United States
Environmental Protection
Agency
Office oF water
(4203)
EPA 833-D-96-001
September 1996
xvEPA
Interim Permitting Approach For
Water Quality-Based Effluent
Limitations In Storm Water Permits
-------
I It I I I nil I III illllll HIM
ill ', I1
:!i' ;i
1 ii
i III1 I 'l1 1 II
111 i
Ml 111 I I 1
l| 111 • I ill 111 * I 111
>l,i", ..... Hill ..... Illlli1
1 1 1,1 III till I III
..... LI ...... ill,! ..... i
-------
INTERIM PERMITTING APPROACH FOR WATER QUALITY-BASED
EFFLUENT LIMITATIONS IN STORM WATER PERMITS
In response to recent questions regarding the type of water quality-based effluent
limitations that are most appropriate for National Pollutant Discharge Elimination System
(NPDES) storm water permits, the Environmental Protection Agency (EPA) is adopting an
interim permitting approach for regulating wet weather storm water discharges. Due to the
nature of storm water discharges, and the typical lack of information on which to base numeric
water quality-based effluent limitations (expressed as concentration and mass), EPA will use an
interim permitting approach for NPDES storm water permits.
The interim permitting approach uses best management practices (BMPs) in first-round
storm water permits, and expanded or better-tailored BMPs in subsequent permits, where
necessary, to provide for the attainment of water quality standards, In cases where adequate
information exists to develop more specific conditions or limitations to meet water quality
standards, these conditions or limitations are to be incorporated into storm water permits, as
necessary and appropriate. This interim permitting approach is not intended to affect those storm
water permits that already include appropriately derived numeric water quality-based effluent
limitations. Since the interim permitting approach only addresses water quality-based effluent
limitations, it also does not affect technology-based effluent limitations, such as those based on
effluent limitations guidelines or developed using best professional judgement, that are
incorporated into storm water permits.
Each storm water permit should include a coordinated and cost-effective monitoring
program to gather necessary information to determine the extent to which the permit provides for
attainment of applicable water quality standards and to determine the appropriate conditions or
limitations for subsequent permits. Such a monitoring program may include ambient monitoring,
receiving water assessment, discharge monitoring (as needed), or a combination of monitoring
procedures designed to gather necessary information.
This interim permitting approach applies only to EPA; however, EPA also encourages
authorized States and Tribes to adopt similar policies for storm water permits. This interim
permitting approach provides time, where necessary, to more fully assess the range of issues and
possible options for the conltrol of storm water discharges for the protection of water quality.
This interim permitting approach may be modified as a result of the ongoing Urban Wet Weather
Flows Federal Advisory Committee policy dialogue on this subject.
-------
-------
Qs & As FOR INTERIM PERMITTING APPROACH FOR WATER QUALITY-BASED
EFFLUENT LIMITATIONS IN STORM WATER PERMITS
Question 1: Must EPA require that storm water- dischargers,
industrial or municipal, be subject to numeric water quality-
based effluent limitations (expressed as concentration and mass)
in order to attain water quality standards (WQS)?
Answer 1: No. Although National Pollutant Discharge
Elimination System (NPDES) permits must contain conditions
to ensure that water quality standards are met/ this does
not require the use of numeric water quality-based effluent
limitations. Under the Clean Water Act (CWA) "and NPDES
regulations, permitting authorities may employ a variety of
conditions and limitations in storm water permits, including
best management practices, performance objectives, narrative
conditions, monitoring triggers, action levels (e.g.,
monitoring benchmarks, toxicity reduction evaluation action
levels), etc., as the necessary water quality-based
limitations,-where numeric water quality-based effluent
limitations are determined to be unnecessary or infeasible.
Analysis:
A. The Clean Water Act does not require numeric -
effluent limitations.
Section 301 of the CWA requires that dis-charger permits
include effluent limitations necessary to meet State or
Tribal WQS. Section 502 defines "effluent limitation" to
mean any restriction on quantities, rates, and
concentrations of constituents discharged from point
sources. The CWA does not say that effluent limitations
need be numeric. As a result, EPA and States have
flexibility in terms of how to express effluent limitations.
B. EPA's regulations do not always require numeric
e f fluent 1imitations.
EPA has, through regulation, interpreted the statute
to allow for non-numeric limitations {e.g., "best management
practices" or BMPs, see 40 CFR 122.2) to supplement or
replace numeric limitations in specific instances that meet
the criteria specified at 40 CFR 122.44(k), This regulation
essentially codifies a court case addressing storm water
discharges. NRDC v. Costle, 568 F.2d 1369 {D.C. Cir. 1977).
In that case, the Court stated that EPA need not establish
numeric effluent limitations where such limitations were
infeasible.
-------
•C. EPA has interpreted the statute and regulations to
allow BMPs in lieu of numeric limitations.
EPA has defended use of BMPs as a substitute for
numeric limitations in litigation involving storm water
discharges (CBE v. EPA, 91-70056 (9th Cir.)(brief on
merits)) and in correspondence (Letter from Michael Coofc,
EPA, to Peter Lehner, NRDC, May 31, 1995). EPA has found
that numeric limitations for storm water permits can be very
difficult to develop at this time because of the existing
state of knowledge about the intermittent and variable
' nature of these types of discharges and their effects on
receiving waters. Some storm water permits, however,
currently do contain numeric water quality-based effluent
limitations where adequate information exists to derive such
limitations.
Question 2; Has EPA provided guidance on a methodology for
deriving numeric water quality-based effluent limitations?
Answer 2: Yes, but primarily for continuous wastewater
discharges at low flow conditions in the receiving water,
not intermittent wet weather discharges during high flow
conditions. Regulations at 40 CFR 122.44(d) specify the
requirements under which permitting authorities establish
water quality-based effluent limitations when a facility has
the "reasonable potential" to cause or contribute to an
excursion of numeric or narrative water quality criteria.
In addition, EPA guidance in the Technical Support Document
for Mater Quality-Based Toxics Control (TSD) and the NPDES
Permit Writers Training Manual, supplemented with total
maximum daily load (TMDL) and modeling guidance, supports
issuing permits that include numeric water quality-based
effluent limitations. This guidance was based on crafting
numeric water quality-based effluent limitations using
TMDLs, or calculations similar to those used in developing
TMDLs, and wasteload allocations (WLAs) derived through
modeling.' EPA expects the Urban Wet Weather Flows Federal
Advisory Committee (60 FR 21189, May 1, 1995) will review
this issue at greater length and may provide recommendations
on how to proceed.
Question 3: Why can numeric water quality-based cfflvsent
limitations be difficult to derive for storm water permits?
Answer 3: Storm water discharges are highly variable both
in terms of flow and pollutant concentrations, and the
relationships between discharges and water quality can be
complex. The water quality impacts of storm water
discharges are related to the uses designated by States and
-------
Tribes in their MJQS, the quality of the storm water
discharge (e.g., conventional,,or toxic pollutants conveyed
to the receiving water) and quantity of the storm water
(eg, erosion and loss of habitat caused by increased flows
and velocity)- Uses may be impacted by both water quality
and water quantity. Depending on site-specific
considerations, some of the water quality impacts of storm
water discharges may be more related to the physical effects
(e.g. stream bank erosion, streambed scouring, extreme
temperature variations, sediment smothering) than the type
and amount of pollutants present in the discharge. For
municipal storm water discharges in particular, the current
use of system-wide permits and a variety of jurisdiction-
wide BMPs, including educational and programmatic BMPs, does
not easily lend itself to the existing methodologies for
deriving numeric water quality-based effluent limitations.
These methodologies were designed primarily for process
wastewater discharges which occur at predictable rates with
predictable pollutant loadings under low flow conditions in
receiving waters. Using these methodologies, limitations
are typically derived for each specific outfall to be
protective of low flows in the receiving water. Because of
this, permit writers have not made wide-spread use of the
existing methodologies and models for storm water discharge
permits. In addition, wet weather modeling is technically
more difficult and expensive than the simple dilution models
generally used in the permitting process.
Question 4: Has EPA previously recognized the technical
difficulty in deriving numeric water quality-based effluent
limitations for storm water discharges?
Answer 4: Yes. EPA recognized the technical difficulty in
deriving numeric water quality-based effluent limitations
for wet weather discharges in its brief on the merits in
Citizens for a Better Environment (CBS) v. United States
Environmental Protection Agency. 91-70056 (9th Cir.) and in
the Great Lakes Water Quality Guidance (58 FR 20841, April
16, 1993).
In the CBE case, EPA explained why it was technically
infeasible to 'derive numeric water quality-based effluent
limitations for the discharge of metals in storm water into
South San Francisco Bay and asserted that a water quality-
based effluent limitation could take the form of a narrative
statement, such as a BMP, if it. was infeasible to derive a
numeric limitation. In explaining its arguments in the CBE
case, EPA cited 40 CFR 122.44 (k.) (2), which provides that
BMPs may be imposed in NPDES permits "to control or abate
the discharge of pollutants when •. . . (2) [n]umeric effluent
limitations are infeasible."
-------
In the Great Lakes Water Quality Guidance, EPA did not
the method for calculating wasteload allocations, the
ba for numeric water quality-based effluent limitations,
to storm water or combined sewer overflow (CSO) discharges
because the varying nature of these discharges is
inconsistent with the assumptions used in developing the
outdance The Great Lakes Water Quality Guidance defers to
national guidance and policy on wet weather and does not
Seek ?o establish a separate and distinct set of wet weather
requirements. EPA expects the Urban Wet Weather Flows.
Advisory Committee to provide recommendations about how to
address the broader'technical issues involved in achieving
compliance with WQS in a wet weather context.
Question 5: What are the potential problems of using standard
methodologies to derive numeric water quality-based effluent
limitations for storm water permits?
Answer 5: Correctly derived numeric water quality-based
effluent limitations provide a greater degree of confidence
that a discharge will not "cause or- contribute to an
exceedance of the WQS, because numeric water quality-based
effluent limitations are derived directly from the numeric
component of those standards. In addition, numeric water
• quality-based effluent limitations can avoid the expense
associated with overly protective treatment technologies
because numeric water quality-based effluent limitations
provide a more precisely guantifed target for permittees.
Potential problems of incorporating inappropriate numeric
water quality-based effluent limitations rather than BMPs in
storm water permits at this time are significant in some
cases. Deriving numeric water quality-based effluent
limitations for any NPDES permit without an adequate
effluent characterisation, or an adequate receiving water
exposure assessment (which could include the use of dynamic
modeling or continuous simulations) may result in the
imposition of inappropriate numeric limitations on a
discharge. Examples of this include the imposition of
numeric water quality criteria as end-of-pipe limitations
without properly accounting for the receiving water
assimilation of the pollutant or failure to account for a
mixing zone (if allowed by applicable State or Tribal WQS) .
This could lead to overly stringent permit requirements, and
excessive and expensive controls on storm water discharges,
not necessary to provide for attainment of WQS- Conversely,
an inadequate effluent characterization could lead to water
quality-based effluent limitations that are not stringent
enough to provide for attainment of WQS. This could result
because effluent characterization and exposure assessments
for discharges with high variability of pollutant
-------
concentrations, loadings, and flow are more difficult than
with process wastewater discharges at low flows.
Question 6: How are water quality-based effluent limitations
developed for combined sewer overflow (CSO) discharges?
Answer 6: The CSO Control Policy issued by EPA on April 19,
1994 (59 FR 18688) provides direction on compliance with the
technology-based and water quality-based requirements of the
CWA for communities with combined sewer systems. The CSO
Policy provides for implementation of technology-based
requirements (expressed as "nine minimum controls") by
January 1, 1997.
In addition, under the CSO Policy, communities are also
expected to develop long-term control plans that will
provide for ^attainment of WQS through either the
^presumption approach" or the "demonstration approach."
Under the presumption approach, CSO controls would be
presumed to.attain WQS if certain performance criteria are
met. A program that meets the criteria specified in the CSO
policy is presumed to provide an adequate level of control
to meet the water quality-based requirements of the CWA,
provided the permitting authority determines that such
presumption is reasonable based on characteriEation,
monitoring, and modeling of the system, including
consideration of sensitive areas. Under the demonstration
approach, the permittee would demonstrate that the selected
CSO controls, when implemented, would be adequate to meet
the water quality-based requirements of the CWA-
The CSO Policy anticipates that it will be difficult in
the early stages of permitting to determine whether numeric
water quality-based effluent limitations are necessary for
CSOs, and, if so, what the limitations should be. For that
reason, in the- absence of sufficient data to evaluate the ,
need for numeric water quality-based effluent limitations,.
the Policy recommends that the first phase of CSO permits
("Phase I") contain a narrative requirement to comply with
WQS. Further, so-called "Phase II" permits would contain
water quality-based effluent limitations, as provided in 40
CFR 122.44(d) {!)• arid 122.44(k), that may take the form of
numeric performance or design standards, such as a certain
number of overflow events or a certain percent volume
capture. Generally, only after the long-term control plan
' is in place and after collection of sufficient water quality
data (including applicable wasteload allocations developed
during a TMDL process) would numeric water quality-based
effluent limitations be included in the permit- This would
likely occur only after several permitting cycles.
-------
Question 7: If BMPs alone are demonstrated to provide adequate
water quality protection, are additional controls necessary?
Answer 7: No. If the permitting authority determines that,
through implementation of appropriate BMPs required by the
NPDES storm water permit, the discharges have the necessary
controls to provide for attainment of WQS and any
technology-based requirements, additional controls need not
be included in the permit. Conversely, if a discharger
(municipal or industrial) fails or refuses to adopt and
implement adequate BMPs , the permitting authority may have
to consider other approaches to ensure water quality
protection.
If, however, the permitting authority has adequate
information on which to base more specific conditions or
limitations, such limitations are to be incorporated, into
storm water permits, as necessary and appropriate. Such
conditions or limitations may include an integrated suite of
BMPs/ performance objectives, narrative standards,
monitoring triggers, numeric water quality-based effluent
limitations, action levels, etc. Storm water permits may
also need to include additional requirements to receive
State or Tribal 401 certifications.
Question 8: What is EPA doing to develop information about the
linkage between BMPs and water quality and to facilitate a
watershed-based approach to storm water permitting?
Answer 8i The Agency has cooperative agreements with WERF
(Water Environment Research Foundation) and ASCE {American
Society of" Civil Engineers) to research which BMPs are most
effective under which .circumstances. The results of this
research should provide permitting authorities and .
permittees with information about how to evaluate the
effectiveness of different kinds of BMPs in different
circumstances and to select the most appropriate controls to
achieve water quality objectives. EPA also has cooperative
agreements with the Watershed Management Institute and other
organizations to conduct re'search over the next two to four
years that will examine the capability of storm water BMPs
to improve redeiving water quality and restore/protect the
bio'logical integrity of those waters. EPA expects the Urban
Wet Weather Flows Federal Advisory Committee to provide
recommendations on how'to permit storm water discharges on a
watershed basis.
Question 9: The interim permitting approach states that permits
should include monitoring programs to generate necessary
-------
information to determine the extent to which permits are
providing for the attainment of water quality standards. What
types of monitoring should be included and how much monitoring is
necessary?
Answer 9: The amount and types of monitoring necessary will
vary depending on the individual qircumstances of each storm
water discharge. EPA encourages dischargers and permitting
authorities to carefully evaluate monitoring needs and storm
water program objectives so as to select useful and cost-
effective monitoring approaches- For most dischargers,
storm water monitoring can be conducted for two basic
reasons: 1) to identify if problems are present/ either in
the receiving water or in the discharge, and to characterize
the cause (s) of such problems; and 2) to assess the
effectiveness of storm water controls in reducing
contaminants and making improvements in water quality.
Under the NPDES storm water program/ large and medium
municipal separate storm sewer system permittees are
required to conduct monitoring. EPA recommends that each
such municipal permittee design the monitoring effort to be
supportive of the goals and objectives of its storm water
management program when developing such a program for the
term of its NPDES permit. To accomplish this, a municipal
permittee may use a variety of storm water monitoring tools
including receiving water chemistry; receiving water
biological assessments (benthic invertebrate surveys, fish
surveys, habitat assessments, etc.); effluent monitoring/-
including chemical, whole effluent and visual examinations;
illicit connections screening; and combinations thereof, or
other methods. Techniques that assess receiving waters will
help to identify the degree to which storm water discharges
are contributing to any water quality problems. Techniques
that assess storm water discharge characteristics will help
to identify potential causes of any identified water quality
problems. The municipal permittee, in conjunction with the
applicable NPDES permitting authority, should determine
which monitoring approaches would be most appropriate given
the objectives of the storm water management program. If
municipal permittees conduct ambient monitoring, it may be
most cost-effective to pool resources with other
organizations (including, for example, other municipalities,
States, and Tribes) conducting monitoring within the same
watershed. This could be best accomplished through a
coordinated watershed monitoring strategy.
For industrial storm water dischargers, monitoring may be
required under the terms of an NPDES permit for storm water
discharges. For those industrial storm water permits that
do require monitoring, this is typically done to
characterize contaminants that might be found in the
-------
industrial runoff and/or to assess the effectiveness of the
industrial- storm water pollution prevention plan in reducing
these contaminants. This typically involves end-of-pipe
chemical-specific monitoring. End-of-pipe monitoring may- be
more appropriate for an industrial facility than for a
municipal permittee/ given the industrial facility's-more
discrete site characteristics, which make management
strategies such as collection and treatment more feasible.
Industries, for the most part, have readily defined storm
water conveyances into which runoff flows from discrete
drainage areas. Industries may more readily identify and
control existing on-site sources of storm water
contamination or provide collection and treatment within
these discrete drainage areas to control pollutant
concentrations in their storm water discharges.
EPA and other organizations are currently working to improve
approaches for monitoring storm water and the potential
effects upon water equality. These new approaches are called
storm water program "environmental indicators."
Environmental indicators are designed to be more meaningful
monitoring tools that storm water dischargers can use to
conduct storm water monitoring for the purposes described
above. A manual describing each of the recommended storm
water "program environmental indicators is being prepared by
the Center for Watershed Protection in Silver Spring,
Maryland. That manual is expected to be ready by the end of
August 1996 and should provide useful information for storm
water dischargers contemplating the need to develop a cost-
effective, meaningful storm water monitoring program. In
addition, EPA expects the Urban Wet Weather Flows Federal
Advisory Committee to provide recommendations on how to
better monitor storm water and other wet weather discharges
using a watershed approach.
Question 10: Does this interim permitting approach apply to both
storm water discharges associated with industrial activity and
storm water discharges from municipal separate storm sewer
systems?
Answer 1O: Yes. The interim permitting approach is
applicable to both discharges from municipal separate storm
sewer systems and storm water discharges associated with
industrial activity ( as defined by 40 CFR 122.26(b) (14)).
The interim permitting approach would not affect, however,
permits that already incorporate appropriately derived -
numeric water quality-based effluent limitations. Since the
interim permitting approach only .addresses water
quality-based effluent limitations, it also does not affect
technology-based effluent limitations, such'as those based
on effluent limitations guidelines or developed using best
-------
.professional judgement,, that are incorporated into storm
water permits. In addition/ particularly for some .
industries/ adequate information may already have been
collected with which to assess the reasonable potential for
a storm water discharge to cause or contribute to an
excursion of a WQS, and from which a numeric water quality-
based effluent limitation can be (or has been) appropriately
derived. An adequate amount of storm water pollutant source
information may also exist with which to assess the
effectiveness of the industrial storm water control measures
in 'complying with the limitations and in reducing storm
water contaminants for protecting water quality.
-------
------- |