Click here for
                             DISCLAIMER

                      Document starts on next page
TITLE: Technical Guidance Manual for Performing Wasteload Allocations,
       Book VI: Design Conditions-
       Chapter 1: Stream Design Flow for Steady-State Modeling

EPA DOCUMENT NUMBER: EPA440/4/86-014      DATE: 1986

ABSTRACT

As part of ongoing efforts to keep EPA's technical guidance readily accessible to
water quality practitioners, selected publications on Water Quality Modeling and
TMDL Guidance available at http://www.epa.gov/waterscience/pc/watqual.html
have been enhanced for easier access.

This manual describes two techniques to estimate the stream design flow at or
above which EPA's two-number aquatic life protection criteria must be met. The
two-number water quality criteria are:  (1) a criteria maximum concentration
(CMC), and (2) a continuous criteria concentration (CCC). Generally, EPA's CMC
represents an acute pollutant concentration that should be exceeded no more
than once in an average of three years for a period of an hour. Similarly, the CCC
represents a chronic pollutant concentration that should be exceeded no more
than once in an average of three years for a continuous period of four days.
Historically, a majority of States  in the USA have required that EPA's aquatic life
protection criteria must be met at all flows that are equal to or greater than a
critical flow  condition of 7-day 10-year low flow (7Q10).

In a steady-state modeling framework, the critical flow conditions can be used to
estimate the maximum amount of any pollutant that can be discharged without
violating water quality  criteria (WQC).  The critical low flow that is used to define
the maximum amount of any pollutant that can be discharged without violating
ambient water quality standards (WQS) is called the stream design flow. The
design flow represents the required level of treatment or the size of the
wastewater treatment  facility.

This manual recommends a biologically-based stream design flow estimating
technique that is strictly consistent with the duration and frequency criteria of
EPA's two-number WQC. The biologically-based stream design flows 1B3 and
4B3 are synonymous with the 1-hour,  3-year and 4-day, 3-year duration and
frequency criteria of CMC and CCC, respectively. The second estimating
technique included in this manual is for a hydrologically-based stream design
flow represented in the xQy format, such as 7Q10, 1Q10, 30Q10, etc.

This hydrologic flow estimation is consistent with the  log Pearson Type  III
frequency curve approach described in USGS Surface Water Branch Technical
Memorandum NO. 79.06, "PROGRAMS AND PLANS - Low-Flow Programs",

-------
available online at http://water.usgs.gov/admin/memo/SW/sw79.06.html. These
two estimating techniques are implemented using EPA's DFLOW computer
program, which is available online at http://www.epa.gov/waterscience/dflow/.
Version 3.1 of DFLOW includes a graphical user interface, and directly
incorporates the USGS implementation of the log Pearson Type III frequency
curve approach and EPA's biologically-based stream design flow technique. The
guidance manual includes a comparison of stream design flows of 60 randomly
selected US streams, calculated using both approaches.

KEYWORDS: Wasteload Allocations, Design Conditions, Design Flow, Steady-
             State, Models, Water Quality Criteria, Acute, Chronic, Biological,
             Hydrologic

-------
          TECHNICAL GUIDANCE MANUAL
                     FOR
       PERFORMING WASTELOAD ALLOCATION
                   Book VI


              Design Conditions


                  Chapter 1




Stream Design Flow  for  Steady-State Modeling
               September  1986
         EPA Publication:  440/4-86-014

-------
Transmittal

                                                      OFFICE OF WATER

                          September  29,  1986
MEMORANDUM


Subject: Technical Guidance Manual for Performing Wasteload
         Allocations Book VI, Design Conditions: Chapter 1 - Stream
         Design Flow for Steady-State Modeling

From:    William A.  Whittington, Director
         Office of Water Regulations and Standards  (WH-551)

To:      Addressees
      Attached, for national use, is the final version of the
Technical Guidance Manual for Performing Wasteload Allocations, Book
VI, Design Conditions: Chapter 1 - Stream Design Flow for Steady-
State Modeling. This manual replaces the interim stream design flow
recommendations included in Appendix D of our Technical Support
Document for Water Quality-based Toxics Control, September 1985.  We
are sending extra copies of this manual to the Regional Waste Load
Allocation Coordinators for distribution to the States to use in
conducting waste load allocations.

      If you have any questions or desire additional information
please contact Tim.  S.  Stuart, Chief, Monitoring Branch,
Monitoring and Data Support Division (WH-553)  on (FTS)  382-7074.
Attachment

Addressees:

Regional Water Management Division Directors
Regional Environmental Services Division Directors
Regional Wasteload Allocation Coordinators

-------
                           ACKNOWLEDGEMENT
The contents of this section have been removed to comply with
current EPA practice.

-------
                          Table of Contents


SECTION 1.  INTRODUCTION 	 1-1
  1.1  Purpose	1-1
  1. 2  Background	1-1
  1. 3  Scope	1-3

SECTION 2.  HYDROLOGICALLY-BASED DESIGN FLOW 	 2-1
  2 .1  Introduction	2-1
  2 .2  Rationale	2-2
  2 . 3  Example Cases	2-2
       Table  2-1.   Hydrologically-based  design flows  (ft3/sec)  for
       60  streams  	2-3

SECTION 3.  BIOLOGICALLY-BASED DESIGN FLOW 	 3-1
  3 .1  Introduction	3-1
       3.1.1  Exceedances  and  Excursions  	 3-2
       3.1.2  Features  of  Calculation  	3-4
       Figure 3-1:  Illustration  of  biologically-based
       design flow  	  3-5
  3 . 2  Procedure	3-6
  3 .3  Rationale	3-7
  3 .4  Example Cases	3-8
       Table  3-1.   Biologically-based  design  flows  (ft3/sec)
       for 63 rivers  	3-9

SECTION 4 . COMPARISON OF THE TWO METHODS	4-1
  4.1  Design Flows	4-1
       Table  4-1.   Comparison of 1Q10  and 7Q10 with 1-day 3-yr and
       4-day  3-yr  low  flows  (all flows in ft3/sec)	4-2
  4 . 2  Excursions	4-3
  4.3  Comparison of the Two Methods	4-4
       Table  4-2.  Comparison  of  number of excursions  of  1Q10  and
       7Q10 with number of excursions  of  1-day 3-yr and  4-day 3-yr
       design flows	4-5

SECTION 5.  RECOMMENDATIONS 	 5-1

SECTION 6 .  REFERENCES	6-1

APPENDIX A. CALCULATION OF HYDROLOGICALLY-BASED DESIGN FLOWS ... A-l

-------
APPENDIX B.  AN EXAMPLE USE OF DFLOW FOR AMMONIA  DISCHARGES  FROM
POTWs	B-l

APPENDIX C. CALCULATION OF BIOLOGICALLY-BASED  DESIGN FLOWS 	 C-l

APPENDIX D. DFLOW 2.0 USER'S GUIDE  (OMITTED)

APPENDIX E. QUESTIONS AND ANSWERS CONCERNING THE  BIOLOGICALLY-BASED
METHOD	E-l

-------
                                 GLOSSARY
IBS - 1-hour 3-year duration  and  frequency criteria of Criterion Maximum
Concentration.
4B3 - 4-day 3-year duration and frequency criteria of Criterion
Continuous Concentration.
7Q10 - EPA's aquatic life protection  criteria to meet all flows equal to
or greater than critical flow condition  of 7-day 10-year low flow.
xQy - Hydrologically-based design flows  expressed as x-day average low
flows whose return period is  y years.  For example, 1Q4 is the daily low
flow that is exceeded once every  four years.  Other xQy values commonly
encountered are 1Q10, 4Q3, 7Q5, and 7Q10.
xBy - Biological design flow  parameter defined in DFLOW 3.
Averaging period - Specified  in national water quality criteria are one
hour for the CMC and four days for the CCC.   The primary use of the
averaging periods in criteria is  for  averaging ambient concentrations of
pollutants in receiving waters in order  that  the averages can be
compared to the CMC and CCC to identify  "exceedances" i.e.,  one-hour
average concentrations that exceed the CMC and four-day average
concentrations that exceed the CCC.
Biologically-based stream design  flow -  Design flow based on the
averaging periods and frequencies specified in water quality criteria
for individual pollutants and whole effluents.

Criteria - Descriptive factors taken  into account by EPA in setting
standards for various pollutants.  These  factors are used to determine
limits on allowable concentration levels,  and to limit the number of
violations per year. When issued  by EPA,  the  criteria provide guidance
to the states on how to establish their  standards.1

Criterion Continuous Concentration (CCC)  - The CCC is the 4-day average
concentration of a pollutant  in ambient  water that should not be
exceeded more than once every three years on  the average.

Criterion Maximum Concentration  (CMC)  -  The one-hour average
concentration in ambient water should not exceed the CMC more than once
every three years on the average.
Design flow - A flow rate resulting from an applied waste load
allocation process, used to ensure ambient water quality compliance.

DFLOW - The DFLOW  (Design FLOW) program  was originally developed by EPA
to support design flow analysis as described  in this manual. At the
time of original publication,  Version 2.0 of  DFLOW was available to
support this analysis on personal computers.  The latest version of
DFLOW at the time of enhancement  of this manual was Version 3.1, which
combines a graphical user interface with the  USGS implementation of the
Log Pearson Type III calculation.  Additional  information may be
available at http://www.epa.gov/waterscience/dflow/.

Duration - The time during which  something exists or lasts
 EPA.  "Terms of Environment: Glossary, Abbreviations, and Acronyms"
http://www.epa.gov/OCEPAterms/cterms.html March 8, 2006.
2
 Merriam-Webster Online Dictionary, http://www.m-w.com/dictionary/duration Accessed April
14, 2006.

-------
Dynamic models  -  Preferred for the  application of aquatic  life  criteria
in order to  make  best use of the  specified concentrations, durations,
and frequencies.  Use of aquatic life  criteria for developing  water
quality-based permit limits and for designing waste treatment
facilities requires the selection of  an appropriate wasteload
allocation model.

Effluent - Wastewater--treated or untreated--that flows out of  a
treatment plant,  sewer, or industrial outfall. Generally refers to
wastes discharged into surface waters.3

Exceedance - Violation of the pollutant levels and average
concentration frequencies that are  permitted by environmental
protection standards

Excursion -  An  unfavorable conditions,  e.g.  low flow or noncompliant
pollutant concentrations
Flow Averaging  Period - Monitor of  water flow for a specific  number of
days, usually 30,  to determine pollutant concentrations.

Frequency -  The number of repetitions of a particular event in  a  unit
of time.

Gaging stations - The locations at  which measurements are  recorded,
generally hydraulic or hydrologic in  nature, usually referring  to
stream flow  gages or rain gages.

Harmonic Mean - Set of numbers that is  the reciprocal of the  arithmetic
mean of the  reciprocals of the numbers.
Hydrologic Flow - The characteristic  behaviour and the total  quantity
of water involved in a drainage basin,  determined by measuring  such
quantities as rainfall, surface and subsurface storage and flow,  and
evapotranspiration.(Source: BJGEO)4

Hydrologically-based design flow  -  Design flow calculated  solely  from
the hydrologic  record.
Log Pearson  Type  III - Statistical  analysis, recommended by the US Water
Resources Council Bulletin #17B,  used to gage natural flood data.4
Model - Either  a  steady-state or  dynamic simulation that uses hydraulic
and biological  criteria to design regulatory compliant waste  loads.
Non-exceedance  -  Relating to steady-state models as flow rates  that are
less than design  flows.
Percentile - One  of a set of points on  a scale arrived at  by  dividing  a
group into parts  in order of magnitude.5

Pollutant -  Generally, any substance  introduced into the environment
that adversely  affects the usefulness of a resource or the health of
humans, animals,  or ecosystems.6
  EPA. "Terms of Environment: Glossary, Abbreviations,  and Acronyms"
http://www.epa.gov/OCEPAterms/.  March 8,2006.
  EPA. "Terminology Reference System: Hydrologic Flow"
http://iaspub.epa.gov/trs/trs proc qry.navigate term?p term id=26386&p term cd=TERMDIS
May 2, 2006.
  Texas Department of Transportation. "Hydraulic Design Manual, Section 10. Statistical
Analysis of Stream Gauge Data".
http://manuals.dot.state.tx.us/dynaweb/colbridg/hyd/@Generic  BookTextView/14106;cs=defau
lt;ts=default;pt=14873 March 2004.
  The American Heritage® Dictionary of the English Language,4th Ed. "Percentile."  Accessed
at www.Dictionary.com July 3, 2006.

-------
Receiving waters  - Bodies of water  into  which effluents are discharged.
Return Period  - Annual x-day average  low flow repeated in y-years.
Site-specific  criteria - Regulatory concentrations, parameters,  or
frequencies  that  are exclusive to proximity or spatially exclusive.
Steady-state models - Models that assume a  constant average flow rate.
Toxicity - The degree to which a substance  or mixture of substances  can
harm humans  or animals. Acute toxicity involves harmful effects  in an
organism through  a single or short-term  exposure. Chronic toxicity is
the ability  of a  substance or mixture  of substances to cause harmful
effects over an extended period, usually upon repeated or continuous
exposure sometimes lasting for the  entire life of the exposed organism.6
Two-number water  quality criterion  -  Criterion Continuous and Criterion
Maximum Concentrations (CCC, CMC) that are  used as the basis for the
both hydrologically- and biologically-based design flows
Variability  -  Reference to the consistency  of pollutant concentrations
in effluent  and ambient waters.
Wasteload allocation - 1.) The maximum load of pollutants each
discharger of  waste is allowed to release into a particular waterway.
Discharge limits  are usually required  for each specific water quality
criterion being,  or expected to be, violated. 2.) The portion of a
stream's total assimilative capacity  assigned to an individual
discharge.7

Water quality  criteria (WQC) - Levels  of water quality expected  to
render a body  of  water suitable for its  designated use. Criteria are
based on specific levels of pollutants that would make the water
harmful if used for drinking, swimming,  farming, fish production, or
industrial processes.

Water quality-based permit - A permit  with  an effluent limit more
stringent than one based on technology performance. Such limits  may  be
necessary to protect the designated use  of  receiving waters  (e.g.
recreation,  irrigation, industry or water supply).
 EPA. "Terms of Environment: Glossary,  Abbreviations, and Acronyms".
http://www.epa.gov/OCEPAterms/pterms.html March 8, 2006.
 EPA.  "Terms of Environment:  Glossary, Abbreviations, and Acronyms"
http://www.epa.gov/OCEPAterms/tterms.html March 8, 2006.
  EPA. "Terms of Environment: Glossary, Abbreviations,  and Acronyms"
http://www.epa.gov/OCEPAterms/wterms.html March 8, 2006.
  EPA. "Terms of Environment: Glossary, Abbreviations,  and Acronyms"
http://www.epa.gov/OCEPAterms/wterms.html March 8, 2006.

-------
                       Section  1.   Introduction

1.1   Purpose

     The purpose of this guidance  is to describe and compare two
methods that can be used to calculate stream design flows for any
pollutant or effluent for which a two-number water quality criterion
(WOC)  for the protection of aquatic life is available.   The two
methods described are:
        1. The  hydrologically-based design  flow method  recommended
          for  interim use  in the Technical  Support  document  for
          Water Quality- based Toxics Control  (1);  and
        2. A biologically-based design flow  method that was developed
          by the Office of  Research and Development of the U.S.
          EPA.

1.2  Background

    National water quality  criteria for aquatic  life  (2)  are derived
on the basis of the best available biological,  ecological and
toxicological information concerning the effects of pollutants on
aquatic organisms and their uses (3,4).   To account for local
conditions,  site-specific criteria may he derived whenever
adequately justified  (4).   In addition,  criteria may be derived from
the results  of toxicity tests on whole effluents (1).   National,
site-specific,  and effluent toxicity criteria specify concentrations
of pollutants,  durations of averaging periods,  and frequencies of
allowed exceedances.   If these criteria are to achieve their
intended purpose,  decisions concerning not only their derivation,
but also their use,  must be based on the biological, ecological,  and
toxicological characteristics of aquatic organisms and ecosystems,
and their uses, whenever possible.

    National,  site-specific, and effluent  toxicity  criteria  are
expressed as two concentrations, rather than one, so that the

-------
criteria can more accurately reflect toxicological and practical
realities (1-4) :

     a.   The lower concentration is called the Criterion Continuous
         Concentration (CCC).   The CCC is the 4-day average*
         concentration of a pollutant in ambient water that should
         not be exceeded more than once every three years on the
         average.
     b.   The higher  concentration  is  called  the  Criterion  Maximum
         Concentration (CMC).   The one-hour average concentration in
         ambient water should not exceed the CMC more than once
         every three years on the average.

Use of aquatic life criteria for developing water quality-based
permit limits and for designing waste treatment facilities requires
the selection of an appropriate wasteload allocation model.
Dynamic models are preferred for the application of aquatic life
criteria in order to make best use of the specified concentrations,
durations,  and frequencies  (2).   If none of the dynamic models can
be used, then an alternative is steady-state modeling.  Because
steady-state modeling is based on various simplifying assumptions,
it is less complex, and may be less realistic, than dynamic
modeling.  An important step in the application of steady-state
modeling to stream is the selection of the design flow.
     One way of using the  CCC and the  CMC in  steady-state modeling
requires calculation of the two design flows  (i.e., a CCC design
flow and a CMC design flow).  Whether the CCC and its design flow or
the CMC: and its design flow is more restrictive, and therefore
controlling, must be determined individually for each pollutant of
  Although a 4-day averaging period should be used for the CCC in
  most  situations,  an  averaging  period  as  long as  30  days  may  be  used
  in  situations  involving  POTWs  designed to remove ammonia when  low
  variability of  effluent  pollutant  concentration  and resultant
  concentrations  in receiving waters can be demonstrated.   In  cases
  where low variability  can  be demonstrated,  longer averaging  periods
  for the  ammonia CCC  (e.g.,  a 30-day averaging period)  would  be
  acceptable because the magnitudes  and durations  of  excursions  above
  the CCC  would  be sufficiently  limited (5).

-------
concern in each effluent because the CCC and CMC are pollutant-
specific,  whereas the two design flows are specific to the receiving
waters.

     Wasteload  allocation  modeling  for  stream usually  uses  flow data
obtained from the United States Geological Survey gaging stations.
If sufficient flow data are not available for a stream of interest,
data must be extrapolated from other streams having hydrologic
characteristics similar to those of the stream of interest.
     This  guidance  is  limited  to  (a)  describing  two  methods  that  can
be used for calculating stream design flows for any pollutant or
effluent for which a two-number aquatic life water quality criterion
is available, and  (b)  making recommendations concerning the use of
these methods in steady-state modeling.
     The water quality criterion  for  dissolved oxygen  was  revised
very recently and the assessment of the appropriate design flow for
dissolved oxygen modeling has not yet been completed.   Therefore,
the state-specified design flows that traditionally have been used
for conventional pollutants should not be affected by this guidance.

-------
     State-specified  design  flows  necessarily  preempt  any  design  flow
that is recommended in this guidance unless the state chooses to use
either of these two methods.  The choice of design flows for the
protection of human health has been discussed in the Technical
Support Document for Water Quality-based Toxics Control (1).
     Aquatic  life criteria of  some pollutants  are  affected by
environmental variables such as water temperature, pH, and hardness.
In addition to the design flow, such other stream variables as pH
and temperature might increase or decrease the allowable in-stream
concentrations of some pollutants (e.g., ammonia).  The need to
consider other variables when determining the design flow for those
pollutants should be emphasized.  This document will provide
guidance for the calculation of design flow; pH, temperature, and
hardness will likely be addressed later.

-------
             SECTION 2.  Hydrologically-Based Design Flow

2 .1   Introduction

     The purpose of this section is to describe the hydrologically-
based design flow calculation method and provide some examples of its
use.  The Technical Support Document for Water Quality-based Toxics
Control (1)  provides Agency guidance on control of both generic and
pollutant-specific toxicity and recommended interim use of the
hydrologically-based method.  In addition, the Agency also recommended
 (1, 2)  that the frequencies of allowed exceedances and the durations
of the averaging periods specified in aquatic life criteria should not
be used directly to calculate steady- state, design flows using an
extreme value analysis.  For example, if a criterion specifies that
the four-day average concentration should not exceed a particular
value more than once every three years on the average,  this should be
interpreted as implying that the 4Q3 low flow is appropriate for use
as the design flow.
     Because a procedure had not been developed for calculating design
flow based on the durations and frequencies specified in aquatic life
criteria,  the U.S. EPA recommended interim use of the 1Q5 and 1Q10 low
flows as the CMC design flow and the 7Q5 and 7Q10 low flows as the CCC
design flow for unstressed and stressed systems, respectively (1).
Further consideration of stress placed on aquatic ecosystems resulting
from exceedances of water quality criteria indicates that there is
little justification for different design flows for unstressed and
stressed system.  All ecosystems have been changed as a result of
man's activities.  These changes have resulted in stress being placed
on the ecosystem before a pollutant stress.  In addition, it is not
possible to predict
                                  2-1

-------
the degree of pollutant stress when one considers both the timing and
variability of flows, effluent discharges, and ecosystem sensitivity
and resilience.

2.2  Rationale

     The  following provides  a  rationale  for  the hydrologically-based
design flow calculation method:

   •  About half of the states in the nation use 7Q10 as the design
      low flow.
   •  The log-Pearson Type III flow estimating technique of other
      extreme value analytical techniques that are used to calculate
      flow statistics from daily flow data are consistent with past
      engineering and statistical practice.
   •  Most users are familiar with the log-Pearson Type III flow
      estimating procedure and the USGS provides technical support for
      this technique.
   •  Analyses of 60 rivers indicate that, on the average, the
      biologically-based CMC and CCC design flows are nearly equal to
      the 1Q10 and the 7Q10 low flows.

2.3  Example Cases

     In order  to illustrate the calculation of hydrologically-based
design flows, sixty rivers with flows of various magnitudes and
variabilities were chosen from around the country.  The 1Q10 and 7Q10
low flow of the sixty rivers are presented in Table 2-1.  The list of
rivers in this table is arranged in increasing magnitude of the 7Q10
low flows.  The estimates of the 1Q10 and 7Q10 low flows were made
using the USGS daily flow database and the FLOSTAT program (6) which
employs the log-Pearson Type III technique.
                                  2-2

-------
     The estimates of 1Q10 and 7Q10 low flows could have been made
using EPA-ORD's DFLOW program, which uses a simplified version of the
log-Pearson Type III method.   The simplified version of the log-
Pearson Type III estimating technique for any xQy design flow is
presented in Appendix A.  Although the Log-Pearson Type III is in
general use it should be recognized that there are other distributions
that may be more appropriate to use on a case-by-case basis.  The
hydrologically-based design flow for ammonia is discussed in Appendix
B.

     Analyses of the 1Q10 and 7Q10 low flow in Table 2-1 indicate that
the mean of the ratios of 7Q10 to 1Q10 is 1.3.  The median of the
ratios is 1.1, whereas the range of the ratios is 1.0 to 3.85.  Thus,
7Q10 low flows are generally 10 to 30% greater than the corresponding
1Q10 low flows,  although in one case the 7Q10 is 3.85 tines greater
than the corresponding 1Q10.

Table 2-1.   Hydrologically-based design flows (ft3/sec)  for 60 streams
Station ID
01657000
02092500
06026000
12449600
05522000
09490800
14372500
05381000
10291500
05585000
12321500
01111500
River Name
Bull Run
Trent
Birch Cr
Beaver Cr
Iroquois
N Fk White
E FK Illinois
Black
Buckeye
LaMoine
Boundary Cr
Branch
State
VA
NC
MT
WA
IN
AZ
OR
WI
CA
IL
ID
RI
Period of
Record
1951-82
1951-82
1946-77
1960-78
1949-78
1966-78
1942-03
1905-83
1911-78
1921-83
1928-84
1940-82
CV*
4 .48
1.77
1.32
1.77
1.33
1.24
2 .03
2 .51
1.30
1.99
1.65
1.16
Design flow (ft 3/sec)
1Q10
0.3
1.4
1.7
2 .4
3 .4
4 .8
6 .4
5.5
7.1
9.3
11.7
8 .8
7Q10
0.4
1.6
2 .4
3 .2
3 .9
5.3
6 .7
6 .7
7 .7
9.9
13 .1
13 .3
7Q10
1Q10
1.33
1.14
1.41
1.22
1.15
1.10
1.05
1.22
1.08
1.06
1.12
1.51
                                  2-3

-------
Table 2-1  (continued)
Station
ID River Name
Period of QV*
State Record
Design flow (ft 3/sec)
1Q10 7Q10
7Q10
1Q10
02138500
05053000
02083000
01196500
02133500
06280300
09149500
02296750
07018500
02217530
01481000
09497500
01144000
01600000
09359500
01403060
02413500
01421000
07298500
07013300
01531000
07096000
09070000
01011000
03528000
13023000
02424000
05515500
02490500
01315500
01610000
05386000
02369000
07378500
06465500
02135000
08110200
02076000
03455000
05333500
06287000
03107500
Linville
Sheyenne
Fishing Cr
Quinnipiac
Drowning Cr
S ho shone
Uncompahgre
Peace
Rig
Middle Oconee
Brandywine
Salt
White
N Br Potomac
Animas
Raritan
L Tallapoosa
E B Delaware
Big Sunflower
Meramec
Chemung
Arkansas
Eagle
Allegash
Clinch
Greys
Cahaba
Kankakee
Bouge Chitto
Hudson
Potomac
Root
Shoal
Amite
Niobrara
Little Pee Dee
Brazos
Dan
French Broad
St. Croix
Bighorn
Beaver
NC
ND
NC
CT
NC
WY
CO
FL
MO
GA
PA
AZ
VT
MD
CO
NJ
AL
NY
MS
MO
NY
CO
CO
ME
TN
WY
AL
IN
MS
NY
WV
MN
FL
LA
NE
SC
TX
VA
TN
WI
MT
PA
1922-
1951-
1927-
1931-
1940-
1957-
1939-
1931-
1922-
1902-
1912-
1925-
1915-
1939-
1946-
1904-
1940-
1915-
1936-
1923-
1915-
1901-
1947-
1932-
1919-
1937-
1902-
1926-
1945-
1908-
1939-
1938-
1939-
1939-
1939-
1942-
1966-
1924-
1901-
1914-
1935-
1957-
84
81
82
84
78
84
80
84
84
84
84
80
84
83
56
83
51
78
80
78
78
81
80
03
78
83
78
78
81
78
83
61
82
83
83
78
70
52
78
81
79
83
1 .
2 .
1.
1.
0.
1 .
0.
1.
2 .
1 .
1.
2 .
1 .
1 .
1.
1.
1 .
1 .
1.
2 .
1 .
1 .
1.
1.
1 .
1 .
2 .
0.
1 .
1 .
1.
1.
0.
1 .
0.
0.
1 .
1 .
0.
0.
0.
1.
.74
.10
.48
.02
.80
.54
.86
.54
.16
.37
.17
.05
.43
.42
.56
.64
.31
.41
.42
.41
.91
.12
.36
.39
.55
.16
.07
.48
.89
.10
.48
.65
.95
.98
.59
.94
.48
.25
.93
.61
.82
.10
13 .
15.
17.
17.
38 .
41.
35.
49.
46 .
49.
61.
64 .
75.
54 .
54 .
54 .
72 .
80.
89.
88 .
89.
107
116
124
120
122
151
179
188
207
209
229
280
298
160
306
311
329
473
505
327
571
4
9
0
5
8
8
6
0
4
4
4
6
3
7
8
2
7
8
4
8
7
.9
.9
.5
.7
.9
.9
.0
.6
.7
.6
.7
.1
.1
.9
.7
.6
.6
.6
.9
.1
.3
16
13
19
32
43
46
50
155
55
57
67
68
85
61
62
67
8 .
89
91
92
97
126
131
134
135
144
156
184
191
211
220
245
291
303
322
322
344
387
532
536
557
594
.4
.3
.4
.3
.4
.8
.8
.3
.3
.4
.2
.7
.2
.6
.3
.1
3
.7
.9
.2
.5
.1
.0
.1
.2
.5
.4
.3
.6
.0
.7
.6
.4
.4
.0
.4
.9
.3
.2
.0
.0
.2
1.22
1.15
1.14
1.85
1.12
1.12
1.43
1.13
1.19
1.16
1.09
1.06
1.13
1.13
1.15
1.24
1.21
1 .11
1.03
1.05
1.09
1.17
1.12
1.38
1.05
1.13
1.03
1.33
1.02
1.02
1.05
1.07
1.04
1.02
2 .00
1.09
1 .11
1.18
1.12
1.06
1.70
1.04
                                      2-1

-------
  Table 2-1  (continued).
Station ID
13341000
07341500
02350500
01536500
01100000
14233430
River Name
N P Clearwater
Red
Flint
Susquehanna
Merrimack
Cowlitz
State
ID
AR
GA
PA
MA
WA
Period of
Record
1927-68
1928-81
1930-58
1901-83
1924-83
1968-78
CV*
1.16
1.41
1.00
1.34
1.01
0.93
Design flow (ft 3/sec)
1Q10
529.2
691.0
207.8
782 .0
270.2
901.5
7Q10
648 .6
769.2
799.8
814 .3
929.3
958 .7
7Q10
1Q10
1.23
1.11
3 .85
1.04
3 .44
1.07
*CV  = Coefficient of Variation
                                           2-2

-------
             SECTION 3.   Biologically-Based Design Flow

3 .1  Introduction

     The purpose of this section is to describe the biologically-
based design flow calculation method and provide some examples of
its use.  This method was developed by the Office of Research and
Development of the U.S.   EPA in order to provide a way of directly
using EPA's two-number aquatic life water quality criteria (WQC)
for individual pollutants and whole effluents to calculate the
design flow for performing a wasteload allocation using steady-
state modeling.  The two-number WQC are in the intensity-duration-
frequency format, in that they specify intensity as criteria
concentrations, duration as averaging periods, and frequency as
average frequency of allowed excursions.   Because the flow of, and
concentrations of pollutants in,  effluents and stream are easily
considered in terms of intensity,  duration, and frequency, use of
this format for expressing WQC allows a direct application to
effluents and streams.

     Because steady-state modeling assumes that the composition and
flow of the effluent of concern is constant, the ambient  (instream)
concentration of a pollutant can be considered to be inversely
proportional to stream flow.  Thus by applying a specified
averaging period and frequency to a record of the historical flow
of the stream of concern, the design flow can be calculated as the
highest flow that will not cause exceedances to occur more often
than allowed by the specified average frequency, based on
historical data.  The allowed exceedances are intended to be small
enough and far enough apart, on the average, that the resulting
small stresses on aquatic organisms will not cause unacceptable
effects, except in those cases when a drought itself would cause
unacceptable effects.

     The averaging periods  specified in national water quality
criteria are one hour for the CMC and four days for the CCC.   The
primary use of the averaging periods in criteria is for averaging
ambient concentrations of pollutants in receiving waters in order
that the averages can be compared to the CMC and CCC to identify
                                3-1

-------
"exceedances" i.e., one-hour average concentrations that exceed the
CMC and four-day average concentrations that exceed the CCC.
However, in steady-state modeling, flow is averaged over a given
period to identify "non-exceedances",  i.e., average flows that are
below a specified flow.

3.1.1 Exceedances and Excursions

     Use of the term  "exceedance" and  "non-exceedance" neither of
which are in the dictionary, can be a cause of confusion.  Water
quality criteria are usually expressed as upper limits on
concentrations in ambient water and the periods of concern are when
the ambient concentration exceeds a criterion concentration, i.e.,
when there is an exceedance.  In steady-state modeling, the
averaging is of flows, not concentrations.  Because a low flow
results in a high pollutant concentration, the period of concern
for flow is when the flow is less than the design flow, i.e., when
there is non-exceedance of a given flow.  A non-exceedance of a
design flow corresponds to an exceedance of a criterion.  Use of
the non-directional term "excursion",  which is in the dictionary,
avoids this confusion.  Use of the term "excursion" also avoids the
problem that some water quality criteria, such as those for
dissolved oxygen and low pH, must be stated as lower limits, not
upper limits.  An exceedance of a dissolved oxygen criterion is
favorable,  not unfavorable.  "Excursions", in this guidance manual,
will henceforth be used to imply
                                3-2

-------
an unfavorable condition, e.g., a low flow or a pollutant
concentration above an upper limit or below a lower limit.

     The national water  quality criteria specify that, if R is the
calculated number of excursions occurring in a period of S years,
then S/R should be equal to or greater than 3 years.  Most
excursions will be small and most aquatic ecosystems will probably
recover from the resulting minor stress in less than three years.
However, the three years is meant to be longer than the average
recovery period so that ecosystems cannot be in a constant state of
recovery even if excursions are evenly spaced over time.

     Although 3 years appears  to be appropriate for small
excursions that are somewhat isolated, it appears to be excessively
long when many excursions occur in a short period of time, such as
would be caused by a drought.  Droughts are rare events,
characterized by long periods of low flow and should not be allowed
to unnecessarily lower design flows.  Although droughts do severely
stress aquatic ecosystems, both directly, because of low flow, and
indirectly, because of the resulting high concentrations or
pollutants, many ecosystems apparently recover from severe stresses
in more than 5,  but less than 10 years (1).   Because it is not
adequately protective to keep ecosystems in a constant state of
recovery 15 years seem like an appropriate stress-free period of
time, on the average to allow after a severe stress caused by a
drought situation.  Because three years are allowed for each
excursion on the average, counting no more than 5 excursions for
any low flow period will
                                3-3

-------
provide no more than 15 years, on the average, for severe stresses
caused by droughts.  Thus, for each low flow period, the number of
excursions cannot be less than 1.0 or greater than 5.0.  The
maximum duration of a low-flow period was set at 120 days because
it is not too uncommon for excursions to occur within 120 days of
each other, whereas it is very rare for excursions to occur during
days 121 to 240 after the beginning of a low-flow period.

3.1.2 Features of Calculation

     Figure 3-1 illustrates the  features of the biologically-based
design flow calculation method.  Intervals a-b and c-d are
excursion periods and each day in these intervals is part of an
average flow that is below the design flow.  The number of
excursions in an excursion period is calculated as the number of
days in the excursion period divided by the duration (in days) of
the averaging period (e.g., 1 day for the CMC and 4 days for CCC).
A low-flow period is defined as one or more excursion periods
occurring within a 120-day interval.  As discussed above, if the
calculated number of excursions that occur in a 120-day low-flow
period is greater than 5, the number is set at 5 for the purposes
of calculating the design flow.

     Because biologically-based  design flows  are based on the
averaging periods and frequencies specified in water quality
criteria for individual pollutants and whole effluents, they can be
based on the available biological, ecological, and toxicological
information concerning the stresses that aquatic organisms,
ecosystems, and their uses can tolerate.   The biologically-based
calculation method is flexible enough to make full use
                                3-4

-------
              Exclusion
                               Time
Figure 3-1:  Illustration of biologically-based design flow
                                  3-5

-------
     The CMC and CCC design flows are calculated in almost the same
manner.  The differences result from the fact that the CMC is
expressed as a one-hour average, whereas the CCC is expressed as a
four-day average.   However, the flow records that are available
consist of one-day average flows.  For streams with naturally
occurring low flows, calculation of the CMC design flow from one-
day averages,  rather than one-hour averages, should be reasonably
acceptable because naturally occurring low flows of receiving
streams are usually very similar from one hour to the next.  In
regulated streams,  such as those affected by hydroelectric or
irrigation projects, hour-to-hour variation of low flows could be
significant and in those situations, use of hourly values, when
available,  is appropriate.  Both the pollutant concentrations and
the flows of most  effluents are expected to change much more from
one hour to the next than the naturally occurring flows of streams.

3.3  Rationale

     The following provides a rationale for the biologically-based
design flow calculation method:

*  It allows the use of the new two-number WQC for aquatic life in
   the calculation of design flow.  If water quality criteria for
   aquatic life are to achieve their intended purpose, decisions
   concerning their derivation and use should be based on the
   biological,  ecological, and toxicological characteristics of
   aquatic organisms and ecosystems and their uses whenever
   possible.
*  It takes into account all excursions in the flow record.
*  It provides the necessary design flow directly without requiring
   any design flow statistics in the xQy format.
*  It is flexible  enough so that any averaging period and frequency
   selected for particular pollutants, effluents,  or site-specific
   criteria can be used directly in design flow calculations.
                                3-7

-------
3 .4  Example Cases

     The sixty flow records that were analyzed using the
hydrologically-based method (see Table 2-1)  were also analyzed
using the biologically-based design flow method.  The CMC design
flow was calculated for a 1-day averaging period and the CCC design
flow was calculated using the 4-day averaging period.  Both were
calculated using a frequency of once every three years on the
average.  Table 3-1 presents biologically-based design flows for
these sixty rivers.

     In addition to the hydrologically-based design  flows, Table B-
1 in Appendix B also includes biologically-based CMC and CCC design
flows for 13 streams for 30-day averaging periods and a frequency
of once every three years on the average.  The purpose of the
biologically-based design flows for ammonia (5)  in Appendix B is to
illustrate how this method might be used for site-specific and
pollutant-specific situations where the durations and frequencies
in aquatic life criteria might be different from those specified in
national two-number aquatic life criteria.

     Analyses of the 1-day 3-year and the 4-day 3-year low flows in
Table 3-1 indicate that the mean ratio of the 4-day 3-year low
flows to the corresponding 1-day 3-year low flows is 1.23.  The
median of the ratios is 1.11,  whereas the range of the ratios is
1.0 to 2.81.  Thus, 4-day 3-year low flows are generally 11 to 23%
greater than the corresponding 1-day 3-year low flows,  although in
one case,  the 4-day 3-year low flow is 2.91 times greater than the
corresponding 1-day 3year low flow.

-------
Table 3-1.  Biologically-based design  flows  (ft3/sec) for 63 rivers
Station ID
01657000
02092500
06026000
12449600
05522000
09490800
14372500
05381000
10291500
05585000
12321500
01111500
02138500
05053000 or
03059000
02083000
01196500
02133500
06280300
09149500
02296750
07018500
02217530
01600000
09359500
01403060
01481000
09497500
01144000
02413500
01421000
07288500
07013300
01531000
07096000
09070000
01011000
03528000
13023000
02424000
River Name
Bull Run
Trent
Birch Cr
Beaver Cr
Iroquois
N Fk White
E FK Illinois
Black
Buckeye
LaMoine
Boundary Cr
Branch
Linville
Sheyenne

Fishing Cr
Quinnipiac
Drowning Cr
S ho shone
Uncompahgre
Peace
Big
Middle Oconee
N Br Potomac
Animas
Raritan
Brandy wine
Salt
White
L Tallapoosa
E B Delaware
Big Sunflower
Meramec
Chemung
Arkansas
Eagle
Allegash
Clinch
Greys
Cahaba
State
VA
NC
MT
WA
IN
AZ
OR
WI
CA
IL
ID
RI
NC
ND

NC
CT
NC
WY
CO
FL
MO
GA
MD
CO
NJ
PA
AZ
VT
AL
NY
MS
MO
NY
CO
CO
ME
TN
WY
AL
Period of
Record
1951-82
1951-82
1946-77
1960-78
1949-78
1966-78
1942-03
1905-83
1911-78
1921-83
1928-84
1940-82
1922-84
1951-81

1927-82
1931-84
1940-78
1957-84
1939-80
1931-84
1922-84
1902-84
1939-83
1946-56
1904-83
1912-84
1925-80
1915-84
1940-51
1915-78
1936-80
1923-78
1915-78
1901-81
1947-80
1932-03
1919-78
1937-83
1902-78
CV*
4 .48
1 .77
1 .32
1.77
1.33
1 .24
2 .03
2 .51
1.30
1 .99
1 .65
1.16
1.74
2 .10

1.48
1 .02
0 .80
1.54
0.86
1 .54
2 .16
1 .37
1.42
1.56
1.64
1 .17
2 .05
1.43
1.33
1.41
1 .42
2 .41
1.91
1.12
1.36
1 .39
1.55
1.16
2 .07
Design flow (itybsec)
1-day 3 -year
0.20
1.40
1. 70
2.80
2.40
4 . 80
5. 80
5.00
7.00
8 . 90
12 .00
10.00
13 .00
15.40

12 .00
14 .90
33 .90
42 .90
29.90
48 .00
45 .00
33 .00
42 .90
60.00
46 .90
55 .80
63 .00
75.90
57.90
82 .00
82 .70
89 .90
85.70
89.90
120.00
134 .00
127.70
124 .80
122 .80
4 -day 3 -year
0.40
1. 60
2 .40
3.40
3.00
5.30
6 . 90
6.10
7.20
9.40
13 . 00
13.20
15.00
17.60

13.50
34 . 00
36 .20
45.80
49.00
55.20
51. 50
45. 70
49.00
61.10
53.60
59.30
59. 50
86.00
70.20
91.40
85.40
92 . 70
92.50
114.00
126.00
138 .40
132.20
135.80
149. 80
ccc
CMC 08
2 .00
1 .14
1 .41
1.21
1.25
1 .10
1 .19
1.22
1.03
1 .06
1 .08
1.32
1.15
1.14

1.13
2 .25
1 .07
1.07
1.26
1 .15
1 .14
1 .38
1.17
1.02
1.14
1 .06
1 .10
1.13
1.21
1.11
1 .03
1 .03
1.08
1.27
1.05
1 .03
1.04
1.09
1 .22

-------
Table 3-1 (continued).
Station ID
05515500
02490500
01315500
01610000
05386000
02369000
07378500
06465500
02135000
08110200
02076000
03455000
05333500
06287000
03107500
13341000
07341500
02350500
01100000
14233430
River Name
Kankakee
Bouge Chitto
Hudson
Potomac
Root
Shoal
Amite
Nebraska
Little Pee Dee
Brazos
Dan
French Broad
St . Croix
Bighorn
Beaver
N P Clearwater
Red
Flint
Merrimack
Cowl it z
State
IN
MS
NY
WV
MN
FL
LA
NE
SC
TX
VA
TN
WI
MT
PA
ID
AR
GA
MA
WA
Period of
Record
1926-78
1945-81
1908-78
1939-83
1938-61
1939-82
1939-83
1939-83
1942-78
1966-70
1924-52
1901-78
1914-81
1935-79
1957-83
1927-68
1928-81
1930-58
1924-83
1968-78
CV*
0.48
1.89
1.10
1.48
1.65
0. 95
1.98
0.59
0. 94
1.48
1.25
0.93
0.61
0.82
1.10
1.16
1.41
1.00
1.01
0.93
Design flow (ft /sec)
1-day 3-year
167.60
187.50
170.00
22.20
239.30
270.50
282.10
199.70
298.70
277.70
321.60
494 .30
477.50
364 .00
539.90
429.60
537.40
262 .50
284.00
934 .70
4-day 3-year
174.20
189.60
191.90
219.60
23937.00
2860.00
295.50
304.30
298.90
305.30
380.40
535.50
508.50
520.20
557.50
613.00
603 .30
731.00
797.30
959.90
ccc
CMC08
1.04
1.13
1.13
1.09
1.00
1.06
1.05
1.52
1.00
1.10
1.18
1.08
1.06
1.43
1.07
1.31
1.12
2 .78
2.81
1.03
 = coefficient of variation
                                3-10

-------
For further clarification of the biologically-based method, refer to
Appendix E, Questions and Answers.
                                   3-11

-------
of special averaging periods and frequencies that might be selected
for specific pollutants (e.g., ammonia)  or in site-specific
criteria.  This method is empirical, not statistical, because it
deals with the actual flow record itself, not with a statistical
distribution that is intended to describe the flow record.

     In  addition, this method provides an understanding of how many
excursions of the CCC or CMC are likely to occur, and during what
time of the year, based on actual historical flow data.  Thus, it
is possible to examine the pattern and magnitudes of what would
have been historical excursions.  This method makes it clear that
criteria concentrations should not be interpreted as values that
are never to be exceeded "at any time or place" in the receiving
waters.  An understanding of what level of protection actually is
provided should aid in the use of criteria.

3 .2  Procedure

     Although the calculation procedure described in Appendix C
might look complicated, it merely consists of a sequence of steps
that are quite simple.  Because flow records usually consist of
daily flows for 20 to 80 years,  manual calculation of design flow
is very time-consuming.  The DFLOW computer program  (Appendix D
(OMITTED) - DFLOW 2.0 has been superseded by newer versions. The
current versions of DFLOW and its documentation are available
online at http://www.epa.gov/waterscience/dflow.) will calculate
biologically-based design flows and display the dates, durations,
and magnitudes of the excursions within each low flow period.
                                3-6

-------
              SECTION 4. COMPARISON OF THE TWO METHODS

4 .1  Design Flows

     Table 4-1 shows the biologically-based 1-day 3-year low flows
and the hydrologically-based 1Q10 low flows for the sixty example
rivers.  The table also presents the difference between 4-day 3-year
low flows and the 7Q10 low flows.

     For 39 of the 60 streams, the 1-day 3-year low flows are less
than the 1Q10 low flows.  For 18 streams,  the 1-day 3-year low flows
are greater than the 1Q10 low flows,  and for the remaining 3 streams
the differences  are less than 0.1%.   Thus,  for the majority of the
streams the 1-day 3-year low flow is lower than the 1Q10 low flow.
For all sixty streams,  the difference between 1-day 3-year low flows
and 1Q10 low flows ((1-day 3-year)-(1Q10))/(1-day 3-year)  ranges
from -50.0% to 20.8%,  with the mean  and median equal to -4.9% and -
3.1%,  respectively.
                                 4-1

-------
       Table  4-1.   Comparison of 1Q10 and 7Q10 with 1-day 3-yr and 4-day 3-
                    yr low flows (all flows in ft3/sec)

River Name State
Bull Run VA
Trent NC
Birch Cr MT
Beaver Cr WA
Iroquois IN
N Fk White AZ
E FK Illinois OR
Black WI
Buckeye CA
LaMoine IL
Boundary Cr ID
Branch RI
Linville NC
Sheyenne ND
Fishing Cr NC
Quinnipiac CT
Drowning Cr NC
Shoshone WY
Uncompahgre CO
Peace FL
Big MO
Middle Oconee GA
N Br Potomac MD
Animas CO
Raritan NJ
Brandywine PA
Salt AZ
White VT
L Tallapoosa AL
E B Delaware NY
Big Sunflower MS
Meramec MO
Chemung NY
Arkansas CO
Eagle CO
Allegash ME
Clinch TN
Greys WY
Cahaba AL
Comparison of CMC Design Flows
1Q10 1-day %DIFF*
3-yr
0.3 0.2 -50.0
1.4 1.4 0.0
1.7 1.7 0.0
2.4 2.8 14 .3
3.4 2.4 -41.7
4.8 4.8 0.0
6.4 5.8 -10.3
5.5 5.0 -10.0
7.1 7.0 -1.4
9.3 8.9 -4.5
11.7 12 .0 2.5
8.8 10.0 12 .0
13 .4 13 .0 -3.1
15.9 15.4 -3.2
17.0 12.0 -41.7
17.5 14.9 -17.4
38.8 33.9 -14.4
41.8 42 .9 2.6
35.6 39.9 10.8
49.0 48 .0 -2.1
46 .4 45.0 -3.1
49.4 33 .0 -49.7
54.7 42.9 -27.5
54 .8 60.0 8.7
54.2 46.9 -15.6
61.4 55.8 -10.0
64 .6 63 .0 -2.5
75.3 75.9 0.8
72.7 57.9 -25.6
80.8 82 .0 1.5
89.4 82 .7 -8.1
88 .8 89.9 1.2
89.7 85.7 -4.7
99.9 89.9 -11.1
116.9 120.0 2.6
124.5 134.0 7.1
128.7 127.7 -0.8
122.9 124.8 1.5
151.9 122.0 -23.7
Comparison of CCC Design Flows
7Q10 4 -day 3- %DIFF*
yr
0.4 0.4 0.0
1.6 1.6 0.0
2.4 2.4 0.0
3.2 3.4 5.9
3.9 3.0 -30.0
5.3 5.3 0.0
6.7 6.9 2.9
6.7 6.1 -9.8
7.7 7.2 -6.9
9.9 9.4 -5.3
13 .1 13 .0 -0.8
13 .3 13 .2 -0.8
16 .4 15.0 -9.3
18 .3 17.6 -4.0
19.4 13.5 -43.7
32 .3 34 .0 5.0
43.4 36.2 -19.9
46 .8 45.8 -2.2
50.8 49.0 -3.7
55.3 55.2 -0.2
55.3 51.5 -7.4
57.4 45.7 -25.6
61.6 49.0 -25.7
62 .3 61.1 -2.6
67.1 53.6 -25.2
67.2 59.3 -13.3
68 .7 69.5 1.2
85.2 86 .0 0.9
88.3 70.2 -25.8
89.7 91.4 1.9
91.9 85.4 -7.6
92.2 92.7 0.5
97.5 92.5 -5.4
120.1 114.0 -9.3
131.0 126.0 -4.0
134.1 138.4 3.1
135.2 132.2 -2.3
144.5 135.8 -6.4
156.4 149.8 -5.4
* %Difference
* %Difference
-day 3-year flow)
day 3-year flow)
 (1Q10)),100 /  (4-day
(7Q10)),100 /  ((4-day
3-year flow)
3-year flow)
                                         4-2

-------
Table 4-1. (continued).
River Name State
Kankakee IN
Bouge Chitto MS
Hudson NY
Potomac WV
Root MN
Shoal FL
Amite LA
Niobrara NE
Little Pee Dee SC
Brazos TX
Dan VA
French Broad TN
St. Croix WI
Bighorn MT
Beaver PA
N P Clearwater ID
Red AR
Flint GA
Merrimack MA
Cowlitz WA
Comparison of CMC Design
Flows
1Q10 1-day 3-yr %DIFF*
179.0 167.6 -6.8
188.6 167.5 -0.6
207.7 170.0 -22.2
209.6 202.2 -3.7
229.7 239.3 4 .0
280.1 270.5 -3.5
298.1 202.1 -5.7
160.9 199.7 19.4
306.7 298.7 -2.7
311.6 277.7 -12.2
329.6 321.6 -2.5
473.6 494.3 4.2
505.9 477.5 -5.9
327.1 364.0 10.1
571.3 539.9 -5.8
529.2 469.6 -12.7
691 537.4 -29.6
207.8 262.5 20.8
270.2 284.0 3.6
901.5 934.7 4.9
Comparison of CCC Design
Flows
7Q10 4 -day 3-yr %DIFF*
184.3 174.2 -5.8
191 .6 189 .6 -1.1
211 . 0 191 . 9 -10 . 0
220.7 219.6 -0.5
245.6 239.7 -2.5
291.4 286.0 -1.9
303.4 295.5 -2.7
322.0 304.3 -5.8
322.4 298.9 -7.9
344.9 305.3 -13.0
307.3 380.4 -1.8
532.2 535.5 0.6
536.0 508.5 -5.4
557.0 520.2 -7.1
594.2 557.5 -6.6
648.6 613.0 -5.9
769.2 603.3 -27.5
799.8 731.3 -9.4
929.3 797.3 -16.6
968.7 959.9 -0.9
* %Difference - ((1
* %Difference - (4-
-day 3-year flow)
day 3-year flow)
 (1Q10)),100 /  (1
(7Q10)),100 /  ((4
-day 3-year flow)
-day 3-year flow)
           Similar comparisons can be made between  the  4-day  3-year  low
      flows and the 7Q10 low flows based on Table 4-1.  For 46 of the 60
      streams, the 4-day 3-year low flows are less than the 7Q10 low
      flows.   For nine streams, 4-day 3-year low flows are greater than
      the 7Q10 low flows, and for the remaining four streams, the
      differences are less than 0.1%.  Thus,  the 4-day 3-year low flow is
      usually lower than the 7Q10 low flow.  For all sixty streams, the
      difference between the 4-day 3-year low flows and 7Q10  low flows
      ((4-day 3-year)  -  (7Q10))/(4-day 3-year))  ranges from -44% to 6%,
      with the mean and median equal to - 7.0% and - 4.4%, respectively.
      4.2  Excursions
                                       4-3

-------
     Table 4-2 presents the calculated number of excursions that



occurred in the 60 streams for the low flows calculated using the



hydrologically- and biologically-based methods.   The table



demonstrates the impact of the choice of one design flow method over



the other in terms of number of excursions.  For any stream, a



higher flow will always result in the same or a greater number of



excursions than a lower flow.   Occasionally, the difference in the



number of excursions of the two design flows is quite dramatic even



if the difference between the two design flows is quite small.  For



example, the 1Q10 and the 1-day 3-year design flow of the Quinnipiac



River in Connecticut are 17.5 ft3/sec  and  14.9 ft2/sec,  respectively,



but the corresponding numbers of excursions were 39 and 13.  Similar



observations could be made for many other streams in Table 4-2.  A



small difference in design flow may not have a significant impact in



wasteload allocations for these streams but may result in a larger



number of excursions that desired during the period of flow record.







4.3  Comparison of the Two Methods





     The comparisons of the design flows show that the magnitudes of



the 1-day 3-year and 1Q10 low flows,  and the 4-day 3-year and 7Q10



low flows are,  on an average basis, similar in magnitude.  Although



these flows are similar on the average, there may be large



differences in the values of these flows for individual streams.



More importantly, there can be a significant difference in the



number of excursions that result, even if the magnitudes of the



flows calculated by the two methods are nearly equal.
                                 4-4

-------
Table 4-2. Comparison of number of excursions of 1Q10 and 7Q10 with
          number of  excursions of 1-day 3-yr and 4-day 3-yr design
          flows.
River Name
Bull Run
Trent
Birch Cr
Beaver Cr
Iroquois
N Fk White
E FK Illinois
Black
Buckeye
LaMoine
Boundary Cr
Branch
Linville
Sheyenne
Fishing Cr
Quinnipiac
Drowning Cr
Shoshone
Uncompahgre
Peace
Big
Middle Oconee
N Br Potomac
Animas
Raritan
Brandywine
Salt
White
L Tallapoosa
E B Delaware
Big Sunflower
Meramec
Chemung
State
VA
NC
MT
WA
IN
AZ
OR
WI
CA
IL
ID
RI
NC
ND
NC
CT
NC
WY
CO
PL
MO
GA
MD
CO
NJ
PA
AZ
VT
AL
NY
MS
MO
NY
Comparison of CMC Design Flows Comparison of CCC Design Flows
1Q10 % Excur 1-day 3-yr % Excur 7Q10 % Excur 1-day 3-yr % Excur
0.3 19 0.2 10 0.4 8.5 0.4 8.5
1.4 9 1.4 9 1.6 9.3 1.6 9.2
1.7 8 1.7 8 2.4 9.3 2.4 9.2
2.4 1 2.8 6 3.2 4.0 3.4 6.0
3.4 18 2.4 9 3.9 16.8 3.0 9.7
4.8 2 4.8 2 5.3 4.0 5.3 4.0
6.4 13 5.8 12 6.7 11.3 6.9 11.5
5.5 27 5.0 21 6.7 26.0 6.1 24.5
7.1 13 7.0 7 7.7 10.0 7.2 8.5
9.3 33 8.9 20 9.9 24.5 9.4 20.5
11.7 15 12.0 15 13.1 15.8 13.0 15.7
8.8 10 10.0 13 13.3 18.3 13.2 14.0
13.4 21 13.0 15 16.4 25.0 15.0
15.9 11 15.4 6 18.3 14.5 17.6
17.0 17 12.0 15 19.4 29.3 13.5 17.2
17.5 39 14.9 13 32.3 11.3 34.0 13.0
38.8 26 33.9 12 43.4 27.8 36.2 12.7
41.8 3 42.9 6 46.8 9.3 45.8 6.3
35.6 7 39.9 13 50.8 17.5 49.0
49.0 17 48.0 16 55.3 17.3 55.2
46.4 23 45.0 15 55.3 27.8 51.5
49.4 25 33.0 11 57.4 23.3 45.7 14.3
54.7 29 42.9 14 61.6 28.0 49.0 14.8
54.8 0 60.0 2 62.3 6.8 61.1 2.5
54.2 25 46.9 13 67.1 24.3 53.6 13.3
61.4 30 55.8 14 67.2 33.0 59.3
64.6 21 63.0 18 68.7 17.3 6935.0
75.3 20 75.9 20 85.2 20.8 86.0 21.5
72.7 6 57.9 3 88.3 7.0 70.2 3.8
80.8 17 82.0 20 89.7 19.0 91.4 20.5
89.4 31 82.7 8 91.9 30.3 85.4 13.8
88.8 17 89.9 18 92.2 16.5 92.7 17.0
89.7 26 85.7 18 97.5 25.0 92.5 20.5
                                 4-5

-------
Table 4-2.  (Continued)
River Name
Arkansas
Eagle
Allegash
Clinch
Greys
Cahaba
Kankakee
Bouge Chitto
Hudson
Potomac
Root
Shoal
Amite
Niobrara
Little Pee Dee
Brazos
Dan
French Broad
St. Croix
Bighorn
Beaver
N P Clearwater
Red
Flint
Merrimack
Cowlitz
State
CO
CO
ME
TN
WY
AL
IN
MS
NY
WV
MN
FL
LA
NE
SC
TX
VA
TN
WI
MT
PA
ID
AR
GA
MA
WA
Comparison of CMC Design
1Q10
107.9
116.9
124.5
128.7
122.9
151.9
179.0
188.6
207.7
209.6
229.7
280.1
298.1
160.9
306.7
311.6
329.6
473.6
505.9
327.1
571.3
529.2
691.0
207.8
270.2
901.5
% Excur
23
9
15
23
10
33
34
13
30
19
7
20
19
4
15
11
11
13
34
12
15
20
28
7
13
0
1-day
115
120
134
127
1234
122
167
187
170
202
239
270
282
199
299
277
321
494
477
364
539
469
537
2625
284
934
3-yr
8
0
0
7
.8
8
6
5
0
2
3
5
1
7
7
7
6
3
5
0
9
6
4
.0
0
7
Flows
% Excur
26
11
17
17
10
10
14
10
29
14
7
12
14
8
12
4
9
18
2 2
14
4
13
17
9
18
2
Comparison of CCC Design
7Q10
126.1
131.0
134.1
135.2
144.5
156.4
184.3
191.6
211.0
220.7
245.6
291.4
303.4
322.0
322.4
344.9
387.3
532.2
536.0
557.0
594.2
643.6
769.2
799.8
929.3
963.7
% Excur
28.0
17.5
13.0
25.0
18.8
24.8
29.5
19.3
27.8
15.0
10.8
19.3
14.0
11.3
15.0
6.8
10.3
16.0
34.5
16.5
13.3
14.8
28.8
20.3
41.8
4.5
1-day
123
126
138
132
135
149
174
189
191
219
239
286
295
304
298
305
380
535
508
520
557
613
603
731
797
959
3-yr
8
0
4
2
8
8
2
6
9
6
7
0
5
3
9
3
4
5
5
2
5
0
3
0
3
3
Flows
% Excur
26.0
11.0

12.0
10.0
16.0
14.0
11.0
24.0
14.0
7.0
17.0
4.0
8.0










19.0
3.0
                                 4-6

-------
     The hydrologically-based design flows may actually provide a



greater degree of protection of water quality in cases where the



value of the design flows are less than that of the corresponding



biologically-based design flows.  Hydrologically-based design flows



have been used successfully in the past in many water quality-based



permits.  In addition, on an average basis, the values of



hydrologically-based design flows are not greatly different from the



corresponding values of biologically-based design flows.






     The biologically-based design flows are not always smaller than



the corresponding hydrologically-based design flows for a given



stream.  Thus, it cannot be stated that choosing one method over the



other will always result in the most protective wasteload allocation



(and therefore the fewest number of excursions over the period of



record).  However, the biologically-based method will always provide



insurance that the design flow calculated will have resulted in no



more than the required number of excursions.






     Based upon the above, both the hydrologically-based and the



biologically-based methods for calculating stream design flows are



recommended for use in steady-state modeling.
                                 4-7

-------
                     SECTION 5.   RECOMMENDATIONS





1.   If steady-state modeling is used, the hydrologically-based or



     the biologically-based stream design flow method should be



     used.  If the hydrologically-based method is used, the 1Q10 and



     7Q10 low flows should be used as the CMC and CCC design flow,



     except that the 30Q10 low flow should be used as the CCC design



     flow for ammonia is situations involving POTWs designed to



     remove ammonia where limited variability of effluent pollutant



     concentrations and resulting concentrations the receiving water



     can be demonstrated.





2.   Other technically defensible methods may also be used.
                                 5-1

-------
                       SECTION 6.  REFERENCES

1.   U.S. EPA 1985. Technical support document for water-quality
     based toxics control. Office of Water, Washington D.C.
     September, 1985.
2.    U.S. EPA.  Water Quality Criteria.  50 FR 30784 July 29, 1985.
3.   Stephan, C.E., D.I. Mount, D.J, Hansen, J.H. Gentile,  G.A.
     Chapman and W.A. Brungs. 1985. Guidelines for deriving
     numerical national water quality criteria for the protection
     of aquatic organisms and their uses. PB85-227049. National
     Technical Information Service, Springfield, VA.

4.   U.S. EPA. 1984. Water Quality Standards Handbook. Office of
     Water regulations and Standards, Washington D.C.

5.   U.S. EPA. 1985. Ambient water quality criteria for ammonia -
     1984. EPA 440/5-85-001. National Technical Information Service,
     Springfield, VA.

6.    U.S. EPA.  1985. STORET User Handbook,  Part FL,  Flow Data File.

-------
   APPENDIX A.   Calculation  of  Hydrologically-Based Design Flows


     Design flows can be calculated as annual x-day average  low

flows whose return period is y years,  i.e., the xQy  low flow.  These

flows can be estimated from a historical flow record of n years

using two different methods.  The first is a distribution-free

method which makes no assumption about the true probability

distribution of annual low flows.  The expression for xQy is


                     xQy =  (1-e) X(ml) +  eX(m2)


where X(m) = the m-th lowest annual low flow of record
       ml =  [(n+l)/y]
       m2 =  [(n+l)/y] + 1
       [z] = the largest integer less than or equal  to z
       e   =
This method is only appropriate when the desired return period is

less than n/5 years (1) .


     The second method fits the historical  low  flow data to  a

specific probability density function and then computes from this

function the flow whose probability of not being exceeded is 1/y.

The log Pearson Type III distribution is a convenient function to

use because it can accommodate a large variety of distributional

shapes and has seen wide-spread use in stream flow frequency

analysis.  However, there is no physically based rationale for

choosing one distribution over another.


     The xQy low  flow based on the  log Pearson  Type III method is

                       xQy = exp(u + K(g,y)  s)

where u = mean of the logarithms (base e) of the historical annual
           low  flows,
      s = standard deviation of the logarithms of the historical low
           flows,
      g = skewness coefficient of the logarithms of the historical
           low  flows,
      K = frequency factor for skewness g and return period y.
                                 A-l

-------
A sample listing of frequency factors  is given  in  Table  A-l.   These
factors can also be approximated as

                K =  (2/g)  [  (1  + (g z)/6 -  g2/36)3 - 1]

for  g  < 3 where z is the standard normal variate with  cumulative
probability 1/y  (2).  Tables of the normal variates  are  available  in
most elementary statistics texts.  An  appropriate  value  (3)  can  be
found from

                     z =  4.91  [(1/y)-14  -(1-1/y)-14]

     To  illustrate  the use of the  two  xQy  low flow estimation
methods, the data in  Table A-2 will be  analyzed for  the  7Q5.   The
flow values in this table represent the lowest  7-day average  flow
for each year of record.  Also shown are the  rankings  of these flows
from lowest (rank 1)  to highest  (rank  45).  The mean,  standard
deviation, and skewness coefficient of  the logarithms  of these
annual low flow are shown at the bottom of the  table.

     For the distribution-free  approach, the  value of  (n+l)/y is
(45+1)/5 or 9.2.  Therefore, the 7Q5 low flow lies between the 9-th
and 10-th lowest annual flow.  The interpolation factor,  e,  is 9.2  -
9=0.2 Thus we have

                 7Q5  = (1. - .20) X(9)  + (.20)  X(10)
                           = (.80(335)  + (.20) (338)
                           = 335.6 cfs
                                 A-2

-------
     For the  log  Pearson Type III method,  the frequency factor K
will be estimated from Table  A-l.   For  skewness  of 0.409 and a 5-
year return period interpolation  results  in K =  -0.956.   The 7Q5 low
flow is:

                    7Q5  = exp(6.01 + (-.856) ( .24) )
                               = 331.8 cfs

For purposes of comparison, K will  be estimated  using the formulae
given above:

                     z =  4.91  [(0.2) -14-  (1-0.2)-14]
                              = -0.840

                           K=  (2/.409) [l+( .409) (-.840)/5-
(.409)/36)3-l]
                             = -.853

                         7Q5 =  exp(6.01+(-.853) ( .24) )
                            =  331.8  cfs
The difference in the three estimates of  the 7Q5 low flow is less
than 2 percent.
                                 A-3

-------
Table A-l.  Frequency Factors  (K) for the Log Pearson Type  III
Distribution
Skewness
Coefficient
3.
2.
2.
2.
2.
2.
1.
1.
1.
1.
1.
0.
0.
0.
0.
0.
-0
-0
-0
-0
-1
-1
-1
-1
-1
-2
-2
-2
-2
-2
-3
0
8
6
4
2
0
8
6
4
2
0
8
6
4
2
0
.2
.4
.6
.8
.0
.2
.4
.6
.8
.0
.2
.4
.6
.8
.0
Return
5
-0
-0
-0
-0
-10
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
-0
Period,
.636
.666
.696
.725
.752
.777
.799
.817
.832
.844
.852
.856
.857
.855
.850
.842
.830
.816
.800
.758
.758
.732
.705
.675
.643
.609
.574
.537
.499
.460
.420
Years
10
-0.660
-0.702
-0.747
-0.795
-0.844
-0.895
-0.945
-0.994
-1.041
-1.086
-1.128
-1.166
-1.200
-1.231
-1.258
-1.282
-1.301
-1.317
-1.328
-1.336
-1.340
-1.340
-1.337
-1.329
-1.318
-1.302
-1.284
-1.262
-1.238
-1.210
-1.180
                                 A-4

-------
Table A-2.  Annual 7-Day Low Flows  (ft3/sec)  for the Amite River Near
            Denham Springs, LA
Year
1939
1940
1941
1942
1943
1944
1945
1946
1947
1948
1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961





Flow
299
338
355
439
371
410
407
508
450
424
574
489
406
291
352
309
322
278
369
483
523
385
474





Rank
5
10
15
30
20
28
27
38
33
29
41
36
26
4
13
7
8
2
19
35
39
21
34
Year
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983


n = 45
u = 6.0
s = 0.23
q = 0.385
Flow
396
275
392
348
385
335
306
280
354
388
357
499
448
650
356
364
648
619
567
445
349
595






Rank
25
1
24
11
22
9
6
3
14
23
17
37
32
45
16
18
44
43
40
31
12
42





                                 A-5

-------
                             References





1.   Linsley, R.K., et al. Hydrology for Engineers, 2nd Edition.



     McGraw- Hill, New York, NY, 1977.





2.   Loucks, D.P., et al., Water Resource Systems Planning and



     Analysis. Prentice-Hall, Englewood Cliffs, NJ, 1981.





3.   Joiner and Rosenblatt, JASA,  66:394,  1971.

-------
   Appendix B.  An Example Use of DFLOW for Ammonia Discharges From POTWs





     The purpose of this Appendix is to illustrate the use of the DFLOW



program to calculate biologically-based design flows for ammonia and compare



them with the hydrologically-based design flows of 30Q10 for the 13 streams



with the lowest coefficients of variations shown in Table 2-1.






     B.1 Introduction







     As stated in the two-number WQC for ammonia  (1) , a CCC averaging period



of as long as 30 days may be used in situations involving POTWs designed to



remove ammonia where low variability of effluent pollutant concentration and



resultant concentrations in receiving waters can be demonstrated.  In cases



where low variability can be demonstrated,  longer averaging periods for the



ammonia CCC (e.g.,  a 30-day averaging period)  would be acceptable because



the magnitudes and durations of excursions above the CCC would be



sufficiently limited (1).






     B.2 Hydrologically-based Design Flow







     The 30Q10 low flows of the 13 streams with the lowest coefficients of



variation (CV)  are presented in Table B-l.
                                     B-l

-------
      Table B-l.  Design flows and resulting number of excursions using 30-day
                  averaging period (all flows in ft3/sec).
River Name
Quinnipiac
Drowning Cr
Uncompahgre
Greys
Kankakee
Hudson
Shoal
Little Pee Dee
St. Croix
Niobrara
French Broad
Bighorn
Flint
State
CT
NC
CO
WY
IN
NY
FL
SC
WI
NE
TN
MT
GA
Coeff
of
Variation
1.02
0.8
0.86
1.16
0.48
1 . 1
0.95
0.94
0.61
0.59
0.93
0.82
1
30Q10
Flow
42.3
54.7
71
160.7
201.8
288
323.5
366.3
571.8
613.2
636.2
913.6
1000
%Excursions
7.8
8.5
6.9
5.7
10
13.4
10.2
7 .4
16.2
6.4
11.9
8.1
6.4
30 -day 3 -year
Flow
46.5
65.5
77 .3
166.9
213.6
340.7
339
450
598.6
673.6
715.7
1103
1097
%Excursions
15
15
14 .6
9.9
16.7
24.3
12.1
11.8
21.9
8.1
20.3
14.3
9.6
%Diff*
9
16.5
8.2
3.7
5.5
13.5
4.5
18.6
4.5
9
11.1
17.2
8.8
*%Difference = ((30-day 3-year flow)  - (30Q10))  * 100 / (30-day 3-year flow)

      B.3 Biologically-based Design Flow

           The 30-day 3-year flows for 13 streams arc presented in Table B-l.  To
      obtain the biologically-based design flow for these streams, an averaging
      period of 30 days instead of 4  days was entered into the DFLOW program
      (Appendix D (OMITTED).  DFLOW 2.0 has been superseded by newer versions. The
      current versions of DFLOW and its documentation are available online at
      http://www.epa.gov./waterscience/dflow).   Table B-l also includes the number
      of  excursions that occurred in  each of 13 flow records for the
      hydrologically and biologically-based design flows.

      B.4 Comparison of Design Flows

           Table B-l shows that for all  13 streams the 30Q10 low  flow is always
      less than the 30-day 3-year low flow.   The difference between the low flows
      ((30-day 3-year - 30Q10)/30-day 3-year))  3.7% to 18.6% with the mean equal
      to  10.2%.  Because the 30Q10 low flow is  always lower, it results in fewer
      excursions than the 30-day 3-year low flow.
                                           B-2

-------
B.5 Use of Biologically-Based Design Flows for Ammonia Discharges from POTWs

     As stated earlier, an averaging period of 4 days and a frequency of
occurrence of once every three years is used for the CCC.   However,  for
ammonia discharges from POTWs, a longer averaging period may be used in
certain cases.  According to the national WQC for ammonia,  an averaging
period as long as 30 days may be used in situations involving POTWs designed
to remove ammonia where low variability of effluent concentrations and the
resulting concentrations in the receiving waters can be demonstrated.  In
cases where low variability can be demonstrated,  longer averaging periods
for the ammonia CCC (e.g.,  a 30-day averaging period)  would be acceptable
because the magnitudes and durations of excursions above the CCC would be
sufficiently limited.

     In Section 4.1, the hydrologically-based design flows have been
compared with the biologically-based design flows for the 4-day averaging
period for all pollutants.   Appendix B shows a comparison between the
biologically-based 30-day 3-year low flows and the hydrologically-based
30Q10 low flows for 13 streams for ammonia. For these 13 streams,  the 30Q10
flow was always less than the 30-day 3-year flow, by an average to 10.2%.
Thus, the use of the 30Q10 as the design flow is relatively more protective
for these streams.

Reference

1.  US EPA. 1985d. Ambient water quality criteria for ammonia.  1984.  EPA
   440/5-85-001.  National Technical Information Service, Springfield, VA.
                                     B-3

-------
     APPENDIX  C.  Calculation  of  Biologically-Based  Design  Flows


     The biologically-based design flow calculation method  is an

iterative convergence procedure consisting of five parts.  In Part

I, Z (the allowed number of excursions) is calculated.  In Part II,

the set of X-day running averages is calculated from the record of

daily flows.  Because the ambient (instream)  concentration of a

pollutant can be considered to be inversely proportional to stream

flow, the appropriate "running averages" of stream flow are actually

"running harmonic means." (The harmonic mean of a set of numbers is

the reciprocal of the arithmetic mean of the reciprocals of the

numbers).  Thus, "X-day running averages "should be calculated as

X/E  (1/F),  not as (E 5)/X,  where F is the flow for an individual

day.  Throughout this Appendix C, the term "running average" will

mean "running harmonic mean."


     Part III  describes the calculation of N  (the total number of

excursions of a specified flow in the flow record).   The

calculations described in Part III will be performed for a number of

different flows that are specified in Parts IV and V.  In Part IV,

initial lower and upper limits on the design flow are calculated,

the number of excursions at each limit are calculated using Part

III, and an initial trial flow is calculated by interpolation

between the lower and upper limits.   In Part V, successive

iterations are performed using the method of false position (1)  to

calculate the design flow as the highest flow that results in no

more than the number of allowed excursions calculated in Part I.


Part I.  Calculation of allowed number of excursions.


     1-1.   Calculate Z = D/[(Y)  (365.25 days/year)]


          where D = the  number of days in  the  flow  record;

                 Y = the average number of years specified in
                     the frequency and
                                 C-l

-------
                  Z = the allowed number of excursions.

  Part  II. Calculation of X-day  running  averages,  i.e., x-day  running
          harmonic means.
     II-l. Where X = the  specified duration  (in days) of  the average
          period, calculate  the set of  X-day  running averages  for  the
          entice flow record, i.e., calculate an  X-day average
          starting with  day  1,  day 2, day 3,  etc.   Each  average will
          have X-l days  in common with  the next average,  and  the
          number of X-day averages calculated from  the flow record
          will be  (D+l-X).
Part III.   Determination of  the  number of excursions of a specified
          flow in a set  of running averages,  i.e.,  running harmonic
          means.
   III-l.  Obtain  a  specified flow of  interest from either Part IV or
          Part V.
   III-2.  In the  set  of  X-day running averages for the  entire  flow
          record, record the date for which the first average  is
          below the specified flow and  record the number of
          consecutive days that are part of at least one or more of
          the X-day averages that are below the specified flow.
           (Note that whether a  day is counted as  an excursion  day
          does not depend exclusively on whether  the X-day average
          for that day  is below the specified flow  of interest.
          Instead, it depends entirely  on whether that day is  part of
          any X-day average  that is below the specified  flow.  Table
          C-l provides  examples of the  counting of  excursion  days.)
                                  C-2

-------
Table C-l.  Counting excursion days for a specified  flow  of  100  ft3/sec using 4-day averages


Date
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18


Daily
Flow
130
120
110
90
90
100
130
150
70
60
130
90
80
110
100
100
200
500


4 -day
avg
flow
112.5
102.5
97.5
102.5
112.5
112.5
102.5
102.5
87.5
90.0
102.5
95.0
97.5
127.5
225.0
>100
>100
>100

Is the
4 -day
average
below
100?
No
No
Yes
No
No
No
No
No
Yes
Yea
No
Yes
Yes
No
No
No
No
No
Date
Is the date of Number of Number of
part of any Date of Number of start excursion excursions
4 -day avg. start of days in of low days in in low
that is below excursion excursion flow low flow flow
100? period period period period period
No
No
Yes 3 4 3 12 3
Yes
Yes
Yes
No
No
Yes 9 8
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
The daily flows and four-day average flows for days  19  to  200  are  all  above  100  ft3/sec
                                                C-3

-------
       Thus the starting date and the duration  (in days) of the
       first excursion period will be recorded.  By definition, the
       minimum duration is X days.
III-3.   Determine  the starting dates  of,  and number of days in,  each
       succeeding excursion period in the flow record.
III-4.   Identify all of the excursion periods that  begin within 120
       days after the beginning of the first excursion period.
        (Although the first excursion period  is often  the only one  in
       the 120-day period, two or three sometimes occur within the
       120 days.  Rarely do any excursion periods occur during days
       121 to 240.) All of these excursion periods are considered  to
       be in the first low flow period.  Add up the total  number of
       excursion days in the first low period and divide the sum by
       X to obtain the number of excursions  in the first low period.
       If the number of excursions is calculated to be greater than
       5.0, set it equal to 5.0.
III-5.   Identify the first  excursion  period  that begins after  the
       end of the first low flow period, and start the beginning of
       the second 120-day  low flow period on the first day of this
       excursion period.  Determine the number of excursion days and
       excursions in the second flow period.
III-6.   Determine  the starting dates  of  and  the  number of excursions
       in each succeeding  120-day low flow period.
III-7.  Sum the  number of  excursions  in all the  low-flow periods to
                               C-4

-------
           determine  S  =  the  total number of excursions of the specified
           flow  of  interest.
Part IV.  Calculation of initial  limits  of  the design  flow  and  initial
           trial  flow.
   IV-1.  Use L = 0 as the initial lower limit.
   IV-2.  Use U = the xQy low flow as the initial upper  limit.
   IV-3.  Use NL =  0  as the number of excursions  (see Part III)  of the
           initial  lower  limit.
   IV-4.  Calculate Nn  =  the  number of excursions (See Part III) of the
           initial  upper  limit.
   IV-5.  Calculate T = the initial  trial flow as T  =  L  +  (Z  -  NL) (U-L)
                                                             (ND-NL)
Part V.   Iterative convergence to the design  flow.
   V-l.   Calculate NT = the number of excursions  (see Part III)  of the
         trial  flow.
   V-2.   If -0.005 <= (NT-Z)/Z)  <= +0.005, use T  as  the design flow and
   stop.
         If NT > Z, set U = T and Nn  = NT.
         If NT < Z, set L = T and NL  = NT.

   V-3.   If ((U-L)/U)<0.005,  use L as the design flow  and stop.
         Otherwise, calculate a new  trial flow as T  =  L  +  (Z  -  NL) (U-L) ,
         and repeat steps V-l, V-2,  and  V-3 as necessary.     (Nu-NL)

                                 REFERENCE
1.    Carnahan,  B., H.A.  Luther,  and J.O.   Wilkes.   1969.   Applied
     numerical  methods.  Wiley,  New  York.
                                    C-5

-------
              APPENDIX  D.   DFLOW  2.0  User's  Guide  (OMITTED)
NOTE: DFLOW 2.0 has been superseded by newer versions. The current
versions of DFLOW and its documentation are available online at
http://www.epa.gov/waterscience/dflow.
                                 D-l

-------
 APPENDIX  E.  Questions and Answers Concerning the  Bi  ologically-Based
              Method
Q #1: New  aquatic life protection criteria specify that the acute
      criteria (CMC) and the  chronic criteria  (CCC)  may be exceeded
      no more than once every three years on the  average by 1-hour
      and  4-day averages, respectively.  They  also state that
      extreme value analyses  may not be appropriate- for estimating
      the  ambient exposure  condition.  What is  an extreme value
      analysis?

A.    This is a very broad  question.  There are many types of
      extreme value analyses.   But all extreme  value analytical
      techniques have something in common.  Let's consider a time-
      series  of daily flow  data in order to explain extreme value
      techniques.

      A low-flow water year starts on April 1  of  each year and ends
      on March 30 of the following year.  If we perform an extreme
      value analysis for a  4-day average condition we should
      estimate 4-day running  averages for each  water year, then
      determine which running average is the lowest (extreme) for
      each water year.  Finally,  we rank the extreme value of each
      year for frequency analyses.
Q #2: Would you explain how  running averages  are  estimated?

A.    Starting with April  1,  our first running  average will be the
      arithmetic mean of flow data for April  1, 2,  3 and 4: the
      second running average will be the arithmetic mean of April 2,
      3,  4  and 5; and the  third running average will be the 3,4,
      etc.   Thus, there will be 362 4-day running averages for each
      water year of 365 days.
Q #3: By  extreme value, do you  mean lowest running average of the
water year?

A.    In  low-flow analyses,  the extreme value  for a water year  is
      the lowest running average for that year.

Q #4: So,  do I have 30 extreme  values from 30  years'  flow record
      considering one extreme value for each water year?

A.    Exactly.

Q #5: You said about ranking the extreme values.   How do you rank
      them and why do you rank  them?

A.    For low flow analysis, ranking can be done  from lowest to
      highest.  For a low-flow  analysis of a 30-year flow record,  we
 The biologically-based design flow method has been supported by an overwhelming
majority of water quality coordinators at Regional and Headquarter  levels.  But  the
method, being totally new, tends to  raise a lot of questions which  we have heard over
time  from many reviewers.  Some of these questions and related answers are listed here
for additional clarification to Appendices C and D of the Guidance.  If this paper
becomes too long, in a way it defeats its purpose.  So we chose questions based  on
their importance.  We encourage our  readers to be critical about our answers and raise
other questions which they may consider important.  This will help  us to improve both
the method itself and its presentation. In this context, readers may contact Hiranmay
Bisuas (FTS-382-7012)  or Nelson Thomas (FTS-780-5702).
                                   E-l

-------
      have 30 extreme values.  If we rank them from the lowest to
      the highest value, and no two extreme values are equal, then
      we have one value for each of 30 ranks, and the return period
      of the first ranked low is approximately 30 years, and that of
      the 10th ranked flow is approximately 3 years.

Q #6: The frequency analysis using the ranked extreme values seems
      to be quite straight forward.  Why are various kinds of
      distribution used for frequency analysis?

A.    If we are concerned with a prediction of low flow for a return
      period that is equal or less than the flow record, then we
      will not have to use any distribution at all.  The
      distribution-free, or non-parametric technique is the best for
      frequency analyses.  But, suppose you need 100- 200- or 500-
      year flood and drought forecasts for the design of a dam (for
      use power production and irrigation) and we do not have a flow
      record of such a long period; then, we need to use some form
      of distribution to extrapolate to 100, 200 or 500 years.
      There are many well known distributions which can be chosen on
      a case-by-case basis.

Q #7: The new WQC also make some reference to the Log-Pearson Type
      III distribution as an example of the extreme value analysis.
      While we are on the subject of distribution, is it the only
      distribution that is currently in use in the water quality
      analytical field?

A.    The United States Geological Survey uses the Log-Pearson Type
      III distribution in low-flow as well as flood-flow analyses.
      They made this choice after conducting a study of flood flow
      analyses using various other techniques.  The choice of
      techniques should be based on the nature of the distribution
      of extreme values.  But, for national consistency of
      estimates, the USGS chose this technique.

Q #8: Extreme value analytical techniques are often used in the
      hydrologic field, and seem to be quite reasonable.  Is there
      any biological/ ecological reason why extreme value analyses
      are not appropriate for estimating design flow using the
      ambient duration and frequency of the new WQC?

A.    Yes, a direct use of extreme value analyses is not appropriate
      because biological effects are cumulative.

Q #9: Would you elaborate how the cumulative nature of biological
      effects is related to extreme value analyses?

A.    In extreme value analytical techniques, only the most extreme
      drought exposure event is considered, but other, less severe
      within-year exposure events are totally ignored, although
      their cumulative efforts could be severe.  The severity of
      those smaller within-year exposure events of extreme drought
      conditions that are ignored may outrank in severity the
      extreme exposure events of other less-than-most severe drought
      conditions.  Since the biological effects are cumulative we
      must find a way to account for all within-year exposures in
      addition to the most extreme event of each year.

Q #10: Your answer is difficult to follow; would you give an
        example?

A.      Hydrologists  know that  we  had,  in  various  parts  of  the US,
        extreme  drought  events  during  the  water  years  1925-1932,
        1955-1956,  and  during a  few years  in the  late  seventies.   In
        other years,  drought  was not as  severe.   Suppose  that  in
                                 E-2

-------
        water  year  1925,  there  were 4  very low 4-day running averages
        of  which only one was acceptable as the extreme value of that
        year;  the 2nd,  3rd,  and the 4th values were ignored.   Similarly,
        one extreme value was estimated for each of the other Water
        years.   But,  some of the extreme values of other water years
        are less severe  than 2nd, 3rd or the 4th running averages of the
        year 1925.   Thus,  by ignoring  these 3 running averages of the
        water  year  1925,  the extreme value method has ignored
        potential severe effects that  may result from those exposure
        events.   In addition, the inclusion of other extreme values
        that are less severe than the  2nd,  3rd, and the 4th running
        averages of the  year 1925,  and exclusion of more severe
        excursion events (2nd, 3rd and 4th excursions of water-year
        1925)  result in  a skewed estimate of low flow.
Q #11: The method described to implement the two-number aquatic  life
        criteria  is  called a  biologically-based method.   What is
        biological about  it?

A.      Almost  every parameter that  is  used in  this method is derived
        on  the  basis of  either biological,  toxicological  or
        ecological considerations.   Whereas the parameters used in
        the extreme  value analysis  are  unrelated to biological,
        toxicological or  ecological  consideration.
Q #12:  Would you name the things that you think are biological,
        toxicological,  or ecological  in nature?

            - Durations of acceptable exposure conditions: 1 hour
              for  CMC  and 4  days  for  CCC are  biologically derived.

            - 3 years on the average is the allowed ecological
              recovery period after a single  excursion  (see Table D-2
              of Appendix D  of  the Technical  Support  Document for
              Water  Quality-based Toxics Control  (TSD)).

            - 15 years is selected for ecological recovery after a
              total  of 5  or  more  excursions within  a  low flow period
              (see reference Table D-2  in Appendix  D  of  TSD).
Q #13:  I see neither 15 years nor 5 exposure events in the
        referenced Table  D-2.   Could you explain the discrepancy?
                                 E-3

-------
A.    It is true that neither 15 years nor 5 excursions are found in
      the reference Table.  But what is available is that rivers and
      streams are fully recovered 5 to 10 years after a severe
      exposure event.  Aquatic biologists consider that repeated
      within-year exposures can result in catastrophic affects.  In
      their judgment, 10 years exposure interval is inadequate
      because under that situation the ecology of the receiving
      system will be under constant stress and recovery.  By the
      same token, a 20-year interval was considered to be
      unnecessarily stringent for attaining healthy biota.  After
      these considerations and debates among biologists and
      wasteload allocation coordinators, we decided to use 15 years
      as an acceptable interval after a severe exposure event
      consisting of several within-year exposures.
Q #14: Have you anything to say about how you decided to allow 5
        excursions  in  an  interval  of  15  years?

A.      WDC  allow an excursion  once every  three  years  on  the  average.
        Since  the effects of  excursions  are  cumulative, ecological
        recovery from  a severe  exposure  event requires about  15  years
        and  the recovery  period from  a single exposure event,
        according to the  national  WDC is 3 years.   Therefore  15/3 or
        5  excursions are  accepted  as  the upper limit of within-year
        excursion counts.
Q #15: Why did you not choose a 12-year interval for 4 within-year
        exposure  events?  Or  could  you  not  choose  an  18-year  interval
        for  6  within-year exposure events  (based  on  the  information
        available in  Table D-2  of  TSD) ?

A.      One  could make  various  other choices  based on  site-specific
        knowledge but we  made our  choice  for  average conditions.


Q #16: If 12- or 18- year intervals are chosen for 4 or 6 within-
        year exposure conditions,  would the design flow  be different
        from that of  the  15-year interval  choice? Do we  have any  idea
        about  how different  the CCC or CMC flow will be  for  the
        choices of  12-  or 13-year  interval?

A.      No,  we did  not  perform  such analyses  or comparisons  but our
        guess  is  that the difference will  not be  substantial.

Q #17: It is understood that,  if a 15-year interval  is chosen for
        ecological  recovery, then  5 within-year exposures may be
        allowed because WQC  specify 1  exposure on the  average of
        every 3 years.  But  some extreme  drought  related low flow
        periods might include less than 5  within-year  exposures,  and
        some more severe  low flow  periods  include more than  5 within-
        year exposures.   If  exposure effects  are  cumulative,  why  not
        include all exposures within a year,  why  limit it to 5?

A.      The  biological  method accounts for all within-year excursions
        when the  number of excursions  during  a low-flow  period is 5
        or less.   So, 5 is the  upper limit, and the  lower limit is 1.
Q #18: What if the within-year excursions for a given flow based on
        the  biological  method  is  naturally greater  than  5  during  say,
        a  50-  or  100-year  drought?  In  those  years,  flows may  remain
        low  for a long  time, such as for  40-50  days,  not necessarily
                                 E-4

-------
        for  just  20  days  for  5  excursions.   After  all,  we  cannot
        change  nature,  can  we?

A.     No, we cannot change nature.  But we can modify our approach
       to suit our objective after understanding the consequences of
       severe events.

       We made a number of analyses to find out what happens if we
        account  for  all,  not  just  5,  excursions  that  one may  expect
        from those most severe  drought  years.  We  found that
        inclusion of all  excursions from  those years  results  in the
        following:

            - Design flows of all return periods of say 3, 5, 10,
              20, 50 years, etc.  are  completely  dominated  by  those
              most severe drought  years;  and

            - This leads to extremely stringent design flows.
Q #19: There is nothing biological in these analyses.  Since the
        exposure  effects  are  cumulative,  should  we  not  count  all
        exposures  regardless  of  how rarely  one may  expect  them, or
        how  stringent  the resulting design  flow  is?

A.      This  is where  a  little understanding  of  ecological recovery
        and  familiarity  with  the North American  aquatic life  are
        necessary  to make a reasonable choice.   The upper  bounds  of
        the  life  cycles  and life spans of most North American aquatic
        species are 2  and 10  years,  respectively.   An exposure  event
        of 20- or  50-year interval  may not  be meaningful,
        particularly when one considers  other ways,  for example
        recruitment from the  surrounding ecosystem,  in  which  recovery
        may  take place.   So,  in  our judgment, a  recovery period of 15
        years  is  adequate for situations where the  number  of
        exposures  in a low flow  period is 5 or more.
Q #20: What is described here in the biological method is similar to
        what  is  done  by  hydrologists  for  partial  duration  series.
        They  address  the problem using  traditional  statistical
        approach.   Why did  you not  use  a  classical  statistical
        method?

A.      First, the  statistical science  of partial duration series,
        particularly  in  the hydrologic  field,  is  not  well  developed.
        Not many people  understand  it.  Although  the  biological
        method lacks  statistical elegance,  it  is  simple  and can  be
        used  and understood by field  biologists and engineers, alike.
        We would not  be  surprised if  a  statistician comes  up with a
        better statistical  answer for the problem that we  have in
        hand.  But  it would be important  for the  regions to
        understand  most  aspects  of  the  method  if  we expected them to
        use it.

Q #21: Over the last 20-25 years, the majority of the states in the
        U.S.  used the 7Q10  low flow as  the design flow for what  we
        essentially had  as  a not-to-be  exceeded single number WQC
        value. It seems  that it  worked  fine, although a  rationale for
        such  a choice is hard to come by.  Why  is  it so important now
        to have  a rational  biolocially-based method to implement the
        two-number  WQC?

A.      It is  important  to  provide  a  rational  method  for three major
        reasons.  First,  lack of a  biologically-based method in  the
        past  led to the  adoption of design flows  such as 3Q20, 7Q10,
        30Q10, 30Q2,  and even the annual  average  flow for  identical
                                 E-5

-------
        water  use.  A  technically  defensible  method  will  bring  about
        technical  consistency  for  any  desired level  of  protection.
        Second,  the introduction of  the  two-number national  WQC,
        whole  effluent toxicity, and the guidance on site-specific
        water  quality  standards have unalterably changed  the
        environment of toxics  control.   In  these situations,  a
        biologically-based  method  is necessary that  can be applied
        not  only to national two-numbered WQC,  but also to other
        sites-and  use-specific durations and  frequencies  of
        pollutants and whole effluent  toxicities.  Third, since WQC
        and  their  field use have become  complex, it  is  very  important
        that we  develop a simple method  that  is easily  understandable
        to field biologists and engineers alike.  In the  past,  very
        few  understood the  relation  between the WQC  and the
        corresponding  7Q10  or  other  xQy  design flow.
Q #22: Why is the biologically-based method considered to be more
        directly  based  on  the  water  quality  criteria  than  the
        hydrologically-based method?

A.      In  the biologically-based  method,  both  the  averaging period
        and the frequency  (for example,  4  days  and  3  years)  are  taken
        directly  from the  criterion,  whereas in the hydrologically-
        based approach,  the two number  in  xQy are not.  Most of  the
        other aspects of the biologically-based approach are also
        based on  biological, ecological, and toxicological
        considerations.  One of the  major  technical differences
        between the  methods is that  the  3  years in  the  biologically-
        based method is  an average frequency, whereas the  10 years in
        the hydrologically-based approach  is a  return period.

Q #23: Does it make any difference whether biologists,  ecologists,
        and toxicologists  understand how design flow  is calculated?

A.      Yes,  for  three  major reasons.   First, these are the  people
        who derive the  aquatic life  criteria.   If the criteria are
        not used  in  a manner that  is consistent with  their
        derivation,  the  intended level  of  protection  will  probably
        not be achieved.   Second,  site-specific frequencies  and
        durations will  not correctly affect  design  flow if the
        duration  and frequency are not  directly used  in the
        calculation.  Third, if they understand what  parameters
        affect design flow, biologists,  ecologists, and toxicologists
        can gather data  that might allow them to refine their
        estimates of such  values as  one  hours,  four days,  three
        years, and fifteen years.
                                 E-6

-------
Q #24: Let us discuss the simplicity of the biologically-based method.
        I  am  not  clear how  an  excursion  is  counted.   Would  you  explain
        how you count excursions  and  estimate  design  flows?

A.      This  is the  key  to  understanding the biologically-based
        method.   Since the  stream flow  is  inversely proportional  to
        instream  concentration, any consecutive  4-day average of  low-
        flow  that is lower  than the design  flow  is counted  as one
        excursion of the CCC.  The following is  the step-by-step
        explanation  of how  excursions are  counted  in  estimating x-day
        y-year design flow:

        1. An excursion  period is defined  as a sequence  of  consecutive
          days where each  day belongs  to  an x-day average  flow that
          is below  the  design flow.  For  example, if the three
          running averages of a  consecutive 6-day period are less
          than the  4-day 3-year  design  flow,  then those 6  days belong
          to an  excursion  period.

        2. The number of excursions in  an  excursion period  is the
          length of the period divided  by  the criteria  averaging
          period.   For  example,  if an  excursion period  is  6 days
          long,  then the number  of excursions for the 4-day averaging
          period for CCC is 6/4  or 1.5.

        3. The total number of excursions  is limited  to  5 within  a low
          flow period.   Usually  a low  flow period lasts 120 days or
          less.   In some rare stream situations, more than one low
          flow period within  a water year  is  possible.

        4. The allowed total number of  excursions over the  period of
          record is the number of years of record divided  by the
          frequency of  aquatic life  criteria  (3 years for  the  CCC of
          the new national two-number  criteria).  For example, if we
          have a 30-year flow record,  then total number of excursions
          that are  allowed for x-day 3-year criteria is equal  to 30/3
          or 10.

        5. The 4-day 3-year design flow  for the  4-day 3-year CCC  based
          on a 30-year  flow record of  a given river  is  equal that
          flow which results  in  no more than  the allowable number of
          excursions.   For example,  the total allowable number of
          excursions for the  given record  is  10.  The design flow is
          the highest flow that  results in no more than 10 excursions
          calculated as defined  in steps  1 through 4 above.

Q #25.  Let  us  take the example printout (from page D-5) for the Amite
        River as  presented  below.  Will  you explain the  procedure
        using this example?

  A.     As shown  in  the  following printout, we have a flow  record from
        1937  to 1983 which  is  approximately 42 years.  Since we are
        allowed to have  no  more than  one excursion in every 3 years on
        the average,  we  have 42/3 or  about  14  excursions.   In October
        1952, we  encountered the  first  excursion for  a continuous
        period of 6  days.   Thus,  we calculate  6/4 or  1.5 excursions
        for that  low flow event.   The next  excursion  period occurs,
        starting  from October  10,  1956,  for 30 consecutive  days.
        Since the upper  limit  of  excursions in a low  flow period  (a
        low flow  period  is  usually 120  days long)  is  5,  we
                                  E-l

-------
        obtained  a  total  of  5  excursions  only,  although  in  reality
        there  were  altogether  30/4  or  7.50  excursions  in that  low flow
        period. Similarly, we  found only  5  excursions  for total  period
        of  30  days  during the  low flow period  of  1963.

Q #26: It seems like the accuracy of the design flow estimates is
        totally   dependent on  the length  of the flow record.   Do you
        agree  with  this observation?

A.      Absolutely.   This is true about any analysis.  More relevant
        data are  necessary to  provide  more  accurate  information.

Q #27: What minimum length of flow record  is recommended?

A.      The longer  the  flow  record,  the more reliable  the estimated
        design conditions will be.   Figure  E-l  (OMITTED)  shows how the
        spread in the  90% confidence limits on  the extreme  value-based
        design load with  10-year  return period  decreases with
        increasing  period of record.   (This figure was derived on the
        basis  of  lognormal statistics,  not  log  Pearson type 3).
        Results are shown for  both  low variability  (CV=0.2)  and  high
        variability (CV=0.8) situations.  Based on the behavior  of
        these  curves,  it  appears  that  20  to 30  years of  record is a
        reasonable  minimum requirement for  extreme value analysis at a
        10-year return  period.

        The case  for the  biologically-based excursion  criterion  is
        less definitive.  However,  since  it considers  all days within
        the period  of  record as its sample  (not just the worst
        condition of each year),  its sample size  is  much larger  than
        that of an  extreme value  analysis.   Thus, it may be possible
        to  use periods  of record  less  than  20  years  with this
        criterion and  still  have  a  good level  of  confidence in the
        results.
                                  E-2

-------
Q #28: What would you do for intermittent streams where low flow is
        zero  during  low  flow periods?  Also,  how will  you use  the
        biologically-based  method  in  situations where flow data are
        not available?

A.      These are  problems  that  are generic  to  all  flow estimating
        techniques.   For intermittent  streams for which the low flow
        is  zero, the design flows  for  CMC  as well as  CCC are  equal to
        zero.   In  situations where flow data are not  available, field
        hydrologists and engineers sometimes use flow data from
        hydrologically comparable  drainage basins.


Q #29: The table given  in Question 23 looks simple.  How much time
        does  it  take to  conduct  a  biologically-based  analysis  for any
        stream of  interest?

A.      The analysis is  performed  in  two steps.  First, daily  flow
        data  are retrieved  from  the daily  flow  file in STORET, by
        submitting a batch  job.  This  will take a few minutes  of time
        at  the computer. However,  the job run  might  take  anywhere
        from  a few minutes  to several  hours, depending on  how  busy the
        computer system  is  at the  time of  submittal.   Once the data
        has been retrieved,  the  analysis can be performed  in  five or
        ten minutes.
Q #30: It seems that the foundation of the information about
        ecological  recovery  periods  for  the  two-number  WQC  is  all  that
        are  listed  in  Table  D-2  of the TSD.   But,  anybody familiar
        with these  references  will tell  you  that  the  recovery  periods
        listed  in that  table are related to  recovery  from catastrophic
        exposures caused by  spills,  not  by effluents  of malfunctioned
        advanced treatment facilities.   Would you agree that this  is
        not  a satisfactory set of information to  make such  an
        important decision?
                                  E-3

-------
A.      This  is  the  best  available  information  that  we  could use  to
        estimate ecological  recovery.   Considering the  complexities
        involved in  the  implementation  of  the two number  WQC,  and the
        site-specific  WQC for  pollutants and whole effluent  toxicity,
        we  could not leave the recovery question open to  anyone's
        interpretation.   Considering  the potential for  misuse of  the
        WQC in their implementation phase,  we had to use  our best
        judgment and the  best  information  available, although we
        recognize that our best judgment would  be debatable.   Since
        the information  base is not as  strong we want to  have,  in
        keeping  with the  Agency policy  and legal background,  we had to
        go  in the direction  of protection  in the over-all decision
        making process.

Q #31: What are you doing to  improve the information base?

A.      ORD is planning  to undertake  a  major effort  before the next
        update of the  WQC.   But,  this is an area in  which success is
        dependent more on cooperative efforts in which  field
        biologists,  ecologists,  toxicologists,  engineers  and
        hydrologists share their experience than doing  mere  literature
        reviews  and/or gathering laboratory-generated information.
                               REFERENCE


1.  Stedinger, J.R., 'Confidence Intervals for Design Events', Jour.

    Hyd.  Eng.  Div.,  ASCE,  Vol.  109,  M.  1,  January  1983.
                                  E-4

-------
                              DISCLAIMER

We have made efforts to ensure that this electronic document is an accurate
reproduction of the original paper document. However, this document does not
substitute for EPA regulations; nor is it a regulation itself. Thus, it does not and
cannot impose legally binding requirements on EPA, the states, tribes or the
regulated community, and may not apply to a particular situation based on the
circumstances. If there are any differences between this web document and the
statute or regulations related to this document, or the original (paper) document,
the statute, regulations, and original document govern. We may change this
guidance in the future.

Supplemental material such as this disclaimer, a document abstract and glossary
entries may have been added to the electronic document.

-------