f/EPA
United States
Environmental Protection
Agency
Project Summary
Exfiltration in Sewer
Systems
Robert S. Amick and Edward Burgess
Many municipalities throughout the
United States have sewerage systems
(separate and combined) that may expe-
rience exfiltration of untreated wastewater
from both sanitary and combined sewers.
This study was conducted to focus on the
estimation of the magnitude of leakage of
sanitary and industrial sewage from sani-
tary sewer pipes on a national basis.
The method for estimating exfiltration
amounts utilized ground-water table in-
formation to identify areas of the country
where the hydraulic gradients of the sew-
age are typically positive, i.e., the sewage
flow surface (within pipelines) is above
the groundwater table. An examination of
groundwater table elevations on a na-
tional basis reveals that the contiguous
United States is comprised of groundwa-
ter regions (established by the U.S. Geo-
logical Survey) which are markedly
different. Much of the northeastern, south-
eastern, and midwestern United States
has relatively high groundwater tables
that are higher than the sewage flow sur-
face, resulting in inflow or infiltration. Con-
versely, a combination of relatively low
groundwater tables and shallow sewers
creates the potential for widespread
exfiltration in communities located in the
western United States.
This study provides information on typi-
cal sewer systems, identifies and as-
sesses the factors that cause or probably
cause exfiltration, presents commonly
used and advanced corrective measures
and their costs for dealing with
exfiltration, identifies technology gaps,
and recommends associated research
needs and priorities. This report also ex-
amines urban exfiltration, including a
case study of Albuquerque, New Mexico.
This Project Summary was developed
by EPA's National Risk Management Re-
search Laboratory, Cincinnati, Ohio, to
announce key findings of "Exfiltration in
Sewer Systems" that is fully documented
in a separate report of the same title (see
Project Report ordering information on
back).
Introduction
Sanitary sewer systems are designed
to collect and transport to wastewater
treatment facilities the municipal and in-
dustrial wastewaters from residences,
commercial buildings, industrial plants,
and institutions, together with minor or in-
significant quantities of ground water,
storm water, and surface waters that inad-
vertently enter the system. Over the years,
many of these systems have experienced
major infrastructure deterioration due to
inadequate preventive maintenance pro-
grams and insufficient planned system re-
habilitation and replacement programs.
These conditions have resulted in dete-
riorated pipes, manholes, and pump sta-
tions that allow sewage to exit the
systems (exfiltration) and contaminate
adjacent ground and surface waters, and/
or enter storm sewers.
Untreated sewage from exfiltration of-
ten contains high levels of suspended sol-
ids, pathogenic microorganisms, toxic
pollutants, floatables, nutrients, oxygen-
demanding organic compounds, oil and
grease, and other pollutants. Exfiltration
can result in discharges of pathogens into
residential areas; cause exceedances of
water quality standards (WQS) and/or
pose risks to the health of the people liv-
ing adjacent to the impacted streams,
lakes, ground waters, sanitary sewers,
and storm sewers; threaten aquatic life
and its habitat; and impair the use and
enjoyment of the Nation's waterways.
Although it is suspected that significant
exfiltration of sewage from wastewater
collection systems occurs nationally, there
is little published evidence of the problem
-------
and no known attempts to quantify or
evaluate it on a national basis. Accordingly,
the objectives of this study were to quantify
through desk-top estimates the magnitude
of the exfiltration problem in wastewater-
collection systems on a national basis;
identify the factors that cause and contrib-
ute to the problem; and document the cur-
rent approaches for correcting the problem,
including costs. The resulting information
was used to identify information and tech-
nology gaps and research priorities.
Procedure
Causative Factors
A search for publications regarding
exfiltration sewage from wastewater collec-
tion systems did not locate any exfiltration-
specific discussion of unique/causative
factors because most factors which cause
inflow/infiltration are identical to those as-
sociated with exfiltration (i.e., they both oc-
cur through leaks in pipes, depending on
the relative depth of the ground water).
Factors that contribute to exfiltration in-
clude:
• size of sewer lines
• age of sewer lines
• materials of construction (sewer
pipe, point/fitting material, etc.)
type and quality of construction
(joints, fittings, bedding, backfill)
• depth of flow in the sewer
Geological conditions that contribute to
exfiltration include:
groundwater depth (in relation to
sewer line/depth of flow of sewage)
• type of soil
• faults
Climate conditions that influence
exfiltration include:
• average frost line in relation to sewer
depth
average rainfall, which helps deter-
mine groundwater depth
The level of ground water and the depth
of flow in the sewer will influence the extent
of exfiltration rates, since the pressure dif-
ferential between the hydraulic head in the
sewer and the groundwater hydraulic head
will force water out of the sewer apertures
into the surrounding soil material.
Health and Environmental Im-
pacts
Ground Water
Little published data is available on spe-
cific incidents of groundwater pollution and
associated health/environmental impacts
arising from leaking sewers, despite the
widespread acknowledgment that these in-
cidents occur. Several studies have indi-
cated widespread pollution of ground
water in urban areas arising from the gen-
eral leakiness of sewers, including bacteria
and ammonium reported from Wisconsin
and general pollution in the San Joaquin
Valley in California.
Transport of the sewage and pollutants
leaking into the subsurface/ ground water
depends on a variety of factors, including
but not limited to the difference in hydrau-
lic head between the sewage surface and
the groundwater table level, the substrate
physical/chemical/biological characteristics
(which determines attenuation potential),
and the sewage pollutants and their con-
centrations. Fecal bacteria contamination is
the most serious health risk associated
with domestic sewage exfiltration. Contami-
nation by viruses, protozoa, and other mi-
croorganisms is also a concern. Increased
concentrations of total organic carbon, ni-
trate, chloride, and sulfate, however, can
also make the water unfit for consumption.
Phosphate and boron are good indicators
of sewage pollution since they are not natu-
rally occurring in ground water.
The solids present in sewage can plug
the porous media beneath the pipe and
rapidly decrease the exfiltration rate. In an
experiment completed to examine this ef-
fect, the leakage was reduced to a steady
state within an hour.
Water Supply Distribution Systems
Because of minimum separation require-
ments for potable water supply distribution
systems and sanitary sewers and vigilant
application of cross-connection control pro-
grams, the opportunity for sewer exfiltration
to contaminate drinking water supplies is
theoretically rather limited. Sewage from
exfiltration can enter a distribution system
through a broken water main or, under re-
duced pressure conditions, through a hole
which leaks drinking water out under nor-
mal positive pressure conditions. Situations
which could allow infiltration of the sewage
through a lowering of water main pressure
primarily involve backflow and surges.
Despite the best efforts of utilities to re-
pair water main breaks using good sanitary
procedures, these breaks represent an op-
portunity for contamination from exfiltration
to enter the distribution system. When a
main breaks, utilities typically isolate the
affected section, superchlorinate, and flush
the repaired pipe. Flushing velocities may
not always remove all contaminated debris,
however, and microbiological testing of the
final water quality may not detect contami-
nating microorganisms. In 1989, Cabool,
Missouri experienced a suspected cross-
connection between sewage overflow and
two major distribution system line breaks
(backflow may have occurred during simul-
taneous repair of numerous water meters)
caused by freezing temperatures, resulting
in 243 cases of diarrhea, 32 hospitaliza-
tions, and four deaths due to E. coli
O157:H7 strain. This town of 2000 was on
an untreated groundwater system and did
not superchlorinate during repairs of the
water main breaks.
Backflow devices to prevent the entry of
contaminated water constitute an important
distribution system barrier. Because of cost
considerations, backflow-prevention de-
vices are primarily installed on commercial
services lines at facilities that use potentially
hazardous substances.
Recent research is focusing on transient
pressure waves that can result in hydraulic
surges in the distribution system. These
waves, having both a positive and negative
amplitude, can draw transient negative
pressures that last for only seconds and
may not be observed by conventional pres-
sure monitoring. Because these waves
travel through the distribution system, at
any point where water is leaking out of the
system, the transient negative pressure
wave can momentarily draw water and
sewage (if present) back into the pipe.
Surface Water
No data or narrative information in the lit-
erature demonstrate, or even suggest, that
sewer exfiltration has directly contaminated
surface waters. Several factors that control
the occurrence of sewer exfiltration may
explain the absence of a linkage between
exfiltration and surface water pollution.
The occurrence of exfiltration is limited to
those areas where sewer elevations lie
above the groundwater table. Since
groundwater elevations near surface water
bodies are typically near the ground sur-
face, sewers near surface water bodies
generally are below the groundwater table,
and infiltration (rather than exfiltration) will
dominate the mode of sewer leakage in
these areas. In areas of steep topographic
conditions, where sewers are located near
surface waters and at elevations that lie
above the surface water, exfiltration impacts
may be possible. However, these situations
are assumed to be sufficiently rare that
exfiltration impacts on surface waters are
not observed.
Exfiltration Magnitude Estimat-
ing Methodology
The process of estimating the magnitude
of the exfiltration problem on a national
scale performed as a series of two inde-
pendent steps:
-------
Qualitatively assessing the portion
of the nation's sewer systems that
are susceptible to exfiltration;
Applying assumptions about
exfiltration rates (percent of base
sewer flow) to the exfiltration sus-
ceptible sewer systems to provide
an assessment of the extent of
sewer exfiltration on a national
scale.
Identification of Exfiltration
Susceptible Sewer Systems
The key factor influencing the occurrence
of exfiltration is the direction of the hydrau-
lic gradient between the sewer flow surface
and the groundwater table (GWT) external
to the sewer. In much of the northeastern,
southeastern, and midwestern United
States, relatively high groundwater tables
typically result in infiltration conditions. Ex-
ceptions include shallow sewers and ser-
vice laterals, and seasonal variation in
GWTs that can significantly change the
spatial extent of the sewer system that lies
above the GWT (i.e., that can be consid-
ered to be "exfiltration susceptible").
Given the importance of first screening
out those areas that are not "exfiltration
susceptible," the initial desktop analysis
task was to perform spatial analysis of
sewer depth relative to regional GWT el-
evations. Existing national-scale ground-
water information was examined, such as
that provided by the U.S. Geological Survey
(e.g., USGS Groundwater Regions of the
United States). As the various national
ground-water data sources were reviewed,
however, it was determined that mapping in
support of the purposes of this study was
not readily available. For this reason, a na-
tional depth-to-ground water map was pre-
pared under this project from groundwater
level data available in the national data-
bases (U.S. EPA STORET and USGS
WATSTORE).
National-scale sewer depth data does
not exist, but for purposes of the desktop
analysis some assumptions about this pa-
rameter can be made. For example, typical
service lateral depth can be assumed to be
8 feet for buildings with basements, and 2
to 4 feet for houses built on slabs. Typical
sewer main depth can be assumed to be 6
to 10 feet; it may be possible for more de-
tailed assessments to develop a typical
depth distribution (i.e., x % 4-10 ft deep, y%
11-15 ft deep, z% > 15 ft deep). Regional
differences should be considered; for ex-
ample, sewer depths typically are shal-
lower in the western United States than in
other areas of the country. Sewer system
density (miles/acre) can be correlated with
readily available national population den-
sity data to create a GIS coverage of sewer
system density.
GIS processing incorporating the gen-
eral spatial (mapped) relationships be-
tween sewer depth and groundwater
elevations allowed the development of a
characterization of the "exfiltration suscep-
tibility" of various areas. This was attempted
at the national level, but the data required
to support this analysis are unavailable;
thus, a representative area (Albuquerque,
New Mexico) for which a recent exfiltration
study had been completed, was selected
on which to perform the analysis. National
exfiltration rate assessments can be ex-
trapolated from this analysis. However,
more detailed identification and inventory of
exfiltration susceptible areas is required to
support a meaningful quantification of na-
tional exfiltration rates.
For purposes of this study, unit rates for
exfiltration (gallons/day/ inch/mile) avail-
able from the 1989 EPA study were used to
generate the assessment of the magnitude
of the national exfiltration problem. These
unit rates were applied to the "exfiltration
susceptible" areas (together with assump-
tions about the inch-miles of sewers/later-
als in those areas) to generate exfiltration
rates in the Albuquerque case study. The
unit rates based on gallons/day/inch/mile
were compared with estimates based on
percent of base sewer flow. Comparisons of
the two methods proved useful in develop-
ing the final estimates.
Corrective Measures
The proper selection of corrective or re-
habilitation methods and materials de-
pends on a complete understanding of the
problems to be corrected, as well as the
potential impacts associated with the selec-
tion of each rehabilitation method. Pipe re-
habilitation methods to reduce exfiltration
(and simultaneously infiltration) fall into one
of the two following categories:
External rehabilitation methods
• Internal rehabilitation methods
Certain conditions of the host pipeline in-
fluence the selection of the rehabilitation
method. It is therefore necessary to assess
these factors to prepare the pipe for reha-
bilitation. Rehabilitation is proceeded by
surface preparation by cleaning the pipe to
remove scale, tuberculation, corrosion, and
other foreign matters.
External Sewer Rehabilitation
Methods
External rehabilitation methods are per-
formed from the aboveground surface by
excavating adjacent to the pipe, or the ex-
ternal region of the pipe is treated from in-
side the pipe through the wall. Some of the
methods used include external point re-
pairs, chemical grouting, and cement
grouting.
Internal Sewer Rehabilitation
Methods
The basic internal sewer rehabilitation
methods include chemical grouting (most
common method for sealing joints), cured-
in-place pipe (insertion of a flexible lining
impregnated with a thermosetting resin),
sliplining (pipes are inserted into an exist-
ing line by pulling or pushing pipes into a
sewer), closed-fit pipe (uses temporarily
deformed new pipe), fold and form pipe
(deformed into a "U" shape before inser-
tion), spiral wound pipe (winds strips of
PVC in a helical pattern to form a continu-
ous pipe), pipe bursting (fragments existing
pipe and replaces with new pipe in void),
and spot (point) repair (used to correct
spot problems).
Results and Discussion
The findings of the Albuquerque case
study were combined with the national
depth-to-groundwater mapping to present
a qualitative assessment of the extent to
which sewer exfiltration represents a risk to
water quality and human health on a na-
tional scale.
National Scale Quantification
Although exfiltration is not a widely stud-
ied phenomenon, several exfiltration stud-
ies and investigations have been
completed throughout the world. These in-
clude work completed in the United States
for the U.S. EPA and several studies in Eu-
rope, the majority of which are focused on
Germany. Some of the more applicable
previous studies are discussed below.
Three basic approaches have been used
to quantify sewer exfiltration rates: (1) direct
measurement of flow in isolated sewer seg-
ments; (2) theoretical estimates using
Darcy's Law and related hydraulic theory;
and (3) water balance between drinking
water produced/ delivered and wastewater
collected/ treated. Each of these ap-
proaches has been applied to the Albu-
querque case study and is described
below.
Estimates Based on Direct
Measurements (U.S. EPA Study)
An EPA study entitled "Evaluation of
Groundwater Impacts of Sewer Exfiltration"
-------
was completed in the late 1980's. The work
measured exfiltration in two California city
sewer systems to develop a correlation be-
tween exfiltration and infiltration. The tests
were conducted in areas of vitrified clay
pipe (VCP) predominance, where older
pipe of known or suspected poor condition
existed. Only those pipe segments located
above groundwater levels were tested.
Water consumption was metered for all
sewer service connections corresponding
with each measured sewer line to deter-
mine the actual quantity of wastewater flow
entering the system. It was assumed that all
internal household water entered the sewer
system. Measurements of sewage flow in
the sewer lines were made by continuous
flow monitoring and by hydrostatic testing.
Calculated sewer exfiltration was reported
in units of gallons per inch diameter per
mile length per day (gpimd).
The study revealed that a large discrep-
ancy exists between the results from the
continuous flow monitoring and the hydro-
static testing at one Santa Cruz location.
The study concludes that the continuous
flow monitoring achieved reliable data and
that the hydrostatic test data was influenced
by the tidal cycle. A correlation model be-
tween exfiltration and infiltration was devel-
oped, but not field tested.
A second evaluation was performed us-
ing field measurements at another location
to verify the correlation model. This evalu-
ation used similar methodologies as the
first task. Exfiltration measurements were
made in the Washington Suburban Sani-
tary Commission (WSSC) sewer system
near Washington, D.C., and in Lexington
Kentucky.
Several problems with the measurement
methodologies were noted, and overall the
hydrostatic test method was judged to be
not successful. It was resolved that the flow
monitoring procedure worked well and
should be applied to areas with a minimum
of 400-500 linear feet of pipe with little or no
service connections.
Estimates Based on Darcy's Law
and Related Theory (European
Studies)
The study of exfiltration has been of
great interest in Germany. This country has
a very old, deteriorated infrastructure. The
cost to complete the necessary repairs to
Germany's sewer systems is estimated to
be nearly $100 billion (U.S.). Therefore,
several exfiltration studies have been con-
ducted to prioritize repair work. These stud-
ies have both applied theoretical (Darcy's
Law) approaches and direct measurements
to estimate sewer exfiltration. Excerpts from
some of the studies are summarized below.
• A report from England provided an
estimate of 300 x 106 m3/yr (793 x
108gal/yr) or approximately 1 liter/
day/m (397 gal/day/mile) for the
exfiltration of the 880,000 km
(547,000 miles) of sewer lines in
Germany, although the basis of the
estimate is not clear. This very low
sewer leakage rate is actually net
exfiltration, which is the difference
between exfiltration and infiltration.
The study indicates that total
exfiltration and infiltration in Ger-
many are nearly equal, but the
amounts are not provided.
To better understand the mechanics
of exfiltration, sewage migration
from leaking pipes to ground water
was correlated in a study using
Darcy's Law (see Equation 1). The
rate of exfiltration is linearly depen-
dent on the area of the pipe
exfiltrating and the pressure head:
(1) Q = LAdh
where Q is the exfiltration rate (ft3/s)
through a pipe leak area A (ft2) at a pres-
sure head of dh (ft), and L is leakage fac-
tor (s-1).
The leakage factor is defined in Equa-
tion 2:
(2) L = K/dl
where K is the permeability of the sur-
rounding soil (ft/s) and dl is the thickness of
the soil layer (ft). This study found that the
settle able solids in the wastewater act to
reduce the permeability of the bedding ma-
terial and lower the exfiltration rate rapidly
at low flows and velocities. This clogging
reduces the rate of exfiltration immediately.
In fact, a steady-state rate of exfiltration was
reached after one hour, even with large
area of joint damage.
A research project undertaken by
the Institute of Environmental Engi-
neering (ISA) at the University of
Technology of Aachen, Germany,
studied the water pollution hazard of
leaking sewers. The ISA developed
and used a special exfiltration mea-
suring device at every joint in sev-
eral sections of sewer pipe on
several tests conducted throughout
Germany. This study determined that
the most significant VCP sewer dam-
ages which permit exfiltration are
leaking service junctions, leaking
sewer joints, pipe cracks, and pipe
fractures. At a pressure head below
the sewer crest, which is typically
the case in gravity flow sewer lines,
exfiltration rates were minimal. At a
pressure head of one pipe diameter,
the exfiltration rate increased dra-
matically, to more than 26 gph per
joint in some segments. This high
leakage rate can in part be attributed
to the generally poor condition of the
old sewer systems.
Estimates Based on Drinking
Water- Wastewater Balance
Exfiltration from Albuquerque's sewer
system was estimated using a water/sew-
age balance calculation, backed up by
some previous local studies on infiltration.
The results are then compared with leak-
age rates calculated from the other meth-
odologies and unit rates derived from the
EPA and European studies presented
above.
A direct method for estimating exfiltration
is to compare water pumpage and usage
with wastewater received at Albuquerque's
Southside Water Reclamation Plant
(SWRP). To make this comparison, it is nec-
essary to identify the base water demand,
which is the indoor component of the total
household use. Demands during mid-win-
ter (January and February) are assumed to
be near base flow because no or very mini-
mal outdoor water usage occurs. Water and
wastewater data obtained from the City for
January 1998 revealed the following:
Average daily influent flow at the
SWRP:51.4mgd
Average daily water pumpage into
transmission/distribution system:
61.2 mgd (this is then considered to
be the daily base flow for that month)
Subtracting wastewater flow from the
pumpage rate yields a difference of 9.8
mgd, which is the first approximation of
sewage leakage. However, several other
factors also impact the water balance in the
water and wastewater systems. These are:
• Sewer infiltration
• In-house water consumption
• Water distribution system leakage
• Sewer exfiltration
City of Albuquerque staff, using a range
of available information (including meter
and billing records, pumpage records, and
other data), have estimated losses in the
water system at about 11 percent of the to-
tal amount pumped. A 1997 study found
water system losses ranging from 8 percent
in Hong Kong, which is considered to have
a relatively "tight" and high-quality system,
to the 20-25 percent range in England,
which has many very old distribution sys-
tems. An 11 percent loss in the system
would account for a daily average loss of
about 6.73 mgd.
-------
In-house consumption is that portion of
the water entering the house that does not
leave as sewage, but is consumed in cook-
ing, drinking, watering plants, cleaning, etc.
National experience indicates that about 3
percent of water entering the home is con-
sumed on an average day in January
1998. With negligible non-domestic con-
sumption, the remaining amount of water,
about 1.4 mgd, represents the net differ-
ence between the two other factors in the
water balance: sewer infiltration and
exfiltration. The net amount is positive, indi-
cating that exfiltration exceeds infiltration by
1.4 mgd, which is plausible given that the
great majority of Albuquerque's sewers,
and particularly those most susceptible to
exfiltration (older VCP), are in exfiltration
areas (well above groundwater levels).
In order to estimate the exfiltration vol-
ume, previous studies addressing infiltra-
tion in the Albuquerque sewer system
were reviewed. One of the studies utilized
several approaches to gain an approxima-
tion of inflow and infiltration in the Albu-
querque system, most of which was
attributed to infiltration in the valley of the
Rio Grande. Some of these methodologies
are described below:
• A flow comparison between winter
water use and sewage flow. This
methodology resulted in an infiltra-
tion flow of 3.7 mgd. However, the
report stated that "this estimation is
probably within + 50 (percent) of the
actual value..."
Early morning sewage flow versus
water use. This methodology resulted
in an infiltration flow of nearly zero.
Sewage flow versus population. Us-
ing a 100-gallons-per-capita-per-
day wastewater flow and a
population of 300,000, infiltration
was estimated at 5 mgd. It was also
noted that the average sewage flow
for Albuquerque at this time was ac-
tually 117 gpcd.
• Influent BOD versus domestic
wastewater BOD. The expected BOD
concentration in the wastewater was
calculated based upon a generally
accepted BOD loading of 0.17 Ib/
cap/day. This BOD concentration
was compared with the average in-
fluent concentration to calculate an
infiltration flow of 5.9 mgd. However,
this was thought to be a high esti-
mate based upon the relatively small
industrial component and the high
institutional contribution.
In addition, the study field-verified the ar-
eas subject to infiltration. Based upon the
above calculations and results of the field
tests, infiltration was thought to be some-
what less than 3 mgd, or 9 percent of the
wastewater flow in 1975. Nine percent of
today's wastewater flow would be in the 5-
mgd range.
Another infiltration analysis was com-
pleted as part of the Albuquerque ASAM
Model Loading and Verification Task. Inter-
ceptor manholes which were within 2 feet
of ground water were identified. Flow moni-
toring was completed in a sewer subbasin,
and the resulting flows were compared with
the predicted flows to determine infiltration.
The infiltration rate for Albuquerque was
calculated at 0.925 mgd, but, again, the im-
pact of exfiltration was not included. There-
fore, the work revealed a net infiltration
rate, indicating that actual infiltration is
about 1 mgd greater than total exfiltration.
From the foregoing investigations, it is es-
timated that the total average infiltration
rate for the Albuquerque system is in the
vicinity of 3.5 mgd. The 9 percent field-veri-
fied rate reported in the Molzen-Corbin re-
port is probably high, given the repair and
replacement of major interceptors in the
valley that have occurred since 1975, as
well as the use of better quality materials
and construction techniques for new pipe-
lines since then. On the other hand, repairs
have generally not been made to the sew-
ers most susceptible to exfiltration (old
VCP pipes).
The total exfiltration rate is obtained by
adding the 1.4 mgd remaining in the water
balance to the infiltration rate, for a total of
4.9 mgd, or approximately 5 mgd.
National Depth to Groundwater
Mapping
In order to extrapolate the Albuquerque
findings to a national scale, a qualitative
assessment of exfiltration susceptibility us-
ing depth-to-ground water information was
made. Since no such mapping at a national
scale suitable for this purpose was readily
available, an initial mapping effort was un-
dertaken as part of this study.
The development of a nationwide depth-
to-ground water atlas is difficult at best due
to the lack of easily obtainable data for most
of the country. Data to determine the depth
to the shallowest water table may be gath-
ered from local, state, federal, and private
sources through well logs, water level mea-
surements, location of wetlands and seeps,
characterization of streams and rivers, and
locations of lakes and other water bodies.
A thorough characterization of the U.S. wa-
ter table is a long and exacting process.
Within the context of this study, a depth-
to-groundwater map was created using
readily available data from the EPA
STORET and USGS WATSTORE data-
bases of depth-to-groundwater parameters.
The data were downloaded from CDROM
databases resident at the COM Hydrodata
Center in Denver, Colorado. The data were
screened to eliminate missing depth-to-
water values, missing latitude and longi-
tude, duplicate data, and easily recognized
anomalous data. The resultant set con-
tained approximately 93,000 data points in
the coterminous United States, Alaska, and
Hawaii (only the coterminous U.S. is
shown below). Since the data retrieved
from STORET and WATSTORE is depen-
dent upon the data owner for accuracy,
there is no comprehensive method of qual-
ity control. USGS data are continually re-
viewed, however, and these data may be
deemed reasonably accurate. The STORET
and WATSTORE databases, while cer-
tainly robust, do not contain all data avail-
able; therefore, data gaps exist.
Despite the large dataset applied to build
the map, many regions of the United States
have relatively limited data; these areas are
unshaded on the map. Areas with the
greatest concentration of valid data points
within the deep ground water range are
generally west of the Mississippi River and
along the Appalachian Mountains.
Conclusions
Most of the urban areas in the northeast-
ern, southeastern, and coastal areas of the
U.S. have relatively shallow groundwater
tables (<15feet). In these areas of the U.S.,
where a significant portion of the population
(and therefore sewer systems) exists, rela-
tively few exfiltration susceptible sewer sys-
tems are expected. One caveat is
exfiltration from service laterals. Even in
these areas of the U.S., many shallow ser-
vice laterals may exist above groundwater
tables. However, the hydraulic head avail-
able to drive exfiltration in these service
lines is generally very low (typically only
one or two inches, and intermittent). Further
study in this area may be warranted to as-
sess the extent of service lateral exfiltration.
Based on a review of the depth-to-
groundwater map, it is expected that wide-
spread exfiltration is probably limited to a
relatively small portion of the total U.S.
population, as relatively few large urban
areas in the U.S. are located in these
deeper groundwater areas. Cities such as
Albuquerque, Phoenix, Tucson, and others
are among the larger urban areas where
significant exfiltration potential exists. Fur-
ther study of exfiltration conditions in cities
such as these, with relatively large areas
with sewers above the groundwater table,
may be warranted on a case-by-case basis
-------
where evidence of exfiltration (e.g., ground-
water contamination) has been observed,
or is revealed by more detailed evaluations.
Areas with extremely deep groundwater
tables probably experience relatively less
risk associated with exfiltration due to the
long subsurface travel times and distances
of the exfiltrated sewage from the sewer to
the groundwater table. Areas with signifi-
cant portions of the system above, but in
close proximity to, the groundwater table
are probably at greatest risk. There is an in-
creased risk in the relatively few areas with
significant exfiltration potential when there
is, for example, a thin soil and fractured
rock hydrogeologic setting which allows
pathogens and other contaminants from
the sewage to reach the ground water
quickly and with minimal attenuation. How-
ever, since public water supplies are treated
with chlorination, ozonation, or other sys-
tems to kill fecal bacterial contamination, an
added measure of protection is provided.
A greater potential problem, albeit iso-
lated, may be exfiltration from sewers car-
rying industrial wastewater. Organic and
inorganic constituents of industrial sewage
can be overall much more persistent than
those of domestic sewage, and therefore
much more likely to reach the ground wa-
ter in areas of significant exfiltration poten-
tial. The disposition of industrial sewage
contaminants which reach ground water
used for drinking water supplies may not
be the same as that of fecal bacteria from
domestic sewage [i.e., the treatment pro-
cesses (flocculation, filtration, chlorination,
activated carbon filtration, et. al) may not
eliminate or reduce these contaminants to
render them harmless]. Untreated well wa-
ter in some rural, small community, com-
mercial, and private-owner drinking water
systems does not enjoy this added protec-
tion. However, these systems are not typi-
cally in close proximity to large
municipalities and associated sewer sys-
tems/exfiltration potential.
The Albuquerque Case Study concluded
that the rate of exfiltration from that sewer
system, expressed as a percentage of base
flow, is on the order of 10% of average daily
base wastewater flow - in absolute terms,
roughly 5 mgd. This rate, expressed as an
average annual rate, is 1825 Mg/yr. An-
other relevant conclusion of the Albuquer-
que study was that there is a greater impact
on ground water from septic tank usage
than from sewer exfiltration. As the forego-
ing depth-to-ground water analysis indi-
cates, however, exfiltration is expected to
vary significantly on a regional basis. Fur-
ther study should expand the initial depth-
to-ground water analysis performed here
and identify more precisely the "exfiltration
susceptible" sewer systems throughout the
U.S. and the extent to which exfiltration im-
pacts ground water in these systems.
In summary, exfiltration does not appear
to be a significant national problem based
on an evaluation of 1) available groundwa-
ter table data to nationally assess the ex-
tent to which sewer systems are susceptible
to exfiltration, 2) past studies of measured
and estimated exfiltration rates, and 3) pro-
tective mechanisms, particularly natural
soil/hydrogeological setting attenuation
and drinking water treatment plants.
Exfiltration may be a regional or more likely
local problem where the GWT lies closely
under the sewage flow surface and/or
where the exfiltrate can reach even deep
ground water through a thin soil/fractured
rock hydrogeologic setting especially
where persistent, potentially toxic contami-
nants are present, such as those often as-
sociated with industrial sewage.
Corrective Measure Costs
Given the relatively high rates of
exfiltration that potentially discharge from
exfiltration susceptible sewer systems in the
U.S., corrective measures may be required
to adequately protect the groundwater re-
sources, and in some limited instances sur-
face waters, in these areas. Given the
site-specific nature of exfiltration problems,
however, a more detailed assessment of
the larger urban areas in the exfiltration
susceptible western U.S. should be com-
pleted before a meaningful estimate of cor-
rective costs can be developed.
Corrective actions to address exfiltration
in those situations where local-level evalu-
ation calls for such action will generally be
accomplished with similar technologies as
those used to address infiltration. Although
an estimate of national-scale costs to ad-
dress exfiltration must follow more detailed
evaluation of exfiltration-susceptible sewer
systems, corrective action costs on a unit
basis [i.e., cost ($) per lineal foot of sewer]
were developed for this study and ranged
from $60 per lineal foot for an 8-inch-diam-
eter sewer to $590 per lineal foot for a 36-
inch-diameter sewer.
Recommendations
This study identified the following data/
technology gaps associated with
exfiltration. Recommendations for research
and development to fill these gaps were
developed for each data/ technology gap
identified.
1. Data Gap - comprehensive national
depth-to-groundwater maps: Although
3.
a large portion of the U.S. has readily
available, accurate depth-to-ground-
water data, many regions of the United
States have relatively limited data.
Recommendation:
An effort to refine the initial depth-to-
groundwater mapping produced in this
study with an expanded and updated
database would support a more de-
tailed national estimate of exfiltration
and the cost of associated corrective
measures.
Data Gap - extent of exfiltration in mu-
nicipalities: There are relatively few
large urban areas in the U.S. which
have the potential for widespread
exfiltration. Western arid U.S. cities
such as Albuquerque, Phoenix, and
Tucson are among the larger metro-
politan areas where significant
exfiltration potential exists and little is
known about it. Albuquerque's
exfiltration has recently been exten-
sively studied.
Recommendation
Further study of localized exfiltration
conditions in cities with high exfiltration
potential may be warranted on a case-
by-case basis where evidence of
exfiltration has been observed, or is re-
vealed by more detailed groundwater
study. This study should be preceded
by assessment using the refined
depth-to-ground water mapping rec-
ommended above to produce a na-
tional inventory of exfiltration
susceptible areas. This localized study
will be of greater value than an attempt
to quantify the problem nationally, due
to the localized nature of the problem.
Data Gap - exfiltration fate and trans-
port: No information is available re-
garding the biological disposition of
sewage exfiltrate. Also, it would be
useful to determine if a biological crust
forms in the bedding below an
exfiltrating sewer that would serve to
insulate/protect ground water and/or
water supply distribution systems.
Recommendation:
Research to fill the exfiltration disposi-
tion data gap could involve the use of
existing sewage systems known or
determined to be leaking in significant
amounts (using carefully excavated
examination of the bedding beneath
and adjacent to the leaking sewer
joints) or by construction of an experi-
mental leaking sewer system (artifi-
cially introducing sewage into the
sewer systems bedding). An analysis
-------
of bedding samples from points at in-
creasing depths and horizontal dis-
tances from the leak would help to
reveal the extent of exfiltrate transport.
4. Combined/Separate Sewer Consider-
ations for Detailed Urban Study
Recommendation
The sewer systems to be considered
in future exfiltration assessments
should include both combined and
separate sewer areas, since combined
sewers are often located in highly ur-
banized areas where imperviousness
is high. The result is a decreased rain-
fall infiltration into the soil and lowering
of the GWTs, making these sewers po-
tentially more susceptible to
exfiltration. Additionally, combined
sewers are often shallower than sepa-
rate sewers, older than separate sew-
ers, and constructed with
less-watertight pipe joints - all factors
that can contribute to higher exfiltration
rates. Another special case that must
be considered in more detailed studies
is force mains. Although they are often
constructed with tighter pipe joints and
more durable pipe material, they none-
theless operate under pressure and
may therefore be more exfiltration sus-
ceptible.
5. Inclusion of Service Laterals
Recommendation
It will be important to more detailed
exfiltration assessments of urban areas
to consider service laterals together
with public sewers in identifying and
evaluating the exfiltration susceptible
sewers, as laterals are the shallowest
portion of the sewer system (largest
hydraulic gradient difference with
GWT) and typically of the poorest con-
struction.
Robert Amick is with Environmental Quality Management Inc,
Cincinnati, OH 45240 and Edward Burgess is with Camp,
Dresser & McKee, Cincinnati, OH 45249.
Ariamalar Selvakumar is the EPA Project Officer (see below).
The complete report, entitled "Exfiltration in Sewer Systems" is
available at http://www.epa.gov/ORD/NRMRL/Pubs/
600R01034/600R01034.pdf.
The EPA Project Officer can be contacted at:
Water Supply and Water Research Division
National Risk Management Research Laboratory
U.S. Environmental Protection Agency
Edison, NJ 08837-3679
-------
vvEPA
Umteii States
National Risk Management
Research Laboratory
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
Please make all necessary changes on the below label,
detach or copy, and return to the address in the upper
left-hand corner.
If you do not wish to receive these reports, CHECK HERE
detach, or copy this cover, and return to the address in the
upper left-hand corner.
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
EPA/600/SR-01/034
March 2003
------- |