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 LaboratoryAthens 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 TreatmentSystem 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 AssessmentA 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
OO 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
XX 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
------- |