EPA
United States
Environmental Protection
Agency
Office Of Water
(4204)
EPA 833-R-99-001
October 1999
           Report To Congress On The
           Phase II Storm Water Regulations

-------
                            Report to Congress
                                    on the
                   Phase II Storm Water Regulations

                                Table of Contents
SECTION
I.     Introduction
 A.   Purpose of Report to Congress
 B.   Purpose of the Phase II Rule
II.    Impact of Phase II Rule on Local Governments
 A.   Summary of Phase II Rule Requirements
 B.   Impacts of the Municipal Minimum Control Measures on Local Governments
 C.   Impacts of the Soil Erosion Control Provision on Local Governments
III.    Rationale for the One Acre Construction Threshold
IV.    Storm Water Problems in Census Designated Urbanized Areas
V.    Rationale for Using a NPDES Approach

-------
I. INTRODUCTION

A. PURPOSE OF REPORT TO CONGRESS

       EPA provides this Report to Congress in compliance with Section 431 (a) of the
Departments of Veterans Affairs and Housing and Urban Development and Independent
Agencies Appropriations Act of 2000, Pub. L. No. 106-74 (1999) ("Appropriations Act"). The
Appropriations Act directs the Administrator of the Environmental Protection Agency ("EPA")
to submit two reports to the Committee on Environment and Public Works in the Senate and the
Committee on Transportation and Infrastructure in the House of Representatives. The first of the
reports is to address several issues related to EPA rulemaking to implement Section 402(p)(6) of
the Clean Water Act ("CWA").  This rulemaking is also referred to as the Storm Water Phase II
rule. Section 431 (a) of the Appropriations Act directs the Administrator to submit a report
containing:

       (1) an in-depth impact analysis on the effect the  final regulations will have on urban,
       suburban, and rural local governments subject to the regulations, including an estimate of
             (A) the costs of complying with the 6 minimum control measures described in the
             regulations; and

             (B) the costs resulting from the lowering of the construction threshold from 5
             acres to 1 acre;

      (2) an explanation of the rationale of the Administrator for lowering the construction site
      threshold from 5 acres to 1 acre, including -

             (A) an explanation, in light of recent court decisions, of why a 1-acre measure is
            . any less arbitrarily determined than a 5-acre measure; and

             (B) all qualitative information used in determining an acre threshold for a
             construction site;

      (3) documentation demonstrating that storm water runoff is generally a problem in
      communities with populations of 50,000 to 100,000 (including an explanation of why the
      coverage of the regulation is based on a census-determined population instead of a water
      quality threshold); and

      (4) information that supports the position of the Administrator that the Phase II  storm
      water program should be administered as part of the National Pollutant Discharge
      Elimination System under section 402 of the Federal Water Pollution Control Act (33
      U.S.C. 1342).
                                          1-1

-------
       Section 431(c) of the Appropriations Act directs EPA to publish the reports in the Federal
Register for public comment. The Appropriations Act does not specify whether EPA should
seek and respond to public comment on the reports prior to submitting them to the Committees.
Section 431 (a) does provide, however, that the Administrator shall not promulgate the Phase II
rule until submitting the Section 431 (a) report to the Committees. EPA is subject to a judicial
consent decree in NRDCv. Browner. (D.D.C., Civ. No. 95-634 PLF) to take final action by
October 29, 1999 on the Phase II rule proposed earlier.  The Appropriations Act does not relieve
EPA from the timing of this rulemaking obligation.  Therefore, EPA will invite public comment
on the Section 431 (a) report after submitting it to the Committees. EPA will carefully review
and evaluate comments received and determine whether the comments warrant further action
regarding the October 29, 1999, final rule.

       As noted above, on October 29,1999, the Administrator of EPA will take final action on
a notice of proposed rulemaking under CWA section 402(p)(6), 33 U.S.C. § 1342(p)(6). On
January 9, 1998, at  63 Fed. Reg. 1536, EPA proposed to expand the National Pollutant
Discharge Elimination System (NPDES) permitting program for storm water to apply to
discharges from certain small municipal separate storm sewer systems (MS4s) and from
construction activity generally disturbing between one and five acres of land surface.  Although
EPA designated for regulation discharges from these two categories, the rulemaking would also
allow for waivers (for subsequent exclusion from regulation of certain sources in these
categories) and designation (for subsequent inclusion of certain sources that fall outside of the
categories).  Waivers would be available based on criteria by which the NPDES permitting
authority would determine a low potential for adverse water quality impact, and the permitting
authority would designate additional sources on a localized basis when necessary to protect or
remedy localized adverse water quality impacts.

       Rulemaking under CWA section 402(p)(6) is to be based on a study that EPA was
directed to provide to Congress under CWA section 402(p)(5).  Section 402(p)(5) provides that:
       The Administrator, in consultation with the States, shall conduct a study for the purposes
       of-
              (A) identifying those stormwater discharges or classes of stormwater discharges
       for which permits are not required pursuant to [CWA sections 402(p)(l) and (p)(2)];
              (B) determining, to the maximum extent practicable, the nature and extent of
       pollutants  in such discharges; and
              (C) establishing procedures and methods to control stormwater discharges to the
       extent necessary to mitigate impacts on water quality.

CWA section 402(p)(5) directed EPA to provide reports to Congress on the different components
of this study. In proposing the regulations under CWA section 402(p)(6), EPA identified the
reports to Congress comprising the study described in CWA section 402(p)(5), specifically,
Storm Water Discharges Potentially Addressed by Phase II of the National Pollutant Discharge
Elimination System Storm  Water Program: Report to Congress (U.S. EPA, 1995, EPA 833-K-
94-002). Today's report under section 431(a) of the Appropriations Act is a supplement to the
study described in CWA section 402(p)(5).

                                          1-2

-------
B. PURPOSE OF THE PHASE II RULE

       The Phase II rule would establish a cost effective, flexible approach for reducing
environmental harm by storm water discharges from many point sources of storm water that are
currently unregulated. Some of the costs of implementing the Phase II rule are discussed in
Chapter II of this report. A summary of the rule's benefits are described below.  EPA's
Economic Analysis of the Final Phase II Storm Water Rule fully analyzes the costs and benefits
expected from implementation of the rule.

       The environmental harm currently caused by discharges from municipal separate storm
sewer systems (MS4s) and construction activity is well documented:

•      Urbanization alters the natural infiltration capability of the land and generates a host of
       pollutants that are associated with the activities of dense populations, thus causing an
       increase in storm water runoff volumes and pollutant loadings in storm water discharged
       to receiving water bodies.

       The National Urban Runoff Program (NURP) Study (U.S. EPA 1983) indicated that
       discharges from MS4s draining runoff from residential, commercial, and  light industrial
       areas carried more than ten times the annual loadings of total suspended solids as did
       discharges from municipal sewage treatment plants that provide secondary treatment, and
       somewhat higher annual loadings of chemical oxygen demand (COD),  total lead, and
       total copper.

•      The National Water Quality Inventory (305(b)), 1996 report to Congress shows that
       urban runoff/storm  sewer discharges affect 13% of impaired rivers, 21% of impaired
       lakes and 45% of impaired estuaries.

•      Urban storm water runoff, sanitary sewer overflows, and combined sewer overflows have
       become the largest causes of beach closings in the United States in the past three years.
       A survey of coastal and Great Lakes communities found that more than 1,500 beach
       closings and advisories were attributable to storm water runoff in 1998  based on EPA
       data supplemented with additional data (Natural Resources Defense Council. 1998.
       "Testing the Waters Volume VIII-Has Your Vacation Beach Cleaned Up  Its Act?" New
       York, NY). Recreational bathers are at the highest risk for contracting  illnesses such as
       gastroenteritis, typhoid, dysentery, hepatitis, skin rashes, and respiratory infections.

•      The MS4 program will address illicit discharges, which can contribute high levels of
       pollutants, including heavy metals, toxic substances, oil and grease, solvents, nutrients,
       viruses and bacteria into receiving water bodies.

•      The NURP study found that pollutant levels from illicit discharges were high enough to
       significantly degrade receiving water quality and threaten aquatic, wildlife, and human
       health.
                                          1-3

-------
•      Discharges from construction activity impact the biological, chemical, and physical
       integrity of receiving waters. A number of pollutants are preferentially absorbed onto
       particles found in fine sediment. Estimates indicate that 80 percent of the phosphorus and
       73 percent of the Kjeldahl nitrogen in streams is associated with eroded sediment from
       construction and other activities.

•      Sediment yields from smaller construction sites are as high as or higher than the 20 to
       150 tons/acre/year measured from larger sites.

•      Siltation is the largest cause of impaired water quality in rivers and the third largest cause
       of impaired water quality in lakes, according to the 305(b) Report to Congress.

       The implementation of the six minimum measures identified for small MS4s should
significantly reduce pollutants in urban storm water compared to existing levels and do so in a
cost effective manner. Similarly, the implementation of best management practice ("BMP")
controls at small construction sites should also result in a significant reduction in pollutant
discharges and an improvement in surface water quality. EPA's Economic Analysis of the Final
Phase II Storm Water Rule details the expected benefits from implementation of the rule.  These
benefits include:

       •      Enhanced Commercial Fishing: Commercial fisheries are a significant part of the
             nation's economy. In 1997, the commercial shellfish catch was worth SI.04
             Billion and the finfish catch was worth $581 Million. 18% of surveyed estuary
             miles identified storm water as a significant source of impairment.

       •      Enhanced Recreational and Subsistence Fishing: The potential value of marine
             recreational fishing is $1.1 Billion to $11.3 Billion annually.  Pollutants in storm
             water may result in eliminating or decreasing the numbers or size of sport fish or
             shellfish in receiving waters.  In September 1996, there were 2,196 fish
             consumption advisories and about 25% of waters designated  for fishing did not
             support that use.

       •      Enhanced Opportunities for Boating: Storm water controls offer benefits to
             boaters by reducing sediment and other pollutants in waters, increasing water
             clarity and enhancing the experience for boating users.  EPA  estimates  that
             pollution reduction due to Phase II controls may result in 3,000 currently non-
             boatable miles of river becoming boatable.

       •      Enhanced Opportunities for Swimming: EPA estimates that Americans
             participated in 1.3 billion non-pool swimming days.  EPA estimates that at least
             28% of these trips are to either marine or fresh water that is impacted by runoff
             from Phase II sources.  For example, in 1998,  storm water runoff caused beach
             closures that resulted in the loss of an estimated 86,000 individual trips to beaches
             impacted by Phase II sources.

                                           1-4

-------
Enhanced Opportunities for Noncontact Recreation: Activities like picnicking,
jogging, biking, camping and hunting do not necessitate direct contact with water,
but water quality affects the ability to enjoy these activities when in close
proximity to water. Storm water controls reduce turbidity, odors, floating trash
and other pollutants and allow waters to be used as focal points for recreation,
enhancing the experience of current and future users.

Enhanced Nonconsumptive Wildlife Uses: An estimated 76.1 million people
participated in observing wildlife and waterfowl in 1991.  Storm water controls
that result in greater numbers or diversity of viewable wildlife species will
produce benefits.

Reduced Flood Damage: Storm water runoff controls may mitigate flood damage
by addressing problems due to the diversion of runoff, insufficient storage
capacity, and reduced channel capacity from sedimentation.

Drinking Water Benefits: Storm water was identified as a major source of
impairment in rivers, streams, lakes, reservoirs and ponds. Pollutants from storm
water runoff, such as solids, toxic pollutants (including pesticides) and bacteria,
may impose additional costs for treatment or even render the water unusable for
drinking.

Water Storage Benefits: Storm Water is a major source of impairment for
reservoirs.  The heavy load of solids deposited by storm water runoff can lead to
rapid sedimentation of reservoirs and the loss of needed water storage capacity.

Navigational Benefits: Storm water also delivers high sediment loads to rivers and
harbors critical to navigation and commerce. Where waters are used for
navigation, solids must be dredged and disposed of to maintain the utility of the
waterway.  An estimated 5% of these sediments (12.6 million cubic yards of
material) is attributed to storm water runoff from roads and constructions sites.
Storm water controls will reduce the rate and amount of sediment loadings.

Reduced Illness from Consuming Contaminated Seafood: Storm water controls
may reduce the presence of pathogens in seafood caught by commercial or
recreational anglers.  Researchers have estimated 2,700 cases of illness annually
from raw or partially cooked contaminated seafood.

Reduced Illness from Swimming in Contaminated Water: Epidemiological studies
have indicated that swimmers in water contaminated by storm water runoff are
more likely to experience illness than those that swim farther from a storm water
outfall.  By reducing  illicit connections and other sources of pathogens in storm
water, EPA estimates that up to 500,000 cases of illness will be avoided annually.
                             1-5

-------
       •      Enhanced Aesthetic Value: When storm water affects the appearance or quality of
              a water body, the desirability of working, living, traveling or owning property
              near that water body is similarly affected.  Improvements in water quality due to
              reductions in storm water pollution will result in benefits as these waters recover
              and become more desirable locations near which people want to live, work, travel
              or own property.

Thus, the rule will result in significant monetized financial, recreational and health benefits, as
well as benefits that EPA has been  unable to monetize.
                                           1-6

-------
 II. IMPACT OF PHASE II RULE ON LOCAL GOVERNMENTS
       This section responds to the Appropriations Act's direction to provide a report
 containing:

 "(1) an in-depth impact analysis on the effect the final regulations will have on urban, suburban,
 and rural local governments subject to the regulations, including an estimate of-
 (A) the costs of complying with the 6 minimum control measures described in the regulations;
 and
 (B) the costs resulting from the lowering of the construction threshold from 5 acres to 1 acre;"
 A. SUMMARY OF PHASE II RULE REQUIREMENTS

       EPA conducted an in-depth impact analysis of the effect of the final Storm Water Phase
 II Rule on local governments.  Two provisions of the Phase II rule are expected to result in
 compliance costs for local governments. These are the provision requiring certain municipalities
 to regulate discharges from their municipal separate storm sewer systems (MS4s) and the
 provision which extends the storm water construction program to cover sites between one and
 five acres in size.  The analysis considers potential cost impacts to all local governments,
 including urban, suburban and rural governments, and provides insight into the differing
 situations of small or very small local governments. Based on this analysis, EPA determined that
 the Phase II rule is not expected to have a significant impact on a substantial number of local
 governments.

 Municipal Storm Water Program:

       The Phase II Rule would automatically designate for regulation discharges from small
 MS4s located in urbanized areas, and require that NPDES permitting authorities examine for
 potential designation, at a minimum, a particular subset of discharges from small MS4s located
 outside of urbanized areas. The MS4 provision would result in costs primarily for local
 governments in urbanized areas.  An urbanized area is defined by the U.S. Census Bureau as an
 area with a population of at least 50,000 and a minimum  average population density of more than
 1,000 people per square mile. Thus, this rule would primarily affect suburban and urban local
 governments, because these MS4s are more likely to be located in urbanized areas.  Rural local
 governments may be designated on a case-by-case basis if the permitting authority determines
 that they have a significant impact on water quality.  The Phase I storm water program addressed
 runoff from "medium" and "large" MS4s, generally those discharges from governmental
jurisdictions serving populations of 100,000 or more people. The Phase II Storm Water
 regulations will address discharges from smaller MS4s.  The rule also would allow MS4s that are
 automatically designated because they are within an urbanized area to obtain a waiver from the
 otherwise applicable requirements if the discharges from small MS4s are not causing impairment
 of a receiving water body. Qualifications for the waivers vary depending on whether the MS4

                                         II-l

-------
 serves a population under 1,000 or a population under 10,000.

        Under the Phase II rule, a storm water discharge control program that meets the
 requirements of six minimum control measures would be administered within the jurisdiction of
 all regulated small MS4s. Small MS4 operators would design and administer the program, or
 would arrange with other government entities (including operators of nearby larger MS4s) to do
 so. These minimum control measures would consist of: public education and outreach on storm
 water impacts, public involvement/ participation, illicit discharge detection and elimination,
 construction site storm water runoff control, post-construction storm water management in new
 development and redevelopment, and pollution prevention/good housekeeping for municipal
 operations.  The Agency provides an analysis of the costs to local governments of implementing
 the six municipal minimum control measures in Section B below.
 Municipal Construction:

       The 1990 Phase I rule required all operators of construction activity disturbing five or
 more acres of land surface to apply for an NPDES permit for any resulting point source
 discharges of storm water.1 The construction provisions of the Phase II rule would extend
 similar requirements to construction projects that disturb between one and five acres of land.
 This provision would impose additional requirements on small construction projects of local.
 governments, regardless of whether the local government is urban, suburban or rural. The rule
 excludes routine road maintenance from the definition of construction, thereby excluding many
 municipal public works projects.  EPA expects that most new one to five acre road construction
 projects are likely to be built in conjunction with either larger development projects or State and
 Federal transportation programs and at least partially funded by these other sources.  The Phase
 II rule would also provide waivers from coverage based on the potential  to discharge storm water
 and cause a significant impact to water quality.  EPA's analysis of construction starts concluded
 that the-additional requirement for municipally constructed projects should not have a significant
 impact on a substantial number of the local governments subject to the regulation. EPA reports
 on its analysis of the costs to local governments of implementing the soil erosion control
 provisions for their construction sites between one and five acres in Section C below.
Regulatory Flexibility:

       In promulgating Storm Water Phase II, EPA examined regulatory flexibility issues and
potential cost impacts on small entities, including small local governments.  In order to solicit
       ' On December 18, 1991, Congress enacted the Intermodal Surface Transportation
Efficiency Act (ISTEA), which postponed NPDES permit application deadlines for most storm
water discharges associated with industrial activity at facilities that are owned or operated by
small municipalities, including construction activity over five acres.

                                          TI-2

-------
input from potentially regulated small entities, EPA convened a Small Business Regulatory
Enforcement Flexibility Act (SBREFA) Panel which included small local government
representatives as well as other stakeholders.  EPA conducted an analysis and determined that
the rule was expected not to have a significant impact on a substantial number of small local
governments or other small entities. However, in order to provide additional flexibility for small
local governments, EPA included several programmatic options and potential waivers for small
governments.

       The rule would allow for a great deal of flexibility by providing various options for
obtaining permit coverage and satisfying the required minimum control measures. For example,
the NPDES permitting authority would be able to incorporate by reference qualifying State,
Tribal, or local programs in a NPDES general permit and recognize existing responsibilities
among different governmental entities for the implementation of minimum control measures. In
addition, a regulated small MS4 could participate in the storm water management program of an
adjoining regulated MS4 and could arrange to have another governmental entity implement a
minimum control measure for them. The rule also provides potential waivers for MS4s serving a
population less than 10,000 and also for construction projects not expected to significantly
impair water quality. Therefore, Storm Water Phase II is not expected to have a significant
impact on a substantial number of small local governments, and offers significant flexibility to
those local governments in implementing provisions of the rule which may result in compliance
cost impacts.
B. IMPACTS OF THE MUNICIPAL MINIMUM CONTROL MEASURES ON LOCAL GOVERNMENTS

       EPA estimated that the overall annual cost to local governments of implementing a storm
water program based on the six minimum measures would be $297 million.  EPA developed this
estimate using actual program cost information from Phase II communities with existing storm
water programs.  The estimate assumes that all of the 5,040 Phase II designated municipalities
would incur program costs and that costs are related to the size of the community served.
Therefore, the Agency probably overestimates national costs because permitting authorities can
waive permitting requirements for MS4s serving up to 10,000 people.

       EPA conducted an in-depth analysis of the potential cost of complying with the six
minimum measures on local governments in urbanized areas. These local governments are
primarily urban and suburban, although a few rural governments may be designated by States to
be included in the program based on potential water quality impacts. While the total regulatory
costs associated with Phase II include all sizes of local government, EPA specifically considered
the impacts to small local governments as required under the Regulatory Flexibility Act, as
amended by the Small Business Regulatory and Enforcement Fairness Act. In preparing the
analysis, EPA compared estimated annual compliance costs with annual municipal revenues for
4,455 small local governments (municipalities with fewer than 50,000 people) and evaluated
cost-to-revenue ratios for indication of significant economic impacts. The results,-which are
reported in the economic analyses prepared for the proposed and final rules, led EPA to conclude
that there would not be a substantial economic impact on a significant number of small

                                         II-3

-------
 governments; EPA expects even less of an impact on larger governments. Below is a summary
 of EPA's analysis.
 Cost Analysis:

       EPA estimated annual costs for the municipal programs based on a fixed cost component
 and a variable cost component.  The fixed cost component included costs for the municipal
 application, record keeping, and reporting activities.  On average, EPA estimated annual costs of
 $1,525 per municipality. Variable costs include the costs associated with annual operations for
 the six minimum measures. EPA reviewed cost data from existing Phase I storm water programs
 and cost data gathered from Phase II communities by the National Association of Flood and
 Storm Water Management Agencies (NAFSMA). These costs reflect the actual operating costs
 of program elements that are comparable to the six minimum measures for municipalities
 representing a wide range of population sizes. EPA estimated costs on a per household basis
 from both data sets. Annual mean costs per household are comparable across the data sets: $8.93
 (NASFMA) and $8.85 (Phase I).

       Total annual cost for each of the 4,455 municipalities was calculated as the sum of the
 $1,525 fixed cost and the urbanized area household estimate multiplied by the per household cost
 based on the NAFSMA data.2 For example, a municipality with 5,000 households would have a
 total program cost of:                                                                    .

                            $46,175 = $ 1,525 + (5,000 * $8.93).

 Small Local Governments:

       EPA estimated municipality revenues based on state-level revenue data collected by the
 U.S. Bureau of the Census 1992 Census of Governments. The Bureau of the Census reports
 municipal government revenues by population size for eight size categories including three used
 by EPA: less than 10,000, 10,000 to 24,999, and 25,000 to 49,999. For every state, EPA
 gathered the aggregate municipal revenue data and aggregate municipal population data reported
 by the Bureau of the Census for these three size categories.3 EPA then divided revenue by
population to obtain revenue per capita for each size category within each state. EPA merged
this data set with the Phase II municipality population data set and multiplied the appropriate per
capita revenue estimates by the Phase II urbanized area populations to obtain 4,455 estimates of
annual municipal revenues.
       " Based on Census data, EPA used a conversion factor of 2.62 people per household to obtain household
estimates for the Phase II communities.

        EPA did not adjust municipal revenue from their 1991 values to 1998 values, which is the unit of
measure EPA used for costs. There is no standard adjustment factor for municipal revenues. Thus, the cost-to-
revenue ratio probably overstates the cost impact.

                                           II-4

-------
       Finally, EPA divided the 4,455 cost estimates by the 4,455 revenue estimates to obtain
cost-to-revenue ratios.  EPA categorized these ratios according to whether they were less than
1% (i.e., cost is less than 1% of revenue), between 1% and 3%, and greater than 3%. Figure A
summarizes the results, showing that the cost-to-revenue ratios were less than 1% for 89% of the
Phase II municipalities and greater than 3% for less than 1% of municipalities.

       Under, the Phase II rule, the permitting authority could waive permitting requirements for
systems serving less than 1,000 people. All of the municipalities with cost-to-revenue ratios that
are greater than 3% have populations less than 1,000 people, and may qualify for a waiver.
Consequently, the flexibility of the rule addresses any potentially significant adverse cost
impacts. Because no Phase II municipality with a population of more than 6,000 had a cost-to-
revenue ratio of more than 1%, EPA does not expect this provision will have significant
economic impacts on the 585 municipalities with populations larger than 50,000.

                   Figure A. Summary of Cost-to-Revenue Ratios for
              4,455 Phase II Municipalities with Populations Less than 50,000
I%to3%
(10.2%)
More than 3%
/~ (0.7%)
                                                 Less than 1 %
                                                   (89.1%)
C.  COSTS OF THE SOIL EROSION CONTROL PROVISION ON LOCAL GOVERNMENTS

       EPA's cost analysis for the soil erosion control provision multiplies cost estimates per
construction site for soil erosion control measures and administrative costs by the number of
construction sites potentially affected by the rule. EPA estimates that the rule would apply to
approximately 110,223 currently unregulated construction starts per year (using 1998 estimates)
out of a total of 528,499 construction starts. Annual costs associated with installing the soil
erosion controls and completing permitting activities is estimated as $505 million. Less than
0.5% (< $500,000) is expected to accrue from local governments.
Cost Analysis:

       Most soil erosion control costs would accrue to the private sector, primarily to
dischargers in the construction industry. However, local governments may also incur soil
                                          II-5

-------
erosion control costs for discharges from public works projects that disturb between one and five
acres (costs borne either directly by the local government or indirectly through a contractor).
Since routine road maintenance is excluded from coverage under Storm Water Phase II, those
public works starts are excluded from analysis. EPA used the site-based estimates of soil erosion
control costs that it developed for the economic analysis of the final rule and Bureau of Census
construction start data to estimate the expected annual impact on local governments. Table A
summarizes the two types of costs by site size that a construction company or public works
department may incur.

               Table A. Summary of Site-Based Soil Erosion Control Costs
Cost
Administrative a
Soil Erosion Control
BMPs b
Total Cost
Annualized Cost
(7%)c
1 Acre Site
$937
$1,206
$2,143
$202
3 Acre Site
$937
$4,598
$5,535
$522
5 Acre Site
$937
$8,709
$9,646
$910
Notes: . •
a. These activities would include costs to submit a notice of intent to be covered by a general permit, to notify the
municipality, to develop a storm water pollution prevention plan, to retain records, and to file a notice of
termination from a general permit.
b. BMPs (best management practices) costs are based on combinations of the following that differ across sites
with different sizes, slopes, and soil types: silt fence, mulch, seed/mulch, stabilized entrance, stone check dam,
earth dike, and sediment traps.
c. Annualized cost assumes a 20-year period and a 7% cost of capital. The capitalization factor is 0.09439.
Small Local Governments:

       There are four categories of local governments which may experience costs of
compliance associated with the Phase II rule. These are:

       1) Phase I jurisdictions (subject to Phase II requirements for construction between one
       and five acres; already required to have a municipal storm water program),
       2) Phase II jurisdiction above 50,000 population (subject to Phase II municipal and
       construction requirements),
       3) Phase II jurisdictions below 50,000 population (subject to Phase II municipal and
       construction requirements; subject to SBREFA review), and
       4) Jurisdictions that are not required to have a municipal storm water program under
       Phase I or Phase II (subject to Phase II requirements for construction).
                                          II-6

-------
 The greatest potential economic impact of the soil erosion control provision is expected to be on
 the third category, because they would incur soil and erosion control costs in addition to annual
 program costs for the six minimum measures, and because of their smaller size. Therefore, the
 analysis below focuses on the impacts on these small local governments.

       To evaluate the severity of potential impact, EPA used the Bureau of the Census
 construction start database to estimate the annual  number of construction starts in Phase II
 municipalities that are classified by the Bureau of the Census as a "public works" start, excluding
 routine road maintenance. The data showed that 2% of municipalities are expected to have a 1-
 acre start, 2% are expected to have a 3-acre start and 1% are expected to have a 5-acre start.
 These results indicate that local governments would not incur soil and erosion control costs on an
 annual basis, because they would not necessarily have Phase II construction starts in any given
 year. As a conservative assumption (i.e., tending  to overstate costs), EPA annualized the costs
 over a 20-year period, assuming a 7% cost of capital (see Table A). The 20-year assumption is
 conservative because it implies higher construction rates than the data suggest [i.e., 1-acre site
 (5%), 3-acre site (5%), and 5-acre site (5%)]. EPA then added  the annualized values to obtain an
 annual cost  of $ 1,634 per municipality for the soil erosion control provision.

       Because the soil erosion control provision of the Phase II rule would apply to discharges
 from construction sites between one and five acres regardless of location; local governments
 other than Phase II designated municipalities could incur costs.  EPA compared the annualized
 value across all site sizes of $1,634 to the national mean estimate of local.government revenues.
 For the smallest municipality size category,  the mean annual revenue was $1.4 million (1991
 dollars; 1992 Census of Governments).  The cost-to-revenue ratio for the smallest size category
 is well below 1%.

       Finally, EPA then added the cost of complying with the Phase II soil erosion program for
 small construction to the cost-to-revenue ratios for the MS4 program discussed above to evaluate
 the combined impact on Phase II municipalities of the municipal minimum  measures and soil
 erosion control costs that may be borne directly or indirectly (passed through from construction
 companies). Based on this revised cost-to-revenue analysis, the combined costs are not expected
 to have a significant economic  impact on a substantial number of designated Phase II
 municipalities.

      .Figure B summarizes the cost-to-revenue impacts for all 4,455 Phase II municipalities
 with populations less than 50,000 (bar on left).  Figure B also summarizes impacts for these
 Phase II municipalities assuming that the municipalities with populations below 1,000 are
 granted waivers so they incur soil erosion control  costs as regulated small construction site
 dischargers but no program costs as small MS4 dischargers (bar on right). In either case, a vast
majority of municipalities would not incur annual costs that are greater than 1% of revenues and
 fewer than 2% of municipalities would incur costs that are greater than 3% of revenues.
Therefore, EPA concluded that the Phase II rule would not have a significant impact on
potentially regulated small local governments.
                                          II-7

-------
 Figure B. Summary of Cost-to-Revenue Ratios Revised to Include Soil
Erosion Control Costs for 4,455 Phase II Municipalities with Populations
                                  under 50,000
                         100%-j

                          80%-

               Percentof    60%"
             Municipalities  40%-

                          20%-
                           0%
                                                      Waivers
                                   No Waivers

                                       Cost-to-Revenue Ratio
                             HI Less than  I %  D'%to3%  QMore than 3%
           * "No Waivers" estimates costs assuming (for the purpose of this analysis) that no small local
           governments with populations below 1,000 receive a waiver and, therefore, are subject to both the
           municipal and the soil erosion provisions of Phase II. Even if this were to occur, the potential
           impacts are not significant.
                                        II-8

-------
III.   RATIONALE FOR THE ONE ACRE CONSTRUCTION THRESHOLD

       This section responds to the Appropriations Act's direction to provide a report
containing:

"(2) an explanation of the rationale of the Administrator for lowering the construction site
threshold from 5 acres to 1  acre, including —
(A) an explanation, in light of recent court decisions, of why a 1-acre measure is any less
arbitrarily determined than  a 5-acre measure; and
(B) all qualitative information used in determining an acre threshold for a construction site;"
BACKGROUND

       In 1990, EPA promulgated the first phase of the NPDES permit application rules for
storm water. (National Pollutant Discharge Elimination System Permit Application
Requirements for Storm Water Discharges, 55 Fed. Reg. 47990 (Nov. 16, 1990), referred to as
the "Phase I" rule).  As directed under CWA section 402(p)(4)(A), the Phase I rule set forth the
permit application requirements for storm water discharges "associated with industrial activity,"
including, applicability provisions defining the term "storm water discharge associated with
industrial activity." Under CWA section 402(p)(2)(B), storm water discharges associated with
industrial activity were excluded from the moratorium against permitting discharges composed
entirely of storm water.

       Among other things, the Phase I rule defined storm water discharge  associated with
industrial activity to include discharges from "construction activity including clearing, grading
and excavation activities except: operations that result in the disturbance of less than five acres of
total land area which are not part of a larger common plan of development or sale." 40 C.F.R.
122.26(b)(14)(x). In 1992, a court ruled that the five acre threshold used for defining
construction activity as "industrial activity" was improper because EPA had failed to identify
information to support its position that construction activities on less than five acres are non-
industrial in nature.

       The Phase II rule would regulate storm water discharges from additional smaller
construction activities.  The rule would regulate these construction-related storm water sources
under CWA section 402(p)(6) to protect water quality rather than under CWA section 402(p)(2).
Designation under 402(p)(6) gives States and EPA the flexibility to waive the permit requirement
for construction activity that is not likely to impair water quality, and to designate additional
sources below one acre that are likely to cause water quality impairment. Thus, the one acre
threshold under the Phase II rule would not be an absolute threshold like the five acre threshold
that applies under the Phase I rule. The one acre threshold is reasonable for accomplishing the
water quality goals of CWA section 402(p)(6) because it results in 97.5% of the total acreage
disturbed by construction being designated for coverage by the NPDES storm water program,
while excluding from automatic coverage the numerous smaller sites that represent 24.7% of the

                                          III-l

-------
total number of construction sites.
RATIONALE FOR FIVE ACRE THRESHOLD IN THE PHASE I RULE

       In the preamble to the Phase I rule, which'regulates storm water discharges from
construction activity disturbing five acres or more as "storm water discharges associated with
industrial activity," EPA had explained that the construction industry should be subject to storm
water permitting because at a high level of intensity, construction is equivalent to other regulated
industrial activities. 55 Fed. Reg. 48033. The Phase I rule regulates storm water "associated
with industrial activity."  EPA had proposed that the Phase I regulations apply to construction
site discharges from sites disturbing down to one acre. EPA increased the size threshold to five
acres for the final rule.

       After a judicial challenge to the Phase I regulations, the Ninth Circuit remanded the
regulation to EPA for further proceedings. NRDC v. EPA. 966 F.2d 1292, 1306 (9lh Cir.  1992).
To support the increased  threshold (from one to five acres), the Agency had explained that larger
sites typically involve heavier equipment for removing vegetation and bedrock than smaller sites.
55 Fed. Reg. 48036. The court found that EPA's rationale for increasing the limit was inadequate
because the Agency cited no information to support its perception that construction activities on
less than five acres are non-industrial in nature. 966 F.2d at 1306. Thus, the Court focused on
the relationship between the size threshold and the statutory reference  to "industrial."
RATIONALE FOR ONE ACRE THRESHOLD IN THE PHASE II RULE

       In lowering the threshold to one acre in the Phase II rule, the Phase II rule would not
regulate discharges from small construction site as "industrial activity."  Instead, EPA interprets
the text of CWA section 402(p)(6) as a basis to designate small construction site discharges as
sources "to be regulated to protect water quality." EPA interprets this language as less restrictive
than the terms "associated with industrial activity" for the purpose of establishing an
applicability threshold that is based on size alone but which may be modified by permitting
authorities to account  for higher and lower threat sources.  In addition to water quality
considerations, the text of CWA section 402(p)(6) allows for designations based on
considerations of administrative feasibility by specifying that the Agency has discretion to
identify sources "to be regulated."

       Though the Phase II rule would not  regulate a discharge from a construction site below
the five acre threshold as a "discharge associated with industrial activity," the Phase II rule
nonetheless responds to the Ninth Circuit's direction to conduct further rulemaking on the matter
of discharges from sites disturbing more than one acre (from the Phase I  proposed rule) and from
sites disturbing less than five acres (from the Phase I final rule). For discharges from sources in
this category, which the Agency still believes present water quality concerns based on the
potential  for water quality impairment due to gross sediment runoff (among other pollutants),

                                          III-2

-------
 CWA section 402(p)(6) rather than CWA section 402(p)(2)(B) and (3) provides a more sensible
 basis to address the sources that threaten water quality. In light of the Agency's decision to
 regulate these sources down to one acre to protect water quality with controls similar to those
 applied to Phase I sources, EPA believes it is unnecessary to examine further whether sites below
 five acres are "associated with industrial activity."

       EPA is regulating storm water discharges from construction activity disturbing between 1
 and 5 acres because the cumulative impact of many sources, and not just a single identified
 source, is typically the cause for water quality impairments, particularly in relation to
 sedimentation-related water quality standards.

       The one acre threshold provides an  administrative tool for more easily identifying those
 sites that are identified for coverage by the rule (but may receive a waiver) and those that are not
 automatically covered (but may be designated for inclusion). Although all construction sites less
 than five acres could have a significant water quality impact cumulatively, EPA is automatically
 designating for permit coverage only those storm water discharges from construction sites that
 disturb land equal to or greater than one acre.  Categorical regulation of discharges from
 construction below this one acre threshold would overwhelm the resources of permitting
 authorities and might not yield corresponding water quality benefits. Construction activities that
 disturb less than one acre make up, in total, a very small percentage of the total  land disturbance
 from construction nationwide (about 2.5%).

       In addition to  the diminishing water quality benefits of regulating all sites below one acre,
 the Agency relies on practical considerations in establishing a one acre threshold and not setting
 a lower threshold. Regardless of the threshold established by EPA, a NPDES permit can only be
 required if a construction site has a point source discharge.  A point source discharge means that
 pollutants are added to waters of the United States through a discernible,  confined, discrete
 conveyance. "Sheet flow" runoff from a small construction site would not result in a point
 source discharge unless and until it channelized.  As the amount of disturbed land surface
 decreases, precipitation is less likely to channelize and create a "point source" discharge
 (assuming the absence of steep slopes or other factors that lead to increased channelization).
 Categorical designation of very small sites may create confusion about applicability of the
 NPDES permitting program to those sites.  EPA's one acre threshold reflects, in part, the need to
 recognize that smaller sites are less likely to result in  point source discharges. Of course, the
NPDES permitting authority could designate  smaller  sites (below one acre, assuming point
 source discharges occur from the smaller designated sites) for regulation if a watershed or other
 local assessment indicated the need to do so.  The Phase II rule would include this designation
 authority at 40  CFR 122.26(a)(9)(i)(D) and (b)(15)(ii).

       Though location-specific water quality studies would provide the  ideal information base
from which to make regulatory decisions, the Phase II rule establishes one acre as a default
standard for regulation in the absence of location-specific studies.  The rule does account for
location-specific water quality information, however, for any deviation from the default standard
through additional designations and waivers.  The rule codifies the ability of permitting
authorities to provide  waivers for sites greater than or equal to one acre and designate additional

                                           III-3

-------
 discharges from small sites below one acre when location-specific information suggests that the
 default one acre standard is either unnecessary (waivers) or too limited (designations) to protect
 water quality.
 OTHER QUALITATIVE AND QUANTITATIVE INFORMATION ON SIZE THRESHOLDS

        EPA had difficulty evaluating the water quality consequences of designating specific size
 thresholds because, while generally proportional to the size of the disturbed site, the water
 quality threat posed by discharges from construction sites of differing sizes varies nationwide,
 depending on the local climatological, geological, geographical, and hydrological influences. In
 order to ensure improvements in water quality nationwide, however, the Phase II rule does not
 allow various permitting authorities to establish different size thresholds except based on the
 waiver and designation provisions of the rule. EPA believes that a national one acre threshold
 for automatic designation, coupled with procedures for waiving sites above one acre and for
 designating sites below one acre based on local water quality considerations, ensures protection
 against adverse water quality impacts from storm water discharges from small construction sites
 while not overburdening the resources of permitting authorities and the construction industry.

        EPA believes that the water quality impact from small construction sites is as high as or
 higher than the impact from larger sites on a per acre basis.  The concentration of pollutants in
 the runoff from smaller sites is similar to the concentrations in the runoff from larger sites. The
 proportion of sediment that makes it from the construction site to surface waters  is likely the
. same for larger and smaller construction sites in urban areas because the runoff from either site is
 usually delivered directly to the storm drain network where there is no opportunity for the
 sediment to be filtered out.

        The expected contribution of total sediment yields from small sites depends, in part, on
 the extent to  which erosion and sedimentation controls are being applied. Because current storm
 water regulations are more likely to require erosion and sedimentation controls on larger sites in
 urban areas, smaller construction sites that lack such programs are likely to contribute a
 disproportionate amount  of the total sediment from construction activities (MacDonald, L.H.
 1997. Technical Justification  for Regulating Construction Sites 1-5 Acres in Size. Unpublished
 report submitted to the U.S. Environmental Protection Agency, Washington). Smaller
 construction  sites are less likely to have an effective plan  to control erosion and sedimentation,
 are less likely to properly implement and maintain their plans, and are less likely to be inspected
 (Brown, W. and D. Caraco. 1997. Controlling Storm Water Runoff Discharges from Small
 Construction Sites: A National Review. Submitted to the U.S. Environmental Protection
 Agency, Office of Wastewater Management, Washington, DC. by the Center for Watershed
 Protection, Silver Spring, MD).

       To confirm its belief that sediment yields from small sites are as high as or higher than
 the 20 to 150 tons/acre/year measured from larger sites, EPA gave a grant to the Dane County,
 Wisconsin Land Conservation Department, in cooperation with the USGS, to evaluate sediment
                                          .111-4

-------
runoff from two small construction sites. The first was a 0.34 acre residential lot and the second
was a 1.72 acre commercial office development.  Runoff from the sites was channeled to a single
discharge point for monitoring. Each site was monitored before, during, and after construction.

       The Dane County study found that total solids concentrations from these small sites are
similar to total solids concentrations from larger construction sites.  Results show that for both of
the study sites, total solids and suspended solids concentrations were significantly higher during
construction than either before or after construction. For example, preconstruction total solids
concentrations averaged 642 mg/L during the period when ryegrass was established, active
construction total solids concentrations averaged 2,788 mg/L,  and post-construction total solids
concentrations averaged 132 mg/L (on a pollutant load basis, this equaled 7.4 Ibs
preconstruction, 35 Ibs during construction, and 0.6 Ibs post-construction for total solids).  While
this site was not properly stabilized before construction, after construction was complete and the
site was stabilized, post-construction concentrations were more than 20 times less than during
construction.  The results were even more dramatic for the commercial site. The commercial site
had one preconstruction event, which resulted in total solids concentrations of 138 mg/L, while
active construction averaged more than 15,000 mg/L and post-construction averaged only 200
mg/L (on a pollutant load basis, this equaled 0.3 Ibs preconstruction, 490 Ibs during construction,
and 13.4 Ibs post-construction for total solids). The active construction period resulted in more
than 75 times  more sediment than either before or after construction (Owens, D.W., P. Jopke,
D.W. Hall, J. Balousek and A. Roa. 1997. "Soil Erosion from Small Construction Sites in Dane
County, Wisconsin." Draft Report. USGS and Dane County Land Conservation Department,
WI).

       Construction  start data indicates that excluding construction sites below one acre from
coverage under the Phase II rule would exclude a significant percentage of sites from automatic
coverage while only excluding a small percentage of the total acreage. As is indicated in
Table B, by choosing a nationwide threshold of one acre, the Phase I and Phase II rules will
together address 97.5% of the national disturbed acreage yet will only regulate 75.3% of the
construction starts. The remaining construction starts (24.7% or 130,435 starts) each occur on
less than one acre of disturbed land and together constitute only 2.5% of total acreage disturbed
by construction.
                                           III-5

-------
 Table B. Percentage of national disturbed acreage and construction starts addressed by
 regulating all construction above different thresholds

all sites
greater than 1 .0 acres
greater than 2.0 acres
greater than 3.0 acres
greater than 4.0 acres
greater than 5.0 acres
percentage of national
disturbed area controlled by
regulating all sites:
100%
97.5 %
92.3 %
87.8 %
83.7 %
78.1 %
number of construction starts
addressed (percent of national
total)
527,774(100%)
397,309 (75.3 %)
301,941 (57.2 %)
253,224 (48.0%)
221,471(42.0%)
188,425(35.7%)
* Table includes all construction starts. It does not exclude starts already regulated by Phase I,
equivalent State programs, or potential Phase II waivers.
       A two acre threshold would have tripled the total number acres that would not be
designated for permit coverage. A threshold below one acre would have significantly increased
the number of sites regulated without significantly increasing the number of acres for which
storm water controls would be required.  Thus, the additional increment in water quality
protection that would be achieved by a lower size threshold would have resulted in a
disproportionately higher burden on the regulated community.
CONCLUSION ON ONE ACRE THRESHOLD

       The Ninth Circuit concluded that EPA arbitrarily defined discharges "associated with
industrial activity" when the Agency established the five acre size threshold, particularly in light
of the Agency's proposal to establish the threshold at one acre.  The Phase II one acre threshold
is not arbitrary because (1) sediment loads from disturbed land surface cause adverse impacts on
water quality, (2) as site size decreases, the likelihood that precipitation will create "discernible,
confined, discrete conveyances" through channelization decreases, (3) the one acre threshold is
not an absolute threshold because NPDES authorities can waive the threshold for sites (and
during seasons) when there is a lower potential for a discharge that would impair water quality
and can designate sources below the threshold where necessary to protect water quality on a
localized basis, and (4) the number of additional sites that would be regulated by a threshold
below one acre is disproportionately high relative to the total number of acres disturbed by those
sites.

       EPA recognizes that the size criterion alone may not be a perfect predictor of the need for
regulation, but effective protection of water quality depends as much on simplicity in
                                          III-6

-------
implementation as it does on the scientific information underlying the regulatory criteria. The
default size criterion of one acre will ensure protection against adverse water quality impacts
from storm water discharges from small construction sites while not overburdening the resources
of permitting authorities and the construction industry to implement the program to protect water
quality in the first place.  Further,  as noted above, NPDES permit authorities can designate
sources below one acre where necessary to protect water quality in a particular area, or waive
sites above one acre where NPDES permit coverage under the Phase II rule is not necessary to
protect water quality.
                                          III-7

-------
 IV. STORM WATER PROBLEMS IN CENSUS DESIGNATED URBANIZED AREAS

 This section responds to the Appropriations Act's direction to provide a report containing:

 "(3) documentation demonstrating that storm water runoff is generally a problem in communities
 with populations of 50,000 to 100,000 (including an explanation of why the coverage of the
 regulation is based on a census-determined population instead of a water quality threshold);


 BACKGROUND

       In 1990, EPA promulgated the first phase of the NPDES permit application rules for
 storm water ("Phase I"). Phase I required NPDES permits for storm water discharges from large
 and medium municipal separate storm sewer systems generally serving populations of 100,000 or
 more.  Areas with a combined sewer were not included in the total population served for Phase I.

       This definition of large and medium MS4s for Phase I created so-called "donut holes."
 Donut holes are unregulated MS4s located within those urbanized areas that include systems .
 covered by the Phase I storm water program, but are not currently addressed by the storm water
 program because the Phase I regulations specify applicability based on political jurisdiction.  In
 other words, donut holes are geographic gaps in the existing NPDES storm water program's
 regulatory scheme. Storm water discharges from donut hole areas present a problem due to their
 adverse impacts on local waters, as well as by frustrating the attainment of water quality goals of
 neighboring regulated communities.

       The storm water Phase II rule designates discharges from small MS4s located in
 urbanized areas for NPDES permit coverage. EPA adopted the Bureau of the Census definition
 of an urbanized area as comprising a place and the adjacent densely settled surrounding territory
 that together have a minimum population of 50,000 people.  A permitting authority may
 designate  additional small MS4s after the authority develops designation criteria and applies
 those criteria to small MS4s located outside of an urbanized area, in particular those with a
 population of 10,000 or more and a population density of at least 1,000 per square mile. The
 permitting authority may waive the requirement for a permit for any small MS4 serving a
jurisdiction with a population of less than 1,000 unless storm water controls are needed because
 the MS4 is contributing to a water quality impairment. The permitting authority may also waive
 permit coverage for MS4s serving a jurisdiction with a population of less than 10,000 if all
 waters that receive a discharge from the MS4 have been evaluated and discharges from the MS4
 do not significantly contribute to a water quality impairment or  have the potential to cause an
 impairment. The Phase II rule also allows States with a watershed permitting approach to phase
 in coverage for MS4s in jurisdictions with populations under 10,000.
                                         IV-1

-------
 EPA's RATIONALE FOR BASING REGULATION ON CENSUS-DETERMINED
 POPULATION RATHER THAN A WATER QUALITY THRESHOLD

       EPA adopted the Bureau of the Census definition of "urbanized area" for the purposes of
 the Phase II rule. The existing storm water Phase I rule already covers discharges from MS4s
 with more than 100,000'population.  Phase II would address the remaining MS4s in urbanized
 areas.

       The Bureau of the Census defines an urbanized area as comprising a place and the
 adjacent densely settled surrounding territory that together have a minimum population of 50,000
 people. The densely settled surrounding territory generally has at least  1,000 people per square
 mile. The Bureau of the Census definition of "urbanized area," adopted by EPA for the purposes
 of the Phase II rule, was published in the Federal Register (55 FR 42592, October 22, 1990).

       EPA is using urbanized areas to automatically designate regulated small MS4s on a
 nationwide basis for several reasons:

 (1) Water Quality Impacts from Urban Runoff

       Studies and data show a high correlation between degree of development/ urbanization
 and adverse impacts on receiving waters due to storm water. See section A below for a full
 discussion of storm water impacts due to urban development.

 (2) Addresses gaps in coverage

       The blanket coverage within the urbanized area encourages the watershed approach and
 addresses the problem of "donut-holes," where unregulated areas are surrounded by areas
 regulated under Phase I.

 (3) Pollution Prevention

       This approach targets present and future growth areas as a preventative measure to help
 ensure water quality protection. Urbanized areas have experienced significant growth over the
past 50 years. According to EPA calculations based on Census data from 1980 to 1990, the
national average rate of growth in the United States during that 10-year period was more than 4
percent. For the same period, the average rate of growth within urbanized areas was 15.7 percent
 and the average for outside of urbanized areas was just more than 1 percent. Table C below
 illustrates the growth of urbanized areas for the past five Census (EPA, 1995). The new
development occurring in these growing areas can provide  some of the best opportunities for
implementing cost-effective storm water management controls.
                                        IV-2

-------
 Table C. Growth of Urbanized Areas in the United States Between 1950 and 1990
Year
1950
1960
1970
1980
1990
Number of
Urbanized
Areas
157
213
273
366
405
Population in Urbanized Areas (millions)
Total
69.2
95.8
120.7
139.2
160.4
Central
Cities
48.4
57.9
65.1
67.0 -
79.7
Urban Fringe
20.9 '
37.8
55.6
72.1
80.7
Land Area
(sq. mi.)
19,728
25,544
35,081
52,017
61,520
 (4) Simplified Designation and Coverage

       The determination of urbanized areas by the Bureau of the Census allows operators of
 small MS4s to quickly determine whether they are included in the NPDES storm water program
 as a regulated small MS4.

       Using urbanized areas as a basis for designation effectively targets resources to the most
 densely developed territory. The 405 urbanized areas in the United States cover only 2 percent
 of the total U.S. land areas yet contain approximately 63 percent of the nation's population.
DOCUMENTATION OF WATER QUALITY PROBLEMS DUE TO STORM WATER
RUNOFF FROM URBANIZED AREAS

       EPA has compiled a number of studies demonstrating that storm water runoff is generally
a problem in urbanized areas.  This information is divided into storm water impacts due to urban
development (section A below) and other discharges to municipal storm sewers (section B
below). The Appropriations Act specifically requested that this report provide "documentation
demonstrating that storm  water runoff is generally a problem in communities with populations of
50,000 to 100,000."  While 50,000 is the population threshold used by the Bureau of the Census
for defining urbanized areas and EPA adopted the Census definition for the purpose of automatic
designation in the Phase II rule, the studies below indicate that water quality impacts will  occur
in these areas and potentially in areas with lower population densities as well. The Phase  II rule
would allow the lower population density areas to be designated on a case by case basis.
                                        IV-3

-------
A.  Storm Water Impacts Due to Urban Development

       EPA's 1995 Storm Water Phase II Report to Congress (EPA, 1995) and the Coastal Zone
Management Measures Guidance (EPA, 1992) describe the impacts from urbanization.
Urbanization impacts water quality principally through changes in hydrology and increases in
pollutant loadings. Increases in population density and imperviousness due to urbanization can
result in significant changes to stream hydrology including:
       - increased peak discharges compared to predevelopment levels;
       - increased volume of urban runoff produced by each storm in comparison to
       predevelopment conditions;
       - decreased time needed for runoff to reach the stream, particularly if extensive drainage
       improvements are made;
       - increased frequency and severity of flooding;
       - reduced streamflow during prolonged periods of dry weather due to reduced level of
       infiltration in the watershed;
       - greater runoff velocity during storms due to the combined effects of higher peak
       discharges, rapid time of concentration, and the smoother hydraulic surfaces that occur as
       a result of development.

       An increase in imperviousness can also significantly decrease the amount of water
infiltration, reducing groundwater recharge.

       The types of pollutants  found in urban runoff include sediment, nutrients, oxygen-
demanding substances, pathogens, road salts, hydrocarbons, heavy metals, and toxics. In
addition, thermal impacts from increased temperature of urban runoff and loss of riparian habitat
can severely impair aquatic organisms that have finely tuned temperature limits.

1. Urbanization and Imperviousness

       Urbanization alters the natural infiltration capability of the land and generates a host of
pollutants that are associated with the activities of dense populations, thus causing an increase in
storm water runoff volumes and pollutant loadings in storm water discharged to receiving
waterbodies (U.S. EPA, 1992). Urban development increases the amount of impervious surface
in a watershed as farmland, forests, and meadowlands with natural infiltration characteristics are
converted into buildings with rooftops, driveways, sidewalks, roads, and parking lots  with
virtually no ability to absorb storm water.  Storm water and snow-melt runoff wash over these
impervious areas, picking up pollutants along the way while gaining speed and volume because
of their inability to disperse and filter into the ground.  What results are storm water flows that
are higher in volume, pollutants, and temperature than the flows in less impervious areas, which
have more natural vegetation and soil to filter the runoff (U.S. EPA, 1997. Urbanization and
Streams: Studies of Hvdrologic Impacts. EPA 841-R-97-009.  Office of Water. Washington, DC).

       Studies reveal that the level of imperviousness in an area strongly correlates with the
quality of the nearby receiving waters.  For example, a study in the Puget Sound lowland
ecoregion  found that when the level of basin development exceeded 5 percent of the total

                                         IV-4

-------
 impervious area, the biological integrity and physical habitat conditions that are necessary to
 support natural biological diversity and complexity declined precipitously (May, C.W., E.B.
 Welch, R.R. Horner, J.R. Karr, and B.W. May. 1997. Quality Indices for Urbanization Effects in
 Puget Sound Lowland Streams. Technical Report No. 154. University of Washington Water
 Resources Series). Research conducted in numerous geographical areas, concentrating on
 various variables and employing widely different-methods, has revealed a similar conclusion:
 stream degradation occurs at relatively low levels of imperviousness, such as 10 to 20 percent
 (even as low as 5 to 10 percent according to the findings of the Washington study referenced
 above) (Schueler, T.R. 1994. "The Importance of Imperviousness." Watershed Protection
 Techniques  1(3); May, C., R.R. Horner, J.R. Karr, B.W. Mar, and E.B. Welch. 1997. "Effects Of
 Urbanization On Small Streams In The Puget Sound Lowland Ecoregion." Watershed Protection
 Techniques 2(4); Yoder, C.O., R.J. Miltner, and D. White. 1999. "Assessing the Status of
 Aquatic Life Designated Uses in Urban and Suburban Watersheds." In Proceedings: National
 Conference on Retrofits Opportunities in Urban Environments. EPA 625-R-99-002, Washington,
 DC; Yoder, C.O and R.J. Miltner. 1999. "Assessing Biological Quality and Limitations to
 Biological Potential in Urban and Suburban Watersheds in Ohio." In Comprehensive Stormwater
 & Aquatic Ecosystem Management Conference Papers, Auckland, New Zealand). Furthermore,
 research has indicated that few, if any, urban streams can support diverse benthic communities at
 imperviousness levels of 25 percent or more.  An area of medium density single family homes
 can be anywhere from 25 percent to nearly 60 percent impervious, depending on the design of
 the streets and parking (Schueler, 1994).

       In addition to impervious areas, urban development creates new pollution sources as
 population density increases and brings with it proportionately higher levels of car emissions, car
 maintenance wastes, pet waste, litter, pesticides, and household hazardous wastes, which may be
 washed into  receiving waters by storm water or dumped directly into storm drains designed to
 discharge to receiving waters. More people in less space results in a greater concentration of
 pollutants that can be mobilized by, or disposed into, storm water discharges from municipal
 separate storm sewer systems. A modeling system developed for the Chesapeake Bay indicated
 that contamination of the Bay and its  tributaries from runoff is comparable to,  if not greater than,
 contamination from  industrial and sewage sources (Colin-Lee, R.  and D. Cameron. 1992. "Urban
 Stormwater Runoff Contamination of the Chesapeake Bay: Sources and Mitigation." The
 Environmental Professional. Vol. 14).

 2.  Large-Scale Studies and Assessments

       In support of Phase IPs regulatory designation of MS4s in urbanized areas, the Agency
relied on broad-based assessments of urban storm water runoff and related water quality impacts,
 as well as more site-specific studies.  The first national assessment of urban runoff characteristics
was completed for the  Nationwide Urban Runoff Program (NURPI study (U.S. EPA.  1983.
Results of the Nationwide Urban Runoff Program. Volume 1 - Final Report. Office of Water.
Washington, D.C.).  The NURP study is the largest nationwide evaluation of storm water
discharges, which includes adverse impacts and sources, undertaken to date.

       EPA  conducted the NURP study to facilitate understanding of the nature of urban runoff

                                          IV-5

-------
 from residential, commercial, and industrial areas.  One objective of the study was to characterize
 the water quality of discharges from separate storm sewer systems that drain residential,
 commercial, and light industrial (industrial parks) sites. Storm water samples from 81 residential
 and commercial properties in 22 urban/suburban areas nationwide were collected and analyzed
 during the 5-year period between 1978 and 1983. The majority of samples collected in the study
 were analyzed for eight conventional pollutants'and three heavy metals.

       Data collected under the NURP study indicated that discharges from separate storm sewer
 systems draining runoff from residential, commercial, and light industrial areas carried more than
 10 times the annual loadings of total suspended solids (TSS) than discharges from municipal
 sewage treatment plants that provide secondary treatment.  The NURP study also indicated that
 runoff from residential and commercial areas carried somewhat higher annual loadings of
 chemical oxygen demand (COD), total lead, and total copper than effluent from secondary
 treatment plants. Study findings showed that fecal coliform counts in urban runoff typically
 range from tens to hundreds of thousands per hundred milliliters of runoff during warm weather
 conditions, with the median for all sites being around 21,000/100  ml. This is generally
 consistent with studies that found that fecal coliform mean values range from 1,600 coliform
 fecal units (CFU)/100 ml to 250,000 cm/100 ml (Makepeace, O.K.,  D.W. Smith, and S.J.
 Stanley. 1995. "Urban Storm Water Quality: Summary of Contaminant Data." Critical Reviews
 in Environmental Science and Technology 25(2):93-139). Makepeace, et al., summarized ranges
 of contaminants from storm water, including physical contaminants  such as total solids (76 -
 36,200 mg/L) and copper (up to 1.41 mg/L); organic chemicals; organic compounds, such as oil
 and grease (up to 110 mg/L); and microorganisms.

       Monitoring data summarized in the NURP study provided important information about
 urban runoff from residential, commercial, and light industrial areas. The study concluded that
 the quality of urban runoff can be affected adversely by several sources of pollution that were not
 directly evaluated in the study, including illicit discharges, construction site runoff, and illegal
 dumping.  Data from the NURP study were analyzed further in the U.S. Geological Survey
 (USGS) Urban Storm Water Data Base for 22 Metropolitan Areas Throughout the United States
 study (Driver, N.E., M.H. Mustard, R.B. Rhinesmith, and R.F. Middleburg. 1985. U.S.
 Geological Survey Urban Storm Water Data Base for 22 Metropolitan Areas Throughout the
 United States. Report No. 85-337 USGS. Lakewood, CO).  The USGS report summarized
 additional monitoring data compiled during the mid-1980s, covering 717 storm events at 99 sites
 in 22 metropolitan areas and documented problems associated  with metals and sediment
 concentrations in urban storm water runoff. More recent reports have confirmed the pollutant
 concentration data collected in the NURP study (Marsalek, J. 1990. "Evaluation of Pollutant
 Loads from Urban Nonpoint Sources." Wat. Sci. Tech. 22(10/11):23-30; Makepeace, et al.,
 1995).4
        EPA notes that it is not relying solely on the NURP study to describe current water quality
impairment. Rather, EPA is citing NURP as a source of data on typical pollutant concentrations in urban
runoff. Recent studies have not found significantly different pollutant concentrations in urban runoff
compared to the original NURP data (see Makepeace, et al., 1995; Marsalek, 1990; and Pitt, et al., 1995).

                                          IV-6

-------
        America's Clean Water - the States' Nonpoint Source Assessment (Association of State
 and Interstate Water Pollution Control Administrators (ASIWPCA). 1985. America's Clean
 Water - The States' Nonpoint Source Assessment. Prepared in cooperation with the U.S. EPA,
 Office of Water, Washington, DC), a comprehensive study of diffuse pollution sources
 conducted under the sponsorship of the Association of State and Interstate Water Pollution
 Control Administrators (ASIWPCA) and EPA revealed that 38 States reported urban runoff as a
 major cause of designated beneficial use impairment and 21 States reported storm water runoff
 from construction sites as a major cause of beneficial use impairment. In addition, the  1996 -
 305(b) Report (U.S. EPA. 1998. The National Water Quality  Inventory.  1996 Report to
 Congress. EPA 841-R-97-008. Office of Water. Washington,  DC), provides a national
 assessment of water quality based on biennial reports submitted by the States as required under
 CWA section 305(b) of the CWA.  In the CWA 305(b) reports, States, Tribes, and Territories
 assess their individual water quality control programs by examining the attainment or
 nonattainment of the designated uses assigned to their rivers,  lakes, estuaries, wetlands, and
 ocean shores.  A designated use is the legally applicable use specified in a water quality standard
 for a watershed, waterbody, or  segment of a waterbody. The  designated use is the desirable use
 that the water quality should support.  Examples of designated uses include drinking water
 supply, primary contact recreation (swimming), and aquatic life support.  Each CWA 305(b)
 report indicates the assessed fraction of a State's waters that are fully supporting, partially
 supporting, or not supporting designated beneficial uses.

       In their reports, States, Tribes, and Territories first identified and then assigned the
 sources of water quality impairment for each impaired waterbody using the following categories:
 industrial, municipal sewage, combined  sewer overflows, urban runoff/storm sewers,
 agricultural, silvicultural, construction, resource extraction, land disposal, hydrologic
 modification, and habitat modification. The 1996 Inventory,  based on a compilation of 60
 individual 305(b) reports submitted by States, Tribes, and Territories, assessed the following
 percentages of total waters nationwide:  19 percent of river and  stream miles; 40 percent of lake,
 pond, and reservoir acres; 72 percent of estuary square miles;  and 6 percent of ocean shoreline
 waters.  The 1996 Inventory indicated that approximately 40 percent of the Nation's assessed
 rivers, lakes, and estuaries are impaired.  Waterbodies deemed as "impaired" are either partially
 supporting designated uses or not supporting designated uses.

       The 1996 Inventory also found urban runoff/discharges from storm sewers to be a major
 source of water quality impairment  nationwide.  Urban runoff/storm sewers were found to  be a
source of pollution in 13 percent of impaired rivers; 21 percent of impaired lakes, ponds, and
reservoirs; and 45 percent of impaired estuaries (second only to industrial discharges). In
addition, urban runoff was found to be the leading cause of ocean impairment for those ocean
miles surveyed.

       In addition, a recent USGS study of urban watersheds  across the United States has
revealed a link between urban development and contamination of local waterbodies. The study
found the highest levels of organic contaminants, known as polycyclic aromatic hydrocarbons
(PAHs) (products of combustion of wood, grass, and fossil fuels), in the reservoirs of urbanized
watersheds (U.S. Geological Survey (USGS). 1998. Research Reveals Link Between

                                          IV-7

-------
Development and Contamination in Urban Watersheds. USGS news release. USGS National
Water-Quality Assessment Program).

       Urban storm water also can contribute significant amounts of toxicants to receiving
waters.  Pitt, et. al. (1993), found heavy metal concentrations in the majority of samples
analyzed. Industrial or commercial areas were likely to be the most significant pollutant source
areas (Pitt, R., R. Field, M. Lalor, M. Brown 1993. "Urban stormwater toxic pollutants:
assessment, sources, and treatability" Water Environment Research. 67(3):260-75).

3. Local and Watershed-Based Studies

       In addition to the large-scale nationwide studies and assessments, a number of local and
watershed-based studies from across the country have documented the detrimental effects of
urban storm water runoff on water quality. A study of urban streams in Milwaukee County,
Wisconsin, found local streams to be highly degraded due primarily to urban runoff, while three
studies in the Atlanta, Georgia, region were characterized as being "the first documentation in the
Southeast of the strong negative relationship between urbanization and stream quality that has
been observed in other ecoregions" (Masterson, J. and R. Bannerman.  1994. "Impacts of Storm
Water Runoff on Urban Streams in Milwaukee County, Wisconsin." Paper presented at National
Symposium on Water Quality: American Water Resources Association; Schueler, T.R.  1997.
"Fish Dynamics in Urban Streams Near Atlanta, Georgia." Technical Note 94. Watershed
Protection Techniques 2(4)).  Several other studies, including those performed in Arizona
(Maricopa County), California (San Jose's Coyote Creek), Massachusetts (Green River),
Virginia (Tuckahoe Creek), and Washington (Puget Sound lowland ecoregion), all had  the same
finding: runoff from urban areas greatly impair stream ecology and the health of aquatic life; the
more heavily developed the area, the more detrimental the effects (Lopes, T. and K. Possum.
1995. "Selected Chemical Characteristics and Acute Toxicity of Urban Stormwater, Streamflow,
and Bed Material, Maricopa County, Arizona." Water Resources Investigations Report  95-4074.
USGS; Pitt, R. 1995. "Effects of Urban Runoff on Aquatic Biota." In Handbook of
Ecotoxicology; Pratt, J. and R. Coler. 1979. "Ecological Effects of Urban Stonnwater Runoff on
Benthic Macro invertebrates Inhabiting the Green River, Massachusetts." Completion Report
Project No. A-094. Water Resources Research Center. University of Massachusetts at Amherst.;
Schueler, T.R. 1997. "Historical Change in a Warmwater Fish Community in an Urbanizing
Watershed." Technical Note 93. Watershed Protection Techniques 2(4); May, C., R. Homer, J.
Karr, B. Mar, and E. Welch. 1997. "Effects Of Urbanization On Small Streams In The Puget
Sound Lowland Ecoregion." Watershed Protection Techniques 2(4)).

       Pitt and others also described the receiving water effects on aquatic organisms associated
with urban runoff (Pitt, R.E. 1995. "Biological Effects of Urban Runoff Discharges" In
Stormwater Runoff and Receiving Systems: Impact, Monitoring, and Assessment, ed. E.E
Herricks, Lewis Publishers; Crunkilton, R., J. Kleist, D. Bierman, J. Ramcheck, and W. DeVita.
1999. "Importance of Toxicity as a Factor Controlling the Distribution of Aquatic Organisms in
an Urban Stream." In Comprehensive Stormwater & Aquatic Ecosystem Management
Conference Papers. Auckland, New Zealand).
                                         IV-8

-------
        In Wisconsin, runoff samples were collected from streets, parking lots, roofs, driveways,
 and lawns.  Source areas were broken up into residential, commercial, and industrial.  Geometric
 mean concentration data for residential areas included total solids of about 500-800 mg/L from
 streets and 600 mg/L from lawns.  Fecal coliform data from residential areas ranged from 34,000
 to 92,000 cfu/100 mL for streets and driveways.  Contaminant concentration data from
 commercial and industrial source areas were lower for total solids and fecal coliform, but higher
 for total zinc (Bannerman, R.T., D.W. Owens, R.B. Dods, and N.J. Homevver. 1993. "Sources of
 Pollutants in Wisconsin Stormwater." Wat. Sci. Tech. 28(3-5):241-59).

        Bannerman, et al. also found that streets contribute higher loads of pollutants to urban
 storm water than any other  residential development source. Two small urban residential
 watersheds were evaluated to determine that lawns and streets are the largest sources of total and
 dissolved phosphorus in the basins (Waschbusch, R.J., W.R. Selbig, and  R.T. Bannerman. 1999.
 "Sources of Phosphorus in Stormwater and Street Dirt from Two Urban Residential Basins In
 Madison, Wisconsin, 1994-95." Water Resources Investigations Report 99-4021. U.S.
 Geological Survey).  A number of other studies have indicated that urban roadways often contain
 significant quantities of metal elements and solids (Sansalone, J.J. and S.G. Buchberger.  1997.
 "Partitioning and First Flush of Metals in Urban Roadway Storm Water." ASCE Journal  of
 Environmental Engineering 123(2); Sansalone, J.J., J.M. Koran, J.A. Smithson,  and S.G.
 Buchberger. 1998. "Physical Characteristics of Urban Roadway Solids Transported During Rain
 Events" ASCE Journal of Environmental Engineering 124(5); Klein, L.A., M. Lang, N. Nash,
 and S.L. Kirschner. 1974. "Sources of Metals in New York City Wastewater" J.  Water Pollution
 Control Federation 46(12):2653-62; Barrett, M.E, R.D. Zuber, E.R. Collins, J.F. Malina,  R.J.
 Charbeneau, and G.H Ward., 1993. "A Review and Evaluation of Literature Pertaining to the
 Quantity and Control of Pollution  from Highway Runoff and Construction." Research Report
 1943-1. Center for Transportation  Research, University of Texas, Austin).

 4. Beach Closings/Advisories
       Urban wet weather flows have been recognized as the primary sources of estuarine
 pollution in coastal communities. Urban storm water runoff, sanitary sewer overflows, and
 combined sewer overflows have become the largest causes of beach closings in the United States
 in the past three years.  Storm water discharges from urban areas not only pose a threat to the
 ecological environment, they also can substantially affect human health. A survey of coastal and
 Great Lakes communities found that more than 1,500 beach closings and advisories where
 attributable  to storm water runoff in 1998 (Natural Resources Defense Council. 1998. "Testing
 the Waters Volume VIII-Has Your Vacation Beach Cleaned Up Its Act?" New York, NY).  Other
 reports also  document public health, shellfish bed, and habitat impacts from storm water runoff,
 including more than 823 beach closings/advisories issued in  1995 and more than 407 beach
 closing/advisories issued in 1996 due to urban runoff (Natural Resources Defense  Council. 1996.
 Testing the Waters Volume VI: Who Knows What You're Getting Into. New York, NY; NRDC.
 1997. Testing the Waters Volume VII: How Does Your Vacation Beach Rate. New York, NY;
 Morton, T. 1997. Draining to the Ocean: The Effects of Stormwater Pollution on Coastal Waters.
 American Oceans Campaign, Santa Monica, CA). The Epidemiological Study of Possible
•Adverse Health Effects of Swimming in Santa Monica Bay (Haile, R.W., et. al. 1996. "An

                                         IV-9

-------
 Epidemiological Study of Possible Adverse Health Effects of Swimming in Santa Monica Bay."
 Final Report prepared for the Santa Monica Bay Restoration Project) concluded that there is a 57
 percent higher rate of illness in swimmers who swim adjacent to storm drains than in swimmers
 who swim more than 400 yards away from storm drains. This and other studies document a
 relationship between gastrointestinal illness in swimmers and water quality, the latter of which
 can be heavily compromised by polluted storm water discharges.
B.  Other Discharges to Municipal Storm Sewers

       In addition to runoff from storm events, municipal separate storm sewer systems may
receive and ultimately discharge other materials introduced into the system. Non-storm water
discharges to storm sewers come from a variety of sources, including:

       - illicit connections and cross connections from industrial, commercial, and sanitary
       sewage sources
       - leaking sanitary sewage systems
       - malfunctioning on-site disposal systems (septic systems)
       - improper disposal of wastes such as used oil, wastewater and litter
       - spills
       - infiltration of ground water contaminated by a variety of sources, including leaking
       underground storage tanks
       - wash waters, lawn irrigation, and other drainage sources.

       Studies have shown that discharges from MS4s often include wastes and wastewater from
non-storm water sources.  Federal regulations (§ 122.26(b)(2)) define an illicit discharge as
"...any discharge to an MS4 that  is not composed entirely of storm water...," with some
exceptions.  These discharges are "illicit" because municipal storm sewer systems are not
designed to accept, process, or discharge such wastes. Sources of illicit discharges include, but
are not limited to,: sanitary wastewater; effluent from septic tanks; car wash, laundry, and  other
industrial wastewaters; improper disposal of auto and household toxics, such as used motor oil
and pesticides;  and spills from roadway and other accidents.

       Illicit discharges enter the system through either direct connections (e.g., wastewater
piping either mistakenly or deliberately connected to the storm drains) or indirect connections
(e.g., infiltration into the MS4 from  cracked sanitary systems, spills collected by drain outlets,
and paint or used oil dumped directly into a drain). The result is untreated discharges that
contribute high levels of pollutants,  including heavy metals, toxics, oil and grease, solvents,
nutrients, viruses and bacteria into receiving waterbodies.  The NURP study, discussed earlier,
found that pollutant levels from illicit discharges were high enough to significantly degrade
receiving water quality and threaten aquatic, wildlife, and human health. The study noted
particular problems with, illicit discharges of sanitary wastes, which can be directly linked  to high
bacterial counts in receiving waters and can be dangerous to public health.

       Because illicit discharges to MS4s can create severe widespread contamination and water

                                          IV-10

-------
 quality problems, several municipalities and urban counties performed studies to identify and
 eliminate such discharges.  In Michigan, the Ann Arbor and Ypsilanti water quality projects
 inspected 660 businesses, homes, and other buildings and identified 14 percent of the buildings
. as having improper storm sewer drain connections.  The program assessment revealed that, on
 average, 60 percent of automobile-related businesses, including service stations, automobile
 dealerships, car washes, body shops, and light industrial facilities, had illicit connections to storm
 sewer drains.  The program assessment also showed that a majority of the illicit discharges to the
 storm sewer system resulted from improper plumbing and connections, which had been approved
 by the municipality when installed (Washtenaw County Statutory Drainage Board. 1987. Huron
 River Pollution Abatement Program).

       In addition, an inspection of urban storm water outfalls draining into Inner Grays,
 Washington, indicated that 32 percent of these outfalls had dry weather flows. Of these flows,
 21 percent were determined to have pollutant levels higher than the pollutant levels expected in
 typical urban storm water runoff characterized in the NURP study (U.S. EPA. 1993.
 Investigation of Inappropriate Pollutant Entries Into Storm Drainage Systems — A User's Guide.
 EPA 600/R-92/238. Office of Research and Development. Washington, DC).  That same
 document reports a study in Toronto, Canada, that found that 59 percent of outfalls from the
 MS4 had dry-weather flows. Chemical tests revealed that 14 percent of these dry-weather flows
 were determined to be grossly polluted.

         Inflows from aging sanitary sewer collection systems are one of the most serious illicit
 discharge-related problems. Sanitary sewer systems frequently develop leaks and cracks,
 resulting in discharges of pollutants to receiving waters through separate  storm sewers. These
 pollutants include sanitary waste and materials from sewer main construction (e.g., asbestos
 cement, brick, cast iron, vitrified clay).  Municipalities have long recognized the reverse problem
 of storm water infiltration into sanitary sewer collection systems; this type of infiltration often
 disrupts the operation of the municipal sewage treatment plant.

       The improper disposal of materials is another illicit discharge-related problem that  can
 result in contaminated discharges from separate storm sewer systems in two ways.  First,
 materials may be disposed of directly in a catch basin or other storm water conveyance.  Second,
materials disposed of on the ground may either drain directly to a storm sewer or be washed into
a storm sewer during a storm event.  Improper disposal of materials to street catch basins and
other storm sewer inlets often occurs when people mistakenly believe that disposal to such areas
is an environmentally sound practice. Part of the confusion may occur because some areas are
served by combined sewer systems,  which are part of the sanitary sewer collection system, and
people assume that materials discharged to a catch basin will reach a municipal sewage treatment
plant. Materials that are commonly  disposed of improperly include used  motor oil; household
toxic materials; radiator fluids; and litter, such as disposable cups, cans, and fast-food packages.
EPA believes that there has been increasing success in addressing these problems through
initiatives such as storm drain stenciling and recycling programs, including household hazardous
waste special collection days.

       Programs that reduce illicit discharges to separate storm sewers have improved water

                                         IV-11

-------
quality in several municipalities.  For example, Michigan's Huron River Pollution Abatement
Program found the elimination of illicit connections caused a measurable improvement in the
water quality of the Washtenaw County storm sewers and the Huron River (Washtenaw County
Statutory Drainage Board, 1987). In addition, an illicit detection and remediation program in
Houston, Texas, has significantly improved the water quality of Buffalo Bayou.  Houston
estimated that illicit flows from 132 sources had a flow rate as high as 500 gal/min.  Sources of
the illicit discharges included broken and plugged sanitary sewer lines, illicit connections from
sanitary lines to storm sewer lines, and floor drain connections (Glanton, T., M.T. Garrett, and B.
Goloby. 1992.  The Illicit Connection: Is It the Problem? Wat. Env. Tech. 4(9):63-8).
                                         IV-12

-------
 V.    RATIONALE FOR USING A NPDES APPROACH
       This section responds to the Appropriations Act's direction to provide a report
containing:

"(4) information that supports the position of the Administrator that the Phase II stormwater
program should be administered as part of the National Pollutant Discharge Elimination System
under section 402 of the Federal Water Pollution Control Act (33 U.S.C. 1342)"

       EPA interprets Clean Water Act section 402(p)(6) as authorizing the Agency to develop a
storm water program for Phase II sources either as part of the existing NPDES permit program or
as a stand alone non-NPDES program such as a self-implementing rule. Although EPA believes
that it has the discretion to not require sources regulated under CWA section 402(p)(6) to be
covered by NPDES permits, the Agency has determined, for the reasons discussed below, that it
is most appropriate to use NPDES permits in implementing the program to address the sources  .
designated for regulation in Phase II. EPA believes that the NPDES program best achieves the
goals of the Phase II rule for the following reasons:

•      Applying an NPDES permit approach to Phase II sources allows for consistent regulation
       between larger MS4s and construction sites regulated under Phase I and  smaller sources
       regulated under Phase II.

•      Use of NPDES permits to regulate Phase II municipalities will allow co-permitting of
       small regulated MS4s with larger MS4s regulated under the existing Phase I storm water
       program.

•      The use of NPDES permits is a familiar regulatory implementation vehicle that is well
       understood by State regulators and potential permittees.

•      NPDES permits provide the flexibility to allow the use of general permits on a watershed
       basis, while also allowing site-specific controls to be developed on a case-by-case basis.

•      NPDES permits allow incorporation by reference of existing State, Tribal and local
       programs.

•      NPDES permit applications and NOIs provide important information to regulatory
       authorities and the public.

•      NPDES permit procedures include beneficial processes for citizen participation and'
       enforcement.

•      NPDES permits are federally enforceable under the CWA.
                                         V-l

-------
 •      NPDES permit coverage provides "permit as a shield" legal protection to the permittee.

 •      NPDES permit coverage provides an established and predictable regulatory regime to
       avoid duplicative regulation under the Resource Conservation and Recovery Act and the
       Comprehensive Emergency Response, Compensation, and Liability Act, due to
       exclusions from regulation for facilities subject to NPDES permits.

       In developing an approach for the Phase II rule, individual members of both the FACA
 Committee and the Storm Water Phase II FACA Subcommittee encouraged EPA to seek
 opportunities to integrate, where possible, the proposed Phase II requirements with existing
 Phase I requirements, thus facilitating a unified and "seamless" storm water discharge control
 program.  EPA believes that using the NPDES framework is the best means of integrating the
 regulation of Phase II sources with the existing storm water program. The NPDES framework is
 already applied to regulated storm water sources and can be extended to the sources to be
 regulated  in Phase II.  This approach facilitates program consistency, public access to
 information, and program oversight.

       Requiring Phase II sources to be covered by NPDES permits would help address the
 consistency problems currently caused by municipal "donut holes." Donut holes are gaps in
 program coverage where a small unregulated MS4 is located next to or within a regulated larger
 MS4 that  is subject to  an NPDES permit under the existing NPDES storm water program. The
 existence  of such "donut holes" creates an equity problem because similar discharges may
 remain unregulated even though they cause or contribute to the same adverse water quality
 impacts.  Using NPDES permits to regulate the unregulated discharges in these areas is intended
 to facilitate the development of a seamless regulatory program for the mitigation and control of
 contaminated storm water discharges in an urbanized area. For example, the Phase II rule would
 allow a newly regulated MS4 to join as a "limited" co-permittee with a regulated MS4 by
 referencing a common storm water management program. Such cooperation should be further
 encouraged by the fact that the minimum control measures to be required in the Phase II rule.for
 regulated small MS4s are very similar to a number of the permit requirements for medium and
 large MS4s under the existing storm water program. The minimum control measures applicable
 to discharges from smaller MS4s under Phase II are described with slightly more generality than
 under the Phase I permit application regulations for larger MS4s, thus enabling maximum
 flexibility for operators of smaller MS4s to optimize efforts to protect water quality.

       The Phase II rule would also apply NPDES permit requirements to construction sites
below 5 acres that are similar to the existing requirements for those 5 acres and above. In
addition, the rule would allow compliance with qualifying local, Tribal, or  State erosion and
sediment controls to meet the erosion and sediment control requirements of the general permits
for storm water discharges associated with construction, both above and below 5 acres.

       Incorporating the CWA section 402(p)(6) program into the NPDES program capitalizes
upon the existing governmental infrastructure for administration of the NPDES program.
Moreover, much of the regulated community already understands  the NPDES program and the
way  it works.

                                         V-2

-------
       Another goal of the NPDES program approach is to provide flexibility in order to
facilitate and promote watershed planning and sensitivity to local conditions. The following are
some of the more significant examples of the flexibility provided by the NPDES approach:

•      NPDES general permits may be used to cover a category of regulated sources on a
       watershed basis or within political boundaries.

•      The NPDES permitting process provides a mechanism for storm water controls tailored
       on a case-by-case basis, where necessary.

•      The NPDES permit requirements of a permittee may be satisfied by another cooperating
       entity.

•      NPDES permits may incorporate the requirements of existing State, Tribal and local
       programs, thereby accommodating State and Tribes seeking to coordinate the storm water
       program with other programs, including those that focus on watershed-based nonpoint
       source regulation.

       NPDES permits generally require an application or a notice of intent to trigger coverage.
This information exchange assures communication between the permitting authority and the
regulated community. This communication is critical in ensuring that the regulated community
is aware of the requirements and the permitting authority is aware of the potential for adverse
impacts to water quality from identifiable locations. The NPDES permitting process includes the
public as a valuable stakeholder and ensures that the public is included and information is made
publicly available.

       Another concern for EPA and several of the individual FACA Subcommittee members
was that the program ensure citizen participation. The NPDES approach ensures opportunities
for citizen participation throughout the permit issuance process, as well as in enforcement
actions. NPDES permits are also federally enforceable under the CWA.

       EPA believes that the use of NPDES permits makes a significant difference in the degree
of compliance with regulations in the storm water program. The Agency does not anticipate that
a self-implementing rule would ensure the degree of public participation needed for the
development, enforcement and revision of the storm water  management program. Citizen suit
enforcement has assisted in focusing attention on adverse water quality impacts on a localized,
public priority basis. Citizens  frequently rely on the NPDES permitting process and the
availability of NOIs to track program implementation and help them enforce regulatory
requirements.

       NPDES permits are also advantageous to the permittee. The NPDES permit informs  the
permittee about the scope of what it is expected do to be in compliance with the Clean Water
Act. As explained more fully in EPA's April 1995 guidance, Policy Statement on Scope of
Discharge Authorization and Shield Associated with NPDES Permits, compliance with an
NPDES permit constitutes compliance with the Clean Water Act (see CWA section 402(k)).  In

                                          V-3

-------
 addition, NPDES permittees are excluded from duplicative regulatory regimes under the
 Resource Conservation and Recovery Act and the Comprehensive Emergency Response,
 Compensation and Liability Act under RCRA's exclusions to the definition of "solid waste" and
 CERCLA's exemption for "federally permitted releases."

       Throughout development of the rule, State representatives sought alternatives to the
 NPDES approach for State implementation of the storm water program for Phase II sources.
 Discussions focused on an approach whereby States could develop an alternative program that
 EPA would approve or disapprove based on identified criteria, including that the alternative non-
 NPDES program would result in "equivalent or better protection of water quality." The State
 representatives, however, were unable to propose or recommend criteria for gauging whether a
 program would provide equivalent protection. EPA also did not receive any suggestions for
 objective, workable criteria in response to the Agency's explicit request for specific criteria (by
 which EPA could objectively judge such programs) in the preamble to the proposed rule.

       EPA also considered suggestions that the Agency authorize Phase II to be implemented
 as a self-implementing rule, which would be a regulation promulgated at the Federal, State, or
 Tribal level to control some or all of the storm water dischargers regulated under the Phase II
 rule. Under this approach, a rule would spell out  the specific requirements for dischargers and
 impose the restrictions and conditions that would otherwise be contained in an NPDES permit. It
 would be effective until modified by EPA, a State, or a Tribe, unlike an NPDES permit which
 cannot exceed a duration of five years. Some stakeholders believed that this approach would
 reduce the burden on the regulated community (e.g., by not requiring permit applications), and
 considerably reduce the amount of additional paperwork, staff time and accounting required to
 administer the proposed permit requirements.

       EPA is sensitive to the interest of some stakeholders in having a streamlined program that
 minimizes the burden associated with permit administration and maximizes opportunities for
 field time spent by regulatory authorities. Key provisions in the Phase II rule would address
 some of these concerns by promoting a streamlined approach to permit issuance by, for example,
 using general permits for coverage of Phase II permittees and allowing the incorporation of
 existing programs. By adopting the NPDES approach rather than a self-implementing rule, the
 Phase II rule also allows for consistent regulation between larger MS4s and construction sites
 regulated under the Phase I rule and smaller sources regulated under Phase II.

      EPA believes  that it is most appropriate to use NPDES permits to implement a program
to address Phase II sources. In addition to the reasons discussed above, NPDES permits provide
a better mechanism than would a self-implementing rule for tailoring storm water controls on a
case-by-case basis, where necessary. A self-implementing rule would not ensure the degree of
public participation that the NPDES permit process provides for the development, enforcement
and revision of the storm water management program. A self-implementing rule also might not
have provided the regulated community the "permit shield" under CWA section 402(k) that is
provided by an NPDES permit. Based on all these considerations, EPA declined to adopt a self-
implementing rule approach and adopted the NPDES approach for Phase II sources.
                                          V-4

-------