1.

1 RBF Process Description

In principle, RBF occurs when a well is placed sufficiently close to a river and part of the
surface water is induced to flow underground towards the cone of depressions caused by the
pumping well (see Figure 1-1). During ground passage, the water quality parameters change
due to microbial and physical- chemical processes (Partinoudi, 2004). After river water
infiltration is intercepted by collection wells, the riverbank filtrate requires additional
treatment steps before it can be pumped to the distribution system.

Figure 1-1 Generalized schematic of an RBF system. Source: Ray et al., 2002.

According to Heij (1989) most contaminants are degraded within the first few centimeters of
their path through the subsoil, but others are persistent and mobile and may move over longer
distances within the aquifer. Generally, two major areas during subsurface passage can be
designated:

a biologically high-active infiltration and clogging zone, where intensive degrading

and sorption processes take place
the successive subsurface passage, with lower degrading and sorption rates and an
increasing impact of dilution processes
The first part of the flow-course within the subsoil passage is the so called clogging zone - a
thin layer at the interface between river water and riverbed characterized by intensive physical,

1
chemical, and biochemical processes. Analyses have shown that an immense part of the
cleaning capacity taks place here (Grischek, 2003).
Clogging leads to the reduction of the permeability, and thus a decrease in the infiltration rate
(Grabs, 1981).
Permeability can be increased again by the self-cleaning power of the river itself. The self-
cleaning capacity of a river depends chiefly on the runoff regime, characterized by amount,
frequency, length, time and rate of change of runoff conditions (Schubert, 2007).

After passing the biologically high-active infiltration zone, surface water mixes with adjacent
groundwater. Dilution with groundwater improves the infiltrated surface water quality
because groundwater is usually a source of higher quality. In addition, dilution compensates
for temperature peaks and provides protection against shock loads (Kuehn and Mueller, 2000).

In most cases there is a flow of oxygen-rich surface water into the subsurface environment,
but it is common for dissolved oxygen to be completely used up by aerobic microorganisms at
some distance from the infiltration zone (Partinoudi, 2004).
Without oxygen, a change from oxidizing to reducing conditions can favor the effect of
heavy-metal remobilization of metals such as iron and manganese (Grischek, 2003). On the
other hand, anoxic conditions (no dissolved oxygen) help in the removal of river water nitrate
through anaerobic microorganisms.

After subsoil passage, the mixture of both groundwater and infiltrated surface water is
intercepted by collection wells. The collection wells can employ horizontal laterals or be
vertical wells. The laterals may or may not extend under the riverbed, depending on pumping
needs, the available budget of the utility and local geohydrological conditions (Partinoudi,
2004). [See section 3.3 for more on wells.]

After extraction, bank filtrate is usually treated depending on the bank filtrate quality,
whereby the quality determines the additional treatment steps required in order to produce
good-quality drinking water (Khn, 1999).
At a minimum, RBF acts as a pre-treatment step in drinking water production and, in some
cases, can serve as the final treatment just before disinfection (Bourg et al., 2002).

2
 1.2 Applicable Regulations
In order to understand the rising interest in design and construction of RBF, it is necessary to
get an insight into the log removal credit system and recent development of regulations in the
United States as the level of treatment and monitoring efforts are highly correlated with
legislative requirements.
This section provides an introduction into the log removal credit system of the United States
and an brief review of Environmental Protection Agency (EPA) regulations with particular
emphasis to the Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR).
Furthermore, the Microbial Toolbox, as a means to comply with governmental requirements,
is presented.

Introduction to the Log Removal Credit System

Log removal is a shorthand term for log10 removal, which refers to the physical and
chemical treatment of water to remove, inactivate, or kill pathogenic organisms such as
Giardia lamblia, Cryptosporidium parvum, and viruses (Ray et al., 2002). The equation of log
removals plays out as follows:

1- log removal equals a 90-percent

2- log removal equals a 99-percent
3- log removal equals a 99.9 percent
4- log removal equals a 99.99 percent target level of reduction, and so on.

Log removal credit is a regulatory term that expresses the amount of pathogens that a water
utility has removed from its water using technologies such as slow sand filtration, RBF and
other conventional types of treatment. Water utilities that employ RBF may receive 1-log
removal credit. This means that the RBF process has removed 90 percent of the initial
concentration of pathogens. However, if the target removal is 99.9 percent (3-logs), the utility
must remove an additional 2 logs using conventional filtration or other alternative techniques.
According to United States law, the granting of log removal credit is, in general, negotiated
between the water utility and primacy agency responsible for enforcing regulations (Ray et al.,
2002).

3
EPA Regulations

The United States Environmental Protection Agency (EPA) Office of Ground Water and
Drinking Water (OGWDW) develops potable water regulations to control microbial
pathogens and disinfectants/disinfection byproducts in drinking water (USEPA, 2005).
A short summary of current drinking water rules is given in table 1-6 (sorted by year of issue).
Basic information and compliance tips for each regulation are located at:
http://www.epa.gov/safewater/regs.html

Table 1-6 Current drinking water rules (by date issued). Source: US EPA Website -
http://www.epa.gov/safewater/regs.html.

Chemical Phase (Chemical Contaminant) Revisions to the Unregulated

Rules (2006)
Ground Water Rule (2006)
Stage 2 Disinfectants and Disinfection
Byproducts Rule (2006)
Long Term 2 Enhanced Surface Water
Treatment Rule- LT2ESWTR (2006)
Long Term 1 Enhanced Surface Water
Treatment Rule- LT1ESWTR (2002)
Filter Backwash Recycling Rule (2001)
Arsenic Rule (2001)
Unregulated Contaminant Monitoring
List 2 Rule (2001)
Radionuclides Rule (2000)
Drinking Water State Revolving Fund
Rule (2000)
Removal of the MCLG for Chloroform
(2000)
Public Notification Rule (2000)
Contaminant Monitoring Rule (1999)
Interim Enhanced Surface Water
Treatment Rule- IESWTR (1998)
Stage 1 Disinfectants and Disinfection
Byproducts Rule (1998)
Consumer Confidence Report Rule
(1998)
Variances and Exemptions Rule (1998)
Drinking Water Contaminant Candidate
List (1998)
Small System Compliance Technology
List for the Surface Water Treatment
Rule (1997)
Information Collection Rule (1996)
Surface Water Treatment Rule SWTR
(1989)

Surface Water Treatment Rule (SWTR)

The SWTR became law in 1989 and established a maximum contaminant level goals (MCL)
of zero for Giardia lamblia, viruses, and Legionella, as well as set filtration and disinfection
requirements for all public water systems using either surface water sources or groundwater
sources under the direct influence of surface water. Both, surface water or GWUDISW, are
considered to be vulnerable to microbial contamination.
According to Ray et al. (2002) the SWTR makes the following distinctions amongst drinking
water sources in the United States:

4
 Groundwater: Subsurface water contained in porous rock strata and/or soil that is not
affected by recently infiltrated surface water.
Groundwater Under the Direct Influence of Surface Water (GWUDISW): Water
beneath the surface of the ground that has a significant occurrence of insects or other
microorganisms, algae, organic debris, large-diameter pathogens like Giardia lambia,
or significant and relatively rapid shifts in water characteristics such as turbidity,
temperature, conductivity, or pH that closely correlate with meteorological or
surface water conditions (USEPA, 2004).
Surface Water: Water from sources open to the atmosphere, such as lakes, reservoirs,
rivers, and streams.

There are several methods to determine whether a well is categorized as groundwater or

GWUDISW (Partinoudi, 2009):

Hydrogeologic investigation
Water quality monitoring (WQM)
Microscopic Particulate Analysis (MPA)

If either the WQM method or a hydrogeologic investigation indicate a hydraulic connection to

nearby surface water, the water source is designated as a groundwater in hydraulic connection
with surface water, but not all wells that are hydraulically connected are automatically
categorized as GWUDISW (Ray et al, 2002a). Typically, further investigations like MPA are
required (Partinoudi, 2009).
MPA is often used as a method to provide evidence of a hydraulic connection between the
surface and groundwater (Ray et al, 2002a). Water facilities need to collect 3 E. coli samples
in a period of 3 months. If the well produces E. colifree water within this timeframe and the
well production depth is over 50 feet deep, then the water is not classified as GWUDISW.
If E. coli bacteria are found, the water is characterized as GWUDISW (Partinoudi, 2009).

The SWTR covers surface water systems and those that are classified as GWUDISW, while
the Ground Water Rule applies to wells determined to contain groundwater only.

The Interim Enhanced Surface Water Treatment Rule (IESWTR)

-5-
The IESWTR builds on the SWTR and requires tighter turbidity standards for systems that
serve more than 10,000 people. In addition, the IESWTR requires unfiltered systems to
include source water monitoring of Cryptosporidium in their watershed control plans (USEPA,
2007).

The Long Term 1 Enhanced Surface Water Treatment Rule (LT1ESWTR)

The LT1ESWTR addresses the concerns covered by the IESWTR as they apply to small
systems (i.e. systems serving fewer than 10,000 people) using surface water or GWUDISW
(USEPA, 2002).

Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR)

The LT2ESWTR builds upon the requirements established by the SWTR, IESWTR, and the
LT1ESWTR. The EPA finalized the LT2ESWTR in the Federal Register on January 5, 2006.
At present, it poses the most important drinking water regulation in the United States.
The EPA believes that implementation of the LT2ESWTR will significantly reduce levels of
Cryptosporidium and improve protection from exposure to other microbial pathogens such as
Giardia lamblia (USEPA, 2007). The LT2ESWTR is being promulgated simultaneously with
the Stage 2 Disinfection Byproduct Rule to address concerns about risk tradeoffs between
pathogens and DBPs.

The following requirements apply to all public water systems that use surface water or
GWUDISW (USEPA, 2005a):

Monitoring:

Under the LT2ESWTR, systems must monitor their water sources to determine treatment
requirements. This monitoring includes an initial two years of monthly sampling for
Cryptosporidium. To reduce monitoring costs, small filtered water systems can first monitor
for E. coli, and be required to monitor for Cryptosporidium only if their E. coli results exceed
specified concentration levels. Systems must conduct a second round of monitoring six years
after completing the initial round to determine if source water conditions have changed
significantly. Systems may use previously collected data in lieu of conducting new monitoring,
and systems are not required to monitor if they provide the maximum level of treatment
required under the rule.
Cryptosporidium parvum treatment:

-6-
Filtered water systems will be classified in one of four treatment categories (bins) based on
their monitoring results see Table 1-7. Systems classified in the lowest treatment bin carry
no additional treatment requirements. Systems classified in higher treatment bins must
provide 90 to 99.7 percent (1.0 to 2.5-log) additional treatment for Cryptosporidium. All
unfiltered water systems must provide at least 99 or 99.9 percent (2 or 3-log) inactivation of
Cryptosporidium, depending on the results of their monitoring.

Table 1-7 Treatment levels in RBF classification. Source: Partinoudi, 2003, 4.

Other requirements:

Systems that store treated water in open reservoirs must either cover the reservoir or treat the
reservoir discharge to provide 4-log virus, 3-log Giardia lamblia, and 2-log Cryptosporidium
parvum inactivation. Furthermore, systems must review their current level of microbial
treatment before making a significant change in their disinfection practice.

-7-
The following sources of information and guidance documents are available to help you meet
the LT2ESWTR requirements (USEPA, 2009a):

Table 1-8 Information and Guidance to meet LT2ESWTR requirements. Source: US EPA, 2009a, 30 ff.

EPA guidance manuals located at:

http://www.epa.gov/safewater/regs.
html
EPA Safe Drinking Water Hotline
at (800) 426-4791 (e-mail: hotline-
sdwa@epa.gov)
State drinking water agencies
National Rural Water Association
American Water Works
Association
Source Water Monitoring Guidance
Microbial Laboratory Guidance
Small Entity Compliance Guidance
Microbial Toolbox Guidance
Manual
Ultraviolet Disinfection Guidance
Manual
Membrane Filtration Guidance
Manual
Simultaneous Compliance
Guidance Manual
Low-pressure Membrane Filtration
for Pathogen Removal: Application,
Implementation, and Regulatory
Issues

-8-
The LT2ESWTR is the first regulation in the USA that specifically recognizes RBF as a
compliance technology option and includes provisions by which RBF could be used as one of
the compliance options for providing Cryptosporidium removal credits. It provides a natural
technology that will help many water treatment utilities to meet the stringent requirements of
the Stage 2 DBP and LT2ESWTR, and explains the recent, heightened interest in the design
and construction of RBF facilities (Partinoudi, 2004).

Microbial Toolbox

All surface water utilities in the United States are required to comply with specific
Cryptosporidium removal targets in the LT2ESWTR. The microbial toolbox is intended to
provide utilities with a wide range of treatment options for meeting LT2ESWTR compliance
requirements (Brown, 2003). The toolbox offers log-removal credits for Cryptosporidium for
various technologies, including riverbank filtration.

Most of the toolbox components require compliance with design and/or required
implementation criteria to receive credit (USEPA, 2007).
The following requirements must be met in order to receive log-removal credit for RBF
(Regli, 2003):

Design Criteria:
25 foot distance between river and well receives 0.5 log credit
50 foot distance between river and well receives 1 log credit
Only vertical and horizontal wells are eligible for removal credit
Only wells in granular aquifers - comprised of sand, clay, silt, rock fragments,
pebbles, or larger particles and minor cement, are eligible for removal credit

-9-
 Demonstration of aquifer characterization:
Sieve analysis of relatively undisturbed core samples from surface to depth > to
bottom of well screen
each recovered cored interval must be < 2 feet
at least 90% of the cored intervals must contain > 10 % fine grain material (grains
< 1.0 mm diameter)

Turbidity Criteria:
Turbidity monitoring for each well at least every 4 hours during operation
Average annual turbidity values (based on daily maximum values) should be less
than 1 NTU

In other European countries, for example Germany, no guidelines or handbooks are available
on where and how to install RBF systems; however, RBF, as an engineering technique, is
widespread throughout Europe and the design and construction are based upon personal
experience (Grischek et al., 2003).
Table 1-9 is a compilation of selected hydrogeologic information for RBF sites in the
United States and in Germany. As can be seen, the conditions vary mainly for capacity,
travel time, and distance between the river and the wells. At most sites in Europe, the distance
between the riverbank and production wells is >50 m and travel times are >50 days. In the
United States, travel times are <50 days and the distance between river and production well is
generally lower.
Table 1-9 Selected Site Data for RBF Systems in the United Sates and Germany. Source: Grischek et al., 2003,
293.

- 10 -
Systems can implement a variety of source, pre-filtration, treatment, additional filtration, and
inactivation toolbox components to receive Cryptosporidium credit, as summarized in Table
1-9 (USEPA, 2007).

Table 1-9 Microbial Toolbox Summary Table: Options, Treatment Credits and Criteria. Source: US EPA, 2007,
77.

- 11 -
Since 2007, an updated Web tool is available to the public and can be found at:
http://www.waterresearchfoundation.org/research/TopicsAndProjects/Resources/webTools/L
T2ESWTR_Index/index.aspx.
This web based microbial toolbox can help utilities determine which treatment is most
suitable for their water treatment plant. Based on the information entered, the program selects
the treatment options suitable to the treatment facility from a broad range of both new and
established technologies and programs. The Toolbox, which was developed by Environmental
Engineering and Technology Inc. as part of a Water Research Foundation project, features
help screens, cost estimates, and other information to evaluate whether selected items are
applicable for a particular site.

1.3 Source Water Quality Concerns

One of the main aims of a RBF system is to improve source water quality by removing a
variety of physical, chemical and microbiological pollutants. Understanding contaminants
present in the source water is essential for regulation, design and operation issues. This
section provides information pertaining to these applicable contaminants and substances that
cause the greatest water quality concerns to the water industry.

- 12 -
 Physical Contaminants

Temperature and turbidity are the physical contaminants of the greatest concern (Ray et al.,
2002) as high temperatures favor regrowth of bacteria in distribution systems, while high
turbidity can negatively impact source water quality.

Temperature

Temperature influences chemical reactions during disinfection (e.g. ozonation).

RBF extraction wells, water is typically at a low and constant temperature.
In addition, the solubility of oxygen also depends on temperature. Water with a high
temperature allows for less oxygen to be dissolved and can cause problems for aquatic species
in surface waters.
According to a 2-year monitoring effort (Wang, 2002) of the temperature of the Ohio River in
Louisville, Kentucky, temperature ranged from a low of 2C to as high as 32C, while the
temperature of the collector well, which is located 30.5 m away from the river, remained
relatively unchanged between 15 and 25C due to the flow through the aquifer and dilution
with groundwater whose temperature ranges around 10C.

Turbidity

Turbidity is a measure of the cloudiness of water- the cloudier the water, the greater the
turbidity. It is used to indicate water quality and filtration effectiveness (USEPA, 2009b).
Source-water can contain suspended solid matter consisting of particles of many different
sizes. While some suspended material will be large, and heavy enough to settle rapidly (the
settable solids), very small particles will settle only very slowly or not at all (the colloidal
particles). These small solid particles cause the liquid to appear turbid.
Turbidity can fluctuate significantly and is a concern for rivers that traverse through clay-rich
formations (Ray et al., 2002) where the river water becomes loaded with colloidal and
suspended particles such as clay, metal- hydroxo- compounds from e.g. Fn2+/ Mn2+.
Another type of particles and source of turbidity are algae and detritus (dead organic material).
The algae grow in the water and the detritus comes from dead algae, higher plants,
zooplankton, bacteria, fungi, etc. produced within the water, and from watershed vegetation
washed in to the water (WOW, 2008).

- 13 -
A variety of negative effects can be listed (Uhl, 2007):

sorption of harmful substances (e.g. hydrocarbons and microorganisms)

corrosion in the distribution system as a result of depositions
favouring transport of microorganisms
aesthetic concerns
may favour formation of disinfection by-products (DBPs) during water treatment

Chemical Contaminants

According to Ray et al. (2002) chemical contaminants can be divided into four major groups:

pH
Inorganics
Synthetic organic compounds (SOCs)
Natural organic matter (NOM)
Pharmaceuticals and personal care products (PPCPs)
Dissolved oxygen

pH
The pH of river water is the measure of how acidic or basic the water is on a scale of 0-14. A
pH of 7 is neutral, below 7 is acidic, and above 7 is basic or alkaline. Acid rain, from auto
exhaust or coal-fired power plants, causes a drop in the pH of water. Pollution from accidental
spills, agricultural runoff and sewer overflows can also change the pH. Buffering capacity is
water's ability to resist changes in pH.
The optimum pH for river water is around 7.4. Extremes in pH can make a river inhospitable
to life. Acidic water speeds the leaching of heavy metals.

Inorganics

Water Hardness originates from the dissolution of minerals containing chemical compounds
such as calcium (Ca2+) and magnesium (Mg2+), which are found in regions where sandstone
and limestone are dominant. The level of (Ca2+) and (Mg2+) determine the level hardness of
river water and groundwater. Hardness removal is a significant treatment process as hard

- 14 -
water can cause operation calcinations on both the distribution system and domestic
appliances such as showerheads, faucets, etc. Hardness can be reduced during peak flow
periods when the contribution from groundwater is low (Ray et al., 2002).

Bromide in high concentrations in bank filtrate can lead to the formation of bromate during
disinfection with ozone. Bromate is carcinogenic so the addition of ozone is limited by the
concentration of bromide in the source water (Ray et al., 2002).

Nitrogen and other forms of fertilizers do not occur naturally. Substances as ammonium
(NH4+), nitrate (NO3-) or nitrite (NO2-) run in rivers because of the sewage addition or by
traversing agricultural watersheds during flood periods. Rivers can receive large amounts of
these substances seasonaly (Ray et al., 2002).

Iron and Manganese Iron are two of the most abundant elements in earth`s crust and are the
two heavy metals of most significance for the water industry.
Although iron (Fe) and manganese (Mn) are not directly health threatening, they may cause
negative aesthetic effects (such as taste, odor, or color) in drinking water.
According to Schmidt et al. (2003), heavy metals can be removed by subsoil filtration for a
long time and they cannot be easily remobilized with one exception: if conditions in the
aquifer become anaerobic (no oxygen), iron and manganese undergo chemical
reduction and appear in the water, necessitating their elimination by additional treatment.

Synthetic Organic Compounds

Synthetic organic compounds, including pesticides and herbicides, are of great concern in
surface water treatment and often coincide with flow peaks. Rivers that run through
agricultural areas receive large loads of pesticides, especially during spring runoff (Ray et al.,
2002).
Ray et al. (1998) reported concentrations of atrazine, the most heavily used herbicide in the
United States used for control of broadleaf and grassy weeds in corn and soybeans (USEPA,

- 15 -
2009), as high as 12 g/l in the Illinois River. This concentration is much higher than the
maximum contaminant level (MCL) of atrazine of 3 g/l required by the EPA.
A wide range of synthetic contaminants, including pesticides, and herbicides, are listed in
table 1-1. Consumer and technical fact sheets of each of these substances are posted by the
USEPA at http://www.epa.gov/ogwdw000/hfacts.html.

Table 1-1 Synthetic organic Contaminants, including pesticides & herbicides. Source: US EPA Website -
http://www.epa.gov/ogwdw000/hfacts.html.

2,4,5 - TP (Silvex) Ethylene Dibromide

Adipate Glyphosate
Alachlor Heptachlor/Heptachlor Epoxide
Aldicarb/Aldicarb Metabolites Hexachlorobenzene
Atrazine Hexachlorocyclopentadiene
Benzo(a)pyrene LindaneChlordane
Carbofuran Methoxychlor2,4 D
Dalapon Oxamyl (Vydate)
Dibromochloropropane Pentachlorophenol
Dinoseb Phthalate, di(2-ethylhexyl)
Dioxin(2,3,7,8-TCDD) Picloram
Diquat Simazine Polychlorinated Biphenyls
Endothall Toxaphene
Endrin

Natural Organic Matter

NOM is a collective term assigned to all broken down organic matter that comes from plants
and animals in the environment. The concentration of NOM in source waters is directly
related to the concentration of disinfection by-products (DBPs) in treated waters, such as
trihalomethanes (THMs) and haloacetic acids (HAAs), all of which are potentially
carcinogenic (Vogt et al., 2003). DBPs form when organic and mineral materials in water
- 16 -
react with chemical treatment agents during disinfection. Therefore, NOM in surface water is
a major concern for utilities that use chlorine as the disinfectant and it can be challenging to
reduce NOM prior to disinfection.

The following parameters are typically used as indicators of NOM in surface water (Ray et al.
2002):
Total organic carbon (TOC)
Dissolved organic carbon (DOC)
Biodegradable organic carbons (BDOC)/ Assimilable organic carbon (AOC)
Specific ultraviolet absorption (SUVA)
Ultraviolet absorbance of water at 254 nanometer (UV254)

DOC is responsible for the majority of reactions of interest in drinking water treatment, e.g.
disinfectant demand, DBP formation, biogrowth and coagulant demand (Drewes and
Summers, 2002). Monitoring efforts in Germany report a mean DOC level of 3mg/l at the
Rhine River, 5.5 mg/l at the Elbe River and 6-8 mg/l at Lake Tegel. These higher levels of
DOC at Lake Tegel are partly due to the discharge of sewage from wastewater treatment
plants (Ray et al., 2002).

Pharmaceuticals and Personal Care Products (PPCPs)

PPCPs is an umbrella term for thousands of chemical substances, including prescription and
over-the-counter therapeutic drugs, veterinary drugs, fragrances, and cosmetics. These
compounds are considered micropollutants because they are detected at very low levels, e.g.
nanogram-per-liter (Ray et al., 2002). Many of these products are found in domestic sewage
and ultimately end up in rivers. Some pharmaceuticals and personal care products are
suspected of causing direct endocrine disruption. They have potentially adverse effects in
natural ecosystems, such as causing abnormal physiological processes and reproductive
impairments of aquatic species (Kolpin et al., 2002).
Analytical determination of pharmaceuticals and personal care products in rivers, lakes and
other water sources is difficult due to the fact that many of these compounds are found in
extremely low concentrations and complex instrumentation is required. As a result, there is
very little data on the concentration of these compounds in surface waters within the United
States (Ray et al., 2002).

- 17 -
Representative classes and members of pharmaceuticals and personal care products can be
found in a presentation by Christin G. Daughton located at: http://www.epa.gov/ppcp/. A
summary of these chemical substances is presented in Table 1-2.

Table 1-2 Selected classes and members of Pharmaceuticals and Personal Care Products found in
environmental samples. Source: Ray et al., 20002, 9.

Dissolved Oxygen

Dissolved oxygen (DO) is found in microscopic bubbles of oxygen that are mixed in the water.
DO is an important indicator of a water body's ability to support aquatic life.
Oxygen enters the water by absorption directly from the atmosphere or by aquatic plant and
algae photosynthesis. Oxygen is removed from the water by respiration and decomposition of
organic matter.
In fast-moving streams streams, if unpolluted, are usually saturated with oxygen due rushing
water is aerated by bubbles as it churns over rocks and falls down. In slow, stagnant waters,
oxygen only enters the top layer of water, and deeper water is often low in DO concentration
due to decomposition of organic matter by bacteria that live on or near the bottom of the
riverbed (BASIN, 2007).
The colder the water, the more oxygen can be dissolved in the water. As a result, DO
concentrations at one location are usually higher in the winter than in the summer.

- 18 -
A lack of oxygen during underground passage due to biological activity in or on the riverbed
can lead to anaerobic conditions over a portion of the flow path, which may result the release
of heavy-metals such as iron and manganese from the bank sediment into the flowing water.

Microbiological Contaminants

According to the AWWA Manual of Water Supply Practices M 48 (2006), biological

contaminants in surface water include protozoa, bacteria, and viruses. These waterborne
pathogens can cause life-threatening disease in the immuno-suppressed populations of the
world and illness in the general population. This section includes information about biological
contaminants and highlights the applicable pathogenic species which can cause major
problems for the water industry.

Protozoa

Waterborne parasites, including various kinds of worms and protozoa, are indicators of fecal
contamination as well as the leading sources of those disease acquired through fecal
contaminants in food and/or drinking water. The protozoa comprise a large group of
extremely diverse unicellular organisms and are described in table 1-3 (AWWA, 2006).

Table 1-3 Parasitic pathogenic agents. Source: AWWA, 2006, 162.

- 19 -
Of particular concern are Cryptosporidium parvum and Giardia lamblia which are known to
be extremely resistant to conventional means of disinfection. Under the Long Term 2
Enhanced Surface Water Treatment Rule (LT2ESWTR), utilities are allowed to choose from a
toolbox of technologies in addition to existing treatment to comply with log treatment
requirements for Cryptosporidium. At present, bank filtration, is given the potential log credit
of 0.5 for a well setback distance of 25 ft and 1.0 for a well setback distance of 50 ft.
Further information on the LT2ESWTR, the log removal credit system and the microbial
toolbox, see section 1.3.

Cryptosporidium parvum and its potential for causing disease has become a major concern to
water treatment personnel since 1984 when the first waterborne outbreak associated with
Cryptosporidium was reported in Milwaukee, Wisconsin. Cryptosporidium oocysts are
ubiquitous parasites that are widely distributed in the water source and remain quite
environmentally stable. Occysts can survive for months in cold, moist environments, such as

- 20 -
lakes and streams. Once a human is infected, they can cause watery diarrhea, abdominal pain,
nausea, fever, and fatigue. In patients with compromised immune systems, the illness may be
life threatening (AWWA, 2006).
Utilities should consider that Cryptosporidium is resistant to chlorine- based disinfectants.
Research literature shows that turbidity is the key parameter to prevent Cryptosporidium
contamination in drinking water. For instance, when filter-effluent turbidity ranged between
0.1 and 0.3 NTU, Cryptosporidium presence was as much as 90 percent (1 log) greater than
when effluent filter turbidity was 0.1 ntu or less. Watersheds should also be managed in a way
that limits the introduction of Cryptosporidium into the drinking water supplies (AWWA,
2006).

Giardia lamblia is a global parasite that infects numerous mammals including humans, dogs,
cats, beavers, muskrats and other warm-blooded animals. It is the most commonly identified
pathogen in waterborne outbreaks in the United States. Giardia cysts have dimensions
ranging from 5 to 18 m. It remains viable in river water for up to 28 days (Regnier et al.,
1989). Giardia cysts are not as resistant to disinfection as Cryptosporidium oocysts thus,
treatment designed to inactivate oocysts can effectively inactivate Giardia cysts, too (EPA,
2009).

Bacteria

Bacteria are microorganisms that are simpler and smaller than parasitic pathogens but larger
and more complex than viruses. Bacteria size ranges from approximately 0.2 to 10 m in
length (Partinoudi, 2004). Natural waters are often contaminated by pathogenic bacteria
excreted by humans and various domestic and wild animals. The main source of bacteria for
entering water sources is sewage (Schijven, 2002). A compilation of bacterial agents that are
known to contaminate public water systems is listed below in table 1-4:
Table 1-4 Bacterial Pathogenic Agents. Source: AWWA, 2006, 73.

Acinetobacter Heliobacter pylori

Aeromonas Klebsiella
Campylobacter Legionella
Cyanobacteria Mycobacterium avium complex
Enterohemorrhagic Escherichia coli Pseudosomonas
Escherichia coli Salmonella
Flavobacterium Serratia
- 21 -
Shigella Vibrio cholerae
Staphylococcus Yersinia

Due to improved hygiene standards the threat of these compounds is considered to be

moderate, although these bacteria are capable of multiplying in water supply storage and
distribution systems (AWWA, 2006). Escherichia coli and Legionella are the two waterborne
bacteria with the most significance to the water industry.

Escherichia coli belong to the group of total coliform bacteria that are used as indicators of
sewage contamination. The presence of fecal coliform bacteria is an indication that a water
source has been contaminated by human or animal waste (Partinoudi, 2004). Compared to
other fecal coliform bacteria, e.g. Enterobacter, Escherichia coli is particularly suitable for an
indicator because it is easily detected and enumerated. In addition, it has the ability to remain
viable outside the bowel of warm-blooded organisms for a long time (AWWA, 2006).

Legionella bacteria are ubiquitous in the aquatic environment and can survive in water system
biofilms. These bacteria are able to colonize artificial environments such as cooling towers,
evaporative condensers, hot-water tanks, whirlpool spas, decorative fountains, and the
drinking water distribution system. The mode of transmission is by inhalation of moist
aerosols contaminated with Legionella bacteria. In immuno-suppressed individuals, the
bacteria can cause Legionnaires Disease and Pontiac fever (AWWA, 2006).

Viruses

Viruses are the smallest and most basic of known life forms, ranging from approximately 18
to 120 nanometers. Viruses have no reproductive system and replicate by taking over a living
cell and usurping cellular machinery (Yates and Yates, 1988).

More than 120 different enteric viruses are known to infect humans. Enteric viruses are
excreted in the feces of infected individuals. Once in the environment, they can survive for
long periods of time, up to several months under cool and moist conditions (AWWA, 2006).
Their extremely small size and resistance to chemical and environmental degradation present
great challenges to the drinking water industry.

22
The following table 1-5 includes groups of viruses identified as sources of waterborne disease
outbreaks or having the potential to cause outbreaks:

Table 1-5 Viral Pathogenic Agents. Source: AWWA, 2006, 251.

Adenovirus groundwater (Gaining & Losing Rivers),

Astrovirus the impact to the aquifer due to fluctuating
Emerging viruses (Parvo, Corona, ) river water characteristics e.g. during flood
Enterovirus and Parechovirus conditions, and points out the correlation
Hepatitis A virus between river morphology and suitability
Hepatitis B virus for a RBF site (Erosion & Deposition).
Human Calicivirus
Reovirus
Rotavirus Gaining & Losing Rivers

The hydraulic connectivity between the

1.4 Generic River & Aquifer river and the adjacent aquifer is a basic

Interactions requirement for the subsurface flow to an

RBF extraction well and must be assessed
in the initial site investigations. This can be
Generic river and aquifer interactions
realized by several measurement methods
include all those phenomena which occur
and instruments such as flow meter
naturally in all river-aquifer systems.
measurement, drilling, ground penetrating
Furthermore, pumping-enforced
radar or tracers (Hoehn, 2002).
interactions, such as clogging or dilution,
gain importance once a RBF plant starts to
Surface water is commonly hydraulically
operate and are discussed in section 2.
connected to groundwater in three possible
Both natural and generated processes are
categories (Hoehn, 2002):
important for understanding the
environment a RBF system is settled in.
Gaining Rivers
Losing Rivers
This section identifies some applicable
Flow Trough Rivers
hydrogeological controls which dominate
the flow of the infiltrating water, such as
the exchange of river water and
23

Gaining Rivers gain water from inflow of
groundwater through the riverbed (Figure
1-2).
This may occur when the altitude of the
groundwater table in the vicinity of the
conditions. Hydraulic conductivity has
great impact to the infiltration capacity
from surface water into groundwater and
must be assessed from pumping tests,
flow-meter measurements or grain-size
distributions (Hoehn, 2002).
Figure 1-3 Losing River. Source: Hoehn, 2002, 21.

stream is higher than the altitude of the

river water stage (Partinoudi, 2004).

Figure 1-2 Gaining River. Source: Hoehn, 2002, 21.

Flow Trough Rivers receive groundwater

through the upgradient bank, and lose
water through the downgradient bank
Losing Rivers lose surface water to (Figure 1-4). This may occur when the
groundwater through the riverbed (Figure river turns at a steep angle to the floodplain,
1-3). and if the rivers surface remains higher
This may occur when the altitude of the then the adjacent down-valley water table
groundwater table in the vicinity of the (Hoehn, 2002).
stream is lower than the altitude of the
river-water stage (Partinoudi, 2004). The Figure 1-4 Flow Trough River. Source: Hoehn,
2002, 21.
geological material below the channel can
either be fully saturated or unsaturated if
the channel is perched above the
underlying water table. The hydraulic
conductivity of an unsaturated leakage is
much lower than under saturated

24
 already in the direction from the river to
the well (Partinoudi, 2009).

All these conditions may exist in the same

river at different locations or times of the
year, but in principle, rivers lose water in
peri-alpine floodplain valleys filled with
coarse and very permeable alluvial
sediments. River channels gain water in
flat regions when the river stage is lower
than the adjacent water table (Hoehn,
2002). In both cases, there must be
permeable material that will allow this
hydraulic head to move water.
Fluctuating River Water
Knowing whether a stream is originally
Characteristics and their Impact
gaining or losing is important (Partinoudi,
2004).
to the Aquifer
Depending on site characteristics it might
be advantageous to be in the vicinity of a River-aquifer interactions are controlled by

gaining river or a losing river. On the one the fluctuating water level of the river.

hand, wells in gaining areas can be The resulting gradients between the

managed to gain higher quality, as the ratio changing river level and the gradual

of high quality groundwater can be adaptation of the groundwater table in the

increased by increased pumping. On the adjacent aquifer, control flow and transport

other hand, wells in losing areas can in riverbank filtration (Schubert, 2001).

increase their yield with a minimum of Figure 1-5 shows an example of surface

energy input, as the natural gradient is water fluctuations of the river Rhine.

25
 Figure 1-5 Surface water level of the river Rhine
(1988-1990). Source: Schubert, 2001, 147.
Sources of information about high flow
and flood data are listed below:

The National Flood Frequency

Program:
http://pubs.usgs.gov/wri/wri024168
/pdf/entirereport.pdf
WaterWatch:
http://waterwatch.usgs.gov/?state=u
s&map%20type=flood&web%20ty
pe&map
Army Corps of Engineers
http://usace.army.mil/Pages/default.
aspx

According to Heij (1989), there exits a

linear relationship between surface water
level and the infiltration rate. There is also
It is well known that fluctuating
an inverse relationship between this level
temperatures and concentration of
and the average time water particles
components in the river water are balanced
require to flow from the surface water to
out during subsoil passage. According to
the bank.
Schubert (2001), the effects of balancing
I.E. higher river water levels, especially
are caused by an age-stratification of
during flood periods, may cause higher
infiltrated river water. Age stratification
infiltration and increased ground water
represents the difference in the residence
flow rates as a result of increased head
time of water in the aquifer. Figures 1-6
gradient. Also, lower log removals are
and 1-7 provide examples of the balancing
expected to occur during and shortly after
effects of subsoil passage. Temperature
floods because protective layers may be
and other abiotic parameters of
removed by flood scour (USEPA, 2003).
groundwater do not fluctuate as suddenly
or as extremely as those of surface water
systems.

26
Figure 1-6 Water temperature in Ohio River and in
the production well. Source: Schubert, 2005b, 3.
Figure 1-7 Chloride concentration in the river
Rhine water compared that in the adjacent well
water. Source: Schubert, 200, 157
Erosion & Deposition
27
According to Schubert (2002) three
different simplified model-regions can be
distinguished along a river:
Upper part (with erosion)
Middle part (with bed load
transport)
Lower part (with deposition)
Although this model assumption does not
truly reflect the natural design of a river, it
can help in selecting RBF sites.
Erosion in upper parts along the river is
characterized by a high river-flow velocity,
a high hydraulic gradient and high shear
force on the riverbed. The grain-size
distribution in these areas is usually limited
to coarse material, as particles of smaller
size and finer material are washed away
downstream. High infiltration capacity and
high hydraulic conductivity of the
streambed lead to the lack of necessary
time for balancing out variations in
temperature and concentration of
pollutants. As a result, no sufficient
protection against sudden contamination
(shock loads), can be secured.
Bed load transport is the movement
(rolling, skipping or sliding) of sediment,
such as soil, rocks, particles, or other
debris along or very near the riverbed by
flowing water. It is instrumental in the self-
cleaning process of the riverbed. Table 1-
10 shows the characteristics of the lower
Rhine valley where 80 percent of the RBF
sites are located. On principle, middle or
lower parts with bed load transport along
the river are suitable for a future RBF site
(Schubert, 2002).
Table 1-10 Characteristics of the Rhine river in the
lower Rhine valley region. Source: Schubert, 2002,
36.
28
Deposition regions occur usually upstream
of dams and near the mouth of the river
and should be avoided when selecting an
RBF site. The deposition of very fine
particles, like silt and fine sand, as well as
slow river flow velocity, can result in a
limited infiltration capacity (Schubert,
2002).
2 Applicable Processes
The treatment effectiveness of RBF results
from a combination of several applicable
processes such as clogging of the riverbed,
the dilution with groundwater after
infiltration, subsurface filtration (filtration,
adsorption, biodegradation, ion exchange,
oxidation/reduction) and additional
treatment steps.
Riverbank filtration is a highly dynamic
process on account of the changes in the
quality of river water due to river water
level and the variations in the physical
(temperature, suspended solids), chemical
(type and concentration of compounds) and
biological (type and concentration of
viruses, bacteria and protozoa) properties
(Schubert, 2005a).
An examination of the basic hydraulic,
physicochemical and biological processes
in bank filtration will help to define several
criteria for the appropriate site for RBF.
This section describes applicable RBF
processes according to the path of the
water from the river until the transfer to the
distribution system. An overview of RBF
processes is given in Figure 2-1.
Figure 2-1 Riverbank Filtration Processes. Source:
Amy et. al., 2006, 104.
2.1 Clogging & Cleaning Processes
Clogging
Clogging is the formation of a clogging
layer on top or in the riverbed and can be
defined as an impediment of flow,
typically as a result of physical, chemical,
and biological processes (Grabs, 1981).
According to Riesen (1975), mechanical
clogging of parts of the riverbed during
29
constant pumping of RBF wells is
unavoidable; however, its effects are not
always harmful. The disadvantage of
clogging is that it can reduce hydraulic
conductivity of the local riverbed and the
aquifer. As a result well-yields are
temporarily or permanently reduced.
On the other hand, some benefits such as
particle- and pathogen removal and
degradation of organic compounds are
positive effects of clogging (Grischek,
2006).
Figure 2-2 shows a picture of a paved and
clogged riverbed.
Figure 2-2 Paved and clogged riverbed near the
outer section of a bend (at Flehe waterworks,
Dsseldorf, well site). Source: Schubert, 2005,3.
Physical clogging results from the
deposition of fine-grained, suspended
sediment at the surface water ground
water interface, and the deposition and
growth of microorganisms. During periods
of low surface water discharge, or in areas
with low flow velocities e.g. the edge of a
river, physical clogging may be
exacerbated (USEPA, 2003).
Chemical clogging results from
precipitation of dissolved surface water
constituents and occurs near the interface
or anywhere along the flow path. This is
due to the change in geochemical
conditions as infiltrating water enters the
riverbed (USEPA, 2003).
Biological clogging results from the
accumulation of bacterial cells in pore
spaces, the production of extra-cellular
polymers, the release of gaseous by-
products from denitrifying bacteria, and
accumulation of insoluble precipitates.
Insoluble sulphite salts can cause clogging
due to the activity of sulphate reducing
bacteria, whereas iron hydroxide and
manganese oxide deposition can be
brought on by bacterial iron-metabolism.
Biological clogging may occur near the
surface water ground water interface
where nutrients are most available (Baveye
et al., 1998).
Cleaning Processes
30
Both the positive and negative effects of
riverbed clogging can be diminished by the
regenerative process of streambed scouring.
Scouring is a result of shear forces
imparted on a riverbed by the motion of
the water passing over the riverbed, and
the resistance to this motion imparted by
the riverbed itself (Hubbs, 2004).
The self-cleaning potential of a river
depends chiefly on the runoff regime,
characterized by amount, frequency, length,
time and rate of change of runoff
conditions. During flooding, the river
channel may be scoured and fine sediments
at the surface water ground water
interface mobilized.
Research efforts by Schubert (2005b) at
the Ohio River in Louisville have shown
that even minor variations in the river level
are able to cause temporary improvements
of the infiltration capacity and flood waves
can even cause significant jumps in the
infiltration rate.
Much of the removal of the contaminants
and microbes discussed above occurs
during the first few centimeters of the flow
path, due to the significant filtering and
sorptive capabilities of sediments in the
riverbed. If this active layer is washed
away or scoured, the effectiveness of bank
filtration may be temporally threatened
(USEPA, 2003).
Riverbed scouring plays an essential role
in determining the sustainable yield in
RBF systems.
Therefore, the US EPA (2003) suggests
evaluating the potential for stream channel
scour as part of the initial RBF site
investigations (see section 3.1).
The extent of riverbed scouring can be
estimated as a function of riverbed shear
stress exerted during high-flow events.
Unfortunately, there is no practical
technique for directly measuring shear
force on the riverbed. However, it can be
estimated by the surface slope of the
stream, vertical velocity profiles in the
stream, and the sediment transport on the
riverbed (Hubbs, 2004). Typically, streams
exert higher shear stresses near the upper
part of a river, with decreasing stresses
exerted near the lower part of a river. This
implies that riverbed scouring will
decrease near the terminus of a stream.
Because of this tendency to deposit fine
materials near the mouth of streams, these
locations are usually not well suited for
RBF systems (Hubbs, 2003).
1.4 Subsurface Filtration
Processes
31
Filtration
Physical filtration is the classical process
for removal of particulate matter and
microbes in water treatment and occurs
primarily by straining and pore
sedimentation (Partinoudi, 2004).
Straining is a purely physical removal
process governed by the size of pore
throats and the size of microbial particles.
Straining occurs when the particles in
suspension in the porous matrix cannot
pass through a smaller pore, and thus their
transport is stopped.
According to Berger (2002), straining of
bacteria and viruses is less effective than
for protozoa because of their smaller size.
However, if the viruses or bacteria are
absorbed onto a solid particle of greater
size than itself, filtration can be an
important removal process (Partinoudi,
2004).
In riverbank filtration the filter medium is
the natural aquifer underneath and adjacent
to the river (Schubert, 2005a). The grain
size distribution has to comply with the
requirements of riverbank filtration. An
important requirement is a sufficient
amount of fine-grained sediments to
achieve adequate pathogen removal. The
determination of the grain size distribution
is part of the initial aquifer characterization
explained in section 3.2.

Pore Sedimentation occurs when the
density of a microbial particle is higher
than that of water (Partinoudi, 2004). Pore
settling is likely where groundwater
velocities are low, such as in finer-grained
riverbed material. These materials are
removed during flood periods, and thus
pore sedimentation is significant during
low flow periods of the river (Berger,
2002).
Adsorption
According to Brandt et al. (1993),
adsorption is the adhesion of molecules of
gas, liquid, or dissolved solids to a surface,
while desorption is the reverse of
adsorption.
The fine-grained sediments on many river
bottoms are generally rich in clay and
organic matter and have the potential to
adsorb a variety of contaminants in river
water (Ray et al, 2002a).
Sorption on solids, e.g. grains in the
aquifer, is an equilibrium reaction. which
may setback peaks of organic substances,
such as humic acids, by adsorption and on
the other hand may cause peaks of
previously adsorbed substances by
desorption due to a rapid change in river
32
water quality parameters. Due to the
permanent infiltration of river water into
the aquifer during bank filtration, the
subsoil tends toward saturation of the
adsorbents (mainly humic acids) and thus
limits the effectiveness of sorption
processes for organic substances in the
bank filtered water (Schubert, 2005a).
Adsorption and desorption are more
important for the removal of
microorganisms. The adsorption of
microorganisms onto the surface of soil
particles is caused by a combination of
electrostatic and Van der Waals forces and
hydrophobic interactions between the
microorganisms and soil particles.
Desorption occurs due to changes in the
ionic strength, the temperature, and pH of
the soil water (Partinoudi, 2004).
Sorption processes will hold back the
transport of microorganisms in the aquifer
significantly. To profit by this process,
sufficient flow path length and flow time is
necessary (Schubert, 2005a).
Biodegradation
Biodegradation is the chemical breakdown
of materials by a physiological
environment usually catalyzed by the
activities of microorganisms. It is a
significant process which starts directly
below the river-aquifer interface (Ray et al,
2002a).
Organic material can be degraded
aerobically with oxygen or anaerobically
without oxygen. There is a flow of oxygen-
rich surface water into the subsurface
environment and this input of oxygen to
the streambed stimulates a high level of
activity by aerobic miccroorganisms. It is
common for dissolved oxygen to be
completely used up at some distance to the
streambed. From there, anaerobic
microorganisms dominate microbial
activity. Anaerobic bacteria can use nitrate,
sulfate, or other solutes in place of oxygen
for metabolism (Partinoudi, 2004).
Ion exchange
Ion exchange is an exchange of ions
between two electrolytes or between an
electrolyte solution and a complex.
Clay minerals, organic substances, and
humic acids have a high exchange capacity
for cations, particularly heavy metals. To
profit from ion exchange, the removal of
33
fine particles during bank filtration is of
significant importance (Schubert, 2005a).
Oxidation/reduction
Aerobic conditions in the aquifer support
high degradation rates of organic
compounds (Schubert, 2005a). As a result,
oxygen depletion by biological activity can
lead to anaerobic conditions over a portion
of the flow path, which may result the
release of heavy-metals such as iron and
manganese from the bank sediment into
the flowing water. This process occurs due
to a redox reaction which reduces iron and
manganese in their water soluble forms
(USEPA, 2003).
At varying distances to the riverbed, the
level of oxygen determines the occurrence
and magnitude of oxidation and reduction
sub-processes as shown in Figure 2-4. The
degree to which substances are reduced
may vary between different rivers and at
different sampling points on a river (Kuehn
and Mueller, 2000).
Figure 2-3 Riverbank Filtration Processes. Source:
Grischek, 2003, 8.
 1.5 Dilution Process

Riverbank filtrate includes both

groundwater and river water that has
percolated through the banks or bed of a
river to an RBF extraction well (Partinoudi,
2004).
The dilution of surface water with
groundwater is considered one of the
advantages of RBF because groundwater is
usually a source of higher quality.
Significant differences exist between the
On the other hand, if the flow path between
two water sources. River characteristics,
the riverbank and the well is long enough,
such as temperature, turbidity and level of
iron and manganese can precipitate onto
contaminants can change significantly
the sediments in the subsurface before
during the year, depending on weather
reaching the extraction well (Tufenkji et al.,
conditions, river flow or emissions by
2002). The aquifer becomes reaerated with
municipal and industrial sewage runoff. In
increasing distance from the riverbed. This
contrast, groundwater remains nearly
is one reason for locating RBF wells
constant (Wang, 2002).
greater than 25 or 50 feet from the river, as
To asses how effectively the RBF process
discussed in section 1.3 (USEPA, 2003).
improves water quality, one must
Sonheimer (1980) considers aerobic
determine the extent of groundwater
conditions in the ground, advantageous, as
mixture to distinguish between true
iron and manganese will not go into
contaminant removal due to RBF process
solution and biological oxidation of
and reduction due to dilution (Partinoudi,
organics occurs under anoxic conditions in
2004).
the presence of nitrate.
The determination is normally based on
water quality parameters:

natural tracers (bromide, chloride)

inorganic parameters (hardness,
temperature, conductivity)

34
 organic parameters (true color,
TOC/DOC)
No tracer is a perfect tracer i.e. there are
advantages and disadvantages to the use of
each of these parameters (Partinoudi,
2004).
It is important that the tracers
concentration is significantly different
between the river water and groundwater
to accurately gauge the dilution ratio
between the two sources. According to
Wang (2002), the best way to determine
the amount of dilution is to find a tracer
that exists at a constant concentration at
one of the sources while remaining absent
from the other source, and that is
considered conservative during subsurface
transport. As a result the most consistent
and reliable parameter has to be selected at
each site.
Once a convenient parameter has been
found, results can be calculated with the
following mass balance formula:
X= ((c(RBF)-c(AQUIFER))/(c(RIVER)-
c(AQUIFER))]*100
35
X: fraction of river water in the RBF
well
c(RBF): concentration of a tracer in
the riverbank filtrate
c(AQUIFER): concentration of a
tracer in the aquifer
c(RIVER): concentration of a tracer
in the river
Example: If the temperature in the river is
21.5C, 10.1C in the aquifer and 13.2C
in the riverbank filtrate, the fraction of the
river water in the RBF well will be
calculated as follows:
X = [(13.2-10.1)/(21.5-10.1)]*100
X = 27.2 %
1.6 Additional Treatment Steps
Bank filtrate is usually treated after
extraction. The level of additional
treatment steps depends on surface water
quality as well as on the cleaning capacity
of the bank passage (Khn, 1999).
One disadvantage of RBF is that an
additional aeration step may be required
during water treatment due to the depletion
of oxygen by microorganisms during
subsoil passage. This oxygen depletion
may lead to the anaerobic conditions which
can result in the release of iron and
manganese from the bank sediment into
the flowing water. This condition may
necessitate the removal of these metals
during additional treatment steps (USEPA,
2003).

The bank filtration has no effect on

recalcitrant substances, such as several
micropollutants. Therefore, the treatment
of bankfiltrate often includes granular
activated carbon filters
(Khn, 1999).

By taking into consideration that the bank

filtrate quality highly depends on the local
environment (e.g. aerobic/anaerobic
conditions or river water quality), the
additional treatment of riverbank filtrate
may include ozonation, nitrification,
rapid/slow sand filtration, granular
activated carbon filters, oxygenation,
removal of iron and manganese, and
disinfection.
At a minimum, RBF acts as a pre-
treatment step in drinking water production
(mostly in large urban areas) and, in some
cases, can serve as the final treatment just
before disinfection (usually in small
communities) (Ray et al., 2002).

36
References

Bourg, A.C.M., Keziorek, M.A.M., Darmendrail, D. (2002). Organic Matter as the Driving Force in the
Solubilization of Fe and Mn during Riverbank Filtration in Ray, C (ed.) Riverbank Filtration:
Understanding Contaminant Biochemistry and Pathogen Removal, Kluwer Academic Publishers,
Netherlands, 43-54.

Grabs, W. (1981). Beitrag zur Beschreibung von Kolmationserscheinungen in einem

organisch belasteten Kleingewsser Beitrge zur Hydrologie 2, 293-311.

Grischek, T. (2003). Zur Bewirtschaftung von Uferfiltratfassungen an der Elbe Institut fr

Grundwasserwirtschaft Technische Universitt Dresden, Heft 4.

Heij, G.J. (1989). River-Groundwater Relationships in the Lower Parts of the Netherlands, J. Hydrol.,
108(1-4), 35-62.

Kuehn, W. (1999). Overview of Riverbank Filtration Issues in Abstracts Riverbank Filtration

Conference, Nov 4-6, Louisville Kentucky.

Kuehn, W., Mueller, U. (2000). Riverbank Filtration: an Overview Journal American Water Works
Association, 92, 60-69.

Partinoudi, V. (2004). Riverbank Filtration as a Viable Pretretment and Treatment Method M.S.Thesis,
Univ. of New Hampshire, Durham, NH.

Schijven, J., Berger, P., Miettinen, I. (2002). Removal of Pathogens, Surrogates, Indicators, and Toxins
Using Riverbank Filtration in Riverbank Filtration: Improving Source Water- Quality, Kluwer Academic
Publishers, Dordrecht, The Netherlands.

Schubert, J. (2007). Significance of Hydrologic Aspects on RBF Performance: Everything is Linked to

Everything else in Hubbs, S. A. (ed.) Riverbank Filtration Hydrology, Vol. 60, 1-20.

Worch, E. (1999). Laboratory Tests for Simulation of Riverbank Filtration Processes in Abstracts
International Riverbank Filtration Conference Louisville, Kentucky.

Figures/ Tables

Figure 1-1
Generalized schematic of an RBF system Source: Ray et al., 2002, page 2.

Amy, G., Collins, R.M., Drewes, J., Grnheid, S., Jekel, M. (2006). Integrated Comparison of
Biofiltration in Engineered versus Natural Systems in Abstracts of International Workshop on
Riverbank/Riverbed Filtration, Korea.

Baveye, P., Vandevivere, P., Hoyle, B.L., DeLeo, P.C., Sanchez de Lozada, D. (1998). Environmental
Impact and Mechanisms of the Biological Clogging of Saturated Soils and Aquifer Materials Critical
Reviews in Environmental Science and Technology, 28(2):926-934.

Berger, P. (2002). Removal of Cryptosporidium using Bank Filtration in Ray, C (ed.) Riverbank
Filtration: Understanding Contaminant Biochemistry and Pathogen Removal, Kluwer Academic
Publishers, Netherlands, 85-121.

Brandt, R.K., Hughes, M.R., Bourget, L.P., Truszkowska, K., Greenler, R.G. (1993). The Interpretation
of CO Adsorbed on Pt/SiO2 of two Different Particle-Size Distributions, Surface Science, Vol. 286, 15-
25.

Grabs, W. (1981). Beitrag zur Beschreibung von Kolmationserscheinungen in einem

Organisch Belasteten Kleingewsser Beitrge zur Hydrologie 2, 293-311.

37
 Grischek, T. (2003). Zur Bewirtschaftung von Uferfiltratfassungen an der Elbe Institut fr
Grundwasserwirtschaft Technische Universitt Dresden, Heft 4.

Grischek, T. (2006). Investigations into River Bed Clogging at RBF Sites along the Elbe River,
Germany International Workshop on Riverbank/ Riverbed Filtration.

Hubbs, S.A. (2003). Plugging in Riverbank-Filtration Systems: Evaluating Yield-Limiting

Factors Program and Abstracts, The Second International Riverbank Filtration Conference, National
Water Research Institute, Cincinnati, Ohio.

Hubbs, S. A. (ed.) (2004). Evaluation Streambed Sources Impacting the Capacity of Riverbank Filtration
Systems Riverbank filtration hydrology: Impacts on System Capacity and Water Quality, NATO Science
Series, IV-Earth and Environmental Science, Vol. 60.

Kuehn, W. (1999). Overview of Riverbank Filtration Issues Abstracts Riverbank Filtration Conference,
Nov 4-6, Louisville Kentucky.

Kuehn, W., Mueller, U. (2000). Riverbank Filtration: an Overview Journal American Water Works
Association, 92, 60-69.

Partinoudi, V. (2004). Riverbank Filtration as a Viable Pretretment and Treatment Method M.S.Thesis,
Univ. of New Hampshire, Durham, NH.

Ray, C., Schubert, J., Linsky, R.B., Melin, G. (2002). Introduction in Riverbank Filtration: Improving
Source Water- Quality Kluwer Academic Publishers, Dordrecht, The Netherlands.

Ray, C., Grischeck, T., Schubert, J., Wang, J., Speth, T. (2002a). A Perspective of Riverbank Filtration
Journal of American Water Works Association, 94(4): 149-160.

van Riesen, S. (1975). Uferfiltratverminderung durch Selbstdichtung an Gewssersohlen (Shrinkage of

bank filtrate by clogging of the riverbed). Dissertation, Fakultt fr Bauingenieur- und
Vermessungswesen, Universitt Karlsruhe.

Schubert, J. (2005). River hydrology and morphology RBF Workshop.

Schubert, J. (2005a). Processes in bank filtration RBF Workshop.

Schubert, J. (2005b). River-aquifer interactions - the clogging process RBF Workshop.

Sontheimer, H. (1980). Experiences with Riverbank Filtration along the Rhine River, Journal AWWA
72, 386-390.

Tufenkji, N., Ryan, N.J., Elimelch, M. (2002). The Promise of Bank Filtration: A Simple Technology
may Inexpensive Clean up Poor-Quality Raw Surface Water Environmental Science and Technology, 36:
422A-428A.

US EPA (2003). LT2ESWTR Toolbox Guidance Manual Office of Water.

Wang, J. (2002). Riverbank Filtration Case Study at Louisville, Kentucky in Riverbank Filtration:
Improving Source -Water Quality, C. Ray, G. Melin, and R.B. Linsky, eds., Kluwer, Academic Publishers,
Dordrecht, The Netherlands.

Figures/ Tables

Figure 2-1
Riverbank Filtration Processes. Source: Amy et. al., 2006, page 104.

38
Figure 2-2
Paved and clogged riverbed near the outer section of a bend (at Flehe waterworks, Dsseldorf, well site).
Source: Schubert, 2005, page 3.

Figure 2-3
Riverbank Filtration Processes. Source: Grischek, 2003, page 8.

Brown, R.A. (2003). Application of the Long Term 2 Enhanced Surface Water Treatment Rule Microbial
Toolbox at Existing Water Plants Program and Abstracts, The Second International Riverbank Filtration
Conference, National Water Research Institute, Cincinnati, Ohio.

Grischek, T., Schoenheinz, D., Ray, C. (2003) Siting and Design Issues for Riverbank Filtration
Schemes in Riverbank Filtration: Improving Source -Water Quality, C. Ray, G. Melin, and R.B. Linsky,
eds., Kluwer Academic Publishers, Dordrecht, The Netherlands.

Partinoudi, V. (2004). Riverbank Filtration as a Viable Pretreatment and Treatment Method M.S.Thesis,
Univ. of New Hampshire, Durham, NH.

Partinoudi, V. (2009). Personal communication.

Ray, C., Schubert, J., Linsky, R.B., Melin, G. (2002). Introduction in Riverbank Filtration: Improving
Source Water- Quality Kluwer Academic Publishers, Dordrecht, The Netherlands.

Ray, C., Grischeck, T., Schubert, J., Wang, J., Speth, T. (2002a). A Perspective of Riverbank Filtration
Journal of American Water Works Association, 94(4): 149-160.

Regli, S., (2003). Potential Uses of Bank Filtration for Regulatory Compliance Regulatory Program and
Abstracts, The second International Riverbank Filtration Conference, National Water Research Institute,
Cincinnati, Ohio.

US EPA (2009a). Complying with the Long Term 2 Enhanced Surface Water Treatment Rule: Small
Entity Compliance Guide Office of Water.

US EPA (2007). The Long Term 2 Enhanced Surface Water Treatment Rule (LT2ESWTR)
Implementation Guidance Office of Water.

US EPA, (2005). Occurrence and Exposure Assessment for the Final Long Term 2 Enhanced Surface
Water Treatment Rule Office of Water.

US EPA (2005a). Rule Fact Sheet - Long Term 2 Enhanced Surface Water Treatment Rule [online].
Available from : http://www.epa.gov/safewater/disinfection/lt2/regs_factsheet.html [Accessed: 1
November 2009]

US EPA (2004). Chapter I-Environmental Protection Agency

Part 141-National Primary Drinking Water Regulations, 40 CFR Ch. I (7104 Edition).

US EPA (2002). Chapter II-Environmental Protection Agency Regulations: Long Term 1 Enhanced
Surface Water Treatment Rule; Final Rule 40 CFR Parts 9, 141, and 142
National Primary Drinking Water.

Figures/ Tables

Table 1-1
Current drinking water rules (by date issued). Source: US EPA Website -
http://www.epa.gov/safewater/regs.html.

Table 1-2
Treatment levels in RBF classification. Source: Partinoudi, 2003, page 4.

39
Table 1-3
Information and Guidance to meet LT2ESWTR requirements. Source: US EPA, 2009a, page 30 ff.

Table 1-4
Selected Site Data for RBF Systems in the United Sates and Germany. Source: Grischek et al., 2003, page
293.

Table 1-5
Microbial Toolbox Summary Table: Options, Treatment Credits and Criteria. Source: US EPA, 2007,
page 77.

Hoen, E. (2002). Hydrogeological Issues of Riverbank Filtration-A Review in Riverbank Filtration:

Understanding Contaminant Biochemistry and Pathogen Removal Ray, C. ed., Kluwer, Academic
Publishers, Dordrecht.

Heij, G.J. (1989). River-Groundwater Relationships in the Lower Parts of the Netherlands, J. Hydrol.,
108(1-4), 35-62.

Partinoudi, V. (2009). Personal communication.

Partinoudi, V. (2004). Riverbank Filtration as a Viable Pretretment and Treatment Method M.S.Thesis,
Univ. of New Hampshire, Durham, NH.

Schubert, J. (2005b). River-aquifer interactions - the clogging process RBF Workshop.

Schubert, J. (2001). Hydraulic Aspects of Riverbank Filtration Field Studies Journal of Hydrology
266, 145-161.

Schubert, J. (2002). German Experience with Riverbank Filtration Systems in Riverbank Filtration:
Improving Source -Water Quality, C. Ray, G. Melin, and R.B. Linsky, eds., Kluwer Academic Publishers,
Dordrecht, The Netherlands.

US EPA (2003). LT2ESWTR Toolbox Guidance Manual Office of Water.

Figures/ Tables

Figure 1-2
Gaining River. Source: Hoehn, 2002, page 21.

Figure 1-3
Losing River. Source: Hoehn, 2002, page 21.

Figure 1-4
Flow Trough River. Source: Hoehn, 2002, page 21.

Figure 1-5
Surface water level of the river Rhine (1988-1990). Source: Schubert, 2001, page 147.

Figure 1-6
Water temperature in Ohio River and in the production well. Source: Schubert, 2005b, page 3.

Figure 1-7
Chloride concentration in the river Rhine water compared that in the adjacent well water. Source:
Schubert, 2002, page 40.

Table 1-11
Characteristics of the Rhine river in the lower Rhine valley region. Source: Schubert, 2002, page 36.

40
41