PB82-260845
River Basin Validation  of the Water Quality
Assessment Methodology  for Screening
Nondesignated  208 Areas.  Volume II
Chesapeake-Sandusky  Nondesignated 208 Screening
Methodology Demonstration
Tetra Tech, Inc,
Lafayette, CA
Prepared for

Environmental Research  Lab.
Athens, GA
May 82
                    U.S. DEPARTMENT OF COMMERCE
                  National Technical Information Service

-------

-------
                                               EPA 600/3-82-057b
                                               May 1982
RIVER BASIN VALIDATION OF THE WATER QUALITY ASSESSMENT
   METHODOLOGY FOR SCREENING NONDESIGNATED 208 AREAS

     Volume II:   Chesapeake-Sandusky Nondesi.gnated
        208 Screening Methodology Demonstration
                          By

                     J.  David Dean
                      Bob Hudson
                   William B.  Mills

                   Tetra Tech, Inc.
               3746 Mt.  Diablo Boulevard
             Lafayette,  California  94549
                Grant  No.  R806315-01-0
                   Project Officer

                  Robert  B.  Ambrose
    Technology  Development and  Applications  Branch
          Environmental Research  Laboratory
               Athens, Georgia  30613
          ENVIRONMENTAL  RESEARCH  LABORATORY
          OFFICE  OF  RESEARCH  AND  DEVELOPMENT
         U.S.  ENVIRONMENTAL PROTECTION AGENCY
               ATHENS. GEORGIA  30613

-------

-------
                                  TECHNICAL REPORT DATA
                           (Please read Instructions on the reverse before completing)
1. .REPORT NO.
   EPA-600/3-82-057b
                                                          3. RECIPIENT'S ACCESSION1 NO.
                               ORD  Report
4. TITLE ANO SUBTITLE
River Basin Validation  of the Water Quality Assessment
Methodology for Screening Nondesignated 208 Areas,
Volume II:  Chesapeake-Sandusky Nondesignaged 208	
                                                          5. REPORT DATE
                                                                  Mav 1982
                                                          6. PERFORMING ORGANIZATION CODE
            Screening Methodology  Demonstration

 J.  David Dean, Bob Hudson, and William B.  Mills
                                                          8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Tetra Tech,  Inc.
 3746 Mt.  Diablo Boulevard
 Lafayette, California
                                                          10. PROGRAM ELEMENT NO.
                                                             ACUL1A
                                                          11. CONTRACT/GRANT NO.
                                                             R806315-01-0
12. SPONSORING AGENCY NAME AND ADDRESS     „  ,
 Environmental Research Laboratory—Athens GA
 Office of Research and Development
 U.S.  Environmental Protection  Agency
 Athens, Georgia  30613
                                                          13. TYPE OF RE
                                                             Final,
                                                                               IOD COVERED
                                                          14. SPONSORING AGENCY CODE
                                                             EPA/600/01
15. SUPPLEMENTARY NQTES  .     „            „                        ,  ,     .
 River Basin Validation of  the  Water Quality Assessment Methodology  for  Screening
 Nondesignated 208 Areas, Volume  I:  Nonpoint Source Load Estimation
16. ABSTRACT
       In earlier work under  the  sponsorship of EPA, a screening methodology was pro-
 duced by Tetra Tech, Inc., for assessing water quality problems in  areas  not covered
 under Section 208 of the  Federal  Water Pollution Control Act Amendments  of 1972, and
 loading functions were developed  by Midwest Research Institute  (MRI)  for estimating
 the quantities of different  diffuse loads entering receiving waters from nonpoint
 sources.  The two methods  had  never been applied together under realistic conditions,
 however, to demonstrate how  the  combined techniques might be used for identification
 of water quality problems  in  U.S.  rivers..  In this volume, the successful application
 of the Tetra Tech-developed  nondesignated 208 screening methodology under field condi-
 tions in five river basins is  described, and the compatibility with the  nonpoint sourcf
 calculator is demonstrated.  Outputs from the nonpoint source calculator were easily
 adapted and in some cases  used directly in the mass balance equations of the screening
 methods.  Loadings predicted with the nonpoint source calculator  in conjunction with
 mass balance techniques employed  by the screening methods provided  reasonably accurate
 predictions of instream,  lake, and estuary water quality constituent concentrations.
 Volume I describes the application of the MRI-developed nonpoint  loading procedures in
 the same river basins  (Sandusky  River in Ohio and the Patuxent, Chester,  Occoquan, and
 Ware Rivers in the Chesapeake  Bay Basin).
17.
                               KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                             b.lDENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Group
18. DISTRIBUTION STATEMENT


 RELEASE TO  PUBLIC
                                             19. SECURITY CLASS (This Report)
                                                UNCLASSIFIED
21. NO. OF PAGES
     245"
                                             20. SECURITY CLASS (Thispage)
                                             .   UNCLASSIFIED
                                                                       22. PRICE
EPA Form 2220-1 (9-73)

-------
                     NOTICE

Mention of trade names or commercial products does not
consititute endorsement or recommendation for use.

-------
                                  FOREWORD

      As environmental controls become more costly to implement and the
penalties of judgment errors become more severe, environmental quality
management requires more efficient analytical tools based on greater knowledge
of the environmental phenomena to be managed.  As part of this Laboratory's
research on the occurrence, movement, transformation,''impact, and control of
environmental contaminants, the Technology Development and Applications
Branch develops management and engineering tools to help pollution control
officials achieve water quality goals through watershed management.

      In earlier work sponsored by EPA, water quality assessment techniques
were developed for characterizing pollution problems in nondesignated 208
areas,and loading functions were developed for estimating quantities of
different pollutants entering receiving water bodies from nonpoint sources.
It appeared that these two tools used in concert might provide an adequate
set of methods for screening nondesignated areas  using simple hand calcu-
lation procedures.  This report describes the application of both methods
to the identification of water quality problems in several river basins
in the United States.

                                       David W. Duttweiler
                                       Director
                                       Environmental Research Laboratory
                                       Athens, Georgia

-------
                                  ABSTRACT

      In earlier work under the sponsorship of EPA, a screening methodology
was produced by Tetra Tech, Inc., for assessing water quality problems in
areas not covered under Section 208 of the Federal  Water Pollution Control
Act Amendments of 1972, and loading functions were developed by Midwest Re-
search Institute (MRI) for estimating the quantities of different diffuse
loads entering receiving water bodies from nonpoint sources.  The two methods
had never been applied together under realistic conditions, however, to
demonstrate how the combined techniques might be used for identification of
water quality problems in U.S. rivers.

      In this volume, the successful application of the Tetra Tech-developed
nondesignated 208 screening methodology under field conditions in five river
basins is described, and the compatibility with the nonpoint source calcula-
tor is demonstrated.  Outputs from the nonpoint source calculator were easily
adapted and in some cases used directly in the mass balance equations of the
screening methods.  Loadings predicted with the nonpoint source calculator
in conjunction with mass balance techniques employed by the screening methods
provided reasonably accurate predictions of instream, lake and esturay water
quality constituent concentrations.  Volume I describes the application of
the MRI-developed nonpoint loading procedures in the same river basins
(Sandusky River in Ohio and the Patuxent, Chester,  Occoquan, and Ware Rivers
in the Chesapeake Bay Basin).

       This report was submitted in fulfillment of Grant No. R806315-01-0 by
Midwest Research Institute under the sponsorship of the U.S. Environmental
Protection Agency.  The report covers the period September 1979 to March 1981,
and work was completed as of November 1981.
                                     IV

-------
                              CONTENTS

Figures 	viii
Tables  	   x
     1.  Introduction 	   1
         1.1  Background	   1
         1.2  Purpose and Scope	   2
         1.3  Format and Organization 	   3
     2.  Results and Conclusions  	   5
         2.1  General Conclusions 	   5
         2.2  Rivers and Streams	10
         2.3  Impoundments	  12
              2.3.1  Stratification	12
              2.3.2  Sedimentation	  12
              2.3.3  Eutrophication	  13
              2.3.4  Dissolved Oxygen 	  13
         2.4  Estuaries	14
              2.4.1  Classification .•	  14
              2.4.2  Flushing Calculations  	  14
              2.4.3  Pollutant Distributions  	 ....  15
              2.4.4  Eutrophication	16
         2.5  The Sandusky River	17
         2.6  The Chester River	17
         2.7  The Patuxent River	19
              2.7.1  Riverine Portion 	  19
              2.7.2  Estuarine Portion  	  19
         2.8  The Ware River	.'  20
         2.9  The Occoquan Reservoir	21
     3.  Demonstration of Methods	23
         3.1  Study Area Description	23
              3.1.1  The Sandusky River	23
                     3.1.1.1  Water Quality 	  25
              3.1.2  The Chester River	  27
                     3.1.2.1  Water Quality 	  29

-------
     3.1.3  The Patuxent River	   29
            3.1.3.1  Mater Quality  	   33
     3.1.4  The Ware River	   34
            3.1.4.1  Water Quality  	   34
     3.1.5  The Occoquan River	   34
            3.1.5.1  Water Quality  	   36
3.2  Demonstration Example:  The Sandusky River 	   37
     3.2.1  Data Collection	   37
     3.2.2  Data Reduction and Supplementation	   39
     3.2.3  River Segmentation	   43
            3.2.3.1  Low Flow	   43
            3.2.3.2  High Flow	   45
     3.2.4  Temperature Profiles  :	   45
     3.2.5  Estimation of BOD Decay  Coefficients and
              Reaeration Coefficipnts   	   50
     3.2.6  BOD Mass Balance	'. ".'	   50
     3.2.7  Dissolved Oxygen Profiles 	   52
     3.2.8  Fecal Coliform Mass Balance 	   59
     3.2.9  Sediment Mass Balance	   61
     3.2.10 Nitrogen and Phosphorus  Balance 	   66
3.3  Demonstration Example:  The Chester River  	   74
     3.3.1  Data Collection	   74
     3.3.2  Data Reduction and Supplementation	   76
            3.3.2.1   Hydro!ogic and Hydraulic Data ....   76
          .  3.3.2.2   Water Quality  Data  	   79
     3.3.3  Point Source Load Estimates 	   86
     3.3.4  Estuarine Classificaion  	   89
     3.3.5  Flushing Calculations 	   93
     3.3.6  Pollutant Distribution  	   98
            3.3.6.1  Low Flow	   98
            3.3.6.2  High Flow	100
     3.3.7  Eutrophi cation	  109
3.4  Demonstration Example:  The Patuxent River 	  112
     3.4.1  Data Collection	  113
     3.4.2  Data Reduction and Supplementation	114

-------
              3.4.3  Fresh Non-Tidal Waters  	  117
                     3.4.3.1  Temperature Profiles 	  120
                     3.4.3.2  Estimation of Reaeration and Deoxy-
                                genation Coefficients-'.	12:1
                     3.4.3.3  BOD Mass Balance	123
                     3.4.3.4  Dissolved Oxygen Profiles  	  126
                     3.4.3.5  Total Coliform Routing 	  134
              3.4.4  Estuarine Waters  	  135
                     3.4.4.1  Flushing Times 	  135
                     3.4.4.2  Pollutant Distribution 	  138
                              3.4.4.2.1  Low Flow	138
                              3.4.4.2.2  High Flow	147
                     3.4.4.3  Estuarine Eutrophication 	  155
         3.5  Demonstration Example:  The Ware River 	  155
              3.5.1  Data Collection	  157
              3.5.2  Data"Reduction and Supplementation-  	  157
              3.5.3  Estuarine Analysis of Fox Mill  Run	158
              3.5.4  Ware River Estuary Flushing Times 	  166
              3.5.5  Pollutant Distribution in the Ware River  .  .  .  166
              3.5.6  Eutrophication	171
                     3.5.6.1  Nutrient Limitation  	  171
                     3.5.6.2  Light Limitation 	  172
         3.6  Demonstration Example:  The Occoquan Reservoir ....  173
              3.6.1  Stratification	174
              3.6.2  Sedimentation	  184
              3.6.3  Eutrophication	  196
              3.6.4  Water Quality High Flow Events	204
              3.6.5  Dissolved Oxygen	213
Bibliography 	  224
Appendix	227

-------
                               FIGURES

Number                                                               Page
3.1-1   The Sandusky River basin	      24
3.1-2   The Chester River basin	      28
3.1-3   The Patuxent River basin	      32
3.1-4   The Ware River basin	      35
3.1-5   The Occoquan River basin	      36
3.2-1   Cross-sectional stream profile of the Sandusky River
          near Fremont	      41
3.2-2   Observed and predicted temperatures in the Sandusky River
          (low flow)	      48
3.2-3   Calculated low flow temperature profile for Spring Run .  .      49
3.2-4   Computed vs. historical dissolved oxygen for the
          Sandusky River 	      57
3.2-5   Dissolved oxygen and temperature profiles for Honey Creek      60
3.2-6   Sediment rating curve for Sandusky River near Fremont,
          October 1-31, 1976	      64
3.2-7   Predicted and observed suspended sediment concentrations
          for the Sandusky River af Bucyrus	      67
3.2-8   Observed and predicted total  phosphorus concentrations
          for the Sandusky River at Mexico	      71
3.2-9   Observed and predicted inorganic nitrogen in the
          Sandusky River 	      75
3.3-1   Frequency analysis of 7-day annual low flows 	      77
3.3-2   Salinity profile for the Chester River, June 30, 1972  .  .      80
3.3-3   Salinity profile for the Chester River, October 30, 1972  .      81
3.3-4   High and low vertically averaged salinities in the
          Chester River estuary  	      83
3.3-5   Empirical relationship between the ratio of modified
          tidal prism and tidal prism methods and mean low tide
          estuary volume 	      97
3.3-6   Schematic of Chester River and point sources (not to scale)     99
3.3-7   Observed and predicted total  nitrogen profiles in the
          Chester River estuary  	     108
3.4-1   Reach segmentation schematic for the Patuxent River  . .  .     11.8
3.4-2   Observed and predicted BOD5 in the Patuxent River  ....     1:27

-------
                      FIGURES (Cont'd)
Number                                                               Page
3.4-3   Observed versus predicted dissolved oxygen profiles
          for the Patuxent River	    130
3.4-4   Frequency histogram of 7-day moving average flows  ....    140
3.4-5   Predicted and observed total nitrogen and observed
          chlorophyll-^ in the Patuxent River, September 27, 1978     142
3.4-6   Predicted and observed total phosphorus in the Patuxent
          River, September 27, 1978	    143
3.4-7   Observed and predicted total nitrogen and observed
          chlorophyll-a in the Patuxent River, July 19, 1978 ...    145
3.4-8   Observed and predicted total phosphorus in the Patuxent
          River, 19 July 1978	    146
3.4-9   Seasonal trend of the N:P ratio in the Patuxent River  .  .    156
3.5-1   Predicted and observed CBOD  in Fox Mill Run	    163
3.5-2   Predicted and observed total nitrogen in Fox Mill Run  .  .    165
3.5-3   Suspended sediment distribution in the Hare River
          during high flow	    168
3.5-4   Total phosphorus distribution in the Ware River during
          high flow	    169
3.5-5   Observed and predicted total nitrogen and BODr in the
          Ware River estuary during high flow	    170
3.6-1   Thermal profile plots for Occoquan Reservoir 	    180
3.6-2   Plot of the Vollenweider relationship showing the position
          of Occoquan Reservoir using calculated total phosphorus
          loads (Source:  Zison e_t aj_., 1977)	    203
3.6-3   Maximal primary productivity as a function of phosphate
          concentration (Source:   Zison et_ a\_., 1977)  	    205
3.6-4   Dissolved oxygen depletion versus time in the
          Occoquan Reservoir	 . .  .  .    221

-------
                              TABLES


Number                                    .                     Page

2.1-1     Water Quality Simulation Results Summary for Rivers..    7

2.1-2     Water Quality Simulation Results Summary for
            Impoundments	    8

2.1-3     Water Quality Simulation Results Summary for
            Estuaries	    9
3.1-1      Major Water Quality Problem Segments  in  the
            Sandusky River	   26

3.1-2      Closed Shellfish Harvesting.Areas  in  Chester River...   30

3.1-3      Comparison of Sewage Treatment Plant  Discharges  in
            Bull Run  Sub-Basin:   1969-1977		    38

3.2-1      Hydraulic Data for Sandusky River  Gaging Stations
            (Low Flow Conditions)	    40

3.2-2      Hydraulic Data for Sandusky River  Gaging Stations
            (High Flow Conditions)	    44

3.2-3      Sandusky River Hydraulic  Data by Stream  Reach for a
            Portion of the System	    46

3.2-4      Deoxygenation Rate Constants for the  Sandusky River..   51

3.2-5      Expected BOD Values at 7Q,n Flow in the  Sandusky
            Ri ver	! T	    53

3.2-6      Expected BOD Values at 7Q10 Flow in Selected
            Sandusky River Tributaries	    54

3.2-7      Reaeration Rates Computed  by Two Methods for the
            Sandusky River	    55

3.2-8      Calculated Fecal Coliform  Concentrations for the
            Sandusky River System	    62
                                    •x .

-------
                         TABLES (Cont'd)
Number                                                             Page
3.2-9     Expected Percent .of Non-Urban Contribution to
            "Worst Case" Concentration of Suspended  Sediment
            at High Flows	   68
3.2-10    Ortho and Total Phosphorus  Relationships  in the
            Sandusky River Basin	   70
3.2-11    Expected Percent of Non-Urban Contribution to  "Worst
            Case" Concentration of Inorganic Nitrogen	   73
3.3-1     Low and High Scenario Flows for Chester River
            Tributaries	   78
3.3-2     Vertically Average Temperatures for the Chester River
            (°C)	   84
3.3-3  •   Vertically Averaged DO Concentrations  (mg  l~ )  for the
            Chester Ri ver	   85
3.3-4     Plant Nutrient and Chi orophy 1.1-a_ Levels in the Chester
            River	   87
3.3-5     Effluent Characteristics for Municipal  STPs and
            Industrial Discharges in  the Chester River Basin	   88
3.3-6     Low Flow Loads to  the Chester River from Municipal  and
            Industrial Point Sources  (Per Tidal  Cycle)	   90
3.3-7     Flushing Times for the Chester River by Three  Methods...   94
3.3-8     Flushing Times for the Chester River and Selected
            Tributaries	   96
3.3-9     Calculated Initial Concentrations in the Chester River
            for Two Water Quality Parameters	  10.1.
3.3-10    Chester River Conservative  Pollutant Distribution
            Coefficient Matrix (High  Flow)	  10-3

-------
                         TABLES (Cont'd)

Number                                                             Page

3.3-11    High Flow Pollutant Distributions  in the Chester
            River Estuary	  106

3.3-12    Correlations for Chlorophyl l-a_ on  Selected Water  Quality
            Parameters in the Chester River	  Ill

3.4-1      Active Dischargers  in the Patuxent River Basin	  115

3.4-2     Estuarine Cross Sections  in the Patuxent River	  116

3.4-3     Patuxent River Hydraulic  Data  for  Free  Flowing Waters
            (Low Flow)	  119

3.4-4     Deoxygenation and Reaeration Rates for  the Patuxent
            River Free Flowing Waters	  122

3.4-5     Major Sewage Treatment Plant Effluent Data in the
            Patuxent Ri ver System	  124

3.4-6     BOD Mass Balance for the  Free  Flowing Waters of the
            Patuxent Ri ver	  125

3.4-7     Dissolved Oxygen Profiles in 'the Patuxent River for  two
            Reaeration Rates	  129

3.4-8     Critical Travel Times,  Distances and Dissolved Oxygen
            Deficits for Some Patuxent STPs  at the 7Q10 Low Flow...  132

3.4-9     Calculation of Flushing Times  for  High  FLows Conditions
            in the Patuxent River Using  the  Modified Tidal  Prism
            Method	  137

3.4-10    Characteristic Data for the Patuxent River Estuary at
            High Flow	  148

3.4-11    Distribution Coefficient  Matrix for the Patuxent  River
            High Flow	  150

3.4-12    Total Nitrogen Calculation in  the  Patuxent River  Estuary
            for High Flow	  151

3.4-13    Upper and Lower Limit Total  Nitrogen and Total Phosphorus
            Concentrations in the Patuxent River  Due to Non-Urban
            NPS Loading	  153

-------
                          TABLES (Cont'd)

Number                                                          Page

3.4-14  Upper Limit Total Nitrogen and Phosphorus Concentra-
          tions in the Patuxent River Due to Urban and
          Non-Urban NPS Loads	    154

3.5-1   Ware River Estuarine Hydraulic Data	    159

3.5-2   Sewage Treatment Effluent and Natural  Water Quality
          in Fox Mill Run August 10-11, 1977	    160

3.5-3   Data For Estuarine Analysis of Fox Mill  Run by
          Modified Tidal Prism Method 	    162

3.6-1   Average Annual Frequency of Wind Speed in Percent .   .    175

3.6-2   Comparison of Geometry of Occoquan Reservoir to
          Parameter Values used to Generate Thermal Plots .   .    177

3.6-3   Mean Monthly Inf.lows to Occoquan Reservoir  .....  ...    178

3.6-4   Thermal Profile Data for Occoquan Reservoir 	    181

3.6-5   Comparison of Modeled Thermal Profiles to Observed
          Temperatures in Occoquan Reservoir  	    183

3.6-6   Annual Sediment and Pollutant Loads in Occoquan
          Watershed in Metric Tons Per Year	    185

3.6-7   Annual Urban Nonpoint Loads in Occoquan Watershed
          in Metric Tons Per Year	    186

3.6-8   Sewage Treatment Plant Pollutant Loads in Bull Run
          Sub-Basin in Metric Tons Per Year	    188

3.6-9   Particle Sizes in Penn Silt Loam	    190

3.6-10  Trap Efficiency Calculations for Lake Jackson  ....    193

3.6-11  Trap Efficiency Caluclations for Occoquan Reservoir  .    195

3.6-12  Calculated Annual Pollutant Loads to Occoquan
          Reservoir	    198

3.6-13  Observed Annual Pollutant Loads to Occoquan Reservoir    199

3.6-14  Calculated and Observed Mean Annual Pollutant
          Concentrations in Occoquan Reservoir  	    201

3.6-15  Nitrogen:Phosphorus Ratios in Occoquan Reservoir  .   .    202
                                 xm

-------
                           TABLES (Cont'd)

Number                                                          Page

3.6-16  High Flow Event Pollutant Loads in Occoquan
          Watershed from Non-Urban Nonpoint Sources 	    207

3.6-17  Stream Flows into Occoquan Reservoir During
          High Flow Events	    208

3.6-18  Trap Efficiency Calculations for Lake Jackson
          During High Flow Events	    210

3.6-19  Total Pollutant Loads to Occoquan Reservoir During
          High Flow Events	    212

3.6-20  Maximum Calculated Pollutant Levels in Occoquan
          Reservoir During High Flow Events (g m"^) 	    213

3.6-21  Hypolimnion Dissolved Oxygen in Occoquan Reservoir  .    222

A-l     Average Characteristics of Municipal Sewage 	    228

A-2     Municipal Wastewater Treatment—System Performance.  .    229
                                 xiv

-------
                              CHAPTER 1
                            INTRODUCTION
1.1  BACKGROUND

     In August, 1977 the U.S. Environmental  Protection Agency (EPA)
published a document entitled. "Water Quality Assessment—A Screening
Method for Nondesignated 208 Areas", EPA-600/9-77-023  (Zison  et aj_.,
1977).  This document is a compendium of techniques  designed  to  aid
in the assessment of water quality problems  in areas other than those
covered under Section 208 of the Federal  Water Pollution Control  Act
Amendments of 1972.  Designated 208 areas are generally characterized
by high concentrations of urban or industrial discharges while non-
designated 208 areas may encompass a wider spectrum  of-human  activities
and, hence, a larger set of water quality conditions.   These  include
agriculture and silviculture, as well as industrial  and municipal
activities.  As a result, methods to assess  water quality in  nondes-
ignated 208 areas must include not only the  capability to predict im-
pacts from point sources but also impacts from diffuse or nonpoint sources,

     In the above document, Tetra Tech, Inc. brought together a number
of methods designed to accommodate both urban and non-urban nonpoint
sources, as well as municipal and industrial point sources of pollutants.
In addition to the assessment of effluent water quality, the  methodology
provided for systematic routing of these pollutants  through rivers and
streams, impoundments, and estuary systems.   All  algorithms were de-
signed to be used as hand calculation tools.

     In 1976 Midwest Research Institute (MRI) developed a document en-
titled, "Loading Functions for Assessment of Water Pollution  from Non-
point Sources," for the U.S. Environmental Protection  Agency  (EPA-
600/2-76-151).  The loading functions described therein are used to

-------
 estimate the quantities of different diffuse loads that enter receiving
 water  bodies.  These methods do not route the pollutants through the
 receiving waters, however.

     Thus,  it appeared that the use of these tools in concert might pro-
vide an adequate  set of methods  for  screening nondesignated 208 areas by
 simple hand calculation procedures.  The methods developed by MRI for
 analysis of diffuse sources of water pollution and the parallel  method-
 ology  developed by Tetra  Tech had never been applied together in an
 actual  field situation.  This study represents an application of
 both methods under realistic situations for the purposes of demon-
 strating how the methodologies may be used for identification of water
 quality problem areas in nondesignated 208 areas.
 1.2   PURPOSE AND SCOPE

      The primary goal of this study is to demonstrate Midwest Research
 Institute's nonpoint calculator and Tetra Tech's nondesignated 208
 screening procedures under authentic field situations.  The demonstration
 is designed to subject the procedures to a wide range of data availability,
 water quality parameters, and hydrologic/hydraulic scenarios.  In
 addition to this primary goal, there are several subgoals.  They are:

      1.  Provide a report demonstrating the 208 screening
         methodology to be used as a guide by planners.
      2.  Show the degree of compatibility between the nonpoint
         loading analysis and the 208 screening methodology.
      3.  Develop firmer insight into the strengths and weak-
         nesses of the nonpoint loading methodology.
      4.  Evaluate the sensitivity of nonpoint load estimates
         to varying degrees of data availability.

-------
     5.  Determine how critical  or necessary the quality  and
         quantity of nonpoint source details are with  regard
         to reliably modeling in-stream processes as they are
         affected by nonpoint loading.
     6.  Demonstrate strengths and weaknesses of the 208
         screening methodology.
1.3  FORMAT AND ORGANIZATION

     The nondesignated 208 screening methods are an  extremely  versatile
set of procedures.  Because of their breadth of scope,  some of the
methods are not applicable in every planning situation.   Consequently,
some techniques are not covered in these demonstrations.   Conversely,
some techniques have been used which are not found in the original
screening procedures.  When this occurs these additional  techniques
have been fully explained.

     The applications pursued in this demonstration  are not exhaustive
of the ways that the methods can be used.   For instance,  the user may
not wish to predict water quality on the basis of a  specified  low flow
as has been done in this document.  Or, the user may choose to evaluate
planning alternatives which have not been investigated  in these demon-
strations.  The approaches presented here are reasonable  ones  for the
water quality constituents under investigation but the  utility of the
methods may be enhanced by innovation coupled with sound  judgment on the
part of the user.
     This document has not been written in the tutorial  style of the ori-
ginal non-designated 208 screening methodology document.   For brevity's
sake, not all calculations are shown.  For in-depth documentation of the
methods, the reader is referred to "Water Quality Assessment:  A Screen-
ing Methodology for Nondesignated 208 areas" (Zison, _et.  a]_., 1977).
Reference is frequently made to this document in these demonstrations
as simply the "screening manual."  To assist in obtaining values of rate

-------
constants, the user may find the following document useful:   "Rates,
Constants, and Kinetics Formulations in Surface Water Quality Model-
ing" (Zison e_t aJ_-> 1978).

     Almost without exception, numerical values appearing in this doc-
ument are given in metric units.  Exceptions are made when:
     •  the values are used in equations which were
        assigned different units in the original  screening
        manual,  or
     •  when non-metric units are used as indices for
        table look-ups in the original screening manual.
This report is divided into three chapters.  The first is the intro-
duction.  The second chapter lists the major results of the demonstrations
and the conclusions reached concerning both the methods themselves and
the systems to which they were applied.  The third chapter deals with
the demonstration of the methods in each of five watersheds.  It beains
with a short description of each system followed by discussions of
the methodology applications.  The most detailed demonstrations for
user orientation are the Sandusky River (for streams), Patuxent River    j
(for estuaries), and the Occoquan Reservoir (for impoundments).  The
remainder of the demonstrations emphasize comparison of predicted and
observed results.       .                             \

-------
                             CHAPTER 2

                      RESULTS AND CONCLUSIONS



     Elements of the non-designated 208 screening methodology were ap-

plied to each of five basins for water quality assessment.   The rivers

and streams methods (Chapter 4 in the screening manual) were applied to

the Sandusky and Patuxent river basins.  Impoundment methods (Chapter 5

in the screening manual) were applied to the Occoquan Reservoir.  Estu-

ary methods (Chapter 6 of the screening manual) were applied to the

Chester, Patuxent, and Ware rivers.  Certain conclusions can be drawn

concerning both the applicability of the methods and the screening

results for each individual basin.  First, however, some general con-

clusions concerning both the screening methodology and the nonpoint

source calculator are presented.  Specific conclusions of the demon-

stration are then listed by screening method (river, lake,  or estuary)

and by each basin studied.



2.1  GENERAL CONCLUSIONS
     0  The nonpoint source calculator (Midwest Research Institute)
        and the non-designated 208 screening methodology (Tetra
        Tech) are highly compatible.  Outputs from the nonpoint
        source (NPS) calculator are easily adapted and in some
        cases are used directly in the mass balance equations of
        the screening methods.  Event-based urban nonpoint loads
        are not readily predictable by the nonpoint calculator,
        but it is questionable if the non-designated 208 screening
        methods are applicable under these high flow - unsteady
        loading scenarios except to provide approximate upper
        and lower limits of instream pollutant levels.

     •  Loadings predicted by the nonpoint source calculator in
        conjunction with mass balance techniques employed by the
        non-designated 208 screening methods provided reasonably
        accurate predictions of instream, lake, and estuary water
        quality constituent concentrations.  No effects due to
        basin size or location were noted that detracted from
        either the applicability or accuracy of the methods.

-------
   Generally, loss of accuracy due to a loss in resolution
   was mitigated by the averaging effects intrinsic to
   larger systems.


   A qualitative assessment of.the rivers, estuaries and
   impoundments methods is shown in Tables 2.1-1 to 2.1-3.
   In general, the tables imply  that the river methods are
   the most accurate followed by estuaries and then impound-
   ments.  Within each method it should be mentioned that
   low flow - steady state conditions are more readily re-
   producible than high flow - unsteady loading situations.
   The impoundment methods probably require the least time
   and background skills to apply.  The riverine methods
   will usually require more time to apply than the estuary
   methods.  The results, however, should be easier to
   interpret for the uninitiated user than the results of
   the estuary methods.

•  Loadings predicted by the nonpoint source calculator in
   which all parameters are assumed to be correlated with
   sediment loss were more accurate for sediment and phos-
   phorus than fornitrogen and BODs.   This is an expected
   result.  In general, predicted nonpoint source nitrogen
	 and BODg loads were too low based on comparison of observed
   and predicted instream concentrations.

•  For conservative parameters, linear increases or de-
   creases in load estimates (either point or nonpoint)
   result in approximately linear changes in the concen-
   trations of those constituents in the water bodies.
   Therefore, an approximate error analysis can be
   performed directly using load estimates.  For non-
   conservative parameters, changes in stream,  lake,
   or estuary concentrations caused by increases or
   decreases in loadings can only be determined' by  routing
   the pollutants through the receiving water system.   An
   error analysis using loading changes and assuming the
   constituents behave conservatively will give an upper
   limit for the concentration changes likely to be en-
   countered.

•  While the methods appear to be a powerful  tool  for
   quickly identifying water quality problem areas, the
   use of the predictive techniques in conjunction with
   observed data further adds to their effectiveness.
   By doing this, the planner can identify specific
   problem areas in which quality cannot adequately be
   described by the simple techniques.  In most cases,
   the planner will  be able to recommend action, based

-------
TABLE 2.1-1.  WATER QUALITY SIMULATION RESULTS SUMMARY FOR RIVERS
                                          SYSTEM
                                SANDUSKY
                          PATUXEflT
      LOW  FLOW
           Temperature
           BOD
           Dissolved  Oxygen
           Coliforms
      HIGH  FLOW
          Sediment

          BOD

          Total  N

          Total  P
           €
           C
               Key:
                C
                *
             (blank)
Results good to excellent
Results fair to good
Simulation performed, no comparative
data available
No simulation performed
                               7  '

-------
     TABLE 2.1-2.   WATER QUALITY SIMULATION RESULTS
                SUMMARY FOR IMPOUNDMENTS
                                     OCCOQUAM
IMPOUNDMENTS
     Temperature                         9

     BOD                                 €)

     Dissolved  Oxygen

     Sediment                •

     Total  N

     Total  P
  Key:
        Results good to excellent
        Results fair to good

-------
TABLE 2.1-3.  WATER QUALITY SIMULATION RESULTS SUMMARY FOR ESTUARIES
SYSTEM
CHESTER PATUXENT
LOW FLOW
BOD . -O.. 	
Col iforms &
Total N .-9f 	
Total P " " Q
HIGH FLOW
Sediment &..
BOD _O
Total N <0 +-
Total P _^ _^._
Key:
0 Results good to excellent
^1 Results fair to good
O Results poor to fair
WARE

_£L

__*_..


JX
O
€
-€.




           .£    Simulation performed, no comparative
                 data availabl e

         (blank)  No simulation performed

-------
        on an understanding of the methods he has already
        applied, to investigate the problem area more
        closely.  These further investigations may include
        sampling programs or the use of a more sophisticated
        analytical tool.
2.2  RIVERS AND STREAMS
        Hydraulic characterization of rivers and streams is
        one of the most error-prone steps in the methods.
        A major reason for this is that flow is in many cases
        a function of subsurface phenomena which are not directly
        estimatable from the surface topography.  Unless the
        user has ground water measurements or detailed potentio-
        metric maps, these effects will not be properly
        characterized in the system description..

        Neither is  the user  given  any  direct guidance  in the
        screening  manuals  as  to appropriate  techniques  which
        can  be used to characterize  his system hydraulically.
        Often  only  limited geometric data  are available for
        a  given river system.   Measurements  of channel  geometry
        and  collection of  stage-discharge  data are  commonly
        done at bridges  or other easily accessible  locations.
        These  locations  are  often  not  representative of the
        conditions  in the  remainder  of the river.   Regardless
        of where these measurements  are made,  the  problem of
        estimating  flows,  geometries,  and  flow depths at
        locations  between  gaging stations  still  persists.
        It was  found in  these  demonstrations  that  simple
        area!  proportioning was  adequate for  interpolating
        and  extrapolating  streamflows.   Hydraulic  depths or
        radii  were  used  successfully in the  mass balance
        equations  for the  flow depth dependent terms.
        Hydraulic  depth  is the preferred characteristic
        depth  to use in  reaeration equations  since  it repre-
        sents  the  ratio  of oxygen  transfer capacity at  a cross
        section of  the river  (the  surface  width) to the oxygen
        storage capacity (the  cross  sectional  area).  The
        hydraulic  radius is  used in  the Manning  equation to
        estimate flow velocities.

        Dissolved oxygen prediction  is far more sensitive  to
        errors in estimating reaeration rates than in  estimating
        deoxygenation rates.   An inspection  of the range of these
       _rates  indicates why  this: is  true.  The methods  used to
        predict reaeration rates can yield answers which vary by '
        an order of magnitude.  Deoxygenation coefficients  are
        generally  predicted with greater accuracy.
                                10

-------
•  Predictive techniques for stream reaeratlon deserve
   further attention.   The three methods primarily used in
   these demonstrations are those of Tsivoglou-Wallace,
   O'Connor and Owens.   Of the three, the O'Connor and
   Owens formulations  are similar, each having the stream
   velocity in the numerator raised to a power and stream
   depth in the denominator raised to a power.  The
   Tsivoglou-Wallace approach takes into account the slope
   and travel time through the reach.  The remainder of the
   commonly used formulations for reaeration are of the
   O'Connor and Owens  type.

   Given that BOD loading rates and stream hydraulics are
   known with some accuracy in the demonstration systems,
   it appears that the O'Connor and Owens formulations give
   estimates which are too high and the Tsivoglou-Wallace
   predicts reaeration rates which are too low.

   The use of the formulations of O'Connor and Owens
   usually kept dissolved oxygen profiles at saturation,
   so it is difficult to determine how much they over-
   estimated reaeration rates.  Use of Tsivoglou*-Wallace
   rates rarely allowed the prediction of anoxic con-
   ditions, however.  A value in between the Tsivoglou-
   Wallace and Owens or O'Connor predictions is more
   likely to be appropriate.

t  Comparison of predicted instream fecal or total
   col 1 form concentrations with observed data is
   impractical.Although the techniques presented in
   the screening manual are adequate, the use of average
   loading data for coliforms does not readily reproduce
   individual instream measurements.  Loadings of fecal or
   total coliforms from sewage treatment facilities are
   extremely unsteady and are subject to very large
   perturbations.  When disinfection equipment is mal-
   functioning, col iform concentrations in the effluent
   may reach 10°/100 ml; when disinfection is successful,
   the concentrations  are negligible.  The methods in the
   manual can be best used for worst case analyses to give
   upper limit concentrations.  Attempts to identify
   bacterial problems  resulting from municipal sewage
   treatment plant and septic tank failures, or combined
   sewer overflows, should probably proceed with a time
   to failure or probability approach for the systems
   involved.
                            11

-------
2.3  IMPOUNDMENTS

2.3.1  Stratification
     •  Thermal  plots from the impoundment thermal  model  accurately
        describe water temperature, thermal  gradients, and time of
        the onset of stratification.  Epilimnion depths are modeled
        less reliably.  Although conditions  for the Occoquan do not
        exactly match the parameter values in any one set of thermal
        plots, the excellent results in the  demonstration show the
        utility of the information that can  be derived from these
        plots.

     0  The greatest difficulty in using the thermal  profiles lies
        in the selection of the correct plot to apply in a border-
        line case.  When several impoundment paramters have to be
        bounded, a large number of plots may have to  be considered.
        Some cases may be eliminated if they predict  a physically
        unreasonable result (e.g., a thermocline depth is
        predicted which is greater than the  mean depth of the
        impoundment).  If the occurrence of  stratification is
        uncertain, the user may want to proceed as  if the impound-
        ment does stratify.  Climatic variation will  almost
        certainly cause stratification to occur in  some years in
        these borderline cases.

     •  In borderline cases, selection of the maximum depth
        parameter may be aided by also considering  mean impound-
        ment depth.  The mean depth represents the  ratio of the
        volume of the impoundment to its surface area.  Because
        the volume and surface area are proportional  to the
        thermal  capacity and heat transfer rate respectively,
        the mean depth should be useful in characterizing the
        thermal  response of the impoundment.

     •  The hydraulic residence time strongly affects the thermal
        profile of an impoundment.  In the demonstration, two- to
        threefold changes in the magnitude of the thermal gradient
        were observed when the residence times varied by 25 percent.
        It was shown that interpolation between thermal plots suc-
        cessfully predicted the effects of variation  in hydraulic
        residence times on the thermal gradient.
2.3.2  Sedimentation
     •  Accuracy of sedimentation calculations depends primarily on
        accurate load estimates.   Predictions based on the Universal
                                 12

-------
        Soil  Loss Equation (USLE)  can vary greatly due to the un-
        certainty in the sediment delivery ratio.   If possible,
        on-site data should be utilized.   However, the good agree-
        ment of the predicted and measured loads in the demonstration
        watershed indicates that the USLE may be used with some con-
        fidence.

        Trapping  efficiencies are most sensitive to particle size.
        This  arises from the quadratic relationship of particle
        diameter  to its settling velocity.  Consequently, accurate
        knowledge of sediment diameters is required.   Often a very
        large portion of the range of particle sizes  is completely
        trapped,  so errors tend to be relatively small.
2.3.3  Eutrophication
        The ability of the methods to quantitatively predict parameter
        values associated with eutrophication is limited.If nutrient
        concentrations are estimated from average annual  loads,  plant
        growth can only be approximated.   Seasonal  effects cannot be
        represented adequately.   Prediction of algal growth may  also
        be confounded by factors such as  toxicants  in the  water.   The
        relationship used in the impoundment methodology  to calculate
        water column total phosphorus levels requires site-specific
        knowledge of rate constants which may only  be determined
        through measurement.
2.3.4  Dissolved Oxygen

     •  The hypolimnion dissolved oxygen calculations are very sen-
        sitive to the BOD loading rate (k^) and decay rates.  The
        dissolved oxygen level has an exponential  dependence on the
        first-order BOD decay rate constants for the water column
        U}) and in the benthic layer (k4)..  The decrease in dissolved
        oxygen at any time is directly proportional to ka<  As a con-
        sequence, any uncertainty in the value of these constants
        will greatly broaden the range of predicted oxygen depletion
        rates.  Because the reported values of the constant vary
        widely or are few in number, on-site measurements of these
        constants are required in order to make quantitative projec-
        tions of dissolved oxygen levels.

     •  Qualitatively useful results are predicted bv the simplified
        hypolimnion dissolved oxygen model even when BOD decay rate
        constants are not accurately known.  By using estimated upper
                                 13

-------
        and lower bounds for these constants, dissolved oxygen
        versus time curves can be obtained.  These curves indi-
        cate the likelihood of experiencing low dissolved
        oxygen levels in the hyplimnion.   When applied to the
        demonstration impoundment, the method predicted a range
        of oxygen depletion rates that was shown to bracket
        the actual  behavior of the impoundment.  This agree-
        ment demonstrates the qualitative value of the model.
2.4  ESTUARIES

2.4.1  Classification
     •  The use of the flow ratio method underestimates the degree
        of vertical  stratification.   According to most sources,
        the Chesapeake Bay and its tributaries are partially mixed.
        The flow ratio method predicts well  mixed conditions for
        both low and high flows in both the  Patuxent and Chester
        Rivers.

     •  The Stratification-Circulation method is preferred for
        estuarine classification, but the required data may not
        be available.   Surface velocity data were available for
        only one estuary (the Chester River).  These data were
        taken from a special  study.   Similar data may not be
        routinely available for other estuaries.  Salinity and
        net fresh water flow rates are usually obtainable or can
        be estimated.   To obtain a complete  picture of the hydro-
        dynamic variation that the estuary might undergo, the
        surface velocity, net fresh  water velocity, and surface
        and bottom salinity should be available for high and low
        fresh water inflows both at the mouth and head of the
        estuary.
2.4.2   Flushing Calculations
        The tidal  prism and modified tidal  prism flushing times
        are related, and their ratio seems to be dependent upon
        the estuary volume"!  The flushing times at the 7Q1Q (seven-
        day low flow that occurs once in 10 years) for several
        estuaries  tributary to and including the Chester River
        were evaluated by both methods.   A log-linear regression
        of the ratio of modified tidal prism method to tidal
        prism method versus the mean low tide volume of the estuary
                                 14

-------
        was performed.   The regression predicted the same ratio
        for the Patuxent and Ware rivers with accuracy.

     •  The fraction of fresh water method is fairly insensitive
        to the number of segments used to estimate flushing times.
        Flushing times  in the Chester River were calculated by
        the fraction of fresh water method using a 1 ppt and
        2 ppt segmentation scheme.   The results were essentially
        equal.

     •  For flushing times derived  by the modified tidal prism
        method that are similar at  highandlow flows, mechanisms
        other than advective flow are more important in  flushing
        the estuary.  This will generally be the case for large
        estuaries with  small drainage basins.  One might also
        infer from this that the water quality in such a system
        is dominated by the quality of the replacement waters
        during tidal exchange rather than the surface runoff
        waters.

     •  The fraction of fresh water and modified tidal prism
        methods predict more similar flushing times for smaller
     •   estuaries.  The'lo'nger the  residence time in the
        estuary, the more likely it is that antecedent runoff
        conditions will affect salinity profiles.  The fraction
        of fresh water  method inherently accounts for ante-
        cedent flow conditions whereas the modified tidal prism
        method does not.  The methods will compare more  con-
        sistently if salinity data  are taken during a period
        of steady inflow.  The salinity profile should be
        averaged over the sampling  period.  The length of the
        period of sampling should be determined by the resi-
        dence time in the estuary.   The problems of estimating
        accurate flushing times in  large estuaries with  short
        term flow or salinity data  should be obvious.  For these
        estuaries a flushing time should be computed from an
        expected quarterly, semi-annual or annual flow rate.
        Flushing times  calculated for long residence estuaries
        using very low  or very high short term flows or
        salinities should be used only for comparisons of
        the relative flushing characteristics.
2.4.3  Pollutant Distribution
        Low flow predictions of pollutant distributions in
        estuaries are good for conservative constituents.As
        long as the steady state assumptions for flow and  loadings
        are met, the fraction of fresh water method is adequate
        to predict distributions.
                                  15

-------
        The modified tidal  prism method must be used for non-
        conservative constituents.  The calculation of decay for
        non-conservative constituents is based on a develop-
        ment using the modified tidal prism method (Officer,
        1976).   If a segmentation other than that determined
        by the modified tidal  prism is used, the results will
        be erroneous.

        Pollutant distributions predicted for unsteady flow or
        unsteady loading represent upper and lower limit concen-
        trations.  Such cases  include prediction of concentrations
        due to storm water carrying nonpoint source contaminants .
        into the estuary or other impulse type discharges (chem-
        ical spills, etc.).  For storm events, the user can assume
        that the full storm load enters the estuary during each
        tidal cycle, providing upper limit concentrations for
        the pollutant.  This is reasonable since the duration
        of most events is less than the approximate 12-hour tidal
        cycle in duration.  Alternatively, the user can assume
        that the storm load is equally distributed over each tidal
        cycle occurring during the time base of the runoff inflow
        hydrograph.  This alternative will give lower limit con-
        centrations for pollutants.  More exact predictions re-
        quire the use of advection-dispersion equations.


        Estuarine contamination from 'tidal exchange with
        polluted background waters can be ascertained.  The
        modified tidal prism and fraction-of fresh  water methods
        assume that the replacement waters on each tidal exchange
        have no residual contamination.  Since this is usually not
        the case, especially for estuaries which may be tributary
        to other estuaries, the degree of contamination in the
        estuary due to replacement waters can be estimated by com-
        paring observed profiles to those predicted by the estuarine
        methods.
2.4.4  Eutrophication
        The two parameter light extinction model  which regresses
        Secchi depth on contaminant concentration data can be
        used to investigate light limitation of algal  growth in
        estuaries.   Values of the background extinction coeffi-
        cient calculated from raw data are of the same magnitude
        as those measured by others (3.4 to 3.6).
                                 16

-------
2.5  THE SANDUSKY RIVER
     •  No critical temperature problems exist in the basin except
        in Spring Run where an industrial effluent enters at a
        high temperature.  In general, temperature is not a
        problem in other subbasins in the system.

     •  High BOD and low dissolved oxygen occur concurrently in
        this system and are bracketed if not accurately predicted by
        reasonable choices of BOD decay rates and reaeration co-
        efficients.  The exception is below Fremont where dissolved
        oxygen levels were not adequately predicted.  Here, the
        methods have indicated that a more detailed analysis
        should be conducted which should include measurement of
        benthic oxygen demand, photosynthesis and respiration
        rates.

     0  Dissolved oxygen profiles support the conclusions drawn
        from field data by the Ohio EPA (1978) with respect to
        river segments in which dissolved oxygen sags normally occur.

     't  Data indicate that suspended sediment is not a problem
        at low flows.

     0  When suspended sediment and total phosphorus are treated
        as conservative constituents, the predicted concentrations
        at high flows are accurate.  Using urban and non-urban
        loads, it appears that urban controls are more appropri-
        ate in the upper basin for reduction of instream
        concentrations while agricultural controls would be more
        appropriate in the lower basin.

     •  Prediction of available nitrogen was poor, but the same
        recommendations as for sediment and phosphorus are
        indicated.
2.6  THE CHESTER RIVER
        Even with the "worst-case" assumption of no enroute decay
        for non-conservative constituents, initial  pollutant con-
        centrations in the Chester due to point sources were small
        Initial concentrations of total nitrogen and total  phos-
        phorus due to sewage treatment plant (STPs) and industrial
        effluents were also negligible.
                                  17

-------
•  Degradations may well occur in small estuaries tributary
   to the Chester into which effluents from STPs or small
   commercial seafood operations flow, but a lack of salinity
   and hydraulic data prevented an in-depth analysis of these
   estuaries.

•  Modified tidal prism flushing times under both high and
   low flows were very nearly equal, indicating that in
   the Chester River tidal action, as opposed to advective
   flow, is the dominant flushing mechanism.

•  Phosphorus concentrations that result from high flow
   events are of sufficient magnitude to cause algal prob-
   lems.  These are more likely to occur in the upper
   estuary where concentrations are higher.  Algal growth
   at the estuary head may be residence time limited due to
   the good flushing characteristics there.

•  Nitrogen also appears to be plentiful enough after high
   flow events to support a large algal crop.

•  The fact that high coliform counts have been observed in
   the estuary should lead naturally to an investigation of
   the-effectiveness of disinfection at the sewage treatment
   plants and the effects of combined sewer overflows from
   municipalities.  No estimates were made of the frequency
   of combined sewer overflows, septic tank, or sewage treat-
   ment plant failures in the basin.  It is felt that due to
   the limited urban or suburban development in the basin,
   other factors must be contributing to the high bacterial
   counts.  Chief among these other factors are probably un-
   treated loadings from boat latrines and waterfowl.

•  Although the methods contain no technique for assessing
   dissolved oxygen levels in the estuary, observed data
   indicate nearly anoxic conditions on occasion near the
   bottom.  Investigations of this problem should include
   estimation of benthic oxygen demand and solids loadings
   from boating and waterfowl which may settle.

0  There appears to be a seasonal shift in the N:P nutrient
   ratios in the estuary based on a limited number of obser-
   vations.  Higher N:P ratios tend to occur in the spring
   with lower N:P ratios occurring in the summer.  The high
   It:P ratio in the spring indicates that spring phosphorus
   control from nonpoint sources may be appropriate for
   managing the size of the algal crop.

•  If it is assumed that the Chester River is a well mixed
   estuary, application of the two parameter light model indi-
   cates light limitation of algal growth.
                                  18

-------
2.7  THE PATUXENT RIVER

2.7.1  Riverine Portion

     •  No temperature problems due to heated effluents are ob-
        served or predicted.

     9  Deoxygenation coefficients calculated by the Bosko equa-
        tion (see Zison, et aj_., 1978, p. 180) were in the range
        of those calculated by graphical analysis of field data.

     t  NBOD estimates based on the type of treatment that the
        sewage facilities employed generally overestimated NBOD
        loadings as compared to estimates based on measured total
        Kjeldahl nitrogen at each plant.

     •  BOD in the Little Patuxent River is higher during low
        flows than in the Patuxent above their confluence.

     •  BOD and dissolved oxygen predictions are not extremely
        sensitive to background values used in the mass balance
        equations.  Usually, loadings from sewage treatment
        facilities are of.great enough magnitude that, the back-
        ground BOD becomes negligible after the first point
        source enters the river.  Similarly reaeration rates
        are sufficiently large so that the initial value
        chosen for the oxygen deficit in the most upstream
        reach is not critical.

     0  As in the Sandusky River, the use of the Tsivoglou-
        Wallace and O'Connor formulations for predicting
        reaeration rates results in a bracketing of observed
        dissolved oxygen profiles.
2.7.2   Estuarine Portion
     •  Non-conservative constituents can be effectively dealt
        with by using simple mass balance with decay in the tidal
        fresh water portion of the Patuxent River.  The tidally
        influenced fresh waters should be segmented as in the
        riverine portion and waste concentrations estimated by
        flow weighting.  These concentrations should then be
        decayed instream by first order kinetics.  Excursion
        times for each segment can be estimated by dividing the
        length of the section by the net velocity in that section.

     •  A problem area was identified in the tidally influenced
        fresh water portion of the Patuxent.  At flows near the
        7Qio f°r tne system, total nitrogen and total  phosphorus
        act non-cpnseryatvvely_in thi_sjDortion.  Just before the
                                 19

-------
        riverine flow meets the density-gradient flow total
        nitrogen and total  phosphorus both drastically decrease.
        This may be explained by macrophyte uptake or algal
        growth with concurrent predation by zooplankton.   At
        slightly higher flow rates earlier in the summer these
        losses of total nitrogen and phosphorus were not observed.

     •  The seasonal trend in N:P ratios is very pronounced in
        the Patuxent estuary.  N:P ratios estimated by the
        screening method adequately predict the seasonal  trend.
        This strong seasonal trend also implies that nonpoint
        controls for phosphorus in the spring and control  of
        nitrogen from point sources in the fall may be used for
        managing the size of the algal crop.   Given the range
        of the N:P ratio observed in the estuary it may be
        concluded that nitrogen control is the more important
        of the two.

     •  Predicted N:P ratios are lower than those calculated from
        observed data in the Patuxent River estuary during high
        flows.  This appears to be due to low predicted total
        nitrogen values.  Even so, the observed N:P ratios and
        those predicted'would lead to the conclusion of'nitrogen
        limitation in the estuary.
2.8  THE WARE RIVER
     0  Estuarine methods were applied to a very small  tidal
        creek, Fox Mill  Run, and yielded reasonable predictions
        for conservative and non-conservative parameters.

     •  Nonpoint source  loads dominate the quality of the Ware
        River.  This is  logical since there is essentially no
        urbanization in  the basin.

     •  Similar values of flushing times for high and low flow
        periods indicate that the Ware River is probably dominated
        by tidal exchange as opposed to advective flow, although
        not to the same  extent as the Chester.

     •  Average nitrogen and phosphorus concentration  predictions
        are good for this estuary at high flows.  Predicted sediment
        concentrations are high while BOD5 concentrations are low.

     •  N:P ratios computed for the estuary were much  lower than
        the observed ratios.   In  this case, these ratios were so
        different that an improper conclusion would have been drawn
        concerning nutrient limitation had data not been available.
                                20

-------
        Observed ratios indicate a trend toward phosphorus limita-
        tion while predicted ratios indicate definite nitrogen
        limitation.

     •  Almost no seasonal  trend is observed in the N:P  ratios for
        this system.   This  is as expected since no major urban areas
        exist in the  basin  and nonpoint sources dominate the  nutri-
        ent input to  the river.

     •  Background extinction coefficients  determined by fitting
        a two parameter light model to Seechi  Disc data  are con-
        sistent with  those  calculated for the  Chester River and
        those observed by other researchers for a  turbid coastal
        inlet.

     •  The  two  parameter  light model predicts that the Ware River
        is light limited for algal  growth,  assuming a fully mixed
        estuary.
2.9  THE OCCOQUAN RESERVOIR
     •  Thermal  profiles predict that stratification in Occoquan
        Reservoir may'occur.  The uncertainty is a result""
        of the mean hydraulic residence time falling between
        10 and 30 days.  During years with low rainfall, stratifi-
        cation should occur.  During years with a higher than
        average rainfall, stratification will be weak or nonexistent.

     •  The Occoquan Reservoir traps approximately 90% of the sedi-
        ment entering the impoundment from the Bull Run sub-basin
        and 80% of the sediment entering from Lake Jackson.  Any
        future land use that would significantly increase sedimenta-
        tion should be carefully examined.

     •  Based on predictions of both the Vollenweider plot and
        the Chiaudani curve, the Occoquan Reservoir is eutrophic.
        This prediction is confirmed by field data.

     •  Water quality should not change significantly during high
        flow events.  The MRI loading functions predict pollutant
        loads will be significantly higher during a high flow
        event than during a seven-day period under average flow
        conditions.  The high loads are offset by a threefold in-
        crease in the volumetric flow rate over that of an average
        seven-day period.

     •  Nutrient loads predicted using the MRI nonpoint calculator
        were quite accurate.  The total nitrogen and phosphorus loads
                                  21

-------
   reported by one source were within twelve percent of those
   calculated using a delivery ratio  of 0.2.  Another source
   reported loads 2.3 and 1.9 times  larger than  the calculated
   loads.

•  Predicted mean nutrient concentrations  were comparable to
   observed summer values.  The ratios of  observed  to pre-
   dicted concentrations of total  nitrogen and phosphorus
   were 1.2 and 0.9, respectively,  for a delivery factor of
   0.1, and 0.7 and 0.5, respectively, for a delivery factor
   of 0.2.  The calculated ratio of  nitrogen to  phosphorus
   concentrations was 27 percent lower than observed in the
   Occoquan Reservoir.

•  Anoxic conditions are likely to occur during  the period of
   stratification.  The dissolved oxygen calculations predict
   that dissolved oxygen will be absent near the bottom for
   20 to 95 days during the summer.

•  The primary cause of oxygen depletion is algal growth,
   the algal BOD contribution being  approximately five times
   greater than the BOD load from the tributaries.
                            22

-------
                              CHAPTER  3
                       DEMONSTRATION OF METHODS
3.1  STUDY AREA DESCRIPTION

     In order to demonstrate the screening methods  under a wide range
of watershed conditions,  types  of discharges,  and date availability,
five watershed systems were  chosen for  analysis.  They were:

     •  Sandusky River, Ohio
     •  Chester River, Maryland and Delaware
     •  Patuxent River, Maryland
     ••  Ware River, Virginia
     t  Occoquan River, Virginia

3.1.1  The Sandusky River

     The Sandusky River rises near Crestline,  Ohio, where  it first
flows west, and then north.   It empties into Muddy  Creek Bay, Sandusky
Bay, and finally into Lake Erie (Figure 3.1-1).  The drainage area  is
       2
4404 km , making it the largest basin under consideration  for this
demonstration.  The river is approximately 209 km long and drops an
average of .74 m/km over that length.   Over most of its length the
river is lined by a vegetative corridor which  provides a buffer for
runoff from agricultural  areas  adjacent to it. About 88% of the land
use in the drainage basin is agricultural.  Crops grown are mostly
corn, soybeans, wheat, oats, hay, orchards and some specialty crops.
There are four municipalities in the study area with populations over
SjOOO.  These towns are Bucyrus, Tiffin,  Upper Sandusky, and Fremont.
                                   23

-------
                               j  Muddy Creek Bay
.  Figure  3.1-1..  The Sandusky River 'basin.
                       24

-------
     The basin lies within the Till  Plains  and  Lake  Plains adjacent
to Lake Erie.  Within the Till  Plains,  covering approximately  the
lower two-thirds of the basin,  end morraines  left  by retreating  ice
control  the surface drainage.   These morraines  are roughly parallel
to .the Lake Erie shoreline.   The northern  third of the Lake  Plains
was formed when the area was inundated  by  an  ancient lake.   Their
topography is flat to gently rolling.

     The climate is humid continental with  an average annual temper-
ature of about 10.6 °C.  The basin receives about  86 cm of precipi-
tation annually.  The months of February and  October typically receive
the lowest amounts of precipitation, while April through July  are the
months with the highest amounts.  Further  climatological and other
descriptive data will be presented in the  demonstration sections as
needed.                  ....                           .....

3.1.1.1  Water Quality

     Table 3.1-1, taken from the Sandusky  River Basin portion  of the
Ohio State Water Quality Management Plan,  shows the  major water  quality
problem areas in the basin (Ohio EPA, 1978).

     The two most seriously affected segments are  immediately  below
the Bucyrus and Fremont Waste Water Treatment Plants. Bucyrus has
secondary treatment facilities, but the fact  that  the plant  discharges
roughly half the streamflow below the effluent  outfall creates problems.
Bucyrus also has combined sewer overflow problems.  It has been  esti-
mated that a 20-minute rainfall of greater than 0.13 cm would  cause
overflows to occur.  The probability of occurrence of this event is
one in every five days (Ohio EPA, 1978).

     The river below Fremont sometimes  shows  anaerobic conditions during
summer low flows.  This apparently is the  result of  overloading  treat-
ment facilities by food processing plants.
                                   25

-------
            TABLE  3.1-1.  MAJOR WATER QUALITY PROBLEM SEGMENTS IN THE SANDUSKY RIVER
 Water Quality
Problem Segment
RKMI* of Segment
 Sub-Basin Name
   Problematic
   Parameters
Source of Problem/RKMI*
Paramour Creek
(211.4 - 208.3)
Upper Sandusky River
Dissolved Oxygen
  '  Ammonia
Crestline STP/211.4
Sandusky River
(177.9 - 170.2)
Sandusky River
(222.5 - 124.5)
Upper Sandusky River
Upper Sandusky River
Dissolved Oxygen
    Ammonia
Dissolved Oxygen
    Ammonia
Bucyrus STP/177.9
Upper Sandusky STP/127.6
Honey Creek
(46.0 - 30.6)
Middle Sandusky River
Dissolved Oxygen
Attica STP/45.4
Spring Run
(9.6 - 0.0)
Tymochtee Creek
Dissolved Oxygen
Carey STP/6.6
Budd CO./9.5
*RKMI - River Kilometer Index.   The distance along the stream/river to
        the confluence with the next  larger water body.

-------
     Otherwise, water quality degradation  as  a  result of  inadequate
waste treatment is not considered  a  problem.  The  river is affected
by secondary pollution problems such as  sediment,  turbidity,  high coli-
form bacteria counts, and high nutrient  levels  resulting  in occasional
algal blooms.  The high coliform counts  are attributable  to animal
wastes, combined sewer overflows,  and poor maintenance of rural septic
tanks.
3.1.2  The Chester River
     The Chester River empties into  the  Chesapeake Bay  between Eastern
Neck and the Northern end of Kent Island (Figure  3.1-2).   Its head-
waters are in Delaware and it meanders  toward  the Chesapeake Bay
                                               1                   2
through Maryland's Eastern Shore.  It drains approximately 1,140  km  .
                                               i
The river is 81 km in length and is  tidally influenced  for 64 km  up-
stream of the mouth.   Agricultural  uses  of land are  predominant in the
basin.  Corn, wheat,  grains, soybeans,  hay, vegetables, and potatoes
are the major crops.   The largest towns  in the study area  are Chester-
town and Centreville.  These municipalities had populations of 3,500
and 1,850, respectively,  as of 1970.  Wetlands are present in the basin,
                      2
comprising about 34 km .   Forestry and  fishery operations  are also
found in the basin, with  the processing  of oysters,  soft shell clams,
blue crabs, and finfish being prevalent.       ',
     The topography is flat to gently rolling.   The uplands  in  the
Wicomico Plain have elevations of 27 to  30 meters  and  the  lower  basin,
which lies in the Talbot Plain, has elevations  from sea  level to  18
meters.
                                               j
     The basin has a humid, temperate continental  climate  with  mild
winters due to climatic moderation by the Chesapeake Bay and  the
Atlantic Ocean.  Summers are warm and humid.  Temperatures average
26°C in July and temperatures of 0°C or  lower occur an average  of 73
                                  27

-------
       x	'  CHESTERTOWN
      yflAOCLIFFECREE
Figure  3.1-2.   The Chester River basin.
                       28

-------
days per year.   Precipitation averages  107  cm  per  year  and  is  dis-
tributed evenly throughout the year.  Snowfall  averages about  54 cm
per year.
3.1.2.1  Water Quality

     Three water quality factors are of primary concern  in  the  Chester
River Basin.  These are:

     •  An unusually high mortality of oysters and other benthic
        organisms,
                                               i
     •  Significant increases in nutrient concentration
        over the last decade,                  !
     0  Bacterial  levels which exceed the State of
        Maryland's shellfish harvesting standards.

     In general, water quality in the Chester is good  with  the  excep-
tion of coliform levels.  High benthic. organism mortality is  suspected
as being caused by a toxicant in the basin.   Efforts by  the Maryland
Department of Natural  Resources are currently underway to identify  this
agent.                                         j
     Nutrient concentrations have been increasing since 1965.   In  1974
phosphorus and nitrogen concentrations were of such magnitude  to support
a large algal bloom in the upper river from Chestertown to  Crumpton
(State of  Maryland,  1975).
     Closures of shellfish waters due to high indicator organism counts
are extensive in the basin.  Table 3.1-2 shows the extent of  these
closures and the suspected  reasons for them.
3.1.3  The Patuxent River
     The Patuxent River is located on Maryland's  western  shore  and
                             2
drains approximately 2,537 km .   It originates in the hilly  Piedmont
                                  29

-------
                             TABLE 3.1-2



       CLOSED SHELLFISH HARVESTING  AREAS IN  CHESTER  RIVER
Area
Chester River
Reed and Grove Creeks
Corsica River
Gray's Inn Creek
Langford Creek
St. Michael 's Harbor
Kent Island Narrows
Queenstown Creek
Cox Creek
Rock Hall Harbor
Spencer and Little Neck
Creeks
Oak Creek
Leeds Creek
County
Kent, Queen Anne's
Queen Anne's
Queen Anne's
Kent
Kent
Talbot
Queen Anne's
Queen Anne's
Queen Anne's
Kent •
Talbot
Talbot
Talbot
Acreage
4,642
469
571
837
531
61
665
316
142
1,591
61
173
387
Conditions for Closure
Sewage treatment plant,
storm water runoff
Storm water runoff, septic
tanks overflowing
Treatment plant
Overflowing septic systems
Agriculture runoff, over-
flowing septic systems
Buffer zone - St. Michael's
Treatment 'Plant
Waste from seafood process-
ing plants
Buffer zone - treatment
plant
Storm water runoff
Storm water runoff, sewage
violations
Storm water runoff, sewage
violations
Storm water runoff, sewage
violations
Storm water runoff, sewage
violations
After U.S. Army Corps of Engineers, Baltimore District, 1977.
                                  30

-------
Plateau between Washington,  D.C.,  and  Baltimore  and  flows  southeasterly,
entering Chesapeake Bay at Solomon's  Island.   The  river  itself can be
divided into four distinct regions as  indicated  on the map in Figure
3.1-3.   These regions are:
        Free flowing waters which  extend  from  the  head-
        waters to approximately Hardesty,  Maryland,
        Tidal  fresh waters which extend from Hardesty  to
        about where Hall  Creek enters  the Patuxent River,
        Estuarine waters  from Hall  Creek  to just below
        Sheridan Point, and
        Embayment waters  which comprise the remainder  of
        the river downstream.
The tidally influenced portion of the river  ends  about  89  km from  the
mouth.

     As of 1977, over 50 percent of tire basin  was forested, with 35  per-
cent being cultivated for agriculture and  the  remainder in  urban or  sub-
urban developments.   Major communities in  the  area include  Laurel, Bowie
and Savage (U.S. Army Corps of Engineers,  1977).

     Two major tributaries contribute flow to  the Patuxent.  The largest
is the Little Patuxent which joins the main  stem  at Bowie.  The other  is
Western Branch which flows into the Patuxent just above Jug Bay.   Two
reservoirs are located on the Patuxent above Laurel.  These are the
Triadelphia and T. Howard Duckett (Rocky Gorge) Reservoirs.  Their com-
bined storage capacity is 13.4 billion gallons.   They are  used primarily
to supply water to the Washington Suburban Sanitary Commission and
Montgomery and Prince George Counties.

     The Patuxent River basin, like the Chester,  has a  humid, temperate
continental climate with warm summers and  mild winters.  The mean  annual
                                 31

-------
                                                      CHESAPEAKE BAY
SCALE IN MiOMf IFH
   SSI::?:*:* PATUXENT
   SSS*** BIVEH BASIN
         M6TROAR6A
          Figure 3.1-3.   The  Patuxent River  basin.
                                    32

-------
precipitation at Washington,  D.C.,  is  103.6 cm,  which is  distributed
fairly evenly by season.   Normal  temperatures range from  2.7°C  in
January to 25.7°C in July with the  annual  average being 13.9°C.

3.1.3.1  Water Quality

     Early water quality studies  in the Patuxent River showed severe
dissolved oxygen violations downstream of  Laurel  during periods  of low
flow and high temperature.  Recently,  due  to sewage treatment plant
modifications these problems  have been mitigated somewhat.   Current
dissolved oxygen problems are attributed primarily to nitrogenous  BOD
demand.  However, total  nitrogen  and phosphorus  levels have continued
to increase.  There are eleven publicly owned sewage treatment  plants
of concern in the basin.

     Industrial  dischargers in the  basin are in  general small and  do
not significantly affect water quality (Pheiffer e_t aj_.,  1976).  The
largest industrial  discharger is  the Potomac Electric Power Company
Chalk Point Electric Plant with a discharge of approximately 720 mgd
of cooling water to the Patuxent  Estuary.   According to a Corps  of
Engineers report (U.S. Army Corps of Engineers,  1977), tidal  portions
of the estuary do not meet Class  II temperature  standards.   However,
studies have not indicated problems attributable to this  heated  ef-
fluent!

     Shellfish waters, as of 1977,  were closed from river km 37.8  to
river km 64.4 because fecal coliform standards were exceeded in  most
portions of the middle section of the main stem  and in the tidal reaches
of the river.  Bacterial  contamination has been  primarily attributed  to
nonpoint agricultural and urban runoff.

     Sedimentation has been and continues  to be  a problem in the basin,
particularly with regard to navigation.
                                      33

-------
3.1.4  The Ware River
     The Ware River is a small  tidal  river  located  in  southern  Virginia
off Mobjack Bay adjacent to Chesapeake Bay.   It  lies between  the  York
                                                             2
and the Rappahannock Rivers on  the western  shore.   The 138  km  area  is
drained principally by two streams,  Beaver  Dam Swamp and  Fox  Mill  Run as
shown in Figure 3.1-4.  These streams are meandering sloughs  with ill-
                                                           2
defined channels.   Fox Mill Run has  a drainage area of 34 km  above  the
outfall  of the Gloucester Sewage Treatment  Plant, the  only  municipal
discharger in the basin.  The land consists  primarily  of  forests  and
swamps with limited agricultural  development.

3.1.4.1  Water Quality
     Water quality problems frequently associated with-the Ware River
are low dissolved oxygen concentrations (<5.0  mg a),  low  pH values
(<6.5), and high fecal  coliform densities  (>allowable log mean of 200/
100 ml MPN).   As of February 1, 1976,  the  Ware River  waters were still
open to shellfishing (U.S.  Army Corps  of Engineers, 1977).

3.1.5  The Occoquan River

     The Occoquan River is  a tributary to  the  Potomac River located  in
Northern Virginia.  The watershed lies entirely in  three  counties: •
Fauquier, Prince William, and Loudoun,.  The map of  Figure 3.1-5 shows
                 2
that the 1,480 km  watershed is drained by three major streams which
form the Occoquan River.  These are Bull Run,  Broad Run,  and Cedar
Run.  The Occoquan River itself is dammed  just below  Hooes Run to form
the only major water supply reservoir  on the east coast downstream
from an urbanized area.

     There are three cities of consequence in  or on the periphery of
the watershed; namely,  Fairfax, Manassas,  and  Warrenton,  Virginia.
                                  34

-------
CO
en .
                                     KILOMETERS
                                         Figure  3.1-4.  The  Ware River  basin.

-------
Dulles Airport is also located on the boundary of the basin.   Aside
from these centers, the basin is relatively undeveloped,  consisting
mainly of agricultural land and forests.   The climate is  similar to
that of the Patuxent basin.

3.1.5.1  Water Quality

     By far the biggest problem area in the watershed is  in the Bull
Run sub-basin.  As of 1977 there were eleven major sewage treatment
plants located there.  Plans were made to  replace these facilities
with an 84 million dollar regional  tertiary treatment plant.   Table
3.1-3 shows an annual history of the combined discharges  of these
plants.

     Apparently, tertiary treatment of municipal  waste has not allevi-
ated the problem of high nutrient levels  and occasional algal  blooms
which have occurred primarily in the Bull  Run arm of the  Occoquan
Reservoir since the late 1960's.  Grizzard e^t aj_. (1977)  have pointed
out that urban nonpoint loads may be the  primary  source of these
nutrients and indicate that higher unit area loads originate from
urban than from agricultural nonpoint sources.

     Nutrient levels in the Occoquan Reservoir as a whole have also
been high enough to create eutrophic conditions.   Total nitrogen con-
centrations averaged 0.9 gm   during the  months of April  through
October between 1973 and 1977.  The mean  total phosphorus concentra-
tion for the same period was 0.08 gm" .   The mean summer  (April  through
October) chlorophyll a_ concentration was  held to  21 mg m~  by the ad-
dition of copper sulfate during the period of measurement (1975-1977)
(Northern Virginia Planning District Commission,  March 1979).

     As might be expected from its trophic status, the Occoquan
Reservoir has oxygen depletion problems as well.    Hypolimnion oxygen
                                  36

-------
concentrations in general begin to decrease with the onset of
stratification in late April and are usually depleted by the end of
May or June.  Oxygen replenishment may not begin until  the end of
September in some cases, leaving the hypolimnion without oxygen for
possibly three to four months.


3.2  DEMONSTRATION EXAMPLE:  THE SANDUSKY RIVER

     The analysis of the Sandusky River basin involved demonstrating
the procedures contained in the Rivers and Streams section of the
screening methodology manual.  Killdeer Reservoir, an impoundment
in the basin for which water quality data were available, is a pumped-
storage reservoir.  As such, most of the methods presented in the
Impoundments chapter of the screening manual are not applicable and'
no analysis was performed."

     The analysis of water quality in .the river and some of its
tributaries was performed for both a high flow and a low flow scenario.
Under high flow conditions, nitrogen, phosphorus, and sediment concen-
trations resulting from both urban and non-urban nonpoint loadings were
analyzed.  Under the low flow scenario, temperature, dissolved oxygen,
BOD, and fecal coliforms were analyzed.  When possible, predicted instream
levels.of these constituents were compared to historial observations made
under similar hydrologic conditions.
 3.2.1 .  Data  Collection

     As a  first  step  in data collection, 7*2-minute topographic maps
 were obtained for the basin.  The 7%-minute maps were necessary to
 obtain  the more  detailed slope and stream mileage information for
 hydraulic  computations.  Larger scale maps were extremely helpful in
 getting a  good perspective on the general basin features,
                                   37

-------
      TABLE 3.1-3.  COMPARISON OF SEWAGE TREATMENT PLANT
        DISCHARGES IN BULL RUN SUB-BASIN:  1969 - 1977
Year
1969
19731
19742
19743
197 54
19753
19764
19763
19774
19773
Jan. -Nov.
Flow
mgd
2.43
5.54
6.09 • -
5.62
6.64
6.69
6.26
6.2
5.1
6.5
2
Jan. -Mar.
BOD5
1 b/day

669
459
260
349 '
315
356
320
315
402
3Jul
Total
Nitrogen
1 b/day
483

560
599
750
608
691
632
575
710
y-Dec.
Total
Phosphorus
1 b/day
250
217
163
75
89
77
85
56
50
72
4
Jan. -June
Source:  Northern Virginia District Planning Commission, 1979.
                               38

-------
     The U.S. Geological Survey (USGS) provided daily flow information
for each of the gaging stations in the basin.  The period of record
varied from over 50 years at some gages to only two at sortie others.
Cross-section profiles and stage-discharge rating curves were also
obtained.

      The USGS provided temperature, pH, conductance, and dissolved
oxygen data for those same stations.  Additional water quality data for
total phosphorus and orthophosphorus, nitrite, nitrate, ammonia, organic
and total Kjeldahl nitrogen, and suspended solids were obtained from the
U.S. Army Corps of Engineers (1978).  Data on the effluent characteristics
of industrial and municipal dischargers were obtained from an Ohio EPA
preliminary report on the Sandusky River basin (Ohio EPA, 1978).


3.2.2  Data Reduction and Supplementation

     Frequency analysis was performed 'on the low flow data to determine
the 7Q^0.  This is the low flow of seven-day duration that occurs once
every ten years (see Haan, 1977, for a comprehensive example).  The
7Qin low flow magnitudes were determined for all gaging stations in the
basin.  These values are shown in Table 3.2-1.

     Once the-7Q,Q low flows were determined, the flow depths were
found from state-discharge curves.  At Fremont, the depth at a flow of
      3   -1
2.35 m sec   is about 0.33 meters.

     The hydraulic radius or hydraulic depth can be used as the char-
acteristic depth of the stream in the mass balance equations for water
quality constituents.  For wide, shallow streams, the hydraulic depth
and hydraulic radius are roughly equivalent.  By using a wetted perimeter
measurement from the cross-sectional profile at the Fremont gage
(Figure 3.2-1) the hydraulic radius was determined for the channel at
that point.  For the Sandusky, use of the hydraulic depth (defined as
cross-sectional area divided by flowing stream width) in lieu of the

                                39

-------
            TABLE  3.2-1.   HYDRAULIC  DATA  FOR  SANDUSKY  RIVER
                GAGING  STATIONS  (LOW  FLOW  CONDITIONS).
     Name
7Q1QFlow

(m3 sec"1)
                                  Character-
                                    istic
                                    Depth
Area of
 Cross
Section

 (m2)
Character-
  istic
 Velocity

 (m sec  )
Bucyrus                0,23

Crawford               0.14
(Tymochtee Creek)

Upper Sandusky         0.33

Mexico                -1.61

Fremont                2.35
                 0.12

                 0.09


                 0.14

                 0.22

                 0.21
  1.17

  1.58
   0.20

   0.09
  1.61        0.20

  8.29        0.19

  4.92        0.48
                                   40

-------
0.0
           15.2       30.4      45.6       60.8      76.0       91.2      106.4      121.6     136.8      152.0
                                                WIDTH
                                                  (m)
       Figure  3.2-1.  Cross-sectional  stream profile  of the Sandusky River near  Fremont.

-------
hydraulic radius (cross-sectional area divided by wetted perimeter)
resulted in a 7 to 10 percent difference in the stream characteristic
depth.  The values shown in Table 3.2-1 are hydraulic radii.

     From the continuity equation (Q = AV), the velocities in these
sections were calculated and are also shown in Table 3.2-1.   This was
done by calculating the cross-sectional area in the channel  correspond-
ing to the 7CLg flow.  Dividing the cross-sectional area into the flow
gives the velocity.

     A word of caution is in order here.  Many times the gaging stations
are installed in sections of the stream which are nontypical  of the rest
of the stream.  For instance, they may be installed in straight, un-
cluttered, narrow sections  for ease of measurement, whereas  most of the
stream may be winding, debris-laden, and wide.  Therefore, the velocities
may be higher and the stream deeper in these sections.  If other infor-
mation is available, these  depths and velocities should be adjusted
accordingly.  If not, there is generally no recourse but to  use the
available information as "representative" of the system.

     For high flow periods  a similar procedure was followed  to char-
acterize the channel system hydraulically.   Instead of choosing the
annual high flow event, however, selected high flow events were singled
out during the period of record.  The reason for this is that generally
the annual high flow seven-day event takes  place in conjunction with
snowmeTt  conditions  in this area.  Under such circumstances it is dif-
ficult to estimate loads of nonpoint source pollutants since the waste
load methods use the USLE (Universal Soil Loss Equation).  The equation
is designed to predict soil loss due to erosive rainfall-runoff conditions.
(The reader is informed that the Screening  Manual does provide a modifica-
tion of the USLE for snowmelt events.)  The "R" (rainfall-runoff erosivity)
factor in the USLE is estimated for each rainfall event and  averaged over
all events, snowmelt events being excluded.  Only those high flow events
occurring from April to September inclusive were considered.  Furthermore,
                                    42

-------
it is desirable that the rainfall events that produced the chosen high
flow periods have a good areal coverage over the watershed.  Because of the
size of the Sandusky basin, the latter requirement was sometimes difficult
to meet.

     Keeping the above constraints in mind, 12 seven-day high flow
periods were chosen.  The loadings represent an average over those
events.  The flow data which represent the average seven-day high flow
over all the events were supplemented as for the low flow periods to
yield the hydraulic information given in Table 3.2-2.
3.2.3  River Segmentation
3.2.3.1  Low Flow

     The Sandusky River system was initially broken down into a total
of 76 reaches.  The Sandusky River proper was made up of 21 reaches
with an average length of 6.3 miles per reach.  The divisions were
made based on either the introduction of significant tributary flow
or the introduction of flow from a point source of contaminants.

     Reach segmentation is a partially subjective procedure.   A point
source may contribute insignificantly to the concentration of a con-
taminant once it is mixed with the flow of the river.  In such cases,
it can be ignored.  It may be that the point source contributes a
certain contaminant at significant concentrations but does not con-
tribute a load of other contaminants under consideration.  Therefore,
the reach segmentation scheme may vary for certain water quality
constituents.

     In whatever manner the river is divided into reaches, slope and
length are necessary to describe each reach.  The user may inter-
polate slopes between the known points (stream gages) in the watershed
or measure them from topographic maps.  Again, slopes at the gaging
stations may be nonrepresentative of the greater part of the river.

                                  43

-------
        TABLE  3.2-2.   HYDRAULIC DATA FOR SANDUSKY  RIVER GAGING
                   STATIONS (HIGH FLOW CONDITIONS)
7-Day High Flow
Name
Bucyrus
Crawford
(Tymochtee Creek)
Upper Sandusky
Mexico
Fremont
(m3 sec"1)
24.0
10.4
74.4
192.3
336.0
Character-
istic
Depth
(m)
1.68
0.79a;
1.40
0.31
1.74
Area of Character-
Cross istic
Section Velocity
2 1
(m ) ••• (m sec )
27.0 0.89
30. S3^ 0.34'3''
69.2 1.07
172.8 1.13
163.5 2.04
a)Estimated because of poor cross section profile
  information at high flows.
                                     44

-------
     The reaches were numbered and the pertinent hydraulic information
was organized into a matrix.  A portion of this matrix is shown in
Table 3.2-3.  The information in this matrix with the exception of
reach length and possibly slope, must be interpolated in some manner
from the information at the stream gages.
3.2.3.2  High Flow

     Because a different set of water quality parameters was under
consideration during high flow events, the watershed was segmented
differently.  Specifically this segmentation scheme was dictated by
the availability and resolution of land use data necessary for the
estimation of nonpoint source loadings.  Because the methodology
treats nitrogen, phosphorus and suspended solids conservatively, much
of the information necessary for mass balance of nonconservative con-
stituents at low flows is not necessary.  For conservative constituent
routing, only flow and stream lengths are needed.  These data are
available from the information matrix orovided in Table 3.2-3.
3.2.4  Temperature Profiles

     Since the rate constants necessary for routing nonconservative
constituents are temperature dependent, a stream temperature profile
was first developed for the entire system.

     The first step in using the methods for stream temperature is the
calculation of the equilibrium temperature.  (See Section 4.4.4 of the
screening manual.)  The month of October is the month when the annual
seven-day low flow event usually occurs.  From the Climatic Atlas of
the United States (U.S. Department of Commerce, 1974) the following
mean monthly climatic data for this month in the Sandusky area were
obtained:
                                   45

-------
                                 TABLE 3.2-3.   SANDUSKY  RIVER HYDRAULIC DATA  BY STREAM REACH
                                                    FOR A PORTION OF THE SYSTEM
cr>
Identifying
Reach Characteristic
Number At Upstream End
1
2
3
4
5
6
7
8
Muskel lunge Creek enters
Sandusky River
Fremont WWTP
Indian Creek enters
Sandusky River
Wolf Creek enters
Sandusky River -
Tiffin WWTP
Willow Creek enters
Sandusky River
Point Source
Honey Creek enters
Hydraulic
Radius (in)
Low High
0.24
0.21
0.20
0.21
0.21
0.21
0.21
0.22
2.13
1.83
1.87
2.40
2.41
2.53
2.53
3.14
Cross
Section (m ) .
Low High '
5.57
4.92
5.20
6.50
6.50
6.78
6.78
8.27
167.2
163.5
164.3
167.9
167.9
168.7
168.7
172.8 :
Flow
(m3 sec"1)
Low High
2.61
2.38
2.35
2.2
2.01
1.95
1.93
1.61
378.5
336.0
323.1
268.5
268.5
255.6
255.6
192.4
Slope
(m km"1)
0.47
1.04
1.91
0.85
1.56
1.56
0.15
0.15
Velocity
(m sec )
Low High
0.46
0.49
0.49
0.30
0.30
0.28
0.28
0.20
2.26
2.04
2.04
1.62
1.62
1.52
1.52
1.10
Length
(km)
15.1
6.0
11.1
3.2
25.7
1.6
1.3
4.8
                     Sandusky River

-------
     Mean daily incoming shortwave radiation =
          280 langleys = 11,147, BTU nf2 day"1 = 1036 BTU ft"2 day"1
     dean cloud cover = .55
     Mean monthly air temperature = 12.8°C (dry bulb) = 55.0°F
     Mean daily wind speed = 12.9 km hr"  = 8.0 miles hr
     Mean daily relative humidity = 71%
     The first four items can be used directly in the equilibrium
temperature calculations.  Mean shortwave radiation has to be converted
                       -2                                      -2
from langleys to BTU ft  .   The conversion factor is 3.7 BTU ft   per
langley.  The equilibrium temperature calculated for. the Sandusky basin
using these inputs is 16.2°C.
     Temperature profiles for the Sandusky River were computed by
beginning at the headwaters of the Sandusky and computing temperature
successively for each downstream reach.  (See Section 4.4.5 of the
screening manual.)  To start the calculations, an upstream temperature
was estimated for the first reach.  During low flow conditions a good
estimate of this temperature was that of the ground water, approximately
10°C.  At the junction of each reach, a resultant temperature was cal-
culated due to the addition of wastewater or tributary inflow.  This
initial (resultant) temperature was used in the heat balance to compute
a temperature at the beginning of the next reach.  The procedure was
repeated until the profiles for the entire river and all the major
tributaries were computed.

     Figures 3.2-2 and 3.2-3 show temperature profiles computed for the
Sandusky River proper and a tributary, Spring Run.  Included in the
Sandusky plot are historical mean stream temperatures ± one standard
deviation for gaged locations along the river.  The computed temperature
is always within the one standard deviation envelope of the historical
observations.  Spring Run had no historical temperature data with which
                                    47

-------
o
IT
UJ
Q_
   17.0
   15.6
    4.2
   12.8
   11.4
       MOUTH
                     I       I
                                          1        T
                 FREMONT
                 WWTP
                                    TIFFIN
                                    WWTP
                                                                                         BUCYRUS
                                                                                         WWTP
                                                                  UPPER SANDUSKY
                                                                  WWTP
                                                                                                   EQUILIBRIUM
                                                                                                  TEMPERATURE.
                                     I
   10.0
                                        OBSERVED LOW FLOW
                                        TEMPERATURE (7-DAY MEAN
                                        ± ONE STANDARD DEVIATION)

                                           I        I       I
             16      32      48      64      80      96      112     128     144

                                                 DISTANCE FROM MOUTH
                                                         (Km)
                                                                                       176     192      208     224
            Figure 3.2-2.   Observed and  predicted  temperatures in  the Sandusky River (low  flow).

-------



o
o
111
CC
Z3
QC
111
O.
H


o/.o
35.0
32.2.
29.4
26.7
23.9
21.1
18.3
15.6
12.8>
10.0
-
-
-
-
y
<
-
-
-
. 	 , 	 -v
-
-
-

• "
;
J
/
' • I
                        8.0

               DISTANCE FROM MOUTH
                       (km)
16.1
Figure 3.2-3.  Calculated low flow temperature
               profile for Spring Run.
                         49

-------
to compare the predicted values but is included because of the presence
of an extremely high temperature effluent from an industrial source.  At
low flows, this effluent comes to equilibrium very rapidly and has no
effect on the stream temperature after approximately three kilometers.
3.2.5  Estimation of BOD Decay Coefficients and Reaeration Coefficients

     Since measured values of NBOD or CBOD (nitrogenous or carbonaceous
biochemical oxygen demand) were not available, the decay coefficients
for these two parameters could not be estimated directly.  Additionally,
no evidence was available to warrant making a distinction between the
rates of decay of CBOD and NBOD.   Therefore the decay of BOD was governed
by a single decay constant.  Using the methods of Hydroscience and Bosko
described in the screening- manual, a range of deoxygen-ation coefficients
was established for each reach.  Since the range of the deoxygenation
rates from the Bosko equation was  not large and the Hydroscience method
predicted values close to the mean of the values from the Bosko equation,
this mean value was used to predict BOD concentrations in the Sandusky
system.  The temperature corrected deoxygenation coefficients which were
used in the routing equations are  shown in Table 3.2-4 for reaches on
the Sandusky River proper.


3.2.6  BOD Mass Balance
     Biochemical  oxygen demand was determined for all  reaches in the
    2m
loads.
system for the 70,0 flow.   Ultimate NBOD plus CBOD values were used as
     At low flows, the primary sources of BOD are sewage treatment
plants discharging treated effluent directly into the river and to a
lesser degree factories and food processing plants.   Most of the in-
dustrial water users around Fremont discharge into the municipal
sewage treatment facility.

                                  50

-------
     TABLE 3.2-4.  DEOXYGENATION RATE CONSTANTS FOR THE SANDUSKY RIVER
                                                       Deoxygenation
                                                                    -1.
  Reach(es)              Location                   Coefficient (day" )
19, 20           Confluence of Paramour Creek      .36, 2.38
                 and Allen Run to Bucyrus WWTP
18               Bucyrus WWTP to confluence of     .47
                 Broken Sword Creek and
                 Sandusky River
15, 16, 17       Confluence of Sandusky River      .46, .37, .37
                 w/Broken Sword Creek to
                 Upper Sandusky WWTP
14               Upper Sandusky WWTP to            .36
                 confluence of Sandusky River
                 w/Tymochtee Creek
12, 13           Confluence of Sandusky River      .28, .28
                 w/Tymochtee Creek to
                 confluence w/Honey Creek
8, 9, 10, 11     Confluence of Sandusky River      .45, .30, .29, .28
                 w/Honey Creek to Tiffin WWTP
3, 4, 5, 6, 7    Tiffin WWTP to Fremont WWTP       .51, .68, .48, .41, .47


1, 2~	  - Fremont-WW-TP-to Muddy— Creek-&a-y-—-.27, .45		
                                    51

-------
     Tables 3.2-5 and 3.2-6 show the computed ranges for BOD in
reaches of the Sandusky and its tributaries on which there are im-
portant point sources of BOD.  The range of values represents the
variability from the upstream to downstream end of the reach or ag-
gregation of reaches.  By far the most degraded portion of the river
considering this parameter is the segment below the Bucyrus Waste Water
Treatment Plant followed by the segment downstream from the Upper Sandusky
Water Treatment Plant.  High values of BOD also occur in the headwater
areas of the Sandusky and its tributaries >/here small sewage treatment
plants are located in reaches with little instream diluting flow during
low flow periods.

3.2.7  Dissolved Oxygen Profiles

     Output from BOD routing and temperature routing were used to
compute dissolved oxygen- profiles for the Sandusky River system.   Re-
aeration rate coefficients were computed by two methods, that of Owens
and that of Tsivoglou and Wallace (1972).  Since the results of these cal-
tions often showed more than an order of magnitude difference in the rate
coefficients, both sets of coefficients were used to determine dissolved
oxygen profiles.  Table 3.2-7 shows the reaeration coefficients computed
by each method for reaches of the Sandusky River proper.  These coeffi-
cients are not corrected for temperature as they appear in the table.

     Generally, reaeration rates computed by the Owens equations are
higher in the upstream portions of the river because of the shallow
depths there.  The depth term.appears in the denominator of that equation,
which drives up the calculated reaeration rate.  Towards the mouth of the
river depths become greater and the rates drop.  Conversely, for the
Tsivoglou-Wallace equation, reaeration rates are higher downstream
generally than they are upstream.  In this expression, stream velocity
and slope are multiplied together.  Although the slope remains relatively
constant throughout the stream, velocity increases towards the mouth, causing
predicted reaeration rates to increase.  Other reaeration rate formulations

                                     52

-------
  TABLE 3.2-5.  EXPECTED BOD VALUES AT 7Q1Q FLOW IN THE SANDUSKY RIVER
                                                          Ultimate
                                                      BOD Concentration
  Reach(es)              Location                      Range (mg £~')


19, 20            Confluence of Paramour Creek           12.7 - 3.4
                  and Allen Run to Bucyrus WWTP
18                Bucyrus WWTP to confluence             43.8 - 20.1
                  w/Brokea.Sword Creek
15, 16, 17     .   Confluence of Sandusky                 14.3 -  5.6
                  w/Broken Sword Creek to
                  Upper Sandusky WWTP
14        -        Upper Sandusky WWTP to                 30.4  -  9.1
                  confluence w/Tymochtee Creek
12, 13            Confluence of Sandusky River            7.50 -  6.1
                  w/Tymochtee Creek to confluence
                  w/Honey Creek
8, 9, 10, 11       Confluence of Sandusky River            5.5  -  4.7
                  w/Honey Creek to Tiffin WWTP
3, 4, 5, 6, 7     Tiffin WWTP to Fremont WWTP              7.4  -  3.3


1, 2              Fremont WWTP to Muddy Creek  Bay          5.8  -  4.9
                                   53

-------
    TABLE 3.2-6.   EXPECTED BOD VALUES AT 7Q1n FLOW IN SELECTED
                      SANDUSKY RIVER TRIBUTARIES
                                                     BOD Concentration
Reach(es)                Location                     Range (mg JT1)


Spring Run-Tymochtee Creek

  54              Carey WWTP to confluence of           10.5-6.7
                  Spring Run w/Tymochtee Creek
  53              Confluence of Spring Run w/            4.1 -  2.8
                  Tymochtee Creek to confluence
                  of Tymochtee Creek w/Sandusky
                  Ri ver
Honey Creek

  26, 27          Attica WWTP to Bloomville WWTP         9.3 -  2.2
  25              Bloomville WWTP to confluence          5.3 -  1.6
                  w/Sandusky River
                                   54

-------
     TABLE 3.2-7.  REAERATION RATES COMPUTED BY TWO METHODS
                      FOR THE SANDUSKY RIVER
 Reach(es)                   Reaeration Rates  (day"  )

                         Owens                       Tsivoglou-Wallace

19, 20               97,  >100                      1.28, 1.92


18                   97                            3.16


15, 16, 17           60,  67, 93                    3.53, 1.61, 1.45


14                   66                            1.5


12, 13               20,  25                        .40, .44


8, 9,  10,  11          28,  28, 25, 20                6.71, .65, .61,  .45


3, 4,  5,  6,  7         45,  45, 42, 30, 30            7.63, 14.01,  5.61,
                                                  3.94, 7.22


1,2                 35,  40                        3.31, 4.64
                                  55

-------
are included in the screening manual (Zison, ejb ^1_., 1977) and in Zison,
e_t aj_., 1978, but were not considered applicable here.

     Figure 3.2-4 shows calculated dissolved oxygen profiles for the
Sandusky River.  The upper profile represents dissolved oxygen com-
puted with the Owens reaeration rates.  The lower shows the dissolved
oxygen profile computed with the substantially lower reaeration rates
calculated by the Tsivoglou-Wallace equation.  Information on photo-
synthetic oxygen production rates and respiration and benthic demand
were not available for any portions of the river.  For this reason, these
factors were not included in the analysis.  These parameters can be
estimated if desired.  For guidance the reader is referred to Zison, ejt
aj_-, 1978.

     An inspection of Figure 3.2-4 shows that the use -of- the Owens
reaeration rates maintains the dissolved oxygen profile for the system
almost always at the saturation dissolved oxygen value.  This value
decreases slightly from upstream to downstream due to temperature in-
creases but is generally on the order of 10.4 to 10.6 mg £~ .  The
only segment in which the dissolved oxygen drops significantly is down-
stream from the Bucyrus Waste Water Treatment Plant (stream-km 177) where
it is 9.0 mg X," .

     Using the Tsivoglou-Wallace reaeration rates, the dissolved oxygen
values remain between 6 and 9 mg SL~  for most of the river.  At about
km 59 the predicted dissolved oxygen begins to climb and almost reaches
the level attained using the higher reaeration rates.  This is due pri-
marily to the fact that the Tsivoglou-Wallace reaeration rates approach
values similar to the Owens rates in these reaches.

     Once errors are introduced into the oxygen balance calculations, they
remain although their influence generally decreases as computations for
additional reaches are done.  These errors may be related, for instance,
to the hydraulic description of the system.  It is also important to
                                   56

-------
                              TT
                              I
                                        I
                                              T Standard Deviation Envelope
                                              I For October Observations

                                              1 Standard Deviation Envelope
                                              | For Low Flow Observations
16       32       48       64       80       96      112      128      144

                                   DISTANCE FROM MOUTH (Km)
160
        176
192
        208
                               224
  Figure 3.2-4.   Computed  vs. historical  dissolved  oxygen  for  the Sandusky  River.

-------
note that in the dissolved oxygen calculation, errors accrued in the
computation of temperature and BOD levels are present.  Therefore,
results of dissolved oxygen calculations are more subject to error
because the calculations are based on information that is estimated
with some degree of uncertainty.  Still, Figure 3.2-4 shows that the
methods used bracket the observed water quality data (with the exception
of the Fremont station at km 19).

     The historical observations merit some discussion here.  There is
some question when making comparisons between predicted and historical
observations regarding which historical data to use.  Historical dis-
solved oxygen data exist for some of the low flow periods that were used
to determine 7Q,Q flows, flow depths, velocities, and deoxygenation and
reaeration rates.  However, temperature predictions were made using
meteorological data excl-us-ively for the month of October.  This temper-
ature information is used in the correction of the rate coefficients
and also has a direct influence on the dissolved oxygen saturation con-
centration.  Therefore, historical mean dissolved oxygen values and
their respective standard deviation envelopes are plotted in Figure 3.2-4
for the observed annual low flow periods (irrespective of when they oc-
curred) and the October low flow data (whether or not they went into the
70-Q computation).  The plotted standard deviation envelopes show that
for the Upper Sandusky gage (stream km 125) these two means are almost
equal with little difference in the dispersion of the observed data.
At the Mexico gage (stream km 77) the observed means are still similar
although the annual low flow dissolved oxygen values are far more dis-
persed than the October low flow observations.  At the Fremont gage
(stream km 20) the mean of the October observations is lower while its
standard deviation is larger.

     While the predicted results at the Upper Sandusky and Mexico gages
are encouraging,the results at the Fremont gage suggest that the compu-
tational approach may be in error.  One possibility for error is that
parameters external to the method used may have a large influence on
                                  58

-------
dissolved oxygen levels in the Tiffin to Fremont segments.  For instance,
there may be a substantial benthic demand which is unaccounted for in
the equation used.  It is also possible that errors in hydraulic infor-
mation have been compounded through repeated use in the estimation of
temperature, BOD levels and rate coefficients.  Additionally, it is
possible that information is missing; i.e., a substantial source of BOD
may have been omitted from the list of point sources, or, perhaps, a
source thought to discharge via the sewage treatment plant discharges
directly into the river.  In instances where the methods are applied
locally by personnel familiar with or having direct access to the study
area, this latter possibility would be substantially reduced.

     The fact that these methods do not reproduce historical observations
in these reaches, however, does not vitiate them.  On the contrary, the
methods have served one .of. the purposes for which they .were designed;
that is, to point out areas requiring special attention and further,
more detailed investigation.

     Figure 3.2-5 shows dissolved oxygen profiles developed for Honey
Creek using Tsivoglou-Wallace and Owens reaeration rates along with its
temperature profile.  The plot shows that neither temperature nor dis-
solved oxygen effects are likely to be of concern at low flows on this
stream with its two small wastewater treatment plants.  It was assumed in
this calculation that the stream temperature above the Attica Waste Water
Treatment Plant was at the temperature of the groundwater and that the
dissolved oxygen was at saturation.  Unfortunately, no historical dis-
solved oxygen data were available for comparison on this stream.
3.2.8  Fecal Coliform Mass Balance

     Total coliform and instream fecal  coliform data were unavailable
for the Sandusky River.  However, effluent data measurements for sewage
treatment plants often included fecal  coliforms.  Since this parameter
was being investigated under low flow conditions sewage treatment plants
                                  59

-------
8
                     O)
                     E
Ul



X
O

Q
LLJ
>
_l
O
                     o
 12



 11



 10

   I

  g



  8



  7






12.8



12.2
                     01
                     DC  11.7

                     I
                     P:  11.1
                        10.6


                        10
                           0
                                           OWENS
                                   TSIVOGLOU -WALLACE
                         16
                                                              BLOOMVILLEWWTP

                                                                        EILBRRJM  TEMPERATURE
                                                             ATTICAWWTP


                                                               I	
                                        32

                               DISTANCE FROM MOUTH
                                       (km)
48
64
                           Figure 3.2-5.   Dissolved oxygen and  temperature profiles for  Honey Creek.

-------
were considered to.be the only source.  In cases where fecal coliforms
were not measured estimates were made based on data from other treatment
plants with similar flow and treatment type characteristics.

     Table 3.2-8 gives the results of the fecal coliform mass balance
for the Sandusky River system.  Computations  show that in many areas
the fecal coliform count due to sewage treatment facilities are near
the drinking water standard (1.0/100 ml) and well below the surface
water standard of 2000/100 ml.  Of course, the estimates of instream
fecal coliform concentrations do not include  the fecal coliform loading
dur to waterfowl and other wildlife or from agricultural operations.
McElroy, ejt ^1_., (1976) show that the background total coliform concen-
tration in this area is approximately 2000/100 ml.  The U.S. Environmental
Protection Agency reported that the fecal to  total coliform ratio for the
Ohio River was in the range of 0.2 to 12 percent (U.S. EPA, 1973), sug-
gesting that background fecal coliform concentrations might be on the
order of 4 to 240/100 ml.

     The survival of fecal coliforms passing  through a sewage treatment
plant can be highly variable.  Hhen chlorination processes are function-
ing properly virtually all coliform bacteria may be eliminated.  However,
 under situations in  which such  a  process fails,  concentrations  on the
 order of 10/100 ml  may be found  in  the  effluent.   Effluent character-
 istics such  as those provided  by  the plant may be quite  nonrepresenta-
 tive of the  actual  stream loadings  over  a  given  7-day period.   Fecal
 coliforms are often  of greater  concern  during high  flow  events  v/hen heavy
 rains may cause  combined  sewers or  animal  feeding operations  to discharge
 raw sewage into  the  river system.
 3.2.9  Sediment Mass Balance

      Previous sections in this example have dealt exclusively with
 parameters which are of concern primarily at low flow conditions.   In
                                   61

-------
     TABLE 3.2-8.  CALCULATED FECAL COLIFORM CONCENTRATIONS
                 FOR THE SANDUSKY RIVER SYSTEM
   Reach(es)
       Location
Fecal CoHform
Concentration
 (MPN/100 mi)
       Sandusky River
19, 20



15, 16, 17, 18



12, 13, 14


8, .9, 10, 11


3, 4, 5, 6, 7

1, 2
Confluence of Paramour
Creek and Allen Run to
Bucyrus WWTP

Bucyrus WWTP to Upper
Sandusky WWTP


Upper Sandusky WWTP to
Confluence w/Honey Creek

Confluence of Sandusky and
Honey Creek to Tiffin WWTP

Tiffin WWTP to Fremont WWTP

Fremont WWTP to Muddy Creek
Bay
   8.6 - 4.6
   3.6 - 0.2
   6.3


   0.4


   2.5

   1.8
0.2


0.03


1.5

1.3
       Honey Creek

26, 27               Attica WWTP to Bloomville
25
WWTP

Bloomville WWTP to Confluence
w/Sandusky River
54-63
53
       Tymochtee Creek - Spring
           Run
Carey WWTP to Confluence of
Spring Run w/Tymochtee Creek

Confluence of Tymochtee Creek
and Spring Run to Confluence
of Tymochtee Creek and Sandusky
River
   2.6 - 1.9


   2.7 - 1.6
   4.1 - 3.8


   2.3 - 0.9
                             :  62

-------
the Sandusky River, sediment, phosphorus and nitrogen were investigated
only at high flows.  Sediment, nitrogen and phosphorus loadings were
provided by the Midwest Research Institute's nonpoint calculator.

     In the Sandusky River system velocities are of such low magnitude
during low flow periods that sediment concentrations are a relatively
unimportant aspect of water quality.  Figure 3.2-6 shows a representative
sediment rating curve for the month of October, 1976, at Fremont.   The
flows shown in the figure are reasonably close to the 7Q,Q low flow esti-
mated for the system.  Most stations in the system for this month had
suspended sediment concentrations under 30 mg £~  with the exception of
the Upper Sandusky gage whose concentrations were consistently in the
20 to 130 mg £-1 range.
     A sediment balance for high flow conditions was performed for the
entire Sandusky system.  The mass balance equations for conservative
constituents were utilized.  Velocities in the Sandusky River ranged
from about 0.88 to 2.3 m sec'l so that the assumption of conservation
of mass throughout the system is reasonable for the particle sizes of
concern.

     One large source of uncertainty when  making  predictions of pollutant
concentrations based on flow frequency data or any method  in which temporal
continuity is ignored is the antecedent condition of the system.   For
instance,  in using the Soil Conservation Service  runoff curve number
method to  predict water yield,  the antecedent soil  moisture must be
estimated.  An analog can be drawn to estimating  sediment  yield from
either an  agricultural or an urban area.

     On agricultural lands the amount of sediment available for trans-
port can depend on a variety of factors.  Among these are the method of
planting,  time elapsed in the growing season, time since last rain-
fall, magnitude of the previous rainfall, and time since the last culti-
vation.  All these factors are time variant.   For instance, in one year
                                     63

-------
  CO
  0)
  Cl
X.
CD
CJ
z
o
<->„
 O7
UJ
Q_
                                                                 +      +
                                                               •K

                                                                     %
  °-l.SD    -0.74    0.02     0.78     1.54     2.30     3.OB     3.82     4.SB
                               LOG  FLOW  tCFS)
   Figure  3.2-6.   Sediment rating curve for Sandusky River  near
                   Fremont, October  1-31,  1976.
                                      64

-------
a farmer may choose to use conventional  tillage methods whereas the next,
he may use a no-till method of planting.   Timing and magnitude of rain-
fall events vary considerably from year to year.  Erosive rainfall  events
not only wash sediment from watershed surfaces but also break up soil
aggregates making more fines available for transport.   Cultivation
also breaks up clods and surface crusts  and generates  soil  fines.
Thus sediment yields for any given event depend largely on  the activ-
ities that occurred perhaps weeks before the event itself.   It is  easy
to see why sediment yields are so variable and consequently difficult
to predict from agricultural areas.

     In urban areas the situation is similar.  Here, deposition of
solid matter onto streets and concrete surfaces may be a function  of
traffic, street cleaning frequency, and  atmospheric conditions.
Timing, and magnitude of previous rainfalls are particularly impor-
tant in urban areas to set initial conditions for sediment  yield
simulation, because deposition is more uniform in time than the
generation of soil fines in agricultural  areas.  Thus  the available
amount of solids for washoff is largely  a function of  the time elapsed
since the last major washoff occurred.

     For non-urban loads, the "R" factor in the USLE was used to predict
loadings for each event.  The "average"  high flow event loads were then
computed.  However, urban loads could only be estimated on an annual
basis.  (Techniques are available to estimate single event urban loads
but are not included in this demonstration.  See Amy,  ejt a_]_., 1974.  The
user should note that if this technique were used, the following assump-
tions would be unnecessary.)  On this level of complexity,  determination
of what portion of the annual urban load can be assigned to the "average"
high flow washoff event is tenuous at best.  As a result, two extreme
cases were considered.  For the first, the annual urban sediment load  is
assumed to enter the stream equally distributed over, each day of the year,
For the second, the annual urban load is assumed to enter the river equal -
                                    65

-------
ly distributed over the single seven-day "average" high flow period.
These two cases should represent reasonable upper and lower bounds for
calculation of the instream concentration.
      The  urban  loads were  superimposed on  the non-urban sediment  loads.
As  a  result  three  concentrations were calculated for each high flow
reach.  Figure  3.2-7 shows  these concentrations predicted for the
Sandusky  River  at  Bucyrus.  The smallest concentration is the contribu-
tion  from non-urban loads  alone.  Next in  magnitude is non-urban  plus
the urban load  equally distributed over each seven-day period in  the
year.  The largest concentration results from the assumption that all
the urban loads are released during  the average seven-day high flow
period.   Historical flow weighted means and standard deviations are
also  shown on the  figure.

      In the  figure the non-urban loads appear to make-up, a substantial
portion of the  maximum "probable" instream concentration (the non-
urban  plus the  seven-day annual urban load release).  Table 3.2-9 shows
the percent  of  the non-urban contribution  to this maximum probable con-
centration for  several gage locations in the system.  Based on this
analysis  it  is  concluded that  the control  of sediments from agricultural
areas  has a  definite impact in the lower reaches of the river and less
impact in the upper reaches.  At the Bucyrus location, controlling urban
washoff appears to have more impact  on local water quality.
 3.2.10  Nitrogen and  Phosphorus Balance

      The methodology  for  routing nutrients at high flows is similar to
 that  used  for  suspended sediment.  The same sources of error are present
 due to  the lack of  definition of initial conditions.  For example,
 application  times for fertilizer and  the formulation of the fertilizers
 used  may change.  The variability of  rainfall also remains a problem.

      With  nutrients,  however, there are additional sources of error that
 tend  to make predictions  more difficult.  The first is that the correct-
                                   66

-------
cc
H
u
V)

Q
LU
Q
Z
LU
Q.
W

to
      500
     400
      300
200
100
            Observed Concentration
            +/- Standard Deviation

            Predicted Non-Urban

            Non-Urban & Urban as
            continuous release

            Non-Urban & Urban as
            7-day plug release
                                                     O
         '0.85
                     2.8
28.3
                                  7-DAY MEAN FLOW

                                      (m3sec~1)
  Figure  3.2-7.   Predicted and  observed suspended  sediment concentrations
                   for the Sandusky River at Bucyrus.
                                         67

-------
   TABLE  3.2-9.   EXPECTED  PERCENT  OF  NON-URBAN
  CONTRIBUTION TO "WORST CASE"  CONCENTRATION  OF
       SUSPENDED  SEDIMENT  AT  HIGH  FLOWS
                                      % Non-Urban
   Location                          Contribution
Bucyrus                                  29%


Upper Sandusky                           57%


Mexico                                   71%


Fremont                                  60%
                        68

-------
 ness  of  nonpoint  source  loading  estimates  is  predicated  on  accurate
 prediction  of  sediment loadings.   Second,  the assumption of conservation  of
 nutrient forms  other  than  total  nitrogen or  total  phosphorus  is  generally
 not valid.

     The nonpoint loadings supplied for the Sandusky system were given
as available nutrient forms; that is, forms available for uptake by
terrestrial or aquatic plants.  Typically,  these forms include the
ammonia, nitrite, and nitrate nitrogen and dissolved orthophosphorus
(PCL) forms.  The question which must be addressed is how to convert
these loadings given as available forms to total nitrogen or total
phosphorus which can then be treated conservatively.

      There  is no definable relationship between total and orthophosphorus
which can be extrapolated  from one watershed  to the next because of the
differences in  soil types, chemical watershed  processes  and degrees of
urbanization that exist  between  different  basins.  For instance  the
total 'and orthophosphorus "in the effluent  from sewage treatment  plants
will  very likely have similar values.  From nonpoint sources, mineralized
forms and organic phosphorus forms will likely be  present in significant
quantities.  Thus,  for a highly  urbanized  watershed, the ratio of instream
ortho- to total phosphorus should be higher than for a non-urban
watershed.

      Fortunately, in  the Sandusky  Basin some  relatively  extensive
 phosphorus  monitoring has  been done.  Both orthophosphate and total
measurements have been made.  The  Ohio EPA (1978)  has presented
weighted average orthophosphorus and total phosphorus concentrations
 for several locations in the watershed.  These are shown in Table
 3.2-10.   These  ratios have been  used for conversion between  the  two
parameters.  The background total phosphorus  concentration in natural
waters was taken to be 0.4 mg H~  for use  in  the mass balance equations.

     Some results of the total phosphorus mass balance are shown in
Figure 3.2-8.   The smallest predicted value represents the non-urban
contribution.   As for the suspended sediment  plots, the next greater
value represents the non-urban plus wastewater treatment plant effluent
and the annual  urban nonpoint loads distributed over the entire year.
                                   69

-------
       TABLE 3.2-10.  ORTHO AND TOTAL PHOSPHORUS RELATIONSHIPS
                     IN THE SANDUSKY RIVER BASIN

Station
Tymochtee
Bucyrus
Upper Sandusky
Mexico
Tiffin
Parameter
Ortho
Total
Ortho
Total
Ortho
Total
Ortho
Total
Ortho
Total
Number
of Obser-
vations
563
593
309
303
570
554
336
360
144
203
Weighted
Mean
Concen-
. tration
(mg £-1)
.071
.499
.230
.563
.139
.580
.098
.563
.102
.550
Ratio
(ortho/
total )
0.14
0.41
0.24
0.17
0.19
After:  Ohio EPA (1978)
                                   70

-------
  2  1.20
  GC
  I-
  LU
  O

-------
The highest value is the sum of the non-urban, waste water treatment plant
effluent and urban diffuse loads assuming the total annual accumulation
is washed off in one seven-day period.

     The estimated contribution from non-urban areas appears to be
small compared to the contribution from urban areas.  However, inspection
of the predicted values for stations upstream from which there is no
substantial urbanization suggests that the predicted values of total
phosphorus are probably low.  Even so, the phosphorus concentrations
appear to be moderately high with regard to eutrophication potential,
even from only the non-urban sources.  The U.S.. EPA (1973) has suggested
that 50 yg £~  may be an upper bound for limiting noxious plant growth
in flowing waters.  However, concentrations as low as 20 ug H~  are not
uncommon in eutrophic lakes.  The total phosphorus predicted concentrations
in the Sandusky system from- non-urban sources are above this value.  Ob-
served water quality data, however, suggest even higher concentrations.
The addition of urban point and nonpoint loads places the Sandusky River
waters well into the potential range for eutrophy.

     As with suspended sediment, the control of urban discharges seems
more critical than the control of non-urban sources for phosphorus con-
trol in this watershed.

     No exact relationships exist for converting available nitrogen to
total nitrogen and no data were available to calculate empirical relation-
ships as was done with phosphorus.  Nitrogen was routed as inorganic
forms only (ammonia and nitrate-nitrogen) because total nitrogen water
quality data were unavailable for making comparisons.

     Nitrogen levels were not predicted as accurately as phosphorus and
suspended sediment.  Table 3.2-11 shows the relative contribution of
non-urban nitrogen to the total predicted concentration assuming all the
annual nonpoint urban loads washed off in one seven-day period.  Once
again, urban sources appear to have the greater influence on possible
                                    72

-------
    TABLE 3.2-11.  EXPECTED PERCENT OF NON-URBAN
     CONTRIBUTION TO "WORST CASE" CONCENTRATION
                OF INORGANIC NITROGEN
                                      -  % Non-Urban
   Location                            Contribution
Bucyrus                                      5%

Upper Sandusky                              15%

Mexico                                      25%

Fremont                                     16%
                           73

-------
available nitrogen concentrations.  Figure 3.2-9 shows predicted
versus observed available nitrogen levels at Bucyrus in the Sandusky
River.  The trend indicated in this figure is representative of the
other locations; that is, predicted nitrogen concentrations were low
compared to observed instream data even using the "worst case"
assumptions.

     Typically the epilimnetic inorganic nitrogen threshold for meso-
eutrophic waters is 0.30 to .650 mg £~1 and for eutrophic waters,
0.50 to 1.50 mg £~1.  Predicted nitrogen levels place the Sandusky
River waters into these categories.  Historical observations place
nitrogen levels in the 4.0 to 9.0 mg £   range.
3.3  DEMONSTRATION EXAMPLE:  THE CHESTER RIVER

     The Chester River system was utilized principally to demonstrate
the estuarine water quality section of the nondesignated 208 screening
manual.  The methodology as set forth in the manual calls for assess-
ment of existing water quality followed by projected water quality under
new conditions.  Typically, this is the direction a 208 planner would
take.  The methodology is demonstrated, however, without fabrication
of "future" scenarios by analyzing existing water quality with regard
to the possible causes of degradation.  The estuarine calculations were
made both under low and high flow scenarios just as for the rivers and
streams demonstration on the Sandusky River system.
3.3.1  Data Collection

      Topographic maps  (7% minute) were obtained for the Chester River
 basin from the U.S. Geological Survey.  The USGS also provided flow
 records and stage-discharge curves for streams flowing into the
 estuary.  Ten years of recent streamflow data were obtained.  The
                                      74

-------
       10
  UJ
  O
Z 2
< O
uj o
V LU  C
<%
Q O
  O
  E
  o
  
-------
U.S. EPA's STORE! system was the primary source of water quality data,
Additional data were also found in a three volume series of reports
on pollutants and sediment in the Chester River done by the Westing-
house Electric Corporation in cooperation with the State of Maryland
(1972).  Listings of industrial and municipal point sources in the
basin and effluent characteristics for those of import were also
obtained from the Maryland Department of Natural Resources.


3.3.2  Data Reduction and Supplementation

3.3.2.1  Hydrologic and Hydraulic Data

     Streamflow data were available for two tributaries to the
Chester:   Unicorn Branch and Morgan Creek.   Unicorn Branch is one
of four creeks with their confluence at or near Millington which
form the  headwaters of the Chester.   Morgan Creek enters the Chester
just above Chestertown.

     For  the purposes of evaluating estuarine conditions under low
flow, the 7Q1(, flows were estimated for these two creeks.   Figure
3.3-1 shows the flow frequency plot.   Although the flow in Morgan
Creek appears smaller, actually it is not.   The gage on Unicorn
Branch is located at the subshed outlet while the Morgan Creek gage
is located at Kennedyville,  several  miles upstream of the outlet.
From this information the 70,Q flows for the major creeks in the
basin were estimated (by area!  weighting) and are shown in Table
3.3-1.

     Because of the paucity of hydraulic data for streams tributary
to the Chester River it was decided to perform a "worst-case" water
quality analysis for the low flow condition.  Loadings for the point
sources in the basin were estimated and were assumed to enter the
Chester  River at full  strength; that  is, with no decay  instream  and
                                 76

-------
   100
   90
   80
   70
   60

   50

   40

   30
   20
I
_    I      I   I   I
      2%   5     10  15  20   30   40   50  60  70   80  85  90    95     98%

               PERCENTAGE OF TIME  FLOW IS LESS THAN SPECIFIED FLOW




     Figure  3.3-1.  Frequency analysis  of 7-day  annual low flows.
                                  77

-------
TABLE 3.3-1.  LOW AND HIGH SCENARIO FLOWS FOR CHESTER RIVER TRIBUTARIES
Creek
Langford Creek
Radcliffe Creek
Corsica River
Reed Creek
Southeast Creek
Red Lion Branch
Sewell Creek
Andover Creek
Cypress Creek
Mills Branch
Gray's Inn Creek
Morgan Creek
Unicorn Branch
7Q1QFlow
(m3 sec"1)
0.23
0.04
0.20
0.07
0.23
0.15
0.10
0.25
0.19
0.07
0.04
0.18
0.12
High Flow
(m3 sec"1)
8.69
1.56
7.62
2.61
8.64
3.34
2.29
5.49
4.28
1.53
1.44
6.51
2.78
Area of
Sub-basin
(km2)
110.0
19.7
109.8
33.1
99.7
63.2
48.7
116.5
90.6
32.6
18.1
82.3
58.8
                   TOTAL        1.88           56.78          883.1
                                 78

-------
only diluted by the flow of the tributary (if any) by which they are
carried.  As such, velocities and depths of flow were not calculated
for these streams.

     High flows were developed by selecting fifteen high flow events.
The criteria for the selection were the same as used in the Sandusky,
i.e., only events occurring from April to September were considered
and storms with good areal  coverage were preferred.  There was no
minimum flow criterion.  The flows from these storms were averaged
for Morgan Creek and Unicorn Branch and areal weighting was used to
determine the flows in other creeks.  The results are also shown in
Table 3.3-1.
 3.3.2.2   Water  Quality  Data

      Water quality data for  the Chester  River, collected principally
 in  1970 and  1972, were  tabulated  to  indicate which water quality
 parameters would be  used  in  the demonstration and where water quality
 problems  existed.  Salinity  profiles were  plotted for each of the dates
 available and were grouped into categories  representing "high" and  "low"
 flow  regimes based on flow records at  the  Unicorn Branch gage.  From
 this  grouping a representative high  and  low flow salinity profile was
 developed.   These profiles indicate  how  pollutants are distributed  in
 the estuary  under certain flow regimes and  are useful for other estuarine
 computations.   Figures  3.3-2 and  3.3-3 show typical  salinity profiles  for
 the high  and low flow conditions, respectively.  The profile of June 30,
 1972  was  taken  after the  passage  of  Hurricane Agnes.  The mean daily flow
 in  Unicorn Branch was 3.82 m^ sec'l  and  followed a period of 14 - 17
 m3  sec'1  mean daily  flows at that gage.  The profile of October 30, 1972
 is  a  typical low flow salinity profile for  this system.  The salinity
 differences  between  surface  and bottom at  this time  are only about
 one-half  ppt while during high flows the difference  may be two to three
 ppt..  The flow  at Unicorn Branch  on  October 30 was 0.42 m3 sec~l.

                                79  "

-------
                 10
              O  6
                                                                                          June 30, 1972
                                                                                               Surface

                                                                                               Bottom (35')
CO
O
              Q.
              a
>-


2
_l
<
                                      16
                                                           I
                                                      I
                                  24         32         40 :

                                     DISTANCE FROM MOUTH (Km)
                                                                              48
                                                                                         56
                                                                                                  64
                                                                                                            72
                          Figure  3.3-2.  Salinity profile for  the Chester  River, June 30,  1972.

-------
00
                12
                10
in
CM
ra
Q.
a

z
_J
CO
                                                                                       October 30, 1972
                                                                                           • Surface
                                                                                           A Bottom
                                                I
                                              I
I
                                      16
                                   24        32         40         48

                                       DISTANCE FROM MOUTH ( Km )
                                                                                        56
                                                                                                 64
                                                                                                            72
                       Figure 3.3-3.   Salinity profile  for the  Chester River, October 30, 1972.

-------
     Vertically averaged salinity profiles for the high and low flow
 scenarios are shown  in  Figure 3.3-4.  For convenience in applying the
 methods and to provide  for smoothness of the data, second degree poly-
 nomials were fit to  the salinity data.  These are shown also in Figure
 3.3-4.  The equations and their corresponding correlation coefficients
 are given in the upper  right-hand corner.  These equations represent
 regressions of salinity (ppt) on river miles from the mouth of the
 estuary (1 mi = 1...61 km).  Because of the averaging used to arrive at
 these profiles the low  flow profile should be indicative of that
 occurring when Unicorn  Branch is flowing at approximately 0.39 m3 sec'1.
 the high flow profile represents salinities corresponding to a flow of
 about 1.98 m3 sec'1  in  Unicorn Branch.  The 7Q,Q low flow at Unicorn
 Branch used for estuarine water quality computations is 0.12 m3 sec'1
and the high flow rate used is 2.78 m3 sec"1.   Therefore these
salinity profiles are reasonably representative of actual  salinities
at those flows.

     Historical  temperature data are shown  in  Table 3.3-2  for dif-
ferent locations in the estuary.   There is  a  pronounced  seasonal  varia-
tion in water temperature with temperatures seeming to  increase slightly
in the landward  direction,  regardless  of season.   Only  slight temperature
variations are observed over depth.

     Table 3.3-3 shows dissolved oxygen data  for  the same  dates and
locations  as for temperature.   Sags are most  noticeable  during  the
summer-autumn low flow periods.   The numbers  in parentheses  are the
standard deviation  of the parameter over readings taken  at varying
depths.   High standard deviations indicate  a  trend towards stratifi-
cation while low standard deviations indicate  a well-mixed system.
Higher deviations seem to occur most often  in  the warmer months.

     Data  for fecal and total  coliform bacteria were not available
for this system  even though high levels of  these  indicator organisms
                                 82

-------
CO |
to j
                   10
                               I           I
                                  Low
                                  Flow
                                  Salinity  <
                                   •

                                 "X
                                                                              VLOW = 9'63 ~ °0593 *'  r = ° 931



                                                                              VHIGH = 6'79 ~ -00366 *J ' = °-856

                 O

                 m
                 CM

                 ra
                                          N
                   —^•^


                  High

                  Flow

                  Salinity


•  Low Flow

A  High Flow

                                                              I
                                                 I
 \l  \
                                         10         15         20         25


                                                  DISTANCE FROM MOUTH (miles)
                                                           30
                                                                     35
   40
             45
Figure 3.3-4
                                        High and  low flow  vertically averaged  salinities in  the

                                        Chester River Estuary.

-------
            TABLE 3.3-2.   VERTICALLY AVERAGE  TEMPERATURES  FOR THE CHESTER RIVER (°C)
RIVER KILOMETER DATE
720301 700312 700402 720522 700604 720619
1.6
6.1
10.8
14.6
g 24.5
29.6
34.6
43.4
48.3
54.7
62.4
(Love Pt. Light) 3.1 3.05 5.85 15.6 21.05 21.5
16.0 - 21.3
3.2 - - 15.6 - ' 21.7
(Long Pt.) 3.2 3.73 5.33 15.65 20.63 21.85
(Boxes Pt.) 3.35 3.93 6.0 18.6 21.93 23.25
(Nichols Pt.) - 4.45 6.25 17.7 • 22.8 23.25
19.2 . - 23.7
(Melton Pt.) - 4.8 6.73 19.35 23.6 23.25
(Chester-town) - 5.65 7.35 19.5 24.6
(Possum Pt.) - 6.55 7.8 - 25.05
(Crumpton Buoy) - 8.1 6.5 - 25.0
a)
720630 720822 700904 721030
19.45 23.7 24.95 13.85
20.0 23.85 - 13.65
21.9 23.7 - 13.35
23.4 24.5 25.1 13.0
24.05 24.1 25.27 13.0
24.5 25.25 13.25
24.5 -
25.0 25.75 13.5
25.3 25.6
25.25
25.1
a;  Dates are given as year/month/day

-------
                  TABLE 3.3-3.   VERTICALLY AVERAGED DO CONCENTRATIONS  (mg jf1)  FOR THE CHESTER RIVERA
00
01
RIVER KILOMETER
720301
1.6
6.1
10.8
14.6
24.5
29.6
34.6
43.4
48.3
54.7
62.4
(Love Pt.) 12.9(.85)
-
12.9(.85)
(Long Pt.) 12.9(.99)
(Boxes Pt.) 13.0(.67)
(Nichols Pt.)
-
(Melton Pt.)
(Chester-town)
(Possum Pt.)
(Crumpton Buoy)
700312
11. 7(
-
-
11. 8(
11. 5(
11. 2(
-
10. 8(
10.2(
10. 3(
.38)


.06)
.06)
.07)

.12)
.14)
.07)
700402 720522
8.8(.71) 6
6
5
9.6(2.4) 5
9.6(1.6) 8
10.3(42) 6
7
9.7(.21) 7
9.6(.42 8
9.3(0)
.4(3.9)
.2(4.6)
.0(4.5)
.5(4.9)
.8(0)
.1(3.2)
.4(.85)
.6(.56)
.0{.71)
-
10.8(0)
DAT
700604
7.0(1.77)
; -
-
5.2(2.9)
6.9(.53)
6.3(.99)
-
6.5(.15)
5.6(.67)
5:.8( .07)
6.4(0)
£
720619 720630
6.2(1
6.1(1
7.0(.
6.2(1
7.7(2
6.8(1
6.3(1
5.9(1
-
-
-
.8) 7.7(.71)
.8) 6.4(2.5)
78) 6.9(1.7)
.5) 6.7(2.5)
.5) 7.0(1.1)
.8) - '
.6)
.3)
•
-
-
720822
4.6(4.9)
6.1(4.2)
4.8(4.2)
6.5(6.1)
5.1(4.4)
5.0(4.4)
5.9(1.8)
7.6(2.2)
8.1(.99)
-
-
700904
7.0(.56)
-
-
3.4(2.6)
4.7(2.2)
6.4(0)
-
6.0(.28)
6.2(1.1)
6.3(.42)
6.3(0)
721030
8.7(.71)
9.K.14)
8.3(.21)
8.6(.28)
8.3(.14)
7.8(.49)
-
9.5(.99)
-
-
-
       a)  Numbers In parentheses are standard deviations over depth.

-------
 have  resulted  in  closure  of many shellfish  harvesting  waters  in  the
 estuary.

      Nutrient  data  were available for  three dates  from the  1970
 investigations on the  river.   Chiorophyll-a_ measurements  were  also
 made  during  those studies.   These data  are  shown  in  Table 3.3-4.
 The parenthetic values are  again standard deviations over depth.
 Chlorophyll-a_  was sampled at  the one-foot depth only.

      Generally, total  nitrogen  profiles  indicate  relatively constant
 levels  of  nitrogen  over the length  of  the estuary  while phosphorus
 levels  increase in  the upstream direction.   There  is also an  increase
 in chlorophyll-a_  production from the mouth  to  the  head of the  estuary.
 Chlorophyll-a_  levels as of  1970 in  the  Chester River have generally
 been  in the  region  characterizing eutrophy  (>8 yg £  }-but  not at
 levels  at  which algal  blooms  are considered a  problem  (>50  yg  &-1).
3.3.3  Point Source Load Estimates

     Temperature, BOD,-, dissolved oxygen, flow, fecal  and total  coli-
form bacteria, total phosphorus, and nitrate and nitrite nitrogen data
were available for most of the sewage treatment plants in the Chester
River basin.  All parameters, except temperature, were averaged over
all available data to determine average plant effluent characteristics.
The temperature of the discharge was taken as the average of those data
taken only during the month in which low flow typically occurred.  Ef-
fluent data for major municipal sewage treatment facilities and indus-
trial dischargers are given in Table 3.3-5.

     Loads to the estuary for low flow periods were calculated in the
following way.  The flows of the discharge and the receiving stream
                                   86

-------
              TABLE 3.3-4.
PLANT  NUTRIENT AND CHLOROPHYLL-a  LEVELS
  IN THE CHESTER  RIVER
VERTICALLY AVERAGED TOTAL NITROGENa; (as N) (mg t~}

Love Pt. (MOUTH)
Long Pt.
Boxes Pt.
Nichols Pt.
Kel ton Pt.
Chester-town
Possum Pt.
Crumpton Buoy
700312
.93 (.06)
.73 (.12)
.85 (.06)
-77 (0)
.89 (.11)
1.55 (.03)
2.04 (.03)
1.57 (0)
700402
1.32 (.06)
1.05 (.21)
.84 (.08)
1.05 (.01)
1.09 (.15)
1.41 (0)
1.58 (0)
1.21 (0)
')
700604
1.00 (.32)
1.20 (.15)
.63 (.04)
.57 (.11)
.31 (.02)
.46 (.06)
.53 (.03)
.89 (0)

700904
_
-
-
-
-
-
-
-
VERTICALLY AVERAGED TOTAL P04 (as P04) (mg l~l)

Love Pt. (MOUTH)
Long Pt.
Boxes Pt.
Nichols Pt.
Melton Pt.
Chester-town
Possum Pt.
Crumpton Buoy


Love Pt. (MOUTH)
Long Pt. •
Boxes Pt.
Nichols Pt.
Helton Pt.
Chester town
Possum Pt.
Crumpton Buoy
700312
.10 (.02)
.07 (.02)
.11 (.02)
.09 (0)
.17 (.02)
.38 (.05)
.47 (.10)
.39 (0)
CHLOROPHYLL-a. (ug
700312
9.0
10.5
7.5
7.5
9.8
11.3
16.5
29.3
700402
.14 (.06)
.10 (.01)
.10 (.01)
.14 (.02)
.20 (.04)
.32 (.09)
.51 (0)
.63 (0)
I'1)
700402
15.0
1.5
7.5
7.5
1.5
6.0
9.0
10.5
700604
.11 (.01)
.17 (.15)
.14 (.01)
.12 (.04)
.26 (.01)
• 3.J5 (.02)
.31 (.01)
.32 (0)

700604
22.5
24.0
29.3
20.3
30.0
32.3
22.5
20.5
700904
.
-
-
-
-
-
-
-

700904
13.5
19.5
7.5
8.3
15.0
24.8
52.5
111.0
a)  The numbers  in parentheses are  the standard deviations over depth.
                                          87

-------
                            TABLE 3.3-5.  EFFLUENT  CHARACTERISTICS FOR MUNICIPAL STPs  AND
                                  INDUSTRIAL DISCHARGES  IN THE CHESTER RIVER BASIN
oo
00
Municipal
STP's
Eastern
Correctional
Canip
Chester town
Rock Hall
Centrevllle
Queens town
Suddlersvllle
Hill Ing ton
Industries
Campbell '$
Soup
Company

Tenneco
Chemicals
Tempera- BQD OQ Fecal Total Total po NQ 4 N0 N11 TKN
, ture 5. , Collform Col 1 form P . 4.2.3 3 . .
(nf sec"1) (°C) (mg t~l) (mg i"1) (MPN/100 ml)(MPN/100 ml)(mg i"1) (ntg l~l) (nig l"1) (mg t'1) (mg l~l)
.014 23e 23 7.4 3 '' 11 6 - .5
.84 23e 22 7.6 10 197 6e .22
.34 23e 24 7.8 7 21 6e
.43 23e 7 7.2 12 105 6e - - -
.10 23e 53 7.0 214 1,196 6e
.06 23e 18 7.5 . 65 848 6.2 6.0 .54 6.6 14.0
.05 23e 63 7.1 105 1,018 3.4

.21 17 24 3e 8.4 - - ...

.05 30e 46 5.2
Suspended
Solids
(nig it"1)
26
103
58
10
56
35
39

32

-
         (e) estimated

-------
were added to yield a total flow.  The temperature, BODg, dissolved
oxygen, fecal and total coliform bacteria loads were computed as flow
weighted averages.  The values of BOD and total coliforms for the
natural waters were taken from background iso-pollutant maps in
McElroy e_t al- (1976).  Background fecal coliform bacterial counts were
assumed to be zero.

     As an example showing how to calculate instream pollutant levels
below a sewage treatment plant,  total coliform bacterial  data from
the Suddlersville Sewage Treatment Plant on Red Lion Branch is used.
The flow from the plant is 0.002 m  sec   and the nautral 7Q,n flow  is
      01                               1-1
0.15 m  sec   for a combined flow of 0.152 m  sec   cfs.  The resultant
total coliform count  is
          TC =   -            °-15  (3°0) = 306 MPN/100 ml
where  848  and  300  are  the  total  coliform counts  of  the  plant  effluent
data and the natural background  count,  from  McElroy,  et al.  (1976),
respectively.

     Loads calculated  in this way tor low flow analysis are shown for
the municipal  sewage treatment plants and major  industrial  discharges
to the Chester River in  Table 3.3-6.
3.3.4  Estuarine Classification

     Geometrically the Chester appears marginal for application
to the screening calculations.  Four creeks provide about one-third
of the fresh water flow at the head of the estuary with the other
two-thirds resulting from fairly well longitudinally distributed
                                 89

-------
         TABLE  3.3-6.   LOW FLOW LOADS  TO THE CHESTER RIVER FROM MUNICIPAL
                    AND  INDUSTRIAL  POINT SOURCES  (PER  TIDAL CYCLE)
Volume
of Water
Source (m )

CBOD
u
(Kg)

NBOD a>
u
(Kg)

Total Coliforms Fecal Coliforms
MPN x 10~9 MPN x 10"8
Queenstown  STP
   126
 9.8
                                 9.4
                1.5
                              2.7
Rock Hall  STP and
Grays Inn  Creek
 2,200
17.8
39.3
                5.4
 .31
Centrevllle STP and
Corsica River
 9.646
18.4
37.5
               27.9
 .65
Eastern Correctional
Camp STP and
Southeast Creek
10,384
16.0
 1.2
                                                                         31.0
                               .0005
Chestertown  STP
and Radcl iff Creek
 2,958
37.0
96.9
                7.8
                                                                                        1.1
Tenneco Chemicals,
Campbell 's Soup
and Morgan Creek
 8,116
25.1
32.9
               24.3
Suddlersville STP
and Red Lion Branch
 6,776
12.0
 4.8
               20.7
4.9
Millington STP, Mills
Branch,  Cypress Branch,
Andover  Branch, Unicorn
Branch,  Sewell; Branch      32,806
Langford Creek
Reed Creek
10,366
 3,160
54.0


15.2


 4.6
                                4.7
               78.7
               31.1
                                                                          9.5
                               .66
   estimated
                                               90

-------
tributaries further seaward.  This violates the assumption of only
one dominant inflow.  No major side embayments exist on the Chester,
however.

     Both classification methods  (flow ratio method and stratifica-
tion-circulation method) were used in attempting to classify the
Chester River estuary.  The flow  ratio was calculated under low and
high flow conditions.  The estimation of these flows has already
been described.  The estuary tidal prism used in the flow ratio
method was calculated in the following way.

     Transects were drawn normal  to the flow at convenient locations'
on the USGS topographic maps.   Bathymetric cross-sectional  pro-
files were then constructed.  Depths given on the maps are for mean
low water.  A mean tidal range is also given.  By assuming that
only depth and not the'wfdth of the channel changes under tidal fluc-
tuations, a mean high water cross sectional area can be estimated.
Using the length between transects the MLT (mean low tide) and MHT
(mean high tide) volumes of the estuary are calculated.  The tidal
prism is the difference of these  two values.  For the Chester River
                                                 7  3
the tidal prism volume is approximately 8.46 x 10  m .

     River flow volumes over the tidal  cycle are computed as

                             V = KQrt
where
     V  = volume of fresh..water inflow per tidal  cycle (m^)
     Qr =-fresh water flow rate (m  sec~ )
     t  = period of the tidal  cycle (hr)
     K  = a units correction (3600 sec hr" )

For the Chester River the tidal  period is M.2.4 hours.   This was deter-
mined from co-spectral  density plots for February tides (State of
                                91

-------
Maryland,  1972).  Qr  is  taken to be the sum of all the fresh water  in-
flows  in the  basin.   Use of  these  values gives a flow ratio of 9.6  x  10
for  low flow  and 2.9  x 10"2  for high flow.  This indicates that  the
Chester River is a well  mixed estuary  under both conditions (FR  <0.1;
see  Section 6.3.5 of  the screening manual).
     The  Stratification-Circulation  (or  Hansen-Rattray) method gives  a
 different result.   Using data  obtained in  the  Chester  River  study
 (State of Maryland,  1972)  the  stratification and circulation parameters
 were computed  for  both  the high  and  low  flow conditions at Love  Point
 Light.  AS in  the  stratification parameter was computed using August
 and September  (low flow) salinity data and salinity after the passage
 of Hurricane Agnes (high flow).   The time  averaged August and September
 surface salinity was  9.6.ppt and the average bottom salinity was 9.85
 giving AS of 0.25.  The averaae  of these two values gives S  , the
 cross-section  mean salinity of 9.7 ppt.  Similarly, for high flow  AS
 is equal  to 3.3 and S  = 2.03.   Using these values gives values  of the
 stratification parameter (AS/S ) of  0.025  for  low flow and 1.64  for
 high flow.  U.c, the mean fresh water velocity, was calculated for  these
 two cases by dividing the  high and low flow rates by the mean tidal
 cross-sectional area  at the river mouth  (Love  Point Light).   The flow
 rates  (from Table  3.3-1) are 1.88 m3 sec"1 (low flow)  and 56.8
 m3 sec'1  (high flow).   The mean  tidal cross section is 62296 nr  at
Love Point.  A mean fresh  water velocity (UJ  is  computed  for each
condition of 3.0 x 10   m sec"    (low flow)  and  9.1 x 10   m sec"
(high flow).  Us,the tidally averaged surface  velocity, was  determined
by measurement (State of Maryland, 1972)  to be  less  than  0.01 knots
or approximately 0.004 m sec"   .  This measurement was  taken  during  a
period of intermediate inflow  in  the tributaries  and as such  represents
a "median" value of this parameter, being neither representative  of the
high or low flow conditions.  However, it is  used as  the  value of U
for both  sets of conditions since it represents  the only  available
measurement.  This gives values for the  circulation  parameter (U  /Uf)
of 133 and 4.4 for low and  high flow, respectively.
                                92

-------
     Plotting the above values on a stratification-circulation diagram
shows that the mouth of the Chester falls into categories 2a and 2b,
indicating a partially mixed estuary.  This result is corroborated by
the Chester River Study (State of Maryland, 1972).  They report that
Chesapeake Bay and its major tributaries belong to the partially
mixed type estuary and cite Pritchard (1967).  A value of U  was not
available for the upper portion of the estuary and calculations could
not be made for this section of the river.
3.3.5  Flushing Calculations

     The tidal prism, modified tidal prism and fraction of fresh water
methods were used to calculate flushing times for the Chester River
and several of its major tributary estuaries.  Flushing times calculated
for the Chester River by each of the three above methods are given in
Table 3.3-7.  A comprehensive example of the flushing time calculation
is given in the Patuxent River section.  The tidal prism method does
not take flow into account and only one value is shown for it.  For
the other two methods, high and low. flow flushing times are given.   Addi-
tionally, the fraction of fresh water method was performed using a
2 ppt and a 1 ppt segmentation scheme to demonstrate the sensitivity
of the method.

     Although the river flow rate is used explicitly in the modified
tidal prism method, the flushing times seem to be fairly insensitive
to flow, the method yielding times of approximately equal magnitude
while the river flow varied by a factor of 30.  This indicates that
the Chester River is flushed primarily by tidal action and not by
advective flow.

     The fraction of fresh water method on the other hand seems
extremely sensitive to flow condition,giving greatly different flush-
ing times of ^380 days'.and V13 days for low and high low, respectively.

                                  93

-------
      TABLE 3.3-7.   FLUSHING TIMES FOR THE CHESTER RIVER
                       BY THREE METHODS
                                   Flushing Time (days)

   Method                       High Flow           Low Flow

Tidal Prism                       5.3                 5.3
Modified Tidal
Prism                           143.                134,
Fraction of
Fresh Water
(1 ppt)a;
(2 ppt)
13.6
12.7
381
382
 'The estuary was segmented using first a 1 ppt salinity
  difference per segment and then a 2 ppt salinity
  difference per segment.
                            94

-------
The results are somewhat afield from those obtained using the modified
tidal prism method possibly because the salinity profiles used were
not measured at the same flow rates used in the modified tidal prism
method.  Even if this were the case, salinity profiles are heavily in-
fluenced by antecedent flow conditions in estuaries with long residence
times.  The high flow flushing times are approximations since this
estuary likely stratifies at high flows, invalidating the assumption of
a well mixed system.

      The fraction of fresh water method shows little improvement when
a one-part-per-thousand segmentation scheme is used instead of two parts
per thousand.  The 1 ppt segmentation requires only a nominal additional
effort above that required using 2 ppt.

     'Flushing  times were  also computed  for  several  of  the major  tidal
 tributaries  to  the  river  for high  and  low  flow  scenarios using both
 the  tidal  prism and modified tidal  prism methods.   The  results are
 given  in Table  3.3-8.  The  low  flow flushing  times  show a generally
 increasing  trend  with  the MLT volume of the estuary with the modified
 tidal  prism  values  being  consistently  higher  than  those computed by
 the  tidal  prism method.

      Because of the relative ease  of applying the  tidal prism method
 as opposed  to  the modified  method,  it  would be  desirable to  use  the
 former if  possible  in  a  screening  procedure.  However,  it  is known
 that this  method  underestimates  the true  flushing  time  considerably
 (Officer,  1976).   For  this  reason,  ratios  of  the  two methods were
 computed and are  also  shown in  Table 3.3-8.   For  the high  flow regime.
 the  ratio  of the  values  produced  by the two methods are consistently
 around 4.0 for the  tributary estuaries.   However  for the entire
 river this ratio  is 22:1.   For  the low flow regime, the ratios are
 variable and seem related to the  MLT volume of  the  estuary.   Figure
 3.3-5 shows  an empirical  relationship  between flushing  time  ratio
                                 '95

-------
TABLE 3.3-8.  FLUSHING TIMES FOR THE  CHESTER  RIVER  AND  SELECTED TRIBUTARIES
Flushing Time
River/Creek
Langford Creek
(East Fork)
(West Fork)
Corsica River
Gray's Inn
Creek
Chester River
MLT
Volume
(m3)
2.07 x 107
1.76 x 107
1.19 x 107
3.68 x 106
7.39 x 108
Tidal
Prism
Method
4.3;
5.0 .
2.2
3.8
6.4
Tidal
|High
16.. 7
17.7
9.8
15.0
140.1
(days)
Modified
Prism Method
Low
69
59
18.1
26.4
136.4


Ratio
MTP/TP Method
High Low
3.9 16
3.5 11
4.5 8
3.9 6
21.9 21
.0
.8
.2
.9
.3

-------
  25
  20
  15
UJ
2
C3
Z
I
  10
                                                                Patuxent River
                                             • Ware River
                                                          y =2.73 (In x) - 33.76
    1.0 x 10'
1.0 x 10'                  1.0 x 10'
   MLT ESTUARY VOLUME (m1)
    Figure 3.3-5.   Empirical relationship  between the  ratio of modified
                     tidal  prism  and tidal prism methods  and mean  low tide
                     estuary volume.
                                       97

-------
 and  MLT  volume.   Also  plotted  on  the  figure  are  points  calculated
 for  the  Ware  and  Patuxent  Rivers.   The  agreement between  the  rela-
 tionship developed  on  the  Chester River and  applied  to  two  other
 estuaries tributary to the Chesapeake Bay  is good.
 3.3.6   Pollutant  Distribution

     Pollutant  distributions in  the  Chester  River  are  analyzed  under
 both high  and low sets  of  flow conditions.   The  low  flow  scenario
 considers  the direct  discharges  of point  sources into  the estuary
 and streams  flowing into the estuary.   The high  flow scenario con-
 siders  primarily  nonpoint  source discharges  into the system.  For
 the purposes of this  analysis, nonpoint loadings distributed along
 the length of the river are treated as discrete point loadings
 into'each  of several  segments  into which  the river has  been
 separated.
3.3.6.1  Low Flow

    Figure 3.3-6 shows a schematic of the Chester River with the major
point source discharges.  The loads from these dischargers have been
listed in Table 3.3-6.  The first step in performing the pollutant
distribution calculations is to determine which of the 30 modified
tidal prism segments receives the discharge.  Having done this, the
initial concentration in the segment of discharge (Cd) can be calculated
for the appropriate segments by dividing the load from the point source
(Kg tidal cycle  ) by the hypothetical fresh water volume passing through
the segment during a tidal cycle.  This is expressed as:

                             r  - W f
                             Cd   R rd
                                98

-------
              Chesapeake
                 Bay
10
<£>
             01
             CD
                            Rock Hall
                              STP
                     -8.9 Km-
                                  Chestertown     Tenneco Chemicals
                                     STP        Campbell's Soup
                                                                    6.4 Km-
                                                                             -4.8 Km-*
•3.2 Km-
            • 13.7 Km •
                                                                                  Millington
                                                                                     STP
                                                                                      I
                              • 11.3 Km-
            Queenstown
               STP
Centreville
   STP
  Eastern
Correctional
   Camp
   STP
Suddlersville
   STP
          8.0 Km
                          Figure 3.3-6.   Schematic of Chester  River  and  point  sources (not to  scale)

-------
    where C . = the initial concentration in the segment of discharge
           d   (Kg m-3)
          W  = mass of pollutant per tidal  cycle from the point source (Kg)
                                                        2
          R  = Volume of river inflow per tidal cycle (m )
          f . = fraction of fresh water in the segment of discharge
               (unitless)

Total coliform bacteria concentrations and ultimate NBOD plus CBOD con-
centrations were calculated for the Chester River.  The results are
shown in Table 3.3-9.  Each initial concentration shown is due solely
to a single point source. From the calculated concentrations it can be
concluded that these point sources have little impact on the water
quality of the main stem of the Chester River, although they quite
possibly may be degrading quality in tributary estuaries into which
they discharge before entering the main river.  Because these concen-'  • •
trations are so low, the technique of estimating their longitudinal
distribution (which is essentially to-superpose individual concentration
profiles) is not demonstrated here but is discussed in the Patuxent
River section.
3.3.6.2  High Flow

     The Chester River basin was divided into sub-basins, and nonpoint
loads under "average" high flow conditions were estimated for each
sub-basin.  In the following calculations, a delivery ratio of 0.1 was
used for all the nonpoint loadings (Midwest Research Institute,  1979).
Since the effects of point sources were judged to be of minor importance
in the previous section, their contributions to the total  loads were
not considered for the high flow analysis.  The parameters analyzed
under this scenario were BOD5, total nitrogen, total phosphorus, and
suspended sediment.  All parameters were treated as conservative
materials with the exception of BOD,-.
                                 100

-------
   TABLE  3.3-9.   CALCULATED  INITIAL CONCENTRATIONS IN THE CHESTER RIVER
                     FOR TWO WATER QUALITY PARAMETERS
  Point Source
    /Segment
Total Coliform
   Bacteria
 (MPN/100 ml)
(NBOD plus  CBOD)i
     (ma a~1}
Queenstown STP/22
Rock Hall  STP/18
Centreville STP/16
    0.11
    0.90
    5.9
   1.37 x 10
                                                                 -2
   9.50 x 10
                                                                 -2
   1.19 x 10
                                                                 -1
Eastern Correctional
Camp STP/11
Chestertown STP/9
   16.7 '


    6.0
   9.20 x 10


   1.03
                                  -2
Campbell's Soup Co.
and Tenneco Chemical
Co./8
Suddlersville STP/3
Millington STP/0
   20.6


   23.7


   93.8
   4.92 x 10
                                  -1
   1.92 x 10
                                                                 -1
   7.00 x 10
                                                                 -1
                                  101

-------
     There are two cases to consider in the treatment of nonpoint
sources.  First, there are sub-basin loads which enter as a point
source to the estuary via the tributaries.  Also, there are distributed
loads from sub-basins adjacent to the estuary.  Tributary inflows are
easily assigned to one of the 23 segments delineated by the modified
tidal prism method.  Usually, distributed inflow loads from adjacent
sub-basins have units of ML"  (mass per unit length).  In this case
a portion of the total load assigned to each  segment  is calculated by
multiplying the load per unit length by segment length.  For instance,
if the loading rate from an area adjacent to the estuary is 3 Kg/Km of
estuary length and the area adjoins three 1 Km- segment, the loading into
each is 3 Kg for a total of 9 Kg in all three segments.  Initial concen-
trations are computed for each high flow segment just as in the low flow
case.

     Once the initial concentrations for each segment are determined,
they are distributed in the estuary in the upstream direction as the
salinity gradient  and downestuary as the fraction of fresh water.
This is most easily accomplished by setting up a coefficient matix
for the system.  This matrix,, called the distribution matrix^ is made
up of coefficients f-j/fj for segments downstream of the discharge segment
(below the main diagonal of the matrix) and S./S . upstream of the dis-
charge segment (above the main diagonal) with coefficients of unity along
the matrix diagonal.  The f.. are fractions of fresh water in the i
segment, and the d subscript denotes the discharge segment.  The same .
subscripts apply to the salinities (S).  Table 3.3-10 shows the high
flow distribution coefficient matrix for the Chester River.

      Conservative pollutant distribution in the estuary is determined
as follows.  First, the following terms are defined:
      ^dm = initial concentration of a pollutant that is discharged
            into segment m
      p   = number of pollutant dischargers into the estuary
                               102

-------
TABLE 3.3-10.
CHESTER RIVER CONSERVATIVE POLLUTANT DISTRIBUTION
  COEFFICIENT MATRIX (HIGH FLOW)


0
1
2
3
4
5
6
7
t-' 8
O .
CO j 9
1 u 10
Ul
I 11
£ 12
5 13
(/I
14
15
16
17
18
19
20
21
22
23
0

1
.82
.67
.54
.44
.34
.22
.18
.16
.15
.13
.10
.09
.06
.04
.03
.03
.01
.01
.01
0
0
0
0
1

0
1
.82
.66
.54
.41
.27
.22
.19
.18
.16
.12
.11
.07
.05
.04
.04
.01
.01
.01
0
0
0
0
2

0
.54
1
.80
.66
.51
.33
.27
.24
.22
.19
.15
.13
.09
.06
.04
.04
.01
.01.
.01
0
0
0
0
3

0
.39
.71
1
.81
.63
.41
.33
.36
.28
.24
.18
.17
.11
.07
.06
.06
.02
.02
.02
0
0
0
0

<:}
0
.31
.58
.82
1
.77
.50
.41
.36
.34
.30
.23
.20
.14
.09
.07
.07
.02
.02
.02
0
0
0
0
5

0
.27
.49
.69
.84
1
.65
.53
.47
.44
.38
.29
.26
.18
.12
.09
.09
.03
.03
.03
0
0
0
0
6

0
.23
.42
.58
.72
.85
1
. -82
.73
.68
.59
.45
.41
.27
.18
.14
.14
.04
.04
.04
0
0
0
0
7

0
.21
.39
.55
.68
.80
.95
1
.89
.83
.72
.56
.50
.33
.22
.17
.17
.06
.06
.06
0
0
0
0
a

0
.21
.39
.54
.67
.79
.93
.98
1
.94
.81
.62
.56
.38
.25
.19
.19
.06
.06
.06
0
0
0
0
S
9

0
.21
.38
.53
.66
.76
.91
.97
.98
1
.87
.67
.60
.40
.27
.20
.20
.07
.07
.07
0
0
0
0
E G M
10

0
.20
.37
.52
.64
.76
.90
.95
.97
.98
1
.77
.69
.46
.31
.23
.23
.08
.08
.08
0
0
0
0
E II T
11

0
.20
.36
.51
.62
.73
.87
.92
.93
.95
'.97
1
.90
.60
.40
.30
.30
.10
.10
.10
0
0
0
0
K U
12

0
.19
.35
.50
.61
.73
.85
.90
.92
.93
.95
.98
1
.67
.44
.33
.33
.11
.11
.11
0
0
0
0
H B E
13

0
.19
.34
.48
.59
.70
.83
.87
.89
.91
.92
.95
.97
1
.67
.50
.50
.17
.17
.17
0
0
0
0
R
14

0
.18
.34
.48
.58
.69
.81
.86
.88
.89
.91
.94
.95
.98
1
.75
.75
.25
.25
.25
0
0
0
0
15

0
.18
.33
.47
.58
.68
.80
.85
.86
.88
.89
.92
.94
.97
.98
1
1
.33
.33
.33
0
0
0
0
16

0
.18
.33
.47
.58
.68
.80
.85
.86
.88
.89
.92
.94
.97
.98
1
1
.33
.33
.33
0
0
0
0
17

0
.18
.33
.46
.57
.67
.79
.84
.85
.86
.88
.91
.93
.96
.97
.99
.99
1
1
1
0
0
0
0
18

0
.18
.33
.46
.57
.67
.79
.84
.85
.86
.88
.91
.93
.96
.97
.99
.99
1
1
1
0
0
0
0
19

0
.18
.33
.46
.57
.67
.79
.84
.85
.86
.88
.91
.93
.96
.97
.99
.99
1
1
1
0
0
0
0
20

0
.18
.32
.46
.56
.66
.78
.82
.84
.85
.87
.90
.91
.94
.96
.97
.97
.98
.98
.98
1
0
0
0
21

0
.18
.32
.46
.56
.66
.78
.82
.84
.85
.87
.90
.91
.94
.96
.97
.97
.98
.98
.98
1
1
0
0
22

0
.18
.32
.46
.56
.66
.78
.82
.84
.85
.87
.90
.91
.94
.96
.97
.97
.98
.98
.98
1
1
1
0
23

0
.18
.32
.46
.56
.66
.78
.82
.84
.85
.87
.90
.91
.94
.96
.97
.97
.98
.98
.98
1
1
1
1

-------
       K..  =  element in  the  distribution matrix found  in row i and
        1J    column  j  (i,  j  = 0,  1,  2,  .  .  .  , n)
       n   =  number  of segments  into which  the estuary is divided.
The pollutant concentration in segment i  due to the single discharge
into segment m, C'is is given as:


                       C'i = Kim Cdm
Then the actual  concentration in segment i,  C.,  is  the  sum  of 'the
contributions from all  the dischargers,  namely:
                    Ci      Kim Cdm
This procedure is repeated for all  segments  i  =  0,  1,2,  .  .  .  n.

The Patuxent River section contains a comprehensive example  of this
calculation for a conservative pollutant.

       Nonconservative  pollutant  distribution must  take into account the
 decay  of the  substance as well as  dilution as it moves away from the
 segment of discharge.   Officer (1976) and Dyer  (1973) develop the
 algorithm for nonconservative substances within the constructs of the
 tidal  prism method  only for  the  case  in which the  exchange  ratios (r.)
 for each segment are equal.   The case for which the r. are  not equal is
 approximated  by  the following expressions:


                        f.  d + a
               c<= c^  , * d Bi

 in  the  downstream direction and
                               104

-------
                       S.   d - a

               Ci  = CH S1    n   Bi
                1     d bd   i = d  1
in the upstream direction
         where C.  = the concentration of nonconservative pollutant
                    in segment "i"
               f.  = fraction of fresh water in segment "i

               f ,  = fraction of fresh water in the segment of
                    discharge

               C ,  = the initial concentration "in the segment of dis-
                    charge,  assuming the pollutant acts conservatively
                                         r.
               B.J  = the decay term (-j  _ ,^_ r \   -k)


               S.  = the salinity in segment "i"

               S,  = the salinity in the segment  of discharge

               a  = an index telling how many segments up or down
                    the estuary from the segment of discharge that
                    segment "i" is located

               r.  = tidal  exchange coefficient (1/T.,  where  T.  is
                    the segment flushing  time)and
               k  = nonconservative constituent  decay  rate (tidal  cycles'  )
For the segment of discharge the expressions reduce to


                         C1 - Cd B.                 (i = d)



     Table 3.3-11 shows calculated pollutant distributions for the
Chester River.  BODr distributions were computed using a high and a

low decay, rate (.05 and .41 per tidal cycle).  Decay coefficients Were
                                 105

-------
                        TABLE 3.3-11.   HIGH FLOW POLLUTANT DISTRIBUTIONS IN THE  CHESTER RIVER  ESTUARY
o
crv
Segment
.0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Distance
of Landward
End from Mouth
71.1
62.7
56.8
51.2
45.6
39.8
32.4
29.6
27.9
26.5
24.6
22.4
20.0
17.4
14.6
12.5
10.6
9.16
7.61
6.16
4.64
3.34
2.11
0.91
BOD5
(Low Decay)
(mqjT1)
4.98
5.40
5.03
4.09
3.41
2.41
1.32
0.70
0.52
0.43
0.24
0.13
0.09
0.04
0.01
•vO
^0
<\-o
-v-0
^0
0
0
0
0
BOD5
(High Decay)
dm,*'1)
3.54
2.80
2.07
1.36
0.81
0.30 •
0.08
0.02
0.03
0.04
0.02
0.01
M)
0
0
0
0
0
o :
0
0
0
0
0
Total
Nitrogen
(mgjT1)
2.68
3.20
3.21
2.97
2.62
2.12
1.52
1.32
1.22
1.13
0.97
0.77
0.70
0.46
0.28
0.22
0.22
0.07
0.07
0.07
%0
0
0
0
Total
Phosphorus
(mgt~l)
0.35
0.43
0.42
0.40
0.34
0.28
0.22
0.17
0.15
0.14
0.14
0.11
0.10
0.06
0.02
0.02
0.02
0
0
0
0
0
0
0
Suspended
Sediment
(mgJT1)
791
949
951
879
784
637
455
393
366
342
298
234
207
139
96
72
72
22
22
22
0
0
0
0
                    AVERAGE
                                                0.13
0.03
0.27
.028
84.2

-------
not adjusted longitudinally for temperature based on the uniformity of
historical temperature observations in the estuary.  At the bottom of
the table, volumetrically weighted average concentrations for the
entire estuary are shown.

     The table shows that the pollutant distributions almost without
exception monotonically decrease towards the mouth of the river. 'This
occurs primarily because the method assumes that the tidal prism of the
most seaward segment is replaced entirely by background waters with no
pollutant after each tidal cycle, which may not really be the case.


     Figure 3.3-7 shows vertically averaged total nitrogen profiles
for the river.  Three of the profiles are observed and were measured
for substantially lower-fr-esh water inflow rates than was used in cal-
culation of the predicted profile.  There are a number of possible •
reasons for the discrepancies between the observed and predicted pro-
files.  The high values at the upper end of the estuary may be due to
the use of a delivery ratio that is too large.  The perturbation of
predicted values by changes in the delivery ratio is approximately linear
for conservative parameters.  Therefore, halving the pollutant load would
result in reduction of instream concentrations by a factor of two, etc.
Second, the methodology assumes that the pollutant is continuously
entering the estuary at a constant rate.  During high flow events,
both the quantity of water and pollutant entering the estuary vary.
Thus, an error is introduced into the calculations by making the constant
inflow assumption, although it is difficult to determine how this error
quantitatively affects the magnitude of the  concentrations in
                               v
each segment.  In reality, when the loadings abruptly stop, dispersion
tends to cause a flattening of the longitudinal gradient over time
as  the high loads in the upper estuary  are flushed towards the mouth.
In  such a case the resultant distribution would  more closely resemble
the observed  profiles.
                               107

-------
O)

E
z
ui
O
O
cc
                                                      • 700312

                                                      * 700402

                                                      * 700604

                                                      • Predicted
8       16       24       32      40       48


                   DISTANCE FROM MOUTH (Km)
                                                                56
64       72
     Figure 3.3-7.   Observed and  predicted  total  nitrogen profiles  in the

                     Chester River Estuary.
                                          108

-------
     The prediction of pollutant levels in the lower estuary is
further complicated by tidal exchanges between Kent Marrows and Eastern
Bay caused by phase differences.  Several  seafood packing plants oper-
ate in this area and contribute both BOD and nutrients to the system.
Description of this exchange within the constructs of the methodology
presented here is not possible.  Therefore, this exchange has not been
accounted for and is a source of error in the predicted profiles.
     Since the total storm load  is assumed  to enter  the estuary  in  a
tidal  cycle, .this i.s~ equivalent to saying  that "in  succeeding  cycles
loads  of  equal magnitude  enter the estuary.  Therefore, the  concentra-
tions  given  in Table 3.3-11 are  necessarily upper limit values.   Lower
limit  concentrations may  be found by assuming that the storm load is
equally distributed over  each tidal cycle during  the length  of time
base of the  inflow  hydrograph.   To estimate initial  concentrations
for this  case the pollutant load must first be divided by  the time  base
(in hours) of the inflow  hydrograph and multiplied by the  number of
hours  in  a tidal cycle.   The resultant load is then  divided  by the
hypothetical flow rate  (per tidal cycle)  through  the receiving segment
as is  normally done to  arrive at the initial concentration.   Alter-
natively, dispersion equations may be used  if the user wishes to treat
the event as a single event discharge.
3.3.7   Eutrophication

     Using the volumetrically averaged  predicted  nitrogen and  phosphorus
concentrations from Table 3.3-11^-!:? ratios are computed for the  river.
The  ratio of  total nitrogen  (as  N)  to total phosphorus  (as  P)  is
0.27:0.028 or 9.64:1.   This  falls  into  the  region where either nitrogen
or phosphorus may  be limiting. '  To  compare  the predicted N:P ratio
with observed data, the ratio of total  soluble inorganic nitrogen
(TSIN  = NH3 + N02  + N03 (as  N)  to:total orthophosphorus  (P04  as  P)

                                 109

-------
was used.  (Total phosphorus data were not available.)  Observed ratios
for the three available dates of March 12, April 2, and June 4, 1970
were computed as 12.6, 11.7, and 4.52 respectively.  These values
in general support the calculated conclusion of either nitrogen or
phosphorus limitation. although it appears from the 1970 data the phos-
phosphorus was limiting early in the calendar year while nitrogen was
limiting during the summer months.   Not only can the limiting nutrient
vary seasonally,  but it may well  change from nitrogen  to  phosphorus in
the less saline portions of the estuary (Porcella  and  Bishop, 1974).

     Correlation coefficients computed for chlorophyll-^ observations .
on selected observed water quality parameters are shown in Table 3.3-12.
They show that chlorophyll -a concentrations are strongly correlated
with surface nitrite and nitrate nitrogen and total nitrogen early in
the year.  This dependence diminishes as time progresses.  However
dependence on phosphorus concentration remains strong throughout the
spring and summer.  There is also a tendency for algal growth to occur in
the .less  saline  waters as indicated by the correlation of chlorophyll-a
on salinity.  Chlorophyll -a_ is also correlated negatively with Secchi
depth which indicates possible light limitation in the estuary.

     To investigate this latter phenomenon a two-parameter light pene=
tration model was fit to Secchi depth vs. chlorophyll-a_ found in
Table 3.3-4.  The model formulation is
D  =
                            -In (0.1)
                             a-+ 6' .('Chi -a.)

 where
       D        = Secchi  depth,  meters

       Chl-a_    = chlorophyll -a  concentration in yg £~
       In (0.1)  = a constant relating Secchi  depth at disappearance to
                  light penetration in seawater
                                110

-------
  TABLE 3.3-12.  CORRELATIONS FOR CHLOROPHYLL-a ON SELECTED WATER
              QUALITY PARAMETERS IN THE CHESTER RIVER
Parameter
Surface
NO 2 + N03
Surface
Total P04
Surface
Total N
Secchi
Depth
Surface
Sal inity
- D A T E a}
700312 700402 700604
+.90 +.50 -.15
'+.76 +.36 +.62
+.65 ' +.23 -.38
-.64 -.47 -.44
-.87 -.38 -.25
a>) year/month/day
                                 111

-------
      a        = background extinction coefficient
      3        = coefficient of incremental extinction due to the
                 algal concentration

     The regression produced values of 3.6 fora and 0.069 for 6.
The incremental extinction coefficient falls outside the typical  range
for algae (.015-.022) (Megard, 1979) and the background extinction is
also quite high.  At typical Chi-a concentrations in the estuary the
magnitude of a (3.6) is three to five times greater than that of
B'Chl-a_.   This indicates possible light limitations due to some con-
stituent other than algae, perhaps suspended s-ediment or detritus.  Mo
suspended sediment data were available concurrent with Chi-a_measurements
to validate this hypothesis.

     The depth of the photic zone has been defined as-the depth to which
1% of surface illumination penetrates (Lorenzen, 1972).  Using average
Chi-a_ concentrations in the light model and substituting In (0.01) for
In (0.1)  gives an average photic zone depth for the estuary of approxi-
mately one meter.  This value is appreciably less than the mixing depth
(assuming that the estuary is fully mixed) indicating that algal  production
is light limited.
 3.4   DEMONSTRATION  EXAMPLE:  THE  PATUXENT RIVER

      The  Patuxent River  system affords a good opportunity to use both
 the  river and  stream  methodology  and  the estuarine methodology  in a
 single  system.   A complicating feature of this system  is a section of
 river which  is  tidally influenced but has no salinity  gradient  (i.e.,
 it consists  entirely  of  fresh water).  This introduces a situation which
 is not  covered  by the river  methodology and is only briefly addressed
 in the  estuary  methodology through application of advection-dispersion
 equations.   A  major tributary to  the  system, Western Branch, enters  the
 Patuxent  in  this section.
                                   112

-------
     Parameters analyzed in the low flow assessments of the previous
systems (temperature, dissolved oxgyen, BOD, and fecal coliforms) are
also investigated for this system.  High flow assessments for the
Patuxent include total phosphorus and total nitrogen.  High flow
analyses were not performed in the riverine portion of the system.
3.4.1  Data Collection

     Quadrangle maps (7% minute series) were obtained from the U.S.
Geological Survey as were flow data and stage-discharge curves at
various locations.  Flow data for approximately 10 years were used
in the analysis.  The most recent available stage-discharge curves
were utilized to determine flow depths.

     .The user is cautio-ned here that use of 10 years of-low flow data
to estimate the 7Q1Q will almost always give a biased estimate of this
statistic.  The 7Qir), like any other statistic, has some distribution
around its mean value.  Necessary sample sizes for estimating the mean
can be determined by the Law of Large Numbers (Haan, 1977).

     Some water quality data were provided by the Maryland Department
of Natural Resources (DNR).  Among the quality data available from the
Maryland DNR were:

     •  Salinity profiles at several flow regimes and times of year
     •  Dissolved oxygen and BOD
     t  Total chlorophyll-a and phaeophyti.n-a_
     •  Phosphorus (total whole, total filtered, orthophosphorus)
     •  Ammonia nitrogen, nitrate nitrogen, total Kjeldahl nitrogen
        (whole and filtered), and
     0  Fecal coliforms.
                                    113

-------
Copious data were also available from the STORE! system at locations
in the estuary and in the tidal fresh water and the riverine portions
of the system.

     Effluent data for municipal point sources were provided both by
the Maryland DNR and individuals in the various municipalities within
the basin.  Industrial sources were judged to be insignificant con-
tributors for the parameters analyzed.  The Maryland DNR has compiled
a list of all point source discharges in the basin.  They are shown in
Table 3.4-1.
3.4.2  Data Reduction and Supplementation

     Gaging stations in the nontidal fresh water portion of the
Patuxent used for the -stream analysis were located at Laurel,  Savage,
and Bowie, Maryland.  Sufficient data were not available at either
the Savage or Bowie stations to estimate the 7Q,Q low flows.  For the
Savage location an upstream station at Guilford was used and the 7Q10
flow estimate there was extrapolated to Savage by area!  weighting.  The
7Q,0 at Bowie was also estimated by areal weighting.

     Gaging stations as well as the major point sources  in the basin
were located on the 7^ minute series maps.  Distances and elevation
changes between these gages and point sources were measured and
average slopes were computed and recorded.  Depths and velocities were
determined at each gage for the 7Q,Q flows using stage-discharge curves
and the continuity equation.

     For the estuarine portion of the system, transects  were drawn at
convenient locations normal to the flow and cross sections were measured.
Lengths between the transects were recorded and the MHT and MLT volumes
of the estuary were  calculated.  Transects were  taken at  the  locations
                        r
shown in Table 3.4-2.  Mean high tide (MHT) cross sections were calculated
                               114

-------
      TABLE 3.4-1.   ACTIVE  DISCHARGERS  IN THE  PATUXENT RIVER  BASIN
    Publicly Owned
   Treatment Plant
           Private
                                        Industrial Wastes
Lower Patuxent
 Academy Natural  Sciences
   of Philadelphia
 Evergreen  Park,  STP
 Natural Resources  Institute
 Village Center,  The
Asher, B.F.  Sand & Gravel
Oenton, Warren & Co.,  Inc.
Lore, J.C. & Sons
Patuxent River Oyster  Co.
Pepco, Chalk Point
Trossback Brothers
Middle Patuxent

Marl ton Temporary STP
 Boone's  Mobile  Estates
 Croom Vocational H.S.
 Edgemeade  of  Maryland
 First Md.  Utilities, STP
 Lyons Creek Mobile Home Pk.
 Md.  Manor  Mobile Homes
 Northern School, STP
•Patuxent Mobile Home Estates
.Patuxent River  4-H Center
 Southern Senior H.S.
 Tucker's Restaurant
 Wayson's Mobile Court
Annapolis Sand & Gravel  Co.
Calvert Meats, Inc.
Davidsonville Sand  & Gravel
First Maryland Utilities
Pepco Flyash - Brandywine
Western Branch

Western Branch STP
 Andrews  Field Motel
 Pointer  Ridge Lagoon
Upper Patuxent

Bowie State College
Bowie City of, STP
Horsepen Branch,  STP
Maryland City, STP
Parkway, STP
 Bowie Race  Course
 City of Capitals, STP
Bio-county Aggregate  Corp.
Bowie Water Filtration  Plant
Electro-Therm,  Inc.
Little Patuxent

Maryland House of
  Corrections
Patuxent, STP
Savage, STP
 John  Hopkins University
 Parkway  Manor Motel
Arctec, Inc.
Barton, Alan  E.,  Inc.
Bercon Parkaging,  Inc.
Columbia Park
Contee Sand & Gravel
Crofton Meadows Water
  Treatment Plant
Crofton Water Treatment
  Plant
                                         115

-------
     TABLE 3.4-2.  ESTUARINE CROSS SECTIONS IN THE PATUXENT RIVER
    Location
Distance From    MLT Cross      MHT Cross
 Mouth  (km)     Section (m?)   Section
Peterson's Point

Broome's Island

Prison Point

Trent Hall Point

Rt. 231 Bridge

Potts Point

Mi 11 town Landing

Hall Creek
    14.5

    18.9

    26.6

    32.3

    36.7

    42.4

    51.89

    59.0
15,205

 8,152

12,028

 5,643

 4,060

 2,843

 1,417

 1,282
16,233

 8,836

13,299

 6,305

 4,543

 3,017

 1,591

 1,816
                                   116

-------
assuming that the banks of the estuary are vertical and using the
mean tidal range given on the USGS topographic maps.
3.4.3  Fresh Non-Tidal Waters

     The free flowing fresh waters of the Patuxent extend to approxi-
mately Hardesty, Maryland, 89 km above the mouth.  Above this point,
along the Patuxent to the Rocky Gorge Reservoir Dam (41 km) and along
the Little Patuxent to Savage (70 km), river methodologies were
applied.  The Patuxent was divided into seven reaches and the Little
Patuxent was divided into seven reaches for the low flow analysis.  •
Reach segmentation was based on the locations of important sewage
treatment plant effluent discharges.  These reaches are schematically
shown in Figure 3.4-1.  The accompanying hydraulic data for each reach
are.shown in Table 3.4-3.  The characteristic depth shown is the
hydraulic depth of the stream.  A range of velocities for each reach
is also shown in the table.  The lower of the two velocities is that
derived by continuity (Q = AV), and the upper value was derived from
the Manning equation using a roughness coefficient (n) of 0.08.  This
coefficient was evaluated using the hydraulic data of Tsivoglou and
Wallace (1972).  It falls into the category of "sluggish river reaches,
rather weedy with very deep pools" (Schwab e_t jfL, 1966).  Tsivoglou
and Wallace (1972) describe the stream as "a typical coastal plains
stream characterized by alternating small pools and riffles."

     The Patuxent and Little Patuxent Rivers in this portion of the
basin flow through a region adjoined by marshy areas.  The 7Q,Q out-
flow at the Bowie gage is greater than the sum of the inflows at the
Savage and Laurel gage and the sewage treatment plant inflows.  The
difference in these is assumed to come from the swampy areas adjacent
to the river.  This flow was distributed incrementally by proportion-
ing it to the length of the reach.  The last column of Table 3.4-3
gives these incremental natural flows.  For the most part the numbers in
this table are estimates only, interpolated from hydraulic data at the
gaging stations.
                               117

-------
        HORSPEN STP
                                        SAVAGE  GAUGE

                                  71 |  SAVAGE  STP
                                  6L,
                                       HAMMOND  BRANCH (MD-VA
                                         MILK PROD. Assoc.)
                                    i
                                 51 ,
                                       DORSEY RUN (MD HOUSE OF
                                         CORRECTION)

                                      FT.  MEADE #2 STP
 FT,  MEADE #1 STP

2L

PATUXENT STP
                                   1L
                                 BOWIE  GAUGE
Figure 3.4-1.   Reach segmentation schematic  for the
                Patuxent  River.
                            118

-------
                 TABLE 3.4-3.   PATUXENT  RIVER  HYDRAULIC DATA FOR  FREE  FLOWING  WATERS (LOW  FLOW)
Patuxent
Reach f
1
2
3
4
5
8
7
Little
Patuxent
1L
2L
3L
4L
5L
6L
7L
Character 1 stic
Descriptor at Upstream Depth
End of Reach (in)
Bowie Gauge
Confl w/Little Patuxent
Bowie STP
Horsepen STP
Parkway STP
Maryland City STP
Laurel Gauge

Patuxent STP
Fort Meade 
-------
3.4.3.1  Temperature Profiles

     Temperature profiles for the system were developed using the
equilibrium temperature (T ) approach (Section 4.4.4, p. 205 of the
screening manual).  Data used to calculate T  for the Patuxent are as
follows:
                  u = 14.5 km hr"1 (9 mi hr"1)
        H$n = 14500 BTU m"2 day"1 (1350 BTU ft"2 day"1

                       T  = 21.1°C (70°F)
                     Relative Humidity = 72%
                   Cloud cover fraction = 0.5
These values represent average conditions for the month of September.
Observed annual low flows typically .occurred in this month.  The
computed T  for the Patuxent basin is 20.2°C.

     Temperature profiles were calculated using equation IV-36,
section 4.4.5, p. 217 of the screening manual.   Two cases were used to
evaluate temperature profiles.  The difference in the two cases involved
assumptions concerning the natural incremental  inflow temperature.
This inflow can be assumed to be at the temperature of the ground water-
(when the inflow is primarily from subsurface sources) or at the
equilibrium temperature (when the inflow is primarily from standing
marsh water).  Incoming flow at Laurel and Savage from the headwaters
was assumed to be at the equilibrium temperature.  Calculations show
that the river essentially does not deviate from the equilibrium
temperature except in reach #5 when the incremental inflow temperature
is assumed to be that of subsurface water (10°C).
                               120

-------
3.4.3.2  Estimation of Reaeratlon and Deoxygenation Coefficients

     Deoxygenation rates for each reach were calculated using the
Bosko equation,

                        k  = k  + n(V/D)
with a mean k, of 0.45 (see Zison e_t ^1_., 1978).  No differentiation
was made between NBOD and CBOD deoxygenation.

     Pheiffer e_t al_., (1976) studied BOD decay rates in the Patuxent
River below the Parkway Sewage Treatment Plant (reach #5).  CBOD decay
rates were determined to be 0.61 day   in 1973 and 0.30 day"  in 1975
(all coefficients are base e).  The difference between these two rates
was attributed to more efficient CBOD removal at the Parkway plant
caused by upgrading in the interim.  For NBOD, rates changed from
0.76 day"1 in 1973 to 0.48 day"1 in 1975.  The mean deoxygenation
rate calculated from Bosko is 0.56 day"  which is greater than both
the current NBOD and CBOD decay rates for this reach, but less than
the rates determined before the upgrading of the treatment facility.
     Reaeration rates were determined by each of three methods, Tsivoglou-
Wallace, O'Connor, and Owens.  Table 3.4-4 shows both the deoxygenation
rates and reaeration rates computed by these three methods.  The table
values are for temperatures of 20°C with the exception of the Tsivoglou-
Wallace rates which are at 25°C.  Tsivoglou and Wallace (1972) used the
gas-tracer method to evaluate reaeration in the Patuxent.  Flows at that
time were similar to the 7Q,n flows used in this investigation.  Their
                                                     -1
value for the reaeration rate in reach #5 was 3.3 day   (base e) at
25°C.  The rate predicted by their equation using interpolated hydraulic
data is 2.31 day"  at 25°C.  The O'Connor and Owens formulations for
this reach give values an order of magnitude higher (20.9 day"  and
37.2 day   respectively).
                                  121

-------
TABLE 3.4-4.  DEOXYGENATION AND REAERATION RATES FOR THE PATUXENT
                    RIVER FREE FLOWING WATERS
Mean Deoxygenation Rate Reaeration Rates
Reach # (day" base e (day" base e)
@20°C)
Tsivoglou-Wallace O'Connor Owens
(@25°C) ((320°C) (020°C)
Patuxent
1
2
3
4
5
6
7
Little Patuxent
1L
2L
3L
4L
5L
6L
7L

0.47
0.52
0.53
0.52
0.56
0.57
0.68

0.51
0.51
0.56
0.50
0.48
0.49
0.67

0.38
1.61
1.71
1.03
2.31
2.45
4.65

1.93
3.09
4.69
1.98
1.11
1.56
19.11

2.37
10.5
12.4
12.2
20.9
22.8
25.4

6.25
5.95
5.56
4.73
4.16
4.00
5.79

2.54
15.9
19.5
19.0
37.2
41.4
47.7

8.49
8.07
7.47
6.04
5.12
4.88
8.02
                              122

-------
3.4.3.3  BOD Mass Balance

     Equation IV-7, section 4.2.4, p. 152, of the screening manual
was used to perform all BOD mass balance calculations for the Patuxent
River.  Three cases were used to demonstrate BOD routing.  In the
first, BODg profiles were developed to determine closeness to observed
BODr data.  The second case used BODr plant effluent data converted to
CBOD  together with estimated NBOD  values for all plants based on
the type of treatment that they employ (U.S. Army Corps of Engineers,
1977).  Lastly, NBOD  and CBODu were routed using NBODU values as
determined using the Kjeldahl nitrogen values from plant effluent
data.  Municipal treatment plant effluent data are shown in Table
3.4-5.  The headwaters above Savage and Laurel and the incremental
inflow waters were assumed to have background BODr values of 1.0
    -1
mg £  .

     The results of these calculations are shown in Table 3.4-6.  The
range given shows the upstream to downstream variation within each
reach.  The cases in which NBOD are determined, one with the total
Kjeldahl nitrogen data for the treatment plants and one assuming a
treatment type for the treatment facility (see Appendix A), are shown
for comparative purposes.  Without actual data to rely on, the user
would likely estimate BOD loads with a percent reduction based on
the type of treatment that the facility employs.  Using this approach,
the predicted instream ultimate BOD concentrations in this example
are often two to three times higher than the profiles predicted using
observed loading data.  No observed ultimate BOD data were available
for comparison with either of these profiles.

     Using routed treatment plant BOD,- data, the predicted BODr values
in reach #5 range from 1.80 to 1.19.  The observed BODr data (retrieved
from STORET) for September in that reach have a mean of 4.9 ± 2.8 which
puts the calculated values below the low end of the one standard devi-
ation range.  One plausible explanation for this discrepancy is that

                              123

-------
                          TABLE  3.4-5.  MAJOR SEWAGE TREATMENT PLANT EFFLUENT DATA  IN THE  PATUXENT
                                                       RIVER  SYSTEMS
ro
Plant
Western Branch
Savage
Bowie
Patuxent
Parkway
MD-VA Mil k
Producers Assn.
Maryland House
of Correction
Maryland City
Horsepen
Ft. Meade #1
Ft. Meade #2
Flow
(m3 sec"1)
.40
.37
.07
.15
.20
.01
.03
.02
.02
.05
.05
Temperature
19.1
18.5
23.8
20.5
24.4
20.7
23.8
22.2
18.7
21.3
20.9
B005
(mg t~ )
3.6
13.6
23.1
41.1
2.2
23.3
5.4
16.3
5.7
22.4
28.0
Dissolved
Oxygen
(mg i.'1)
9.2 '
8.9
10.0
8.6
8.0
4.9
7.9
5.5
9.8 .
6.9
7.9
NH3
(mg 4"1)
5.2
9.0
18.2 .
• 10.8
4.6
817.6
0.60
16.6
7.5
9.0
8.2
Total
Kjeldahl
Nitrogen
(mg i' )
8.8
14.7
25.0
16.7
7.7
23.8
2.3
22.3
9.8
12.2
12.9
Total
Phosphorus
(mg i'1)
5.7
8.7
9.4
6.7
3.8
28.3
3.2
9.4
4.3
6.8
6.8
Fecal /Total
Coliforms
(MPN/100 mil)
12147/-
4/84
2/26
1847/126552
7/242
39/2732
4/9
723/13200
2/1204
177/1005
6/116
           a)  All values are averages of composite samples except dissolved oxygen (grab sample).

-------
TABLE 3.4-6.  BOD MASS BALANCE FOR THE FREE FLOWING WATERS
                  OF THE PATUXENT RIVER
Reach # CBODg
1 4.44 -
2 5.11 -
3 3.50 -
4 1.30 -
5 1.80 -
6 2.10 -
7 1.0 -
1L 7.80 -
2L 4.50 -
3L 5.30 -
4L 5.00 -
5L 6.60 -
6L 7.20 -
7L 1.0 -
3.03
4.44
2.79
1.12
1.19
1.57
0.68
6.41
2.81
3.14
4.15
4.99
6.43
0.87
(CBOD + NBOD)u
(NBOD Determined
by Estimation)
31.1 -
35.8 -
47.7 -
42.1 -
64.9 -
13.5 -
1.47-,
42.5 -
35.6 -
48.3 -
50.0 -
63.0 -
69.1 -
1.47-
21.1
31.1
37.7
34.9
39.9
9.8
. i.o
34.8
22.0
28.4
41.3
47.3
62.0
1.29
(CBOD + NBOD)u
(NBOD Determined
from Plant TKN)
19.9 -
22.9 -
26.6 -
14.1 -
20.9 -
11.9 -
1.47-
29.3 -
21.3 -
28.4 -
29.3 -
39.1 -
43.4 -
1.47-
13.5
19.9
21.0
11.8
12.9
8.7
1.0
24.0
13.2
16.7
24.3
29.4
39.0
1.29
                             125

-------
while the BOD,- test of plant effluent represents primarily carbonaceous
demand, samples taken from streams may contain the effects of some
nitrogenous deoxygenation.  This should be small, however, since the
travel time through the riverine system is short.  It is also probable
that the small incremental inflows from the marshy areas have concen-
trations greater than the 1.0 mg £~1 assumed in this exercise.  The
user is warned, however, against drawing such conclusions based on
only a point observation in the system.

     Figure 3.4-2 shows observed and predicted BODr concentrations in
the Patuxent River from below Rocky Gorge Dam to just below the con-
fluence of the Patuxent and Little Patuxent rivers (approximately
km 130 to km 100).  The observed data were measured on September 26,
1978 by the Maryland Department of Natural Resources.  The flow in
                   3    -1
reach #3 was 0.71 m  sec  .  The comparison between predicted and
observed results is quite good.  Hov/ever, the background BOD,-
                      -1
assumption of 1.0 mg £   seems too low as evidenced by the discrepancy
in the upper reaches where this assumption has greatest effect.

     The background condition is damped out after a substantial source
of BOD enters the system.  The large spike at river km 104 is due to
the inflow of the Little Patuxent River whose quality is more degraded
than that of the Patuxent.
3.4.3.4  Dissolved Oxygen Profiles

     A dissolved oxygen mass balance for the Patuxent River was performed
using both Tsivoglou-Wallace and O'Connor reaeration rates.  The O'Connor
formulation was developed on rivers having depths of 0.3 to 9 meters
and consequently produces reaeration rates lower than the Owens equation
which was developed on faster, shallower streams.  Because the O'Connor
method yielded dissolved oxygen concentrations greater than those
observed instream, and close to saturation, profiles using the Owens
                             126

-------
                                  REACH NUMBER
 0>
Q
O
CD
      0 _
      135
                                                 4   3
                                                            O—O Observed BODc
                                                                 780926
                                                                 Predicted
                                                                        BODc
130     125     120      115      110     105     100      95      90

                DISTANCE FROM MOUTH

                        (Km)
          Figure  3.4-2.  Observed and  predicted  BODc in  the
                          Patuxent River.
                                      127

-------
method for reaeration were not calculated.  Ultimate BOD from the case
in which NBOD  was estimated from plant total Kjeldahl nitrogen data
was used to perform all dissolved oxygen calculations.  Temperature
profiles showed that 20°C was a good approximation of temperature
throughout the system.  The corresponding dissolved oxygen saturation
is 9.2 mg if .   NBOD and CBOD decay rates were not differentiated so
equation IV-18 (Streeter-Phelps), section 4.3.6, p. 170 of the screen-
ing manual was used.
     One complicating factor in this calculation is the presence of
the Rocky Gorge Reservoir just upstream from the Laurel gage.  If
stratification occurs in the lake and hypolimnion water is released
into the Patuxent, very low initial values of dissolved oxygen might
be observed at Laurel.  On the other hand, turbulent flow releases
might drive waters with initially low dissolved oxygen to near satura-
tion.  (See section 4.3.4 of the screening manual, Effect of Dams on
Reaeration.)  Dissolved oxygen profiles were computed using both
assumptions.  For the first case, the initial dissolved oxygen level
at the Laurel gage was assumed to be at saturation.  For the latter,
the boundary dissolved oxygen value was selected to be 5.0 mg £~ .
Table 3.4-7 shows computed dissolved oxygen profiles for the Patuxent
and Little Patuxent Rivers using the assumption that waters released
from Rocky Gorge Reservoir .were at saturation.  Again, the ranges
represent the longitudinal variation within each reach.  Average
observed dissolved oxygen levels at Duvall Bridge (in reach #5) for
the month of September v/ere 6.5 ± 0.52 mg £~ .  From this point obser-
vation, it appears that perhaps the Tsivoglou-Wallace rates are better
indicators of actual reaeration rates.  However, Figure 3.4-3 shows
that this may not be the case.  This figure illustrates an observed
dissolved oxygen profile taken on 26 September, 1978 at a flow rate not
greatly different from that used in the analysis.  Predicted profiles
were computed using the assumption that release waters from Rocky
Gorge were at 5 mg 5,   dissolved oxygen.  The observed data validate
this assumption.  Although they are not plotted, predicted dissolved
                               128

-------
TABLE 3.4-7.  DISSOLVED OXYGEN PROFILES IN THE PATUXENT RIVER
                  FOR TWO REAERATION RATES
Reach #
1
2
3
4
5
6
7
1L
2L
3L
4L
5L
6L
7L
Dissolved
O'Connor
8.3 -
8.2 -
8.7 -
8.9 -
9.0 -
9.2 -
9.2 -
7.6 -
6.7 -
5.5 -
5.5 -
7.0 -
9.2 -
9.2 -
6.4
8.3
8.2
8.7
8.9
9.0
9.2
7.1
7.6
6.7
6.0
5.5
7.0
9.2
Oxygen (mg l~ )
Tsivoglou-Wallace
2.7 -
3.4 -
5.0'-
5.9 -
7.5 -
9.1 -
9.2 -
5.8 -
5.4 -
1.8 -
2.0 -
6.6 -
9.2 -
9.2 -
0.0
2.7
3.4
5.0
5.9
7.5 ""
9.1
4.0
5.8
5.4
1.8
2.0
6.6
9.2
                            129

-------
                                       REACH NUMBER
   12
LLJ
>
d
    10
£   8


01
CJ
X
                                                   I
       •—• Observed DO
            26 Sept. 1978
            Flow = 0.71 rrWsec.
            ©Bowie

       X—X Predicted DO
            (O'Connor Rates)

       v—* Predicted DO    ~
          •  (Tsivoglou-Wallace
            Rates)
            Flow = 0.56 nV/sec
            @ Bowie
      135     130       125       120       115      110

                                     DISTANCE FROM MOUTH
                                             (Km)
105
100
95
                                                                                       90
      Figure 3.4-3.  Observed  versus predicted  dissolved oxygen profiles for
                       the Patuxent  River.
                                          130

-------
oxygen values for reaches #1 and #2 are essentially the same as those
values shown in Table 3.4-7 for reaches #1 and #2.  The plot also shows
that the Tsivoglou-Wallace formulation yields reaeration rates that
appear to be too low because the predicted dissolved oxygen profile
is considerably below the observed profile.  However, this is indirect
evidence since there may be other causes for the discrepancy, such as
inaccurate BOD data.

     For the purposes of selecting between alternatives, the use of
lower reaeration may be more appropriate.  For instance, if the
planner has to choose between reach A or B as a possible site for a
sewage treatment facility, the use of very high predicted reaeration
rates which maintain dissolved oxygen at or near saturation may make
A and B appear equally attractive.  The use of lower reaeration rates
may show that one reach is less desirable than the other.

     The use of 9.2 or 5.0 mg £,"  for background conditions at the
upstream end of reach #7 made essentially no difference when using
O'Connor rates since the predicted dissolved oxygen was driven to
saturation in both cases.  Using Tsivoglou-Wallace rates, the differ-
ence in predicted dissolved oxygen levels was about 1 mg i~  at the
end of reach #7, about 0.3 mg £~  at the downstream end of reach #6
and negligible thereafter.  Selection of the boundary dissolved oxygen
value will tend to be more important when boundary flows are large
compared to incoming flows from waste sources in the system.

     Critical travel times and critical dissolved oxygen deficits were
also computed for several of the waste treatment plants using the
7Q1Q low flow for the Patuxent.  Equations IV-22 and IV-24 (section
4.3.7, pp. 175-176) in the screening manual were used in this evalu-
ation.  Alternatively, Tables IV-10 and IV-11 in the same section can
be used for this purpose.  Table 3.4-8 shows critical travel times
and critical deficits calculated by the equations.  If T  (time of
                                                       \f
travel  to critical  deficit)  is  undefined or zero,  then  the  answer
                              131

-------
TABLE 3.4-8.  CRITICAL TRAVEL TIMES, DISTANCES AND DISSOLVED OXYGEN
       DEFICITS FOR SOME PATUXENT STPs AT THE 7Q1Q LOW FLOW


                        Travel      Distance to   Critical
                         Time        Deficit      Deficit
      STP Name/         (days)         (km)        (mg  A-l)
     Reach  Number      0     T-W    0     T-W     0   T-Wa^


    Savage/6L        0,59    1.15   6.9    13.4   4.00  8.6

    Ft. Meade #2/    0.00i;  0.00   0.0    0.0   3.70  7.4
    3L

    Ft. Meade #!/    0.00    0.00   0.0    0.84  3.60  3.8
    21

    Patuxent/lL      0.00    0.73   0.0    9.0   3.40  5.9

   • Maryland City/   0.15    0.81   1.50   8.4   0.27  1.9
    6
Parkway/5
Horsepen/4
Bowie/3
0.
0.
0.
00
00
00
0.70
0
0
.93
.71
P.
0.
0.
0
0
0
7
7
6
.2
.9
.8
1.
3.
4.
70
30
20
3.8
4.9
6.3
      0 - O'Connor reaeration rates
      T-W - Tsivoglou-Wallace reaeration rates.

      A zero entry indicates that the deficit occurs at the
      STP outfall.
                              132

-------
obtained from Equation IV-24 for D  (the critical deficit) is not
valid.  In such cases 0  occurs at the location of the waste water
discharge and can be determined by calculating a flow weighted average
dissolved oxygen deficit using the upstream deficit and the deficit in
the incoming waste water.

     In general, Tables IV-10 and IV-11 are simpler to use than
Equations IV-22 and IV-24.  As an example, for the stretch of the
river below the Parkway Sewage Treatment Plant, the user first cal-
culates D /L  (L  is the instream BOD concentration just below the
sewage treatment plant, and D  is the initial  deficit) and k /k,  (the
                             o              .                a  L
reaeration constant over the deoxygenation constant for the reach).
The results are:
                      .o._ 1.7 mq &
                     1	          -1
                      o   20.9 mg. £
and
                                     = 3.696 -3.7
                      L   .056 day
L  is taken from the calculated BOD concentrations in Table 3.4-6 for
reach #5.  D  is calculated from the dissolved oxygen concentration in
Table 3.4-7 for reach #5, using Tsivoglou-Wallace reaeration rates and
a saturation concentration of 9.2 mg i~ .   The rate constants k  and
                                                               d
k.  are for reach #5 also.  The reaeration constant k  is computed from
 L                                                  a
the Tsivoglou-Wallace formulation adjusted to 20°C to be consistent with
the temperature base of the deoxygenation rate:

                  k  = 2.31 (1.022)(~5) = 2.07
Using these values as the abscissa and ordinate in Table IV-10 of the
screening manual, the value for D /L  is found to be 0.18.  When this
                               133

-------
is multiplied by L  (20.9), D  is found to be 3.8 mg H~l.   This is
identical to the value found, in Table 3.4-8 obtained using Equation
IV-24 of the screenina manual.
3.4.3.5  Total Coliform Routing

     Total coliform bacteria levels in the free flowing portion of the
Patuxent River system were computed using average plant effluent con-
centrations from Table 3.4-5.  As was previously discussed in the
Sandusky River example, coliform bacteria loadings can be quite
variable depending on the reliability of the chlorination process
at each sewage treatment plant.  The predicted concentrations could
vary several orders of magnitude under an identical set of flow con-
ditions.  Calculations showed that the likely problem areas with
respect to bacterial concentrations at low flow would be downstream
from the Patuxent sewage treatment facility where the calculated con-
centrations reached 16,477 MPN/100 m.1.  Initial concentrations of
1317 and 562 MPN/100 ml were predicted instream at the outfalls of
the Maryland City and Parkway sewage treatment plant facilities.  The
high predicted value at the Patuxent sewage treatment plant kept the
concentrations of total coliforms in the 7600 to 3600 MPN/100 ml
range from the confluence of the Patuxent and Little Patuxent to the
end of the free flowing portion of the river at Hardesty.

     A background value of 300 MPN/100 ml (McElroy ej: al_., 1976) was
used in the mass balance equation (case 2) for total coliforms.  (See
section 4.6.2, p. 237, of the screening manual.)

     The State of Maryland standards for fecal coliforms are 200 MPN/100 ml
for Class I waters and 70 MPN/100 ml for Class II waters.  The Patuxent
River Class I waters extend from source to 63 kilometers above the
mouth of the estuary.  Below this point the waters are categorized
as Class II.  Sampling on 26 September 1978 indicated violations of
Class I standards at ten locations with concentrations as high as
                                134

-------
4000 MPN/100 ml.  No violations occurred in Class II waters on
this date.
3.4.4  Estuarine Waters

     The Patuxent River estuary is an excellent choice for demonstra-
tion of the estuarine methods because of its geometry and drainage
characteristics.  It has no major side embayments, and the assumption
of one fresh water inflow at the head of the estuary is very nearly
met.  Flow ratio calculations using a tidal prism volume of 3.51 x 10
 3                                                           43
m  and estimated flows at the head of the estuary of 2.3 x 10  m  and
3.58 x 10  m  yield flow ratios of 0.004 and 0.103 for the low and
high flows cases investigated here.  These values indicate that the
estuary is well mixed for both flow rates.  However, historical data
indicate that the Patuxent River estuary is partially stratified at
high flows.  Unfortunately, sufficient velocity data were not available
to check the classification of the estuary using the stratification-
circulation method.

     The estuarine waters of the Patuxent extend from Hall Creek to
Sheridan Point, a distance of approximately 46.6 km.  Above these
waters is a 37.5 km segment of fresh tidally influenced waters and
below are embayment waters which are essentially of the same salinity
and quality as the Chesapeake Bay.
3.4.4.1  Flushing Times

     Flushing times for the Patuxent River estuary were calculated by
each of the tidal prism, modified tidal prism, and fraction of fresh
water methods.  Both high and low river flow volumes per tidal cycle
at the Bowie gage were extrapolated to the location identified as the
beginning of the estuary (Hall Creek) by area! proportioning.  These
                                      5  3
volumes were estimated to be 1.42 x 10  m  for the low flow conditions
                                135

-------
(7Q,Q) and 3.6 x 10  m  for high flow conditions.  The volume of
tidally influenced fresh water immediately upstream of the first
modified tidal prism segment was assumed not to influence the flushing
characteristics of segments down the estuary.

     The results of the flushing calculations are as follows:
     •  The tidal prism method gives a flushing time of 6.1
        days for both scenarios.
     •  The modified tidal prism method gives flushing times
        of 119 days for low flow and 36 days for the high flow
        case.
     •  The fraction of fresh water method gives flushing times
        of 203 days and 14.3 days for the low and high flow
        scenarios, respectively.
As an example, the calculations for flushing times using the modified
tidal prism method to estimate flushing times in the Patuxent River
under high flow conditions are summarized in Table 3.4-9.  The "0"
segment begins when Hall Creek enters the Patuxent River.

     The analysis shows the modified tidal prism method when applied
to the Patuxent River is more sensitive to flow rate changes than it
is for the Chester River, even though the flow rates are comparable
for the two systems.  The drainage basins of the two estuaries are
shaped quite differently, and the Patuxent more nearly meets the
"single inflow at the estuary head" assumption than does the Chester.
The Patuxent drainage basin is also more than twice as large but the
volume of the Chester estuary is greater.  It can be concluded that
flow rate changes of approximately equal magnitude have less impact
on the larger estuary's flushing times.

     The fraction of fresh water method produced flushing time estimates
that closely approximated those attained by the modified tidal prism
method.  This was not the case for the Chester River.  Some allowance
                                136

-------
TABLE 3.4-9.  CALCULATION OF FLUSHING TIMES FOR HIGH FLOWS CONDITIONS
     IN THE PATUXENT RIVER USING THE MODIFIED TIDAL PRISM METHOD
Segment
0
1
2
3
4
5
6
Segment
Length
15
9
5
3
4
3
3
.4
.6
.9
.9
.1
.9
.8
3
5
4
5
5
4
4
P.
I.3
.6
.7
.1
.8
.0
.8
.6
x
x
X
X
X
X
X
)
106
106
106
106
106
106
106

3
3
4
4
4
5
5
Vi
(.3
.0 x
.4 x
.0 x
.4 x
.9 x
.4 x
.9 x
)
107
107
107
107
107
107
107
P
" (tidal
9
7
10
8
10
12
13
1 +Vi
Pi
cycles)
.3
.0
.7
.6
.8
.3
.8
                                               T.C. = 72.5 tidal cycles
                                                     or ^ 36 days
                                  137

-------
must be made for the fact that salinity data used in the fraction of
fresh water method were not always measured at the exact flow rates
used in the modified tidal prism method analysis.
3.4.4.2  Pollutant Distribution

3.4.4.2.1  Low Flow

     A significant problem in estimating pollutant loads into the head
of the Patuxent estuary is the long reach of tidally influenced fresh
water between the free flowing waters and the estuary head (defined
by the location where salinity gradient begins).  In this reach, the
distribution of pollutants cannot be determined by the estuarine
methods due to the lack of a salinity gradient.  Neither can it be
treated with the river methods because of the unsteady flow induced
by tidal action and the attendant increased dispersive effects.  In
effect, it exhibits both river and estuary characteristics.

     Since pollutants entering the estuary from the river must pass
through the fresh water tidal reach, a method must be devised to route
pollutants through the reach.  Conservative pollutants loads to the
estuary can be determined by a simple mass balance.  For non-conservative
pollutants, decay must be taken into account.  This was done as follows.
The reach containing the tidal fresh waters was divided into two seg-
ments, that above the confluence of Western Branch with the Patuxent
and that portion below this confluence to the head of the estuary.
A parabolic shape was assumed for the channel at Bowie where the closest
reliable hydraulic information existed in the upstream direction.  Using
this assumption the cross sectional area was calculated, knowing the
top width and flow depth at the low flow rate.  (For a parabola, the
cross section is two-thirds of the product of the top width and
maximum flow depth.)  The estuary at Hall Creek is also roughly para-
bolic in shape.  The flow in the river just above the confluence of
Western Branch was determined by linearly interpolating between the net
fresh water flows at Bowie and Hall Creek.  Net velocities at each of

                                138

-------
the three locations (Bowie, confluence of Western Branch and confluence
of Hall Creek) were estimated by dividing calculated flows by the
respective cross sectional areas.  The cross section at Hall Creek
was taken to be the average cross section between high and low tide.
Tidal variation at Bowie was assumed to be zero.  From these net
velocities, the excursion time for a plug was determined for each of
the two segments.  The excursion times under the conditions of the
70, Q low flow were 60 days from Bowie to the confluence of Western
Branch and 64 days from the mouth of Western Branch to the confluence
of Hall Creek with the Patuxent River.  Because the excursion times
are so much longer than the period for which flow was averaged, the
use of this flow and these excursion times to estimate pollutant decay
and subsequent loadings to the estuary is unrealistic.  A more reason-
able method for calculating a "normal" load to the estuary is
demonstrated below using BOD as an example.

     Consider Figure 3.4-4, which shows the relative frequency of the
seven-day moving average of mean daily flows of the Patuxent River at
Laurel, Maryland.  If it is assumed that the frequency of seven-day
moving average flows at the head of the estuary has the same probability
mass function (pmf) as those occurring at Laurel, then the expected
seven-day average flow at the head of the estuary can be found by
scaling the class width of the histogram and finding the.expected value
of the rescaled pmf.  The expected value is given by
                        E(x) =£ 7-f(x).
                               •i  .  1     '
where x.    = designates the midpoint of the class interval, and

      f(x)i = is the relative frequency in class "i".

The class interval is rescaled by area! proportioning using drainage
area above each location which makes the new class interval
                               139

-------
  a
  a
U
z
UJ a
^^
a a'
UJ
cc
U.

UJ a
a:
_J
UJ
      1    t    J    4   3   «    7    J    9    ID   U   It    13   1*   13

                              CLRSS NO.

         PflTUXENT  RIVER NERR  LRUREL, HO-

                           CLflSS  WIDTH =  181.33 ft3  sec"1  = 5.13 m3 sec'1
     Figure 3.4-4.  Frequency histogram of  7-day moving average flows.
                                   140

-------
                    182     x         = 855 ft3/sec -
                              110 	• ^-
                              132 mi
            2
where 620 mi  = the drainage area above the Hall Creek transect, and
            2
      132 mi  = the area above the Laurel gage.

The expected value of flow is 573 ft^/sec or 16.23 m /sec.  Using this
flow and dividing by the average of the Bowie and the Hall Creek cross
               2
sections (775 m ) gives an estimate of the average velocity  in the
                                        _i
tidal fresh, water section of 0.021 m sec A. "Using this velocity and
the length of the section givens the "expected excursion time of 20.7
days.  Using a first order decay rate of O.I/day and a travel time of
20.7 days gives an ultimate biochemical oxygen demand of 4 mgl~  at the
head of the estuary.  'This, in a sense, is the flow averaged concentra-
tion of BOD  that will occur at that location.  This value is substantially
greater than the value predicted using the 7Q,Q flow but not large enough
to adversely impact dissolved oxygen levels in the estuary.

     Concentrations of total nitrogen and total phosphorus were
estimated at the head of the estuary for the 70,0 low flow condition.
These predicted concentrations are solely the result of treatment plant
effluent discharges in the upper basin.  Because total nitrogen and
total phosphorus are treated conservatively, they are distributed as
the fraction of fresh water in the estuary.  Figures 3.4-5 and 3.4-6
compare predicted total nitrogen and total phosphorus to observed data
collected by the Maryland Department of Natural Resources on 27
September 1979.  On this date, the flow at the Bowie gage was 0.71 m  sec" .
The 7Q,Q flow used in the analysis was 0.53 m  sec" .

     The most obvious disparity between predicted and observed values is
the decrease in total nitrogen and total phosphorus concentrations up-
stream of the observed location of the dilution gradient.  This dilution
                              141

-------
   90
  80
  70
  60
3.
<0
I
Q.
O  •
§40
_i
I'
O

  30
  20
  10
                                                      Observed
                                                      Total Nitrogen
                                                      (TKN + NO3)-N
                                                      780927
                                                     Observed
                                                     Chlorophyll -a
                                                     780927
                          DISTANCE FROM MOUTH
                                 (Km)
      Figure  3.4-5.
Predicted and observed total  nitrogen and
observed chlorophyll-a_ in  the Patuxent
River,  September  27,  1978.
                                 142

-------
                  Tidal Fresh Water
                                                       Estuarine Water
     30
      25
•      20
oc
O

Q.
w
O
X
Q.
      10
                                                            Observed Total

                                                            Phosphorus (whole)
                                                            780927

                                                            Predicted Total

                                                            Phosphorus
100      90      80      70       60       50       40


                        DISTANCE FROM MOUTH (Km)
                                                                30
                                                                   20
10
Figure 3.4-6.
                      Predicted  and observed  total  phosphorus in  the
                      Patuxent River, September 27,  1978.
                                        143

-------
should occur where salt water begins to  intrude on the fresh water out-
flow from the watershed.  This  indicates  that total  nitrogen and
phosphorus are not acting conservatively  as the fraction of fresh
water method assumes.  The reason for this behavior  may be copious
algal production at  the interface of the  tidal fresh and estuarine
waters as evidenced  by elevated chlorophyll-a_ concentrations there.
.(See Figure 3.4-5.)  Nitrate nitrogen decreases rapidly in this zone
but total Kjeldahl nitrogen (whole sample) increases only slightly.
As a result total nitrogen (shown here as TKN (whole) + NO., as N)
decreases drastically.  Macrophyte uptake or phytoplankton utilization
followed by settling may account for the  overall total nitrogen reduction
in the waters.  Total phosphorus likewise decreases  in this segment of
the river.

     The zone in which this action occurs probably corresponds to the
null zone described  by Officer  (1976) as  the zone at which riverine
type flow occurs up  the estuary and estuarine or density gradient flow
occurs down the estuary.  This  zone-is also the area of longest particle
residence time and,  consequently, the zone in which  the phytoplankton
crop and turbidity usually achieve their  maximums.

     In Figure 3.4-7 the peak chlorophyll-a_ concentration has moved
downstream approximately 14 km  in comparison with the September study.
This chlorophyll-a_ data was taken earlier in the summer (July 19,
1978) at a higher upstream flow rate.  Areally extrapolated fresh
                                                 3    _i
water flow at the Hall Creek transect was 11.84 m  sec   versus
4.17 m  sec   at the same location on September 26,  1978.  Predicted
total nitrogen levels are much  closer to  observed values for this
date.  Total nitrogen behaves more conservatively and consequently
the fraction of fresh water method yields better estimates.  The
same is true for the total phosphorus predictions in Figure 3.4-8.
                                144

-------
en
                          70
                          60
                          50
                        01

                        CO
                         i40
X
Q_
o
OC
O
X
o
                          30
                          20
                          10
                            ~5
                            o>
04
O
oc

z
_i
< i
i- J
O
                                          -Tidal Fresh Water-
                                                             -Estuarine Water-
                                                                                 Observed
                                                                                 Total Nitrogen
                                                                                 (TKN-f N03)-N
                                                                                 780719

                                                                                 Predicted
                                                                                 Total Nitrogen

                                                                                 Observed
                                                                                 Chlorophyll • a
                                                                                 780719
                                                                           I
                               100     90       80       70       60        50       40
                                                               DISTANCE FROM MOUTH
                                                                        (Km)
                                                               30       20
                                                                                      10
                           Figure  3.4-7.   Observed and predicted  total  nitrogen and  observed chlorophyll-a_
                                            in  the Patuxent River,  July  19, 1978.

-------
                                   Tidal Fresh Water
                                                                               Estuarine Water
S r
en I
                                                                                                Observed Total
                                                                                                Phosphorus (whole)
                                                                                                780719
                                                                                                Predicted Total
                                                                                                Phosphorus
                                                          60       50       40
                                                         DISTANCE FROM MOUTH
                                                                 (Km)
10
                         Figure  3.4-8.   Observed and  predicted total  phosphorus  in the  Patuxent River,
                                          19 July 1978.

-------
3.4.4.2.2  High Flow

     There were essentially no observed water quality data available
at a flow rate comparable to that used in the Patuxent River high
flow analysis.  Therefore, high flow analyses were only done for
total nitrogen and total phosphorus in order to draw conclusions con-
cerning eutrophication.  (See section 3.4.4.3.)

     The total nitrogen and total phosphorus loads provided by the
Midwest Research Institute's nonpoint source calculator were divided
into two groups:
     0  Loads entering the estuary with the upstream river
        flow, and
     •  Loads entering the estuary laterally from adjacent
        land areas.
The former were distributed according to the fraction of fresh water
in each high flow modified tidal prism segment.  The latter were
distributed through use of the distribution coefficient matrix.  The
nonpoint loads due to lateral inflows were provided as a single value.
This value was proportioned to each modified tidal prism segment by
dividing the length of the segment by the total length of the estuary.
Resultant loads were assumed to be discrete point loads entering at
the center of each segment.

     The Patuxent River estuary under the high flow scenario was
divided into seven segments.  The segment characteristics and the
related information needed to perform the analysis are given in
Table 3.4-10.  The hypothetical flow rate through the segment is the
riverine flow per tidal cycle divided by the fraction of fresh water
in the segment (Officer, 1976).  The approximate net fresh water flow
                            3    _i
rate past Hall Creek is 83 m  sec   for this segmentation scheme or
        fi  3
3.6 x 10  m  per tidal cycle.  Salinity data for high flow was obtained
                               147

-------
TABLE 3.4-10.  CHARACTERISTIC DATA FOR THE PATUXENT RIVER
                  ESTUARY AT HIGH FLOW
Segment
0
•1
2
3
4
5
Total
.Length
(km)
15.4
9.6
5.9
3.9
4.1
3.9
46.6
Hypothetical Flow
Through Segment
(m3 tidal cycle"1)
4.32 x 106
• -5.78 x 106
6.52 x 106
1.24 x 107 '
1.63 x 107
1.79 x 107
8.3 x 107
Salinity
(ppt)
1.6
3.6
4.3
6.7
7.4
7.6
Su = 9.5
D
Fraction
of Fresh
Water (-)
0.83
• 0.62
0.55
0.29
0.22
0.18
                            148

-------
                                                          3    -1
from STORE! and was measured at a flow rate of only 15.9 m  sec   at
the same location.  Thus, concentrations predicted will be lower than
what would be predicted if coincidental salinity and flow data were
available.

     The distribution coefficient matrix, necessary for treatment of
lateral loads, was constructed from the salinity and fraction of
fresh water data in Table 3.4-10 and is shown in Table 3.4-11.  As
in the Chester River example, the entries in the upper right hand
corner are computed by dividing the salinity in the i   segment by
the salinity in the segment of discharge.  For example, entry C0 /\ =
                                                               L JT- •
0.58 is calculated by dividing the salinity in segment 2 by the salinity
in segment 4 (4.3/7.4 = 0.58).  Similarly, the entries in the lower
left hand corner are determined by ratios of fraction of fresh water.

     Total nitrogen is used to provide a comprehensive example.  The
load carried into the estuary in the riverine flow is 1.18 x 10  kg-N.
This is distributed by the fraction-of fresh water divided by the river
flow rate per tidal cycle to give a concentration and is shown as
Column 2, Table 3.4-12.  Column 3 is the fraction of the total length
of the estuary that is attributed to each segment.  This column multi-
plied by the total lateral load into the estuary (4.22 x 10  kg) and
divided by the hypothetical flow rate in each segment  gives the
initial concentration due to lateral loads in each segment (Column 4).
These loads are multiplied by the diagonal entries of .unity on the
distribution coefficient matrix.  These are then multiplied by the
coefficients in each column to obtain the concentration distribution
in the estuary (Columns 5).  These entries are summed for each row over
all columns and added to the river borne loads in each segment (Column
2) to give Column 6, the total steady state concentration in each seg-
ment in mg £~ .  The average concentration in the estuary also shown
in Table 3.4-11 is found by computing a flow weighted average concen-
tration using each segment.
                               149

-------
TABLE 3.4-11.  DISTRIBUTION COEFFICIENT MATRIX FOR THE
               PATUXENT RIVER HIGH FLOW
Segment Number


^
-Q
3
C
O)
CD
O)



0
1
2
3
4
5
6


0
0
0
0
0
0
0
1
.75
.66
.35
.27
.24
.22

0

0
0
0
0
0
1
.44
1
.39
.47
.35
.32
.29
2
0.37
0.84
1
0.53
0.40
0.36
0.33
3
0.24
0.54
0.64
1
0.76
0.69
0.62

0
0
0
0

0
0
4
.22
.49
.58
.91
1
.91
.82
5
0.21
0.47
0.57
0.88
0.97
1
0.90
6
0.21
0.46
0.55
0.86
0.95
0.97
1
                           150

-------
TABLE 3.4-12.  TOTAL NITROGEN CALCULATION IN THE PATUXENT RIVER
                      ESTUARY FOR HIGH FLOW
(1)
Segment
0
1
2
3
4
5
6
(2)
Contribution from
Riverine Flow
(mg £-1)
27.2
20.3
18.0
9.5
7.2
6.6
5.9
(3)
0.33
0.21
0.13
0.08
0.09
0.08
0.08
(4)
3.2
1.5
0.83.
0.27
0.23
0.19
0.17
(5)
:0 1 2 3
3.2 0.7 0.3 0.1
2.4 1.5 0.7 0.1
2.1 1.3 0.83 0.2
1.1 0.7 0.4 0.27
0.9 6.5 0.3 , 0.2
0.8 0.5 0.3 0.2
0.7 0.4 0.3 0.2
456
0.1 0.0 0.0
0.1 0.1 0.1
0.1 0.1 0.1
0.2 0.2 0.1
0.23 0.2 0.2
0.2 0.19 0.2
0.2 0.2 0.17
AVERAGE
(6)
Total
Concentration
(nig £-J)
31.6
25.3
22.7
12.5
9.7
7.8
7.1
12.0

-------
     Results of the calculations for total nitrogen and total phos-
phorus are shown in Table 3.4-13.  The upper limit numbers assume
that all of the storm nonpoint source loads enter during a tidal
cycle and remain at that steady-state loading rate during succeeding
tidal cycles.  The lower limit concentrations assume that the storm
loads enter the estuary equally distributed over the seven-day high
flow period.  This is equivalent to assuming that the total storm load
is introduced into the estuary approximately every 14 tidal cycles.
The true concentration should lie between these limiting values.

     The impact of urban nonpoint source runoff contributions from
the towns of Laurel, Bowie, and Columbia on total nitrogen and total
phosphorus concentrations in the Patuxent River estuary was also
determined.  As in the Sandusky analysis, the urban loads were sup-
plied on an annual basis.  Therefore, two cases were investigated.
The'first assumed that urban nonpoint source runoff entered the
river system distributed evenly throughout the year.  This amounted
to loads of 406 kg of total N and 194 kg of total P being supplied
to the stream in the high flow runoff event.  The alternative
assumption was that all of the total annual load was washed off by a
single high flow event.  Under this assumption loads to the estuary
were 21182 kg of total N and 10091 kg of total P per event.  These
quantities were added to the riverine load and distributed by the
fraction of fresh water method in the estuary.  The resultant con-
centrations are given in Table 3.4-14 for the case that all the
annual urban load is released in a single high flow event.  The
addition of urban total N and total P loads under the other assumption
had no impact on estuarine water quality.

     Comparison of the mean estuarine concentrations indicates that
even for this "worst case" assumption the urban nonpoint source loads
have only a small effect on water quality in the estuary.  The mean
total nitrogen concentration increased 12% which the total phosphorus
mean concentration increased by 36%.  The N:P ratio using only

                       	  152   	

-------
        TABLE  3.4-13.   UPPER AND  LOWER  LIMIT  TOTAL  NITROGEN AND TOTAL
               PHOSPHORUS  CONCENTRATIONS  IN THE  PATUXENT  RIVER
                       DUE TO NON-URBAN NPS LOADING
Upper Limit
Total N
Segment (m9 &'1)
0 31.6
1 25.3
2. 22.7.
3 12.5
4 9.7
5 7.8
6 7.1
Lower Limit
Total N
(mg £-1)
2.26
1.81
1.62
0.89
0.-69
0.56
0.51
Upper Limit
Total P
(mg £-1)
7.1
5.4
4-8.
2.5
1.9
1.6
1.4
Lower Limit
Total P
(mg £-1)
0.51
0.39
0.34
0.18
0.14
0.11
0.10
Average Estuary
 Concentration
12.0
0.88
2.5
.18
                                    153

-------
TABLE 3.4-14.  UPPER LIMIT TOTAL NITROGEN AND PHOSPHORUS
   CONCENTRATIONS  IN THE  PATUXENT  RIVER  DUE  TO URBAN
                AND NON-URBAN NPS  LOADS
Segment
0
1
2
3
4
5
6
Average Estuary
Concentration
Total N
(mg a"1)
36.5
28.9
25.9
14.2
11.0 '
9.0
8.1
14.1
Total P
(nig i~ )
9.4
7.1
6.3
3.3
2.5
2.2
1.9
3.4
                           154

-------
non-urban loads (based on predicted values) were 4.8:1.  The same ratio
computed using both urban  and  non-urban loads decreased to 4.15:1.


3.4.4.3  Estuarine Eutrophication

     The ratios just presented would tend to indicate that as urban-
ization of the watershed continues, algal growth may become more
nitrogen limited.   Accompanying this trend is also a seasonal trend
in the N:P ratio.   Figure 3.4-9 shows N:P ratios for a three-year
period from May 1978 to November 1970 taken near the PEPCO Chalk
Point Power. Plant.  The periodic trend is that higher N:P ratios
occur in the spring corresponding to periods of high runoff and con-
sequently high nonpoint source loadings.  The lower N:P ratios occur
during the typical low flow autumn period.  Based on predicted total
nitrogen and phosphorus for the dates 26 September 197-8 and 19 July
1978, this ratio would be 2.60.  Therefore, the predicted N and P
reflect this seasonal periodicity, low N:P ratios during low flow
with higher N:P ratios occurring during high flows.

     N:P ratios were computed from the Maryland DNR data for the
low flow dates mentioned above.  The ratios were calculated for several
locations in the estuary and averaged to give values of 5.85 and
4.79, respectively.  These values are roughly twice the predicted N:P
ratio but are still in the region that one could conclude that algal
growth is nitrogen limited.  Longitudinal variation in the N:P ratios
showed no particular trend.


3.5  DEMONSTRATION EXAMPLE:  THE WARE RIVER

     The Ware River is the smallest of the watershed-river systems used
to demonstrate Midwest Research Institute's nonpoint source calculator
and Tetra Tech's non-designated 208 screening methodology.  Because'
of its small size, a unique feature exists here not found in the other
systems.
                                 155

-------

-------
     The Ware River is composed of several creeks that drain through
marshy land into an estuary that is tributary to the Chesapeake Bay.
(See Figure 3.1-4.)  During high flows the salinity gradient exists
only in the main estuary.  However, during dry periods, the salinity
gradient moves up into .the tidal portions of the tributary creeks,
and the salinity in the main estuary becomes essentially that of the
Chesapeake Bay.  This situation affords the opportunity to apply the
estuarine methods to a very small estuary (V.,,j - 4.66 x 10  m ).
3.5.1  Data Collection

     The USGS provided 7% minute series topographic maps, as well
as flow data and stage-discharge curves for the one existing gage in
the basin, Beaver Dam Swamp near Ark, Virginia.  Ten years of flow
data were used to estimate the 7Q,Q flow for low flow analysis.

     Water quality data were available from the EPA STORET system,
although they were of marginal value for this demonstration.  Of
greater use were data being collected in a monitoring program conducted
by the Virginia Institute of Marine Sciences.  Additionally, the
Tidewater Regional Office of the Virginia Water Control Board pro-
vided some useful data in the form of two documents relating the
results of two special water quality studies carried out in Fox Mill
Run (a tributary to the Ware River) during 1977.


3.5.2  Data Reduction and Supplementation

     The available flow information at the USGS gage was used to
evaluate 7Q,n flow.  However, no salinity data in the estuary were
                                                 3-1
available to coincide with this low flow (0.003 m  sec" ).  There-
fore, a water quality analysis corresponding to 7Q,Q flow conditions
was not done for this estuary.
                                157

-------
     Bathymetric information and mean tidal ranges were obtained from
USGS topographic maps of the basin.  From this information, estuarine
cross sections at various transects were determined at mean high and
mean low tides.  This information was subsequently used to evaluate
estuarine flushing times and to determine the distribution of pollutants
in the system.  The data are shown in Table 3.5-1.

     Some hydraulic data for the estuarine analysis of Fox Mill Run
were given in the special studies report of the Water Control Board
(Virginia Water Control Board, 1977).  Using these data, together with
information from topographic maps, this creek was hydraulically
characterized from the outfall of the Gloucester Sewage Treatment
Plant to its mouth.  Where insufficient data existed, cross-sectional
areas for the creek were determined by linear interpolation.
3.5.3  Estuarine Analysis of Fox Mill Run

     The only permitted discharger in the Ware River basin is the
Gloucester Sewage Treatment Plant, which is located on Fox Mill  Run
approximately 4.5 km upstream of the mouth of the creek.   The tidal
portion of the creek begins about 0.5 km downstream from the outfall.

     On August 10 and 11, 1977, flow and water quality samples were
taken from the plant effluent and at several stations in the creek,
one of which was upstream of the outfall.  These data were collected
by the Tidewater Regional Office of the Virginia Water Control Board.
The observed data were compared to concentrations predicted by the
estuarine screening methods.  The application and results are described
below.

     The quality of the plant effluent and the quality of the natural
waters upstream of the outfall were tabulated.  Table 3.5-2 shows these
data.  Using this information, flow-averaged concentrations of quality
                               158

-------
          TABLE 3.5-1.   WARE RIVER ESTUARINE HYDRAULIC DATA
Location
of
Transect
Ware Neck Pt.
Windmill Pt.
Jarvis Pt.
Horse
Hall
Confluence of Beaver
Dam Swamp and Fox
Mill Run
Distance
from
Mouth
(km)
0.0
4.0
' 7.3
8.9
10.4

11.1
MLT
Cross
Section
(m2)
10,038
4,129
2,192
878
325

362
MHT
Cross
Section
(m2-)
15,063
5,289
3,017
1,279
637

942
Width
(m)
3,383
1,585
1,128
549
427

792
Hydraulic
Depth^
(m)
3.71
2.97
2.31
1.97
1.12

0.67
a'Computed from the tidally averaged cross-sectional  area.
                                  159

-------
 TABLE 3.5-2.  SEWAGE TREATMENT EFFLUENT AND NATURAL WATER QUALITY IN
                    FOX MILL RUN AUGUST 10-11, 1977


3 -1
Flow (m sec. )
Temperature ( C)
CBODu (mg r 1)
NBODU (mg a'1}
Total N (mg r1) as N
Total P (mg jr1) as P
Total Suspended Solids (mg £~1)
N02 + N03 (mg z'1) as IT
Gloucester STP
Effluent -
0.006
30.0
45.3
66. 3C (119)b
19. 9d
10.0
27.6
5.4
Fox Mill Run
(Upstream of Effluent
Outfall)
.024
29.5
1.83C
0.46d
0.10
4.5
0.06
aValue reported Is BOD   with suppressed nitrification.
 Estimated by type of treatment facility.
cDerived from total Kjeldahl  nitrogen data.
dSum of TKN, N02 and N03 (as N)
                                  160

-------
parameters were calculated and used as inputs to the head of the tidal
portion of the Fox Mill Run estuary.

      Modified tidal prism segmentation was used to determine the
distribution of both conservative and nonconservative substances in
this small estuary.  The tidal portion of Fox Mill  Run was described
by a series of four estuarine segments with a fresh water inflow at its
head of 1.32 x 10  m  per tidal cycle.  Table 3.5-3 shows the data
required to perform the estuarine analyses for Fox Mill Run.  The
distances given are measured from the mouth of the creek to the center
of each segment.   Salinities were interpolated from observed salinity
profiles of August 10-11, 1977 and represent the average profile over
several tidal cycles.  Fractions of fresh water values were computed,
using a background salinity of 20 ppt (Lippson, 1973).  Exchange ratios
(r-) were computed as the inverse of the flushing times for each
individual segment.  (Flushing times produced by the modified tidal
prism method and the fraction of fresh water method compared very
favorably, being 5.2 tidal cycles and 5.3 tidal cycles, respectively.)
Two sets of B. (decay correction terms) were computed for the estuary
using the above exchange ratios.  One set was calculated using a high
deoxygenation rate (0.8 day"  at 20°C) and the other using a low rate
(0.1 day"  at 20°C).  These rates were corrected for temperature.  The
average water temperature over the study period was 29.5°C with no
appreciable longitudinal variation.

      In Figure 3.5-1 the observed CBOD3Q with suppressed nitrification
are plotted against three predicted CBOD  profiles.  The two lower
predicted profiles were computed using average plant effluent character-
istics from January 1978 to December 1979.  The upper predicted profile
was computed using a CBOD  concentration of 74 mg i~  which represents
the average of the effluent CBOD3Q measured on 10 and 11 August 1979.

      The figure leads to two important conclusions.  First, the dif-
ference in decay coefficients between the lower two curves to  greater

                                 161

-------
                                TABLE  3.5-3.   DATA FOR ESTUARINE ANALYSIS OF

                                 FOX MILL RUN BY MODIFIED TIDAL PRISM METHOD
cr>
r\>
Segment
0
1
2
3
Distance
from
Mouth
(km)
3.72
3.20
2.36
1.08
Si
(PPt)
3.5
8.6
13.4
17.8
f.
( - )
0.83
0.57
0.33
0.11
ri :
(tidal cycles ~l)
0.76
0.83
0.71
0.77
Bi
(low decay)
0.98
0.98
0.97
0.98
B.
(high decay)
0.87
0.91
0.84
0.88

-------
CTi
CO
                        35
                        30
                        25
                    CD
                    E
<    20
5
ui
Q
UI

1    '•
O
                   P    10
                                                           •	• Observed BODJO
                                                                  Suppressed

                                                           »	• Predicted CBODUi
                                                                  low decay, average effluent

                                                           AC---A. Predicted CBODU, high decay,
                                                                  average effluent

                                                           T- — T Predicted CBODU, low decay,
                                                                  upper limit effluent
                                                       DISTANCE FROM MOUTH
                                                                (Km)
                             Figure  3.5-1.   Predicted  and  observed  CBOD  in  fox  Mill  Run.

-------
than an order of magnitude, but the effects of decay compared to the
effects of dilution even in this small estuary are minor.  Second, it
appears that the modified tidal prism method, using the method of
Officer (1976) to decay the nonconservative pollutants, attenuates
them too quickly.  It is possible that another source of CBOD enters
the stream between kilometers two and three causing the predicted and
observed profiles to diverge.  Contamination from replacement waters
could partially explain the high observed values at the estuary mouth.
However, the observed concentrations increase in the upestuary direction
supporting the postulate that an unidentified source is contributing
CBOD.  In fact, a stream  fed by a small pond does enter Fox Mill Run
between kilometers two and three.

      Treating total nitrogen as a conservative material, levels were
predicted in the Fox Mill Run estuary using the fraction of fresh
water method and the segmentation scheme in Table 3.5-3.  Figure 3.5-2
shows the predicted distribution versus observed data.  The observed
data represent the sum of total Kjeldahl nitrogen and nitrite- and
nitrate-nitrogen expressed as N.  The two profiles compare favorably
with the exception of the discrepancies at the head and mouth of the
estuary.  Because of the good reproduction of observed values every-
where else, the tendency is to believe that these deviations represent
sampling errors.

      Using the predicted concentrations in segment 3, .the loads to
the Ware River estuary can be calculated by multiplying these con-
centrations by the effective downestuary transport rate.  The following
low flow loads per tidal cycle result:
       »  Total nitrogen, 5.9 kg
       »  Total phosphorus, 2.7 kg and
       i^  Ultimate oxygen demand, 51.8 to 45.4 kg.  (The range
         is determined by the choice of decay coefficient for BOD.)
                                  164

-------
CD    4
6
z
LLJ

O
                                                                      • Observed
                                                                        Total Nitrogen
                                                                        (TKN + NO, + N
                                                                        770810

                                                                      • Predicted
                                                                        Total Nitrogen
                                         I
                                    DISTANCE FROM MOUTH
                                             (Km)
      Figure 3.5-2.   Predicted and  observed total  nitrogen in Fox  Mill Run,
                                            165

-------
3.5.4  Ware River Estuary Flushing Times

      Both the tidal prism and modified tidal prism methods were used
to determine flushing times for the Ware River estuary.  The flushing
time produced by the tidal prism method was 4.3 tidal cycles.  This re-
sult is insensitive to flow rate changes.  Using the modified tidal
prism method, flushing times of 58.2 and 39.3 tidal cycles were cal-
culated for low and high flow conditions, respectively.  It was shown
in the Chester River example that the relationship between mean low
tide volume and the ratio of low flow modified tidal prism to tidal
prism flushing times for the Ware River was consistent with the results
found in the' Chester River.  The small difference in these two flushing
times indicates that advective flow has only minor impact on the flush-
ing processes in the estuary.
3.5.5  Pollutant Distribution in the Ware River

      A comparison of the loads of total nitrogen and total phosphorus
calculated during low flow with those loads predicted during a typical
high flow event using the nonpoint calculator shows that loadings occurring
during low flows are almost negligible.   For instance, Fox Mill Run con-
tributes approximately 600 kg of total nitrogen and 80 kg of total
phosphorus (assuming a delivery ratio of 0.1) to the Ware Estuary during
an everage high flow event.  This load can be assumed to enter the
estuary in one tidal cycle.  The calculated loads during low flow were
only 5.9 and 2.7 kg of total nitrogen and total phosphorus per tidal cycle
respectively.  (See section 3.5.3.)   Because the concentrations of total
nitrogen and phosphorus during low flows were small compared to the high
flow loads, they were not considered in this analysis.  High flow analyses
only were performed on the Hare River estuary for the sediment, total
nitrogen, total phosphorus, and BOD,, parameters.  These analyses were
performed assuming that the total "average" storm load enters the estuary
on each tidal cycle giving upper limits for the constituent concentrations
                                  166

-------
in each estuarine segment.  Loads were provided by the nonpoint source
calculator.  A delivery ratio of 0.1 was used in all subsequent calcula-
tions.

      From Figure 3.5-3 it does not appear that the prediction of sediment
distribution in the Uare estuary is particularly good.  It has been
pointed out previously that the distribution shown is an upper limit
because of the assumptions that the nonpoint source loads are steady
and continuous and that settling of the suspended material is negligible.
It should be noted, however, that the large concentrations predicted in
the upper estuary are associated with segments having small volumes.
The volumetrically weighted mean sediment concentration (also shown in
Figure 3.5-3) is not unreasonably greater than the oberved profile.
The volumetric mean is actually calculated as a weighted average with
the weights in each segment equal to the hypothetical flow rate (R/f.)
                       .. .-                 3         . ...  i          '
in each segment when R is the river flow (m  tidal cycle  ) and f. is
the fraction of fresh water in the segment.
      The total phosphorus distribution (Figure 3.5-4) closely resembles
the observed total phosphorus profile throughout most of the estuary.
The volume-weighted mean of 0.05 mg £   is representative of actual total
phosphorus concentrations in the estuary during and immediately following
high flows.

      Predictions for sediment and phosphorus were in general better
than predictions for total nitrogen and BODg.  Figure 3.5-5 shows total
nitrogen (TKN + N02 + NO.,) and BOD,- observed in the estuary at high
water slack on May 15, 1979 three days after a major streamflow event.
The volumetric mean of total N and BOD,- (low decay) are also illustrated.
The BODr profile was computed using a decay rate of O.I/day (20°C) adjusted
to the mean water temperature of 23.1°C on that date.  Although the
prediction of the BOD- and total nitrogen distributions were poor, the
mean estuarine concentration is reasonably predicted by the estuary
screening techniques.  The trend observed in the phosphorus, nitrogen,
                                  167

-------
cr>
CO
                      350 r
                      300
                      250
                      200
CD






I
o
                   2 150
                   a.
                   w

                   to
                      100
                       50
                          817  538
                                    \
                                                                            -• Predicted Suspended

                                                                               Sediment


                                                                            -• Observed Suspended

                                                                               Solids 790515
                                                                      Calculated Mean
                                                                      (•••••••••^•••••••••••••i

                                                                           Concentration
                                                                                        -•—-*-
                                                                                                             10
                                                                                                                     11
                                                              DISTANCE FROM HEAD

                                                                     (Km)
                       Figure 3.5-3.   Suspended sediment distribution in  the Ware River during  high  flow.

-------
UD
                       .35
                       .25
                       .2
CD


O-
                   cr
                   o
o
a.
<
O
                       .15
                          0.40
                                \
                                 \
                                  \
                                       \
                                         \


                                             \

                                                                                Predicted Total
                                                                                Phosphorus
                                                                                Observed Total
                                                                                Phosphorus 790515
                                                                       Calculated Mean
                                                                       Concentration
                                                                           I
                                                                   5        6
                                                             DISTANCE FROM HEAD
                                                                     (Km)
                                                                                        10
                                                                                                                    11
                         Figure  3.5-4.   Total  phosphorus  distribution  in the Ware  River  during  high  flow.

-------
    3.5   1.4
    3.0
    2.5
o>

10

&  1'5
m
     .5
         1.2
         1.0
       <  .4
          .2
                                                           Observed BODc (Suppressed)
                                                           790515
                                                           Observed Total Nitrogen
                                                           790515
                                                                                    Concentration
                                                                                  (Low Decay)
                                                                                 Calculated Mean
                                                                            Total Nitrogen Concentration
                                            4567

                                             DISTANCE FROM HEAD

                                                     (Km)
10
11
                 Figure  3.5-5.   Observed  and predicted total  nitrogen  and 6005 in
                               the Ware River estuary during  high  flow.

-------
and BOD profiles of an increase in these parameters at the mouth of
the estuary is probably due to contamination from the Chesapeake Bay.


3.5.6  Eutrophication
3.5.6.1  Nutrient Limitation

     Using observed slack water data collected at biweekly intervals
from 11 April to 22 August, 1979, N:P ratios were computed for the
Ware River estuary.  Several dates were not used because total phos-
phorus was reported only as less than .10 mg £~ .   Uhen a "less than"
value appeared infrequently in an otherwise usable data set the upper
limit was used in computations.  The N:P ratios using total nitrogen
(TKN (whole) + N02 + N03) as N and total phosphorus (whole) as P
were computed for the dates 11 April, 25 April, 9 May, 15 May, 26 July,
8 August and 22 August, 1979.   The average N:P ratio for those dates
was 10.2.  The spring dates gave an average of 11.2, while the summer
dates had a mean N:P ratio of 8.73.  'This indicates a slight seasonal
influence in the ratio.

     The N:P ratios based on predicted values of total nitrogen and
total phosphorus for the high flow scenario was 6.8, which is lower than
the observed ratios (recall the low mean total nitrogen predictions).
This value might lead to the improper conclusion that nitrogen is
limiting.

     Based on the low flow loadings to the estuary from Fox Mill Run
(5.9 kg - N and 2.7 kg - P per tidal cycle), the N:P ratio in the
estuary is 2.2.  Using total N and P concentrations of the natural
waters (see Table 3.5-2), this ratio is 4.4.  Thus, it appears that
addition of sewage treatment plant effluent tends to shift the N:P
ratio towards nitrogen limitation.
                                171

-------
     Based on the predicted high flow and low flow N:P ratios there is
less seasonality evident in this basin than in the Patuxent.  This is
because point sources do not dominate water quality during the low flow
periods as they do in the Patuxent.
3.5.6.2  Light Limitation

     The two parameter light model
                        n  _ -In (0.1)
                         s   a + gx
where
     D  = the Secchi disc depth
     a  = background extinction coefficient
     6  = incremental extinction coefficient
     x  = a water quality constituent

was used to analyze the estuary for light limitation.  Suspended solids
and chlorophyll-a_ were used as independent variables.  The least squares
estimates of the incremental extinction coefficient, 8, for both independent
variables were found to be negative.  The values are -0.008 and -0.056
for solids and chlorophyll-a_, respectively.  These estimates for g are
meaningless since it is not expected that the water column transmits
more light at higher concentrations of these parameters.  Notice, how-
ever, that the coefficients are very close to zero.  This is because
large values of the independent variables are not found in the data set.
Therefore, in this particular data set these variables do not occur in
the light limitation range.  The regression equation is essentially
analyzing background noise; hence, the values are close to zero for the
incremental extinction coefficient.  This hypothesis is supported in a
study done by Thompson et_ aj_., 1979.  In their work, extinction coeffi-
cients were determined for light of different wavelengths in a turbid
                                 172

-------
 coastal  inlet.   Values  of extinction coefficients  were determined
 for light of different  wave lengths  during  the summer  and winter season.
 These ranged from 2.03  for light of  630 mm  wave length during  the
 summer to a high of 4.42 for light with a 445 mm wavelength  also
 during the summer season.  The a coefficients for  the  Ware-River are
 3.5 for chlorophyll-a_ and 3.4 for suspended solids,  and fall within
 the range reported by Thompson et_ al_.,  1979.   Thompson et_ al_.,  1979
 further  report  that particulates,  as opposed  to dissolved materials,
 dominate extinction.  The property most highly correlated with
 extinction in the visible light region  was  particle  cross-sectional
 area, indicating that turbidity may  be  a good choice for an  independent
 variable in the light model.
     Using 3.5 as  the extinction coefficient and substituting -In  (0.01)
for -In  (0.1). in the light parameter model, the depth of the euphotic
zone is  calculated to be  1.32 meters for the Hare River.  Comparisons
of this  depth with the hydraulic depths in Table 3.5-1 show that the
estuary  is probably light limited for algal growth  in all except the
most landward portions.
3.6  DEMONSTRATION EXAMPLE:  THE OCCOQUAN RESERVOIR

     The Occoquan basin was used in this demonstration to test the im-
poundment section of the non-designated 208 screening manual.  Because
the Occoquan Reservoir is a public drinking water supply downstream
from metropolitan areas (see Figure 3.1-5), large quantities of water
quality data were available to compare to the screening method's outputs.
This example follows the sequence of methods described in Chapter 5
of the screening manual with the exception of the discussion of water
quality during high flow events.
                                 173

-------
3.6.1  Stratification

     Using the screening manual, the first step in assessing impoundment
water quality is to determine whether the impoundment thermally strati-
fies.  This requires knowledge of local  climate, impoundment geometry,
and inflow rates.  Using this information, thermal plots likely to
reflect conditions in the prototype are selected from the screening
manual (Appendix D).

     For the thermal plots to realistically describe the thermal behavior
of the prototype, the plots must be selected for a locale climatically
similar to that of the area under study.  Because the Occoquan Reservoir
is within 32 kilometers of Washington, D.C., the Washington thermal
plots should best reflect the climatic conditions of the Occoquan water-
shed.

     The second criterion for selecting a set of thermal plots is the
degree of wind stress on the reservoir.   This is determined by evaluating
the amount of protection from wind afforded the reservoir and estimating
the intensity of the local winds.  Table 3.6-1 contains the average
annual wind speed frequency distribution for Washington, D.C. and
Richmond, Virginia.  The data suggest that winds in the Occoquan area
are of moderate intensity.

     Predicting the extent of shielding from the wind requires use of
topographic maps.  The reservoir is situated among hills that rise 25
meters or more above the lake surface within 200 meters of the shore.
The relief provides little access for wind to the lake surface.  The
combination of shielding and moderate winds implies that low wind stress
plots are appropriate.

     The geometry of the reservoir is the third criterion used in the
selection of thermal plots.  Geometric data for the Occoquan Reservoir
                                  174

-------
en
                           TABLE 3.6-1.   AVERAGE ANNUAL FREQUENCY OF WIND SPEED IN PERCENTa;

State Station
D.C. Washington
VA Richmond
Hinder 1%
Wind Speed Categories
:(km hr'1)
Mean
75 and Speed
0-5 6-12 13-20 21-29 30-39 40-50 51-61 62-74 Over (km hr-1)
11 26 35 22 5 1 * * * 15.6
14 37 36 11 1 * * * * 12.6

          Source:   U.S.  Department of Commerce,  1968.

-------
are summarized in Table 3.6-2.  The volume, surface area, and maximum depth
are all nearly midway between the parameter values used in the 40-foot and
75-foot maximum-depth plots.  However, the mean depth is much closer to the
mean depth of the 40-foot plot.

     The mean depth represents the  ratio  of  the  volume  of  the  impoundment
to its surface area.  Because the volume and surface area are propor-
tional to the thermal capacity and heat transfer rates respectively, the
mean depth should be useful in characterizing the thermal response of the
impoundment.  It follows that the 40-foot thermal profiles should match
the temperatures in the Occoquan Reservoir more closely than the 75-foot
profiles.  However, it is desirable to use both plots in order to bracket
the actual temperature.

      Flow data provide the final information needed to determine which
 thermal plots should be. used.  Most of the inflow comes, from two
 tributaries whose confluence form the upper end of the impoundment.
 The flows in these two creeks are listed in Table 3.6-3.

      The hydraulic residence time can be estimated by using the expres-
 sion:

                  r  =V=3.71xl07m3	=2i.4 days
                   w   "         m           sec
                           20.09EL_x864oolfJ.

 Since the residence time is midway between the thermal plot parameter
 values of 10 and 30 days, both should be used to bracket the mean
 hydraulic residence time in the prototype.  It should be noted that
 these flow estimates do not include runoff from the area immediately
 around the lake.  However, the upstream Occoquan watershed is large
 enough to justify the assumption that the contribution of the immediate
 area is not significant.

      The likelihood that the Occoquan Reservoir thermally stratifies can
 now be evaluated.  For a hydraulic residence time of ten days, the
 thermal plots show that stratification is not likely for maximum depths
 of 40 or 75 feet.  In the case of a 30-day hydraulic residence time,

                                  176

-------
           TABLE 3.6-2.  COMPARISON OF GEOMETRY OF OCCQUAN RESERVOIR TO PARAMETER
                           VALUES USED TO GENERATE THERMAL PLOTS
Impoundment
Occoquan Reservoir
@ Occoquan Dam
@ High Dam
Mean
40-foot Max. Depth Plots
75-foot Max. Depth Plots
Data
Source
a
b
b
c
c
Maximum
Depth
(m)

17.1
7.92
12.5
12.2
22.9
Volume
(m3)
3.71 x 107


1.74 x 107
1.14 x 108
Surface
Area
(m2)
7.01 x 106


3.08 x 106
1.08 x 107
Mean
Depth
(m)
5.29


5.6
10.6
  Northern Virginia Planning District Commission: . HSP Model Idealized Channel  Geometry.

^Maximum depths were taken from water quality profiles
  retrieved from EPA STORET System.
  Screening Manual, Section 5.2.2.1, p. 279.

-------
   TABLE 3.6-3.   MEAN MONTHLY  INFLOWS  TO OCCOQUAN  RESERVOIRa;
Month
January
February
March
April
May
June .
July
August :
September
October
November
December
Mean
Occoquan River
(m3 sec"1)
18.24
23.35
14.91
17.02
10.66
•15.40
4.35
2.99
9.35
3.55
10.47
19.32
12.47
Bull Run River^
(m3 sec"1)
10.26
12.41
7.92
11.03
5.44
13.12
2.48
1.52
6.45
2.90
6.32
VI. 63
7.62
Total
(m3 sec"1)
28.50
35.76
22.83
28.05
16.10
28.52
6.83
4.51
15.8
6.45
16.79
30^95
20.09
a)Based on daily mean flow data for October, 1968 to September,  1976.
  Source:  USGS Regional Office, Richmond, VA

wFlow data for Bull Run River @ Clifton  are estimated from another upstream
  gage by area! extrapolation.
                                       178

-------
the profiles suggest that the reservoir develops a thermal gradient be-
tween 1°C m~  and 3 C°nf  for either value of maximum impoundment depth.
The 40-foot plots indicate stratification occurs from May to August
(see Figure 3.6-1).  However, the  75-foot plots predict that the im-
poundment will have a thermal gradient greater than 1°C m~  only at
depths greater than 17 meters.  Since the Occoquan Reservoir is only
17.1 meters deep, this suggests that the impoundment does not stratify.
     The mean hydraulic residence time can be computed using either the
average annual flow rate or the flow rate just prior to stratification.
In order to use the latter method, the flow rate during the months of.
March and April should be computed.  The flow rate for this period,
25.4 m  sec"  , reduces the hydraulic retention time to 17 days.  Since
the model predicts no stratification for a ten-day residence time, the
judgment as to whether stratification occurs becomes difficult.
     Because ten- and 30-day residence times do bracket both calculated
residence times and because the 30-day plots predict stratification
while the ten-day plots do not, it may be concluded that stratification
is possible, but not certain.  In borderline cases such as this, the
reservoir will almost certainly stratify during some part of the summer.

     The accuracy of predictions made using the impoundment screening
methodology can be assessed by comparing them with available temperature
depth data.  Temperature profiles retrieved from the EPA STORET system
are listed in Table 3.6-4.  These profiles show that stratification oc-
curs at both ends of the reservoir.

     At the upper end of the reservoir (High Dam) the thermal gradient
remained near 1°C m"  between the surface and depths of 3.0 and 4.6
meters on the two dates shown.  Below this region of gradually decreasing
temperatures, the thermal gradient increased sharply to 3.3 and 2.9°C
m"  on the June and July dates respectively.  These temperature profiles
demonstrate that the distinctly stratified conditions predicted by the
                                 179

-------
4 •
=
Z
CL
O
0 •
uo
0
4 •

Z
0.
UJ
a
8 •
0
4 •
Z
a.
UJ

8
17







10 20
TEMP. C
JUl







10 20
TEMP. C
NQV





3







3

4 •
21
Z
a.
a
8 •
12
D G
0
4 •
3=
Z
Q_
UJ
a
8 •
o uc
0
4 •
r:
z
a.
UJ

!?

/
/
/
/
1












10 20
TEMP. C
OUC



.. .-
















,
10 2'a
TEMP. C
OEC











3







4 •
C
z
a.
a
0 •
1 2
0 C
0
4 •
SZ
Z
0-
UJ
0
8 •






10 20 3
TEMP, C
SEP






/
4 •
3=
z
a.
a
8
0 C
0
4 •
=
Z
a.
UJ
Q
8 •






10 20 31
TEMP, C
on







30 "0 10 20 30 "0 10 20 3
TEMP. C TEMP. C

WASHINGTON, D,C,
40' INITERL MRXIMUM DEPTH
30 Oflr HrOR. RES. TIME

MINIMUM MIXING
0   10   20   30
   TEMP. C
                0    10   ZO   30
                   TEMP.  C
Figure 3.6-1.  Thermal profile  plots  for  Occoquan Reservoir.
                            180

-------
TABLE 3.6-4.  THERMAL PROFILE DATA FOR OCCOQUAN RESERVOIRa;
Location Date
High Dam 6/11/70
(upper end of
reservoir)
.. ..

High Dam 7/01/70


Occoquan Dam 7/19/73'
(Lower end of
reservoir)



Depth
(m)
0
1.5
3.0
4.6
5.8
0
3.0
6.1
0
1.2
4.6
7.6
16.8
Temperature
26
25
22
20
16
26
23
15
28.3
26.8
22.2
19.1
17.0
 a) Source:  U.S. EPA STORET System.
                          181

-------
thermal plots do occur.  The good agreement may be largely due to the
fact that relatively low flows occurred during 1970.  As a result, the
mean hydraulic residence time increased to 26 days, a value much closer
to the 30-day residence time Figure 3.6-1 is based on.

     At the lower end of the reservoir (Occoquan Dam), the thermal
gradient remains between 1.0 and 1.4°C m   from the surface to a depth
of 7.6 meters.  At greater depths, the gradient is very small.  Although
the initial gradient is steep enough to meet the criterion for stratifi-
cation (thermal graident _> 1°C m~ ), it is not as steep as the thermal
plots predict.  The reason for the poor agreement in this case is
probably that 1973 was a slightly wetter than average year.  The mean
hydraulic residence time using the annual flow for 1973, 20 days, was
substantially lower than the value of 26 days that resulted in more
strongly stratified conditions during 1970.

     These two cases demonstrate that the thermal plots can be used
successfully to predict the time and 'the degree of stratification in
impoundments.  The epilimnion depths predicted using the model are
somewhat less reliable.  In both cases, the model did not predict the
observed 1°C m~  gradient beginning at the surface.
     The temperatures predicted by the thermal plots match those
actually measured in the reservoir quite closely.  A comparison of
predicted and observed monthly mean temperatures (1974-1976) in both
the epilimnion and hypolimnion can be made using data in Table 3.6-5.
The difference between the two epilimnion temperatures averages 1.0°C
and varies between 0.2 and 1.8°C.  The difference in the hypolimnion
temperatures averages 1.0°C and ranges from 0.2 to 2.7°C.

     The close agreement of the predicted and observed impoundment tem-
peratures probably results from the relatively long hydraulic residence
times observed in two of the three years on which the averages are based.
                                182

-------
 TABLE 3.6-5.   COMPARISON  OF MODELED THERMAL  PROFILES "TO  OBSERVED  TEMPERATURES  IN OCCOQUAN  RESERVOIR
Month
March
April
May
f June
i — i
Co July
co
August
September
October
November
December
Hean Epilimnion Tanp.
40-foot Plot (°C)a> Observedc;
7
13.5
19
24
26
26
22
17
11
7
8.4
12.6
20.5
24.8
26.6
26.5
23.8
17.2
12.2
6.2
Mean Hypol imnion Temp.
40-foot Plot (°C)b> Observed0''
6
: 10
15
18
20
21
20
16
10
7
6.3
9.2
14.4
. 17.2
21.2
23.7
20.2
15.8
11.6
5.8
Epil imnion Depth
(m)
40-foot PlotW
._
-- •
4.5
5.0
6.5
7
—
--
--
--
a> Hean temperatures  in epil imnion from thermal plots with T  =  30 days and  a maximum depth of 40 feet.


W Mean temperatures  in thermocline and hypol imnion from thermal  plots with  T  = 30 days and a maximum
  depth of 40 feet.
c> Means of observed  temperatures in "upper" and "lower" layers of Occoquan  Reservoir for 1974-1976,
  at Sandy Run.

  Source:  Northern Virginia Planning  District Commission,  January ,  1979.

-------
In 1974, 1975, and 1976, the mean hydraulic residence times were 31, 18,
and 25 days, respectively.  The 30-day thermal plots should predict
results relatively close to the two low-flow years.  The differences
expected for 1975 would be less pronounced when averaged with the other
two.
3.6.2  Sedimentation

      The second step in the water quality screening method is to esti-
mate the sedimentation rate in the impoundment.  The computations used
here require knowledge of sediment loading rates, the sediment size
distribution, and the physical properties of the sediment and water.
Using this information, the trap efficiency of the impoundment is com-
puted which, along with the annual load, determines the amount of
sediment accumulation.

     A number of simpler methods of computing trapping efficiences and
sediment loads are contained in the screening manual.  The alternative
means of computing trapping efficiencies will not be used in this
demonstration due to their lower accuracy.  Since the MRI loading
functions provide sediment loads, other load estimation procedures given
in the manual will only be used for comparisons.

     The necessary sediment loading estimates were provided by the
Midwest Research Institute's nonpoint source calculator.  Table 3.6-6
contains the sediment and pollutant loads carried by major rivers and
streams in the Occoquan watershed.  Before they are~used in further com-
putations, a delivery factor must-be applied to thesa values.  This factor
(the sediment delivery ratio or SDR) accounts for the fact that not all
the sediment removed from the land surface actually reaches the water-
shed outlet.  Additional nonpoint loads from urban sources are listed
in Table 3.6-7.  They are presumed to enter the reservoir through Bull
Run River since most of the urbanized portion of the watershed lies in
                                  184

-------
                           TABLE 3.6-6.  ANNUAL SEDIMENT AND POLLUTANT LOADS IN OCCOQUAN
                                       WATERSHED IN METRIC TONS PER
00
en
Type of Load
Sediment
Total Nitrogen
Available Nitrogen
Total Phosphorus
Available
Phosphorus
BODK
D
Rainfall Nitrogen
Kettle
Run
46,898
164.46
16.45
39.01
2.18
328.92
0.72
Cedar
Run
396,312
1,457.42:
145.74
341.95
• 14.95
2,925.63
5.50
Broad
Run
142,241
518.91
51.89
114.22
5.57
1,042.45
2.00 ,
Bull
Run
232,103
789.24
78.92
202.71
12.50
1,578.47
3.92
Occoquan
River
139,685
469.46
46.05
119.42
8.43
925.85
2.48
                    ^Estimates provided by Midwest Research Institutes Nonpoint Source Calculator.
                      These values have not yet had a sediment delivery ratio  (SDR) applied to
                      them.  We will use 0.1 and 0.2 as lower and upper bounds.  The SDR does not
                      apply to rainfall nitrogen.

                      Note:  A large number of significant figures have been retained in these
                             values to ensure the accuracy of later calculations.

-------
        TABLE 3.6-7.  ANNUAL URBAN NONPOINt LOADS IN OCCOQUAN
                 WATERSHED IN METRIC TONS PER YEAR-^
Suspended     Total      Available      Total       Available
Sediment     Nitrogen     Nitrogen     Phosphorus    Phosphorus
  12,699        12.88         5.38          2.59  -        1.270       77.47
  Estimates provided by Midwest Research Institute's Nonpoint
  Source Calculator.
                                  186

-------
this sub-basin.   Pollutant loadings from sewage treatment plants are
shown in Table 3.6-8.

     Computing the annual  sediment load into Occoquan Reservoir is
complicated by the presence of Lake Jackson immediately upstream from
the reservoir.  The trap efficiency must be computed for Lake Jackson
as well in order to determine the amount of sediment entering the
Occoquan Reservoir from Lake Jackson.

     The calculation of the trap efficiency of a reservoir requires
first that the fall velocities of sediment particles be computed and,
second, that the flow pattern be modeled.  Fall velocities for spherical
particles can be computed using Stokes' law:
                   max
                   max
                                           -  D ^ d
                        = 4'71 x  10  X    - ~
where  v^ = settling velocity, m day"1
      Vi    = viscosity of water, centipoise
      D    = density of sediment particle, g cm

      DW   = density of water, g cm

      d    = diameter of sediment particle, nun

Stokes' law holds satisfactorily for Reynolds' numbers between 0.0001
and 0.5.  Generally this requirement is met for particles less than 0.7
in diameter.  Corrections for larger particles are unnecessary since
impoundment residence times are rarely so small that particles in the
range of 0.7 mm or greater are not trapped completely.
                                187

-------
TABLE 3.6-8.  SEWAGE TREATMENT PLANT POLLUTANT LOADS IN BULL RUN
              SUB-BASIN IN METRIC TONS PER YEAR*;
        Total Nitrogen      Total Phosphorus       BOD,-

             108.0                11.92            54.80

 a)Averages for July 1974 - December  1977
   Source:  Northern Virginia  Planning District  Commission,
            March 197?: -
                                188  :

-------
     Soil types provide an indication of the particle sizes in the basin
under study.  Soils in the Occoquan basin are predominately silt loams.
Particle size data on the principal variety, Penn silt loam, are given
in Table 3.6-9.  (The size fractions of water-borne sediments would be
more appropriate than j_n situ size fractions.  This information should
be used if available.)

     Some effort can be conserved by first calculating the smallest
particle size that will be completely trapped in the impoundment.
To do so, P, the trap efficiency, must first be computed.  Because both
reservoirs are long and narrow and have relatively small residence times,
the flow will be assumed to approximate vertically mixed plug flow.  In
this case, P is found from the expression:

                        "  "p =  max Tw
                                 D1

 where D1  = mean flowing layer depth, m.
     To calculate the smallest particle that is trapped in the impound-
ment, P is set equal to unity and the above equation is solved for v
                                                                    max
This expression for v    is then substituted into the fall velocity
                     max                                          j
equation (Stokes1 law), which in turn is solved for d.   The resulting
expression is:
                      •1
D'  V
                          4'71xl°
     The trap efficiency of Lake Jackson is  calculated first.   The data
required for these calculations are:
                                 189

-------
TABLE 3.6-9.   PARTICLE SIZES IN PENN SILT LOAM

Particle Size
(mm) .
4.76
2.00
0.42
0.074
0.05
0.02
0.005
0.002
% of Particles Smaller Than
(By Weight)
100
99
93 '
84
78
50
26
16
                  190

-------
        V  =  1.893  x  106 m3
        Q  =  12.37  m3  sec'1
        "D  =  3.34 m

        y  =  1-llcP        (Assuming  T= 16°C as in Occoquan Reservoir)
       T  = ^ - 1 71
       Tw   Q-1-7'
       The  minimum particle  size  for  100  percent  trapping  is  computed  as:
             d = |3.34 m x 1.11.	  =   5<18  x  10-3 mm
                 T  4.71 x 10  (2.66 -  1.0)  •  1.77
      Sediment size fractions for in situ soils are known.   However,
 particle sizes delivered to the reservoir via the stream channels are
 not determined by the screening methods, therefore,  a composite trap
 efficiency for all particle sizes is needed.   This is computed as follows
                       c
                      P,, = 1 - /100
where Wj - weight percent of sediment entering the reservoir in
           incompletely trapped particle size range (i.e ,  below
           minimum particle size 100% trapped)
      Pd = trap efficiency for particle size range d

                                                            the
                                  191

-------
      Weight fractions  in each  size range are estimated  from data  in
 Table 3.6-9 using linear interpolation.   It is  assumed  that virtually
 all  of the sediment mass consists  of particles  greater  than 0.001 mm
 in diameter.   For each size range, a mean trapping  efficiency  is  cal-
 culated.   The sedimentation calculations for Lake Jackson  are  summarized
 in Table  3.6-10.

      Substituting the  values from  Table  3.6-10  into the above  equation
 yields P   = 0.80.

      The  total  sediment accumulation in  Lake Jackson i   determined from
 the  expression:
                          V
where dp = sediment delivery ratio from USLE

      P  = composite trap efficiency

      S. = sediment load from tributary i .
       St = (0.1, 0.2) x 0.8 [46898 + 396312 + 14224lJ metric tons/year
          = (46836, 93672) metric tons/year.

 Data listed in Appendix F of the screening  manual  show  that  the  rate  of
 sedimentation in Lake Jackson is 56153  metric  tons/year.

      The next step is to compute the  sedimentation in Occoquan Reservoir.
 The minimum particle  size that is completely  trapped  is  computed using
 the following values:
                                  192

-------
                              TABLE 3.6-10.  TRAP EFFICIENCY CALCULATIONS FOR LAKE JACKSON
CO


Particle Size
(mm)
0.00518

0.005

0.0035

0.002

0.0015

0.001

Settling Velocity
(m day" )
1.89

1.76

0.861

0.281

0.158

0.070

P
1.00

0.932

0.456

0.149

0.084

0.037

-a)
P

0.966

.694

.303

.117

.061

Wt. % Particles
in Size Range
Into Lake Jackson

.288

5.0

5.0

8.0

8.0

Wt. % Particles
in Size Range
Out of Lake Jackson

.05

7.81

•17.78

36.04

38.32


                     P represents the mean trap efficiency for the given size range.

-------
        D1   =   5.29 m
        y    =   1.1111  cp  (@ T  =  16°C, mean of Table  3.6-5)
        D    =   2.66 g  cnr3

        D    =   l.Ogcm'3


Under stratified conditions,  the epilimnion  thickness should be used for
D1.  Since stratification is  uncertain in this  case and the predicted
hypolimnion thickness, 5.75 m, is greater than  the mean depth,  the latter
value will be used.   All  particles with diameter,  d, such that:
           H     /  5.29  x  1.11	      0_      -3
           d  =  /  	%	  =  1-87  x  10  mm
               V  4.71  x  10  (2.66  -  1.0)  •  21.54

will be completely trapped in the Occoquan Reservoir.  By the same
techniques utilized for Lake Jackson, the trap efficiency of the
Occoquan Reservoir may be computed.  The calculations are summarized
in Table 3.6-11.  The fraction of material from each source can now be
evaluated.  For the sediment from Lake Jackson:

           P   = 1  - (64.99 -  .822 x 26.67 -  .465 x 38.32)/100 = .748
            c

And for the sediment from the Bull  Run and Occoquan rivers:

          P  = 1 - (13.92 - .822 x 5.92 - .286 x 8.)/100 = .932


     Finally, the total annual sediment accumulation in Occoquan
Reservoir may be estimated.  Using equation  2:
                                 194

-------
Ul
                          TABLE 3.6-11.  TRAP EFFICIENCY CALCULATIONS FOR OCCOQUAN RESERVOIR

Particle Size
(mm)
0.00187

0.0015

0.001
Settling
Velocity
(m day-*) P
.246 1.00

0.158 0.644

0.070 .286
Wt. % Particles in Size Range
— a. From Bull
P : From Lake Jackson Occoquan

.822 26.67 5.92

.465 38.32 8.0

Run and
Rivers






                 P represents the mean trap efficiency for the given size range.

-------
       S  = (0.1, 0.2)  x [(46898 + 396312 +142241)  (1 - .80) x .748 +(232103+13968^1 x 0.932J

                        sediment from Lake Jackson            sediment from
                                                        Bull Run and
                                                         Occoquan
           + 12699 x 0.932
             urban
             load
        St = (55200, 98700) metric tons/year
3.6.3  Eutrophication


     In addition  to  the  assessment of impoundment thermal characteristics and

sedimentation rates, estimating nutrient levels is a major  concern.   The con-

centrations of nutrients directly affect plant growth rates, which  in
turn affect dissolved  oxygen  levels and impoundment aging.  Since nitro-

gen and phosphorus are the  macronutrients most commonly in  limited

supply, the screening  methods  focus on them.  Nutrient concentrations

depend primarily  on  the  amounts of each carried into the reservoir  by

tributaries and point  sources.   Several assumptions concerning  pollutants
in the watershed-reservoir  system are necessary in order to calculate

the desired annual loads:
     •  The unavailable  phosphorus  is adsorbed on sediment par-
        ticles.  Therefore,  of the  unavailable forms coming  into
        Lake Jackson, only  the fraction (1-P          ) is delivered
        to the Occoquan  Reservoir;            [Jackson]

     •  All of the rainfall  nitrogen is in available form;

     •  All of the phosphorus  and nitrogen from the sewage treat-
        ment plants  (STPs)  is  in  available form;

     •  The output of STPs  outside  the Bull Run sub-basin is
        negligible compared to that of the STPs in Bull Run.
        This is justified by the  fact that during the period
        under study, the plants in  Bull Run had a combined
        capacity several times larger than the few plants out-
        side the sub-basin.
                                   196

-------
     By applying these assumptions to nonpoint source data generated by
the MRI loading function and point source data reported in the liter-
ature (see Tables 3.6-6, 3.6-7, and 3.6-8) the total load of each
pollutant type may be calculated.  The computation for the total  annual
phosphorus load in Occoquan Reservoir is shown below.  First, the quantity
of total phosphorus coming into the Occoquan Reservoir through Lake
Jackson is calculated by:

     TP,  .     = (1 - P        )  x [Total P - Available P] + Available P
       Jackson         cjackson

The total phosphorus from Broad Run, Cedar Run-, and Kettle Run are summed
and the available phosphorus loads are subtracted to give the unavailable
load.  This load is multiplied by the trap efficiency of the lake, P ,
which yields the unavailable load passing through.  This value, plus
the available load, is an -estimate of the total phosphorus entering
Occoquan Reservoir from Lake Jackson.  This quantity is 117.2 metric
tons yr" .  This value is added to the non-urban, nonpoint source loads
from .Bull Run and areas adjacent to the Occoquan Reservoir (see Table
3.6-6):

                 TPNRNU = 202.71 + 119.42 + 117.21
                        = 439.34 metric tonsyr   .

This quantity is modified by the sediment delivery ratio.  The urban
nonpoint loads and STP loads are added to complete the calculation:

                TP = (0.1, 0.2) (439.34) + 2.59 + 11.92
                   = (58.44, 102.38) metric tons yr"1.

     The results of load calculations are summarized in Table 3.6-12.

     The calculated annual total  phosphorus and nitrogen loads may be
compared with the observed loads listed in Table 3.6-13.  The loads observed
                                197

-------
   TABLE 3.6-12.  CALCULATED ANNUAL POLLUTANT LOADS TO OCCOQUAN RESERVOIR
Sediment Del

Total Nitrogen
Available Nitrogen
Total Phosphorus
Available Phosphorus
BOD5
0.1
% of Load
From Source
Nonpoint Point
77" 23
33 67
80 20
32 68
93 7

Loada;
474.56
161.91
58.44
17.56
812.40
ivery Ratio
0.2
% of Load
From Source
Nonpoint Point
87 13
45 55
88 12
46 54
96 4


Loada;
813.61
195.82
102.38
21.92
1492.53
a)Units of metric tons year"
  Note:  A large number of significant figures have been retained
         in these values to ensure the accuracy of later
         calculations.
                                    198

-------
                            TABLE 3.6-13.  OBSERVED ANNUAL POLLUTANT  LOADS TO OCCOQUAN  RESERVOIR
10
«£>




Period
10/74
7/75
7/76
- 9/75
- 6/76
- 6/77
Mean Flov/''
Rate
(m3 sec"1 )
24.7
24.0
10.4

Total Nitrogen Load
(metric tons year" )
805"
1905c;
4763c;

Total Phosphorus
(metric tons year
110"
188^
454c;

Load
-1)



                ^Source:  USGS Regional Office, Richmond, Virginia.



                ^Grizzard et al_., 1977



                c/>Northern Virginia Planning District Commission, March,  1979.
                  Data gathered by Occoquan Watershed Monitoring Laboratory.

-------
by Grizzard et_ al_., (1977) are within seven percent of those calculated
using a delivery ratio of 0.2.  On the other hand, the Occoquan Water-
shed Monitoring Laboratory (OWML) reported values 1.9 to 5.9 times higher
than highest calculated loads.  Comparison of loadings (kg/ha year) with
literature values suggest that Grizzard is most accurate (Likens et al.,
1977).

     By dividing the total annual load by the total annual flow rate,
the pollutant concentrations may be estimated.  For example, the avail-
able phosphorus concentration is:

     PAV = Annual Available Phosphorus Load/Annual Flow
                              c. g
         = (17.56, 21.92) x ]
                  3
           20.09 ^— x 86400
                 5 SC

     Calculated and observed pollutant concentrations are listed in Table
3.6-14.  The mean summer concentrations of phosphorus and nitrogen are
closer to the concentrations calculated using a delivery ratio of 0.1 than
0.2, although this would not be expected on the basis of the previous
comparison of annual loads.  The discrepancy could arise from a large
seasonal variation in concentrations or processes such as adsorption onto
settling sediment that reduce the water column concentrations.

     The screening manual does present a relationship to compute the
water column total phosphorus level, C :

                                  Cin
                           w " 1 +
(See the screening manual section 5.4.5, p. 348.)

     Since the model has not been widely tested and the values of k.. and
k., are unknown, no attempt will be made to refine the previous calculation
                                 200

-------
                     TABLE  3.6-14.  CALCULATED AND OBSERVED MEAN ANNUAL  POLLUTANT CONCENTRATIONS
                                                IN OCCOQUAN RESERVOIR
r\>
o

Delivery Ratio
0.1
0.2
Observed Values'3^
Mean
Max.
Min.
Total
Nitrogen
fa m~3}

0.76
1.3

0.88
1.50
0.35
Available
Nitrogen
(q nf )

0.26
0.31

0.16
0.24
0.10
Total
Phosphorus
(q rrT3)

0.093
0.16

0.08
0.12
0.04
Available
Phosphorus
(q m"3)

0.028
0.035




BODC
b
(q m-3'

1.3
2.4




                    Averages for April-October between 1973 and 1977.
                    Source:  Northern Virginia Planning District Commission,
                             March, 1979.

-------
by using this method.  If the product k,k3 is in the range expected under
steady state conditions, between 20 and 40, the actual concentrations will
be 18 to 30 percent lower than that calculated without this correction.
Thus, the result obtained by dividing the annual load by the annual flow
should be considered an upper bound to the actual concentration.

     The ratio of nitrogen to phosphorus concentration in the reservoir
can be used to estimate which nutrient will limit the rate of plant
growth.  For the Occoquan Reservoir, the N:P ratios are shown in Table
3.6-15.  The calculated nutrient ratios care between 5 and 10, where
either nutrient could be considered limiting.  "The N:P ratio of the
observed data indicates more conclusively that phosphorus is growth
limiting.
   TABLE 3.6-15.  NITROGEN:PHOSPHORUS RATIOS IN OCCOQUAN RESERVOIR


                                    Delivery Ratio
                                      0.1     0.2        Observed
Total N: Total P
Available NrAvailable P
8.1
9.2
7.9
8.9
11.0
     The first method of predicting algal growth is known as the
Vollenweider Relationship.  In the graph of total phosphorus load
    -2   -1
(g m   yr  ) versus mean depth (m) divided by hydraulic retention time
(yrs) (see Figure 3.6-2), areas can be defined that roughly correspond
to the nutritional state of the impoundment.  For the Occoquan Reservoir,
the values of the parameters are:
                                  202

-------
 100.0 =
N
E
•X
  10.0
or
o>
c
o
O 1.0
Q.
M
   0.1
O
o
        Eufrophic

Occoquan (Delivery Ratio • 0.2):-  _•-»--

Occoquan (Delivery Ratio • 0.1).-
         I   III
                         I  I I I I III
                                      O = Oligofrophic'
                                      A = Mesotrophic
                                      Q = Eutrophic
                                    Open  Symbols = P-limited
                                    Solid  Symbols = N-limited
                               != Present Load
                               = Present Load Minus  50% MSTP Load
                               = Present Load  Minus  80% MSTP Load
                            l  l l i i i III    I  I l l  l l ill    i  ill
                  1.0
                       10.0         100.0
                    Mean  Depth  (m)
                                                         1000.0
                     Hydraulic  Retention  Time  (yrs)
 Figure 3.6-2.   Plot of the  Vollenweider relationship  showing
                   the position of Occoquan Reservoir using
                   calculated total  phosphorus  loads  (Source:
                   Zison  et al., 1977).
                                  203

-------
         Lp =  (58.4,  102.4)  x  106  g/yr =  f^ ]4>g\   m-2 yf-l
                   7.01  x 10b  rrT          \

         D_ _    5.29  m   x 365 days  _  90  m yr~l
         TW "  21.4 days     1  yr    "       y
     According to the Vollenweider Relationship, Occoquan Reservoir is°
well into the eutrophic region for both estimates of the total phosphorus
load (see (Figure 3.6-2).  Based on these predictions a more in-depth
study of the algal productivity seems to be in order.

     The available data also permits the estimation of the maximal primary
production of algae from the Chiaudani and Vighi Curve (Figure 3.6-3).
The theoretical phosphate (available phosphorus) concentration should
                        • •  -  _o                        • • • •
be between .028 and .035 g m   according to these calculations.   The
maximal primary production of algae is found from Figure 3.6-3 to be
between 1850 and 2000 mgC m   day, or 1.85  and  2.0 gC m   day"1.  This
level of algal production is roughly 75 to 80 percent of the maximum pro-
duction shown on the curve.  Both this method and the Vollenweider
Relationship suggest algal growth will contribute significantly  to the
BOD load in the impoundment (see section 3.6.5).
3.6.4  Hater Quality High Flow Events

     A substantial fraction of the pollutant loads into the Occoquan
Reservoir come from nonpoint sources.  Seventy-seven percent or more
of the total nitrogen and total phosphorus loads to the reservoir are
derived from nonpoint sources (Table 3.6-12).   Since most of the non-
point loads into the reservoir are carried by runoff, it is important to
determine the magnitude of nonpoint source pollutants, especially nutrients,
during or immediately following a high flow event.
                                 204

-------
o  :
               500
                                              P04= (as  P,  mg/l)
            Figure 3.6-3.
Maximal  primary productivity as a function of phosphate concentration (Source:
Zison et al_., 1977).

-------
     To accomplish this, 15 seven-day high flow events between October
1978 and September 1976 were selected.  The Midwest Research Institute's
nonpoint source calculator was used to estimate the loads to the
reservoir during these events.  These results are summarized in
Table 3.6-16.

     Loads during the seven-day high flow event from urban nonpoint an
point sources may be estimated by taking a fraction -FF of the annual
loads.  As in the calculations of the annual  average loads, unavailable
phosphorus is assumed to be adsorbed on the sediment.  Because of flow
differences, it is necessary to determine the trapping efficiency of
Lake Jackson during  high  flow  events.   Because  of  the lake's  long,
shallow geometry, the fluid motion is assumed to be approximated by
vertically mixed plug flow.

     'Computation of trapping efficiencies requires that hydraulic
residence times be known. The average inflow rates during the selected
high flow periods are listed in Table- 3.6-17.

     The hydraulic residence time of Lake Jackson is:


                    r  -     1.893 x 106 m3     '   _..  ,
                    Tw	3	 = 0.546 days
                         40.10 Lx 86400


     The second step in the compuation of trap efficiencies involves deter-
mining the minimum sediment size that is completely trapped in the reservoir.
The values of constants used to determine this are:

       TW = 0.546 day   = hydraulic residence time

       D  = 3.34 m      = mean depth
       y  = 1.11 cp      = viscosity
                     -3
       D  = 2.66 g cm   = particle density
       DW = 1.0  g  cm"   = water  density

                                 206

-------
                    TABLE 3.6-16.  HIGH FLOW EVENT POLLUTANT LOADS IN OCCOQUAN WATERSHED
                                     FROM NON-URBAN NONPOINT SOURCES3>
Kettle
TVDB of Load Run
Sediment, (metric tons/event) 2981
Total Nitrogen,
(metric tons/event) 10.52
Available Nitrogen,
(kg/event) 1050
Total Phosphorus,
(kg/event) 3046
Available Phosphorus,
(kg/event) 170.4
BOD5, (metric tons/event) 21.04
SUB-BASIN
Cedar Broad Bull
Run Run Run
25758 9211 14207
95.08 ' 33.75 48.48
9508 3375 4848
26718 8937 15813
1168.8 436.6 975.2
190.16 67.50 96.98
Occoquan
River
8622
28.58
2858
9295
657.1
57.17

;These values are gross loads (i.e., delivery ratio has not been applied)
 Values supplied by Midwest Research Institutes Nonpoint Event Load
 Calculator.

-------
TABLE 3.6-17.   STREAM FLOWS INTO OCCOQUAN RESERVOIR
             DURING HIGH FLOW EVENTS^
                                       Flow Rate,
     Tributary	(m3  sec"1)
   Bull  Run @ Clifton                     28.40

   Occoquon River @ Manassas               40.10

                          TOTAL           68.50
   a)Averages of 15 high flow events from
     October, 1968 to September, 1976.
     Source:  USGS Regional  Office, Richmond, VA.
                       208

-------
Substitution  of these constants  into the following  equation:
                      d =
                                ...71 104
yields  a  minimum diameter of:
                      d = 9.32  x 10~3 mm.
     The  fraction of the  total  load in size ranges  smaller than the
minimum diameter is estimated  by interpolation  between values listed
in Table  3.6-18.  The computations for the trap efficiency in each  size
range  are also summarized  in Table 3.6-18.


     The  composite trap efficiency is:


     PC = 1 -(32.9  -  6.9 x .64 -  5.(.21  + .093) - 8.(.036 +  .019)) /lOO

        = 0.735


     There is now sufficient information to calculate the total sediment
and pollutant loads to Occoquan  Reservoir during  high flow events.  The
computational steps for the total phosphorus  load are shown below:
Total  Phosphorus            Availabl e Phosphorus
                               JL.
                       t               \
                                       .6
          	   J
          __     Y

           Unavailable Phosphorus
                          I" (	*      1
                          [(3046 + 26718 + 8937)
TPJackson = (1 ' °-735)xl (3046 + 26718 + 893?) ^^  ~  (170-4 + 1168-8 + 436-5'
               ,[170.4^ 116fr+ 436.6)

                Available Phosphorus
                                   209

-------
       TABLE 3.6-18.  TRAP EFFICIENCY CALCULATIONS FOR
             LAKE JACKSON DURING HIGH FLOW EVENTS
Particle
Size
(mm)
0.00932

0.005.

0.0035

0.002

0.0015

0.001
Settling wt_ %
Velocity Particles in
(m day ) P ~ P Size Range
6.11 1.00
0.64 6.9
1.76 - 0.287
.21 5.0
0.861 0.141
0.093 5.0
0.281 0.046
0.036 8.0
0.158 0.026
0.019 8.0
0.070 0.011
  represents the mean trap efficiency for the given
size range.
                               210

-------
Total Nonpoint Non-Urban Load:
     TPNpNU = (11560 + 15813 + 9295)
     TPNpNU = (3667, 7334)
 Urban  Nonpoint  Load:
TP  =
i r
                   = 49.7 kg event
                                  "
                                                Sediment
                                                Del ivery
                                                Ratio
                          ka vr'1 x   1 yr   x  7 da;/s
                          Kg. yr    x          x
 Point-Source  Load:
               TPps  . ,1920
                     = 228.6 kg event'1
 Total  Load  from  All  Sources:

                 TP  = (3667,  7334)  + 49.7 + 228.6

                    = (3945,  7612)  kg event'1.


     The  total  load  for  each  pollutant type is  listed  in  Table  3.6-19.

     Pollutant  concentrations  in  the reservoir  during  high  flow periods
 may  be estimated by dividing  the total  event load  by  the total  event
 flow.  This is an accurate estimate if flow in the reservoir is char-
 acterized by vertically mixed plug flow.   Otherwise,  the estimate can
 be considered  to be an  upper bound for the concentration that  would be
 attained in a  partially or completely mixed impoundment.

                                 211   :

-------
TABLE 3.6-19.  TOTAL POLLUTANT LOADS TO OCCOQUAN RESERVOIR

                 DURING HIGH FLOW EVENTS
                                  Sediment Delivery Ratio
    Pollutant,                  	:	~	
(metric tons event"•*•)	0.1	0.2


 Sediment                         3532               6821


 Total Nitrogen                   23.96              45.60


 Available Nitrogen               4.339              6.503


 Total Phosphorus                 3.945              7.612


 Available Phosphorus             0.594              .935


 BODC                             45.8               89.1
                            212

-------
    Calculated maximum nitrogen and phosphorus concentrations for the
Occoquan Reservoir during high flow events are given in Table 3.6-20.
     TABLE 3.6-20.  MAXIMUM CALCULATED POLLUTANT LEVELS IN OCCOQUAN
            RESERVOIR DURING HIGH FLOW EVENTS (g nT3)
Pollutant
Total Nitrogen
Available Nitrogen
Total Phosphorus
Available Phosphorus
Sediment
0.1
0.58
0.10
0.095
0.014
Del ivery Ratio
0.2
1.1
0.16
0.18
0.023
     The total phosphorus concentration is one to 12 percent higher during
 high flow events than the annual means.  However, the remaining nutrient
 concentrations are lower for high flows.  In spite of the high nutrient
 loads during these events, most concentrations are lower because of the
 high flow rates.  It may be safely concluded that water quality will not
 worsen during high flow events.  Any planning for future water quality
 can be based on annual average loads.
 3.6.5  Dissolved Oxygen

     The final water quality parameter to be examined using the screening
 methods is hypolimnion dissolved oxygen.  If stratification does not
 occur, reaeration of deep waters occurs via mixing with surface waters.
 When a hypolimnion exists, the deep waters are cut off from oxygen
                                  213

-------
sources, e.g., surface reaeration and inflowing water.   Organic matter
decay, benthic uptake, and other oxygen demands can seriously deplete
dissolved oxygen levels.  As was determined earlier,  the Occoquan
Reservoir may stratify during the months of May through August.  The
time period is long enough that hypolimnion oxygen depletion could be
a problem.

    The simplified model used to predict hypolimnion  dissolved oxygen
levels assumes that the only substantial dissolved oxygen sinks are
water column and benthic deposit BOD.  Additionally,  all sources of
oxygen, photosynthesis, etc., are neglected in the hypolimnion after
the onset of stratification.  Thus, the procedure requires that pre-
stratifi cation levels of BOD and dissolved oxygen be  estimated in order
to compute the post-stratification rate of oxygen disappearance.  The
pre-stratification concentration of water column BOD  is determined
first.'  A simple mass balance leads to the following  relationship, if
steady state conditions are assumed:
 where GSS  =  steady  state  concentration  of  BOD  in water column, mg
       k    =  mean  rate  of  BOD  loading  from  all  sources, g m   day"1
                               kb    "  ks  "  kl " V
 where  k$  =  V/D= mean  rate of  BOD  settling out onto impoundment bottom, day
       k-|  =  mean  rate  of decay of water column BOD, day
       Q   = mean  export  flow  rate, m  day~
       V   = impoundment  volume, m
       Vs  = settling  velocity, m day"
       D   = impoundment  mean  depth, m .
                               214

-------
    The BOD load to the impoundment originates in two principal sources:
algal growth and tributary loads.  The algal BOD loading rate is computed
from the expression:

                             ka(algae)
     S   =  stoichiometric  conversion  from  algal  biomass  as  carbon to  BOD  =  2.67
     M   =  proportion of algal  biomass expressed  as oxygen  demand
     P   =  primary algal production,  g m    day
     D   =  mean  impoundment depth, m
    Since the Chiaudani curve (see section 3.6.3) gives the maximal
algal  production, a correction should be made for the actual epilimnion
temperature.  If the .maximal rate occurs at 30°C and the' productivity
decreases by half for each  15°C decrease in temperature, the algal pro-
duction can be corrected for temperature using the expression:
                            =  P(30)
According  to  the  data  in  Table  3.6-5,  the  epilimnion  temperature  during
the  month  priorto stratification  is  approximately  13°C.   Thus:

            P(13o}  =  (1.85, 2.0) gC m"2 day"1  x  i.o47(13°C-30°C)
                   =  (0.85, 0.92)  gC  m"2 day"1.


     If  M is assumed  to have  lower and  upper limits  of 0.7 and  1.0,  then:

            ka(algae) - 2.67  x (0-.7.-1;Q) x (.85,  .92) gC m~2 day"1
                                       5.293 m
                      =   f0.30,  0.46J  g  m"3 day"1
                                  215

-------
   The BOD load borne by tributaries is found by the expression:
           k  /,  .. v _ Mean Daily BOD from Tributaries
            a^ r   '         Impoundment Volume
                    =  (812.40. 1492.53) x 106 g yr-lx  1
                              3.71 x
                         1.060, O.ll)
3.71 x 107 m3           365 days
       g m"  day
   The total BOD load to Occoquan Reservoir is then:
        ka  =  ka(algae)  + ka(trib)

           =  (0.30,  0.460)  g m"3  day"1  + (0.06,  .11)  g  m"3  day"1
           =  (0.36,  0.57)  g m"3  day"1
    Before the water column BOD concentration can be computed,  the
constants comprising kfa must be evaluated.   The first of these, kg,
requires knowledge of the settling velocities of BOD particles.  Ideally
these would be determined by using values of the physical properties of
the particles and the water in the settling velocity equation,  V-6
(screening manual section 5.3.3.1, page 398).  Because such data are
lacking, a settling velocity of 0.2 m day"1 reported for detritus will
be substituted.  The reported values lie between 0 and 2 meters day   ,
with most values close to 0.2 m day"1 (Zison et a]_., 1978).  Then,
                       1.2 m day"1/5-29 m = -0378
                                 216

-------
    The second constant comprising kb is the first-order decay rate
constant for wtaer column BOD.  Reported values of k1 vary widely
depending on the degree of waste treatment.  Zison et_ a1_. (1978) presents
data for rivers, but contains only two values for k1 in lakes and estuaries.
Both are k, = 0.2 day"1.  Camp  (1968) reports values from 0.01 for slowly
metabolized industrial wastes to 0.3 for raw sewage.  Because there is
considerable sewage discharge into the Occoquan Reservoir, ^ may be
assumed to be in the upper range of these values, between 0.1 and 0.3
day"1.  Like the algal production rate, k1 must be corrected for the
water temperature.  In April, the mean water temperature is about 11°C.
Then:
                   k1 =  (0.1, 0.3) day"1 x  1.047(11°C"20°C)
                     =  (0.066, 0.20) day"1
 Finally,  kb  is evaluated as follows:

    kh = - 0.0378 day"1- (0.066, O.ZOjday-1 	m"3 sec"1 x 86400 sec day"1
     D                                             3.71 x 10' mj
       = -  f 0.15, 0.28J day"1
Next, k  and kK may be substituted into equation V-27 (screening manual
       a      D
section 5.5.2.1, page 362) to obtain Cs$.


             c   a (0.36. 0.57) 9 m"3 day"1
              SS     (0.15, 0.28) day"1

             c   = (1.3,  2.0) g m"3£  k
                                                         "1
                    (2.4,  3.8) g m"3E  k1(2QOC)  =  0.1  day
                                  217

-------
    Once the water column BOD concentration is "known, the benthic BOD
is computed from the expression:
                                k  C   D~
                         L   =  s  ss	
                          ss      k.

where  k, = mean rate of benthic BOD decay, day" .

    Values for the benthic BOD decay rate constant span a greater range
than those for water column BOD.  Camp(1968),  however, reports values
of k. very near 0.003 day"  for a range of benthic depth from 1.42 to
10.2 cm (see page 366 of the screening manual)..  Assuming this to be a.
good values, a temperature-corrected value of k. may be computed at an
April hypolimnion temperature of 10°C:

               k4 = 0.003 day"1 x 1.047(10"20) = 0.0019 day"1

Then,
               L   = 0.038 day"1 x (1.3, 2.0;  2.4. 3.8) g m"3 x 5.29 m
                SS              0.0019 day"1
                                   r\
                   = (138,212) g m"   for  k, = 0.3 day"1
                     (254,402) g m"2  fnr  k, =0.1 day"1   .

     Prior  to stratification the impoundment  is assumed  to be  fully mixed
and  saturated with oxygen.  During April,  the  hypolimnion temperature  is
10°C.  Saturated water at this  temperature contains  11.17 ppm oxygen.

     Finally, the dissolved oxygen level in the hypolimnion may be
predicted  during the period of  stratification.  The  applicable expressions
are:
          °t - °o ' AOL - i0c
         ao.
                                  218

-------
where 0. = dissolved  oxygen at time t

      0  = dissolved  oxygen at time t = 0

      D  = hypolimnion  depth

     In order to illustrate the use of these expressions,  the  computation
of a dissolved  oxygen concentration for the case  k  =  0.57  g  m"   day"
                        1                           ^
and k,/pQop) =0.1  day"  is shown below.  First,  the BOD  decay rate
constants must  be adjusted to account for their temperature dependence.
During the period from  May through August, the mean temperature  in the
hypolimnion and thermocline ts 19°C according to the thermal plots (see
Table 3.6-5).   The  temperature-corrected rate constants are:
         k4 =  0.003  day"1 x 1.047^19"20^ = 0.0029 day"1

     Next, the settling  coefficient, k  , must be reevaluated  using the
mean depth of  the  hypolimnion and thermocline.  The mean  depth (3.38 m)
was approximated by  assuming the reservoir has a triangular cross section
and an average maximum depth of 12.5 meters.  Then:

                 ks  =  0.2 m day'Va.SS m = 0.06 day"1

     Finally,  the  dissolved oxygen level 26 days after the  onset of
stratification is  calculated by substituting these parameters into the
oxygen uptake  equations.   The results are:

         \'n  /  402   ,     OT06 x 3.8    V,   -.0029 x 26\        "     ~
         iUL =1\3.38 m  0.06 + 0.096 - O.OOZ9/\            /
               /   0.06 x 3.8     \ /  0.0029  \ /,   -(0.06 + :096)x26^ . ft „   -3
            •  \QM + 0.096 - 0.0029/ ^0.06 + 0.096/ \                 /

    -     *°C - 8:S: o:§b fr - e'(0-096 + °-06)x 26) •2-319 ^3
         0  -  11.27  - 8.72 -  2.31  = 0.24 g nT3
                                  219

-------
     By performing many similar calculations, the dissolved oxygen-time
curves presented in Figure 3.6-4 were generated.  Each of the four curves
represents the predicted dissolved oxygen levels for one combination of
values of the BOD loading rate, k , and water column BOD decay rate con-
                                 d
stant, k,, expected in the Occoquan Reservoir.  These curves indicate
that the dissolved oxygen will be completely depleted in 25 to 100 days
after the onset of stratification.  If stratified conditions last four
months, as predicted by the thermal plots, there could be as many as
20 to 95 days when water close to the reservoir bottom contains no
oxygen.

     The wide range of calculated times required to deplete the oxygen
supply demonstrates the sensitivity of the oxygen level  to the BOD
loading and decay rates.  This can also be seen from the equations used
in the methodology presented here.  The dissolved oxygen level has an
exponential dependence on the first-order BOD decay rate constant in
the water column, k,, and the benthic layer, k,.  The decrease in the
dissolved oxygen concentration at any-time is directly proportional
to the BOD loading rate, k .   Because of the uncertainty inherent in the
                          a
use of reported or calculated values of these constants, they should be
measured in situ if quantitatively certain projections of dissolved oxygen
levels are desired.

     The accuracy of the hypolimnion dissolved oxygen model can be
determined by a comparison of predicted and observed dissolved oxygen
levels (see Table 3.6-21).  At Occoquan Dam, the hypolimnion oxygen was
depleted within 85 days of stratification in 1973.   Since the reservoir
is deepest at this location and a higher-than-average flow rate in 1973
resulted in weakly stratified conditions (see section 3.6.1)), the period
in which oxygen is depleted in this case should be one of the longest
likely to occur in the impoundment.  The model predicted a maximum of 99
days would be required to consume the dissolved oxygen.   Thus, the time re-
quired to consume the dissolved oxygen should fall  below the predicted
upper limit in nearly all cases.

                                220

-------
ro
ro I
                                                                                       'k,,w (day-)   Ka(g m- day-)
                                                                                                      0.57
                                                                                                      0.36
                                                                                                      0.57
                                                                                                      0.36
                                          0.1
                                          0.1
                                          0.3
                                          0.3
                                       High Dam (1970)
                                       Occoquan Dam (1973)
                                                                                            --•Calculated Points
                                                                                            —H Observed Points
                                       20
30       40       50        60       70

    TIME AFTER STRATIFICATION (DAYS)
                                                                                              60
90
                                                                                                                100
                                Figure 3.6-4.   Dissolved oxygen depletion versus  time in  the
                                                          Occoquan Reservoir

-------
        TABLE 3.6-21.  HYPOLIMNION DISSOLVED OXYGEN IN
                     OCCOQUAN RESERVOIR5'1
Location
Occoquan Dam









High Dam





Depth
(m) Date
7.6-16.8 4/9/73
4/18/73
4/24/73
5/4/73
5/9/73
5/16/73
5/29/73
6/19/73
6/26/73
7/3/73
4.3- 7.9 4/8/70
4/16/70
4/29/70
5/14/70
5/22/70
6/5/70
Dissolved Oxygen
(9 m-3)
9.9
9.0
8.7
8.0
6.3
4.8
2.3
1.0
0.6
0.1
10.3
10.7
6.2
8.6
4.3
0.1
a;Source:  U.S. EPA STORET System.
                            222

-------
     Interpretation of the dissolved oxygen-time data at High Dam
in 1970 presented in Table 3.6-21 is complicated by the introduction of
fresh oxygen after the onset of stratification.  Although a direct com-
parison of oxygen depletion times is not possible, the rates of oxygen
level follows curve 2 of Figure 3.6-4 very closely, while during the
second period of oxygen consumption the oxygen concentrations closely
match those of curve 1.  Since the reservoir is shallowest at High Dam
and the substantially lower than average flow rate in 1970 resulted in
strongly stratified conditions, the oxygen depletion rates in this case
should be among the highest likely to be observed in the impoundment.
Curve 1 represents the fastest decay rates predicted by the model.
Thus, the observed oxygen consumption times should be greater than the
lower limit predicted by the model in nearly all cases.

     The above agreement of the observed with the predicted limits for
the range of oxygen depletion times in Occoquan Reservoir implies that
the typical or average time must also fall'within the predicted range.
Since it was for "average" conditions that the impoundment was modeled, .
it may be concluded that the model does accurately describe the behavior
of the Occoquan Reservoir.
                                 223

-------
                           BIBLIOGRAPHY


Amy, G., e_t a]_., 1974.  Water Quality Management Planning for Urban
     Runoff.  EPA 440/9-75-004.

Camp, T. R. 1968.  Water and Its Impurities.   Reinhold  Book  Corporation.
     New York.

Delucia, R. and J. Smith.  1973.  Effluent Charges:   Is the Price Right?
     Meta Systems, Inc.  September

Dyer, K.R.  1973.  Estuaries:  A Physical  Introduction.  John Wiley and
     Sons, London.

Eckenfelder, W.W.  1970.  Water Quality Engineering  for Practicing
     Engineers.  Barnes and Noble, Inc., New York.

Grizzard, T.J., J.P. Hartigan, C.W.  Randall,  A,S. Librach,  J.I.  Kim,  -
     and II. Derewianka.  1977.  Characterizing "Runoff Pollution - Land
     Use" Relationships in Northern Virginia's Occoquan Watershed.
     Presented at the MSDGC-AMSA Workshop on Water Quality Surveys for
     208:  Data Acquisition and Interpretation of Nonpoint Runoff.
     Chicago,  IL.  April 20-22,  1977.

Haan,'C.T.  1977.  Statistical Methods in Hydrology.  The Iowa State
     University Press, Ames, Iowa 50010.

Likens, E., F. Bormann, S. Pierce, S. -Eaton,  M. Johnson.  1977.
     Biogeochemistry of a Forested Ecosystem.  Springer-Verlag.   New
     York, Heidelberg and Berlin.

Lippson, A.J.  1973.  The Chesapeake Bay in Maryland.  An Atlas  of
     Natural Resources.  The Johns Hopkins University Press.  Baltimore
     and London.

Lorenzen, C.J.  1972.  Extinction of Light in the Ocean by Phytoplank-
     ton.  J. Cons. Int. Explor. Mer.  34:262-267.

McElroy, A.D., S.Y. Chiu, J.W. Nebgen, A.  Aleti, F.W. Bennett.  1976.
     Loading Functions for Assessment of Water Pollution from Nonpoint
     Sources.  EPA-600/2-76-151.  U.S. Environmental Protection Agency,
     Washington, D.C.

Megard, R.O.   19-79.  Mechanisms and Models for the Transmission  of Light
     Rates of  Photosynthesis, and the Regulation of  Phytoplankton
     Populations.  In Phytoplankton-Environmental Interactions in
     Reservoirs (Lorenzen, M.W., ed.).  Vol.  I U.S.  Army Corps of
     Engineers, Waterways Experiment Station, Vicksburg, MS. (In Press).

Midwest Research Institute.  1979.  Personal  Communication.

Northern Virginia Planning District Commission, Regional Resources
     Division.  1979a.  Water Quality Goals for Nonpoint Pollution
     Management Program:  Background Information.  Falls Church,
     Virginia.


                                   224

-------
Northern Virginia Planning District Commission,  Regional  Resources
     Division.   1979b.   Occoquon Basin Computer  Model:   Summary of
     Calibration Results.   Falls Church,  Virginia.

Officer, C.B.  1976.   Physical  Oceanography of Estuaries (and Assoc-
     iated Coastal  Waters).   John Wiley and Sons,  Inc.,  New  York.

Ohio Environmental  Protection Agency.   1978.   State  Water Qual ity
     Management Plan,  Sandusky River Basin  Part  II.   Preliminary
     Report.

Pheiffer, T.H., L.J.  Clark,  and N.L. Lovelace.   1976.   Patuxent River
     Basin Model  Rates  Study.   Proceedings  of the  Conference on
     Environmental  Modeling and Simulation, April  19-22, 1976,
     Cincinnati,  Ohio.   U.S.  Environmental  Protection Agency.

Porcella, D.B., A.B.  Bishop.   1974.  Comprehensive Management of Phos-
     phorus Water Pollution.   Ann Arbor Science, Ann Arbor,  MI.

Pritchard, D.W.  1967.   Observations of Circulation  in  Coastal  Plain
     Estuaries.  American  Association for the Advancement of Science
     Publication No.  83* "-Estuaries".

Schwab, G.O., R.K.  Frevert,  T.W. Edminster, K.K. Barnes.  1966.  Soil
     and Water Conservation Engineering.   John Wiley and Sons, New
     York, Second Edition.

State of Maryland and the  Westinghouse (Electric)  Corporation.  1972.
     Chester River Study.   November.

State of Maryland.   1975.   Maryland Water Quality  75.   Joint Report of
     the Water Resources Administration,  Environmental  Health Adminis-
     tration and the Maryland Environmental Service.

Thompson, M.J., L..E.  Gilliland, L.K. Rosenfeld.  1979.   Light Scatter-
     ing and Extinction in a Highly Turbid  Coastal  Inlet.  Estuaries
     2(3):  164-171.   September.

Tsivoglou, E.G.,  J.R.  Wallace.   1972.   Characterization of Stream
     Reaeration Capacity.   EPA-R3-72-012.  U.S.  Environmental  Protection
     Agency, Washington, D.C.   October.

U.S. Army Corps of Engineers,  Baltimore District.   1977.  Chesapeake Bay
     Future Conditions Report,  Vol. 6, Water Quality.   December.

U.S. Army Corps of Engineers,  Buffalo District.  1978.   Water Quality
     Data, Sandusky River  Material Transport. Buffalo,  N.Y.

U.S. Department of Commerce.   1965.  Climatic Atlas  of  the United States.
     U.S. Dept. of Commerce,  Environmental  Science Services  Administration,
     Environmental  Data Service; Washington,  D.C.
                                   225

-------
U.S. Environmental  Protection Agency.   1973.   Water Quality Criteria,
     1972.  EPA-R3-73-033,  March,  1973.   Washington,  D.C.

Virginia Water Control  Board, Tidewater  Regional  Office.   1977.   Glou-
     cester STP-Fox Mill  Run/Ware  River  Special  Studies  #7-0178,
     October,  1977.  Unpublished.

Zison, S.W., K. Haven,  W.B. Mills.   1977.   Water Quality Assessment:
     A Screening Methodology for Nondesignated 208 Areas.   EPA-600/9-
     77-023.  U.S.  Environmental  Protection Agency, Athens,  GA.

Zison, S.W., W.B. Mills,  D. Deimer,  C. Chen.   1978.  Rates,  Constants,
     and Kinetics Formulations in  Surface Water Quality  Modeling.
     EPA-600/3-78-105.   U.S. Environmental  Protection Agency,  Athens, GA.
                                      226

-------
                             APPENDIX A
                   MUNICIPAL DISCHARGE SUPPLEMENT
     If effluent data is not available for sewage treatment
plants in the study area waste loads can be roughly estimated by
using survey data and reducing influent concentrations by knowledge
of the type of treatment that the facility applies.  In addition to
the raw sewage data presented in the screening manual the following
are also shown.  Table A-l is taken from Eckenfelder (1970) and
shows typical municipal sewage characteristics.  Once the influent
characteristics of raw sewage to the treatment plant have been de-
termined some effluent concentrations may be estimated by the
following method.
     Assume, for instance that the reduced nitrogen load for medium
strength municipal sewage is 40 mg i~ .   The BODK of medium strength
                        -1       '
sewage is about 200 mg a • .   The ratio of the two times the influent
BODj. (x) gives an estimated reduced nitrogen load (y) in the unknown
influent.
40 m-N £-                  1             '1
           200 mg-BOD
                             x mg-BOD, r1   = y mg-N i
                                     °
By multiplication by the stoichiometric factor for converting total
oxidizable nitrogen to biochemical oxygen demand the MBODj- for the
influent is estimated.  Then, consulting Table A-2 (Delucia and
Smith, 1973)  a treatment efficiency for the particular type of
treatment can be ascertained for BODr.  Subtracting this fraction
from unity and multiplying by the influent MBODr gives an estimate
of the effluent NBOD|- from the sewage treatment plant.  This propor
tioning procedure can be used for any of the parameters listed in
Table A-2 (BOD,-, COD, suspended solids, total phosphorus, total
nitrogen) to estimate effluent concentrations.
                                 227

-------
      TABLE A-l.  AVERAGE CHARACTERISTICS OF MUNICIPAL SEWAGE
  Characteristics              Maximum         Mean        Minimum
pH
Settleable solids, mg £~
Total solids, mg Jl
Volatile total solids, mg £
Suspended solids, mg l~
Volatile suspended solids,
mg £-1
Chemical oxygen demand,
mg A~1
Biochemical oxygen demand,
mg H ~ '
Chlorides, mg £~^
7.5
6.1
640
388
258
208
436
276
45
7.2
3.3
453 '
217
145
120
288
147
35
6.8
1.8
322
118
83
62
159
75
25
After Eckenfelder (1970)
                                 228

-------
      TABLE A-2.   MUNICIPAL  WASTEWATER  TREATMENT --  SYSTEM PERFORMANCE3"*
Effluent Concentrations (mg t~)
(2 Total Removal Efficiencies )

Scheme
Number2'1'
0, Raw
Wastewater
1

2

3
4

5

6

7


BOD5
200
(0%)
130
(35%)
40
(802)
25
(88%)
18
(91%)
18

13
(94%)
2
(99%)

COD
500
(0%)
375
(25%)
125
(75%)
100
(80%)
70
(86%)
70
(86%)
60
(88%)
15
(97%)

SS
200
(0%)
100
(50%)
30
(85%)
12
(94%)
7
(96%)
7
(96%)
1
•(99.5%)
1
(99.5%)
pr
(mgP l~ )
10
(02)
9
(10%)
7.5
(25%)
(302)
1
(90%)
• y
(902)
1
(902)
1
(902)
NT,
(mgN J,"1)
40
(0%)
32
(20%)
26
(35%)
24
(40S)
22
(«2)
4
(90%)
3
(92%)
2
(95%)
a;  Influent  is  assumed to be raw-medium strength domestic sewage.  See scheme
    number 0  for characteristics.
b)  Efficiencies for wastev;ater treatment are  for the approximate concentration
    range, as measured by B00r, of 100 <_ 800-  <_ 400  (mg l~^).
c)  Scheme No.       Process
      0	  Ho treatment
      1	  Pn>,ary
      2	  Primary, plus Activated Sludge  (Secondary  Treatment)
      3	  Primary, Activated Sludge, plus  Polishing  Filter  (High Efficiency
                    or Super Secondary)
      4	  Primary, Activated Sludge, Polishing  Filter,  plus  Phosphorus^
                    Removal and Recarbonation
      5	  Primary, Activated Sludge, Polishing  Filter,  Phosphorus
                    Removal, plus Nitrocen Stripping and  Recarbonation
      6	  Primary, Activated Sludge, Polishing  Filter,  Phosphorus
                    Removal, Nitrogen Stripping Recarbonation,  plus Pressure Filtration
      7	  Primary, Activated Sludge, Polishing  Filter,  Phosphorus
                    Removal, nitrogen Stripping Recarbonation,  Pressure Filtration,
                    pi us Activated Carbon Adsorption
                                                 229

-------