Water Quality: Better Data and Evaluation of Urban Runoff	 
Programs Needed to Assess Effectiveness (29-JUN-01, GAO-01-679). 
								 
The Environmental Protection Agency (EPA) regards the		 
contaminants in storm water runoff as a significant threat to	 
water quality across the nation and considers it to be one of the
most significant reasons that water quality standards are not	 
being met nationwide. Prompted by Congress, EPA has responded	 
with a variety of programs, including the National Pollutant	 
Discharge Elimination System Storm Water Program, which requires 
more than 1,000 local governments to implement storm water	 
management programs. Those municipalities that are currently	 
involved in Phase I of the program have been attempting to reduce
pollutants in storm water runoff for several years. It is time to
begin evaluating these efforts. However, EPA has not established 
measurable goals for this program. In addition, the agency has	 
not attempted to evaluate the effectiveness of this program in	 
reducing storm water pollution or to determine its cost. The	 
agency attributes this problem to inconsistent data reporting	 
from municipalities, insufficient staff resources, and other	 
competing priorities within the Office of Wastewater Management. 
Although Phase I municipalities report monitoring and cost data  
to EPA or state regulatory agencies annually, these agencies have
not reviewed this information to determine whether it can be of  
use in determining the program's overall effectiveness or cost.  
GAO's analysis shows that the reported cost information will be  
difficult to analyze unless EPA and its state partners set	 
guidelines designed to elicit more standardized reporting. Better
data on costs and program effectiveness are needed--especially in
light of the Phase II program that will involve thousands more	 
municipalities in 2003. EPA's planned research grant to the	 
University of Alabama and its pilot project to analyze data from 
annual reports and develop baseline indicators is a step in the  
right direction and could point the way for a more comprehensive 
approach.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-01-679 					        
    ACCNO:   A01269						        
  TITLE:     Water Quality: Better Data and Evaluation of Urban Runoff
             Programs Needed to Assess Effectiveness                          
     DATE:   06/29/2001 
  SUBJECT:   Cost analysis					 
	     Program evaluation 				 
	     Reporting requirements				 
	     Water quality					 
	     Water pollution control				 
	     Performance measures				 
	     Clean Water State Revolving Fund			 
	     EPA National Pollutant Discharge			 
	     Elimination System 				 
								 
	     EPA National Pollutant Discharge			 
	     Elimination System Storm Water Program		 
								 
	     EPA Total Maximum Daily Loads Program		 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-01-679
     
A

Report to Congressional Requesters

June 2001 WATER QUALITY Better Data and Evaluation of Urban Runoff Programs
Needed to Assess Effectiveness

GAO- 01- 679

Letter 3 Results in Brief 4 Background 6 Volume of Urban Runoff Increases
With the Expansion of Urban

Development and Can Affect Water Quality 12 Local Governments Take Actions
to Manage Urban Storm Water

Runoff, but Information Is Limited on the Cost and Effectiveness of These
Actions 26 Conclusions 37 Recommendation 37 Agency Comments and Our
Evaluation 38

Appendixes Appendix I: The Storm Water Program in Baltimore City, Maryland
40 Appendix II: The Storm Water Program in Boston, Massachusetts 42 Appendix
III: The Storm Water Program in Los Angeles County, California 44

Appendix IV: The Storm Water Program in Milwaukee, Wisconsin 48 Appendix V:
The Storm Water Program in Worcester, Massachusetts 51 Appendix VI: Comments
From the Environmental Protection Agency 55 Appendix VII: GAO Contacts and
Staff Acknowledgments 58

Tables Table 1: Storm Water Pollutants in Urban Runoff, Including Sources
and Potential Impacts 21

Table 2: Comparative Health Outcomes for Swimming in Front of Drains Versus
400 or More Yards Away 24 Table 3: Summary of Fiscal Resources Projected for
Los Angeles

County and Its Co- permittees, Fiscal Years 1997- 99 46 Table 4: City of
Worcester?s Capital Expenditures for Storm Water

Management 53 Figures Figure 1: Urban Runoff Flows in Different Types of
Sewer Systems 10

Figure 2: Increase in Urbanized Land in Selected Cities, 1960- 90 14 Figure
3: Percentage of Paved Public Road Miles, 1945- 97 15 Figure 4: Impact of
Impervious Surfaces on the Amount of Storm

Water That Runs Off, Infiltrates, and Evapotranspires 17 Figure 5: Damage
Caused by Storm Water Runoff From Urbanized

Areas in the Gingerville Creek Subbasin 20

Figure 6: Comparison of Town Lake PAHs and Traffic Trends 25

Abbreviations

BMP best management practices CSO combined sewer overflow DOT Department of
Transportation DPW Department of Public Works EPA Environmental Protection
Agency FHWA Federal Highway Administration MCTT Multi- Chambered Treatment
Tank MDE Maryland Department of Environment MS4 Municipal Separate Storm
Sewer System NPDES National Pollutant Discharge Elimination System PAH
polycyclic aromatic hydrocarbons PCB polychlorinated biphenyls TEA- 21
Transportation Equity Act for the 21 st Century TMDL total maximum daily
load USGS U. S. Geological Survey WDNR Wisconsin Department of Natural
Resources WPDES Wisconsin Pollutant Discharge Elimination System

Lett er

June 29, 2001 The Honorable Olympia Snowe United States Senate

The Honorable Sherrod Brown The Honorable Martin Meehan The Honorable James
Oberstar The Honorable Jack Quinn House of Representatives

Nonpoint source pollution- that is, pollution from contaminants picked up
and carried into surface water by water running over land- is known to be
one of the leading causes of water quality problems in the United States.
Water that runs over developed areas, including paved surfaces such as roads
and parking lots, before reaching a water body is known as urban

runoff and is an increasingly important category of water pollution. As
urban areas have expanded over the past several decades, the amount of urban
runoff has also increased. Although the overall quality of the nation?s
waters has improved since the passage of the Clean Water Act in 1972, a
significant number of water bodies still suffer from poor water quality.
Because the act brought discharges from ?point sources,? such as industrial
plants and municipal treatment plants, under control, the continuing

pollution of these waters suggests that other sources, including urban
runoff, are contributing to water quality problems. As a result, the
Environmental Protection Agency (EPA) now classifies urban runoff as a

significant cause of impairment to water quality. The Water Quality Act of
1987, which amended the Clean Water Act, required EPA, among other things,
to regulate as a point source urban runoff that reaches municipal sewer
systems. EPA?s National Pollutant Discharge Elimination System Program for
storm water requires that certain local governments take measures to control
storm water runoff.

Concerned about the degradation of water quality in urban areas, you asked
us to report on (1) the amount of runoff from urban areas, particularly from
roads, highways, and other impervious surfaces, 1 and its effects on water
quality and (2) the programs that federal regulations require local
governments to develop to address urban runoff, and the

costs and effectiveness of those programs. To address these issues, we
reviewed federal and other studies and interviewed experts on the
relationship between the amount of paved and other impervious surfaces and
the amount of runoff, and on the types of materials typically contained in
urban runoff. We also reviewed studies and interviewed experts on the
sources of these materials and any actual or potential effects on water
quality from urban runoff. We visited five urban areas and organizations

that are affiliated with their watersheds 2 to obtain site- specific
information about urban runoff problems, programs these areas have
implemented in response to federal requirements, and the costs and
effectiveness of these programs. Finally, we reviewed studies and estimates
of the costs and investment requirements associated with implementing storm
water management programs. Because this report focuses on local governments?
actions, we did not review the portions of the National Pollutant Discharge

Elimination System Storm Water Program that address industrial facilities
and construction sites. We performed our review from August 2000 through May
2001 in accordance with generally accepted government auditing standards.
Results in Brief The volume of urban storm water runoff increased throughout
the United States in the last half of the 20 th century because of the
growth in

impervious surfaces that resulted from the development of urban and suburban
areas. According to the U. S. Department of Agriculture, between 1945 and
1997, land devoted to urban areas in the United States has increased by
about 327 percent; according to EPA, paved road mileage has

increased by 278 percent. Because paved surfaces are almost impervious, they
allow little storm water to infiltrate the ground; therefore, the storm
water runs off into creeks, rivers, and lakes. As storm water runs across
these impervious surfaces and land, it picks up pollutants from these
surfaces and carries them to receiving bodies of water- either directly or

1 An impervious surface keeps water from soaking into soils. 2 A watershed
is an area of land in which all surface water drains to a common point.

through conveyances such as gutters, storm sewers, and culverts. EPA?s 1998
National Water Quality Inventory Report to Congress showed that certain
rivers, streams, lakes, and estuaries are impaired in terms of their

ability to support such uses as aquatic life, swimming, and fish
consumption, and concluded that urban runoff was a major source of this
impairment. Studies have shown that urban runoff and the pollutants it
carries can cause increases in sedimentation, water temperature, and
pathogen levels and decreases in dissolved oxygen levels in bodies of water.
These changes can lead to the degradation of habitat in these water bodies
and a decline in diversity of aquatic life and can endanger public health.
For example, metals, a pollutant typically found in urban runoff, can be
toxic to aquatic organisms. Pathogens, such as bacteria from animal waste,
another pollutant commonly found in urban runoff, can pose public health
problems when present in waters used for recreational purposes.

The magnitude and nature of these effects vary by region, depending on the
type and concentration of pollutants in storm water, rainfall
characteristics, land use, and other factors.

Local governments are required to address urban runoff through EPA?s
National Pollutant Discharge Elimination System Storm Water Program. Under
permits that EPA and states issue through this program, over 1,000 local
governments must meet EPA?s requirements to implement storm water management
programs to reduce contaminants in storm water to the ?maximum extent
practicable.? EPA recommends that these cities use ?best management
practices? to reduce contaminants in storm water runoff. The most typical
practices included controlling runoff through a combination of structural
means, such as detention ponds, and nonstructural means, such as increasing
the frequency of street sweeping and educating the public about how to
prevent pollutants from reaching storm sewers. Cities also used specialized
practices to address specific local runoff problems. For example, Baltimore,
Maryland, has focused on reducing the level of nutrients, such as
fertilizers, in its runoff because of its proximity to the Chesapeake Bay,
which suffers from high nutrient levels.

Neither the overall costs of implementing the storm water program nor the
program?s effectiveness has been determined. EPA estimated in a 1996 report
to congress that the potential need for spending on storm water runoff and
overflows of sewage resulting from runoff was over $50 billion over 20
years, but the agency also believes this estimate will increase when it
issues its next report in 2002. EPA?s regulations require that permitted
cities annually report the costs of implementing their storm water

programs, along with the results of their monitoring of storm water runoff
and water quality. However, in part because EPA has not established
guidelines for reporting costs, these data have not been calculated or
reported consistently and, therefore, are not currently useful in
characterizing the program?s overall cost. EPA, state, and city officials
generally believe that managing storm water runoff will reduce the volume of
runoff and concentrations of pollutants in the runoff, as well as improve

water quality, but no systematic effort to evaluate the program?s results
has been started. EPA and the states have generally been unsuccessful in
developing measurable program goals and in demonstrating program
effectiveness through the review of water quality data reported by local

governments. We believe it is time for EPA to begin evaluating this program,
which is directed at one of the nation?s most significant water quality
problems. Therefore, this report includes a recommendation to EPA to work
with states to develop program goals, establish standards for reporting on

program costs and effectiveness, and review reported water quality data to
determine whether the current storm water management programs are having the
intended effect of improving the quality of the nation?s waters and how much
the programs cost. We provided a draft of this report to EPA and the
Department of Transportation (DOT). EPA generally agreed

with the report and plans to take action to implement several parts of the
recommendation; the agency did not comment on the other parts of the
recommendation. DOT generally agreed with the report. (See the Agency
Comments and Our Evaluation section of this report.)

Background Nonpoint source pollution can result when water, such as
precipitation, runs over land surfaces and into bodies of water. Significant
nonpoint sources of pollution can include paved urban areas, agricultural
practices,

forestry, and mining. However, in urban and suburban areas, this runoff
generally enters a sewer system that can be regulated as a point source of
water pollution. For example, precipitation from rain or snowmelt may run
into a municipal separate storm sewer system (MS4 or storm sewer) that
eventually discharges into a body of water. The precipitation may also run
into a combined sewer system, which carries a combination of storm water
runoff, industrial waste, and raw sewage in a single pipe to a sewage

treatment facility for discharge after treatment. Lastly, the precipitation
may run off of land or paved surfaces directly into nearby receiving waters.

EPA?s Office of Wastewater Management, which is within the Office of Water,
implements the National Pollutant Discharge Elimination System (NPDES)
Program. The program was created in 1972 with the passage of the Clean Water
Act. Created to control water pollution from point sources- those sources,
such as a factory or wastewater treatment plant, that contribute pollutants
directly into a body of water from a pipe or other

conveyance- the NPDES Program did not specifically address storm water
discharges. In 1987, the Congress amended the Clean Water Act with the Water
Quality Act, which directed EPA to also control storm water discharges that
enter MS4s- essentially requiring EPA to treat such storm water as a point
source. 3 MS4s are defined as those sewers that collect and

convey storm water; are owned or operated by the federal, state, or local
government; and are not part of a publicly owned treatment (sewage)
facility. To regulate urban storm water runoff, EPA published regulations in
1990 that established the NPDES Storm Water Program and described permit
application requirements. According to EPA, the program?s objective, in
part, is to preserve, protect, and improve water quality by, among other

things, controlling the volume of runoff from paved surfaces and by reducing
the level of runoff pollutants to the maximum extent practicable using best
management practices (BMP). 4 The 1987 act also authorized

EPA to implement a program that provides federal funds and technical
assistance to states to develop their own nonpoint source pollution
management programs. States can use the federal funds they receive for
nonpoint source programs to address nonpoint sources of pollution as well as
urban runoff.

Currently, EPA manages NPDES Storm Water programs in six states (Alaska,
Arizona, Idaho, Massachusetts, New Hampshire, and New Mexico) and has
delegated authority to the remaining 44 states to manage these programs. The
storm water program is being implemented in two phases. Local governments
meeting the following criteria must comply with EPA?s storm water program
regulations. First, Phase I of the program requires that municipalities with
a population of 100,000 or more obtain a permit for their MS4 system;
second, the program requires that entities obtain a

3 Section 402( p) of the Clean Water Act. 4 According to EPA, a best
management practice is a device, practice, or method for removing, reducing,
retarding, or preventing targeted storm water runoff constituents,
pollutants, and contaminants from reaching receiving waters.

permit if they discharge storm water from sites with industrial activities,
including construction activities that disturb 5 acres or more of land. In
addition, NPDES permitting authorities may also bring other municipalities
and industrial entities into the program if they deem it necessary.

Municipalities that meet these conditions must submit a permit application
to EPA or the governing regulatory state agency. In 1990, the regulations
specifically identified 220 municipalities throughout the United States that
were required to apply for a Phase I permit. According to EPA, as of April
2001, about 256 Phase 1 MS4 permits had been issued and about 17 more still
needed to be issued. Because some permits cover more than one

municipality, these permits cover about 1, 000 medium and large
municipalities nationwide. The final rule for Phase II of the program was
issued in December 1999.

Phase II extends Phase I efforts by requiring that a storm water discharge
permit be obtained by (1) operators of all MS4s not already covered by Phase
I of the program in urbanized areas 5 and (2) construction sites that
disturb areas equal to or greater than 1 acre and less than 5 acres of land.

As with Phase I of the program, permitting authorities may require
additional small MS4s and construction sites to obtain a permit if they are
a significant contributor of pollutants. Currently, EPA anticipates that
about 5,000 municipalities may be subject to permitting requirements under

Phase II of the storm water program. These municipalities are required to
obtain permits no later than March 10, 2003.

5 The Bureau of the Census generally defines an urbanized area as a land
area comprising one or more places- central place( s)- and the adjacent
densely settled surrounding area- urban fringe- that together have a
residential population of at least 50, 000 and an overall population density
of at least 1,000 per square mile.

EPA also regulates combined sewer overflows (CSO) that can be caused by
urban storm water runoff. Combined sewer systems, in which storm water
enters pipes already carrying sewage, may overflow when rain or snowmelt
entering the system exceeds the system?s flow capacity. In the CSO that

results, the mixture of untreated sewage and runoff bypasses the water
treatment facility and is diverted directly into receiving waters. (See fig.
1 for an illustration of combined and separate sewer systems.) These
combined systems generally serve the older parts of approximately 900 cities
in the United States. Pipes carrying sewage and storm water separately
generally serve newer parts of cities. EPA?s 1994 CSO policy requires
communities with combined sewer systems to take immediate and

long- term actions to address CSO problems. The policy contains provisions
for developing appropriate, site- specific NPDES permit requirements for all
combined sewer systems that overflow because of wetweather events. The Wet
Weather Water Quality Act of 2000 requires that any permit, order, or decree
issued for a CSO conform to the 1994 policy. Under this act, EPA is also
required to submit a report to the Congress by September 2001 on the status
of the program. 6

6 Sanitary sewer overflows, which are illegal under the Clean Water Act, can
also result from rainfall. A sanitary sewer overflow may occur when
rainwater or snowmelt leaks into sanitary sewage pipes, thereby exceeding
the pipes? capacity and causing them to overflow. This discharge of raw
sewage from municipal sanitary sewer systems can release untreated sewage
into places such as streams, basements, and streets. EPA proposed
regulations to require municipalities to reduce the number of overflows.
However, these regulations have been withdrawn for further review.

Figure 1: Urban Runoff Flows in Different Types of Sewer Systems Combined
Sewer System

Sanitary wastewater Storm water Storm Storm drain

drain Snowmelt Sanitary wastewater

Wastewater treatment plant Fully treated Combined effluent sewer overflow
Sanitary sewage/ wastewater Storm water runoff with potential contaminants

Separate Sanitary Separate Storm

Sewer System Sewer System Storm

drain Sanitary wastewater

Storm drain Storm

Sanitary wastewater drain

Storm water

Erosion Storm water

point source Lawn Care

Nonpoint source pollution

Source: GAO illustration based on EPA data.

The Total Maximum Daily Load (TMDL) Program, established under the Clean
Water Act, is intended to address water bodies that do not meet water
quality standards because of pollutant loadings from point and nonpoint
sources. Currently, it is unclear how and when this program will affect
EPA?s and states? issuance of storm water permits. A TMDL is a calculation
of the maximum amount of a pollutant that a body of water can receive and
still meet the water quality standard set by the state. Under EPA?s
regulations, the state is to allocate this ?pollutant load? among the point
and nonpoint pollutant sources that flow into the water body and then take
steps to ensure that no source exceeds its assigned load. In 1996,

EPA issued a policy that outlined an interim approach to including water
quality standards in storm water permits. The policy promoted the use of
BMPs in the first 5- year term permits, followed by a tailoring of BMPs in
the second round of permits as necessary to comply with water quality
standards. Until recently, few TMDLs had been established, and citizen
organizations sued EPA for its lack of action. EPA issued a new set of
regulations for the TMDL Program in 2000, but the Congress prevented

EPA from spending money to implement the rule in 2000 and 2001. It is
possible that establishing a TMDL for a body of water could result in the
application of a numeric effluent limit to outfalls 7 that release storm
water into that body of water. Some city officials we spoke with generally
felt that numeric effluent limits would significantly increase the cost of
managing storm water.

Volume of Urban Since World War II, urban runoff has increased throughout
the United Runoff Increases With

States. This increase is directly related to growth in the amount of
impervious surfaces due to urban and suburban development and the the
Expansion of

construction of roads, highways, and other impervious surfaces. Urban
Development

Coinciding with this growth in impervious surfaces has been a reduction in
and Can Affect Water wetlands and in the amount of storm water that
infiltrates the ground to recharge aquifers. Moreover, the loss of
vegetation due to development

Quality and related runoff can cause major erosion. Ultimately, much of this
runoff

is channeled into gutters, storm drains, and paved channels, and vegetation
and sediment removed with the runoff may end up in receiving waters. EPA has
identified urban storm water runoff as one of the leading sources of
pollution to the nation?s rivers, streams, lakes, and estuaries. Runoff from
impervious surfaces picks up potentially harmful pollutants and

7 An outfall is an outlet, such as a pipe, that allows storm water to flow
into a river, lake, or other body of water.

carries them into receiving waters. Studies have shown that urban runoff and
the pollutants it carries can negatively affect water quality, aquatic life,
and public health.

Paved Surfaces Have According to the U. S. Department of Agriculture,
between 1945 and 1997, Increased With Urban and

urban land area increased by almost 327 percent, from 15 million acres to
Suburban Expansion and

about 64 million acres in the contiguous 48 states. From 1992 through 1997,
Growth in Automobile Use the annual rate of development averaged about 1
million acres per year.

The land developed between 1945 and 1997 came primarily from forestland and
pasture and range. 8 For example, according to the Bureau of the Census,
between 1960 and 1990, the amount of land used for urban

purposes in Baltimore, Maryland, and Washington, D. C., grew by about 170
percent and 177 percent, respectively. As a result, urbanization, with its
accompanying expansion of impervious surfaces like sidewalks , roofs,
parking lots, and roads, has significantly increased the nation?s total
developed land and paved surface area. 9 Figure 2 demonstrates the growth in
the urbanized areas of Baltimore and Washington, D. C., over the last half
of the 20 th century. 8 Agricultural Resources and Environmental Indicators,
2000, U. S. Department of Agriculture, Economic Research Service, Resource
Economics Division. 9 Our Built and Natural Environments, A Technical Review
of the Interaction Between Land Use, Transportation and Environmental
Quality, U. S. Environmental Protection Agency (EPA 231- R- 00- 005, Nov.
2000).

Figure 2: Increase in Urbanized Land in Selected Cities, 1960- 90

Source: U. S. Geological Survey.

The increase in paved surfaces has been spurred not only by urban and
suburban development, but also by a steady increase in the use of
automobiles, the primary mode of daily transportation for most Americans.
Roads also play an important role in the economy of the United States,

since trucks carry about 75 percent of the value of all goods shipped.
According to EPA, paved road mileage in the United States increased by 278
percent from 1945 to 1997. In 1945, 19 percent of the public roads in the
country were paved; by 1997, that percentage had increased to 61. (See

fig. 3.) According to a 1999 study, motor- vehicle infrastructure, such as
roads and parking lots, accounts for close to half of the land area in U. S.
urban cities. 10 10 Stormwater Strategies, Community Responses to Runoff
Pollution, Natural Resources

Defense Council (May 1999).

Figure 3: Percentage of Paved Public Road Miles, 1945- 97 Percent

70 60 50 40 30

A 20

10 0

1945 1955 1965 1975 1985 1997 Source: EPA.

Increase in Impervious The increase in impervious surfaces over the past
several decades has led

Surfaces Leads to Increased to an increase in storm water runoff. In part,
this has occurred because Runoff

highways and other developments have reduced the amount of wetlands and
other undeveloped land. Wetlands mitigate the effects of storm water runoff
by acting as a natural form of flood control, facilitating sediment
replenishment, and improving water quality by removing excess nutrients and
other chemical contaminants before the contaminants can affect receiving
waters. According to a 2000 EPA report, 11 of the 12 states that listed
wetland losses, six reported that they had significant losses due to highway
construction, and 10 reported that they had significant losses due to
residential growth and development. However, the effect of road building on
wetland loss has been reduced in recent years. According to a

Federal Highway Administration (FHWA) official, since 1996, wetlands have
been replaced and restored under the Federal- Aid Highway Program 11 See
footnote 9.

at an average rate of 2. 7 acres for every acre lost to highway building.
Other undeveloped land with vegetation also performs some of the roles that
wetlands play in managing runoff, although to a lesser extent.

Furthermore, as impervious surfaces increase, less storm water is able to
infiltrate through the soil to groundwater. Impervious areas allow only a
very small amount of initial infiltration compared with unpaved areas whose
infiltration capacity varies, depending on the soil type. Figure 4
demonstrates EPA?s estimates of the impact of impervious surfaces on the
percentages of storm water that runs off, infiltrates the ground, and is
lost through evapotranspiration. 12 When natural ground cover is present
over an entire site, normally 10 percent of precipitation runs off the land
into

nearby creeks, rivers, and lakes. In contrast, when a site is 75- to
100percent impervious, 55 percent of the precipitation runs off into these
receiving waters. However, according to an FHWA official, the runoff rates
can be reduced if developers take mitigating actions to develop and
implement BMPs to control flooding or runoff.

12 Evapotranspiration represents water loss from evaporation and the
absorption and eventual release into the atmosphere of water that plants and
trees have collected. The extent to which evapotranspiration occurs is
dependent primarily on the solar energy available to vaporize the water. As
a result, the effect of evapotranspiration varies greatly across the
country.

Figure 4: Impact of Impervious Surfaces on the Amount of Storm Water That
Runs Off, Infiltrates, and Evapotranspires

Source: EPA.

The decrease in storm water infiltration that accompanies urbanization also
reduces the amount of water that is available to recharge groundwater
supplies. For this reason, reduced infiltration may lead to problems with
the water table in certain urban areas. For example, a Massachusetts

Department of Environmental Protection official noted that a low recharge
rate affects water quality because it can result in a loss of wetlands and
adversely affect aquatic habitat as water- table levels fall during dry
weather. 13 In addition, officials from the Charles River Watershed
Association in Massachusetts are concerned that the lack of infiltration

might cause some communities to run short of drinking water in the next 20
years. Urban Runoff Has the

Urban runoff can adversely affect the quality of the nation?s waters, and
Potential to Impair Water

urban storm water runoff has been identified as one of the leading sources
of pollution to rivers, streams, lakes, and estuaries. 14 Section 305( b) of
the Quality and Disrupt

Clean Water Act requires states and other jurisdictions to report on the
Biological Integrity

quality of their waters to EPA every 2 years. The 1998 National Water
Quality Inventory Report to Congress showed that 35 percent of assessed
river and stream miles, 45 percent of assessed lake acres, and 44 percent of
assessed estuarine square miles were impaired in terms of their ability to
support uses such as aquatic life, swimming, and fish consumption. 15 The

report identified urban storm water runoff as one of the leading sources of
impairment to the assessed waters. 13 Dry weather is defined as a period
when rainfall measuring at least 0.10 of an inch has not occurred for 72
hours. 14 Other leading sources of pollution include agricultural runoff,
municipal point sources, hydrologic modifications, and atmospheric
deposition.

15 Information contained in the 1998 report reflects only those waters
assessed by states and other jurisdictions and cannot be used to
characterize nationwide water quality. Furthermore, water quality standards
among states are not identical, and the monitoring design used to collect
data differed among states.

Studies have shown that as the percentage of impervious cover increases
within a watershed, biodiversity also declines. Research conducted by the
Center for Watershed Protection found that, generally speaking, when a
watershed has 10 percent or less impervious cover, the associated stream can
be categorized as sensitive. 16 Sensitive streams are characterized as
having high fish diversity and good water quality. Once the percentage of
impervious cover exceeds 25 to 30 percent of the watershed, however,

streams tend to become nonsupporting. Nonsupporting streams are highly
unstable, have poor diversity of fish and aquatic life, and have poor water
quality. For example, one study evaluated the relationship between the
extent of impervious cover in watersheds to the number and diversity of fish
populations in 47 small streams in southeastern Wisconsin between the 1970s
and 1990s. 17 The results revealed that the number of fish species per site
was highly variable for drainage areas that had less than 10- percent
imperviousness. In contrast, sites that had greater than 10- percent

imperviousness had consistently low numbers of fish species. Other studies
have associated urban runoff with basic changes in the receiving body of
water. Runoff can carry sediment into surface water, and this sediment can
carry contaminants, harm aquatic plants, and smother organisms. Runoff can
also be warmed by the impervious surfaces it flows across. When sufficient
amounts of warmed runoff enter a water body, the

water temperature can rise. Less oxygen is then available for aquatic
organisms because water holds less oxygen as it becomes warmer. These
combined factors lead to the degradation of aquatic habitat. According to
EPA, the common effects of these types of pollution on aquatic life include
a decline in biodiversity and an increase in invasive species.

An increase in the volume of storm water runoff also increases the
likelihood of erosion, which allows for transport of eroded sediment
downstream into receiving waters. For example, during a site visit, we

observed extensive erosion along the Gingerville Creek Subbasin in Anne
Arundel County, Maryland, that was caused by urban runoff channeled into the
creek. Figure 5 depicts the eroded banks and channel of this creek.

16 ?The Importance of Imperviousness,? Watershed Protection Techniques, v.
1: 3, Fall, 1994. The article reviews 18 studies on the relationship between
urbanization and stream quality. 17 L. Wang and others, ?Watershed
Urbanization and Changes in Fish Communities in

Southeastern Wisconsin Streams,? Journal of the American Water Resources
Association, Oct. 2000, Vol. 36, No. 5.

Figure 5: Damage Caused by Storm Water Runoff From Urbanized Areas in the
Gingerville Creek Subbasin

Source: Anne Arundel County, Maryland, Department of Public Works.

Contaminants in Urban There have been several efforts to characterize the
chemicals and other

Runoff Can Affect Aquatic constituents in urban runoff. The Nationwide Urban
Runoff Program,

Life and Human Health conducted by EPA between 1978 and 1983, examined the
characteristics of urban runoff. Another federal effort to characterize
urban runoff is an ongoing joint project of the U. S. Geological Survey
(USGS) and the FHWA to evaluate guidelines for highway runoff. As table 1
indicates, these studies and others have shown that the principal
contaminants found in

urban runoff include nutrients, solids, pathogens, metals, hydrocarbons,
organics, salt, and trash. Water flowing over various surfaces, such as
streets, parking lots, construction sites, industrial facilities, rooftops,
and lawns, carries these pollutants to receiving waters . The contaminants
have the potential to impair water quality, degrade aquatic ecosystems, and
pose health risks to swimmers.

Table 1: Storm Water Pollutants in Urban Runoff, Including Sources and
Potential Impacts Contaminant Source Potential impact Nutrients

Nitrogen, Animal waste, fertilizers, failing septic systems, Nutrient
enrichment can cause an excessive growth of

phosphorous atmospheric deposition, a CSOs algae. Nuisance levels of algae
are associated with dissolved oxygen deficiencies leading to fish kills,
loss of submerged aquatic vegetation that serves as a habitat for aquatic
organisms, and loss of natural biodiversity.

Solids

Sediment Construction sites, other disturbed and/ or Sediment can cause
infection and disease among fish,

nonvegetated lands, eroding banks, road sanding scour submerged aquatic
vegetation, prevent sunlight from reaching aquatic plants, and bury bottom-
dwelling aquatic organisms. Pathogens

Bacteria, viruses Animal waste, failing septic systems, illicit Pathogens
entering waters used for recreational purposes

connections and discharges to storm sewer can pose human health risks.

system, CSOs

Metals

Lead, cadmium, Industrial processes, normal wear of automobile Metals can
cause acute or chronic toxicity for aquatic

copper, zinc, mercury, brake linings and tires, automobile emissions,
organisms. chromium, aluminum,

automobile fluid leaks, metal roofs and others

Hydrocarbons

Oil and grease, Industrial processes, automobile wear, automobile
Hydrocarbons have the potential to be acutely toxic for

polycyclic aromatic emissions, automobile fluid leaks, waste oil aquatic
organisms and several are suspected carcinogens. hydrocarbons Organics

Pesticides, Pesticides (herbicides, insecticides, fungicides,

Low concentrations of some organics have the potential to polychlorinated

rodenticides, etc.), industrial processes bioaccumulate in the food chain.
biphenyls (PCB), synthetic chemicals

(Continued From Previous Page)

Contaminant Source Potential impact Salt

Sodium Road salting and uncovered salt storage Salt can damage roadside
vegetation, transport high levels Chlorides of chlorides to receiving
waters, and degrade aquatic ecosystems. Chloride can be harmful to some
species of fish.

Trash

Street refuse and improperly discarded waste Trash impairs water quality by
inhibiting the growth of material aquatic vegetation and conveys nutrients,
toxic substances,

and other pollutants to aquatic ecosystems. a Atmospheric deposition occurs
when pollutants in the air fall on land or water.

Sources: Massachusetts Department of Environmental Protection Stormwater
Policy; EPA reports and guidance, including Preliminary Data Summary of
Urban Storm Water Best Management Practices, Combined Sewer Overflow Control
Policy, Innovative Urban Wet- Weather Flow Management Systems,

and the 1998 National Water Quality Inventory Report to Congress; the
California Regional Water Quality Control Board; the Natural Resources
Defense Council?s Stormwater Strategies: Community Responses to Runoff
Pollution; ?Accretion of Pollutants in Roadway Snow Exposed to Urban Traffic

and Winter Storm Maintenance Activities - Part I,? Draft; 18 and USGS?
National Water Quality Assessment Program.

18 J. J. Sansalone and D. W. Glenn, ?Accretion of Pollutants in Roadway Snow
Exposed to Urban Traffic and Winter Storm Maintenance Activities Part I,?
DRAFT.

In our visits to cities with Phase I permits and their watersheds, we
identified specific instances in which these contaminants had affected water
quality. The Chesapeake Bay, for example, has been polluted with the
nutrients nitrogen and phosphorus and with excess sediment caused, in part,
by urban runoff. The excess nutrients cause algae blooms that block

sunlight from reaching bay grasses- which are a source of food, shelter, and
nursery grounds for many aquatic species. In an effort to control nutrient
pollution in the Chesapeake Bay, the Executive Council of the Chesapeake Bay
Program 19 established a goal to reduce the nitrogen and phosphorus entering
the Chesapeake Bay by 40 percent, including through control of runoff from
urban areas. In addition, an assessment of the status of chemical
contaminant effects on living resources in the bay?s tidal rivers found ?hot
spots? of contaminated sediment. As a result, the Baltimore Harbor and the
Patapsco River in Maryland; the Anacostia River in Washington, D. C.; and
the Elizabeth River in Virginia were designated as

?regions of concern.? Urban storm water runoff is a significant source of
contaminants in the three regions. The Chesapeake Executive Council has
committed to reduce by 30 percent the chemicals of concern in the regions of
concern by 2010 through pollution prevention measures and other voluntary
means. 20

Pathogens such as bacteria and viruses, which are often present in urban
runoff, can pose public health problems. For example, the Santa Monica Bay
Restoration Project conducted a study to identify adverse health effects of
untreated urban runoff by surveying over 13,000 swimmers at three bay
beaches. 21 The study established a positive association between

an increased risk of illness and swimming near flowing storm- drain outlets.
Table 2 explains health outcome measures at various distances from storm
drains. For example, the study found a 1- in- 14 chance of fever for
swimmers in front of the drain versus a 1- in- 22 chance at 400 or more
yards away.

19 The Chesapeake Executive Council includes the governors of Maryland,
Pennsylvania, and Virginia; the Administrator of the U. S. Environmental
Protection Agency; the mayor of the District of Columbia; and the chair of
the Chesapeake Bay Commission.

20 Chesapeake Bay Program Office, Toxics 2000 Strategy: A Chesapeake Bay
Watershed Strategy for Chemical Contaminant Reduction, Prevention, and
Assessment, Dec. 2000. 21 R. W. Haile and others, ?The Health Effects of
Swimming in Ocean Water Contaminated by Storm Drain Runoff,? Epidemiology,
July 1999, Vol. 10, No. 4.

Table 2: Comparative Health Outcomes for Swimming in Front of Drains Versus
400 or More Yards Away

400 or Health outcomes 0 yards more yards

Fever 1: 14 1: 22 Chills 1: 26 1: 42 Ear discharge 1: 68 1: 143 Coughing
with phlegm 1: 20 1: 33 Significant respiratory disease (fever and

1: 12 1: 22 nasal congestion, fever and sore throat, and cough with phlegm)

Note: This table includes the statistically significant health outcomes.
Source: GAO analysis of data from ?The Health Effects of Swimming in Ocean
Water Contaminated by Storm Drain Runoff,? Epidemiology, July 1999, Vol. 10,
No. 4.

Metals and polycyclic aromatic hydrocarbons (PAH) in urban runoff can
present a threat to aquatic life. Studies have found the following:

 Storm water runoff from an urban area proved to be toxic to sea urchin
fertilization in the Santa Monica Bay, and dissolved zinc and copper were
determined to be contributors to this toxicity. 22  Brown bullheads (a
bottom- dwelling catfish) in the Anacostia River developed tumors that were
believed to be caused by PAHs associated in part with urban runoff. 23 
High PAH and heavy metal concentrations were found in crayfish tissue

samples from several urban streams in Milwaukee. The study associated these
contaminants with storm water runoff. 24

22 Southern California Coastal Water Research Project, Study of the Impact
of Stormwater Discharge on Santa Monica Bay Executive Summary, Nov. 1, 1999.
23 Chesapeake Bay Program Office. 24 J. P. Masterson and R. T. Bannerman,
?Impacts of Stormwater Runoff on Urban Streams in Milwaukee County,
Wisconsin,? National Symposium on Water Quality, American Water Resources
Association, Nov. 1994.

In addition, USGS tracked trends in the concentrations of PAHs found in
sediment in 10 lakes and reservoirs in six metropolitan areas over the last
several decades. This study found that PAH concentrations in developed
watersheds are increasing and that these increases may be linked to the
amount of urban development and vehicle traffic in urban and suburban

areas. 25 For example, from 1982 to 1996, PAH concentrations in the sediment
core in Town Lake (Austin, Texas) and total miles driven in greater Austin
both increased by about 2.5 times. Figure 6 illustrates this correlation.
Figure 6: Comparison of Town Lake PAHs and Traffic Trends

Total miles driven on PAHs detected in sampled core sediments

Austin roads (micrograms per kilogram)

(in thousands)

12, 000 14, 000

10, 000 12, 000

10, 000 8,000

A 8,000

6,000 6,000

4,000 4,000

2,000 2,000

0 0 1960

1962 1963

1964 1966

1972 1978

1982 1984

1986 1990

1992 1994

1995 1996

1997 1998

Date

Note: According to USGS, irregularities in the date pattern are due to
intervals at which sediment samples were collected. Source: USGS National
Water Quality Assessment Reconstructed Trends Program.

25 P. Van Metre, B. Mahler, and E. Furlong, ?Urban Sprawl Leaves Its PAH
Signature,? Environmental Science and Technology, Vol. 34, No. 19, 2000.

Although the studies we reviewed show that certain contaminants are likely
to be present in urban runoff, factors such as land development practices,
climate conditions, atmospheric deposition, and traffic characteristics all
can affect the characteristics of runoff from a particular area. Therefore,
given the diffuse nature of many storm water discharges and the variability
of other contributing factors, characterizing the

concentrations of pollutants contained in storm water runoff has been
challenging. Recent USGS reports also suggest that improvements are needed
in the methods used to analyze sediment and metals in runoff. 26 Local
Governments

To comply with federal and state storm water management for Phase I Take
Actions to

permitting requirements, permitted municipalities must create and implement
storm water management programs. The three primary Manage Urban Storm

activities used in these programs include efforts to characterize storm
Water Runoff, but

water runoff; BMPs aimed at reducing pollutants in storm water runoff to
Information Is Limited the maximum extent practicable; and reporting program
activities, monitoring results, and costs of implementing the program. Some
BMPs on the Cost and are structural- meaning that they are designed to trap
and detain runoff

Effectiveness of These until constituents settle or are filtered out. Other
BMPs are nonstructural- meaning that they are designed to prevent
contaminants Actions

from entering storm water through actions like street sweeping and
inspections. Many permitted municipalities use specialized BMPs tailored to
address particular runoff problems in their locations. Over 1,000 cities are
undertaking these efforts under the NPDES Storm Water Program, but
information on the overall costs of managing urban runoff and the
effectiveness of the actions taken is limited. EPA?s attempts to forecast
costs have not encompassed the entire program or are out of date. In
addition, the permitted municipal agencies we visited estimated their

annual storm water management costs and reported them to state agencies or
EPA, but the approaches they used to calculate these estimates varied
considerably, making it difficult to draw any conclusions. Although EPA and
state agencies believe that the program will be effective in improving water
quality, EPA has not made a systematic effort to evaluate the

program. Without such an effort, EPA cannot tell what effect the program is
having on water quality nationally.

26 The USGS reports indicate that certain methods used to analyze sediment
and metals samples can be unreliable. For example, sample collection and
processing methods can have an effect on measured concentrations of metals.

Municipalities Comply With The NPDES Storm Water Program requires
municipalities operating under

Federal and State a Phase I MS4 permit to characterize and monitor storm
water runoff,

Requirements Through implement BMPs to reduce pollutants to the maximum
extent practicable, Monitoring, Best

and report costs and monitoring results to the permitting authorities.
Because of these requirements, local governments have generally shifted
Management Practices, and

the focus of their storm water management from water quantity control or
Reporting

flood management to water quality concerns. Besides following the basic
federal requirements, municipalities must follow any additional regulations
developed by states that have been delegated the authority to manage the
NPDES Storm Water Program. For example, Wisconsin?s Department of Natural
Resources broadened the

requirements for determining which municipalities must get permits. The
state requires local governments with storm sewer systems in priority
watersheds (based on the significance of storm water runoff as a pollutant
source) that serve a populace of 50,000 or more 27 to obtain a permit with
requirements similar to those for a Phase I permit. Wisconsin?s Department
of Natural Resources also requires municipalities that are located in one of
the state?s five Great Lakes Areas of Concern 28 to obtain a state permit.
Furthermore, in line with specific criteria in Wisconsin?s Administrative
Code, the state requires other municipalities to obtain a permit if the

municipality is found to significantly contribute storm water pollutants to
waters of the state. These various requirements increased the number of
municipalities that must get permits from the two under federal requirements
to over 70 under the states? requirements.

The local governments we reviewed were undertaking three primary activities
when applying for permits and implementing their storm water management
programs. Specifically, these activities were (1) characterizing storm water
runoff; (2) developing BMPs to reduce discharges of pollutants to the
maximum extent practicable; and (3) reporting program activities, monitoring
results, and reporting program

costs. First, to characterize runoff, applicants are to provide quantitative
data that describe the volume and quality of discharges from municipal storm

27 For example, we visited West Allis, Wisconsin, which has a permit even
though its population is under 100,000. 28 Areas of concern have persistent
water quality problems, which impair beneficial uses.

sewers. For example, cities must map all storm sewer outfalls- an
undertaking that one group representing cities described as significant.
After the permit application is approved, additional monitoring is required
throughout the life of the permit to facilitate the design of effective
storm water management programs and to document the nature of the storm
water. The local governments we visited were all monitoring for a variety of
purposes, including characterizing runoff from different types of land

use in order to target their BMPs, testing the effectiveness of a particular
BMP, or establishing a baseline for their storm water quality evaluations.

Second, the storm water management programs that local governments develop
focus on implementing BMPs. While active treatment, such as sending storm
water through a treatment facility, is a possible BMP, the cities we visited
were generally not using active treatment. EPA?s February 2000 report 29 on
the Phase I program described the program as based on the ?use of low- cost,
common- sense solutions.? The five cities we visited were generally using
similar types of structural and nonstructural BMPs, as follows:  Structural
BMPs are designed to separate contaminants from storm

water. For example, detention ponds temporarily hold storm water runoff to
allow solids and other constituents in the runoff to settle before the water
is released at a predetermined rate into receiving waters. In addition,
catch- basin inserts, placed in a storm drain, catch trash and other debris,
and particle separators, placed beneath the surface of an impervious area
such as a parking lot, separate oils from runoff and allow sediment and
debris to settle. Structural devices such as these require regular
maintenance to function properly and remain effective.  Nonstructural BMPs
are primarily designed to minimize the

contaminants that enter storm water. These nonstructural BMPs include 
?good housekeeping? practices by the local government, such as oil
collection and recycling, spill response, household and hazardous

waste collection, pesticide controls, flood control management, and street
sweeping; 29 Report to Congress on the Phase I Storm Water Regulations, U.
S. Environmental Protection Agency, February 2000. This report includes
information on the program for local governments, industries, and
construction sites.

 public education programs, such as storm- drain stenciling, to remind the
public that trash, motor oil, and other pollutants thrown into storm drains
end up in nearby receiving waters; 30

 new ordinances to control pollution sources, such as prohibiting the
disposal of lawn clippings in storm drains and requiring pet owners to clean
up after their pets; 31  requirements that developers comply with storm
water regulations

and incorporate erosion and sediment controls at all new development sites;
 requirements that runoff from properties owned or activities

sponsored by the municipality be properly controlled; and  efforts to
identify and eliminate illicit connections and illegal

discharges to the storm sewer systems, such as those from pipes carrying
sewage.

We found that the NPDES Program?s requirements allowed local governments to
tailor their storm water management efforts to prioritize local concerns,
such as a particular type of contaminant, a particular climatic condition,
or a particular body of water. Some cities also

developed specialized BMPs to address these concerns. The following
information highlights specific storm water- related concerns in the five
cities we visited and the specialized BMPs these municipalities have
developed to address these particular concerns. (See apps. I to V for
additional information on these cities? storm water management programs.)

30 Other public education programs we observed included in- school education
programs, partnerships with grassroots organizations concerned with water
quality issues, and the identification of commercial businesses and
industries to educate owners on methods to control storm water runoff.

31 According to Worcester, Massachusetts? April 2000 City of Worcester DPW
Stormwater Management Program Annual Report, the city has proposed
ordinances that prohibit the disposal of lawn clippings and other yard waste
in catch basins and that require pet owners to clean up after their pets. As
of April 2001, neither ordinance had been implemented.

In Baltimore, Maryland, excessive levels of nutrients, particularly
phosphorus and nitrogen, are among the city?s major water quality concerns
because of the city?s participation in the Chesapeake Bay Program. Baltimore
City agreed to assist the state in reaching the Chesapeake Bay Program?s
goal to reduce nutrients discharged to the bay by 40 percent by the year
2000. According to a Chesapeake Bay Program Office representative, 32 as of
March 2001, the program has not met this goal but expects to reach it within
the next several years. In Boston, Massachusetts, the Boston Water and
Sewer Commission,

which holds the permit for Boston?s storm sewer system, is concerned about
runoff from roadways, especially runoff containing salt and sand used in the
winter months and dissolved metals (copper and zinc) from automobiles. In
September 2000, the commission began a 3- year program to develop and
implement a citywide catch- basin inspection, cleaning, and preventive
maintenance program. The program will also include the development of a
database and map that can be linked to the commission?s Geographic
Information System.  Los Angeles County, California, is responding to a
TMDL for trash in

the Los Angeles River Watershed that will require the county, over a 10year
period, to eliminate trash in runoff. The county is testing a variety of
devices that remove trash from runoff and specialized catch- basin devices
that are designed to prevent trash from ever reaching the storm

sewers.  Milwaukee, Wisconsin, changed its monitoring and public education

activities in its recent permit to test the effectiveness of a BMP targeting
public education efforts to a specific community. The new permit also
requires a monitoring program aimed at the community, its associated

watershed, and city employees who work in the area.  Worcester,
Massachusetts, had a significant problem with illicit

connections to its storm sewers and with flow in these sewers during dry
weather. Worcester?s Department of Public Works (DPW) screened 71 of its
storm water outfalls and determined that 32 of them had drainage areas that
carried both sanitary sewage and storm drainage in separate conduits through
common manholes. DPW has retrofitted over 65 percent of the manholes to
prevent sewage from mixing with storm

water. 32 The Chesapeake Bay Program Office, U. S. EPA Region III, was
founded in 1983 with the formation of the Chesapeake Bay Program. The
program is a voluntary regional partnership that leads and directs
restoration of the Chesapeake Bay. Members of the Chesapeake Bay Program
include Maryland, Pennsylvania, Virginia, the District of Columbia, the
Chesapeake Bay Commission (a tristate legislative body), EPA, and
participating citizen advisory groups .

Third, local governments participating in the Phase I program are required
to report annually to EPA or the state regulatory agency on their storm
water programs. These reports are to include a status report on the program;
a summary of data, including monitoring results collected during the
reporting year; information on annual expenditures on the program and

a budget for the coming year; and a description of any water quality
improvements or degradation.

Information on the Costs of Good information about the cost of implementing
federal storm water Addressing Storm Water requirements is limited. EPA
conducted a survey to estimate the nation?s

Runoff Is Limited future water infrastructure needs over a 20- year period-
from 1996 to 2016.

In its 1996 report, 33 EPA estimated that states would require over $50
billion to meet their current (as of 1996) water infrastructure needs. The
estimate consists of storm water management needs (at $7.4 billion) and CSO
needs (at $44.7 billion). 34 EPA noted, however, that estimated storm water
management needs are likely too low and could increase following an

analysis of data collected to prepare the agency?s 2000 clean water needs
survey- to be released in 2002. According to EPA, many cities have
implemented the Phase I program since EPA reported to the Congress in 1996,
and municipalities should now be better able to provide documented cost
data. As a result, EPA will need to rely less on modeled storm water needs
than it did in the 1996 needs survey. EPA did not project the costs

and benefits of the program when it was initiated; therefore, no initial
cost estimates are available. When EPA promulgated the Phase I program
regulations in 1990, the agency decided that the storm water program did not
meet the requirements for preparing a benefit/ cost analysis.

33 1996 Clean Water Needs Survey Report to Congress, U. S. Environmental
Protection Agency (Sept. 1997). EPA?s estimate represents the estimated
capital costs for water quality projects eligible for state revolving fund
support. 34 EPA also estimates that $81.9 billion of its 20- year water
infrastructure needs cost can be attributed to sanitary sewer overflows.
These overflows may occur when rainwater or snowmelt leaks into sanitary
sewage pipes, exceeding the pipes? capacity and causing them to overflow.
This overflow can release untreated sewage from municipal sanitary sewer
systems into streams, basements, and streets.

The costs to local governments of complying with the Phase I program have
generally been portrayed as high. However, because of inconsistencies in
cost accounting and reporting practices, we could not determine the cost of
the program to several of the cities we visited. Although municipalities are
required to provide information on the expenditures that they anticipate
will be needed to implement their storm

water management programs for each fiscal year covered by the permit, EPA
has not issued any cost reporting guidelines. Consequently, while the
reported fiscal year 1999 total cost to manage and treat storm water runoff
across the five municipalities in our review ranged from less than $1
million (Milwaukee) to $135 million (Los Angeles County), 35 these numbers
are not comparable because the municipalities did not have consistent cost
accounting and reporting practices and did not fully express storm water

management costs. 36 For example, some cities reported only the costs of
activities that were funded by the city department that held the permit.
Significant activities funded by other city departments were not reported,
even if they were important components of the storm water program. Officials
in the Milwaukee Department of Infrastructure Services and the Boston Water
and Sewer Commission told us that other city departments perform and fund
activities such as street sweeping and flood control. The costs of these
activities are not reported as storm water program costs because the
activities serve other purposes besides preventing storm water

pollution. In addition, according to some city officials, these activities
were in place before the permit was issued and, therefore, cannot be
characterized solely as storm water costs. The cost of street sweeping can
be significant- for fiscal year 1999, Baltimore City and Worcester, which
did include streetsweeping costs in their storm water program?s cost
estimate, stated that their street- sweeping expenses totaled about $9.5
million and $1. 2 million,

respectively. Similarly, Milwaukee did not report the cost of a significant
project related to storm water runoff because it was mostly funded by the
state of Wisconsin.

35 Los Angeles County?s cost was projected by the municipal permit holder
and represents the cost of the 85 cities covered by the permit. 36 We were
unable to obtain comprehensive information on the total cost to the Boston
Water and Sewer Commission of managing storm water, so their fiscal year
1999 costs could not be included in this range.

An EPA official told us that the agency had not yet made a national effort
to analyze the information that Phase I permittees submitted on the costs of
their storm water programs. This official cited the inconsistent formats of
the annual reports as a reason that the information was not readily

available at the national level and also indicated that adequate staff are
not available to analyze the data. In addition, other EPA officials informed
us that the Office of Wastewater Management must divide its resources among

a number of issues that will challenge the agency?s water program over the
next decade. Several officials in the cities we visited said that their
annual costs are likely to increase. A number of factors could affect the
costs. For example, a Baltimore City official explained that the
anticipated, future program costs depend on several factors, including (1)
requirements in watershedmanagement plans currently being developed, (2)
pollution- reduction goals the city will be required to achieve, (3)
requirements of the state regulatory agency in future permits, and (4)
requirements the city may have to meet if TMDLs or numeric effluent limits
are incorporated into NPDES storm

water permits. Other city officials also expressed concern about the extent
to which TMDLs could affect their future costs. These city officials are
concerned that when and if TMDLs are established, their future storm water
permits may require that storm water runoff meet specific water quality
standards. For example, Los Angeles County?s trash TMDL could potentially
drive the county?s storm water management costs upward, and the county
expects additional TMDLs to be imposed. On the other hand,

Worcester officials estimated that their future storm water costs would be
about the same as they were at the time of our review- about $4.5 million
per year. In a separate analysis, EPA estimated in 1999 that it will cost
Phase II municipalities about $848 million to $981 million per year (in 1998
dollars) to manage storm water runoff. Because Phase II permits have not
been

issued as of May 2001, we did not gather any cost information on them from
these cities. Funding for Managing The five cities we visited had not
generally obtained federal funds for their Storm Water Runoff Is

storm water management efforts. They used local sources, including Available
From Local and general revenues, bonds, revenue from specifically created
storm water

Federal Sources utilities, state grants, and inspection and permit fees.

While several sections of the Clean Water Act provide funding that can be
used for municipal storm water control, relatively few federal funds have
been directed to these types of projects. The most significant source of
funds is the state revolving loan funds administered by states. 37 These
revolving loan funds provide loans for eligible storm water control
projects. In some cases, nonpoint source projects may also qualify for
funding when storm water permits are not required or issued. However,
municipal storm water management is generally a low priority in these

programs. Specifically, in the year 2000, revolving fund loans were made in
the ?storm sewers? category in the amount of $38.76 million for 44 different
projects. These funds represented less than 1 percent of the amounts loaned
from these revolving funds that year. Activities eligible for revolving fund
loans include constructing BMPs to control runoff, but support for ongoing
operations and maintenance is not eligible. Revolving fund loans can also be
used for eligible CSO control projects. In 2000, Clean Water State Revolving
Fund Program loans were made in the ?CSO Correction? category of a national
EPA database in the amount of $411.3

million for 69 different projects and could have been used for CSO or
sanitary sewer overflow projects. This amount represented about 9 percent of
the funds loaned in 2000. According to EPA, the agency also issues grants to
universities and other research institutions to help implement the storm
water program. Some of these grants provide training and guidance to Phase I
permittees on

watershed protection and the proper selection of BMPs. Other sources of
funding may be available to local governments beginning in 2002. In December
2000, the Congress authorized programs for fiscal years 2002 through 2004 to
provide grants to local governments for (1) pilot projects for managing
municipal CSOs, sanitary sewer overflows, and storm water discharges on a
watershed basis and for testing BMPs and (2) controlling pollutants from
MS4s to demonstrate and determine costeffective,

innovative technologies for reducing pollutants from storm water discharge.
EPA?s proposed budget does not request funds for these programs. In
addition, the Congress authorized programs for fiscal years

2002 and 2003 to provide grants to local governments for planning,
designing, and constructing treatment works to intercept, transport, 37
Under the Clean Water State Revolving Fund Program, the federal government
provides grants to capitalize states? funds. States provide loans to local
governments for wastewater projects.

control, or treat municipal CSOs and sanitary sewer overflows. EPA?s
proposed budget requested $450 million for this program.

EPA, States, and Local EPA, state, and municipal officials generally believe
that the NPDES Storm Governments Believe the Water Program will improve
water quality. These officials believe that the NPDES Storm Water program
will result in more bodies of water that meet water quality Program Is
Effective, but It

standards, improved aesthetic conditions, reduced risk from bacterial
contamination, and improvements attributable to the discovery and Has Not
Been Evaluated management of pollutants in storm water that otherwise would
have gone unnoticed. EPA attempted to put a dollar value on these benefits
in its benefit/ cost analysis prepared for the Phase II storm water
regulations,

estimating that such benefits could range from $672 million to $1. 1 billion
per year (in 1998 dollars). 38 However, little information is currently
available on the benefits of the storm water program or its general
effectiveness. There is no doubt that it will take time for the results of
the Phase I program to be demonstrated. As EPA notes in its February 2000
report to the Congress, pollution control efforts under water quality
management programs produce long- term

changes, and the agency expects water quality improvements attributable to
the Phase I program to become evident in the future, as the program matures.
In this report, EPA concluded that the program has improved

storm water management at the local level, improved water quality, and
decreased pollutant loads in storm water. However, EPA relied on a survey of
only nine Phase I cities in making these conclusions and, therefore, also

reported that the agency could not provide national estimates on water
quality protection and improvements generated by Phase I of the program. To
evaluate the entire program, EPA would have to establish goals for the

program that are based on its mission; obtain information about the
program?s results; compare the results with the goals; and make changes to
the program, if warranted, to get closer to achieving the agency?s goals.
EPA and the states also have not taken advantage of information that is

available to evaluate the program. Each city we visited was regularly
monitoring its storm water to establish baseline information on pollutant
levels and was reporting this information to EPA or the regulatory state

agency each year. Although cities with Phase I permits are required to
report on their storm water monitoring results and changes in water

38 Using another method, EPA estimated the benefits at $1. 6 billion per
year.

quality, overall, EPA and the states have not successfully developed
measurable goals for the program or demonstrated its effectiveness through
the review of municipal reports. An EPA official said that some states had
requested funding to analyze program data because they did not have the
resources to do so, and that EPA had provided the funding in a few cases.
EPA also has not established any guidelines for how these data should be
reported. Therefore, the reports may be as variable as the cost information
we obtained in our five site visits.

EPA has not yet taken any of these data- analysis steps because, according
to EPA officials, other program challenges within the Office of Wastewater
Management compete with storm water management efforts for priority. For
example, EPA officials stressed that available resources within the office
must address other significant wet- weather pollution problems, such as CSOs
and sanitary sewer overflows, and nonpoint source pollution

problems, such as agricultural practices, forestry, and mining. One agency
official noted that the highest priority is addressing needs that the agency
and local governments have identified for improving wastewater
infrastructure, such as sewage treatment facilities. The program also has
relatively few staff assigned- about five in the headquarters office and
about 10 in the regional offices- for the municipal, industrial, and

construction portions of the program. In a program plan recently prepared
for the storm water program, EPA estimated that nine to 10 staff would be
needed in EPA headquarters to evaluate the program and implement other

program requirements. EPA officials described two efforts that may be the
first steps in developing better information about the program. First, EPA
intends to issue a grant to the University of Alabama in June 2001 to
evaluate monitoring data submitted by a sample of municipalities with Phase
I permits. This effort will (1) determine the different types of monitoring
being conducted by

Phase I municipalities, (2) assess water quality in and around permitted
municipalities and determine any correlation between program implementation
and impacts on water quality, and (3) recommend approaches for improving the
effectiveness of municipal storm water monitoring programs. EPA expects the
results of this study in 2003. Second, an EPA official stated that the
agency would like to establish a system for analyzing program findings,
incorporating necessary changes that are based on these findings, and
evaluating the program?s effectiveness. The agency plans to implement a
pilot project in 2001 in the agency?s Atlanta Region IV office for analyzing
data reported in annual

reports and developing key indicators for the program. If this project is
successful and resources are available, the project could be expanded.

Conclusions EPA regards urban runoff as a significant threat to water
quality across the nation and considers it to be one of the most significant
reasons that water

quality standards are not being met nationwide. Prompted by the Congress,
EPA has responded with a variety of programs, including the NPDES Storm
Water Program, which requires more than 1, 000 local governments to
implement storm water management programs. Those municipalities that are
currently involved in Phase I of the program have been attempting to reduce
pollutants in storm water runoff for several years. It is time to begin
evaluating these efforts. However, EPA has not

established measurable goals for this program. In addition, the agency has
not attempted to evaluate the effectiveness of this program in reducing
storm water pollution or to determine its cost. The agency attributes this
problem to inconsistent data reporting from permitted municipalities,

insufficient staff resources, and other competing priorities within the
Office of Wastewater Management. Although Phase I municipalities report
monitoring and cost data to EPA or state regulatory agencies annually, these
agencies have not reviewed this information to determine whether it can be
of use in determining the program?s overall effectiveness or cost.

Our analysis shows that the reported cost information will be difficult to
analyze unless EPA and its state partners set guidelines designed to elicit
more standardized reporting. Better data on costs and program effectiveness
are needed- especially in light of the Phase II program that will involve
thousands more municipalities in 2003. EPA?s planned research grant to the
University of Alabama and its pilot project in the agency?s Region IV to
analyze data from annual reports and develop baseline indicators is a step
in the right direction and could point the way for a more

comprehensive approach. Recommendation To determine the extent to which
activities undertaken through the NPDES

Storm Water Program are reducing pollutants in urban runoff and improving
water quality, and the costs of this program to local governments, we
recommend that the Administrator, EPA, direct the Assistant Administrator
for the Office of Water to  establish measurable goals for the program;

 establish guidelines for obtaining consistent and reliable data from local
governments with Phase I permits, including data on the effects of the
program and the costs to these governments;  review the data submitted by
these permittees to determine whether

program goals are being met and to identify the costs of the program; and 
assess whether the agency has allocated sufficient resources to oversee and
monitor the program.

Agency Comments and We provided a draft of this report to EPA and DOT for
their review and

Our Evaluation comment. EPA generally agreed with the report and with the
recommendation, although it did not explicitly comment on all parts of it.

(EPA?s comments appear in app. VI.) In response to our recommendation that
EPA set measurable goals for the storm water program, EPA stated that under
the second phase of the program, local governments will establish their own
goals. Although this is an important activity, EPA will have difficulty
evaluating the program?s effectiveness at a national level without setting
goals that reflect the program?s mission of improving water

quality. The agency (1) agreed that it should establish guidelines for
obtaining consistent and reliable data from local governments about their
programs and (2) plans to award grants to two universities for reviews of
monitoring data reported by local governments. EPA did not comment on
whether local governments should report on the costs of their programs. EPA
also agreed that it and its state partners should review data reported by
local governments to determine whether the program?s goals are being met. In
April 2001, EPA officials told us that the agency planned to undertake a
project in the Region IV (Atlanta) office to evaluate the

methods local governments are using to control storm water. EPA?s letter
indicates that the agency now plans to implement this project in three
regional offices and 10 states. EPA did not comment on the part of our
recommendation that the agency review the level of resources devoted to
overseeing and monitoring the program. EPA also provided technical

comments that we incorporated where appropriate. DOT generally agreed with
the draft report and provided technical comments that we incorporated where
appropriate. In particular, DOT suggested that we revise several references
in the draft report to paved surface area and its relationship to increases
in urban runoff, to emphasize that impervious surfaces, of which paved
surfaces are a significant subset, cause increases in runoff. We revised the
language in these places.

As agreed with your offices, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 7 days after
the date of this report. At that time, we will send copies of this report to
the Administrator, Environmental Protection Agency, and the Secretary of
Transportation. We will make copies available to others on request. If you

or your staff have any questions about this report, please call me at (202)
512- 2834. Key contributors to this report are listed in appendix VII. Peter
F. Guerrero

Director, Physical Infrastructure Issues

Appendi xes The Storm Water Program in Baltimore City,

Appendi x I

Maryland Baltimore City?s municipal separate storm sewer system (MS4) is
regulated by the Maryland Department of Environment (MDE) and, according to
a city official, services the entire city. The city is currently
implementing its second, 5- year National Pollutant Discharge Elimination
System (NPDES) permit, issued on February 8, 1999. Before obtaining the
first NPDES

storm water permit in 1993, Baltimore City addressed the adverse affects of
storm water runoff by implementing Maryland?s Storm Water Management Program
and Erosion and Sediment Control Program. According to the 2000 census,
Baltimore City?s population is about 651, 000. Urban Runoff

Baltimore City?s urban runoff discharges to four major areas- Gwynns
Problems in Baltimore Falls, Jones Falls, Herring Run, and the Patapsco
River- and then ultimately to the Chesapeake Bay. In 1990, the Environmental
Protection City

Agency?s (EPA) 319( a) report 1 implicated urban runoff as the main source
of pollution in these waters. Moreover, Baltimore City was one of the areas
studied in EPA?s Nationwide Urban Runoff Program in the 1980s. This study
reported that urban runoff contributed over 60 percent of the total
nitrogen, phosphorus, and organic carbon; over 70 percent of the chemical
oxygen demand; and over 80 percent of the total suspended solids, lead,

and zinc in local water bodies. An MDE official told us that nutrients,
zinc, and suspended solids are among the constituents most commonly found in
urban runoff, but the quantitative contribution to water quality impairment
in the state?s waters was not known. Also, in 1996, the Chesapeake Executive
Council designated the Baltimore Harbor as one of three toxic regions of
concern in the Chesapeake Bay. The harbor suffers from sediment contaminated
by banned substances (such as the termiticide chlordane) and contaminants
currently being released (such as metals and organics). Furthermore,

according to the Chesapeake Bay Program Office, data collected from Phase I
permittees indicate that storm water runoff can be a significant source of
metals and organics in the harbor. A Baltimore City official told us that
some portions of Maryland?s waters are impaired because of unacceptable
levels of nutrients, metals,

1 Section 319( a) of the Clean Water Act requires, among other things, that
states identify and report to EPA the navigable waters that cannot
reasonably be expected to maintain water quality standards (e. g.,
established water body uses) without additional action to control nonpoint
source pollution.

suspended sediments, and chlordane. Moreover, this official noted that the
state does not consider data that municipalities collect under their NPDES
storm water permits during the 303( d) listing process. Therefore, he
believes that streams in Maryland are much more impaired than indicated by
the listing process. Baltimore City?s Use of Like other NPDES storm water
permit holders, Baltimore City uses a Best Management

variety of best management practices (BMP) to reduce the amount of
pollutants in runoff to the maximum extent practicable. These BMPs Practices

include detention ponds, shallow marshes (which use the biological and
naturally occurring chemical processes in water and plants to remove
pollutants), sand filter devices, public education programs, and the
identification of illicit discharges to the MS4 system. Furthermore,

Baltimore City participates in Maryland?s effort to reduce nutrient levels
in the Chesapeake Bay. Refer to the section of this report describing local
government efforts to manage storm water for details concerning this
nutrient- reduction goal. One other BMP includes the following:

 Baltimore City has incorporated the 2000 Maryland Storm Water Design
Manual?s management policies, principles, methods, and practices into its
current NPDES storm water discharge permit. The purpose of the design manual
is to (1) protect the waters of the state from the adverse effects of urban
storm water runoff; (2) provide design guidance on the most effective
structural and nonstructural BMPs for

development sites; and (3) improve the quality of BMPs that are constructed
in the state, with particular attention to their performance, longevity,
safety, ease of maintenance, community acceptance, and environmental
benefit. Costs Associated With

We were not able to obtain comprehensive information on the total cost to
Managing Storm Water Baltimore City of managing storm water. Therefore, we
do not present that information here.

Funding Sources Baltimore City funds its storm water management control
efforts with city water and sewer user fees and with state funds.

The Storm Water Program in Boston,

Appendi x II

Massachusetts The Boston Water and Sewer Commission received a NPDES storm
water permit in October 1999. The commission is a separate entity from the
city of Boston and, therefore, does not manage some storm water controls
that are common in Phase I permits, such as street sweeping, winter deicing,
and many of the urban runoff controls required for new developments. Boston
has combined sewer systems as well as separate sanitary sewers and storm
drains. The commission maintains 206 storm water outfalls and

serves approximately 33 percent of the city through its separate MS4 system.
In addition to the resident population of about 589,000, this system also
almost daily serves 340,000 commuting workers; 70, 000 shoppers, tourists,
and business people; and 75,000 commuting students. The commission?s
sanitary and combined flows are transported to the

Massachusetts Water Resources Authority at Deer Island. The commission is
also the permittee for EPA?s Combined Sewer Overflow Program.

Urban Runoff The commission considers the identification and elimination of
illegal Problems in Boston

sanitary sewer connections as the most effective means of improving water
quality and protecting public health . It is also concerned with the washoff
of animal wastes from residential and open land, which is another major
contributor to the impairment of water quality because it can cause an
increase in coliform levels in the storm water discharges to the receiving
waters.

The commission has contracted for various studies to determine the impact of
storm water runoff. The following two studies identified sources of
bacterial contamination and characterized the quality of storm water
discharged from different types of land uses. The studies included metering
storm water flows, collecting and analyzing the storm water and receiving
water quality samples, and identifying and remediating illegal sewer
connections. Observations from the studies include the following:

 A 1996 study determined that pet waste, rather than sanitary sewage, was a
key contributor of bacteria to the storm drain system that had possibly led
to beach closings in the area.  A 1998 study identified several illegal
connections to the storm drain

system. Furthermore, the study showed that deicing and sanding efforts
resulted in levels of sodium, chloride, total dissolved solids, and cyanide
that exceeded EPA?s acute (high dose) toxicity levels.

Boston?s Use of Best To meet the NPDES permit?s requirements, the
commission, like other Management Practices

permittees, continued BMPs, such as identifying illegal connections, and
implemented new BMPs aimed at preventing the discharge of pollutants to
storm drains and receiving waters. Refer to the section of this report

describing local government efforts to manage storm water for details
describing the commission?s citywide catch- basin inspection cleaning and
preventative maintenance program. Other efforts include the following:

 The commission has placed particle separators, which remove oil, grease,
and sediments from storm water flows, throughout the city. The commission
requires particle separators to be installed by developers on all newly
constructed storm drains that serve outdoor parking areas. Fuel- dispensing
areas not covered by a canopy or other type of roof enclosure must also have
a particle separator.

 The commission requires developers to consider on- site retention of storm
water for all new projects, wherever feasible. On- site retention aids in
controlling the rate, volume, and quality of storm water

discharged to the commission?s storm drainage system. Costs Associated With

We were not able to obtain comprehensive information on the total cost to
Managing Storm Water the commission of managing storm water because the
commission does not separate the cost of its storm water program from the
cost of its sewer operations. Therefore, we do not present that information
here.

Funding Sources The commission funds its storm water management control
efforts primarily with city water and sewer user fees and bond proceeds.

The Storm Water Program in Los Angeles

Appendi x II I County, California Under the NPDES Storm Water Program, the
Los Angeles Regional Water Quality Control Board issues 5- year permits to
Los Angeles County for its municipal storm water program. The Los Angeles
County permit, issued in July 1996, is the county?s second storm water
permit. This permit includes Los Angeles County as the principal permittee
and 85 cities as permittees. According to the 2000 census, Los Angeles
County?s population is about 9.5 million.

Urban Runoff The effects of urban runoff on the ocean are of particular
concern in Problems in Los southern California. Contaminated sediments,
impaired natural resources, and potential human illness could threaten the
county?s tourism economy, Angeles County

estimated to be about $2 billion a year. The following three studies have
shown that urban runoff can pose health risks to swimmers near storm drains
and contribute toxic metals to receiving water sediments:  The Santa Monica
Bay Restoration Project conducted a study to assess the possible adverse
health effects of swimming in waters contaminated by urban runoff. 1 This
study revealed that there is an increased risk of illness associated with
swimming near flowing storm drain outlets and

an increased risk of illness associated with swimming in areas with high
concentrations of bacteria indicators. Furthermore, illnesses were reported
more frequently on days when the samples were positive for enteric viruses.
Refer to the section of this report describing the effects

of runoff on aquatic life and human health for more details.  he Southern
California Coastal Water Research Project coordinated a

study that assessed microbiological water quality and found that the
majority of shoreline waters exceeded water quality standards during wet-
weather conditions. Furthermore, the ocean waters near storm

water outlets demonstrated the worst water quality regardless of the
weather. 2  The Southern California Coastal Water Research Project also
compared the runoff from an urban area and a nonurban area in the Santa
Monica

1 R. W. Haile and others, ?The Health Effects of Swimming in Ocean Water
Contaminated by Storm Drain Runoff,? Epidemiology, July 1999, Vol. 10, No.
4. 2 Southern California Coastal Water Research Project, Southern California
Bight 1998 Regional Monitoring Program, Volume 3: Storm Event Shoreline
Microbiology, 2000.

Bay Watershed. 3 The results of the study indicated that storm water plumes
extended up to several miles offshore and persisted for a few days.
Furthermore, the runoff from the urban area proved to be toxic to sea urchin
fertilization, and dissolved zinc and copper were determined to be
contributors to the toxicity. The study also found that in urban areas,
sediments offshore generally had higher concentrations of contaminants such
as lead and zinc.

Los Angeles County?s As in the other sites we visited, the county is
managing its runoff through

Use of Best the use of conventional BMPs. These BMPs include the elimination
of illicit connections and discharges to the storm sewer system,
construction

Management Practices control measures, routine inspections, staff training,
pollution prevention

plans for public vehicle maintenance and material storage facilities,
sweeping and cleaning public parking facilities, street sweeping, catchbasin
cleaning, and public education.

The Los Angeles Regional Water Quality Control Board recently adopted a
Total Maximum Daily Load (TMDL) Program to reduce trash loads to the Los
Angeles River. As a result, the county is exploring a number of trash

reduction BMPs, which are discussed in the section of this report describing
local government efforts to manage storm water. Costs Associated With

Table 3 indicates that the county and the other permittees have allocated
significant funding for storm water management activities over the years.
Managing Storm Water

For example, for fiscal year 1999, 4 projected funding for storm water
management activities for the county and the other permittees amounted to
over $134 million. 5 The largest projections for both went toward public
agency activities. For example, during fiscal year 1999, the principal
permittee and the permittees together projected almost 67 percent of storm
water management funds to public agency activities. The activities in this 3
Southern California Coastal Water Research Project, Study of the Impact of
Stormwater Discharge on Santa Monica Bay Executive Summary, Nov. 1, 1999.

4 The county?s fiscal year begins July 1 and ends June 30. 5 According to an
official with the Los Angeles Regional Water Quality Control Board, this
figure may also include activities that are outside the scope of the permit
.

program include staff training, inspections of construction projects, street
sweeping, and catch- basin cleaning.

Table 3: Summary of Fiscal Resources Projected for Los Angeles County and
Its Co- permittees, Fiscal Years 1997- 99

(Dollars in thousands) a

Fiscal year 1997 Fiscal year 1998 Fiscal year 1999 Activity County Others b
County Others b County Others

Program $2, 225 $6, 195 $1, 856 $4,874 $1, 466 $6, 187

Management Illicit 1, 620 3, 515 1,017 3, 075 764 2, 901 Connection, Illicit
Discharge Program

Development 784 6, 208 1,300 3, 769 1, 452 5, 743

planning and construction

Public agency 38, 544 40, 915 40, 256 31,992 43, 316 46, 657

activities Public 2, 840 5, 538 4,360 3, 856 4, 629 6, 177 information and
participation

Monitoring 2, 018 619 1,768 729 1, 598 737 Other 187 13, 991 490 8, 656
1,318 11, 834

Total $48, 218 $76,981 $51, 048 $56,950 $54,543 $80, 237

a Totals may not add up because of rounding. b Does not include 17
permittees for fiscal year 1998 and 13 permittees for fiscal year 1997 for
the following reasons: The permittee operated on a different budget cycle,
the final document was not available at the time of the annual report, or
the information submitted by the permittee was not complete.

Source: GAO?s analysis of cost data provided by the Los Angeles County
Department of Public Works.

As shown in table 3, the county maintains primary responsibility for
monitoring activities, having projected over $2 million for storm water
monitoring activities in fiscal year 1997, almost $2 million in fiscal year
1998, and over $1.5 million in fiscal year 1999. Conversely, the permittees?

projected funding levels for monitoring activities amounted to only $619,000
in fiscal year 1997, $729,000 in fiscal year 1998, and $737,000 in fiscal
year 1999. According to an official with the Los Angeles Regional Water
Quality Control Board, the County has consistently maintained

primary responsibility for monitoring activities required under the permit.

Funding Sources The primary source of funds for the county?s storm water
program is flood control assessments collected throughout the district.
Although the county has not applied for any state revolving funds, it has
applied for and received

approval for federal funds through the Transportation Equity Act for the 21
st Century (TEA- 21) for a pilot study of an engineering device that would
remove trash from storm water. Additionally, the county has received partial
funding through Proposition A of the Safe Neighborhood Parks of 1992 and
1996 6 for two Vortex Separation Systems- a Continuous

Deflective Separation unit and a Stormceptor unit. Additionally, the county
received grant money from the Metropolitan Transit Authority, which
partially funded catch- basin screens, a Continuous Deflective Separation
unit, and 120 catch- basin inserts. 7

6 The Los Angeles County Regional Park and Open Spaces District (a district
within the Parks Department) received this funding from Proposition A and,
in turn, made grants to the Los Angeles County Department of Public Works
for the BMP devices.

7 The Metropolitan Transit Authority receives TEA- 21 funds from the
California Department of Transportation.

The Storm Water Program in Milwaukee,

Appendi x I V

Wisconsin The Wisconsin Department of Natural Resources (WDNR) has the
authority to regulate the discharge of storm water from municipalities,
construction sites, and industries under Natural Resources Code 216 . This
rule identifies Wisconsin municipalities that are required to obtain a storm
water discharge permit under the Wisconsin Pollutant Discharge Elimination
System (WPDES). Milwaukee completed its application process in 1994, and
WDNR issued a WPDES permit to the city in October 1994. This was the first
municipal storm water permit issued to a

municipality in EPA?s Region 5 covering the midwest. In July 2000, WDNR
reissued Milwaukee?s storm water permit. According to the 2000 census,
Milwaukee?s population is about 597,000.

Urban Runoff Milwaukee has a combined sewer system as well as a separate
sanitary

Problems in Milwaukee sewer system. The Milwaukee Metropolitan Sewerage
District implemented a rehabilitation program that cost over $2 billion to
reduce the number of combined sewer overflow (CSO) events each year. The

rehabilitation program involved the construction of deep tunnels to store
untreated wastewater and rainwater for later treatment at a wastewater
treatment plant. Since 1996, the deep tunnels have significantly reduced the
number of overflow events from an average of 50 to 60 per year before

the construction to an average of two per year afterwards . Urban runoff has
been identified as a leading source of pollution to the Milwaukee River
basin?s streams, lakes, and wetlands and the Milwaukee River estuary. To
address pollution from urban runoff, WDNR issues storm

water permits to municipalities with MS4s serving areas with populations of
100, 000 or more, municipalities in Great Lakes ?areas of concern? where
water quality has been identified as a serious problem, municipalities with
populations of 50,000 or more that are located in priority watershed

planning areas, and designated municipalities that contribute to the
violation of a water- quality standard or are significant contributors of
pollutants to state waters.

Milwaukee?s Use of In addition to BMPs such as the elimination of illicit
connections and

Best Management discharges to the storm sewer system, the reduction of
pollutants in storm

water runoff from construction sites, public education, catch- basin
Practices cleaning, street sweeping, and the use of detention basins,
Milwaukee has explored the use of innovative BMPs. Refer to the section of
this report describing local government efforts to manage storm water for
more

details about an educational campaign directed at a specific watershed.
Additional BMPs include the following:

 An innovative storm water control device was installed in a parking lot at
a heavily used municipal public works yard that was found to discharge
significant amounts of storm water pollutants. Termed the Multi- Chambered
Treatment Tank (MCTT), this device is suitable for areas with limited space,
cleans up polluted runoff close to its source, removes pollutants that are
not susceptible to other treatment methods, and is hidden from view. The
MCTT consists of a catch basin, a settling chamber, and a filter. Although
the results of the monitoring studies have revealed that the device has a
positive effect on water quality, officials with the Department of Public
Works explained that it is costprohibitive

and suitable only for sites with limited space.  The permittee has also
been working with WDNR, the Department of Transportation, the U. S.
Geological Survey, and a neighborhood

association in a joint effort to develop a storm water monitoring assessment
program consisting of two innovative storm water treatment devices. One
device removes grit, contaminated sediments, heavy metals, and oily floating
pollutants from surface runoff. The other device removes a broad range of
pollutants from runoff, such as bacteria, heavy metals, nutrients, petroleum
hydrocarbons, and

suspended solids. The devices are to be installed along a new reach of the
Milwaukee Riverwalk through the third ward of Milwaukee. Costs Associated
With

Reliable data on the total cost to manage storm water in Milwaukee were
Managing Storm Water not available and cannot be presented here because
certain activities are not reported as program costs in the city?s annual
report. These activities include street sweeping; leaf collection; catch-
basin and inlet cleaning; maintenance of public boulevards, parks, and
public green spaces; and the recycling of waste oil and antifreeze.
Therefore, the program costs reflected in the annual report do not take into
account many of the nonstructural BMPs employed by the city nor do the
totals include activities funded through grants. The storm water management
activities that were included in the city?s 2000 budget request were
estimated to cost $460,000. Funding Sources Milwaukee?s storm water program
is primarily funded through the city?s sewer maintenance fund. Unlike the
general revenue account, which is

based on property taxes, the sewer maintenance fund is based on water
consumption. The city has also received supplemental funding from the
Wisconsin Nonpoint Source Water Pollution Abatement Program in the form of
WDNR grants. The city has received over $1 million since 1991 for a wide
variety of storm water management activities.

The Storm Water Program in Worcester, Appendi x V

Massachusetts Worcester?s Department of Public Works (DPW) received a NPDES
permit on November 1, 1998. The Sewer Operations Division, within the DPW,
is directly responsible for operating and maintaining the city?s separate
storm sewer system, along with the sanitary and combined sewer system. Since
1993, the Sewer Operations Division has had a full- time storm water
coordinator, reflecting Worcester?s increased emphasis on meeting NPDES
program requirements. Worcester has a population of about 173 , 000. Its
water system covers an extensive area, including 371 miles of sanitary
sewers, 340 miles of storm sewers, 56 miles of combined sewers, 27,000

manholes, over 14,000 catch basins, and 263 outfalls. Worcester?s separate
storm drain systems consist of 93 main drainage areas covering approximately
6,680 acres.

Urban Runoff The constituents that are typically found in urban runoff in
Worcester are

Problems in Worcester the same as those normally found in urban runoff in
older cities. Because virtually all of the paved surfaces in the Worcester
area are devoted to the

city?s transportation infrastructure, the constituents generated include
automobile- related petroleum products, such as total petroleum
hydrocarbons, oil and grease, along with total suspended solids. Also,
coliform, silt, and sediment have been identified in the city?s runoff.

Worcester?s Use of Best Like other permittees, the DPW has implemented BMPs
under the major

Management Practices areas of education outreach, pollution prevention and
source controls,

storm- drainage system maintenance, regulatory efforts, and storm- drainage
system infrastructure. Additionally, to reduce storm water pollution , the
DPW has retrofitted a number of twin manholes in the city as discussed

below. BMPs that are specific to Worcester include the following:  The DPW
implemented a demonstration project to determine the effectiveness of an oil
and grit separator installed on a street drain. The drain is a major surface
sewer main that services approximately 226 acres of heavily urbanized area
with a typical mix of residential, commercial, and industrial use. The drain
discharges into Lake

Quinsigamond, which is a large lake used for recreational purposes such as
swimming and boating. In its April 2000 annual plan submitted to EPA, the
DPW noted that because of drought conditions, it currently did

not have sufficient sampling data to determine the effectiveness of the
project.

 The DPW has embarked on a comprehensive program to minimize the
possibility that sewage and storm water will be mixed in its twin invert
manholes. Since the program began, the DPW has installed hold- down devices
on over 1, 680 of the approximately 2, 580 twin invert manholes in the city.
The DPW expects to continue the program until all of the

manholes have been retrofitted.  The DPW is also working closely with the
Massachusetts Department of

Environmental Protection in its ongoing tracking efforts to ensure that
industries in Worcester are doing their part to reduce storm water
pollution.  To improve its storm- drainage infrastructure, the city has
established a voluntary plan to reduce the number of unpaved private roads.
The dirt from these roads, especially after rain storms, causes sediment to
build

up in the drainage system. The DPW has developed a plan to pave the streets
at a lower grade than would be necessary to meet the legal requirements for
a public street. Under this plan, residents would not have to pay the
additional betterment taxes that are now required to cover the costs of
sediment removal and less sediment would be

transported in runoff. Costs Associated With Since 1993, the DPW has
allocated significant funding from the water and Managing Storm Water

sewer utility fees it collects for controlling the effects of runoff,
especially through catch- basin cleaning, street sweeping, and correcting
illegal connections. For example, its fiscal year 1993 budget for storm
water programs included about $1.6 million for specific programs and another
$1 million for capital improvement programs, such as inflow/ infiltration
and flood control. The DPW also spent $500,000 to develop and submit its
permit application. Furthermore, as shown in table 4, Worcester made
extensive capital expenditures during fiscal years 1994 through 1999 on
pertinent storm water projects to improve the quality of storm water runoff
emanating from the city?s storm water sewer system.

Table 4: City of Worcester?s Capital Expenditures for Storm Water Management

(Dollars in thousands)

Fiscal year Activity 1994 1995 1996 1997 1998 1999

Sewer construction $0 $500 $500 $300 $300 $300 Infiltration control 0 400
400 100 100 100 Pump station

200 200 200 200 200 200 rehabilitation Sewer rehabilitation 300 750 300 750
750 1, 500

Landfill closeout 150 1, 200 200 500 0 0 Belmont Drainage 0 100 600 100 0 0
project Beaver Brook 0 500 100 100 300 100

Culvert project Surface drain control 40 150 200 200 200 200 Geographic 0 0
0 125 125 125 Information System Other 0 70 10 0 0 0

Total $690 $3,870 $2,510 $2,375 $1,975 $2, 525

Note: The Belmont Drainage project involved enlarging the drain to eliminate
surcharging and siltation and moving the outfall to eliminate stagnation.
The Beaver Brook Culvert project involved repairing the culvert and
conducting a study that included a detailed hydraulic analysis of the
drainage basin.

Source: Worcester Department of Public Works.

Furthermore, during fiscal year 1999, the DPW spent approximately another
$2. 1 million to operate and maintain storm water activities. Key
expenditures included about $1. 2 million for street sweeping, about
$617,000 for catch- basin maintenance, $52,000 for root control, and another
$48,000 for street paving. Also included was $40,000 per year for sampling
five outfalls around the city three times per year as required by the
permit. According to a DPW official, in previous fiscal years, the DPW
funded the

same or similar operation and maintenance activities to help control storm
water runoff. As a result, the costs since 1994 were similar to those for
1999, except for annual adjustments for inflation. Therefore, the annual

operation and maintenance expenditures ranged from about $1.7 million for
1994 to about $2. 1 million for 1999.

According to a DPW official, the department expects to spend from $3 million
to $4. 5 million annually over the next several years on storm water

related activities. The amount of the cost increase will depend on whether
EPA asks the city to increase its spending.

Funding Sources The DPW funds its storm water management controls effort
from the water and sewer user fees it assesses to homes and businesses.

Comments From the Environmental

Appendi x VI Protection Agency

Appendi x VII

GAO Contacts and Staff Acknowledgments GAO Contacts Peter Guerrero (202)
512- 2834 Katherine Siggerud (202) 512- 2834 Staff

In addition to those named above, Jennifer Clayborne, Richard LaMore,
Acknowledgments

Sally Coburn, Elizabeth McNally, Charles Bausell, and Timothy Guinane made
key contributions to this report. (348248) Lett er

Ordering Information The first copy of each GAO report is free. Additional
copies of reports are $2 each. A check or money order should be made out to

the Superintendent of Documents. VISA and MasterCard credit cards are
accepted, also. Orders for 100 or more copies to be mailed to a single
address are discounted 25 percent.

Orders by mail: U. S. General Accounting Office P. O. Box 37050 Washington,
DC 20013

Orders by visiting: Room 1100 700 4th St. NW (corner of 4th and G Sts. NW)
U. S. General Accounting Office Washington, DC

Orders by phone: (202) 512- 6000 fax: (202) 512- 6061 TDD (202) 512- 2537

Each day, GAO issues a list of newly available reports and testimony. To
receive facsimile copies of the daily list or any list from the past 30
days, please call (202) 512- 6000 using a touchtone

phone. A recorded menu will provide information on how to obtain these
lists.

Orders by Internet: For information on how to access GAO reports on the
Internet, send an e- mail message with ?info? in the body to:

info@ www. gao. gov or visit GAO?s World Wide Web home page at: http:// www.
gao. gov

To Report Fraud, Contact one:

Waste, or Abuse in  Web site: http:// www. gao. gov/ fraudnet/ fraudnet.
htm

Federal Programs

 e- mail: fraudnet@ gao. gov  1- 800- 424- 5454 (automated answering
system)

GAO United States General Accounting Office

Page 1 GAO- 01- 679 Water Quality: Urban Runoff Programs

Contents

Page 2 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 3 GAO- 01- 679 Water Quality: Urban Runoff Programs United States
General Accounting Office

Washington, D. C. 20548 Page 3 GAO- 01- 679 Water Quality: Urban Runoff
Programs

Page 4 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 5 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 6 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 7 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 8 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 9 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 10 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 11 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 12 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 13 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 14 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 15 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 16 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 17 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 18 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 19 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 20 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 21 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 22 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 23 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 24 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 25 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 26 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 27 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 28 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 29 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 30 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 31 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 32 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 33 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 34 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 35 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 36 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 37 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 38 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 39 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 40 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix I

Appendix I The Storm Water Program in Baltimore City, Maryland

Page 41 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 42 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix II

Appendix II The Storm Water Program in Boston, Massachusetts

Page 43 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 44 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix III

Appendix III The Storm Water Program in Los Angeles County, California

Page 45 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix III The Storm Water Program in Los Angeles County, California

Page 46 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix III The Storm Water Program in Los Angeles County, California

Page 47 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 48 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix IV

Appendix IV The Storm Water Program in Milwaukee, Wisconsin

Page 49 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix IV The Storm Water Program in Milwaukee, Wisconsin

Page 50 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 51 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix V

Appendix V The Storm Water Program in Worcester, Massachusetts

Page 52 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix V The Storm Water Program in Worcester, Massachusetts

Page 53 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix V The Storm Water Program in Worcester, Massachusetts

Page 54 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 55 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix VI

Appendix VI Comments From the Environmental Protection Agency

Page 56 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix VI Comments From the Environmental Protection Agency

Page 57 GAO- 01- 679 Water Quality: Urban Runoff Programs

Page 58 GAO- 01- 679 Water Quality: Urban Runoff Programs

Appendix VII

United States General Accounting Office Washington, D. C. 20548- 0001

Official Business Penalty for Private Use $300

Address Correction Requested Presorted Standard

Postage & Fees Paid GAO Permit No. GI00
*** End of document. ***