science Letters 139 (1996)
79-90.
[23] F.D. Morgan, G. Wadge, J. Latchman, W.P. Aspinall, D.
Hudson, F. Samstag, The earthquake Hazard alert of
September 1982 in southern Tobago, Bull. Seismol. Soc.
Am. 78 (4) (1988) 1550-1562.
[24] R.M. Russo, R.C. Speed, Oblique collision and tectonic
wedding of the South America continent and Caribbean
terrains, Geology 20 (1992) 447-450.
[25] J.F. Flinch, V. Rambaran, W. Ali, V. de Lisa, G.
Hernandez, K. Rodrigues, et al., Structure of the Gulf of
Paria pull-apart basin (Eastern Venezuela-Trinidad), in: P.
Mann (Ed.), Caribbean Sedimentary Basins, Elsevier
Basins of the World, Elsevier, 2000.
[26] J. Weber, J. Saleh, S. Balkaransingh, T. Dixon, W.
Ambeh, T. Leong, et al., Triangulation-to-GPS-to-GPS
Geodesy in Trinidad, West Indies: Neotectonics, Seismic
risk, and Geological Implications, Marine and Petroleum
Geology, 2009.
[27] S. Algar, J. Pindell, Structure and deformation of the
northern range of Trinidad and adjacent areas, Tectonics
12 (1993) 814-829.
[28] O.J. Prez, R. Bilham, R. Bendick, J. Velandia, C.
Moncayo, M. Hoyer, et al., Velocity fields across the
southern Caribbean Plate boundary and estimates of the
Caribbean/South American plate motion using GPS
geodesy 1994-2000, Geophysical Research Letters 28 (15)
(2001) 2987-2990.
[29] A. Suppasri, T. Futami, S. Tabuchi, F. Imamura, Mapping
of historical tsunamis in the Indian and Southwest Pacific
Oceans, International Journal of Disaster Risk Reduction 1
(2012) 62-71.
[30] R.K. McGuire, Seismic Hazard and Risk Analysis,
Monograph 10, Earthquake Engineering Research
Institute EERI, 2004.
[31] N. Gregor, W. Silva, I. Wong, R. Youngs,
Ground-motion attenuation relationships for Cascadia
subduction zone megathrust earthquakes based on
stochastic finite-fault model, Bull. Seismol. Soc. Am. 92
(2002) 1923-1932.
[32] R.R. Youngs, S.J. Chiou, W.J. Silva, J.R. Humphrey,
Strong ground motion attenuation relationships for
subduction zone earthquakes, Seismological Research
Letters 68 (1) (1997) 58-73.
[33] G.M. Atkinson, D.M. Boore, Empirical ground-motion
relations for subduction-zone earthquakes and their
application to Cascadia and other regions, Bull. Seismol.
Soc. Am. 93 (4) (2003) 1703-1729.
[34] J.X. Zhao, J. Zhang, A. Asano, Y. Ohno, T. Oouchi, T.
Takahashi, et al., Attenuation relations of strong ground
motion in Japan using site classification based on
predominant period, Bull. Seismol. Soc. Am. 96 (3) (2006)
898-913.
[35] T. Kanno, A. Narita, N. Morikawa, H. Fujiwara, Y.
Fukushima, A new attenuation relation for strong motion
in Japan based on recorded data, Bull. Seismol. Soc. Am.
96 (3) (2006) 879-897.
[36] P.S. Lin, C.T. Lee, Ground-motion attenuation
relationships for subduction-zone earthquakes in
Northeastern Taiwan, Bull. Seismol. Soc. Am. 98 (1)
(2008) 220-240.
[37] N. Abrahamson, W. Silva, Summary of the Abrahamson
& Silva NGA ground-motion relations, Earthquake
Spectra 24 (1) (2008) 67-97.
[38] D.M. Boore, G.M. Atkinson, Ground-motion prediction
equations for the average horizontal component of PGA,
PGV and 5%-damped PSA and spectral periods between
0.01 s and 10.0 s, Earthquake Spectra 24 (1) (2008)
99-138.
[39] K.W. Campbell, Y. Bozogorgnia, NGA ground motion
model for the geometric mean horizontal component of
PGA, PGV, PGD and 5% damped linear elastic response
spectra for periods ranging from 0.01 s to 10 s, Earthquake
Spectra 24 (1) (2008) 139-171.
[40] K. Sadigh, C.Y. Chang, J.A. Egan, F. Makdisi, R.R.
Youngs, Attenuation relationships for shallow crustal
earthquakes based on California strong motion data,
Seismological Research Letters 68 (1) (1997) 180-189.
[41] B.S.J. Chiou, R.R. Youngs, An NGA model for the
average horizontal component of peak ground motion and
response spectra, Earthquake Spectra 24 (1) (2008)
173-215.
[42] D. Wells, J. Coppersmith, New empirical relationships
among magnitude, rupture length, rupture width, rupture
area, and surface displacement, Bull. Seismol. Soc. Am.
84 (2) (1994) 974-1002.
[43] F. Strasser, M. Arango, J. Bommer, Scaling of the source
dimensions of interface and intraslab subduction-zone
earthquakes with moment magnitude, Seismological
Research Letters 81 (2010) 941-950.
[44] M. Chapman, A probabilistic approach to ground-motion
selection for engineering design, Bull. Seismol. Soc. Am.
85 (3) (1995) 937-942.
[45] P. Bazzurro, A. Cornell, Disaggregation of seismic hazard,

1422
Bull. Seismol. Soc. Am. 89 (1999) 501-520.
[46] G. Woo, Kernel estimation methods for seismic hazard
area source modelling, Bull. Seismol. Soc. Am. 86 (2)
(1996) 353-362.
[47] S. Molina, C. Lindholm, H. Bungum, Probabilistic
seismic hazard analysis: Zoning free versus zoning
methodology, Bulletin of Applied Theoretical
Geophisycs 42 (1993) 19-39.
[48] E. Zuccolo, M. Corigliano, C. Lai, F. Bozzoni, L.
Scandella, W. Salazar, Zone-free approach applied to
Subduction zone in the Eastern Caribbean Islands, in: 29th
National Convention, National Group of Geophysics
Trieste, 2010.
[49] ASCE 7-10, Minimum Design Loads for Buildings and
Other Structures, American Society of Civil Engineers,
2010, p. 658.
[50] Strategic Standardization, IBC (International Building
Code), Inc., U.S., 2012.

Use of Geophysical Techniques for the Localization of
the Restricting Zones of Permeability in the Bottom of
Basin: Forecasting the Zones of Clogging
Ibrahim Alimi Ichola
1
, Salima Bouchemella
2
and Claudio Fernando Mahler
3

1. Laboratoy of Civil Engineering and Environment, Villeurbanne 69621, France
2. Department of Civil Engineering, Guelma and Infrares Laboratory, Guelma 24000, Algeria
3. Waste Technology Study Group, Federal University of Rio de Janeiro, Rio de Janeiro 22470-050, Brasil

Abstract: Measurements of dielectric parameters followed by permeability tests are performed on soil samples of infiltration basin.
The dielectric parameters are obtained by TDR (time domain reflectrometry) measurements from which it is found that the
measurement of the permittivity, the electrical conductivity and the relaxation time of compacted soil allows the mapping of the
values of the coefficient of permeability at the surface of an infiltration basin. With the distribution of the coefficient of permeability,
the areas of water stagnation can be detected before the basin filling. The study proves that the TDR measurements for the detection
of these zones can be used for the management of infiltration basins for sustainable working and their remediation can be undertaken
before the rainy seasons.

Key words: Permeability, infiltration pond, soil conductivity, TDR method.

1. Introduction

The geophysical techniques were mainly developed
for detection of in-depth anomalies of sites. For the
past few decades, these techniques have been used to
detect or localize the pollution in the soil and the
ground water within old sites of industrial activities.
Analysis of the parameters used in the most of the
geophysical techniques can be described by the Debye
relaxation function representing the expression of the
complex permittivity. The success of the geophysics
methods for the site survey rests mainly on the fact
that the soil is a dispersive medium and that its
dependence on frequency has the advantage that the
contribution of the bulk material and the grain
boundaries in the measurements can be distinctly
separated. Thus different models of relaxation
dielectric are proposed for analysing the dielectric

Corresponding author: Ibrahim Alimi Ichola, Ph.D.,
associate professor, research fields: geotechnic-geophysic in
unsaturated soils, water and pollutant migration in vadose zone.
E-mail: ibrahim.alimi-ichola@insa-lyon.fr.
medium response. These models introduce the
relaxation frequency or time, the real static
permittivity and the real permittivity at infinity
frequency.
TDR (time domain reflectrometry) measurements
are developed to detect faults in the transmission line.
Topp et al. [1] and several other authors that pursued
similar research proposed methods of determination of
dielectric constant and electrical conductivity of soil
surrounding a transmission line. As already shown [2],
there exist realistic possibilities of determining the
hydraulic conductivity from the results of vertical
electric soundings. Consequently, many calibration
tests are carried out on samples from the infiltration
basin by dielectric and electrical conductivity
measurements applying the TDR method. Standard
permeability tests are then performed on the samples
after TDR measurements. Subsequently, a mapping of
the infiltration basin bottom is carried out according to
a predetermined grid, by adopting TDR measurements.
DAVID PUBLISHING
D
Use of Geophysical Techniques for the Localization of the Restricting Zones of
Permeability in the Bottom of Basin: Forecasting the Zones of Clogging

1424
2. Soil Dielectric Properties
2.1 Complex Permittivity of Dielectric Material
Every material has electrical characteristics which
depend on the dielectric properties. Measurement
techniques of these properties offer to research
engineers the means to ameliorate their quality or
control their fabrication. Basically, the dielectric
properties of such materials are represented by the
electrical resistivity, the permittivity and by the
magnetic permeability. These properties change with
temperature, pressure, composition and the molecular
structure of the material. Fundamental definitions of
the basic properties are as follows:
Electrical conductivity concerns the ionic flow in
the material, high resistivity characterizes insulator
material, whereas weak resistivity characterizes
electrical conductor material;
Material dielectric constant: A material is called
dielectric when it is able to store energy and to restore
it, the dielectric constant is the ratio between the
capacitor capacity in void and in dielectric
material;
Material permittivity describes the interaction
between a material and an applied electric field, the
dielectric constant measures the ratio between the
permittivity of the material and the permittivity of the
void;
Loss phase: occurs when the permittivity of the
material is a complex number represented by a vector.
The relative energy loss is the ratio between the
restored energy and the stored energy. In this case, the
real part of the material permittivity describes the
energy storage capacity and the imaginary part
describes the energy loss.
All geophysical methods use the electrical flow or
the EM (electromagnetic) propagation to reveal the
dielectric properties of soil. The change in the value of
the measured parameters allows for contrast
identification.
The complex permittivity of the material where
electrical conductivity is represented as o is expressed
as:
|
|
.
|
\
|
+ =
0
"
'
*
ec
o
c c c
d
j (1)
where, c' is the real part of the complex permittivity
and c
d
'' is the dielectric loss through energy
dissipation.
= 2tf is the angular frequency and c
0
the void
permittivity.
The Debye relaxation function allows writing the
complex permittivity of insulator material as:
r
f jf
f
/ 1
*
+
A
+

=
c
c c
(2)
( )

= A
f f
c c c
0
defines the difference
between the polarization under static field c
f0
, and
the polarization distortion c
f
. f is the frequency of
the electrical field and f
r
the relaxation frequency.
From the relaxation function, the real and imaginary
parts of the complex permittivity are given by:
( )
2
/ 1
'
r
f f
f
+
A
+

=
c
c c
(3)
( )
( )
0
2 2
/ 1
/
"
c t
o
c
c
f
r
f f
r
f f
+
+
A
=
(4)
The derivative of the Eq. (4) as a frequency
function, proves that the loss c'' is maximum at the
frequency of relaxation f
r
or at a relaxation time t
r
=
1/f
r
.
The Debye relaxation function proves that the
dielectric properties of materials can be described in
frequency domain or in time domain. Each description
domain presents its advantage and its inconveniency.
2.2 Dielectric Property Analysis in Frequency Domain
The variation of the electrical characteristics is
provided by the study of the microscopic properties of
the materials. For most types of soil, the real part of
the permittivity c' is constant for a frequency less than
2 GHz. The variation of the electrical conductivity o
depends on the frequency bandwidths. As a

1425
consequence, there is a need to choose the study
frequency in these respective frequency bandwidths to
detect anomalies in the soil. The spectra of the
material must be determined to highlight the
measurement frequencies. Devices for such
measurement are expensive and difficult to implement
in situ.
2.3 Dielectric Property Analysis in TDR
A system TDR can be broken up into a signal u(t)
injected into a cable represented by a function h
2
(t), a
probe and the medium tested, h(t) and a cable of exit
for recording h
2
(t). So Gemert [3] suggests, the
reflected signal r(t) is the convolution product given
by:
}

=
t
d t h t u t r t t ) ( ) ( ) ( (5)
The use of the Fourier transform results in EM
propagation within the frequency domain expressed as
follows:
) ( ) ( ) ( f U f H f R = (6)
It is derived that most part of the signal is valid for
a frequency in the range of minimum frequency f
min

and maximum frequency f
max
. f
max
is related to the rise
time of the signal which depends on the generator
capacity and on the pulse velocity in the cable. H(f) is
the transfer function of the probe in the soil.
According to Clarkson et al. [4], H(f) depends on soil
dielectric properties and the cable impedance and can
be obtained from the Eq. (6). Using the model as the
Debye model, the soil permittivity can be computed
for each frequency. Based on the spectrum of the
permittivity thus obtained and the use of the
Cole-Cole representation c' = f(c''), the four
parameters of the Debye model can then be
determined. The curve c' = f(c'') and these parameters
characterize the dielectric properties of the soil and
reveal any change in their structure and their texture.
Fig. 1 presents three types of responses of TDR
system corresponding to an open circuit where the
coefficient of reflection is 1, a short circuit where the
coefficient is -1 and the reflection of the signal in

Fig. 1 TDR response r(t) for three media.

water. The transform of Fourier applied to the such
curves gives the spectrum H(f) of the response of the
medium tested. The real part of the dielectric
permittivity, the relaxation frequency and the electric
conductivity of the medium tested are the parameters
selected to differentiate the mediums. While referring
to the works of Topp and Davis [5], Nadler et al. [6],
Dasberg and Dalton [7] and Heimovaara et al. [8], the
three parameters are directly determined from signal
TDR without using the spectrum.
2.4 Dielectric Property Measurement in TDR
Instead of seeking the transfer function of a probe
and a medium H(f), the TDR technique consists of
launching a high EM wave in the probe and
determining the transit time of this wave. A reflection
is created at the top of the probe and a second
reflection is created at the end of the probe. Since the
probe length is fixed, the propagation velocity can be
computed as twice the length divided by time required
for a pulse to travel along the probe and back. The
apparent dielectric constant of the medium is simply a
ratio between this measured velocity and the speed of
light in free space. The apparent dielectric constant
which is the relative real part of the permittivity can
be computed by:
2
2
2
|
.
|
\
|
|
.
|
\
|
= =
L
ct
v
c
r
c

(7)

TDR response
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
0 1.5 3 4.5 6 7.5 9 10.5 12 13.5 15 16.5 18
Time (ns)
R
e
f
l
e
c
t
i
o
n

c
o
e
f
f
i
c
i
e
n
t

Open circuit
Water
Short circuit

1426
where, t is the transit time, L the length of the probe
and c light speed in free space.
The load Z
L
of the transmission line embedded in
the investigated soil can be calculated from the
voltage reflection coefficient defined by:
0
0
Z
L
Z
Z
L
Z
+
=

(8)

where, Z
0
is the characteristic impedance of the cable,
The values can be read directly from the signal
curve for longer propagation time. On the other hand,
the load Z
L
can be calculated and converted to o
a
by
using probes geometric constant. The geometric
constant K
c
, is experimentally obtained by immersing
the probe in a solution of known salinity o
a
,
measuring the resistance of Z
L
by TDR, and using an
equation identical to that proposed by Rhoades and
Schilfgaarde [9]:
f
L
Z
ref c
K / ) 25 ( = o
(9)
where, f is temperature correction, o
ref
(25
o
C) is the
permeate electrical conductivity.
The signal TDR response results in a step pulse
which undergoes a transformation with the rise in time
corresponding to the maximum frequency that the
spectrum of the signal contains. The rising part of the
signal corresponds to a gain of energy similar to the
loading of a capacitor. At the half of the relaxation
time the 63% of the maximum loading of the
capacitor is attained. It is observed that the point of
measurement of the time or the frequency of
relaxation is at the 2/3 of the maximum gap between
the incident voltage and reflected voltage.
To map out the infiltration basin surface, standard
permeability tests based on TDR measurements are
carried out in laboratory to determine how electrical
parameters correlated to compacted soil permeability.
3. Devices and Soil Tested in Laboratory
3.1 Infiltration Pond
A pond of 4,800 m
2
of bottom surface with storage
volume of 44,800 m
3
is built at 10 km from Lyon
(France) to infiltrate the storm water runoff. It
receives water from a catchment of 119 ha and the soil
layer is glacio-fluvial alluvia. The water table depth in
this area is 13 m from the basin bottom. Fig. 2
presents a photo of the pond.
Clogging zones observed on the infiltration basin
bottom reveal the surface variation of the hydraulic
conductivity of the subsoil. Infiltration surfaces
without clogs and with clogs are shown in the
Figs. 3a and 3b. Sampling was carried out in such
zones for laboratory tests.
3.2 Soil Identification
Laboratory tests to determine geotechnical
engineering parameters were carried out on the
different types of soil extruded from the pond. The
results of the tests facilitate for the identification of
the samples of soils which characterize the geological
formation of the site including the state of pollution of
the sampling zones.
3.2.1 Grain Size Distribution
Unscreened samples taken from the basin are
defined within the 0/80 mm range in size. The color
and the texture of the samples prove that they can be
categorized as clean alluvia, polluted alluvia and
polluted sands. Three tests were carried out on each
type of sample. Grain size distribution curves are
shown in Fig. 4. The grain size distribution curves
presented in Fig. 4, show that the clean and polluted
unscreened soils (0/80 mm) are clean poorly graded

Fig. 2 One of the perspective of the infiltration pond.

1427

(a)

(b)
Fig. 3 Infiltration surfaces without clogs and with clogs:
(a) infiltration surface without clogs; (b) clogging zone of
the infiltration surface.

gravel. The sand is clean sand poorly graded. It can be
noted that the difference between the two gravels is
exhibited in the range of grain size 0.1 mm-10 mm
(Fig. 4). The three types of soils, however, contain
virtually equal quantities of fines with the sandy soil
having a slightly higher content. It is observed, from
the grains size distributions that the two alluvia have a
high coefficient of permeability. However, the value
of this coefficient depends on the unit weight of the
earth in place.
3.2.2 Compaction Tests
In order to determine the maximum bulk dry
density that can reach the layers of alluvia and of sand,
compaction tests were carried out on the three types of
samples. The values of the maximum dry density
(
dOPN
) with the corresponding optimal water contents
(w
OPN
) are given in Table 1.
0
10
20
30
40
50
60
70
80
90
100
0.001 0.01 0.1 1 10 100
Particle diameter (mm)
G
r
a
i
n

s
i
z
e

p
e
r
c
e
n
t

(
%
)
Clean gravel sample
Polluted gravel sample
Polluted sand sample

Fig. 4 Grain size distributions for the three types of soil.

Table 1 Compaction test results.
Tested sample w
n
(%) (
d
/
w
)
n
w
OPN
(%) (
d
/
w
)
OPN
Clean alluvia 2.6 1.990 5.8 2.060
Polluted alluvia 5.0 2.000 7.5 2.056
Polluted sand 7.8 1.666 10.0 1.700
w
n
: the natural water content of the soil; (
d
/
w
)
n
: the bulk dry
density at the natural water content; w
OPN
: the water content at
the maximal dry density; (
d
/
w
)
OPN
: the maximal bulk dry
density.

It can be noted that the two samples of the alluvia
have approximately the same maximum bulk dry
density but varying optimum water contents, where it
is higher for the polluted alluvia. Proctor curves of the
two alluvia have the same shape that of the polluted
soil is being shifted towards the higher compaction
water contents.
Polluted fine sand is characterized by a curve
stretched with the rather low value of maximum dry
density (1.700) and high corresponding water content
(10%).
The permeability of the compacted soils and the
dielectric properties are correlated based on TDR
measurements and permeability tests carried out in the
laboratory on compacted samples.
3.3 Permeability and TDR Measurements
In order to characterize the zones of permeability
restriction on the surface of the basin, the samples of

1428
soil are compacted in moulds at different water
contents and at standard proctor energy. After a TDR
measurement, the permeability test on the sample is
carried out.
3.3.1 Laboratory TDR Measurements
A TDR probe with five stems able to be inserted in
ground is used for the tests. Calibration in water and
in a saline solution was undertaken. A cable tester of
model Tektronix 1502C connected to a computer
allows the production of a signal which is recovered
by the units automatic acquisition program. The
results of the calibration are presented in Fig. 1.
Fig. 5 presents an example of the TDR signals
which allow for the determination of parameters that
characterize dielectric properties of soils. In this figure,
the stationery points of inflection indicated are
characteristic for the determination of the different
parameters. Point A indicates the exit of the signal
from the cable while point B corresponds to the
contact point between the probe and the soil. The
distance BD represents the duration of the rise of the
EM signal that traverses the soil. On the other hand,
point C, which is the inflection point of the rising part
of the curve, corresponds to the exit of the signal from
the probe. The time elapsed between A and C signifies
the transit time of the wave in the probe. Point D
represents the relieve signal towards the end,
indicating the time necessary for the loads to locate
before the application of the electric fields.
With the Tektronix 1502C cable tester, the
minimum and the maximum frequencies are
respectively given by:
xN
p
cV
f
2 min

(10)

and
ds dd
p
cV
x
p
cV
f
. 2
250
2
250
max

(11)

where, Vp is the rate of propagation of the EM wave
in the cable, (Vp = 0.659), dd is the scale chosen to
display the TDR curve, d
s
is the number of points
contained in the length of the scale (d
s
= 25).
When the scale in distance of 0.25 m is used, the
minimum and the maximum frequencies are:
f
min
= 15.8 MHz and f
max
= 3.95 GHz (12)
The TDR signal for the saline solution resulted the
estimated geometric constant of the probe: K
c
= 736.
The minimum and maximum values obtained for
the three types of samples compacted in the ring of the
test of permeability are given in Table 2.
In all cases, the following conclusions can be
derived from the results of the TDR measurement:
(1) The dielectric constant increases with the
increase in water content, increases when the dry
density decreases and is higher in sand than in the
alluvial soils;
(2) The relaxation time increases with the increase
in water content, decreases when the dry density
increases and is higher in sand than in alluvial soil,
with higher values recorded in polluted alluvial than
in clean alluvial soils;
(3) Electrical conductivity increases with the
increase in water content as well as with the dry
density and is higher in sand and in polluted alluvial
than in clean alluvial soils.
The correlation of these results with the coefficient
of permeability is to enable us to differentiate the

Fig. 5 Model of waveform to determine dielectric
parameters.

Table 2 TDR measurements after compaction test.
Values (%)
r
(ns) (mS/m)
d
/
w

Minimum 6 1.42 3.3 425 1.63
Maximum 26.5 2.12 14.3 958 2.06
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
19 21.5 24 26.5 29 31.5 34 36.5 39 41.5 44 46.5 49 51.5 54
Elapsed time (ns)
R
e
f
l
e
c
t
i
o
n

c
o
e
f
f
i
c
i
e
n
t
Srie1
A
B
C
D
E

1429
measurement points of the surface of the basin.
3.3.2 Laboratory Permeability Tests
The permeability of each material was given
according to the water content and dry density of the
material and a correlation between the permeability of
soil layers and TDR measurements was established.
Single values of water content on either of the dry side
and wet side were chosen on the Proctor curve of each
material in order to limit the number of the
permeability tests. The values of the coefficients of
permeability are then related to the values of the
dielectric parameters. To appreciate this correlation,
the various parameters are summarized in Table 3.
The results presented in Table 3 also show the
difficulty of saturating the three types of soils by
infiltration. But the values of the coefficient of
permeability thus determined are indicative of the rate
of the infiltration of the water in the basin. When the
dielectric parameters are compared to the coefficients
of permeability it can be noted that the coefficient of
permeability increases when:
(1) the initial water content decreases;
(2) the dielectric constant decreases;
(3) the electrical conductivity decreases;
(4) the relaxation time increases;
(5) the dry density increases.
Among the dielectric parameters which can be used
to locate the zones of weak permeability, the
permittivity and electrical conductivity have the
widest field of variation with change in the
permeability coefficient. These two dielectric
parameters can help detect in situ, the nature of the

Table 3 Comparison of the different parameters.
Soils Clean alluvia Polluted alluvia Polluted sand
wi (%) 5.1 7.4 5.0 9.0 7.8 14.5
wf (%) 10.2 7.8 9.7 10.0 19.8 15.2
Sr (%) 94.1 66.6 79.8 81.5 89.2 65.9
d
/
w
2.053 2.018 2.00 1.99 1.67 1.63
4.83 5.89 5.52 8.29 6.25 14.35
(mS/m) 454 425 515 668 804 958
r
(ns) 1.57 1.62 1.57 1.87 1.67 2.12
10
5
k (m/s) 5.1 2.4 4.7 1.5 6.1 3.1

soil from the range of the dry density obtained and by
giving the order of magnitude of the coefficient of
permeability of the ground in place.
4. In Situ Dielectric Parameter
Measurements
For reference and comparative analysis of the state
of compactness of the soil layer in the infiltration
basin, the measurements of dry density and water
content were also carried out using the gamma ray
probe method. To fix the measurement points, a grid
of the surface of the basin was drawn in plan form of
localization. Fig. 6 presents the map of the basin of
infiltration with the grid of measurements. The grid on
this chart is characterized by its shape in L, one side
considered the small and the other the large wing of
the chart for purposes of locating and describing the
results.
4.1 Bulk Dry Density Mapping by Gamma Ray
Measurement
Measurements of the dry density and of the water
content also allow an estimate of the permittivity of
the soil layer. It was noted that the values obtained for
the density varied from 1.4 to 2.2 and from 3% to
40% for the water content. But these values do not
make it possible to distinguish the clean alluvia from
the polluted alluvia.

Fig. 6 Presentation of the infiltration basin map.

1430
It was observed that in the small wing of the grid
corresponding to the north-west of the basin, the dry
density was between 1.4 and 1.7 while the water
content was very high from 20% to 40%. This means
that the soil layer of this zone is mainly made up of
sand. In the central zone, the dry density value is 1.8
and the water content varies from 3% to 7%. These
values characterize the nature of the alluvial soils.
The measurements carried out within the large wing
of the basin, in the north-south direction enable us to
appreciate the presence of various zones. Mainly, in
this regard, the existence of high to very high values
of density with the relatively low corresponding water
contents (between 5%-6%) and exceeding 8% in some
case, can be observed intermittently. The presence of
alluvia is noted in the greater part.
It is then necessary to compare the various
observations with the indicative results given by the
dielectric parameters.
4.2 Basin Surface Mapping by Dielectric Constant
Measurement
The varying colors of the grid points plotted in
Fig. 7 indicate the values of the permittivity within the
tested zone. The yellow color corresponds to the
lowest value (2 to 4), the red color (from 4 to 6), the
green color (high values of 6 to 8) and the blue color
(very high value greater than 8). The distribution of

Fig. 7 Distribution of the dielectric constant on the
surface of the infiltration basin.
the points of color in the north-west part of the basin,
covered by polluted fine sand, is characterized by
rather high permittivity (color red or green). It can be
concluded, according to the already established
correlations, that this is a zone of weak dry density
with high water content. The whole of the north-south
part of the basin (large wing of the basin), covered
mainly by alluvia, exhibits various values of
permittivity. The colors of the points of measurements
confirm the fact that the values of the permittivity are
low to medium. Comparison of these results with
measurements from the gamma ray probe indicates a
good correspondence between the values of the
permittivity and the dry density measured by the
gamma ray probe. One can thus use these
measurements of permittivity to appreciate the
compactness of the soil layer.
4.3 Basin Surface Mapping by Electric Conductivity
Measurement
The distribution of the values of electrical
conductivity is characterized by three level of color.
The green color corresponds to the low value (of 0 to
700 mS/m), the dark blue, to a median value (of 700
to 1,200 mS/m) and the light blue color to values
ranging between 1,200 and 1,300 mS/m (Fig. 8). In
the north-west and extreme east zones of the basin the
electrical conductivity lies between 1,200 and 1,300

Fig. 8 Distribution of the electrical conductivity on the
surface of the infiltration basin.

1431
mS/m. That corresponds to high values of electrical
conductivity. The central zone of the large wing of the
basin is characterized by low values of electrical
conductivity.
4.4 Basin Surface Mapping by Relaxation Time
Measurement
In Fig. 9, the yellow color represents a relaxation
time lower than 1.4 ns, the red color corresponds to
the values ranging between 1.4 ns and 1.6 ns, the
green color is for values laying between 1.6 ns and 1.8
ns and the dark blue color represents values higher than
1.8 ns.
The sandy zone located at the north-west of the
basin exhibits higher relaxation times analogous to the
water content, permittivity and electrical conductivity.
Some of the arrears within this zone, which have not
been polluted by rather recent fine sand deposits have
a relatively mean relaxation time. Measurements
indicate values of weak to medium relaxation times in
the central part of the northern edge of the basin. In
the large wing of the basin, located at the east and
extending from the north to the south, covered by
alluvia and punctuated by sands, the values of the
given relaxation times are generally close to medium
values, save for some exceptional points which are
associated with weak relaxation times.
5. Distribution of Permeability Coefficient
on Pond Surface
The correlations established in the laboratory,
between the dielectric parameters and the coefficient
of permeability of compacted soil, allow for the
determination of the corresponding distribution on the
surface of the infiltration basin. This distribution of
the coefficient of permeability allows for the
localization of the zones of low permeability which
delay the infiltration of the storm water which flows
in the basin. These zones undergo the deposit of fine
grains from which results surface clogging.
According to the results from the laboratory tests,
the values of the dielectric parameters facilitate for the
prediction of permeability coefficient varying from 2
10
-5
to 10
-4
m/s. The yellow color corresponds to a
coefficient of permeability lower than 2 10
-5
m/s,
the red color for the coefficients ranging between 2
10
-5
and 4 10
-5
m/s, the green color for the
coefficients ranging between 4 10
-5
and 6.1 10
-5

m/s and the dark blue color corresponds to
coefficients higher than 10
-4
m/s.
The distribution of the colors on the grid of the surface
of the basin proves that the most permeable zones are
more significant in the large wing of the basin while in
the small wing, the points represented by yellow, red
and green colors are more dominant (Fig. 10).

Fig. 9 Distribution of the relaxation time on the surface of
the infiltration basin.

Fig. 10 Distribution of the coefficient of permeability on
the surface of the infiltration basin.


1432
It is observed that the clogging can appear in the
small surface of the large wing of the basin. However,
the fact that these zones are surrounded by more
permeable zones indicates that stagnation of water and
delays in clogging can be partially mitigated.
Implicitly, the small wing of the basin undergoes
clogging after each rainy season.
6. Conclusions
The various tests performed in this study showed
that it is possible to associate the measurements of the
dielectric parameters with the permeability of a
compacted soil as has been reported by some
researchers. In addition to electrical conductivity, the
permittivity and relaxation time parameters can be
adopted for this estimate.
This study shows that it is possible to describe the
direction of variation of the coefficient of permeability
when the dielectric parameters vary. One can thus
know the evolution of the coefficient of permeability
by following the evolution of the dielectric parameters.
The determination of the distribution of the
coefficients of permeability on the surface of the
infiltration basin allows for the prediction of the zones
of stagnation of water after infiltration, depicting
zones that characterize clogging. Furthermore, TDR
measurements can be used for the detection of these
zones. Consequently, the permeability of these zones
can be improved by excavation before the rainy
season. These measurements can also be used for the
management of the infiltration basins for sustainable
flood mitigation measures.
References
[1] G.C. Topp, J.L. Davis, A.P. Annan, Electromagnetic
determination of soil water content: Measurement in
coaxial transmission lines, Water Resour. Res. 16 (1980)
574-582.
[2] O. Mazac, I. Landa, On determination of hydraulic
conductivity and transmissivity of granula acquifers by
vertical electrical sounding, Journal of Geological
Science 16 (1977) 123-129.
[3] M.J.C. van Gemert, High frequency time domain
methods in dielectric reflectrometry, Philips Research
Reports 28 (1973) 530-572.
[4] T.S. Clarkson, L. Glasser, R.W. Tuxwordth, G. Williams,
An appreciation of experimental factors in time-domain
spectroscopy, Adv. Mol. Relaxation Interact. Processes
10 (1977) 173-202.
[5] G.C. Topp, J.L. Davis, Time-domain reflectrometry
(TDR) and its application to irrigation scheduling, in: D.
Hillel (Ed.), Advances in Irrigation, Vol. 3, Academic
Press Orlando, USA, 1985, pp. 107-127.
[6] A. Nadler, S. Dasberg, I. Lapid, Time domain
reflectometry measurements of water content and
electrical conductivity of layered soil columns, Soil Sci.
Soc. Am. J. 55 (1991) 938-943.
[7] S. Dasberg, F.N. Dalton, Time domain reflectrometry
field measurements of soil water content and electrical
conductivity, Soil Sci. Soc. Am. J. 49 (1985) 293-297.
[8] T.J. Heimovaara, E.J.G. de Winter, W.K.P. van Loon,
D.C. Esveld, Frequency-dependent dielectric permittivity
from 0 to 1 GHz: Time domain reflectrometry
measurements compared with frequency domain network
analyzer measurements, Water Resourc. Res. 32 (1996)
3603-3610.
[9] J.D. Roades, V. Schilfgaarde, An electrical conductivity
probe for determining soil salinity, Soil Sci. Am. J. 40
(1976) 647-651.

Sustainable Methods and Techniques in Old and New
Public Constructions
Mauricio Noya
1
, Ana Lucia Torres Seroa da Motta
1
and Wagner Abreu
2

1. Department of Civil Engineering, Universidade Federal Fluminense (UFF), Niteri 24210-240, Brazil
2. Brazilian Institute of Geography and Statistics, Rio de Janeiro 20021-120, Brazil

Abstract: This article is about the methods and techniques used in the building retrofit in modernist style designed by the architect
Affonso Eduardo Reidy in the fifties, and also about the project of the new Rio de Janeiro MIS (Museum of Image and Sound),
under construction, by The North-American office Diller Scofidio + Renfro, both located in the city of Rio de Janeiro. These
constructions present a wide view of the concept of sustainability, not commonly used in public buildings in Brazil. In the building
designed by Reidy, which belongs to the State of Rio de Janeiro, the regeneration to host the new headquarters of the State
Department of Taxation and Finance respected its original characteristics and used smart and sustainable design techniques, as well
as recyclable materials. On the other hand, the project of the new museum at Copacabana Beach was entirely planned on sustainable
bases and it includes the application of building techniques concerning that purpose. This article presents a study of these two cases
and evaluates the techniques and materials used in one another. Observing the principles of sustainability in construction is a
pioneering initiative of the State Government. This article aims at making the scientific community aware of the needs of immediate
adoption of sustainability techniques in Civil Engineering, not only in the construction of new buildings, but also in the regeneration
of the existing ones, as well as the need of imperious adoption of sustainable techniques when it comes to maintainability.

Key words: Sustainability, building techniques, retrofit, public buildings.

1. Introduction

In most countries, the civil construction industry
corresponds to a great share of the national GDP
(gross domestic product). In Brazil, the productive
chain of the sector is responsible for more than 9.2%
of the GDP, corresponding to 40% of the total
investments and generating more than 10 million jobs
[1]. However, the civil construction industry is one of
the economy sectors which has great impact on the
natural environment [2]. This impact is due to the
great consumption of materials, about 75% of the
available natural resources, as it also occurs in the
United States [3, 4]. In countries such as Japan, it is
estimated that this consumption reaches 50% of raw
material inputs [5].
The sector is responsible for 12% to 16% of the

Corresponding author: Mauricio Noya, master, Ph.D.
candidate, research field: sustainability in civil engineering.
E-mail: mauricionoya@hotmail.com.
total water consumption, for 25% of the use of forest
wood, for 30% to 40% of energy, and for 40% of
extractive raw material. The sector is also responsible
for the generation of 40% of residues of all human
activitiesduring the execution of a construction,
15% of the materials become residuesamong these,
from 15% to 30% are deposited in landfills. The civil
construction industry is also responsible for the
production of 20% to 30% of GHEs (greenhouse
gases), like carbon dioxide (CO
2
), the main pollutant.
In the world, CO
2
emissions by the sector reach 3%,
mostly from the decarbonization of limestone during
the production of cement [6, 7]. In Brazil, CO
2
emissions reach 6% and the productive chain of the
civil construction sector represents 8% of the total
emissions of GHEs in the country [8]. A great share of
the burning of fossil fuels must be added to these
figures.
Another relevant fact in terms of environmental
DAVID PUBLISHING
D
Sustainable Methods and Techniques in Old and New Public Constructions

1434
impact is due to the buildings and how they work.
According to Watson [9], buildings are the greatest
energy consumers of the planet. It is estimated that at a
global level, the working of buildings is responsible for
25% of CO
2
emissionsincluding not only direct
emissions, such as the burning of fossil fuels for
environmental conditioning, water heating and cooking,
but also the indirect ones, such as the emissions
associated to electricity purchased by others [10].
Therefore, the civil construction industry represents
one of the most inefficient models of production and
consumption there is ISAIA [6]. In Brazil, despite the
growing concern in relation to environmental issues,
civil construction is still at the beginning of what we
could call the first stage of environmentally correct
production. In this respect, we can point out legal
documents such as Resolution n. 307/2002 by
CONAMA (National Environment Council in Brazil)
which sets guidelines, criteria and procedures for the
management of residues in civil construction [11], and
initiatives of sustainable construction like the two
buildings which are the objects of study of this paper.
Agopyan and John [10] state that there is no
sustainability without durability, in relation to the
lifespan of the components, it would be essential for
the classification of a construction as sustainable.
Hence, the importance of choosing durable materials
to be used in the constructions. When specified in the
project phase, they will also define in quantitative
terms the maintenance resources, as well as their
impacts. The CIBs (International Council for
Building) W82 Commission, upon giving concepts for
sustainable construction, and also taking into
consideration many other countries, has translated
sustainable as durable [12]. This placement meets the
concepts used in the refurbishing and maintenance of
the building designed by Reidy, one of the objects of
study of this article.
On the other hand, if all building projects demand
changes in the pre-existing natural systems as well as
the consumption of energy and inputs, which means
that a completely green project is not possible, the
ones oriented to sustainability represent an
opportunity of improving environmental performance.
Sustainability is a complex challenge, it comprehends
not only how the resources are obtained, how they are
used, but also, how to use them up to their full
potential, considering non-wastefulness [13]. This is
the case of the constructive project of the new MIS,
which has also included techniques for reverse
construction. According to the Technical Report 1 by
Casa do Futuro, associated with the U.S. Green
Building Council, the project has been developed in
order to achieve better results in terms of
environmental sustainability and the construction site
must follow criteria and a set of concepts which will
grant it the LEED (leadership in energy and
environmental design) seal, a worldwide reference in
sustainable certification in the last few years [9, 14].
The article draws a parallel between the
refurbishing of an existing building, which uses
sustainable techniques of environmental preservation,
and the construction of a new building, based on
environmental, economic and social criteria.
2. The Life Cycle of Buildings
In order to talk about preservation, demolition and
new building, it is necessary to understand the concept
of Life Cycle of Buildings, which comprehends: the
manufacturing of materials, their transportation to the
construction site, the construction itself, the durability
of the building, its maintenance and the demolition
and recycling of the materials [15].
The acquisition of sustainability conditions in the
three cases (preservation, demolition and building)
depends on the assessment of the environmental
impacts throughout these phases, such analysis is
called LCA (life cycle assessment) of the building, a
methodology of environmental management which
evaluates, through qualitative and quantitative
parameters, the potential impacts associated with a
determined final product, so as to make it possible to

1435
have control of these impacts, their reduction or
elimination [16]. The LCA, which has been used in
civil construction since the nineties, has become an
important tool for the evaluation of buildings in their
two forms of application: The first one focused on the
Whole Process of the Construction, and the second
one focused on the Building Material and Component
Combination [17].
According to Agopyan and John [10], the durability
of the materials directly influences the length of time
during which the building will render services, as well
as the amount of resources to be used in its
maintenance, defining environmental, social and
economic impacts. The longer the lifespan of the
building, the longer the preservation of the materials
and the environment due to the reduction of the speed
of flow of materials extracted from nature, which
serve as raw materials for the constructions. On the
other hand, if the longer the lifespan of the project,
greater the risk of obsolescence, with techniques of
sustainable recovery and the introduction of new
concepts of use and building maintenance, it is
possible to recover, use and manage the projects in a
sustainable way [10].
These new concepts demonstrate that despite the
importance of preservation during the life cycle, it is
possible to opt for the new techniques and build with
sustainability and social responsibility through reverse
construction, which becomes a sustainable demolition,
within the reuse of all material which otherwise would
be dumped in a landfill. The recycling of residues and
their utilization as raw materials is a measure of
fundamental importance for the preservation of
natural resources; not only does it collaborate to the
reduction of the consumption of energy, but also
guarantees the preservation of the environment [18].
3. Case Studies
The study refers to two individual constructions, an
initiative of the Government of the State of Rio de
Janeiro, which aim at only one goal: the sustainability
in civil construction. Respecting the characteristics of
the building, the first one (Fig. 1) proposed changes in
the pre-existing natural systems and also in the
consumption of energy and inputs. The retrofit project
has presented preservation material sustainable
techniques, as well as the substitution of equipment
and systems so as to acquire better environmental
development. The second case, the construction of the
new MIS, Fig. 2 was totally projected based on
sustainability principles. For that matter, it was
initiated undergoing the process of reverse
construction, having selective demolition, as well as
the recycling of the withdrawn material. The project
of the new museum anticipates the reduction of the
energy charge used in the heating, lighting and
cooling systems, with emphasis on indoor air quality.

Fig. 1 Retrofit project of the modernistic architecture
[19].

Fig. 2 The construction of the new MIS [21].


1436
3.1 The Old IPEG Building
The old IPEG (Instituto de Previdncia do Estado
da Guanabara) building is located on the corner of
Presidente Vargas Avenue and Andradas Street,
downtown Rio de Janeiro. The 22-story building
designed by Affonso Eduardo Reidy, became a
landmark in Rio de Janeiro architecture in the fifties
because it was the first building in the city to use a
metallic structure (Figs. 3 and 4) [20]. The modernist
building has its frontage facing south and west. As the
south face is almost the whole year in the shade, it is
totally made of glass while the west face, which is
subject to intense sun rays throughout the day, has a
brises soleil system which protects the building
against solar radiation (Fig. 5). This kind of forecast
in the project demonstrates the architects concern, at
that very time, about issues such as environmental
comfort and cooling efficiency, which can be
translated as energy efficiency, since the thermal
conditioning is one of the greatest villains, responsible
for the electric power consumption in the operational
phase of the building, as well as in electric power
consumption in the construction phase [22].
The construction that hosts the State Department of
Taxation and Finance has its own team of architects.
They have remodeled all the floors (Fig. 6) according
to new needs, this remodelling was performed by
EMOPa Public Construction Company. In
accordance with the new desires of the administration
and corresponding to the need of economic
operational results and maintainability oriented to
sustainability, the retrofit was really important for the
city, especially because it aimed at preserving an
awarded architecture. Because of this challenge, the
insertion process of new techniques and concepts
environment oriented was initiated.
The buildings are responsible for 42% of the total
consumption of electric power generated in Brazil.
Recent studies have shown that residential buildings
consume 23% of energy, commercial ones 11% and
public ones 8%. In office buildings, air-conditioning

Fig. 3 The construction in 1950, west facade [20].

Fig. 4 The construction in 1950 south faade [20].

Fig. 5 Model of the building, west and south facades [20].

1437
consumes 48% of energy [23]. For that reason, the
cooling system Fig. 7 was completely replaced. The
old system used to have a cooling tower, which makes
the cooling obsolete, causes a high consumption of
electric power and a great waste of water (through
evaporation and natural system loss), besides the high
cost of water chemical treatment required by this kind
of equipment. The new cooling system, Fig. 8 has
chillers with internal tanks made of serpentines where
the water circulates producing the cooling. These
internal tanks are not only responsible for water
economy, but also for stopping its evaporation,
providing greater energy efficiency and less loss, thus
making the system more economical, efficient and
sustainable.
Regarding the rational water use, it stands out the
replacement of all restroom units. In the original
project, in Fig. 9, the adopted system used pressure
valves, which are responsible for high consumption
and water waste. In the refurbishing, a Dual Flush
system was used (Fig. 10), it allows the liberation of
three liters of water for liquids and six for solids. Data
about the real economy of this system vary
considerably, however, using it, we can reach an
economy of 6.480 L of water per person annually.
That allows us to classify it as being sustainable [25].
The original faucets, Figs. 11 and 13 were also
replaced by faucets with timer system Figs. 12 and 14,
which limits the water flow and saves up to
40%.

Fig. 6 New layout of the floors.
Source: State department of taxation and finance architecture (2011).

Fig. 7 Cooling system before [19]. Fig. 8 Cooling system after [19].
FLOOR PLAN

1438

Fig. 9 Bathrooms before: pressure valve system [19]. Fig. 10 Bathrooms after: dual flush system [24].

Another relevant aspect was the implementation of
the selective waste collection. With 100% of the
materials of consumption properly dumped, the
recyclable waste receives selective treatment.
Selective trash cans were also placed in all areas of
the building, including the offices. The trash, which is
separated in specific bags, is collected by registered
and authorized non-governmental organizations for
the purpose of recycling. This measure is according to
Decree n. 5.940/2006, which issues the separation of
recyclable residues discarded by Public
Administration Entities, as well as their destination to
recycling [26].
Disposable cups were replaced by customized mugs.
Besides coming from fossil materials, they have
cancer-stimulating effects in their formula, with an
amount of styrene above the one recommended by the
Ministry of Health (20 ng/mL-L) [27].
The project prioritized suppliers that perform their
activities in a sustainable way, observing the life cycle
of their products. The supplier of the raised flooring
installed in all floors uses raw-material from the
plastic waste processor (Fig. 15).
The conventional solutions use aluminum plates
which weigh about 45 kg/m
2
. The 12 kg solution used
in the building is formed by recycled polypropylene
plates inlet on paving slabs. The covering of this floor
was made of carpet tiles which reduces waste
generation when compared to roll carpet: something
like 3.9% against 13.1% of the traditional roll carpets.
The strands of the carpet tiles are made of recyclable
materials, which eliminate the use of virgin
raw-materials, while the office furniture was made of
recycled material. The supplying company uses 100%

Fig. 11 Bethrooms before: urinals with original faucets [19]. Fig. 12 Bathrooms after: urinals with timer system [24].

1439

Fig. 13 Bethrooms before: original faucets [19]. Fig. 14 Bethrooms after: timer system faucets [24].

Fig. 15 Raised flooring and carpet tiles used installed in all floors [24].

reforested wooden planks, certified by FSC (Forest
Stewardship Council). Besides, it also has a program
of waste management, machinery which reduces the
consumption of energy and the emission of pollutant
gases, it also uses procedures in order to make good
use of daylight in the factory, it has selective waste
collection among other ecological initiatives that are
really important for the environment preservation.
Fig. 16 shows the original furniture replaced by a new
conception (Fig. 17).
Finally, the project has developed an alternative
transportation program, which provides parking
bicycle racks and creates special parking spots for
vehicles used for carpooling.
The refurbishing attempted at inserting new
techniques in the operational and maintenance system
in a consistent measure because they positively
interfere in the habits and behaviors, meaning an
improvement in salubrity conditions for the personnel,
the maximization of operational efficiency and
minimization of environmental impacts.

Fig. 16 Office furniture before [19].


1440

Fig. 17 Office furniture after [19].

3.2 MIS (Museum of Image and Sound)
The construction of the new museum of image and
sound is a partnership between Rio de Janeiro State
Government and Roberto Marinho Foundation. It is
the base of the sustainable revitalizing process of
Copacabana beach and we hope it will become a
reference in sustainability for the whole Brazilian
Patrimony and Culture field [9]. Diller Scofidio +
Renfro office was the winner of the contest
concerning the choice of the new museum
architectonic project, which, composed of seven floors
located on Atlntica Avenue, will take the space once
occupied by Boate Help, which was completely
demolished to give place to this new construction. The
project will be registered at USGBC (United States
Green Building Council), a non-governmental
organization created in 2007, and it may be the first
institution of the kind to receive the LEED (leadership
in energy and environmental design) certificate [9].
Fig. 18 illustrates the eletronic model of the museum
frontage.
Differently from the first case study, in which
obsolescence was overcome by specific strategies of
the project, such as the insertion of new techniques
and concepts based on sustainability, the sustainability
of MIS started to be thought since the project design
and in a integrated form. In order to be granted with
the LEED seal, an American design office specialized
in environmental consultancy was hired, as well as
another consultancy company from Rio de Janeiro [9].
The work, based on the sustainability of the
construction, has already started at the demolition day
of the old Boate Help building (Fig. 19). The sustainable

Fig. 18 MIS electronic model frontage [21].

1441

Fig. 19 Demolition of the old boate help building [28].

demolition (or reverse construction) tried to reduce
problems such as dust, noise and dirt [9]. This
different demolishing process certainly provided more
safety to the surrounding neighborhood. The principle
of reverse construction depends on the selection of
the in loco material. According to CONAMA
Resolution n. 307, all the material must be carefully
separated at the site [11]. Therefore, materials such as
wood, shingles, bricks, glass, aluminum, metals,
among other things that otherwise would become
debris, are carefully separated at the site. Despite
taking more time than the conventional discard, the
apparent initial loss of this process is worth it, since
the additional costs are avoided. The selective practice
of waste segregation allows better management at the
site, besides the commercialization easiness [10].
It was exactly what happened at the demolition
phase of the old building at the MIS construction site:
There was separation and selection of materials that
might be reused or recycled, either in the site or
elsewhere. Whatever would serve as raw material for
the making of new materials have remained at the
construction site, for example, the concrete plaques
which were ground for re-utilization in paving
material. While the recyclable or reusable material
was donated to NGOs and cooperatives in order to be
turned into other diverse products. This way
significantly contributed to the reduction of
transportation costs and emission of greenhouse gases,
besides helping those who make a living of this kind
of commerce [9].
This procedure is very relevant, considering that the
amount of waste generated in civil construction due to
demolitions is the same or at least superior to the
amount of urban garbage, approximately 400
kg/inhabitants/year [7]. Obtaining new materials from
recyclable ones means reducing the rate of discarding
and consequently, the exploitation of natural resources
for the transformation of raw material for the industry,
besides creating new business opportunities and jobs
for the population. Sustainable demolition creates
direct and indirect beneficiaries, environment, society,
workers involved (who take advantage of these new
techniques and acquire new sustainable habits), and
people who develop similar activities, like craftsmen
or garbage collectors. According to the United States

1442

Fig. 20 Evolution of the construction site [29].

Green Building Council, an outstanding
environmental undertaking is characterized for
minimizing or even eliminating its negative impact
on environment and its users. The council
evaluates the environmental performance of buildings
in five concepts: sustainable planning of the
constructed area, water economy and efficiency in its
use, power efficiency and use of renewable energy,
conservation of materials and sources of resources,
and quality of indoor environment [30]. The project of
the new MIS aims at better results in terms of
environmental sustainability and the construction site
must follow the same criteria which will allow the
achievement of the LEED certification. In order to
reduce the negative impacts in the environment, the
building strictly follows all the regulating
determinations for achieving sustainable development
in civil construction, in a pioneering initiative in the
city. Fig. 20 shows the evolution of the construction.
Forecasting the reduction of energy consumption,
the consultancies made a simulation that enabled them
to elaborate a thermal picture of the building, taking
into consideration solar radiation, relative humidity,
wind frequency, the localization (since it is across
from the sea, it makes the use of natural ventilation
difficult and it may damage some electronic
equipment used in the exhibitions of the museum);
and the climate of the city throughout the seasons of
the year [9]. According to Lodi [9], former
coordinator of projects in the Roberto Marinho
Foundation and current project supervisor at
IPHAN-Rio (National Historic and Artistic Heritage
Institute), criteria optimization of natural lighting, use
of ideal glasses, air conditioning systems that meet the
environmental comfort of the workers and users, solar
power for water heating and the transformation of the
air-conditioning water into energy, as well as a green
roof, are under study.
The project establishes new sustainability
techniques and promotes good practices that can be
applied to other cultural assets of the city, since it is
considered a pilot project by Rio de Janeiro State
Government, not only with the intention of giving
sustainability to the buildings but also with the
purpose of educating Brazilian society.
4. Conclusions
The objective of this article was to approach the

1443
sustainability in civil construction theme by reporting
clean techniques of pre-production, production and
utilization of resources used in both individual cases.
In the retrofit work of the modernist style building,
designed by the architect Affonso Eduardo Reidy, it
was demonstrated that it is possible to preserve old
and important buildings in a sustainable way taking
into consideration not only new concepts, techniques
and sustainable habits, but also efficient and more
economic equipments in their work, optimizing
resources utilization, reducing the CO
2
emissions and
providing environmental comfort for their users.
These concepts can also be applied to the construction
of new undertakings, as the case of the MIS. If the
construction is based on sustainability fundamentals
and on the eventual application of their principles in
the pre-construction or demolition phase, represent the
concretization of that which is an important
perspective for the civil construction and architecture
fields nowadays. The application of new project and
constructive techniques oriented to the preservation of
the environment make the involved professionals have
more responsibility, not only because of the cost of
the undertaking, but also because of the impact that
the new buildings cause to the environment. Studies
demonstrate that it is possible to adopt new
sustainability techniques, from the making of the
products to the construction site, with the reuse of the
materials, through the selective demolition of inputs.
Despite not reaching the total reuse in some cases, the
study shows that the waste which is not used may be
discarded in a correct way so as to do no harm to the
environment.
It is worth mentioning that, in spite of having clear
objectives to be accomplished within the
implementation of a cleaner production, there are still
some obstacles to be overcome, particularly when it
comes to cultural aspects, lack of information and also
the adherence to past habits and old behaviors.
Balancing social, economical and environmental
development is still an enormous challenge. Hence,
there is huge urge for commitment among
professionals and the society as a whole towards
sustainability in order to assure society a better future
with superior quality.
References
[1] CBIC (Brazilian Chamber for the Construction Industry),
Civil Construction: Analyses and Perspectives, 2010,
http://www.cbcs.org.br/ (accessed Oct. 6, 2013). (in
Portuguese)
[2] G.C. Grigoletti, M.A. Sattler, Environmental strategies
for Brazilian red ceramic industry in Rio Grande do Sul
State, Built Environment 3 (3) (2003) 19-32. (in
Portuguese)
[3] V.M. John, S.C. ngulo, S.E. Zordan, Sustainable
development and waste recycling in civil construction, in:
Proceedings of 4th Workshop on Sustainable
Development and Recycling in Construction, So Paulo,
2001. (in Portuguese)
[4] G. Matos, L. Wagner, Consumption of materials in
United States 1900-1995, Annual Review of Environment
and Resources 13 (1998) 107-122.
[5] Y. Kasai, Barriers to the reuse of construction
by-products and the use of recycled aggregate in concrete
in Japan, in: R.K. Dhir, N.A. Henderson, M.C.
Limbachiya (Eds.), Use of Recycled Aggregate, Tomas
Telford, 1998, pp. 433-444.
[6] G.C. Isaia, Civil Construction Materials and Science
Principals and Materials, Vol. 2, Ibracon (Brazilian
Institute of Concrete), So Paulo, 2007. (in Portuguese)
[7] V.M. John, Waste recycling in construction: A
contribution towards research methodology and
development, Habilitation Thesis, Polytechnic School of
the University of So Paulo, So Paulo, 2000. (in
Portuguese)
[8] McKinsey and Company, Ways towards Low Carbon
Emission Economy in Brazil, So Paulo, 2009. (in
Portuguese)
[9] C. Lodi, Social, Economic and environmental
sustainability in the new museums of Rio de Janeiro:
Museum of tomorrow, MARMuseum of art of Rio de
Janeiro and the new museum of image and sound, in:
Proceedings of 2th International Workshop on
Museography and Museum Architecture: Identity and
Communication, Rio de Janeiro, 2010. (in Portuguese)
[10] V. Agopyan, V.M. John, Sustainability Challenges in
Civil Construction, Edgard Blucher, So Paulo, 2011. (in
Portuguese)
[11] Ministry of Environment, 2002, CONAMA (National
Environmental Council of Brazil) Resolution Number

1444
307, July 5, 2002, Federal Official Gazette of Brazil 136,
Brasil, July 17, 2002, pp. 95-96. (in Portuguese)
[12] L. Bordeau, Agenda 21 on Sustainable Construction,
Rotterdam, CIB, 1999, p. 120. (in Portuguese)
[13] Doerr Architecture, Definition of Sustainability and the
Impacts of Building, Colorado, EUA,
doer.org/html/GreenChecklistResidential.doc (accessed
Jan. 1, 2012).
[14] Museum of Image and Sound, Technical report 01:
Worksite and Assembly Monitoring, Rio de Janeiro, June
20, 2011. (in Portuguese)
[15] S.F. Tavares, R. Lamberts, Energy consumption for
construction, operation and maintenance of residential
buildings in Brazil, in: Proceedings of Brazilian Meeting
on Built Environment Comfort, Brazil, 2005. (in
Portuguese)
[16] J.R.B. Chehebe, Product Life Cycle Assessment: ISSO
14000 Management Tool, Qualitymark, Rio de Janeiro,
[17] O. Ortiz, F. Castells, G. Sonnemann, Sustainability in the
construction industry: A review of recent developments
based on LCA, Construction and Building Materials 23
(2009) 28-39. (in Portuguese)
[18] V.M. John, S.E. Zordan, Research and development
methodology for recycling residues as building materials,
Waste Management Series 21 (3) (2001) 213-219.
[19] M. Coelho, Photographic Collection of the Latter IPEG
Building, State Department of Taxation and Finance, Rio
de Janeiro, 2011. (in Portuguese)
[20] N.G. Bonduki, Affonso Eduardo Reidy, Brazilian
Architecs, Editora Blau, Lisboa, 2000. (in Portuguese)
[21] Diller Scofidio + Renfro, 2011, MIS,
http://www.dsrny.com/ (accessed Jan. 1, 2012).
[22] J.W.M. Kaehler, M.B. Garcia, L.A. Pereira, Civil
construction industry, in: Proceedings of International
Workshop on Energy Efficiency, Campinas, 2006. (in
Portuguese)
[23] C. Escudero, News: Energy Efficiency, Construction &
Bussiness Magazine, So Paulo, Jan. 13, 2012. (in
Portuguese)
[24] D. Normando, Photographic Collection of Rio de Janeiro
State Department of Taxation and Finance, Rio de Janeiro,
[25] C.G.S. Dusse, Company Runs a Campaign for Saving
Water in Restrooms, Brazilian Newspaper O
DebateScience and Environment, July 3, 2008,
www.odebate.com.br (accessed Jan. 1, 2010). (in
Portuguese)
[26] Decreto n 5.940, It Defies That All Disposed Recyclable
Waste Must Be Direct and Indirectly Managed by the
Federal Government Not Only at the Generating Sources,
but also at Associations and Cooperatives of Recyclable
Materials Collectors, Brasil, Oct. 25, 2006. (in
Portuguese)
[27] P.A.P. Pereira, R.F.S. Oliveira, J.B. Andrade,
Determination of styrene content in polystyrene cups by
purge-and-trap followed by HRGC-FID, American
Laboratory News 36 (15) (2004) 16-18.
[28] Rio de Janeiro State Department of Taxation and Finance,
Architecture Department, Rio de Janeiro, 2011. (in
Portuguese)
[29] Photographic Collection, Rio de Janeiro State Department
of Culture 2010/2012, Rio de Janeiro, 2012. (in
Portuguese)
[30] An Introduction to the US Building Council and the
LEED Green Building Rating System, Washington DC,
USGBC (United States Green Building Council), Mar.
2002.

Social Participation in the Context of the Urban Public
Space Renewal: The Case of Lagarteiro Neighborhood
in Oporto
Paulo Tormenta Pinto
Department of Architecture and Urbanism, ISCTE-Lisbon University Institute, Lisbon 1649-026, Portugal

Abstract: In 1980, Giancarlo de Carlo wrote in the Perspecta: Yale Journal, An Architecture of the Participation. Carlo, proposed
a sequence of procedures: the problem definition, the solution and the results evaluation, in all the three steps, the social communities
should be integrated in as part of a share and dynamic process. This text corresponds to a new way of understanding the importance
of the social communities in the architecture definition. From de Carlos text, the main goal of this paper is to present the results of
an intervention in public space renewal of the Lagarteiro neighborhood (in Oporto city) and the relation of this process with the
participation of the resident population. The urban context of Lagarteiros neighborhood presented before the intervention the typical
and classic signs of disqualification of environmental and urban peripheral areas of resettlement. The term disadvantaged areas is
associated with these areas in urban or peri-urban case, was framed in a specific Portuguese policy, called Initiative Critical
Neighborhoods. Being a recent intervention, the renewal of the public space in Lagarteiro is a complete case study that allows
comprehending the participation phenomenon in nowadays, at the same time that seduces a reflexion about technical and
architectural solutions for critical neighborhood.

Key words: Participation, architecture, public space, Lagarteiro.

1. Introduction

As the democratic process in Portugal has matured
over the last 40 years (after the Carnation
Revolution, on the 25th of April 1974, that deposed
the dictatorship of the Estado Novo that had been
governing Portugal since 1933), the relationship
between architecture and participation has gradually
faded in the architecture debate. Some civic
movements have been more active and have made
their presence felt mainly through online social
networks, mostly as a way of expressing opinions
about works that have a strong impact.
Gone are the days when SAAL (Ambulatory
Support to Local Residents Programa program
introduced by Nuno Portas as Secretary of State for

Corresponding author: Paulo Tormenta Pinto, Ph.D.,
assistant professor, researcher, director, research fields:
architecture project, theory and history of architecture of the
21th century and metropolitan culture studies. E-mail:
paulo.tormenta@iscte.pt.
Housing and Urban Planning during the 1st
Provisional Government, following the Revolution of
1974) was where brigades of architects sat through
endless meetings alongside neighborhood committees
discussing direct interaction in the typological debate
about housing programs. It was in this context that the
objective of works was defined on the assumption that
when the basic needs of a section of the Portuguese
population were met in the so called PREC period
(Ongoing Revolutionary Process), it would lead to
housing being built with an unprecedented spirit and
methodology of participation. In the short period
between 1974 and 1976, when there was a ministerial
dispatch that allowed the work of SAAL to be
developed, the architects were forced to find
representational mechanisms to support their
arguments on the proposals made. One report that
stands out from that era came from Siza Vieira in the
documentary entitled Operaes SAAL [1] (SAAL
DAVID PUBLISHING
D
Social Participation in the Context of the Urban Public Space Renewal: The
Case of Lagarteiro Neighborhood in Oporto

1446
Operations) in which he states that a resident
committee came to his defense in the Bairro da Boua
in Oporto when some municipal technicians
questioned the oxblood color that had been planned to
paint some of the masonry. The residents were of
humble stock and have been seduced by the
architects discourse, they explained that the color in
question represented Siza Vieiras anxiety about the
work of Bruno Taut. Another extraordinary report was
shared by Manuel Vicente with Architectures
students at ISCTE-IUL and told of Bairro da Quinta
do Bacalhau in Lisbon. Vicente claimed that
everything was explained by the proximity with which
each person presented himself, given that it took more
than different financial resources to create barriers
between the architect and the future residents.
According to Manuel Vicente, the residents would
live with just as much affluence as the rich ones if
they could, and he suggested that this was what made
him come to the site in his red Alfa-Romeo, seducing
the support and intrigue of people who had asked him
for a house that was built in the knowledge that it
could be occupied by the architect himself.
The SAAL (Table 1) process contained the
momentum of the revolution and was linked to the
optimism of a new generation of architects that were
willing to do proximity work. To a certain extent, the
program implemented immediately after the PREC
saw its role as one of finding ways to bring the ideas
of equality advocated by democracy to fruition. In this
sense, there was a change from the methodology
adopted in Housing Development Fund (FFH)
procedures of the so-called Marcelist spring era in
the final years of the Estado Novo regime led by
Marcelo Caetano.
The decade that preceded the revolution of 1974
was characterized by the opening up of the economy
which triggered a surge in large scale construction,
also motivated by the Colonial War (which begun in
1961 with the insurrection in Angola). An attempt was
made to respond to the need for housing and the
effects of the rural exodus which was already very
significant at that time. This was when large
consortiums
1
were responsible for localized
construction with the aim of providing an alternative
to the city centers that were already saturated and
unable to meet the demands of the population. The use
of innovative construction techniques allowing the
rapid completion of large scale constructions was the
common denominator for this wave of optimism among
everyone involved at the time. This was when the
structural systems of reinforced concrete appeared, and
in particular the prefabricated industry. In its idealism,
the fascination with technology contained a
mechanistic conviction in tune with European
development and growth in the aftermath of the
Second World War and the application of the Marshall
Plan.
With the end of the dictatorship of the Estado Novo,
the flux from country to city intensified and this
coincided with the arrival of Portuguese citizens
returning from the violence that accompanied
independence in the Portuguese overseas territories at
the time. After Portugal became a member of the EEC
(European Economic Community) in 1986, the first
signs of some control were seen to tackle the heavy
legacy of the conflicts resulting from the installation
of democracy. European adhesion meant that public
practices had to be more tightly controlled and
procedures therefore became more institutionalized,
this was accompanied by a steady change in the direct
relationship between architects and communities. This
was also linked to the political pressure to resolve the
problem not only of settlements of clandestine
housing but also the illegal construction that was
building up in the periphery of the cities in general.
The 1980s and 1990s in particular saw an
exponential growth in construction and it included
both the supply of housing by private enterprises that

1
Consortiums such as: Soltia in the Carnaxide area, J Pimenta
in Pao dArcos, Icesa in Santo Antnio dos Cavaleiros. These
are just some of the most striking examples in the area of
Metropolitan Lisbon.

1447

Table 1 Main national housing programs in Portugal after the revolution of 1974.
Program Period Description
SAAL 1974-1976
SAAL (Ambulatory Support to Local Residents) was a program implemented following the military
coup of 25 April 1974, by the Ministry of Social Equipment and the Environment, and of Internal
Administration of the 1st Provisional Government, when Nuno Portas was the Secretary of State for
Housing and Urban Planning. This program that figured until 1976 was created to provide support for
people that lived in precarious conditions, SAAL was a decentralized service to build new houses and
infrastructures, through a wide participation process
PIMP 1987-1993
PIMP (Middle Term Intervention Program) was a municipal program implemented before the entrance
of Portugal in the European Union in 1986. Through this program, new urbanizations of social houses
were implemented, in large interventions in the periphery of the main Portuguese cities
PER 1993/2003-()
PER (Special Re-housing Program) had the main goal to give conditions to the municipalities to
proceed to the eradication of the slums and consequent re-housing of the inhabitants, in houses of
controlled costs. This program was created in 1993 to Lisbon and Oporto metropolitan areas. This
program was coordinated by the INH Housing National Institute (Instituto Nacional de
Habitaonow called IHRU)
IBC 2005-2013
IBC (Critical Neighborhoods Initiative) was a program launched in 2005 with the aim of
compensating for some of the shortcomings in the previous processes mostly in terms of participation
of inhabitants. The aim of this program, coordinated by the IHRUHousing and Urban Renewal
Institute (Instituto da Habitao e da Reabilitao Urbana), was to act in several urban areas with
critical factors of vulnerability

took advantage of easier access to credit, and also
work done by the State and the Local Authorities with
the aid of the European cohesion funds. This expansion
as a whole was also based on the opening up of new
fronts of housing as a result of the new infrastructures
developed as part of the European process.
As a result of shortages and the need to resolve
precarious living situations of certain social sectors,
the Central Administration took measures in the
terrain that were rapidly implemented, doing away
with all proximity and mediation work with future
residents. This phenomenon was most apparent during
the 1990s when PER (special re-housing programa
program supported by the Portuguese government
with European funds to eradicate all the slums
existing in Portuguese metropolitan areas in the last
decades of the twentieth century) was the mechanism
used to frame the process (Table 1).
Nevertheless, the change in the life styles from the
shanty towns to the generalized verticalization
imposed by the new neighborhoods benefited from the
fact that the residents now adapted more easily to the
urban context, this had not been the case following the
revolution in 1974 when the populations bond with
their rural roots was still very strong. Another
situation that contributed to the peaceful
implementation of this process was the application of
diverse architectural models with more generous
designs and typologies (normally based on the
right/left model) and the importance given from the
outset to the care of public outdoor spaces. The belief
that the whole process would be redeemed through
discipline, in some cases, extended beyond the need to
encourage any advance mediation with future
residents, though later the management of the units
was assured usually by means of municipal
companies set up particularly in local authorities to
work in proximity with the people who had been
re-housed
2
.
2. Giancarlo de Carlo in An Architecture of
Participation versus Architecture Cannot
Be Put to the Vote! by Manuel Graa Dias
While the first steps were being taken that
unleashed the strong urbanization process, in 1980
An Architecture of Participation by de Carlo
(1919-2005) [2], rubbed salt into the wound of
architects and political power in relation to the
participation issue. As an active member of Team 10,
he believed that an architectural operation (involved)

2
Cases of GEBALIS in Lisbon and GOP in Porto.

1448
three phases: the definition of the problem, the
development of the solution and the assessment of
the results. The users should be present throughout
the entire operation and this should reflect directly
and indirectly on the whole process, in other words:
each stage of the operation would be like a phase of a
project, the use would also be interpreted as a
phase of the operation and, therefore, of the actual
project, if the different phases merged, the operation
would cease to be linear, unidirectional or
self-sufficient.
Giancarlo de Carlo emphasizes the importance of
the programmatic definition in particular, and it is
noted that the architect should be involved in the
project at this stage when the specific objectives of the
operation are defined and a plan can really be drawn
up that meets the needs of the community of users. To
a certain extent his principle contrasts with the central
administrations programmatic definition which is
pre-defined by legislative mechanisms or the typified
mediation work with the communities. Carlo
underlines the importance of adjusting to the
specificity of the cases, this position is set out clearly
in 2002 in one of his last interviews, when he stated:
It is necessary to find the right path in each case.
There is no one way that can be written down and
applied to all situations. Some practice participation in
this way, but it is not participation, it is demagogic
bait. We must invent [3]. The phase in which the
problem is defined is therefore seen as part of the
actual project insofar as the objectives of the operation
and the resources made available become the subject
of discussion with the future users.
The readiness to have a head-on encounter with the
future users implies that the planners are willing to
bring all divergences to the table so that they can
explain their contradictions, and at the same time take
advantage of the explosive potential in the
resolution, or at least in the attempt to find a
resolution to all the misunderstandings and conflicts.
According to de Carlo, this procedure will stimulate
the search for the most suitable and carefully
considered solutions. The phase in which the architect
comes up with the solution is therefore the
production of final and unalterable solutions, but
extracting solutions from an ongoing encounter with
the people who will use his work.
The idea of anarchy can be found in de Carlos
discourse, he considered that true anarchy believes in
human energy and emphasized that anarchy is often
misunderstood, once anarchists are always described
as the people who put bombs in theaters. De Carlo
was interested in the urbanist trends at the turn of the
century, which was set aside by zoningthe
authoritarian trend which was mainly German and has
a capitalist stamp, and he noted Patrick Geddes in
particular and the Tel-Aviv plan as the reference of
urbanistic anarchy. This ethical and idealistic vision
recognizes the power of the people as a strong force
that can intimidate all ideological dimensions of
political power the use of criticism stimulates direct
intervention, and undermines the fixed principle of
delegation once and for all.
According to de Carlo:
Political power, even left-wing, doubts the validity
of the architecture of participation because it is scared
by it, it extends to criticizing the encouragement of
direct intervention, undermining the fixed principle of
delegation once and for all.
Local authorities hesitate because participation
tends to lead to conflict, it confirms an intention to
intervene in future programs, makes them accountable
for what has already been done and they find it hard to
accept and understand negotiations, transactions and
compromises.
Public administrators and technical staff are against
participation because it implies detailed analysis,
re-composition and promotion. It breaks the routine
and upsets bureaucratic immobility.
The professionals are against participation because
it destroys the hidden privileges of the specialization,
reveals professional secrecy, uncovers and exposes

1449
incompetence, multiplies the responsibility and
changes them from the private to the social sphere.
Academia is against it because participation
invalidates the systems on which teaching and
research are based [2].
Carlo also manifests the acceptance that
architecture is an ongoing process, setting aside the
idea of the closed work and advocating direct and
continuing action with communities, this is clearly
visible in his participative experiences in the Detailed
Plan for Rimini (1970-1972), in Bairro Matteotti
(1964-1974) in Terni and the fishermens houses in
Mazzorbo (1980-1986) in Venice. Carlos humanist
dimension is in tune with a low-tech approach that
everyone can easily understand, like that of the
Australian boomerang, that he compares to a
computerIt is made with simple technology, the
Australian native knew what it was, how to make it, to
carve it, to refine it: Perhaps if he took off another
millimeter it would fly better.
Carlos position is to some extent consistent with
that of Manuel Graa Dias in Architecture can not
be put to the vote! the editorial of JA (Jornal dos
Arquitectos) in an issue entitled Ser Populista (Being
Populist). However, the underlying assumptions and
the reflection on the topic are not exactly the same.
Firstly, when intercepting the two positions, we find
two distinct temporal contexts and a social
organization based on the democratic achievements
of the right to vote. The vote serves as a barometer of
balance and defines the ideological line manifested
by Central and Local Administrations. Dias article
gives an account of what happened during a debate
in the Association of Portuguese Architects in
relation to a project by Manuel Aires Mateus and
Frederico Valssassina for Largo do Rato, in Lisbon.
At the end of a heated discussion on the validity, or
not, of the new building to be placed on the corner of
Rua Alexandre Herculano and Rua do Salitre, Nuno
Teotnio Pereira got to the heart of the matter,
clearly, succinctly and elegantly, according to
Teotnio:
I come from the days when we were forbidden to
discuss things, to exchange ideas and even to have
meetings. So I am delighted to be here, to listen to
and debate topics on architecture; but be careful!
Discussing architecture does not mean putting
architecture to the vote! Each and every one of us has
the responsibility to discuss, to exchange opinions, to
understand the opinions of others, to agree or
disagree; but the responsibility for the approval of
the projects lies with the architects, the technicians
and the official departments that the law provides for
this purpose. Architecture can not be put to the
vote [4]!
Nuno Teotnio Pereiras undoubtedly lucid position
was formulated in light of his experience as a political
leader and active architect in processes governed by a
strong proximity principle like that of the
abovementioned SAAL for example.
It is the populist dimension that Dias strived to
isolate in participative processes. Populism that is
always one step away from being realized when, in
relation to architecture we come across sentences or
thoughts like: The City belongs to the citizens, and it
is for them to decide what they want for their city,
according to Graa Dias, these debates have
half-truths that may exalt meetings but, in fact, if
taken literally can prove completely useless and
inconsequential. This position does not refute the
importance and use of public debates because they are
what feed the project.
This openness does not however take away the
architects role as mediator, he is responsible for
understanding, interpreting, processing, adapting,
displaying and delivering in ways that include,
comprehend and enable them as well as other
unmentioned reasons and doubts that are brought to
the surface in the future. The moment when this
whole process is summed up culminates in the
interaction with technicians from the official
departments that the law provides for this purpose,

1450
and ultimately they are the ones that have been
legitimately elected to represent the interests of the
community in an informed manner.
Dias balanced position is reinforced by the
limitations imposed by the norms and regulations
which, from the start, guide and limit much of what
an architect does. On the other hand, considering all
the limiting effects imposed by the norms, namely
with regard to architectures freedom to
experiment, we can not ignore the fact that this legal
root is necessarily the result of a democratic process
discussed by the peoples representatives
in its forumthe Assembly of the
Republic.
However, the proximity between the populations
and their political representatives has not yet been
assessed. It is necessary to reflect on the quality of a
democratic process which, in the case of Portugal,
goes back 39 years, this has been the root of solid
public institutions, throughout this period, they have
managed legal instruments so that they themselves
have become the place for mediation between
designers and citizens, and the proximity ties
between them are filtered by suitably established
procedures.
This situation converges with the critical debate
mostly among architects since the 1990s. Disciplinary
matters are essentially at the core of the debate.
Post-modern designs created and strengthened an
architectural culture that, to a great extent, was
supported by historical recognition, which in turn,
allowed the quality of project work to be
benchmarked and valued. This incidental hermeticism
does not pertain only to architects, nevertheless, it can
be said that specialized jargon requires the definition
of limits that only encompass the discourse of each
disciplinary field, be it in the scope of architecture,
sociology, economics, technology or art. The
professional and/or scientific corporations that have
gained shape within the democratic context have
contributed greatly to this.
3. IBC (Critical Neighborhoods Initiative),
Participation Criteria in the Context of the
Urban Public Space Renewal of Lagarteiro
Neighborhood, Borough of
CampanhaOporto
The Secretary of State for Land Use and Town
Planning launched the Critical Neighborhoods
Initiative (IBC) in 2005 (Table 1) with the aim of
compensating for some of the shortcomings in
processes prior to participation in the promotion of
housing. The aim of this new program, which was
coordinated by the Housing and Urban Renewal
Institute (IHRUInstituto da Habitao e da
Reabilitao Urbana), was to act in several urban
areas with critical factors of vulnerability. Thus, three
priority areas were defined: the Cova da Moura
neighborhood (in Amadora municipality), the
Lagarteiro neighborhood (in Oporto municipality)
and Vale da Amoreira (in Moita municipality)
[5].
Specifically in the case of the Lagarteiro
Neighborhood, it was advocated that the whole
residential area and public space was in need of
complete renovation, thus providing a response to the
concrete problem of renovating the extensive
municipal estates built mostly in the 1970s, which was
in an appalling state of degradation just like
Lagarteiro at the start of this century (built in two
phases: 1973 and 1977). The urban isolation of these
areas had led to significant delinquency problems and
the buildings and public space reflected the lack of
conviviality among residents (Fig. 1).
The Critical Neighborhoods Initiative (IBC) was
also a response to renovation dynamics that stands in
contrast to all the other above-mentioned urban
planning programs, for which new construction is the
common denominator. But this program not only
provides an innovative solution within its field of
action, it also includes a renewed participative
methodology. It is the very concept of participation
that provides the slogan for the preliminary diagnostic

1451

Fig. 1 Lagarteiro neighborhood in the 1970s and in 2009 where can be seen the degradation of the public space and
buildings (photos from the municipality archive and by PTP).

report on the neighborhood, coordinated by Teresa S
Marques, with the title: Operation Lagarteiroan
intervention based on participation [3], and it was the
starting point for the architectural intervention and
consequent work with the population.
From the outset, the report defined an idealist
approach to the type of urban and social organization,
which strived above all to describe the collective
reas as clearly urban, qualifying them as places
where people could be, could meet and move about
(squares, alleys, parks, etc.), and that the people
using them would remember and feel an empathy
for. It followed an urban model, similar to that of
the standard town, and its structure was defined by
conventional typologies of urban design, valorizing a
network of proximity services and local trade within
the perimeter of the neighborhood, as well as small
collective support facilities for collective and family
life.
Just as in the originals plans, Lagarteiro
neighborhood is based on a principle of functional
separation, widely used at the time, which was
basically a simple residential zoning model, supported
by two facilities: a primary school and a sports area.
In time, and aimed at extending the range of social
support facilities available, several spaces were
allocated to Private Social Charity Institutions
(IPSSInstituies Privadas de Solidariedade Social)
where support services were provided for children and
the elderly. In this way, unused areas of buildings,
such as basements, were taken over and adapted for
their new purposes.
To follow up the intervention program being
prepared, a partnership protocol was established

1452
which would implement the intervention and funding
plans and management models during the process. To
this end, a local management model was defined,
which included: An executive committee (responsible
for the overall results of the intervention program and
ensuring that all the planned initiatives were suitably
coordinated), a monitoring committee (of an advisory
nature to ensure synergies were created among all the
local partners), an inter-ministerial work group (with
follow-up functions, including representatives of the 8
ministries involved in the IBC) and a project team
(comprising IHRU technicians and responsible for
implementing the action program with the
population).
The proximity work carried out by the so-called
IHRU project team is extremely revealing within the
intervention context. This team, which worked
directly with the community, was set up in a
residential unit within the Neighborhood
3
. It is
through this team made up of social workers that a
plan of activities was developed to build citizenship
among the resident population.
Lagarteiro was estimated to have 1892 residents, all
of whom had very little schooling. Although the
average age was roughly 35 years, only 6.4% of those
over 15 had completed secondary education [6]. The
number of unemployed was also very high, i.e., about
16.6%. The activities carried out with the population
were based on a number of actions targeting the
general population, but focused more on the younger
age groups, trying to bring them closer together
through participation in a variety of workshops, such
as: journalism, drama, photography, music and others
related to sports or training for inclusion. Special note
goes to the project O Meu Bairro na Cidade (My
Neighborhood in Town), for example, which took
place in November 2010 and involved the children
from the after-school recreational center, and
Lagarteiros Pre-School Social Center, these children
were encouraged to get to know their neighborhood so

3
Block 9, Door 152, Home 22.
as to make a model of the buildings and the
surrounding area. The activities included
interpretative readings and representations of the built
up area (Fig. 2).
All this social work was coordinated by specialized
teams with a specific social focus, as a result, the local
management model separated the proximity actions
with the population from the actions related to
interventions in the public and built areas of
Lagarteiro Neighborhood. This separation meant that
Office of Public Works of the Oporto Town Council
(GOP-CMP) managed the whole urban regeneration
process for the buildings, it launched three public
tenders using the consultation, by direct adjustment
model: one for the urban layout project of the public
space, and the other two for the work to be carried out
on the buildings.
Specifically in relation to the intervention in the
urban layout of the public space, which we will deal
with more details, the architectural decisions were
backed by criteria described in the preliminary
intervention program defined by the GOP-CMP, this
filtered the above-mentioned Operation Lagarteiro
diagnosis from a technical point of view. The
projects were also coordinated by the different
departments of the Oporto Town Council, which
gave their expert opinions leading to the licensing of
the projects.
Within the defined parameters, there was room for
debate which allowed certain proposals to be included
that were made by the team of architects (coordinated
by the author of the present text) and established in
light of specific intervention criteria. The fundamental
principle was the need to connect the neighborhood to
its surroundings, introducing it into a network of
urban flows of metropolitan interaction. The aim of
this principle was to overcome the negative effect of
its peripheral location, this was aggravated by
geographical factors that involved the Lagarteiro
Neighborhood in the same system as the long
Campanh valleyA geo-morphological fact that

1453

Fig. 2 Workshop My Neighborhood in Town, with children from Lagarteiros Pre-school Social CenterModel of the
buildings and the surrounding area of the neighborhood, November 2010 (photos taken from the IHRU activities report).

separates the Oporto municipality from the Gondomar
municipality (Fig. 3).
4. Architectural Intervention in the Public
Space of Lagarteiro Neighborhood
There is no real idea of urban center in Lagarteiro
Neighbourhood. Instead, there are long rows of
adjoining buildings along the streets, some set back,
others set forward, or misaligned within the different
lots, creating a diversity that could only be understood
if looked at with the green area near the Towns
Parque Oriental as a whole.
In its genesis, the neighborhood seemed to have a
humanist dimension that could be seen in the type of
buildings. Even as a project-type, the rows of
buildings strive to establish an archetype idea of
house, manifest in the sloping rooftops, and the
brick facing used in several lots.
Inadequate municipal access deprived the
neighborhoods poorer residents of normal social
integration, this was worsened by the somewhat
impermeable configuration of the urban setting as
neither the structure itself nor the proximity of
neighbors fostered social interaction.
The inner area of Lagarteiro was divided into two
sectors corresponding to the two periods of
construction (1973 and 1977) and they were separated
by topographic differences. Although physically close,
these two areas were only randomly connected by
rough paths prior to the intervention. The poor flow of
traffic was also apparent, in most cases, the same road
was used to enter and exit the neighborhood which
became like a cul-de-sac where it was difficult to
circulate freely. As a result, it was isolated and
non-residents were pushed away or discouraged from
using the public areas, this in turn meant that these
areas were only used by residents which also
contributed to the social fragility found in this
Problematic Neighborhood in Oporto.
The basic question underlying renovation project of
the public area in the Lagarteiro Neighborhood
involved the redefinition all the Neighborhoods links
with road network of the Oporto metropolitan area,
making it more permeable and more in line with
Oportos Municipal Master Plan; the main aim of the
Qualification of the Land Charter of the urban planning


1454

Fig. 3 Location of the Lagarteiro neighborhood close to the Campanh valley that in the past segregated the neighborhood
from the city center (photo by Google Earth).

instrument is to create a series of streets crossings to
connect the Lagarteiro Neighborhood with a Hub for
Urban Structuring and Inter-municipal Articulation.
Anchoring the neighborhood with its surroundings by
means of the planned streets was essential to the start
of the urban regeneration process.
However, the project was based on a new system
of streets linking two sectorial areas in Lagarteiro
Neighborhood (connecting Rua Diogo de Macedo
with Alameda Arq. Carlos Ramos). This hub was
considered essential to articulate the internal
connections with its surroundings (Fig. 4). The new
street also permitted pedestrian access for people
with limited mobility. Hence, it was the area running
through the center of the neighborhood that required
most expertise, particularly for the construction of a
series of walls for ground contention that would
control the irregular topography of the Lagarteiro
Neighborhood and stabilize the area of the two
existing facilities: the sports court and the school
(Fig. 5).
A very limited range of materials was used
consisting mainly of yellowish granite for the
pedestrian areas and grey for the roads. This material
was chosen due to the cost and ease of supplying it in
large quantities, it was also easy to maintain by a
relatively unskilled workforce. Another consideration
was the way in which the stone ages, resilient to wear
and tear and any inappropriate use. A 2.5 m mesh was
used throughout the neighborhood to highlight the
joints between the granite cubes. The control base of
the entire project was in the use of this mesh as it
ordered the different compositions in terms of the
design of the pedestrian areas and car parks.

1455

Fig. 4 The new road in Lagarteiro (foto by Ins dOrey).

Fig. 5 Detail of the pedestrian ways in Lagarteiro (foto by Ins dOrey).

The very marked topographic differences
characterizing the neighborhood meant that a series of
support walls had to be put in place. These support
walls were built with the help of cofferdams prepared
specially for the purpose. The final appearance of the
bare concrete is defined by triangular sections of
high-relief which, together, create a kinetic effect
when exposed to the light (Fig. 6). As a whole, this
solution created a strong expressive identity which
was fundamental in a regenerative process but, thanks
to its texture, also one that could prevent the negative
impact of the most common acts of urban
vandalism.
5. Conclusions
The urban intervention in the Lagarteiro
Neighborhood is an example of participation fostered
by social work in which the architectural project
serves as an instrument, as part of a policy defined
upstream (in the diagnosis of the Operation
Lagarteiro), it sought to fit in with the general criteria
of the intervention. On the other hand, the entire

1456

Fig. 6 Detail of the concrete walls, in Lagarteiro, done with special cofferdams (foto by Ins dOrey).

institutional structure involved in developing the
architectural projects stipulates regulatory
mechanisms that themselves guide and limit the
general lines of the intervention.
This underpins the architects involvement in a
broader process of recognition and proximity in
relation to the community of residents and in this way
indirectly places the architectural intervention in the
participative context.
It could be said that the intervention process in
Lagarteiro is in line with Nuno Teotnio Pereiras
vision when he claimed that but the responsibility for
the approval of the projects lies with the architects, the
technicians and the official departments that the law
provides for this purpose. Architecture can not be put
to the vote!
The establishment of democracy institutionalized
human relations, stipulating intervention criteria that
are duly oriented and systematized in line with good
practices. In the contemporary context in particular,
the debate on the role of the architect should be open
to parallel processes. Although the process described
above Lagarteiro intervention involves a pacific space
for the development of projects, the architects role is
to some extent peripheral in direct relation to the
transformation advocated for this urban settlement.
The more lateral positioning protects both the
architect and the project itself in the different phases
of the process. And yet this implies a withdrawal from
direct contact with the people, this in turn implies a
loss of contact with the arguments that foster a more
idealistic approach to how human activities in the
territory can be organized
4
.
The anarchy described by Giancarlo de Carlo for
the purposes of an Architecture of Participation thus
becomes an important base in the reflection on the
architects role in urban and social regeneration
processes. This role is perhaps even more relevant for
interventions in inherited territory which require a
decodifcation of the space into different scales, and
the respective crossing of this procedure centered on
morphological matters with other less tangible social
questions.
References
[1] J.A. Bandeirinha, The SAAL Process and the
Architecture in the Period of the April 25, 1974, Coimbra
University, 2007.
[2] G. Carlo, An Architecture of the Participation, Critical
Theory of Architecture in the Twentieth Century,
Caleidoscpio, Casal de Cambra, 2011, pp. 753-758.

4
See the minutes of the architects association in the Statute of
the Portuguese Architects Association.

1457
[3] J. Piza, The Participative Experience of Giancarlo de
Carlo, Interview with Giancarlo de Carlo at His Studio in
Milan, Nov. 2007.
[4] M.G. Dias, Architecture is not a Voting Matter!, Being
Populist, JA 234, 2009.
[5] P. Tormenta, Absent peripheryTwo cases of urban
housing clusters of controlled cost in the Contemporary
Portugal, in: Proceedings of SILACC 2010, So Carlos,
Brazil, Aug. 29-Sep. 1, 2010.
[6] Data from the Government Portal,
http://www.portugal.gov.pt/pt/GC17/Governo/Ministerios
/MAOTDR/Documentos/Pages/20061002_MAOTDR_D
oc_Bairros_Criticos.aspx (accessed May 20,
2010).


Analyzing Methods of Using Sustainable Development
Indicators
Fani Samara, Olga Christopoulou and Athanasios Sfougaris
Department of Planning and Regional Development, University of Thessaly, Volos 38334, Greece

Abstract: Sustainable development indicators, aim to measure sustainable development over longer periods of time. The sustainable
development indicators are grouped into 10 subject categories: socioeconomic development, sustainable consumption and production,
social inclusion, demographic changes, public health, climate change and energy, sustainable transport, natural resources, global
partnership and finally good governance. In this paper, the authors examined a number of researches and compared the methods they
used in order to determine the advantages and disadvantages of each method. The methods used a large number of different indicators.
The comparison of the indicators could be made through mathematical models and also through theoretical approach.

Key words: Sustainable development indicators, methodology, urban planning.

1. Introduction

In Chapter 40 of Agenda 21, the development of
indicators for sustainable development is called for. In
particular, countries at the national level, international
governmental and non-governmental organizations at
the international level are requested to develop the
concept of indicators of sustainable development in
order to identify such indicators. This issue was raised
during the first two sessions of the CSD (Commission
on Sustainable Development), at which time a large
number of countries emphasized the urgent need for
these indicators. Other countries expressed some
concern and insisted that indicators would be
developed in cooperation with governments. Pursuant
to the schedule adopted by the commission at its first
session, the progress achieved on developing these
indicators, in the context of Chapter 40 of Agenda 21,
will be discussed by the commission during its third
session. The objective of this schedule is primarily to
make the indicators for sustainable development
accessible to decision-makers, at the national level, by
defining them, elucidating their methodologies and

Corresponding author: Fani Samara, engineer, research
fields: regional planning and sustainable development. E-mail:
fanisamara@gmail.com.
providing training and other capacity-building
activities. Indicators, used in national policies, may
also be used in the national reports of the CSD and
other intergovernmental bodies [1].
A large number of researches have been taken place
in the entire world about the sustainable development
indicators. A lot of researchers use the analytic network
process for criteria selection which is the most
widespread nowadays. Multi criteria analysis methods
can help to optimize the power of criteria and indicator
approach to have a comprehensive analysis for
sustainable development. Also, another method which
uses a lot of researchers is the GIS and the
multi-criteria decision analysis. Through GIS could
construct clusters and rasters and finally compare the
results for choosing the best scenario [2].
2. Methodology
In this paper, a number of researchers and methods
were analyzed and compared with four case studies.
Specifically, the first case study has been talked in the
paper analyzing the role of sustainable development
indicators in accounting for and constructing a
sustainable Scotland [3]. The second case study has
been discussed in the paper a method to assess the
DAVID PUBLISHING
D
Analyzing Methods of Using Sustainable Development Indicators

1459
relevance of sustainability indicators: Application to
the indicator set of the Czech Republics Sustainable
Development Strategy [4]. The third case study has
been given in the paper environment and sustainable
development indicators in Lebanon: A practical
municipal level approach [5], and the fourth case
study has been talked in the paper perspectives of
sustainable development in countries of southeastern
Europe [6].
2.1 Case Study I
The first case study has been talked in the paper
analyzing the role of sustainable development
indicators in accounting for and constructing a
sustainable Scotland.
The main objective of the above paper was to
analyze the impact of sustainable development
indicators on the integration of sustainable
development into the governing of Scotland. A major
concern was whether an accounting technology could
represent this complex multi-dimensional and
interdisciplinary concept. They analyzed the
relationship between the official sustainable
development strategy of the Scottish Executive and the
associated indicator set using an analytics of
government framework. They observed a lack of
alignment between these sustainable development
indicators and the visions, fields of visibilities, forms of
knowledge and techniques of government contained in
this strategy.
The empirical evidence used in this paper was drawn
from a wider longitudinal analysis where they applied
the analytics of government framework to a number of
key strategic data and sustainable development
indicators reports produced by the Scottish Executive.
This paper was based upon a critical reading of
Choosing Our Future and related documents and
websites. Choosing Our Future was chosen to be the
focus of this paper because it was the most
comprehensive of all the sustainable development
strategies.
The documents were analyzed in four stages: First, a
cognitive map of the content of each document was
created; Second, these individual maps were reviewed
to derive a common cognitive mapping structure and
each document was remapped using this common
structure; Third, the maps were represented using key
elements of the analytics of government framework.
These elements were: the objective and vision of the
document, problematisations contained within, fields
of visibility, techniques of government, forms of
knowledge and identity formation; Fourth, these
revised maps were overlaid and interpreted using the
characterization of practices and rationalities of
government.
In the research a multilayered analytical framework
has been used and this section will be structured
according to these different layers. The analysis and
evaluation of the data is presented as follows: Firstly,
the analysis of the sustainable vision, problems and
actions of Choosing Our Future has been presented;
Secondly, the analysis of vision, visibilities and
dividing practices of the sustainable development
indicators has been presented; Thirdly, the alignment
of forms of knowledge in Choosing Our Future and the
sustainable development indicators have been analyzed;
Fourthly, the alignment between the sustainable
development indicators and Choosing Our Futures
technologies of government has been analyzed; Finally,
the summary of the different layers of analysis has been
presented.
The paper was motivated by the possibility that
sustainable development indicators could negatively
impact upon the integration of sustainable development
into the everyday governing of Scotland. A major
concern was the effectiveness of accounting
technologies on representing this complex
multi-dimensional and interdisciplinary concept. These
concerns were grounded in findings from prior
research into social and environmental accounting,
sustainable development indicators and govern
mentality studies of accountancy [3].

1460
2.2 Case Study II
The second case study has been discussed in the
paper a method to assess the relevance of
sustainability indicators: Application to the indicator
set of the Czech Republics Sustainable Development
Strategy.
The goal of this article is to critically review the
state-of-the-art in assessing the quality of sustainability
indicators and contribute to the development of a
suitable methodology for that. They start with a broad
review of a vast body of work in this fieldin both
practice and academic research. They show that both
scientists and practitioners have sought developing and
using methods for assessing quality of the indicators.
They have usually defined some criteria for that,
however, neither science nor practitioners provided
major support by developing reliable as well as
practical and operative methods for indicator
assessment. Therefore, they propose an innovative new
method for indicator assessment from the perspective
of their relevance. They operationalize this criterion
and apply it to the environment-related indicators from
the set used for the evaluation of the Czech Republics
Sustainable Development Strategy.
In this context, by success they mean massive spread,
frequent use and high popularity. Therefore, they
delimited the domain of their interest to providing a
method (tool) for indicator developers and users so that
they can readily identify what is pertinent and what is
not. They narrowed the wide choice and potential
application of the assessment criteria down to one of
them: relevance. In accordance with the findings of the
literature survey shown in the previous sections, they
started with a notion that relevance is a term used to
describe how pertinent, connected or applicable
something is to the indicators.
The criteria and the indicators used in this research
are public transport (transport intensity), use of energy
(energy intensity, primary energy supply, share of
renewable energy), climate change (greenhouse gas
emission per capita, greenhouse gas emissions per
GDP), use of resources (material consumption), waste
treatment (material use of waste), agriculture
(consumption of mineral fertilizers, consumption of
pesticides, organic farming), forestry (defoliation),
biodiversity (bird index) and environmental protection
(expenditures on environmental protection).
There has been no agreement or consensus on a
common set of scientific and management criteria for
evaluating or assessing indicators from several points
of view. The results of the assessment were statistically
processed and discussed. The discussion investigated
the suitability of the proposed procedure and the
recommendations stemming from the experimental
assessment conducted for further work on the progress
reports [4].
2.3 Case Study III
The third case study has been given in the paper
environment and sustainable development indicators
in Lebanon: A practical municipal level approach.
In close coordination and collaboration with the
Lebanese Environment and Development Observatory
at the Ministry of Environment, the Faculty of Health
Science at the University of Balam developed an action
plan to establish a decentralized environment and
sustainable development monitoring network through
local authorities using agreed upon environment and
development indicators. This effort was part of a pilot
regional project involving six Mediterranean countries
with funds provided by the Ministry of Environment,
Greece through the University of Athens to establish a
Mediterranean Environmental Reporting, Monitoring
and Information System for Mediterranean countries.
According to a well-defined selection process, a total
of 44 municipalities out of more than 700
municipalities belonging to three unions representing
three different regions in Lebanon qualified for
inclusion in the project with 17 choosing to participate
in MED-ERMIS-Lebanon. Meetings with concerned
public officials at all levels were then held to ensure the
support of all relevant governmental institutions and

1461
therefore the success of the project. Applying a
participatory approach, local government officers and
representatives from relevant ministries, universities,
research centers and non-governmental organizations
were invited to attend the different workshops at the
municipality level to produce the appropriate lists of
indicators. A total of 110 indicators were generated and
grouped into four major categories adopted by the
national indicator system: (1) population and
socio-economic; (2) economic activities; (3)
environment; (4) sustainable development activities
and policies. Indicator data sheets, a door-to-door
questionnaire and a municipality archive survey
were developed for data compilation, calculation and
presentation, and a total of 6,250 questionnaires were
filled (20%). Field surveying was carried out by trained
teams using the Multistage Cluster Systematic
Random Sampling technique and data analysis. Data
is being currently interpreted, and a user friendly
software application hosted on the University of Balam
and web-site to facilitate the updating of indicators and
the dissemination of indicator data on the World Wide
Web has been developed allowing access to
information by all interested parties [5].
2.4 Case Study IV
The fourth case study has been talked in the paper
perspectives of sustainable development in countries
of Southeastern Europe.
Countries of SE Europe are at the very beginning of
planning their development in accordance with
theoretical concept of sustainable development.
Sustainable development is defined as the basic
imperative and the only model of development in a
longer period of time. The basic indicators of
sustainable development represent a very useful and
qualitative way for measuring and monitoring the state
of sustainable development, as in every country
individually as well as in the regions and globally as a
whole. Values of the indicators of all four subsystems
in observed countries are within expectation. Also the
current level of development of every country
separately is taken into account and in accordance to
that, it the current position of any country in respect to
its international requests is defined. The special value
results from the comparison of obtained results of the
state in the developed EU countries, according to
which there is an impression of equal level of
sustainable development in France and Germany.
Whereas, the data regarding the level of sustainable
development in Greece and Slovenia has been
improved between two analysis. Taking into account
the results of the previous and current research there is
a clear need for regular observation of every indicator
individually but also the need to observe the
relationship among them as a whole. Special attention
needs to be paid to the intensive monitoring of
indicators of economic and ecological subsystem. The
research showed, among other things, the fragile
relationship between values of indicators of economic
and ecological subsystem, so that any change in values
of one indicator consequently leads to changes in the
final result. Both conducted researches point to the fact
that the value of the mentioned indicators changes at
the fastest pace [6].
3. Discussion
A different model was used in each described
method. Table 1 presents a comparison of used
methodologies.
In the first case study, the researchers used the
longitudinal method. One of the advantages of this
research is that this program is a governmental program,
so there are a lot of data and collaboration between the
organizations. Also, the researchers collected data,
every year, so they can compare their results. On the

Table 1 A comparison of used methodologies.

Case studies Methods
Case study I Longitudinal analysis
Case study II Practice and academic research
Case study III
Indicator data sheets
Door to door questionnaire
Municipality archive survey
Case study IV
Comparison between developed and
undeveloped countries

1462
other hand, despite the containing number of radical
elements there is a lack of an overarching concise
conceptual definition of sustainable development in the
program Choosing Our Future.
In the second case study, the researchers present a
practice and an academic research. The advantages of
this research are that they used keywords for analyzing
the results, so they can compare them. Also, they
distinguish between the relevance of the theme of an
indicator and the relevance of the indicator itself for the
theme. Moreover, they made a pretest research.
However, besides lack of operationalization of the
validation criteria, another weakness of the indicator
assessment methods proposed so far has been the fact
that they have often been qualitatively based on expert
evaluation which hampered their reproducibility.
In the third case study, the research is part of a pilot
regional project involving six Mediterranean countries.
Also, there is collaboration between the organizations
and the Ministry of the Environment. Moreover, the
collection of the random sampling was different for
each method. In addition, this research would allow the
municipalities to contribute the compiled indicator data
to the database of the Ministry of Environment in the
future. However, there are some difficulties to the
research, for example, Lebanon lacked baseline data,
both on the national and local scales and decisions are
taken based on either outdated or erroneous baseline
information. Nevertheless, community-based projects,
more specifically those introducing new
methodologies for planning and development,
encounter resistance and challenges from stakeholders.
Difficulties were encountered in obtaining clear
commitments from stakeholders in relation to
attendance, participation and involvement, mostly due
to the lack of knowledge about indicator science,
limited resources and the concern of the level of work
that needs to be undertaken. In addition, the absence of
the culture of indicators exacerbated by the lack of
baseline data has also led to drawbacks in the surveying
method ranging from securing the needed percentage
of respondents to the absence of archival data, to the
disinterest by municipal employees in providing the
available data for the municipal archive survey.
In the fourth case study, the goal of the research was
the comparison between developed and undeveloped
countries. In order to provide more precise results the
weight coefficients in range from 0 to 100 were used.
Also, this research used new values of indicators, but
the methodological approach was the same as the one
used in a previous research in 2006. The results of
research in the countries of Southeastern Europe, and
their comparison with the state of indicators in
Germany and France lead to significant conclusions.
However, the research that dealt with the similar
problem (monitoring, comparison and determining the
degree of importance of chosen indicators of
sustainable development in countries of Southeastern
Europe) has been performed by the authors at 2006. In
that way, the authors were able to conduct a
comparison with the previous results. The researches
that deal with the comparison of the overall state of
sustainable development in countries of Southeastern
Europe with some of the most developed countries of
the European Union have not been found either, so the
authors used only their research results for comparison
[7].
4. Conclusions
In above articles, a different method is used as they
aim for a different goal. Also, in the paper different
approaches of the indicators due to the level of
development of the case study country were presented.
Additionally, observe that the countries with higher
economic growth use more complex methods and
indicators than the less economic developed countries.
Moreover, the researches which use a mathematical
model can calculate more complex indicators. In the
other side, the researches which have used
questionnaires and data collection are more
representative.
All methods need data collection that is one of the

1463
most difficult parts of each research. Also, the four
models use different criteria as well as
indicators.
In this paper, the authors compare different methods
about the indicators and criteria. The goal of the paper
was to create a new model which will take place in
Greece. In the new model which would be constructed
the collection of the data would be made through
questionnaires and statistics data from the Ministry of
Economics. Also, the indicators would be separated in
three categories (economic, environmental and
demographical). Moreover, the results of the research
would be set in maps through graphs so that the
comparison would be easier. Finally, with the help of
GIS raster images will be constructed.
Acknowledgments
This research has been cofinanced by the European
Union (ESF (European Social Fund)) and Greek
national funds through the Operational Program
Education and Lifelong Learning of the NSRF
(National Strategic Reference Framework)Research
Funding Program: THALES. Investing in knowledge
society through the European Social Fund.
References
[1] M. Golusin, O. MunitlakIvanovic, Definition,
characteristics and state of the indicators of sustainable
development in countries of Southeastern Europe,
Agriculture, Ecosystems and Environment 130 (2009)
67-74.
[2] C.C. Boggia, Measuring sustainable development using a
multi-criteria model: A case study, Journal of
Environmental Management 91 (2010) 2301-2306.
[3] S. Russell, I. Thomson, Analysing the role of sustainable
development indicators in accounting for and constructing
a sustainable Scotland, Accounting Forum 33 (2009)
225-244.
[4] T. Hak, J. Kovanda, J. Weinzettel, A method to assess the
relevance of sustainability indicators: Application to the
indicator set of the Czech Republics sustainable
development strategy, Ecological Indicators 17 (2012)
46-57.
[5] M. Nader, B. Salloum, N. Karam, Environment and
sustainable development indicators in Lebanon: A
practical municipal level approach, Ecological Indicators
8 (2008) 771-777.
[6] O. MunitlakIvanovic, M. Golusin, S. Dodic, J. Dodic,
Perspectives of sustainable development in countries of
southeastern Europe, Renewable and Sustainable Energy
Reviews 13 (2009) 2079-2087.
[7] J. Wilson, P. Tyedmers, R. Pelot, Contrasting and
comparing sustainable development indicator metrics,
Ecological Indicators 7 (2007) 299-314.


Economy of RHA (Rice Husk Ash) in Concrete for
Low-Cost Housing Delivery in Nigeria
Olonade Kolawole Adisa
Department of Civil Engineering, Obafemi Awolowo University, Ile-Ife 220005, Nigeria

Abstract: Delivery of housing units in adequate quantity and quality has been the pursuits of individuals and successive governments
in Nigeria. Still, the gap between demand and supply is becoming wider on daily basis due to exorbitant cost of building materials that
is beyond the reach of average Nigerians. Concrete being the most acceptable construction material is expensive due to high cost of
cement. Efforts made to reduce the cost of cement were to no avail, thus there is need to look elsewhere. Past researches showed that
substituting cement with 15% of RHA (rice husk ash) in concrete improves the performance of concrete and reduces global warming as
a result of emissions from cement production. This paper looks into the effect of using RHA as partial substitute of cement on the cost
of a low-cost housing unit. Results showed that cement based construction materials are responsible for 82.58% of the total cost in
which cement is responsible for 42%. When 15% RHA was used to substitute cement over N90,000 was saved, amounting to about 7%
of the total cost of the building. In addition, depletion of natural resources was reduced.

Key words: Housing delivery, pozzolan, rice husk ash, global warming, low-cost housing.

1. Introduction

Acquiring a befitting housing unit at an affordable
cost is the desire of every man, irrespective of his
location and social status. Housing serves as shelter
and haven against unfavorable weather conditions,
protection from dangerous animals and affords us the
opportunity of raising our family. To others, it serves as
one of the best indicators of a persons standard of
living and his or her place in the society [1]. These
features and others have made the demand for housing
to skyrocket.
The limiting factors to owning a home includes some
of the following: cost of land acquisition, construction
cost, style, design, purpose and use. These however,
have little or no effect on the desires of the elites, the
rich and the well-to-do in the society on developing and
owning property. In recent years, home ownership for
the middle and low income earners of the society is
turning into a mirage as building materials,
construction costs and other factors have constantly put

Corresponding author: Olonade Kolawole Adisa, M.Sc.,
research field: cementitious material. E-mail:
makkolawole@yahoo.com.
housing development at a very high cost.
In Ref. [2], Mabogunje assesses housing situation in
Nigeria and put existing housing stock at 23 per 1,000
inhabitants and also put housing deficit at 15 million
houses which will require N12 trillion to finance.
Assessing this output shows that 43 persons will
occupy a housing unit of two-bedroom flat which is
about 21 persons per room.
The problem of qualitative housing has been a
concern for both the government and individuals.
Appreciating these problems, both public and private
sector developers make effort through various
activities to bridge the gap between housing supply and
demand, but the cost of building materials, deficiency
of housing finance arrangement, stringent loan
conditions from mortgage banks, government policies
amongst other problems have affected housing delivery
significantly in Nigeria [3]. Other factors include lack
of fund or misappropriation of available fund,
bureaucracy on land acquisition, continuous
rural-urban drift and geometrical increase in
population.
Concrete is one of the most versatile construction
DAVID PUBLISHING
D
Economy of RHA (Rice Husk Ash) in Concrete for Low-Cost Housing Delivery in Nigeria

1465
materials because of its excellent resistance to water,
fire, cyclic loading, ease of maintenance and its
potential to be formed into various shapes and sizes.
Cement, being the only manufactured constituent of
concrete, is the most expensive component which
makes concrete for housing construction unaffordable
by average Nigerians. Seeking alternative cheap
materials that have been considered as waste in partial
substitute of cement in concrete will reduce cost and
minimise environmental hazards associated with
production of cement and also ensure sustainable
environment which is the goal No. 7 of the MDGs
(millennium development goals).
In this paper, effort is made to look into the state of
housing as well as cement production and consumption
in Nigeria. The effect of high cost of cement on the cost
of building is viewed. An attempt has also been made
to assess the utilization of agro-by-products as partial
substitutes of cement and their advantages. Finally, a
case study of a two-bedroom flat is used to substantiate
the relative decrease in cost of building if RHA (rice
husk ash) is used as pozzolan in concrete.
2. State of Housing Delivery in Nigeria
Adequate housing delivery in Nigeria is still an
illusion despite efforts being made by successive
governments to face out what could be considered as
national embarrassment.
The National Rolling Plan of 1990-1992 estimated
housing deficit at 4.8 million. The 1991 housing policy
estimated that 700,000 housing units are to be built
each year if housing deficit is to be cancelled. The
documents indicated that not less than 60% of the new
houses are to be built in urban centres. In 2006, the
Ministry of Housing and Urban Development declared
that the country needs about 10 million housing units
before all Nigerians can be sheltered.
Between 1975 and 1980, there were plans to deliver
202,000 housing units to the public but only 28,500
units, representing 14.1% was achieved. Also, out of
200,000 housing units planned to be delivered between
1981 and 1985, only 47,200 (23.6%) was constructed.
Under the NHF (National Housing Fund) programme
initiated in 1994, to produce 121,000 housing units, it
was believed that less than 5% was achieved [4, 5]
Of recent the Federal Government created a new
Ministry of Lands, Housing and Urban Development
with a view to tackling the housing deficit in Nigeria
put at 54,500 housing units across the country [6].
What could have been responsible for these shortfalls?
One factor that is always being attributed to the
failure of the past programmes and policies is lack of
enough funds to finance them. This can be linked to
high cost of construction materials in which cement
takes a lion share.
3. Cement Consumption in Nigeria and Its
Attendant Effect on Adequate Housing
Delivery
In Nigeria, annual cement consumption was put at
19.5 million metric tons out of which 9.5 million tons
are produced locally [7]. As a developing country with
much construction work to do, our cement demand will
even go beyond this projection in view of our ever
growing population and desire for shelter and other
infrastructure.
Presently, more than 60% of our local cement
consumption is imported. This situation is a potential
threat to our yearning for rapid infrastructural
development and desire to meet the housing need of the
nation estimated at 14 millions.
Despite the increase in Nigerias cement
consumption which has increased from 8 million tons
in 2001 to 11 million tons in 2006 and now 19.5 million
tons, the per capital cement consumption in Nigeria is
still low when compared to some other countries. For
instance, South Africas cement consumption per
capita is 280 kg which is about four times that of
Nigerias 75 kg. China has 610 kg, Mauritius 600 kg,
Saudi Arabia 900 kg and Angola 105 kg [8].
Even at that, cement prices in Nigeria remain among
the highest in the world. The price of 50 kg (a bag) of

1466
cement has moved from N625 in 2002 to between
N1530 and N1800 in 2010 depending on the location.
The price of cement in the country is said to be four
times the price in Egypt and South Africa and ten times
the price in the United States. Such situations required
interventions.
In order to bring down materials costs and stimulate
construction, as well as make housing more affordable
to the Nigerian population, the government should
continue its reconsideration of restrictions on the
importation of cement and other building materials.
Another alternative we could explore in the long-run is
to conduct more research on how we can use local
materials such as clay and other local building
materials [9].
4. RHA as Pozzolan in Concrete for Low Cost
Housing and Environmental Sustainability
It is known that the incessant generation of solid
waste materials represents a serious problem. For that
reason, it is very important to study and develop any
technology, procedure or method that may help to
exploit them efficiently.
At present, pozzolanic materials are, in most cases,
industrial by-products and solid wastes that can be used
as active additions to portland cement. This is due to
their capacity for reacting with CH (calcium
hydroxide), produced during the hydration of the
portland cement. It is well known that the hydrated
compounds formed during the pozzolanic reaction
improve the performance of new cements elaborated
with them [10, 11].
The incorporating agricultural wastes, once calcined
at high temperatures, as pozzolans of high activity,
such as rice husk ash, have been studied with positive
results in the manufacture of mixed cement. Other
materials like sugar cane straw ash, rice husk ash, and
silica fume were used as pozzolans [12, 13]. Ketkukah
and Ndububa [14] establish that groundnut husk ash is
a potential pozzolanic material. These ashes are
characterized by their very low density and very
high volume.
Similarly, for a variety of reasons, the concrete
construction industry is not sustainable. Firstly, it
consumes huge quantities of virgin materials which can
remain for next generations. Secondly, the principal
binder in concrete is portland cement, the production of
which is a major contributor to greenhouse gas
emissions that are implicated in global warming and
climate change. From 1880 to 1996, the worlds annual
consumption of portland cement rose from less than 2
million tons to 1.3 billion tons. Thirdly, many concrete
structures suffer from lack of durability which may
waste the natural resources.
In Nigeria, cement is expensive, whereas, United
Nations Millennium Development Goal Number 1 is to
eradicate extreme poverty and hunger. Since more
than a billion people still live on less than US$1 a day,
there is a need for low cost and locally available
substitutes for portland cement. Therefore, if a
community obtains materials locally, their labor can
usually be considered as an in-kind cost to the projects
sponsors[15]. Hence, people can afford to build houses
at relative low costs.
The use of RHA in concrete was patented in the year
1924. Up to 1978, all the researches were concentrated
on utilizing ash derived from uncontrolled combustion.
Later, it was discovered that the type of RHA which is
suitable for pozzolanic activity is amorphous rather
than crystalline. Results of researches conducted on
RHA in concrete, as reported in Ref. [16], showed that
RHA is responsible for:
increased compressive and flexural strengths;
reduced permeability;
increased resistance to chemical attack;
increased durability;
reduced effects of ASR (alkali-silica reactivity);
reduced shrinkage due to particle packing, making
concrete denser;
enhanced workability of concrete;
reduced heat gain through the walls of buildings;
reduced amount of super plasticizer;

1467
reduced potential for efflorescence due to reduced
calcium hydroxide.
On what amount of RHA will be adequate to be used
as partial replacement of cement to cause the potential
asset, Habeeb and Fayyadhin [17] investigated the
effects of concrete incorporating 20% RHA as partial
replacement of cement at three different particle sizes.
They found that the compressive strength of the
resulting concrete was on the increase. The same
percentage of 20% was recommended by Oyekan and
Kamiyo for low cost housing development [18]. In the
same vein, Oyetola and Abdullahi favored an optimum
replacement level of 20%. In the works of Zhang and
Mohan [19] and Mahmud et al. [20], 10% RHA and
15% of RHA optimum replacement were recommended,
respectively.
5. Cost Assessment of a Two-Bedroom Flat
Using RHA Concrete
BEME (Bill of Engineering Measurement and
Estimates) of a 2-bedroom flat shown in Fig. 1 was
carried out. The amount of concrete and sandcrete
blocks required for complete construction of the
housing unit was estimated.
The quantity of cement to be used for the production
of sandcrete blocks of mix ratio 1:6 and that of concrete
and mortar of mixes 1:2:4 and 1:5 respectively were
estimated. Then cost comparison analysis was carried
out between pure concrete, mortar and sandcrete with
their corresponding products where 15% of RHA were
used.
Table 1 shows the quantity of materials and cost.
44.22 m
3
and 59.43 m
3
of concrete and mortar will be
needed at costs of N1,017,060.00 and N385,522.41
respectively while 1990 sandcrete blocks will be
needed at cost of N358,433.54. Thus, the building will
cost a total sum of N2, 059,515.00. The cost of which is
beyond the reach of an average citizenry in Nigeria.
This was acknowledged by the new minister of the
newly created Ministry of Lands, Housing and Urban
Development, Nduese Essien, who said that the
human development index shows that over 86% of
Nigerians live on $2 or less per day. This translates to
less than N10, 000 a month. The implication of this is
that over 86% of Nigerians can not afford a house
valued at N1.5 million even if they invested 100% of
their earnings over a period of 10 years. Obviously,
this is a very unlikely scenario as no one can save 100%

Fig. 1 Plan view of a typical two-bedroom low cost housing unit.
Kitchen
Sitting room

Bedroom
Verandah
Bath
& WC
Bath
& WC

Bedroom
Visitors
Toilet

1468
Table 1 Quantity of materials and cost implication for a 2-bedroom flat.
Materials Quantity Cost (N) Percent (%)
Concrete 44.22 m
3
1,017,060.00 49.38
Mortar 59.43 m
3
385,522.41 18.71
Sandcrete blocks 1990 nos 298,500.00 14.49
Others 358,433.54 17.72
Total 2,059,515.95 100

Fig. 2 Cost of materials in percentage of total cost.

Table 2 Quantity/cost of cement consumed and cost saved with RHA.
Materials
Cement
Cost saved with 15% RHA substitute (N)
Quantity (m
3
) Qty. (No. of bags) Cost (N)
Concrete 6.32 182.02 300,333.00 45,049.95
Mortar 9.91 285.54 471,141.00 70,671.17
Sandcrete blocks 1.97 56.74 93,621.00 14,043.15
Total 18.20 524.30 865,095 129,764.27

of their earnings. However, the above picture offers a
vivid view of the enormous task that lies ahead of
us [6].
It could be seen that concrete, mortar and sandcrete
block are responsible for 82.58% (Fig. 2) of the total
cost (N1,761,015.95). To produce these materials
(concrete, mortar and sandcrete blocks), 524.3 bags of
50 kg cement will be needed which will cost N865, 095
(about 50% of the cost of total cost of producing
concrete, mortar and sandcrete blocks) as shown in
Table 2. When comparing the cost of cement with the
overall cost of the building, cement alone is responsible
for 42% of the total cost.
Also in Table 2, when 15% of RHA is used as partial
substitute of cement N129, 764.27 will be saved. If
20% of the total cost saved is considered to be
responsible for the production of RHA, then
N90,834.99 (7% of cost of building) will be preserved.
Monetarily, it may appear to be insignificant but
considering its role in sustaining our environment it
will be highly appreciated. Some associated
environmental costs of cement production are
Ref. [21]:
After vehicle and utility emissions, cement
manufacturing is the largest industrial producer of CO
2

and accounts for over 50% of all industrial CO
2

49.38
18.71
14.49
17.72
0
10
20
30
40
50
60
Concrete Mortar Sandcrete
Blocks
Others
P
e
r
c
e
n
t
o
f
t
o
t
a
l
c
o
s
t
Materials

1469
emissions;
For every ton of cement produced, 1 t to 1.25 t of
CO
2
are produced;
Approximately 7,040 kg of raw materials is
required to manufacture 4,400 kg of cement.
This shows that use of RHA as partial replacement
of cement will reduce the consumption of cement and
subsequently reduce associated environmental
problems linked with the production of cement. In view
of this, a two bedroom flat, which required 26.215 t of
cement (Table 2) have potential to generate between
26.215 t and 32.769 t of CO
2
while 41.94 t of raw
materials will be consumed. If this trend continues
unchecked, then our desire to meet the MGD Goals
No. 7 to ensure environmental sustainability will be a
mirage. Other environmental problems such as
non-renewable resource depletion, acid rainwater and
contamination by sedimentation resulting from cement
production could be limited if consumption of cement
is reduced. Local cement production may also
contribute to noise, vibrations and scenic ecological
degradation. All these are reduced for every kilogram
of cement substituted for by RHA.
Furthermore, the use of RHA provides other benefits
of scientific, technical and economic nature such as the
improvement in the mechanical performance and
durability of cement, as well as the reduction of energy
consumptions associated with the production of clinker
[16].
6. Conclusions
The study investigated the effect of using RHA as
partial replacement of cement on the cost of
construction of a typical low-cost housing delivery.
The following conclusions were made:
Cement is responsible for the high cost of building
as 42% of total cost of a two-bedroom flat housing unit
is spent on cement;
Partial replacement of cement by 15% RHA saves
N70,000 of the cost of construction and thus, make
housing relatively affordable;
The use of RHA as partial substitute of cement in
the production of cement-based products will improve
their performance;
The use of RHA as partial substitute for cement
reduces cement consumption, hence, environmental
problems associated with cement production are
curtailed.
Reference
[1] Nubi, Affordable housing delivery in Nigeria, in: The
South African Foundation International Conference and
Exhibition, Capetown, Oct. 2008, pp. 1-18.
[2] Mabagunje, Housing Delivery Problems in Nigeria, Punch,
Nigeria, May 2007.
[3] Raji, Public and private developers as agents in Urban
Housing delivery in sub-saharan Africa: The situation in
lagos state, Humanity of Social Sciences Journal 3 (2)
(2008) 143-150.
[4] Olomolaiye, Rural Housing in Nigeria: Concept, Problems
and Functional Approach, Nigerian Institute of Town
Planners, Nigerian, 1999, pp. 20-35.
[5] T. Agbola, The Housing of Nigerians: A Review of Policy
Development and Implementation, Research report,
Development Policy Centre, Ibadan, 1998.
[6] Iroegbu, A. Kehinde, FG to Deliver 55,000 Housing Units
in 10 Months, Thisday Online, Apr. 13, 2010.
[7] Franklin, Nigeria: Minister Forecasts 2013 Self
Sufficiency in Cement Production, Vanguard, Nov. 26,
2009.
[8] International Cement Review, 8th ed., The global cement
report, 2009, pp. 120-250.
[9] A.A. Akeju, Challenges to providing affordable housing in
Nigeria, in: The 2nd Emerging Urban Africa International
Conference on Housing Finance in Nigeria, Nigeria, 2007,
pp. 1-5.
[10] H.F.W. Taylor, Cement Chemistry, Thomas Telford
Publishing, London, 1997.
[11] S. Wild, J.M. Khabit, A. Jones, Relative strength,
pozzolanic activity and cement hydration in super
plasticised MK concrete, Cem. Concr. Res. 26 (10) 1996
1537-1544.
[12] M. Fras, E. Villar Cocia, E. Valencia-Morales,
Characterization of sugar cane straw waste as pozzolanic
material for construction: Calcining temperature and
kinetic parameters, Journal of Waste Management 27
(2007) 533-538.
[13] Cizer, K. van Balen, J. Elsen, D. van Gemert, Carbonation
and hydration of calcium hydroxide and calcium silicate
binders with rice husk ash, in: 2nd International

1470
Symposium on Advances in Concrete through Science and
Engineering, Quebec City, Canada, Sep. 11-13, 2006.
[14] T.S. Ketkukah, E.E. Ndububa, Groundnut husk ash (GHA)
as pozzolana material in mortar, International Journal of
Science and Technological Research 3 (2) (2006) 209-214.
[15] P.K. Metha, Roles of cementitious material in sustainable
development of the concrete industry, in: Proceeding of
the sixth CANMET/ACI International Conference on the
Fly Ash, Silica Fume, Slag and Natural Pozzolanas in
Concrete, Bangkok, 1998, pp. 1-20.
[16] N.G. Alireza, A. Suraya, N.A.A. Farah, M. Mohamad
Amran, Contribution of rice husk ash to the properties of
mortar and concrete: A Review, Journal of American
Science 6 (3) (2010) 157-165.
[17] G.A. Habeeb, M.M. Fayyadh, Rice husk ash concrete: The
effect of RHA average particle size on mechanical
properties and drying shrinkage, Australian Journal of
Basic and Applied Sciences 3 (3) (2009) 1616-1622.
[18] G.L. Oyekan, O.M. Kamiyo, Effect of Nigerian rice husk
ash on engineering properties of sandcrete blocks and
concrete, Research Journal of Applied Sciences 3 (5)
(2008) 345-351.
[19] M.H. Zhang, M.V. Mohan, High-performance concrete
incorporating rice husk ash as a supplementary cementing
material, ACI Materials Journal 93 (6) (1996) 629-636.
[20] H.B. Mahmud, E. Majuar, M.F.M. Zain, N.B.A.A. Hamid,
Mechanical properties and durability of high strength
concrete containing rice husk ash, Journal of Advanced
Concrete Technology 79 (1) (2009) 21-30.
[21] ICLEI-Local Governments for Sustainability, Cement
Glossary, http://ww.iclei.org/us/cement/glossary.html
(accessed Mar. 5, 2004).

Issue 11, 2013 Journal of Civil Engineering and Architecture

Transféré par

Informations du document

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

Issue 11, 2013 Journal of Civil Engineering and Architecture

Transféré par

Droits d'auteur :

Formats disponibles

Journal of Civil Engineering

=(2 cover +diameter of bar)

Vous aimerez peut-être aussi