&EPA
United States
Environmental Protection
Agency
American Society of Civil Engineers
Urban Storm water
BMP Performance Monitoring
A Guidance Manual for Meeting the National
Stormwater BMP Database Requirements
April 2002
-------
Urban Stormwater BMP Performance
Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
Prepared by
GeoSyntec Consultants
Urban Drainage and Flood Control District
and
Urban Water Resources Research Council (UWRRC) of ASCE
In cooperation with
Office of Water (4303T)
US Environmental Protection Agency
Washington, DC 20460
April 2002
EPA-821-B-02-001
vvEPA
United States
_ . .-..,,. . Environmental Protection
American Society of Civil Engineers Aaencv
-------
TABLE OF CONTENTS
List of Tables viii
List of Figures x
Acknowledgements and Disclaimer xi
1 INTRODUCTION 1
1.1 Scope 1
1.1.1 State of the Practice 2
1.1.2 The Need for Guidance 2
1.1.3 National Stormwater Best Management Practices Database 2
1.2 Format and Content of This Document 2
2 BMP MONITORING OVERVIEW 4
2.1 Context of BMP Monitoring in the Regulatory Environment 4
2.2 BMP Monitoring Goals 5
2.3 Physical and Chemical Characteristics of Stormwater Runoff 7
2.4 Stormwater Quality Monitoring Challenges 8
2.5 Complexities Specific to BMP Monitoring 9
2.5.1 Considerations for Evaluating BMP Effectiveness 10
Load Versus Water Quality Status Monitoring 10
Consideration of Parameters for Monitoring 12
2.6 BMP Types and Implications for Calculation of Efficiency 13
2.7 Relationship Between Monitoring Study Objectives and Data Analysis 14
2.8 Physical Layout and Its Effect on Efficiency and Its Measure 15
2.9 Relevant Period of Impact 16
2.9.1 Concentrations, Loads, and Event Mean Concentrations 17
2.9.1.1 Concentrations 17
2.9.1.2 Loads 17
2.9.1.3 Event Mean Concentrations 18
2.9.2 Measures of BMP Efficiency 18
2.9.2.1 Historical Approaches 21
Efficiency Ratio 21
Definition 21
Assumptions 23
Comments 23
Example 24
Summation of Loads 24
Definition 24
Assumptions 24
Comments 25
Example 25
Regression of Loads (ROL) 25
Definition 25
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
ii
-------
Assumptions 26
Comments 27
Mean Concentration 29
Definition 29
Assumptions 29
Comments 30
Efficiency of Individual Storm Loads 30
Definition 30
Assumptions 31
Comments 31
Summary and Comparison of Historical Methods 32
2.9.2.2 Other Methods and Techniques 32
"Irreducible Concentration" and "Achievable Efficiency" 32
Percent Removal Relative to Water Quality Standards 36
"Lines of Comparative Performanceฉ" 37
Multi-Variate and Non-Linear Models 40
2.9.2.3 Recommended Method 40
Effluent Probability Method 40
2.9.2.4 Reference Watershed Methods 43
2.9.3 BMPs and BMP Systems 44
3 DEVELOPING A BMP MONITORING PROGRAM 45
3.1 Phase I - Determine Objectives and Scope of BMP Water Quality Monitoring
Program 46
3.1.1 Monitoring and Literature Review to Assess BMP Performance 47
3.1.2 Monitoring to Assess Compliance with Surface Water quality criteria 49
3.1.3 Criteria for the Protection of Aquatic/Marine Life 49
3.1.4 Human Health 50
3.1.5 Application of Water quality criteria to Stormwater 50
3.1.6 Groundwater and Sediment Standards 51
3.1.7 Scope of Work for BMP Monitoring Program 51
3.1.8 Information Needs to Meet Established Goals of BMP Monitoring 55
3.2 Phase II - Develop BMP Monitoring Plan 56
3.2.1 Recommendation and Discussion of Monitoring Locations 56
Integration of BMP Monitoring into a Municipal Monitoring Program 57
Sampling from a Well Mixed Location 58
3.2.1.1 Upstream 59
3.2.1.2 Downstream 60
3.2.1.3 Intermediate Locations 60
3.2.1.4 Rainfall 61
Site Proximity 61
Number of Gauges 62
3.2.1.5 Groundwater 62
3.2.1.6 Sediment Sampling 63
3.2.1.7 Dry Deposition 63
3.2.1.8 Modeling Methods 64
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
iii
-------
Estimates of Water Quality Parameters 64
Estimates of Flow 67
Estimates of Rainfall 67
3.2.2 Recommendation and Discussion of Monitoring Frequency 68
3.2.2.1 Statistical Underpinnings of Study Design 68
3.2.2.2 Factors Affecting Study Design 69
Number of Samples 69
Determining the Number of Observations Needed 70
3.2.3 Recommendation and Discussion of Water Quality Parameters and Analytical
Methods 76
3.2.3.1 Selecting Parameters 76
3.2.3.2 Dissolved vs. Total Metals 79
3.2.3.3 Measurements of Sediment Concentration 79
3.2.3.4 Analytical Methods 81
3.2.4 Recommendation and Discussion of Monitoring Equipment and Methods 83
3.2.4.1 Equipment 83
Data Loggers 83
Power Requirements 87
Flow 89
Volume-Based Methods 91
Stage-Based Methods 91
Manning's Equation 92
Other Empirical Stage-Flow Relationships 93
Stage Based Method Using Weirs and Flumes 93
Stage-Based Variable Gate Meters 94
Velocity-Based Methods 94
Tracer Dilution Methods 95
Constant Injection Rate Tracer Dilution Studies 95
Total Recovery Tracer Dilution Studies 95
Pump Discharge Method 95
3.2.4.2 Automatic Sampling Techniques 96
Selection of Primary Flow Measurement Device 96
Types of Primary Flow Measurement Devices 96
Weirs 97
Flumes 97
Considerations for Selection of Primary Flow Measurement Device 99
Range of Flows 99
Flow Rate 100
Accuracy 100
Cost 100
Head Loss and Flow Characteristics 101
Sediment and Debris 101
Construction Requirements 101
Selection of Secondary Flow Measurement Device 102
Float Gauge 103
Bubbler Tube 103
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
iv
-------
Ultrasonic Depth Sensor 104
Pressure Probe 105
Ultrasonic "Uplooking" 106
Radar/Microwave 106
Equipment for Measuring Velocity 107
Methods Suitable for Calibration 107
Tracer Studies 108
Rotating-Element Current Meters 108
Pressure Sensors 108
Acoustical Sensors 108
Float-and-Stopwatch Method 109
Deflection (or Drag-Body) Method 109
Methods Most Suitable for Continuous Velocity Monitoring 109
Ultrasonic (Doppler) Sensors 109
Electromagnetic Sensors 110
Acoustic Path Ill
Water Quality Sample Collection Techniques Ill
Grab Samples Ill
Composite Samples 112
Automatic Sampling 114
Automatic Sampling Equipment 115
Overland Flow Sampler 118
In-situ Water Quality Devices, Existing Technology 119
In-situ Water Quality Devices, Future Technologies 121
Ion-Selective Electrodes 121
On-Line Water Quality Analyzers 121
Particle Size Analyzers 122
In-situ Filtration and Extraction System 123
Remote Communications with Automatic Equipment 123
Manual Sampling 124
Manual Grab Sampling Equipment 125
Manual Composite Sampling Equipment 125
3.2.4.3 Error Analysis and Measurement Accuracy 126
3.2.5 Recommendation and Discussion of Storm Criteria 127
3.2.5.1 Storm Characteristics 127
3.2.6 Recommendation and Discussion of QA/QC 129
3.2.6.1 Sampling Methods 132
Contamination/Blanks 133
Reconnaissance and Preparations 134
Site Visits 134
Laboratory Coordination 134
Sample Containers/Preservation/Holding Times 135
Recommended Field QA/QC Procedures 135
Field Blanks 135
Field Duplicate Samples 135
Field Sample Volumes 136
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
V
-------
Chain of Custody 136
Recommended Laboratory QA/QC Procedures 136
Method Blanks 136
Laboratory Duplicates 136
Matrix Spike and Spike Duplicates 136
External Reference Standards 136
3.2.7 Recommendations for Data Management 137
3.2.7.1 Database Requirements 137
Analysis of Database Links 138
Analysis of Outlying Records 138
Sample Comparisons Between Original Documents and Final Data Set 139
Digital Conversion of Data 139
Double Data Entry and Optical Character Recognition 139
3.3 Phase III - Implementation of Monitoring Plan 139
3.3.1 Training of Personnel 139
3.3.2 Installation of Equipment 140
3.3.3 Testing and Calibrating Equipment 141
3.3.4 Conducting Monitoring 141
3.3.5 Coordinate Laboratory Analysis 143
3.4 Phase IV - Evaluation and Reporting of Results 144
3.4.1 Validate Data 144
3.4.2 Evaluate Results 144
3.4.2.1 Preliminary Data Evaluation 145
3.4.2.2 Definitive Evaluations 145
3.4.3 Report Results 146
3.4.3.1 National Stormwater BMP Database Requirements 147
3.4.3.2 Standard Format Examples 156
General Test Site Information 156
Watershed Information 159
Structural BMP Information 164
Non-Structural BMP Information 166
Detention Basin Design Data 169
Retention Pond Design Data 172
Percolation Trench and Dry Well Design Data 175
Media Filter Design Data 178
Grass Filter Strip Design Data 181
Wetland Channel and Swale Design Data 183
Porous Pavement Design Data 186
Infiltration Basin Design Data 189
Hydrodynamic Device Design Data 192
Wetland Basin Design Data 194
Monitoring Station Information 198
Precipitation Data 201
Flow Data 203
Water Quality Data 205
3.4.3.3 On-line Information 207
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
vi
-------
References 208
Index 214
APPENDIX A
APPENDIX B
APPENDIX C
APPENDIX D
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
vii
-------
List of Tables
Page
Table 2.1: Objectives of BMP implementation projects and the ability of comprehensive
water quality monitoring studies to provide information useful for
determining performance and effectiveness 6
Table 2.2: Examples of water quality parameters and relevant monitoring
period 17
Table 2.3: Summary of historical, alternative, and recommended methods for
BMP water quality monitoring data analysis 20
Table 2.4: Example of ER method results for TSS in the Tampa Office Pond 24
Table 2.5: Example of SOL method results for TSS in the Tampa Office Pond 25
Table 2.6: Example of ROL method results for TSS in the Tampa Office Pond 27
Table 2.7: Example of Individual Storm Loads Method results for TSS in the
Tampa Office Pond 32
Table 2.8: Comparison of BMP efficiency methods 32
Table 2.9: "Irreducible concentrations" as reported by Scheuler, 2000 33
Table 2.10: Example TSS results for typical ER Method 33
Table 2.11: Example TSS results for demonstration of Relative Efficiency
approach 35
Table 2.12: Example of percent removal relative to receiving water quality limits
approach 36
Table 3.1: Typical urban stormwater runoff constituents and recommended detection
limits 79
Table 3.2: Flow measurement methods 91
Table 3.3: Equipment for measuring depth of flow 103
Table 3.4: Velocity measurement methods suitable for calibration 108
Table 3.5: National Stormwater BMP Database requirements for all BMPs 149
Table 3.6: National Stormwater BMP Database requirements for structural
BMPs 150
Table 3.7: National Stormwater BMP Database requirements for
Non-structural BMPs 150
Table 3.8: National Stormwater BMP Database requirements for individual
structural BMPs 151
Table 3.9: National Stormwater BMP Database requirements for
non-structural BMPs and structural BMPs that are based on
minimizing directly connected impervious areas 156
Table 3.10: National Stormwater BMP Database requirements for structural
BMPs that are based on minimizing directly connected
impervious areas 157
Table 3.11: General test site form data element descriptions 158
Table 3.12: Watershed form data elements description 160
Table 3.13: Structural BMP form data elements description 165
Table 3.14: Non-structural BMP form data elements description 167
Table 3.15: Detention Basin design form data elements list 170
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
Vlll
-------
Table 3.16: Retention Pond design form data elements list 173
Table 3.17: Percolation trench and dry well design form data
elements list 176
Table 3.18: Media filter design form data elements list 179
Table 3.19: Grass filter strip form data elements list 182
Table 3.20: Wetland channel and swale form data elements list 184
Table 3.21: Porous pavement form data elements 187
Table 3.22: Infiltration basin form data elements list 190
Table 3.23: Hydrodynamic device form data elements 193
Table 3.24: Wetland basin form data elements list 195
Table 3.25: Monitoring station form data elements 199
Table 3.26: Precipitation form data elements 202
Table 3.27: Flow form data elements 204
Table 3.28: Water quality form data elements 206
Table A. 1: Example of inputs for estimation of errors in flow measurement
Devices A-5
Table A.2: Summary of examples demonstrating the propagation of errors in
flow measurement A-7
Table D.I: Relationships of log-normal distributions D-l
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
IX
-------
List of Figures
Page
Figure 2.1 ROL plot for use in calculating efficiency for TSS
using the Tampa Office Pond (1990) 28
Figure 2.2 ROL plot for use in calculating efficiency for TSS
using the Tampa Office Pond (1993-1994) 28
Figure 2.3 ROL plot for use in calculating efficiency for TSS using the Tampa
Office Pond (1994-1995) 29
Figure 2.4 Removal Efficiency (ER Method) of TSS as a function of influent
concentration 38
Figure 2.5 Removal Efficiency (ER Method) of total phosphorous as a
function of influent concentration 38
Figure 2.6 Removal Efficiency (ER Method) of total zinc as a function of influent
concentration 39
Figure 2.7 Percent removal as a function of influent concentration for randomly
generated, normally distributed influent and effluent concentrations 39
Figure 2.8 Probability plot for Suspended Solids 42
Figure 2.9 Probability plot for Total Dissolved Solids 42
Figure 2.10 Probability plot for Chemical Oxygen Demand 42
Figure 3.1 Nomograph relating coefficient of variation of a samples set to the
allowable error in the estimate of the population mean 71
Figure 3.2 Number of samples required using a paired sampling approach to observe a
statistically significant percent difference in mean concentration as a function
of the coefficient of variation
(power of 80% and confidence of 95%) 75
Figure 3.3 Datalogger with weatherproof housing 84
Figure 3.4 Data logger without housing 85
Figure 3.5 Data logger summary 88
Figure 3.6 Parshall flume 98
Figure 3.7 H-flume 98
Figure 3.8 Bubbler flow meter 104
Figure 3.9 Ultrasonic-depth sensor module 105
Figure 3.10 Pressure transducers 106
Figure 3.11 Area velocity sensors module 110
Figure 3.12 Automatic sampler 116
Figure 3.13 VOC sampler 117
Appendix B Figures: Number of samples required for various powers, confidence
intervals, and percent differences B-l
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
Acknowledgements and Disclaimer
The authors, Eric Strecker, P.E. of GeoSyntec Consultants, Ben Urbonas, P.E. Urban
Drainage and Flood Control District, Denver, Marcus Quigley, P.E., Jim Howell, and
Todd Hesse of GeoSyntec Consultants would like to thank Jesse Pritts, P.E. and Eric
Strassler of the Environmental Protection Agency and Tom McLane and Lorena Diaz of
the American Society of Civil Engineers (ASCE) for their support for and participation in
the ASCE/EPA National Stormwater Best Management Practices Database Project and the
development of this guidance. The authors would also like to thank the following
members of ASCE's Urban Water Resources Research Council for their thorough review
and contributions to this guidance:
Robert Pitt, P.E., Ph.D. (University of Alabama, Birmingham)
Eugene Driscoll, P.E.
Roger Bannerman, P.E. (Wisconsin Department of Natural Resources)
Shaw Yu, P.E., Ph.D. (University of Virginia)
Betty Rushton (Southwest Florida Water Management District)
Richard Field (EPA), P.E.
Jonathan Jones, P.E. (Wright Water Engineers)
Jane Clary (Wright Water Engineers)
Tom Langan (Wright Water Engineers)
Sections of this manual were developed by the authors concurrently with the Federal
Highway Administration's (FHWA) "Guidance Manual For Monitoring Highway Runoff
Water Quality." Although the focus of the FHWA manual is on highway runoff
monitoring, much of the information on equipment selection, use, and installation is
applicable to best management practice monitoring and thus was adapted for this
guidance.
In addition, portions of this document were adapted from work originally conducted for
the Washington State Department of Ecology's (DOE) November 1995, "Stormwater
Monitoring Guidance Manual" by an author of this document (Eric Strecker) and Mike
Milne (Brown and Caldwell), Terry Cook (URS Group, Inc.), Gail Boyd (URS Group,
Inc.), Krista Reininga (URS Group, Inc.), and Lynn Krasnow. The thoroughness and
specific insight provided in the DOE Manual were useful in assembling this guidance.
The authors would also like to thank Joan LeBlanc, and Kathy Staffier (GeoSyntec
Consultants) for editorial review and edits of the final document.
Disclaimer:
Mention of trade names or commercial products does not constitute endorsement by EPA
or ASCE, or recommendation for use.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
xi
-------
1 Introduction
1.1 Scope
Existing guidance is available for assessing the effectiveness of stormwater best
management practices (EPA 1997; FHWA 2000). However, few existing documents
provide targeted practical assistance in conducting and reporting data from a water quality
based monitoring program that results in data that are useful for assessing BMP
effectiveness on a broader scale.
This guidance has been developed by integrating experience gleaned from field
monitoring activities conducted by members of ASCE's Urban Water Resource Research
Council and through the development of the ASCE/EPA National Stormwater Best
Management Practices Database. The manual is intended to help achieve stormwater BMP
monitoring project goals through the collection of more useful and representative rainfall,
flow, and water quality information. Many of the recommended protocols (particularly
those for reporting monitoring, watershed, and design information) are directly related to
requirements of the National Stormwater Best Management Practices Database.
This manual is intended to improve the state of the practice by providing a recommended
set of protocols and standards for collecting, storing, analyzing, and reporting BMP
monitoring data that will lead to better understanding of the function, efficiency, and
design of urban stormwater BMPs. This manual provides insight into and guidance for
strategies, approaches, and techniques that are appropriate and useful for monitoring
BMPs.
This document addresses methods that were in use at the time it was written. As the state
of the practice and the design of monitoring equipment progress, new monitoring
approaches and techniques, more sensitive devices, and equipment based on new
technologies will likely be employed. Although the technology may change somewhat
from that described herein, most of the basic flow and water quality monitoring methods
discussed in this document have a long history of use and will most likely remain viable
even as new and different technologies emerge.
This manual focuses primarily on the collection, reporting, and analysis of water quantity
and quality measurements at the heart of quantitative BMP efficiency projects. It does not
address, in detail, sediment sampling methods and techniques, biological assessment,
monitoring of receiving waters, monitoring of groundwater, streambank erosion, channel
instability, channel morphology, or other activities that in many circumstances may be as,
or more, useful for measuring and monitoring water quality for assessing BMP efficiency.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
1
-------
1.1.1 State of the Practice
Many studies have assessed the ability of stormwater treatment BMPs (e.g., wet ponds,
grass swales, stormwater wetlands, sand filters, dry detention, etc.) to reduce pollutant
concentrations and loadings in stormwater. Although some of these monitoring projects
conducted to date have done an excellent job of describing the effectiveness of specific
BMPs and BMP systems, there is a lack of standards and protocols for conducting BMP
assessment and monitoring work. These problems become readily apparent for persons
seeking to summarize the information gathered from a number of individual BMP
evaluations. Inconsistent study methods, lack of associated design information, and
reporting protocols make wide-scale assessments difficult, if not impossible. (Strecker et
al. 2001; Urbonas 1998) For example, individual studies often include the analysis of
different constituents and utilize different methods for data collection and analysis, as well
as report varying degrees of information on BMP design and flow characteristics. The
differences in monitoring strategies and data evaluation alone contribute significantly to
the range of BMP "efficiency" that has been reported in literature to date.
1.1.2 The Need for Guidance
Municipal separate storm sewer system owners and operators need to identify effective
BMPs for improving stormwater runoff water quality. Because of the current state of the
practice, however, very little sound scientific data are available for making decisions about
which structural and non-structural management practices function most effectively under
what conditions; and, within a specific category of BMPs, to what degree design and
environmental static and state variables directly affect BMP efficiency. This guidance
addresses this need by helping to establish a standard basis for collecting water quality,
flow, and precipitation data as part of a BMP monitoring program. The collection, storage,
and analysis of this data will ultimately improve BMP selection and design.
1.1.3 National Stormwater Best Management Practices Database
The National Stormwater BMP Database (Database) serves two key purposes: (1) to
define a standard set of data reporting protocols for use with BMP monitoring efforts; and
(2) to assemble and summarize historical and future BMP study data in a standardized
format. The software consists of a data entry module for reporting data on new BMP
studies and a search engine module to allow users to retrieve data. The Database is a user-
friendly, menu-driven software program developed in a run-time version of Microsoftฎ
Access 97 and Access 2000. The software has been distributed on CD-ROM and is now
also accessible via the Internet at www.bmpdatabase.org.
1.2 Format and Content of This Document
This document is broken down into two main sections following this introduction:
Section 2 provides an overview of BMP monitoring. Discussion is provided on the context
of BMP monitoring, difficulties in assessing BMP performance, and understanding the
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
2
-------
relationship between BMP study design and the attainment of monitoring program goals.
Useful analysis of data collected from BMP monitoring studies is essential for
understanding and comparing BMP monitoring study results. A summary of historical
and recommended approaches for data analysis is provided in this section to elucidate the
relationship between the details and subtleties of each analysis approach and the
assessment of performance.
Section 3 discusses the specifics of developing a monitoring program, selecting
monitoring methods and equipment, installing and using equipment, implementing
sampling approaches and techniques, and reporting information consistent with the
National Stormwater Best Management Practices Database.
In addition, four appendices have been included in this guidance document. The first
appendix describes methods for calculating expected errors in field measurements. The
second provides detailed information about the number of samples required to obtain
statically significant monitoring data. The third appendix includes charts for estimating
the number of samples required to observe a statically significant difference between two
populations for a various levels of confidence and power. The final appendix is a table for
estimating arithmetic descriptive statistics based on descriptive statistics of log-
transformed data.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
2 BMP Monitoring Overview
This section provides an overview of BMP monitoring program context and execution,
including a discussion of approaches used for quantifying BMP efficiency.
2.1 Context of BMP Monitoring in the Regulatory Environment
BMP monitoring is conducted by researchers, public entities, and private companies for
meeting both regulatory and non-regulatory needs. This section briefly discusses some of
the regulatory programs that drive BMP monitoring programs.
A number of environmental laws exist for implementation of stormwater and BMP
monitoring programs including:
The Clean Water Act (CWA) of 1972:
Section 208 of 1972 CWA requires every state to establish effective BMPs to
control nonpoint source pollution. The 1987 Water Quality Act (WQA) added
section 402(p) to the CWA, which requires that urban and industrial stormwater
be controlled through the National Pollutant Discharge Elimination System
(NPDES) permit program.
Section 303(d) of WQA requires the states to list those water bodies that are not
attaining water quality standards including designated uses and identification of
relative priorities among the impaired water bodies. States must also develop
TMDLs (Total Maximum Daily Loads) that quantify the pollutant load or the
impairing pollutants that will bring the waterbody back into attainment.
The Endangered Species Act:
The Endangered Species Act of 1973 protects animal and plant species currently
in danger of extinction (endangered) and those that may become endangered in
the foreseeable future (threatened). It provides for the conservation of ecosystems
upon which threatened and endangered species of fish, wildlife, and plants
depend, both through Federal action and by encouraging the establishment of state
programs.
Coastal Zone Act Reauthorization Amendments (CZARA) of 1990:
CZARA was passed to help address nonpoint source pollution in coastal waters.
Each state with an approved coastal zone management program must develop and
submit to the EPA and National Oceanic and Atmospheric Administration
(NOAA) a Coastal Nonpoint Pollution Control Program (CNPCP), which
provides for the implementation of the most economically achievable
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
management measures and BMPs to control the addition of pollutants to coastal
waters.
CZARA does not specifically require that states monitor implementation of
management measures and BMPs. They must, however, provide technical
assistance to local governments and the public in the implementation of the
management measures and BMPs, which may include assistance to predict and
assess the effectiveness of such measures.
CZARA also states that the EPA and NOAA shall provide technical assistance to
the states in developing and implementing the CNPCP, including methods to
predict and assess the effects of coastal land use management measures on coastal
water quality and designated uses:
1. Protection of stream and water body designated use (meet fishable and
swimmable goals)
2. Antidegradation policies designated to protect water quality when the
water quality already is higher than existing standards
3. Other state, county, and local regulations or ordinances
As regulations and the application and enforcement thereof change over time, details
about the above environmental laws and their implications for specific sites and
watersheds are best obtained from current EPA, state, county, and local resources.
2.2 BMP Monitoring Goals
BMP monitoring projects are initiated to address a broad range of programmatic,
management, regulatory, and research goals. Goal attainment is often focused on the
achievement of water quality objectives downstream of the BMP. However, there are
many other objectives that have been established as part of BMP implementation projects
that cannot be measured using a water quality monitoring approach alone. Table 2.1
below describes the relationship between BMP implementation objectives and the ability
of water quality monitoring studies to address the attainment of these objectives.
Studies directed at addressing the efficiency of BMPs in attaining water quality goals are
usually conducted to obtain information to help answer one or more of the following
questions:
What degree of pollution control or effluent quality does the BMP provide under
normal conditions?
How does this efficiency vary from pollutant to pollutant?
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
How does this normal efficiency vary with large or small storm events?
How does this normal efficiency vary with rainfall intensity?
How do design variables affect efficiency?
How does efficiency vary with different operational and/or maintenance approaches?
Does efficiency improve, decay, or remain stable over time?
How does this BMP's efficiency compare with the efficiency of other BMPs?
The ability of a specific BMP monitoring program to answer these questions and
ultimately address the desire to measure goal attainment is a vital planning stage
component of setting up a meaningful BMP monitoring program.
Table 2.1: Objectives of BMP implementation projects and the ability of
comprehensive water quality monitoring studies to provide information useful for
determining performance and effectiveness
Category Goals of BMP Projects
*ฐ Performance and Effectiveness
Hydraulics Improve flow characteristics upstream and/or downstream
of BMP
Hydrology Flood mitigation, improve runoff characteristics (peak /
shaving)
Water Quality Reduce downstream pollutant loads and concentrations of /
pollutants
Improve/minimize downstream temperature impact */
Achieves desired pollutant concentration in outflow */
ป Removal of litter and debris -
Toxicity Reduce acute toxicity of runoff ^
Reduce chronic toxicity of runoff /
Regulatory
Implementation
Feasibility
Cost
Aesthetic
Compliance with NPDES permit
Meet local, state, or federal water quality criteria
For non-structural BMPs, ability to function within
management and oversight structure
Capital, operation, and maintenance costs
Improve appearance of site
S1
-
-
-
Maintenance Operate within maintenance, and repair schedule and
requirements
Ability of system to be retrofit, modified or expanded -
Longevity ป Long-term functionality */_
Resources Improve downstream aquatic environment/erosion control
Improve wildlife habitat
ป Multiple use functionality -
Safety, Risk and Function without significant risk or liability
Liability Ability to function with minimal environmental risk
downstream
Public Information is available to clarify public understanding of ^
Perception runoff quality, quantity and impacts on receiving waters
^ can be evaluated using water quality monitoring as primary source of information
/l can be evaluated using water quality monitoring as the primary source of information combined with a secondary source of
comparative data
- cannot be directly evaluated using water quality monitoring, but in some cases may be supported by work associated with collecting
water quality information (i.e., detailed flow data)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
2.3 Physical and Chemical Characteristics of Stormwater Runoff
In this guidance manual, the term "stormwater" refers to more than just storm-driven
surface runoff. Here the term is expanded to cover water and other substances that are
transported through stormwater conveyance systems during, after, and between storm
events. In addition to the runoff from rainfall or snowmelt, a typical stormwater sample
may contain materials that were dumped, leaked, spilled, or otherwise discharged into the
conveyance system. The sample may also contain materials that settled out in the system
toward the end of previous storms and were flushed out by high flows during the event
being sampled. Stormwater also can include dry weather flows such as pavement
washing, pavement cutting wash water, or irrigation. Loads from dry weather flows, in
some cases, can greatly exceed wet weather loads over the course of a year and must be
taken into account.
Stormwater quality tends to be extremely variable (EPA 1983; Driscoll et al. 1990). The
intensity (volume or mass of precipitation per unit time) of rainfall often varies
irregularly and dramatically. These variations in rainfall intensity affect runoff rate,
pollutant washoff rate, in-channel flow rate, pollutant transport, sediment deposition and
re-suspension, channel scour, and numerous other phenomena that collectively determine
the pollutant concentrations, pollutant forms, and stormwater flow rate observed at a
given monitoring location at any given moment. In addition, the transitory and
unpredictable nature of many pollutant sources and release mechanisms (e.g., spills,
leaks, dumping, construction activity, landscape irrigation runoff, vehicle washing
runoff), and differences in the time interval between storm events also contribute to
inter-storm variability. As a result, pollutant concentrations and other stormwater
characteristics at a given location should be expected to fluctuate greatly during a single
storm runoff event and from event to event.
In addition, the complexity of introducing a structural management practice can greatly
affect hydraulics and constituent concentrations in complex ways. For example, flows
from detention facilities are often not confined only to the period of wet weather, as drain
time can be significant.
Numerous studies conducted during the late 1970s and early 1980s show that stormwater
runoff from urban and industrial areas are a potentially significant source of pollution
(EPA 1983; Driscoll et al. 1990). As a result, federal, state and local regulations have
been promulgated to address stormwater quality (see Section 2.1 above).
The impacts of hydrologic and hydraulic (physical as opposed to chemical) changes in
watersheds are increasingly being recognized as significant contributors to receiving
waters not meeting beneficial criteria. These impacts include stream channel changes
(erosion, sedimentation, temperature changes) as well as wetland water level fluctuations.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
7
-------
2.4 Stormwater Quality Monitoring Challenges
Information collected on the efficiency and design of BMPs serves a variety of goals and
objectives as discussed in Section 2.2. The principal challenge facing persons
implementing BMP monitoring programs is the great temporal and spatial variability of
Stormwater flows and pollutant concentrations. Stormwater quality at a given location
varies greatly both between storms and during a single storm event, and thus a small
number of samples are not likely to provide a reliable indication of Stormwater quality at
a given site or the effect of a given BMP. Therefore, collection of numerous samples is
generally needed in order to accurately characterize Stormwater quality at a site and BMP
efficiency (see Section 3.2.2).
Collecting enough Stormwater samples to answer with a high level of statistical
confidence many of the common questions regarding BMP efficiency is generally
expensive and time-consuming. A poorly-designed monitoring program could lead to
erroneous conclusions and poor management decisions, resulting in misdirected or
wasted resources (e.g., staff time, funds, credibility, and political support). Therefore,
before one begins a BMP monitoring program, it is critical to clearly identify and
prioritize the goals of the project, determine the type and quality of information needed to
attain those goals, and then compare this list of needs to the resources available for
monitoring. If the available resources cannot support the scale of monitoring needed to
provide the quality of information deemed necessary, then consider the following options
to obtain useful results within your resource limitations (e.g., funds, personnel, time):
A phased approach wherein you address only a subset of the overall geographic area,
or only the most important Stormwater questions.
Limiting the number of constituents evaluated as an alternative to reducing the
number of samples collected.
Utilizing available data from other locations to support decision-making.
The key question should be: "Will the information provided from the monitoring program
I am considering (and would be able to implement) significantly improve my
understanding of the effectiveness of the BMP being monitored?" If the answer is no,
reconsider the monitoring program.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
2.5 Complexities Specific to BMP Monitoring
Monitoring BMPs introduces a number of specific difficulties into the already complex
task of monitoring storm water runoff water quality.
In many ways a structural BMP system is best viewed as an environmental unit process
with a large number of static and state variables affecting functionality of the process.
For example, static variables that can directly affect BMP system function include:
BMP design (e.g., length, width, height, storage volume, outlet design, upstream
bypass, model number, etc.)
Geographical location.
Watershed size.
Percent imperviousness.
Vegetative canopy.
Soil type.
Watershed slopes.
Compaction of soils.
State variables that directly affect BMP function may include:
Rainfall intensity.
Flow rate.
Season.
Vegetation.
Upstream non-structural controls.
Inter-event timing.
Settings for control structures such as gates, valves, and pumps.
Maintenance of the BMP.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
The inconsistent use of language in reporting BMP information can compound the
difficult task of assessing physically complex systems. In order to provide a consistent
context for discussion of monitoring approaches in this guidance, the following
definitions are provided:
Best Management Practice (BMP) - A device, practice, or method for removing,
reducing, retarding, or preventing targeted stormwater runoff constituents, pollutants,
and contaminants from reaching receiving waters.
BMP System - A BMP system includes the BMP and any related bypass or overflow.
For example, the efficiency (see below) can be determined for an offline retention
(Wet) Pond either by itself (as a BMP) or for the BMP system (BMP including
bypass).
Performance - measure of how well a BMP meets its goals for stormwater that the
BMP is designed to treat.
Effectiveness - measure of how well a BMP system meets its goals in relation to all
stormwater flows.
Efficiency - measure of how well a BMP or BMP system removes or controls
pollutants.
Researchers often want to determine efficiency of BMPs and BMP systems and to
elucidate relationships between design and efficiency. Efficiency has typically been
quantified by "percent removal". As is discussed in the following sections, "percent
removal" alone is not a valid measure of the functional efficiency of a BMP (Strecker et
al. 2001). As a result the definition of "efficiency" in this manual can mean any measure
of how well a BMP or BMP system removes or controls pollutants and is not restricted
by the historical use of the term referring to "percent removal."
2.5.1 Considerations for Evaluating BMP Effectiveness
Load Versus Water Quality Status Monitoring
The choice between monitoring either (a) the status or condition of the water resource or
(b) the pollutant load and event mean concentrations discharged to the water resource
should be made with care (Coffey and Smolen 1990). Monitoring of loads and event
mean concentrations is focused on obtaining quantitative information about the amount
of pollutants transported to the receiving water from overland, channel and pipe,
tributary, or groundwater flow. Load and concentration monitoring can be used to
evaluate pollutant export at a stormwater BMP.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
10
-------
Water Quality Status Monitoring
Water quality status can be evaluated in a number of ways, including:
Evaluating "designated use" attainment1.
Evaluating Water Quality Standards violations.
Assessing ecological integrity.
Monitoring an indicator parameter.
Monitoring water quality status includes measuring a physical attribute, chemical
concentration, or biological condition, and may be used to assess baseline conditions,
trends, or the impact of treatment on the receiving water. Monitoring water quality status
may be the most effective method to evaluate the impact of the management measure
implemented, but sensitivity may be low (Coffey and Smolen 1990). When the
probability of detecting a trend in water quality status is low, load monitoring may be
necessary.
When deciding between measuring load or water quality status (i.e., it is not clear
whether abatement can be detected in the receiving resource), a pollutant budget may
help to make the decision (Coffey and Smolen 1990). The budget should account for
mass balance of pollutant input by source, all output, and changes in storage. Sources of
error in the budget should also be evaluated (EPA 1993a).
Pollutant Load and Event Mean Concentration Monitoring
Load monitoring requires considerable effort and should include the protocols that are the
primary intent of this document. Because of potentially high variability of discharge and
pollutant concentrations in watersheds impacted by both point and non-point sources,
collecting accurate and sufficient data from a significant number of storm events and base
flows over a range of conditions (e.g., season, land cover) is important. This manual
describes several methods for collecting and analyzing meaningful pollutant loading and
event concentration data. Most of these methods are also applicable to water quality
status monitoring where specific chemical concentrations must be monitored.
Monitoring for designated use attainment or standards violations should focus on those
parameters or criteria specified in state water quality standards. Where the monitoring
objective includes relating improvements in water quality to the pollution control
activities, it is important that the parameters monitored are connected to the management
See Clean Water Act, Section 303(c)(2)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
11
-------
measures implemented. For violations of standards, the choice of variable is specified by
the state water quality standard (EPA 1993a).
Consideration of Parameters for Monitoring
Many studies have been conducted to assess the effectiveness of stormwater treatment
BMPs to reduce pollutant concentrations and loads in stormwater runoff. Unfortunately,
inconsistent study methods and reporting make assessment and comparison of BMP
efficiency studies difficult. The studies often analyze different constituents with varying
methods for data collection and analysis. These differences can contribute considerably
to the range of BMP effectiveness observed (Strecker 1994).
Several protocols for parameter selection have been used in the past. The most widely
applied was developed as a part of the Nationwide Urban Runoff Program (NURP).
NURP adopted consistent data collection techniques and analytical parameters so that
meaningful comparisons of gathered data could be made. NURP adopted the following
constituents as "standard pollutants characterizing urban runoff (EPA 1983):
SSC - Suspended Solids Concentration
BOD - Biochemical Oxygen Demand
COD - Chemical Oxygen Demand
CU - Copper
Pb - Lead
Zn - Zinc
TP - Total Phosphorous
SP - Soluble Phosphorous
TKN - Total Kj eldahl Nitrogen
NO2 + NO3 - Nitrate + Nitrite
The following factors were considered for including a parameter in the list of
recommended monitoring constituents (Strecker 1994):
The pollutant has been identified as prevalent in typical urban stormwater at
concentrations that could cause water quality impairment.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
12
-------
The analytical test used can be related back to potential water quality impairment.
Sampling methods for the pollutant are straight forward and reliable for a moderately
careful investigator.
Analysis of the pollutant is economical on a widespread basis.
Treatment is a viable option for reducing the load of the pollutant.
Similar considerations should go into the planning of water quality constituents and
analytical methods to be used in monitoring the effectiveness of stormwater BMPs. The
NURP parameters are a starting point and may or may not represent constituents of
concern for discharges from specific BMPs. As mentioned previously, there is often a
tradeoff between the breadth and depth of a monitoring program given a fixed cost and,
as a result, narrowing the list of constituents monitored can dramatically improve the
ability to quantify the efficiency of the BMP.
Large volumes of data have been collected over the past 20 years on the performance of
many structural stormwater BMPs, with most of the data relating to the performance of
detention basins, retention ponds, and wetlands. Less data are available on the
effectiveness of other types of BMPs (Urbonas 1994). Many of the reported results do
not demonstrate a clear relationship between the efficiency of similar BMPs among the
sites in which they were investigated. Sufficient parametric data has generally not been
reported with the performance data to permit a systematic analysis of the data collected
(Urbonas 1994).
There are a number of important parameters that need to be measured and reported
whenever BMP performance is monitored (Urbonas 1994). A detailed discussion on this
subject is provided in Section 3.4 of this manual.
2.6 BMP Types and Implications for Calculation of Efficiency
The issues involved in selecting methods for quantifying efficiency, performance, and
effectiveness are complex. It would be difficult, at best, to find one method that would
cover the data analysis requirements for the widely varied collection of BMP types and
designs available. When analyzing efficiency, it is convenient to classify BMPs
according to one of the following four distinct categories:
BMPs with well-defined inlets and outlets whose primary treatment depends upon
extended detention storage of stormwater, (e.g., retention (wet) and detention (dry)
ponds, wetland basins, underground vaults).
BMPs with well-defined inlets and outlets that do not depend upon significant storage
of water, (e.g., sand filters, swales, buffers, structural "flow-through" systems).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
13
-------
BMPs that do not have a well-defined inlet and/or outlet (e.g., full retention,
infiltration, porous pavement, grass swales where inflow is overland flow along the
length of the swale).
Widely distributed (scattered) BMPs where studies of efficiency use reference
watersheds to evaluate effectiveness, (e.g., catch basin retrofits, education programs,
source control programs).
Any of the above can also include evaluations where the BMP's efficiency was measured
using before and after or paired watershed comparisons of water quality.
The difficulty in selecting measures of efficiency stems not only from the desire to
compare a wide range of BMPs, but also from the large number of methods currently in
use. There is much variation and disagreement in the literature about what measure of
efficiency is best applied in specific situations, however it is generally accepted that event
mean concentrations and long-term loading provide the best means for observing the
effects of the BMP respectively on acute and chronic pollution.
It has been suggested that intra-storm monitoring could be used to establish paired
inflow/outflow samples during the storm based upon average travel times. However, this
method would only be valid if a BMP were functioning as a perfect plug-flow reactor,
which is rarely the case.
2.7 Relationship Between Monitoring Study Objectives and Data Analysis
In selecting a specific method for quantifying BMP efficiency, it is helpful to look at the
objectives of previous studies seeking such a goal. BMP studies are usually conducted to
obtain information regarding one or more of the following objectives:
What degree of pollution control does the BMP provide under typical operating
conditions?
How does effectiveness vary from pollutant to pollutant?
How does effectiveness vary with various input concentrations?
How does effectiveness vary with storm characteristics such as rainfall amount,
rainfall density, and antecedent weather conditions?
How do design variables affect performance?
How does effectiveness vary with different operational and/or maintenance
approaches?
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
14
-------
Does effectiveness improve, decay, or remain stable over time?
How does the BMP's efficiency, performance, and effectiveness compare to other
BMPs?
Does the BMP reduce toxicity to acceptable levels?
Does the BMP cause an improvement in or protect downstream biotic communities?
Does the BMP have potential downstream negative impacts?
The monitoring efforts implemented most typically seek to answer a small subset of the
above questions. This approach often leaves larger questions about the efficiency,
performance and effectiveness of the BMP, and the relationship between design and
efficiency, unanswered. This document recommends monitoring approaches consistent
with protocols established as part of the National Stormwater Best Management Practices
Database project and useful for evaluating BMP data such that some or all of the above
questions about BMP efficiency can be assessed.
2.8 Physical Layout and Its Effect on Efficiency and Its Measure
The estimation of the efficiency of BMPs is often approached in different ways based on
the goals of the researcher. A BMP can be evaluated by itself or as part of an overall
BMP system. The efficiency of a BMP when bypass or overflow are not considered may
be dramatically different than the efficiency of an overall system. Bypasses and
overflows can have significant effects on the ability of a BMP to remove constituents and
appreciably reduce the efficiency of the system as a whole. Researchers who are
interested in comparing the efficiency of an offline wet pond and an offline wetland may
not be concerned with the effects of bypass on a receiving water. On the other hand,
another researcher who is comparing offline wet ponds with online wet ponds would be
very interested in the effects of the bypass. Often in past study reports detailed
information about the bypass flows is not available. In some cases, comprehensive
inflow and outflow measurements allow for the calculation of a mass balance that can be
used to estimate bypass flow volumes. Estimations of efficiency of a BMP system can be
based on these mass balance calculations coupled with sampling data.
The effect of devices in series is often neglected in the analyses of BMPs. BMPs are
often used in conjunction with a variety of upstream controls. For example detention
ponds often precede wetlands, and sand filters typically have upstream controls for
sediment removal such as a forebay or a structural separator or settling device.
Depending on the approach used to quantify BMP efficiency, the effects resulting from
upstream controls can have a sizable impact on the level of treatment observed.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
15
-------
The efficiency of a BMP system or a BMP can be directly affected by the way in which
an operator chooses to physically manage the system. This is the case where parameters
of a design can be adjusted (e.g., adjustments to the height of an overflow/bypass weir or
gate). These adjustments can vary the efficiency considerably. In order to analyze a
BMP or BMP system thoroughly, all static and state variables of the system must be
known and documented for each monitoring period. The protocols established for the
National Stormwater Best Management Practices Database (Database) provide a
framework for reporting the static and state variables thought to most strongly contribute
to BMP efficiency and provide flexibility for non-standard situations.
2.9 Relevant Period of Impact
The period of analysis used in carrying out a monitoring program is important. The
period used should take into account how the parameter of interest varies with time. This
allows for observation of relevant changes in the efficiency of the BMP on the time scale
in which these changes occur. For example, in a wetland it is often observed that during
the growing season effluent quality for nutrients improves. The opposite effect may be
observed during the winter months or during any period where decaying litter and plant
material may contribute significantly to export of nutrients and, potentially, other
contaminants. Therefore, monitoring observations may need to be analyzed differently
during different seasons. This variation of performance and more specifically efficiency
on a temporal scale is extremely important in understanding how a specific BMP
functions.
In addition to observing how factors such as climate affect BMP efficiency as a function
of time, it is important to relate the monitoring period to the potential impact a given
constituent would have on the receiving water. For example, it may not be useful to
study the removal of some heavy metals (e.g., mercury) for a short period of record when
the negative impacts of such a contaminant are generally expressed over a long time scale
(accumulation in sediments and biota). Likewise, some parameters (e.g., temperature,
BOD, DO, pH, TSS and metals) may have a significant impact in the near term.
Toxicity plays a major role in evaluating the type of monitoring conducted at a site as
well as the time period that should be used to analyze efficiency. Specific constituents
that are acutely toxic may require a short-term analysis on an "intra-storm" basis. Where
dilution is significant and/or a constituent is toxic on a chronic basis, long-term analysis
that demonstrates removal of materials on a sum of loads or average EMC basis may be
more appropriate. Many contaminants may have both acute and chronic effects in the
aquatic environment. These contaminants should be evaluated over both periods of time.
Similarly, hydraulic conditions merit both short and long-term examination. Event peak
flows are examples of short-term data, while seasonal variations of the hydrologic budget
due to the weather patterns are examples of long-term data. Examples of water quality
parameters and their relationship to the time scale over which they are most relevant are
given in Table 2.2.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
16
-------
Table 2.2: Examples of water quality parameters and relevant monitoring period
Time Scale for Analysis
Short-term
Long-term
Both Short- and Long-term
Water Quality Parameter
BOD, DO
Organics, Carcinogens
Metals, TSS, Nitrogen,
Phosphorous, Temperature,
pH, Pesticides
2.9.1 Concentrations, Loads, and Event Mean Concentrations
A variety of tools are available for assessing and quantifying the amount of pollutant
conveyed to and from a BMP. Three primary measures are used most commonly:
concentrations of stormwater at some point in time, the total load conveyed over a
specified duration, or the event mean concentration (EMC).
2.9.1.1 Concentrations
Concentrations measured at a point in time can be useful for BMP efficiency evaluation
in a number of circumstances. Concentrations resulting from samples collected at specific
times during an event allow the generation of a pollutograph (i.e., a plot of the
concentration of pollutants as a function of time). The generation of pollutographs
facilitates the analysis of intra-event temporal variations in runoff concentration. For
example, pollutographs can be used to determine if the "first-flush" phenomenon was
observed for a specific event. Detailed concentration data is one of the approaches for
assessing concentrations of pollutants that have acutely toxic effects, particularly where
runoff from storm events constitutes a significant proportion of downstream flow. Under
some circumstances, reduction of peak effluent concentrations may be more important
than event mean concentration reduction. The cost of implementing a monitoring
program that collects sufficient data to evaluate the temporal variation in runoff and BMP
effluent concentration can be high. The trade-off between collecting data from a larger
number of events versus collecting detailed concentration data from intra-storm periods
often limits the utility of studies that collect detailed concentration data. This type of
detailed monitoring is best focused on outflow monitoring rather than inflow and
outflow.
2.9.1.2 Loads
Loads are typically calculated by the physical or mathematical combination of a number
of individual concentration measurements, which have been assigned by some means an
associated flow volume. A variety of methods are available for estimation of loads. The
method employed is dependent on the sampling and flow measurement techniques used.
Sampling approaches include collection of either timed samples, flow weighted samples,
or some combination of both. Likewise, flow can be collected continuously,
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
17
April 25, 2002
-------
intermittently, or modeled from other hydrologic information such as rain gauge
information, or gauging conducted in a nearby watershed. Many BMP monitoring studies
focus efforts on water quality sample collection and neglect flow measurement. Accurate
flow measurement or well-calibrated flow modeling is essential for loading
determination.
Loads are often most useful for assessing the impact of a BMP where receiving waters
are lakes or estuaries where long-term loadings can cause water quality problems outside
of storms. Where the effluent flow rate from a particular BMP is small compared to the
flow rate of the receiving water body, potential downstream impairments are typically not
dependent on concentrations, but the absolute load of pollutant reaching the receiving
water. For example, loads are the central issue in BMP studies that have direct links to
receiving water bodies that are regulated under the Total Maximum Daily Load (TMDL)
program, particularly where the concern is pollutants deposited in slow moving systems.
Dry weather flows can also contribute substantially to long-term loading. In addition,
"on-line" BMPs (ponds and possibly filters) that have appreciable dry weather flows
passing through them, may have reduced "capacity" for storage of wet weather
pollutants. For example, pond performance may also be affected by the amount of water
in the pond before the event, and filters may have some of their adsorption capacity
consumed by pollutants and other constituents during dry weather flows.
2.9.1.3 Event Mean Concentrations
The term event mean concentration (EMC) is a statistical parameter used to represent the
flow-proportional average concentration of a given parameter during a storm event. It is
defined as the total constituent mass divided by the total runoff volume. The calculation
of EMCs from discrete observations is discussed in detail in Section 2.5.3. When
combined with flow measurement data, the EMC can be used to estimate the pollutant
loading from a given storm. The EMC approach to understanding BMP efficiency is
primarily aimed at wet weather flows.
Under most circumstances, the EMC provides the most useful means for quantifying the
level of pollution resulting from a runoff event. Collection of EMC data has been the
primary focus of the National Stormwater Best Management Practices Database Project.
2.9.2 Measures of BMP Efficiency
The efficiency of Stormwater BMPs (how well a BMP or BMP system removes
pollutants or results in acceptable effluent quality) can be evaluated in a number of ways.
An understanding of how BMP monitoring data will be analyzed and evaluated is
essential to establishing a useful BMP monitoring study. The different methods used to
date are explained in this section to illustrate historical approaches and provide context
for the method recommended in this manual (Effluent Probability Method), which is
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
18
-------
presented at the end of this section. The following table (Table 2.3) summarizes all of the
methods examined by this guidance.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
19
-------
Table 2.3: Summary of historical, alternative, and recommended methods for BMP water
quality monitoring data analysis
Category
Historical
Methods
Alternative
Methods
Recommended
Method
Method Name
Efficiency Ratio (ER)
Summation of Loads
(SOL)
Regression of loads
(ROL)
Mean Concentration
Efficiency of Individual
Storm Loads
Percent Removal
Exceeding Irreducible
Concentration or
Relative to WQ
Standards/Criteria
Relative Efficiency
"Lines of Comparative
Performanceฉ"
Multi-Variate and Non-
Linear Models
Effluent Probability
Method
Recommendation
Not recommended as
a stand-alone
assessment of BMP
performance. More
meaningful when
statistical approach is
used.
Not recommended as
a stand-alone
assessment of BMP
performance. More
meaningful when
statistical approach is
used.
Do not use
Do not use
Do not use
Not recommended -
May be useful in
some circumstances
Not recommended -
May be useful in
some circumstances
Do not use
Possible future use
Recommended
Method
Comments
Most commonly used method to
date. Most researchers assume this
is the meaning of "percent
removal". Typical approach does
not consider statistical significance
of result.
Utilizes total loads over entire
study. May be dominated by a
small number of large events.
Results are typically similar to ER
method. Typical approach does not
consider statistical significance of
result.
Very rarely are assumptions of the
method valid. Cannot be
universally applied to monitoring
data.
Difficult to "track" slug of water
through BMP without extensive
tracer data and hydraulic study.
Results are only for one portion of
the pollutograph.
Storage of pollutants is not taken
into account. Gives equal weight to
all storm event efficiencies
Typically only applicable only for
individual events to demonstrate
compliance with standards.
Typically only applicable only for
individual events to demonstrate
how well a BMP perfoms relative
to how well it would perform if it
Spurious self-correlation. Method
is not valid.
Additional development of
methodology based on more
complete data sets than are
currently available.
Provides a statistical view of
influent and effluent quality.
This is the method recommended
in this guidance manual.
Benefits over other approaches
that are described in this section
of the Guidance.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
20
April 25, 2002
-------
2.9.2.1 Historical Approaches
A variety of pollutant removal methods have been utilized in BMP monitoring studies to
evaluate efficiency. This section describes and gives examples of methods employed by
different investigators. Historically, one of six methods has been used by investigators to
calculate BMP efficiency:
Efficiency ratio
Summation of loads
Regression of loads
Mean concentration
Efficiency of individual storm loads
Reference watersheds and before/after studies
Although use of each of these methods provides a single number that summarizes
efficiency of the BMP in removing a particular pollutant, they are not designed to look at
removal statistically, and thus, do not provide enough information to determine if the
differences in inflow and outflow water quality measures are statistically significant.
Efficiency Ratio
Definition
The efficiency ratio is defined in terms of the average event mean concentration (EMC) of
pollutants over some time period:
average outlet EMC average inlet EMC - average outlet EMC
h,K = 1 =
average inlet EMC average inlet EMC
EMCs can be either collected as flow weighted composite samples in the field or
calculated from discrete measurements. The EMC for an individual event or set of field
measurements, where discrete samples have been collected, is defined as:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
21
-------
EMC =
where,
V: volume of flow during period i
C: average concentration associated with period i
n: total number of measurements taken during event
The arithmetic average EMC is defined as:
average EMC = -
m
where,
m: number of events measured
In addition, the log mean EMC can be calculated using the logarithmic transformation of
each EMC. This transformation allows for normalization of the data for statistical
purposes.
Mean of the Log EMCs =
m
Estimates of the arithmetic summary statistics of the population (mean, median, standard
deviation, and coefficient of variation) should be based on their theoretical relationships
(Appendix A) with the mean and standard deviation of the transformed data. Computing
the mean and standard deviation of log transforms of the sample EMC data and then
converting them to an arithmetic estimate often obtains a better estimate of the mean of
the population due to the more typical distributional characteristics of water quality data.
This value will not match that produced by the simple arithmetic average of the data.
Both provide an estimate of the population mean, but the approach utilizing the log-
transformed data tends to provide a better estimator, as it has been shown in various
investigations that pollutant, contaminant, and constituent concentration levels tend to be
well described by a log-normal distribution (EPA 1983). As the sample size increases,
the two values converge.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
22
-------
Assumptions
This method:
Weights EMCs from all storms equally regardless of relative magnitude of storm.
For example, a high concentration/high volume event has equal weight in the average
EMC as a low concentration/low volume event. The logarithmic data transformation
approach tends to minimize the difference between the EMC and mass balance
calculations.
Is most useful when loads are directly proportional to storm volume. For work
conducted on nonpoint pollution (i.e., inflows), the EMC has been shown to not vary
significantly with storm volume. Accuracy of this method will vary based on the
BMP type.
Minimizes the potential impacts of smaller/"cleaner" storm events on actual
performance calculations. For example, in a storm by storm efficiency approach, a
low removal value for such an event is weighted equally to a larger value.
Allows for the use of data where portions of the inflow or outflow data are missing,
based on the assumption that the inclusion of the missing data points would not
significantly impact the calculated average EMC.
Comments
This method is taken directly from non-point pollution studies and does a good job
characterizing inflows to BMPs but fails to take into account some of the
complexities of BMP design. For example, some BMPs may not have outflow EMCs
that are normally distributed (e.g., media filters and other BMPs that treat to a
relatively constant level that is independent of inflow concentrations).
This method also assumes that if all storms at the site had been monitored, the
average inlet and outlet EMCs would be similar to those that were monitored.
Under all circumstances this method should be supplemented with an appropriate
non-parametric (or if applicable parametric) statistical test indicating if the
differences in mean EMCs are statistically significant (it is better to show the actual
level of significance found, than just noting if the result was significant, assuming a
0.05 level).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
23
-------
Example
The example calculations given below are for the Tampa Office Pond using arithmetic
average EMCs in the efficiency ratio method.
Table 2.4: Example of ER Method results for TSS in the Tampa Office Pond
Period of Record
1990
1993-1994
1994-1995
Average EMC In
27.60
34.48
131.43
Average EMC Out
11.18
12.24
6.79
Efficiency Ratio
59%
64%
95%
ER is rounded, but the other numbers were not (to prevent introduction of any rounding errors in the calculations)
Summation of Loads
Definition
The summation of loads method defines the efficiency based on the ratio of the
summation of all incoming loads to the summation of all outlet loads, or:
SOL = 1-
sum of outlet loads
sum of inlet loads
The sum of outlet loads are calculated as follows:
sum of loads =
Assumptions
Removal of material is most relevant over entire period of analysis.
Monitoring data accurately represents the actual entire total loads in and out of the
BMP for a period long enough to overshadow any temporary storage or export of
pollutants.
Any significant storms that were not monitored had a ratio of inlet to outlet loads
similar to the storms that were monitored.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
24
April 25, 2002
-------
No materials were exported during dry periods, or if they were, the ratio of inlet to
outlet loads during these periods was similar to the ratio of the loads during the
monitored storms.
Comments
A small number of large storms typically dominate efficiency.
If toxics are a concern then this method does not account for day-to-day releases,
unless dry weather loads in and out are also accounted for. In many cases long-term
dry weather loads can exceed those resulting from wet weather flows.
Under all circumstances this method should be supplemented with an appropriate
non-parametric (or if applicable parametric) statistical test indicating if the
differences in loads are statistically significant (it would be better to show the actual
level of significance found, rather than just noting if the result was significant,
assuming a 0.05 level).
Example
The example calculations given in Table 2.5 are for the Tampa Office Pond using a mass
balance based on the summation of loads.
Table 2.5: Example of SOL Method results for TSS in the Tampa Office Pond.
Period of Record
1990
1993-1994
1994-1995
Sum of Loads
In (kg)
134.60
404.19
2060.51
Sum of Loads
Out (kg)
39.67
138.44
130.20
SOL Efficiency
71%
66%
94%
SOL Efficiency is rounded, but the other numbers were not (to prevent introduction of any rounding errors in the calculations)
Regression of Loads (ROD
Definition
The regression of loads method as described by Martin and Smoot (1986) defines the
regression efficiency as the slope (/3) of a least squares linear regression of inlet loads and
outlet loads of pollutants, with the intercept constrained to zero. The zero intercept is
specified as an "engineering approximation that allows calculation of an overall
efficiency and meets the general physical condition of zero loads-in (zero rainfall) yield
zero loads-out". The equation for the ROL efficiency is:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
25
April 25, 2002
-------
, o T i a Loads out
Loads out = p Loads in = p
Loads in
The percent reduction in loads across the BMP is estimated as:
Loads out
Percent Removal = 1 - j8 = 1 - -
Loads in
Due to the nature of stormwater event monitoring, it is rare that all of the assumptions for
this method are valid, particularly requirements for regression analysis. The example
calculations and plots provided in this section are from one of the better studies available
at the time this manual was written, and as can be seen from the ROL plots, the data does
not meet the requirements for proper simple linear regression analysis.
Assumptions
Any significant storms that were not monitored had a ratio of inlet to outlet loads
similar to the storms that were monitored. The slope of the regression line would not
significantly change with additional data.
No materials were exported during dry periods, or if they were, the ratio of inlet to
outlet loads during these periods was similar to the ratio of the loads during the
monitored storms.
The data is well represented by a least squares linear regression, that is:
o The data is "evenly" spaced along the x-axis.
o Using an analysis of variance on the regression, the slope coefficient is
significantly different from zero (the p value for the coefficient should
typically be less than 0.05, for example).
o A check of the residuals shows that the data meets regression requirements.
The residuals should be random (a straight line on probability paper) and the
residuals should not form any trend with predicted value or with time (i.e.,
they form a band of random scatter when plotted).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
26
-------
Comments
A few data points often control the slope of the line due to clustering of loads about
the mean storm size. Regressions are best used where data is equally populous
through the range to be examined. This is readily observed in the examples that
follow (See Figures 2.1 through 2.3).
The process of constraining the intercept of the regression line to the origin is
questionable and in some cases could significantly misrepresent the data. It may be
more useful to apply the Regression of Loads method over some subset of the data
without requiring that the intercept be constrained to the origin. The problem with
this alternative approach is that a large number of data points are required in order to
get a good fit of the data. Often a meaningful regression cannot be made using the
data that was collected. This is well illustrated by the very low R values in the table
below. Forcing the line through the origin, in these cases, provides a regression line
even where no useful trend is present.
There is sufficient evidence that this first order polynomial (straight line) fit is not
appropriate over a large range of loadings. Very small events are much more likely to
demonstrate low efficiency where larger events may demonstrate better overall
efficiency depending on the design of the BMP.
Table 2.6: Example of ROL Method results for TSS in the Tampa Office Pond.
Period of Record
1990
1993-1994
1994-1995
Slope of
Regression
Line
0.21
0.18
0.05
R2
0.06
-0.06
0.46
Percent Removal
79%
82%
95%
Percent Removal is rounded, but the other numbers were not (to prevent introduction of any rounding errors in the calculations)
The regressions used to arrive at the above slopes are given in Figures 2.1-2.3.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
27
April 25, 2002
-------
8
7
eT 6
^
H 5
O 4
Q
< >.
O 3.
2
1
5.0 10.0 15.0 20.0
LOAD IN (KG)
25.0
Figure 2.1: ROL Plot for use in Calculating Efficiency for TSS using the Tampa Office
Pond (1990) (Slope = 0.2135, R2 = 0.0563, Standard Error in Estimate =
2.176, one point is considered an outlier with a Studentized Residual of
3.304). All points were used for regression. Method is not valid due to
failure of simple linear regression assumptions.
o
60
50
40
30
i i i i
0 10 20 30 40 50 60 70 80
LOAD IN (KG)
Figure 2.2: ROL Plot for use in Calculating Efficiency for TSS using the Tampa Office
Pond (1993-1994) (Slope = 0.1801, R2 = -0.0562, Standard Error in
Estimate = 10.440, one point is considered an outlier with a Studentized
Residual of 13.206 and one point has a high Leverage of 0.323). All points
were used for regression. Method is not valid due to failure of simple linear
regression assumptions.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
28
April 25, 2002
-------
100 200 300 400 500 600 700
LOAD IN (KG)
Figure 2.3: ROL Plot for use in Calculating Efficiency for TSS using the Tampa Office
Pond (1994-1995) (Slope = 0.0492, R2 = 0.4581, Standard Error in Estimate
= 5.260, three points are considered outliers (Studentized Residuals of
3.724, 8.074, and -4.505, the point to the far right on the graph has large
Leverage (0.724) and Influence, Cook Distance = 36.144). All points were
used for regression. Method is not valid due to failure of simple linear
regression assumptions.
Mean Concentration
Definition
The mean concentration method defines the efficiency as unity minus the ratio of the
average outlet to average inlet concentrations. The equation using this method is:
Mean Concentration = 1 -
average outlet concentration
average inlet concentration
This method does not require that concentrations be flow weighted. This method might
have some value for evaluating grab samples where no flow weighted data is available or
where the period of record does not include the storm volume.
Assumptions
The flows from which the samples were taken are indicative of the overall event.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
29
April 25, 2002
-------
Comments
This method might be useful for calculating BMP's effectiveness in reducing acute
toxicity immediately downstream of the BMP. This is due to the fact that acute
toxicity is measured as a threshold concentration value of a specific constituent in the
effluent at or near the point of discharge.
This methods weights individual samples equally. Biases could occur due to
variations in sampling protocols or sporadic sampling (i.e., collecting many samples
close in time and others less frequently). The sample collection program specifics are
not accounted for in the method and estimated efficiencies are often not comparable
between studies.
There is appreciable lag time for most BMPs between when a slug of water enters a
BMP and when the slug leaves the BMP. Unless this lag time is estimated (e.g.,
through tracer studies) results from this approach can be quite inaccurate. Results of
this method may be particularly difficult to interpret where lag time is ignored or not
aggressively documented.
This method does not account for storage capacity. Typically BMPs will have an
equal or lesser volume of outflow than of inflow. On a mass basis this affects
removal, since volume (or flow) is used with concentration to determine mass for a
storm event,
CoutFOM( average outlet concentration
CinVin average inlet concentration
where,
C;n: Concentration In
Cout: Concentration Out
Vin: Volume In
Vout: Volume Out
In this respect, it is often more conservative (i.e., lower removal efficiency stated) to use
a concentration rather than mass-based removal approach.
Efficiency of Individual Storm Loads
Definition
The Efficiency of Individual Storm Loads (ISL) method calculates a BMP's efficiency
for each storm event based on the loads in and the loads out. The mean value of these
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
30
-------
individual efficiencies can be taken as the overall efficiency of the BMP. The efficiency
of the BMP for a single storm is given by:
Storm Efficiency = 1 - Lฐadฐut
Loadm
The average efficiency for all monitored storms is:
m
^ Storm Efficiency j
Average Efficiency =
m
where,
m: numb er of storm s
Assumptions
Storm size or other storm factors do not play central roles in the computation of
average efficiency of a BMP.
Storage and later release of constituents from one storm to the next is negligible.
The selection of storms monitored does not significantly skew the performance
calculation.
Comments
The weight of all storms is equal. Large storms do not dominate the efficiency in this
scenario. The efficiency is viewed as an average performance regardless of storm
size.
Some data points cannot be used due to the fact that there is not a corresponding
measurement at either the inflow or the outflow for a particular storm, and thus
efficiency cannot always be calculated on a storm-by-storm basis. This is not true for
the ER method, however it is a limitation of the Summation of Load Method.
Storm by storm analysis neglects the fact that the outflow being measured may have a
limited relationship to inflow in BMPs that have a permanent pool. For example, if a
permanent pool is sized to store a volume equal to the average storm, about 60 to 70
percent of storms would be less than this volume [from studies conducted using
SYNOP (EPA 1989)].
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
31
-------
Table 2.7: Example of Individual Storm Loads Method results for TSS in the Tampa
Office Pond.
Period of Record
1990
1993-1994
1994-1995
Efficiency
29%
-2%
89%
Summary and Comparison of Historical Methods
The table below shows the results of the various historical methods shown above for
calculating efficiency for the Tampa Office Pond. The four methods demonstrated (mean
concentration method was not applicable to data available from the Tampa Office Pond
study) vary widely in their estimates of percent removal depending on the assumptions of
each method as discussed above.
Table 2.8: Comparison of BMP efficiency methods.
Design
1990
1993-1994
1994-1995
Method
Efficiency
Ratio (ER)
59%
64%
95%
Summation
of Loads
(SOL)
71%
66%
94%
Regression of
Loads (ROL)
79%
82%
95%
Efficiency of
Individual Storms
29%
-2%
89%
2.9.2.2 Other Methods and Techniques
"Irreducible Concentration" and "Achievable Efficiency"
As treatment occurs and pollutants in stormwater become less concentrated, they become
increasingly hard to remove. There appears to be a practical limit to the effluent quality
that any BMP can be observed to achieve for the stormwater it treats. This limit is
dictated by the chemical and physical nature of the pollutant of concern, the treatment
mechanisms and processes within the BMP, and the sensitivity of laboratory analysis
techniques to measure the pollutant. This concept of "irreducible concentration" has
significant implications for how BMP efficiency estimates are interpreted. However, it is
possible to get concentrations as low as desired, but in most cases achieving extremely
low effluent concentrations may not be practical (i.e., would require treatment trains or
exotic methods). For example, colloids are typically viewed as "never" being able to be
removed in a pond (settling is the primary mechanism for treatment in ponds), despite the
fact that they could be further removed through chemical addition.
The term "irreducible concentration" (C*) has been used in stormwater literature
(Schueler 2000) to represent the lowest effluent concentration for a given parameter that
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
32
April 25, 2002
-------
can be achieved by a specific type of stormwater management practice. Schueler
examined the effluent concentrations achieved by stormwater management practices from
published studies for several parameters. From this research, the following estimates of
"irreducible concentrations" for TSS, Total Phosphorous, Total Nitrogen, Nitrate-
Nitrogen, and TKN for all stormwater management practices were proposed:
Table 2.9: "Irreducible concentrations" as reported by Scheuler, 2000.
Contaminant
TSS
Total Phosphorous
Total Nitrogen
Nitrate-Nitrogen
TKN
Irreducible Concentration
20 to 40 mg/L
0.15to0.2mg/L
1.9 mg/L
0.7 mg/L
1.2 mg/L
Recent research (ASCE 2000) indicates that achievable effluent concentrations vary
appreciably between BMP types. For example, in many cases, well-designed sand filters
can achieve lower effluent concentrations of TSS than well-designed detention facilities
or grassed swales. However, sand filters have issues with long-term maintenance of flow
treatment volumes.
The typical approach to reporting the ability of a BMP to remove pollutants from
stormwater entails comparing the amount of pollutant removed by the BMP to the total
quantity of that pollutant. The concept of irreducible concentration, however, suggests
that in some cases it may be more useful to report the efficiency of the BMP relative to
some achievable level of treatment (i.e. express efficiency as the ability of the BMP to
remove the fraction of pollutant which is able to be removed by a particular practice.)
The following example illustrates this approach. Suppose that two similar BMPs have
been monitored and generated the following results for TSS:
Table 2.10: Example TSS results for typical ER Method
Percent TSS Removal Using Absolute Scale
Influent Concentration
Effluent Concentration
Efficiency Ratio
BMP A
200 mg/L
100 mg/L
50%
BMPB
60 mg/L
30 mg/L
50%
Clearly, the effluent from BMP B is higher quality than that from BMP A, however
comparing percent removals between BMPs alone would indicate that both BMPs have
an equal efficiency. Methods have been suggested for quantifying the dependence of
BMP efficiency on influent concentration. The following section presents one such
method advanced by Minton (1998).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
33
April 25, 2002
-------
In order to account for the dependence of BMP efficiency on influent concentration,
Minton (1998) suggests a method of evaluating BMP efficiency that would recognize the
relationship between influent concentration and efficiency. The relationship is
summarized as follows:
Achievable Efficiency = (C^fiuent - Ciimit)/ Client
where,
Influent Concentration of Pollutant; and
The lower attainable limit concentration of the BMP (e.g., "irreducible
concentration" or value obtained from previous monitoring of effluent
quality)
For example, if a BMP had a lower treatment limit of TSS at 20mg/L concentration, then
at an influent TSS concentration of 100 mg/L, it would be assigned an equivalent
performance of 80%, while at an influent TSS concentration of 50 mg/L the equivalent
performance would be 60%.
This method relies on the ability to determine the lower attainable limit concentration,
which is analogous to the "irreducible concentration" for a specific BMP, however
effluent quality is best described not as a single value, but from a statistical point of view
(See the Effluent Probability Method).
The Achievable Efficiency may be useful in better understanding the results of the ER
method in cases where the influent concentration is lower than is typically observed.
Alternately, a single factor (dubbed the Relative Efficiency here) can be used to report
how well a BMP is functioning during some period relative to what that BMP is
theoretically or empirically able to achieve (as defined by the Achievable Efficiency).
As shown below, the Relative Efficiency can be found by dividing the Efficiency Ratio
by the Achievable Efficiency, thus yielding an estimate of how well the BMP performed
relative to what is "achievable".
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
34
-------
Relative Efficiency =
Efficiency Ratio [(C influent - C effluent)/C influent]
Achievable Efficiency [(C influent - C iimit)/ C influent]
Or simplifying:
Relative Efficiency = (C influent - C effluent)/(C influent - C iimit)
If applied to the example presented earlier in this section, the following results are
obtained:
Table 2.11: Example TSS results for demonstration of Relative Efficiency approach.
Influent Concentration
C limit
Effluent Concentration
Relative Efficiency
BMP A
200 mg/L
20mg/L
100 mg/L
56%
BMPB
60 mg/L
20 mg/L
30 mg/L
75%
For this example, the results indicate that BMP B is achieving a higher level of treatment
than BMP A and this approach may be more useful as a comparative tool than the
Efficiency Ratio for some data sets. The Relative Efficiency for a BMP's effectiveness is
still influenced by influent concentration but less so than is the Efficiency Ratio.
As C influent approaches C iimit the Relative Efficiency goes to infinity, which is not a very
meaningful descriptor. However, if the influent concentration is near the "irreducible
concentration" for a particular pollutant, very little treatment should occur and C influent -
C effluent should approach zero. C effluent, at least theoretically, should always be higher
than C iimit and the numerator of the equation should approach zero faster than the
denominator. If C influent is less than C limit, the Relative Efficiency approach should not
be used. As is always the case, any of the percent removal efficiency approaches
(including the Efficiency Ratio Method) should not be employed if there is not a
statistically significant difference between the average influent and effluent
concentrations.
If this method is used to represent data from more than one event (i.e., mean EMCs are
calculated) it should be supplemented with an appropriate non-parametric (or if
applicable parametric) statistical test indicating if the differences are statistically
significant (it would be preferred to show the actual level of significance found, instead
of just noting if the result was significant, assuming a 0.05 level).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
35
-------
Percent Removal Relative to Water Quality Standards
From a practical or programmatic perspective, it may be more useful to substitute the
water quality limit for the "irreducible concentration" as a measure of how well the BMP
is meeting specific water quality objectives. A measure of efficiency can be calculated to
quantify the degree to which stormwater BMPs employed are meeting or exceeding state or
federal water quality criteria or standards for the runoff they treat.
Standards are enforceable regulations established within the context of an NPDES permit
or a TMDL and are usually specific to the receiving water. Water quality criteria are
more general guidelines expressed as constituent concentrations, levels, or narrative
statements, representing a quality of water that supports a particular beneficial use.
By showing that stormwater is being treated to a level that is higher than standards require
or criteria recommend, a permitee may be able to demonstrate to regulators or stakeholders
that their current stormwater management practices are adequate for a particular constituent
of concern. The equation to calculate the Percent Removal Relative to Receiving Water
Quality Limits is as follows:
Percent Removal Relative to Receiving Water Quality Limits =
(\^ influent ~ ^ effluent)'(^ influent ~~ ^ standard/criterion)
The following example illustrates the application of this approach for reporting efficiency:
Table 2.12: Example of percent removal relative to receiving water quality limits
approach.
Influent Concentration (EMC)
^ standard/criterion
Effluent Concentration (EMC)
Percent Removed Relative to Established WQ Limits
BMP A
1.65ug/l
0.889 ug/1
0.635 ug/1
133 %
The results indicate that the BMP for the given event is meeting the water quality
standard or criterion for dissolved lead. In fact the BMP is functioning to remove in
excess of the amount needed to bring the influent concentration below the water quality
limit (as indicated in the example by a value greater than 100%). Use of this method is
only recommended for specific event analysis. As mentioned for previous analyses, if
this approach is taken for a series of events it should be supplemented with an appropriate
non-parametric (or if applicable parametric) statistical test indicating if the differences
are statistically significant (it would be better to show the actual level of significance
found, than just noting if the result was significant, assuming a 0.05 level)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
36
-------
"Lines of Comparative Performanceฉ"
For many stormwater treatment BMPs, the efficiency of the BMP decreases as a function
of the influent concentration. Methods have been recommended that integrate this
concept into efficiency evaluations. The "Lines of Comparative Performanceฉ" (Minton
1999) is one such method.
In this method, plots of percent removal as a function of the influent concentration for
each storm are generated for each pollutant monitored. The results of these plots are
overlain on plots of data collected from studies of similar BMPs within a region.
"Lines of Comparative Performanceฉ" are generated for the data from similar BMPs
based on best professional judgment by examining the likely "irreducible concentration"
for a particular pollutant, the detection limit for that pollutant, and knowledge of expected
maximum achievable efficiency for a BMP type.
This method has primarily been suggested as an approach to evaluate the efficiency of
innovative and "unapproved" stormwater technologies. "To be accepted, the
performance data points of an unapproved treatment technology must fall above and to
the left of the 'Line of Comparative Performanceฉ'."
This approach has several major problems. The most significant flaw is the use of
"spurious" self-correlation. Plots such as those shown in Figures 2.4 through 2.6 can be
generated using random, normally distributed influent and effluent concentrations as seen
below in Figure 2.7. As such, it is strongly recommended that this approach not be
employed in BMP monitoring evaluation studies. This approach may lead to overly
complicated analysis methodologies without providing additional useable information on
BMP functionality.
Figures 2.4-2.6 below show work conducted by Minton in the development of the
Achievable Efficiency approach.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
37
-------
Pacific Northwest - TSS
Field - individual storms
100%
-40%
-60%
Caveats: faeiiiiies are uf various
designs iNsi #r& not neof^anly
" cofisistcnt with cmnrRt design crilc
0 20 40 60 80 100 120 140 160 180 200
Influent concentration (mg/L) MS- ป,'
iMllRg
Guy
Figure 2.4: Removal Efficiency (ER Method) of TSS as a Function of Influent
Concentration (Minton 1999)
- TP
-
100%
80%
60%
>
c 40%
m
"o
E 20%
o
f 0%
-40%
Swale
* -- ^ :
0 50 100 150 200 280 300 350 400 4iO SQQ
Influent
Pfenning Assoeiaics
Owy 1, Mimoa
-
Figure 2.5: Removal Efficiency (ER Method) of Total Phosphorous as a function of
influent concentration (Minton 1999)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
38
April 25, 2002
-------
Pacific Northwest Data - Zinc
Field studies - individual storms
100%
80% -
Two data points
above 500 tng/l. for __f
sand fillers ,
-i-
-40%
-60%
-80%
Negative values
Grass swale -21*% ill 22 ug/l.
Filter -368% si 66 ug/L and
-222% ซ] 134 ug/l.
-100%-
0 50 100 150 200 250 300 350 400 450 500
Influent (ug/L)
Resource Planning Associates.
Oar.' R. Minton
Swale
Wetpond
Sand fitter
July 13, 1999
Figure 2.6: Removal Efficiency (ER Method) of Total Zinc as a Function of Influent
Concentration (Minton 1999)
100
200 300
Influent Concentration
400
500
Figure 2.7: Percent removal as a function of influent concentration for randomly
generated, normally distributed influent and effluent concentrations. Any
number of similar charts can be generated from randomized data.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
39
April 25, 2002
-------
An alternate method which does not include the serious problems associated with the
"Lines of Comparative Performanceฉ", but presents relatively the same information can
be generated using a simple plot of effluent concentration as a function of influent
concentration with "rays" (or curves on a log plot) originating from the plot origin for
several levels of control (e.g., 0, 25, 50, 75, and 90%). The plot may need to be a log-log
plot for data with a large range of values typical of stormwater monitoring data.
Multi-Variate and Non-Linear Models
Reporting efficiency as a percent removal that is calculated based on the difference
between influent and effluent concentrations will always make a BMP that treats higher
strength influents appear to be more efficient than one treating weaker influents if both
are achieving the same effluent quality. A more useful descriptor of efficiency would
take into consideration that weaker influents are more difficult to treat than concentrated
ones. A multi-variate equation that includes corrections to compensate for this
phenomena or a non-linear model may be worth considering for reporting efficiency.
A model that approaches pollutant removal in a manner similar to the reaction rates for
complex physical and chemical batch and plug-flow processes may be useful. To date
calibration of such a model for all but the most elementary situations (e.g., settling of
solids in relatively simplistic flow regimes) is difficult given the complexity of the real-
world problem. As more high quality data becomes available, other approaches to
evaluating BMP efficiency may become apparent.
Currently, effluent quality, as discussed below, is the best indicator of overall BMP
performance.
2.9.2.3 Recommended Method
The following method is recommended for use in analyzing new and existing monitoring
studies.
Effluent Probability Method
The most useful approach to quantifying BMP efficiency is to determine first if the BMP
is providing treatment (that the influent and effluent mean EMCs are statistically different
from one another) and then examine either a cumulative distribution function of influent
and effluent quality or a standard parallel probability plot.
Before any efficiency plots are generated, appropriate non-parametric (or if applicable
parametric) statistical tests should be conducted to indicate if any perceived differences in
influent and effluent mean event mean concentrations are statistically significant (the
level of significance should be provided, instead of just noting if the result was
significant, assume a 95% confidence level).
Effluent probability method is straightforward and directly provides a clear picture of the
ultimate measure of BMP effectiveness, effluent water quality. Curves of this type are the
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
40
-------
single most instructive piece of information that can result from a BMP evaluation study.
The authors of this manual strongly recommend that the stormwater industry accept this
approach as a standard "rating curve" for BMP evaluation studies.
The most useful approach for examining these curves is to plot the results on a standard
parallel probability plot (see Figures 2.8-2.10). A normal probability plot should be
generated showing the log transform of both inflow and outflow EMCs for all storms for
the BMP. If the log transformed data deviates significantly from normality, other
transformations can be explored to determine if a better distributional fit exists. Figures
2.8-2.10 show three types of results that can be observed when plotting pollutant
reduction observations on probability plots. The data was taken from the Monroe St. wet
detention pond study in Madison, WI, collected by the USGS and the WI DNR. Figure
2.8 for suspended solids (particulate residue) shows that SS are highly removed over
influent concentrations ranging from 20 to over 1,000 mg/L. A simple calculation of
"percent removal" (ER Method) would not show this consistent removal over the full
range of observations. In contrast, Figure 2.9 for total dissolved solids (filtered residue)
shows poor removal of TDS for all concentration conditions, as expected for this wet
detention pond. The "percent removal" (ER Method) for TDS would be close to zero and
no additional surprises are indicated on this plot. Figure 2.10, however, shows a wealth of
information that would not be available from simple statistical numerical summaries,
including the historical analysis approaches described in this manual. In this plot, filtered
COD is seen to be poorly removed for low concentrations (less than about 20 mg/L), but
the removal increases substantially for higher concentrations. Although not indicated on
these plots, the rank order of concentrations was similar for both influent and effluent
distributions for all three pollutants (Burton and Pitt 2001).
Water quality observations do not generally form a straight line on normal probability
paper, but do (at least from about the 10th to 90th percentile level) on log-normal
probability plots. This indicates that the samples generally have a log-normal distribution
as described previously in this document and many parametric statistical tests can often
be used (e.g., analysis of variance), but only after the data is log-transformed. These plots
indicate the central tendency (median) of the data, along with their possible distribution
type and variance (the steeper the plot, the smaller the COV and the flatter the slope of
the plot, the larger the COV for the data). Multiple data sets can also be plotted on the
same plot (such as for different sites, different seasons, different habitats, etc.) to indicate
obvious similarities (or differences) in the data sets. Most statistical methods used to
compare different data sets require that the sets have the same variances, and many
require normal distributions. Similar variances are indicated by generally parallel plots of
the data on the probability paper, while normal distributions would be reflected by the
data plotted in a straight line on normal probability paper. (Burton and Pitt 2001)
Probability plots should be supplemented with standard statistical tests that determine if
the data is normally distributed. These tests, at least some available in most software
packages, include the Kolmogorov-Smirnov one-sample test, the chi-square goodness of
fit test, and the Lilliefors variation of the Kolmogorov-Smironov test. They are paired
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
41
-------
tests comparing data points from the best-fitted normal curve to the observed data. The
statistical tests may be visualized by imagining the best-fit normal curve data (a straight
line) and the observed data plotted on normal probability paper. If the observed data
crosses the fitted curve data numerous times, it is much more likely to be normally
distributed than if it only crosses the fitted curve a small number of times (Burton and Pitt
2001).
1 10 100 1000
Particular^ Residue (SS) (mg/L)
n Inlet
A Outlet
Figure: 2.8: Probability plot for Suspended
Solids
15 1 CO IBM
Filtered Residue (TDS)
Inlet
Outlet
Figure: 2.9 Probability plot for Total
Dissolved Solids
a
TD
c.
10
Inlet
Outlet
Filtered COD (mg/L)
Figure: 2.10: Probability plot for Chemical
Oxygen Demand
(Originally by Burton and Pitt 2001)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
-------
2.9.2.4 Reference Watershed Methods
Many BMPs do not allow for comparison between inlet and outlet water quality
parameters. In addition, it is often difficult or costly, where there are many BMPs being
installed in a watershed (e.g., retrofit of all catch basins), to monitor a large number of
specific locations. A reference watershed is often used to evaluate the effectiveness of a
given BMP or multiple BMPs of the same type. The database allows for a watershed and
all associated data to be identified for use as a reference watershed. One of the primary
reasons for using a reference watershed is that for some BMPs there is no clearly defined
inlet or outlet point at which to monitor water quality. Such is the case with many non-
structural BMPs, porous pavements, and infiltration practices.
The difficulty in determining the effectiveness of a BMP using a reference watershed
approach stems from the large number of variables typically involved. When setting up a
BMP monitoring study, it is advantageous to keep the watershed characteristics of the
reference watershed and the test watershed as similar as possible. Unfortunately, finding
two watersheds that are similar is often quite difficult, and the usefulness of the data can
be compromised as a result. In order to determine the effectiveness of a BMP based on a
reference watershed, an accurate accounting of the variations between the watersheds,
and operational and environmental conditions is needed. The Database explicitly stores
some of the key parameters required for normalization of watershed and environmental
conditions.
The most obvious parameter used to normalize watershed characteristics is area. If the
ratio of land uses and activities within each watershed is identical in both watersheds then
the watershed area can be scaled linearly. The loads found at each downstream
monitoring station for each event can be scaled linearly with area as well. Difficulty
arises when land use in the reference watershed is not found in the same ratio. In this
case, either the effects of land use must be ignored or a portion of the load found for each
event must be allocated to a land use and then scaled linearly as a function of the area
covered by that land use. In many cases, the differences in land use can be ignored, (e.g.,
between parking lots with relatively small, but different unpaved areas). The effect of the
total impervious area is relevant and should always be reported in monitoring studies.
The ratio of the total impervious areas can be used to scale event loads. Scaling the loads
based on impervious areas would be best used where the majority of pollutants are from
runoff from the impervious areas (e.g., parking lots), or the contaminant of interest results
primarily from deposition on impervious surfaces, (e.g., TSS in a highly urban area).
Methods that attempt to determine BMP performance from poorly matched watersheds
yield poor results at best. As the characteristics of the two watersheds diverge, the effect
of the BMP is masked by the large number of variables in the system; the noise in the
data becomes greater than the signal.
The analysis of BMPs utilizing reference watersheds also requires incorporation of
operational details of the system, (e.g., frequency of street sweeping, type of device used,
device setup). Monitoring studies should always provide the frequency, extent, and other
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
operational parameters for nonstructural BMPs. If the BMP is an alteration of the
frequency of a certain practice, the system can be viewed in two ways, (1) as a
control/test system, or (2) as a series of data aimed at quantifying the continuous effect of
increasing or decreasing BMP frequency. In the first case, the BMP can be analyzed in a
manner similar to other BMPs with reference watersheds. In the second case, the loads
realized at the monitoring stations need to be correlated with the frequency using some
model for the effectiveness of the practice per occurrence.
2.9.3 BMPs and BMP Systems
Overflow and bypassing of treatment BMPs affect the long-term performance of the
pollution control measure. Many types of BMP structures, such as detention or filtration
basins, are designed to treat specific volumes of stormwater runoff. Runoff volumes (or
flows) exceeding the designed storage volume or maximum flow rate are bypassed
untreated or partially treated. In order to accurately assess the long-term efficiency of the
BMP system, the bypass flow needs to be taken into consideration. Ideally, a third flow
monitor should be installed to measure by-passed flow directly (Oswald and Mattison
1994).
If monitoring data is not cost effective or physically difficult to collect, estimates of
bypass can be made using inflow / outflow water balance calculations or modeled from
local rainfall data, watershed hydrology, and BMP system hydraulics. The volume
treated by a BMP for each event can be compared to a measured or modeled runoff
volume yielding the volume of bypass.
Estimates of BMP system efficiency should always be calculated for the entire BMP
system (in addition to the BMP). Mass balance checks should be performed in all cases to
help verify monitoring data and/or modeled flow rates.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
-------
3 Developing a BMP Monitoring Program
This chapter describes the steps involved in developing and implementing a monitoring
program to evaluate BMP effectiveness. Regardless of the scope and objectives, designing
a monitoring plan generally involves four phases:
Phase 1: Determine the objectives and scope of your monitoring program
Phase 2: Develop the monitoring plan in view of your objectives
Phase 3: Implement the monitoring plan
Phase 4: Evaluate and report the results of monitoring
The activities associated with each phase are listed below.
Phase 1: Determine Objectives and Scope
Identify permit requirements and/or information needs
Compile and review existing information (maps, drawings, results from prior sampling,
etc.) relevant to permit requirements and/or information needs
Develop monitoring program objectives and scope
Phase 2: Develop Monitoring Plan
Select monitoring locations
Select monitoring frequency
Select parameters and analytical methods
Select monitoring methods and equipment
Select storm criteria (i.e., size, duration, season)
Develop mobilization procedures
Prepare a quality assurance/quality control plan
Prepare a health and safety plan
Prepare a data management plan
Phase 3: Implement Monitoring Plan
Install equipment (and modify channels, if applicable)
Test and calibrate equipment
Conduct training
Conduct monitoring (collect samples)
Conduct analyses (field and/or laboratory)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
45
-------
Phase 4: Evaluate and Report Results
Validate chemical data quality
Evaluate results
Report the results
Several of the steps in developing a monitoring program are dependent on one another.
Consequently, earlier steps may need to be revisited and refined throughout the planning
process. For example, if it is determined in Phases 2 or 4 that monitoring more storms is
needed to achieve objectives, revisiting the "select monitoring location" task and
selecting a lower number of sampling locations and/or a different analytical scheme may
be needed to keep within the schedule and budget.
Determine Key Study Parameters
Key parameters of the monitoring project are determined using the information gathered
in the previous steps of the systematic planning process. Key study parameters include
site selection, number of monitored storm events and their temporal distribution,
characteristics of target storm events, types of samples (composite, grab, etc.), and
analytical constituents. The better these characteristics are understood, the more
efficiently the monitoring data can be collected (Caltrans 1997).
The planned number of sites and monitoring events are often constrained by fiscal
factors, such as the cost of sample collection and analysis. For this reason, the list of
analytical constituents is often considered in the early stages of project planning (see
Section 3.2.3), so that costs of the appropriate sample collection and analysis can be
factored into the expected cost per monitoring event. The analytical constituents are
often prescribed by regulatory or legal mandate.
3.1 Phase I - Determine Objectives and Scope of BMP Water Quality
Monitoring Program
It is particularly important that the objectives of a BMP monitoring program be clearly
stated and recorded. The process of writing them down generally results in careful
consideration being given to the possible options. Written objectives help avoid
misunderstandings by project participants, are an effective way of communicating with
sponsors, and provide assurance that the monitoring program has been systematically
planned.
Studies of BMP performance are usually conducted to obtain information regarding one
or more of the following questions:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
46
-------
What degree of pollution control or effluent quality does the BMP provide under
normal conditions?
How does this performance vary from pollutant to pollutant?
How does this normal performance vary with large or small storm events?
How does this normal performance vary with rainfall intensity?
How do design variables affect performance?
How does performance vary with different operational and/or maintenance
approaches?
Does performance improve, decay, or remain stable over time?
How does this BMP's performance compare with the performance of other
BMPs?
Does this BMP help achieve compliance with water quality standards?
Many BMP monitoring programs have been established to satisfy requirements prescribed
by permits to monitor the effectiveness of BMPs, but often the wording of such
requirements is vague. Local program-specific objectives are likely to provide the soundest
basis for planning a BMP monitoring study.
A well-designed BMP monitoring program may help address specific monitoring
questions, thereby enabling better decisions regarding allocation of resources to address
stormwater quality issues. The ultimate use of the monitoring results should be kept in
mind throughout the monitoring program planning process.
3.1.1 Monitoring and Literature Review to Assess BMP Performance
Typically, structural BMPs have well-defined boundaries and are relatively easy to monitor.
Other types of BMPs, especially non-structural BMPs (e.g., street sweeping, catch basin
cleaning, sewer cleaning, illicit discharge elimination), are more difficult to monitor partly
because they tend to be geographically interspersed with many pollutant sources and can be
influenced by many factors that cannot be "controlled" in an experimental sense. Some
non-structural BMPs, such as public education programs, oil recycling programs, and litter
control programs are virtually impossible to monitor or at best can be evaluated using trend
monitoring.
It is assumed that many stormwater quality management programs will consider the
possibility of implementing some structural BMPs by experimenting with them on a pilot-
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
47
-------
scale by testing and demonstrating their performance, their costs, and their practical
implications before committing to larger-scale implementation. Programs that already have
structural BMPs in place may also test their performance for a variety of reasons.
Before obtaining BMP performance data or establishing the objectives and scope of the
BMP monitoring program, it is useful to investigate other regional BMP monitoring
programs to learn from their successes and/or failures in implementing the BMP,
establishing their objectives and scope of their BMP monitoring program, and obtaining
meaningful results. This research will also provide some level of foresight in developing a
meaningful monitoring program that will produce results that will be useful in achieving
project goals and comparable to other programs.
Nationally, many stormwater programs need BMP performance data, and many are
planning or conducting performance monitoring. The concept of sharing monitoring results
is very appealing but could be seriously constrained if pre-planning to maximize the chances
of yielding comparable/compatible monitoring approaches, analytical protocols, and data
management are not implemented. Some of the guidance provided in this manual and
referred to in literature citations is intended to facilitate exchanges of more transferable data
among programs.
As an example, in a review of the use of wetlands for stormwater pollution control (Strecker
et al. 1992), a summary of the literature was prepared regarding the performance of wetland
systems and the factors that are believed to affect pollutant removals. The studies reported
in the reviewed literature were inconsistent with respect to the constituents analyzed and the
methods used to gather and analyze data. Several pieces of information were improperly
collected and recorded, which decreased their usefulness for evaluating the effectiveness of
stormwater wetlands. Furthermore, the lack of such basic information limits the
transferability of the studies' findings into better design practices.
The technical literature has many reports of monitoring programs to evaluate BMP
performance. Those that address conceptual and strategic aspects of monitoring (e.g.,
Strecker 1994; Urbonas 1993) could be of particular value during the planning stage. In
addition, EPA and ASCE's Urban Water Resources Research Council have compiled a
National Stormwater Best Management Practices Database (ASCE 1999) (on the world
wide web at http://www.bmpdatabase.org/). The purpose of this effort is to develop a
more useful set of data on the effectiveness of individual BMPs used to reduce pollutant
discharges from urban development. Review of the protocols established for the database
is useful in determining what and how information should be collected.
It is also valuable to review the monitoring methods and findings of other reported programs
because they may contain transferable concepts (or even data). In considering the use of
data collected elsewhere, critical attention must be given to differences that might lead to
erroneous conclusions (e.g., weather, soil types, role of specific sources of pollutants).
Particular care should be taken to avoid errors that are often introduced by assuming (rather
than determining) that certain pollutants are associated with certain sediment fractions.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
48
-------
These associations of pollutants with particles are very important (in fact they are the reason
why most BMPs are effective), but they vary dramatically from place to place and must be
determined based on careful local studies of relevant factors. When reviewing data from
relatively early studies, it is important to remember that state of the art of analyses has
advanced considerably in the past decade or so. For example, many data entries that report
"non-detect" may not be relevant.
3.1.2 Monitoring to Assess Compliance with Surface Water quality criteria
A main objective of BMP monitoring is to determine if the BMP helps reduce
concentrations of constituents of concern and therefore achieves compliance with water
quality criteria set forth by state and federal regulations.
Water quality standards may include bacteria, dissolved oxygen, temperature, pH,
turbidity, and toxic organic and inorganic compounds in marine and freshwater bodies.
The water quality standards for toxic compounds (e.g., metals, pesticides) are intended to
protect aquatic organisms, terrestrial animals, and humans who drink the water and/or
consume shellfish and fish from the waterbody. In addition, the water quality bacterial
standards are intended to guard against human health risks associated with recreational
activities such as swimming, wading, boating, fishing, and shellfish consumption.
State water quality standards often include the federal water quality criteria for the
protection of human health and aquatic life (40 CFR 131.36). Federal water quality
criteria may include a number of additional compounds not listed in state water q
uality standards.
Note that water quality criteria are guidelines, whereas water quality standards are
enforceable regulations. In this section, water quality criteria are used to encompass both
state standards and the federal guidelines.
There are two general categories of water quality criteria: aquatic (or marine) criteria, and
human health criteria. These are summarized below.
3.1.3 Criteria for the Protection of Aquatic/Marine Life
Criteria for the protection of aquatic and marine life were developed based on laboratory
toxicity tests of representative organisms using test solutions spiked with pollutants to
simulate exposure. In order to apply the results of these tests, EPA has classified aquatic
life standards as either "acute" or "chronic" based on the length of time the organisms are
exposed to the listed concentrations.
Criterion maximum concentrations (CMC - acute) are intended to protect against short-
term exposure. Criterion continuous concentrations (CCC - chronic) are designed to
protect against long-term exposure. In deriving the acute criteria, the laboratory
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
49
-------
organisms were exposed to pollutant concentrations for 24 to 48 hours. EPA suggests
one hour as the shortest exposure period, which may cause acute effects and recommends
the criteria be applied to one-hour average concentrations. That is, to protect against
acute effects, the one-hour average exposure should not exceed the acute criteria. EPA
derives chronic criteria from longer term (often greater than 28-day) tests that measure
survival, growth, reproduction, or in some cases, bioconcentration. For chronic criteria,
EPA recommends the criteria be applied to an averaging period of 4 days. That is, the 4-
day average exposure should not exceed the chronic criteria.
water quality criteria for aquatic life were developed based on an allowable exceedance
frequency of once every three years, based on the theory that an ecosystem is likely to
recover from a brief water quality exceedance, provided it does not occur too often.
3.1.4 Human Health
Water quality standards for the protection of human health contain only a single
concentration value and are intended to protect against long-term (chronic) exposure. For
carcinogenic compounds, a lifetime exposure over 70 years is generally used to calculate
the criteria. For non-carcinogens, exposure periods are more chemical specific and
depend on the particular endpoint and toxic effect.
EPA has defined two levels of protection for human health criteria. The first criteria
were derived based on cumulative risks associated with drinking water and eating
organisms that live in the water. The criteria for carcinogenic compounds are the
calculated water-column concentrations that would produce a one in a million (10~6)
lifetime cancer risk if water were consumed by humans and a given amount of organisms,
like fish or shellfish, living in that water was eaten every day. The second set of criteria
is based on consumption of organisms alone (the water is not consumed by humans).
These standards apply to saltwater or other water that is not a drinking water source but
does support a fishery, and that is used as food. The standard for carcinogenic
compounds in the consumption of organisms only criteria is the calculated water
concentration that would produce a one in a million (10~6) lifetime cancer risk if a person
were to consume a given amount of fish or shellfish from that waterbody (without
drinking the water).
3.1.5 Application of Water quality criteria to Stormwater
The water quality criteria are intended to protect the beneficial uses of streams, lakes, and
other receiving water bodies. Most of the man-made conveyances within a near-highway
stormwater drainage system do not support these beneficial uses. Thus, monitoring to
assess compliance with water quality criteria is usually conducted in a receiving water
body (rather than in the stormwater conveyance system that discharges into it) in order to
provide a direct measure of whether the beneficial uses of the waterbody are impaired or
in jeopardy.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
50
-------
Direct comparisons between stormwater quality and the water quality criteria should be
interpreted with caution because the effects of receiving water hardness levels do not
account for mixing and dilution in the receiving waters or for such comparisons on heavy
metals. This is especially true when the stormwater discharge is very small relative to the
receiving waterbody.
The variable nature of stormwater quality further complicates comparison to water
quality standards. Stormwater quality varies both between and during storm events, so it
is very difficult to extrapolate data from one storm to another or to generate statistically
representative data for all types and combinations of storms.
In spite of the limitations mentioned above, comparisons between stormwater quality and
water quality standards can provide valuable information for stormwater management.
Water quality standards can be used as screening criteria, or "benchmarks," for assessing
stormwater quality problems and establishing management priorities. Direct comparisons
with the water quality criteria can over-estimate the potential impact of the stormwater
discharges on the receiving water bodies because mixing and dilution are not taken into
account. However, the relative frequency and magnitude of water quality standards
exceedances within storm sewer systems can help prioritize additional investigations
and/or implementation of control measures. Frequent large exceedances are a clear
indication that further investigation and control measures are warranted. Marginal or
occasional exceedances are more typical and more difficult to interpret.
3.1.6 Groundwater and Sediment Standards
In addition to surface water quality standards, stormwater discharges may affect
compliance with standards for groundwater quality and/or marine sediment quality.
However, stormwater monitoring is typically of limited value with regard to assessing
compliance with groundwater and/or sediment quality standards. Compliance with the
groundwater standards is generally assessed through groundwater monitoring (rather than
stormwater monitoring) because stormwater quality is likely to change substantially
while percolating through soils, and the extent of the change is very difficult to predict
without a great deal of site-specific information. Similarly, compliance with sediment
quality standards is generally assessed through sediment monitoring within receiving
water bodies. This is because numerous storms would need to be monitored in order to
develop useful estimates of total annual sediment loads, and the particulate portion of
each sample would need to be divided into particle size fractions prior to chemical
analysis to allow even a qualitative evaluation of potential sediment transport/deposition.
For these reasons, this manual does not address stormwater monitoring to assess
compliance with groundwater or sediment quality standards.
3.1.7 Scope of Work for BMP Monitoring Program
Once monitoring objectives have been defined, the scope of the monitoring program must
be determined. It is important to balance information needs with the resources available,
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
51
-------
and to consider alternative means for obtaining information. To that end, consider the
following:
How accurate or representative do the monitoring results need to be in order to
support forthcoming management decisions?
If objectives include determination of stormwater quality trends or evaluation of BMP
effectiveness, numerous storms may need to be monitored in order to account for the
variability inherent in stormwater quality data. It can be difficult and expensive to obtain
truly definitive stormwater data. For example, one of the City of Fresno's monitoring
programs (15 storms per year) has a 20% probability of detecting a 20% change in
stormwater quality at a confidence level of 95%. This monitoring program was expected
to cost about $1.55 million over 10 years, which was about 21% of Fresno's total budget
for stormwater management during that period. To attain an 80% probability of detecting
a 20% change at a 95% confidence limit, the monitoring cost would have risen to about
$5.84 million, or 41% of the total stormwater management budget (Harrison 1994).
Note that the BMPs necessary to reduce stormwater contamination from built-out areas
by 20% would probably be costly and challenging to implement. Cave and Roesner
(1994) estimated that typical non-structural BMPs are likely to result in stormwater
pollutant reductions on the order of 5%-10%, while structural measures may reduce some
stormwater pollutants by 50%-90%. They suggested that a fully implemented municipal
stormwater management program is likely to result in pollutant load reductions of 25% or
less for built-out areas. This number, however, has been cited by others to be closer to
40% (Bannerman 2001).
Devoting large amounts of time and money to achieve a high level of accuracy may not
be the best use of stormwater program resources. It might be more cost effective to
spend less on trend monitoring and more on source identification, sediment monitoring,
and/or control measures. In some cases, a simple, screening-type monitoring program
may be sufficient to meet needs.
Are sufficient staff and financial resources available to obtain the needed information
at the desired level of accuracy? If not, can additional resources be obtained?
This is a critical consideration. BMP monitoring is generally expensive and time-
consuming. This question can be addressed by developing an overview of monitoring
required and reviewing general cost information of other programs.
In assessing personnel resources, consider staff size, technical background, physical
condition, and ability (and willingness) to respond to storm events with little advance
notice. These factors are discussed below.
Staff Size. Few organizations can afford to have many personnel whose sole
responsibility is stormwater monitoring. In most cases, monitoring duties are assigned to
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
52
-------
certain people in addition to their regular responsibilities. Back-ups are needed in case
the designated personnel are sick, on vacation, or otherwise unavailable when a storm
monitoring event occurs. The assigned people must be able and willing to drop what they
are doing and mobilize for a storm event on short notice. In some organizations,
personnel are not allowed to perform work that is not specified in their job descriptions.
Insurance and liability may also be considerations. Because of these staffing issues,
some agencies elect to hire contractors to perform monitoring.
Technical Expertise. Some technical expertise is needed to properly conduct monitoring,
especially if automated equipment is used. Special training is required for any personnel
that enter confined spaces, such as manholes, to collect samples. In addition, the person
directing a monitoring program should be familiar with how the results will be used, so
that effective decisions are made regarding storm selection, when to cancel a monitoring
event, etc.
Physical Condition/Health. Stormwater monitoring can be physically demanding.
Monitoring personnel may be required to work in slippery or otherwise challenging
conditions at night.
Ability to Respond to Storm Events. Storms often occur outside of normal working
hours when it is more difficult to contact and mobilize monitoring personnel.
If resources are not sufficient to sample enough storms and/or enough locations to meet
tentatively identified program objectives, monitoring program objectives and scope
should be scaled back until they are commensurate with resources. This can sometimes
be accomplished by using a phased approach where only one or two areas or questions
are addressed at a time so that useful results can be obtained within budget limitations.
Supplementing existing resources should also be considered. It may be worthwhile to
contact neighboring municipalities or facilities to find out if they are willing to pool their
resources in order to fund a joint BMP monitoring program. If objectives cannot be met
with the available resources, possible alternatives to stormwater monitoring should be
considered (discussed below), or monitoring resources should be allocated to additional
pollution control measures.
Can some of the information needed be obtained without conducting BMP
monitoring?
Because of the typically high cost of BMP monitoring, it may be desirable to evaluate
alternative means for addressing some information needs (assuming that BMP monitoring
is not required to comply with a permit). Depending on the situation, sediment sampling,
biological sampling, and/or visual surveys of the stormwater conveyance system may be
cost-effective alternatives to stormwater quality monitoring. Literature reviews may also
help address some stormwater management issues.
Who is going to use the monitoring data and what is the intended use?
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
53
-------
Develop specific monitoring objectives and scope based on answers to these questions.
At this point, the objectives should still be considered flexible because they may need to
be re-considered and revised as the monitoring program is developed.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
54
-------
3.1.8 Information Needs to Meet Established Goals of BMP Monitoring
Generally, the more information that is available, the easier it is to design a practical
monitoring program. For BMP monitoring programs, compile and review the following
information, if available:
Results from prior surface water and groundwater quality studies, other BMP
monitoring studies in the local area, sediment quality studies, aquatic ecology
surveys, dry weather reconnaissance, etc.
Drainage system maps.
Land use maps (or general plan or zoning maps).
Aerial photographs.
Precipitation and streamflow records.
Reported spills and leaks.
Interviews with public works staff.
Literature on design of structural BMPs to understand functionality and pollutant
removal processes.
For BMPs monitored in industrial areas, the following information may also be relevant:
BMP performance data for similar industries in region.
Facility map(s) showing locations of key activities or materials that could be exposed
to stormwater.
Lists of materials likely to be exposed to stormwater.
Reported spills and leaks.
Interviews with facility staff and others who are knowledgeable about the facility.
In addition to gathering information about the study area and BMP design, some
forethought should be given to the expected data characteristics and subsequent data
analysis methods in order to optimize collection of data within the limitations of the
proposed study and ensure that useful results will be provided to fulfill study objectives
(Caltrans 1997).
Essential data characteristics include the type of data to be collected (e.g., constituents
and concentrations), the variables affecting the data (e.g., antecedent conditions, rainfall
intensity, site type and location) and the expected variability of the data (derived from
previous studies when available). Statistical techniques such as power analysis can then
be used to determine key study parameters, such as the number of monitoring locations
and storm events to be monitored (Caltrans 1997).
Prior to the initiation of environmental sampling, a strategy should be developed for
analysis of the data, directed to answering the specific study questions. The selected data
analyses technique(s) may influence the types and quantities of data required to satisfy
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
55
-------
study objectives. The analysis methods applied to data collected for BMP evaluations or
characterization studies typically involve straight-forward statistical operations.
3.2 Phase II - Develop BMP Monitoring Plan
3.2.1 Recommendation and Discussion of Monitoring Locations
The number of locations to be monitored depends on program objectives, permit
requirements (if applicable), the size and complexity of the drainage basin(s), and the
resources (time, personnel, funds) allocated to monitoring. In addition, the frequency of
sampling at each location must be considered. Depending on objectives, resources, and
logistical considerations, many locations may be sampled infrequently, or fewer locations
more frequently. The former approach is generally better for evaluating place-to-place
variability; the latter approach is generally better for evaluating storm-to-storm variability
and for characterizing the monitoring location more accurately. If the effectiveness of a
specific structural BMP needs to be evaluated, monitoring locations should be located
immediately upstream and downstream of the structure.
In general, choose monitoring sites that facilitate representative sampling and flow
measurement. Consider the criteria listed below in the selection of monitoring sites:
The contributing (upgradient) catchment should be completely served by a separate
storm drain system or, if it is served by a combined sewer system, carefully consider
the possibility that stormwater samples would be contaminated by sanitary sewage.
The storm drain system should be sufficiently well understood to allow a reliable
delineation and description of the catchment area (e.g., geographic extent,
topography, land uses).
For monitoring stations that will be used to measure flow in open channels, the flow
measurement facilities need to be located where there is suitable hydraulic control so
that reliable rating curves (i.e., stage-discharge relationships) can be developed. In
other words, the upstream and downstream conditions must meet the assumptions on
which the measurement method is based.
Where possible, stations should be located in reaches of a conveyance where flows
tend to be relatively "stable" and "uniform" for some distance upstream
(approximately 6 channel widths or 12 pipe diameters), to better approach "uniform"
flow conditions. Thus, avoid steep slopes, pipe diameter changes, junctions, and
areas of irregular channel shape due to breaks, repairs, roots, debris, etc.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
56
-------
Locations likely to be affected by backwater and tidal conditions should be avoided
since these factors can complicate the reliable measurement of flow and the
interpretation of data.
Stations in pipes, culverts, or tunnels should be located to avoid surcharging (pressure
flow) over the normal range of precipitation.
Stations should be located sufficiently downstream from inflows to the drainage
system to better achieve well-mixed conditions across the channel and to favor the
likelihood of "uniform" flow conditions.
Stations should be located where field personnel can be as safe as possible (i.e.,
where surface visibility is good and traffic hazards are minimal, and where
monitoring personnel are unlikely to be exposed to explosive or toxic atmospheres).
Stations should be located where access and security are good, and vandalism of
sampling equipment is unlikely.
Stations should be located where the channel or storm drain is soundly constructed.
If an automated sampler with a peristaltic pump is to be used, and the access point is a
manhole, the water surface elevation should not be excessively deep (i.e., it should be
less than 6 meters, or 20 feet, below the elevation of the pump in the sampler, and
preferably less than 4.5 meters or 15 feet deep).
If automated equipment is to be used, the site configuration should be such that
confined space entry (for equipment installation, routine servicing, and operation) can
be performed safely and in compliance with applicable regulations.
Each potential sampling station should be visited, preferably during or after a storm to
observe the discharge. A wet-weather visit can provide valuable information regarding
logistical constraints that may not be readily apparent during dry weather.
Integration of BMP Monitoring into a Municipal Monitoring Program
In most cases, it is not practical to monitor water quality at every BMP within a
municipality. Therefore, most municipal monitoring programs are designed to yield
estimates of effluent water quality for other similar BMPs by extrapolating data collected
at a small number of locations.
Many municipal stormwater monitoring programs use stations that monitor relatively
small, homogeneous land use catchments (so called "single land use" or upland stations).
Data from a study site may then be extrapolated to other catchments within the project
area that are thought to have similar sources and pollutant-generating mechanisms. This
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
57
-------
approach may also be useful for BMP monitoring studies. However, extrapolations
should be interpreted with caution because it is difficult to ascertain the degree to which
catchments and BMP functionality are truly similar. Also, previous studies have shown
that stormwater quality within a given land use category can vary considerably; thus, the
correlation between land use and stormwater quality, and thus the utility of a particular
BMP, may not be as strong as is typically assumed.
Other municipal programs use stations that sample relatively large catchments
representing a composite of land uses. These stations are typically located in streams or
other stormwater conveyances at the lower end of a watershed and are sometimes
referred to as "mixed land use" stations or "stream stations." If possible, choose stream
stations that receive runoff from catchments with a land use composition similar to that of
the project area as a whole. This will make it easier to apply BMP monitoring results to
similar watersheds. A geographic information system (GIS) can be very helpful in
characterizing land uses and identifying stormwater monitoring locations.
Care must be taken to locate flow measurement and sampling sites in places that are likely
to yield good data over diverse operational conditions. For performance monitoring
approaches that are intended to compare changes in pollutant loads (i.e., "loads in" versus
"loads out" of the BMP), it is especially important to use accurate flow measurement
methods and to site the points of measurement at locations that maximize the attainment of
credible data (see Section 3.2.1). The added cost of a weir or flume, as opposed to less
sophisticated flow measurement methods, is almost always worthwhile because
measurement errors propagate through various aspects of the analysis. Propagation of errors
due to inaccurate measurement is discussed in detail in Section 3.2.4.3.
It is often difficult to identify large, homogeneous land use catchments that satisfy all of
the above criteria. As a result, compromises will typically need to be made. Refer to
basic texts on hydraulics and flow measurement and the instructions provided by
monitoring equipment manufacturers to guide judgment.
Sampling from a Well Mixed Location
The location of a permanent sampling station is probably the most critical factor in a
monitoring network that collects water quality data. If the samples collected are not
representative of the water mass, the frequency of sampling as well as the mode of data
interpretation and presentation becomes inconsequential. The following paragraphs
describe the theory of mixing within a river cross-section, which is applicable to
stormwater flows within stormwater conveyance systems. Typically these calculations
are not needed for stormwater monitoring design, but they are presented here to bring
attention to the need to be aware of mixing problems, particularly in wide conveyances.
(Saunders 1983)
The representativeness of a water quality sample is a function of the uniformity of the
sample concentrations in a river's cross sectional area. Wherever the concentration of a
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
58
-------
water quality variable is independent of depth and lateral location in a river's cross
section, the river at that point is completely mixed and could serve as a desirable
sampling location (Saunders 1983).
Well mixed zones in a river for representative water quality sampling can be defined,
given that several assumptions will apply. By assuming that a pollutant distribution from
an instantaneous point source is normally distributed on both the lateral and vertical
transect and applying classical image theory, a theoretical distance from an outfall to a
well mixed zone in a straight uniform river channel is a function of 1) mean stream
velocity, 2) location of the point source and 3) the mean lateral and vertical turbulent
diffusion coefficients (Saunders 1983).
There are several models available that are functions of the mixing coefficients, which
have been shown to apply for predicting a zone of relatively complete mixing. Ruthven
(1971) derived an expression for a mixing distance utilizing the solution to the steady-
state, two-dimensional advection and dispersion equation. Assuming that complete
vertical mixing is assured in a relatively short distance, he established a relationship from
the two-dimensional solution to predict the mixing distance to a point where the
concentration variation in the cross section does not exceed ten percent. The approach
taken by Ruthven is shown in the following equation:
Equation 3.1
where,
L: mixing distance
w: width of channel
u: mean stream velocity
Dy: lateral turbulent diffusion coefficient
The distance needed for complete mixing using the above approach results in great
distances for most situations. In addition, many upstream discharges normally exist and it
is rarely possible to get far enough below all of them. Because of the distance required
for complete mixing, there is often a need to composite samples across wide streams.
Extensive discussion on this subject can be found in Fischer et al. (1979).
3.2.1.1 Upstream
Monitoring stations established upstream of a BMP can give results that reveal the
influent concentration or load of pollutants before they flow through the BMP. Upstream
water quality is indicative of concentrations and pollutant loads that would be observed
downstream if no BMP were implemented. It is important to monitor only waters that
flow into the BMP to be able to use the resultant data to compare upstream water quality
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
59
-------
with downstream locations. Upstream monitoring locations can also be useful to
determine bypass water quality. Where bypass is present, accurate flow measurement is
highly important. Where sufficient funds are available and the physical layout of the
control structures allow, bypass and flow to the BMP should be monitored directly. In
situations where direct measurement is not practical, modeling of bypass flows can be
substituted, particularly where the hydraulics of the bypass structure are well known or
can be calibrated to flow rates. Typically a mass balance approach is used to model
bypass flow rates and volumes.
Upstream monitoring stations should be located far enough away from the BMP to ensure
that samples are independent of the BMP. Immediately upstream from a BMP,
contributing runoff could be affected by backflow, slope, vegetation, etc. Upstream
monitoring should be representative of conditions that existed before the BMP was
implemented.
3.2.1.2 Downstream
Monitoring stations established downstream of a BMP can indicate water quality of flows
that are treated by the BMP. Downstream monitoring is essential for establishing:
That the BMP provides a measurable and statistically significant change in water
quality.
That the BMP provides effluent of sufficient quality to meet water quality criteria.
A comparison of effluent concentrations with similar BMPs to determine if the BMP
is achieving typical effluent water quality.
Monitoring stations should be located immediately downstream so that BMP effluent is
sampled before it is introduced into the receiving waters or is exposed to factors that may
affect constituent concentrations. Where bypass is present and one wants to understand
the efficiency of the BMP in addition to the BMP system, it is important to monitor water
quality of the bypass flows and the effluent separately. In some cases where influent
water quality is not expected to be appreciably different than bypass water quality,
upstream data may be used to determine water quality. This approach does not, however,
obviate the need for accurate estimates of bypass flow rates and/or volumes from
monitoring or flow modeling. In some cases, bypass flows may be very difficult to
separate from treated effluent (e.g., in hydrodynamic devices).
3.2.1.3 Intermediate Locations
BMPs are often designed as a group of devices or chambers that target specific processes.
For example, a filter might have a settling chamber to quickly remove large settlable
solids before flowing into the filter media chamber. A treatment train approach is
sometimes taken to combine various BMPs in order to maximize removal of specific
constituents. Intermediate monitoring locations in the interior of the BMP are useful for
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
60
-------
investigating how various sections of the BMP are working and establishing mid-BMP
concentrations. Monitoring stations are also useful in between treatment train BMPs to
assess effectiveness of each individual BMP in addition to monitoring
upstream/downstream stations to determine overall BMP efficiency.
Intermediate monitoring locations should be located either interior to the BMP or in
between BMPs linked in a treatment train. For interior monitoring, such as in the middle
of a wetland or detention pond, stations should be established in a location that is
representative of the BMP. For example, monitoring within a wetland should be done in
the middle section, where the slopes, vegetation, channel width, etc., are uniform and
similar to the rest of the wetland, avoiding any microcosms of unique vegetation, basins,
or slopes. To monitor in between treatment train BMPs, stations should be established to
capture effluent from the upstream BMP or inflow to the downstream BMP, or both.
Monitoring should not be conducted in a place where backflow or mixing is occurring, as
these processes do not allow for isolated sampling of direct BMP discharge or inflow.
During high flow conditions, this may be difficult because many BMPs overflow,
reducing the distinction and separation between BMPs. Intermediate treatment train
BMP monitoring stations need to be carefully evaluated to determine if samples taken
during high flows are representative of water quality of flow between the BMPs and not
backflow or some other phenomena.
3.2.1.4 Rainfall
Rainfall monitoring can be an essential piece of the monitoring puzzle. Rainfall data may
help determine when to start sampling as well as provide information to calculate rainfall
characteristics such as intensities. The importance of accurate rainfall data, however,
decreases as the accuracy and reliability of flow information is improved. Rainfall data
are relatively inexpensive to collect and therefore, even in cases where rainfall data may
not be required for a detailed analysis of BMP efficiency, it is usually worthwhile to
monitor for validation of flow monitoring results.
Site Proximity
Rainfall gauges should be established as close as possible to the monitoring stations. In
many regions, rainfall is highly variable within a small area due to orographic effects,
elevation, and proximity to water bodies. The US Geological Survey, National Weather
Service, and many municipalities have networks of rain gauges, some with real-time rain
data available over the Internet. These established stations are convenient to use if they
are in close proximity to the monitoring site, or as a general estimate of rainfall if they are
not in close proximity to the monitoring site.
Rain gauges may need to be installed near the site to obtain accurate rainfall data where
established gauges are not available. Proper installation and maintenance of the rain
gauge is as important as gauge proximity to the monitoring site. Installation of rain
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
61
-------
gauges is often a straightforward matter. Manufacturers provide guidelines on the
appropriate mounting of the devices. The main concerns during installation are:
Leveling the device.
Making sure that vegetation (trees) or structures are not obstructing rainfall.
Providing enough height above the ground to prevent vandalism.
Locating the rain gauge in close proximity to other monitoring equipment to provide
required connections for recording of rainfall depths and/or representative records.
Number of Gauges
The number of precipitation gauges installed in a system directly affects the quality of
precipitation data. Generally, the higher the number of precipitation gauges, the better
the estimate of precipitation amounts. Locating a gauge at each monitoring site for small
catchments is imperative, because local variations in total rainfall and rainfall intensity
can have significant effects on runoff when the watershed is minimal in size. Nearby
locations may not be useful in estimating rainfall at the actual site.
In addition to the network of rain gauges accessed for monitoring, it is also useful to
install manual rain gauges at the monitoring site to check accuracy, consistency and
proper functioning among different gages. It is not difficult to discover a gauge that
produces different rainfall data than that observed at the site due to the location of the
gauge at a different elevation or microclimate, improper installation or placement, or
natural interferences (birds resting on the gauge, for example).
3.2.1.5 Groundwater
Although most BMPs are designed to treat surface water runoff, some BMPs also
promote groundwater infiltration. BMPs incorporating infiltration should not process
large quantities of certain constituents (petroleum products, pesticides, solvents, etc.) that
could be mobilized in groundwater or pose a drinking water hazard to those who rely on
downstream wells.
Groundwater monitoring wells should be established if contamination of groundwater is
suspected. Groundwater flow, direction and elevation as well as soil types should be
established before monitoring sites are chosen. Monitoring stations should be located
sufficiently down gradient from the BMP where infiltrated water from the BMP is
accessible. A series of monitoring stations could be established: a station upstream of the
BMP, one a short distance downstream from the BMP, another a longer distance
downstream, and another even further downstream from the BMP. This will indicate if
there is any contribution of constituents to the groundwater from the BMP, and where
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
62
-------
there is a contribution, if the concentrations decrease with increasing distance from the
BMP.
3.2.1.6 Sediment Sampling
Many constituents either settle out of the water column or prefer not to be in the water
column (due to hydrophobicity) and become incorporated in the sediment. Sediment can
store significant amounts of certain constituents, such as BTEX, PCBs, metals, and
microbes. During high flows, these sediments are stirred up and can release their
potentially high concentrations of accumulated constituents. Many BMPs are designed to
remove the sediment from runoff, theoretically removing the associated constituents as
well.
Sediment sampling can determine concentrations of constituents not necessarily found
through water column monitoring. Sediments can be sampled upstream and downstream
of BMPs as well as internal to the BMP to assess removal and effluent efficiencies as
well as internal accumulation of sediment and associated constituents.
When sampling for suspended sediments in the water column, it is important to take the
sample well below the surface of the water, ideally in the middle portion of the water
column where the average concentration of suspended sediment is found. When
sampling sediment from the creek bed or internal to the BMP (e.g., sampling the filter
media or detention pond bottom sediments) sediments should be collected minimizing
disturbance or resuspension of the sediment bed so that the original settled material is
captured in the sample apparatus. Depth of sediment sample should also be noted as
constituent concentrations can vary with depth.
3.2.1.7 Dry Deposition
Many constituents are quite volatile, including mercury, BTEX, PCBs, and some
pesticides. Atmospheric deposition has been pointed to as a significant source of certain
constituents to water bodies in some areas. These constituents are continuously being
deposited out of the atmosphere either by coming into contact with the surface and
sorbing to it, settling out of the air, or through rainfall. Constituents are deposited onto
surfaces, such as roads, rooftops, and driveways and then incorporated into runoff during
storm or low flow events. Therefore, atmospheric deposition may contribute some
material to those BMPs that are exposed to the atmosphere, such as detention ponds and
wetlands.
In order to assess the contribution of atmospheric deposition to constituent concentrations
and to isolate influent and effluent concentrations, dry deposition can be monitored in
conjunction with BMP monitoring. Pans can be set out near BMPs to capture dry
deposition of these volatile constituents much in the same way that rainfall gauges are
installed to capture rainfall. After a period of time the deposited material can be analyzed
to determine constituent concentrations. It is recommended that dry deposition sampling
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
63
-------
should only be conducted as a follow-up investigation where sufficient evidence indicates
that dry deposition may be contributing appreciably to stormwater pollution.
It is important to note that very little of the total watershed dry deposition actually
contributes to stormwater runoff. The only contributions to water quality impairment that
currently can be directly attributed to dry deposition fall on the receiving waters
themselves (such as PCBs and DDT measurements for the Great Lakes) (Pitt 2001).
Otherwise, most is incorporated in soils or may not wash off paved areas during rains.
Fugitive dust from nearby sources is usually comprised of relatively large material that is
poorly washed off, while particulates from regional air pollution sources (particularly
power generation and autos) are mostly very small and are typically incorporated in soils;
however, these smaller particles are much more easily washed off from pavements and
might be a quantifiable source of pollutants where depositional rates are relatively large
compared to other sources.
3.2.1.8 Modeling Methods
When monitoring is not feasible due to a limited budget or lack of sampling staff,
estimates of water quality parameters, flow, and rainfall can be made using various
models and assumptions. The use of modeling to estimate these parameters may limit
usability of the data depending on the validity of the assumptions made, the accuracy of
the model itself, and accuracy of the information input into the model.
Estimates of Water Quality Parameters
Certain water quality parameters can be estimated by monitoring instead for related
parameters that are simpler or less expensive. These related or surrogate parameters are
statistically correlated to the more complicated or expensive parameters. Some common
surrogate parameters and represented parameters are:
Surrogate Parameter Parameter Represented by Surrogate
Turbidity TSS
Fecal Coliform Pathogens
Chemical oxygen Demand (COD) Biological Oxygen Demand (BOD)
In addition to monitoring for surrogate parameters at each monitoring site, water quality
models can be used to estimate constituent concentrations at monitoring sites using
available monitoring data, upstream land use, hydrology, geology, and history to
calculate a mass balance for each constituent. Water quality models are a tool for
simulating the movement of precipitation and pollutants from the ground surface through
pipe and channel networks, storage treatment units, and finally to receiving waters. Both
single-event and continuous simulation may be performed on catchments having storm
sewers and natural drainage for prediction of flows, stages and pollutant concentrations.
Each water quality model has its own unique purpose and simulation characteristics. It is
advisable to thoroughly review downloading and data input instructions for each model.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
64
-------
The applicability and usefulness of these models is dependent upon a number of
assumptions. The degree of accuracy of these assumptions determines the usefulness of
the output data. For example, one assumption could be based on certain types of land use
contributing certain constituents to the catchment runoff. The constituents associated
with each land use have been well studied by many monitoring programs, but are still
highly variable, depending on specific activities on each parcel, history of spills, age of
infrastructure, climate, and many other factors. Although modeling of water quality
parameters is a useful tool to estimate parameter concentrations, model results should not
be interpreted as exact data. Confirmation of water quality model results should be done
by monitoring a few storms and/or a few sites, then running the model with the observed
conditions as input variables and comparing the results.
A variety of modeling tools are available for modeling water quality these include, but
are not limited to, the following:
Enhanced Stream Water Quality Model, Windows (QUAL2E)
Simulates the major reactions of nutrient cycles, algal production, benthic and
carbonaceous demand, atmospheric reaeration and their effects on the dissolved oxygen
balance. It is intended as a water quality planning tool for developing total maximum
daily loads (TMDLs) and can also be used in conjunction with field sampling for
identifying the magnitude and quality characteristics of nonpoint sources.
AQUATOX: A Simulation Model for Aquatic Ecosystems
AQUATOX is a freshwater ecosystem simulation model. It predicts the fate of various
pollutants, such as nutrients and organic toxicants, and their effects on the ecosystem,
including fish, invertebrates, and aquatic plants. AQUATOX is a valuable tool for
ecologists, water quality modelers, and anyone involved in performing ecological risk
assessments for aquatic ecosystems.
SWMM: Storm Water Management Model
The EPA's Storm Water Management Model (SWMM) is a large, complex model
capable of simulating the movement of precipitation and pollutants from the ground
surface through pipe and channel networks, storage/treatment units, and finally to
receiving water. Both single-event and continuous simulation may be performed on
catchments having storm sewers, combined sewers, and natural drainage for prediction of
flows, stages and pollutant concentrations (EPA 1995). See
http://www.ccee.orst.edu/swmm/ for more information on this model.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
65
-------
HSPF: Hydrologic Simulation Program - Fortran
The HSPF Model is an EPA developed application for simulation of watershed hydrology
and water quality. The HSPF model uses historical rainfall, temperature and solar
radiation data; land surface characteristics such as land use patterns; and land
management practices to simulate the processes that occur in watersheds. The result of
this simulation is a continuous recreation of the quantity and quality of runoff from urban
or agricultural watersheds. Flow rate, sediment load, and nutrient and pesticide
concentrations are predicted. The HSPF model incorporates the watershed-scale
Agricultural Runoff Model (ARM) and Non-Point Source (NFS) models into a basin-
scale analysis framework that includes pollutant transport and transformation in stream
channels.
WASPS: Water Quality Analysis Simulation Program
The Water Quality Analysis Simulation Program (WASP) is a generalized framework for
modeling contaminant fate and transport in surface waters. WASPS is the latest of a
series of WASP programs. Based on the flexible compartment modeling approach,
WASP can be applied in one, two, or three dimensions. WASP is designed to permit easy
substitution of user-written routines into the program structure. Problems that have been
studied using the WASP framework include biochemical oxygen demand and dissolved
oxygen dynamics, nutrients and eutrophication, bacterial contamination, and organic
chemical and heavy metal contamination (James 2001).
SLAMM: Source Loading and Management Model
The Source Loading and Management Model (SLAMM) was developed to assist water
and land resources planners in evaluating the effects of alternative control practices and
development characteristics on urban runoff quality and quantity. SLAMM only
evaluates runoff characteristics at the source areas In the watershed and at the discharge
outfall; it does not directly evaluate receiving water responses. However, earlier versions
of SLAMM have been used in conjunction with receiving water models (HSPF) to
examine the ultimate effects of urban runoff.
SLAMM is different from other urban runoff models. Beside examining land
development practices and many source area and outfall control practices, it contains two
major areas of improvements. These are corrected algorithms for the washoff of street
dirt and the incorporation of small storm hydrology. Without these corrections, it is not
possible to appropriately predict the outfall responses associated with source area
controls and development practices. (James 2001)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
66
-------
Estimates of Flow
Flows entering and leaving a BMP may be useful to model if actual monitoring is
prohibitive. Flow can be estimated at varying levels of detail using approaches ranging
from simple spreadsheets to complex hydraulic simulations of extensive urban drainage
networks. Many of the water quality models presented in the previous section are also the
best choices for modeling flows.
The simplest approach is to use the volumetric runoff coefficient approach described
below.
Volumetric Runoff Coefficient
The Volumetric Runoff Coefficient is an empirical relationship that provides an estimate
of total volume of runoff based on total volume of rainfall according to the following
equation:
Volume of Runoff = Volume of Rainfall x Rv - Depression Storage
where,
Rv: Volumetric Runoff Coefficients
This method is usually applied to smaller catchments such as parking lots, rather than
entire watershed areas.
Where monitoring data have been collected for some calibration period such that an
accurate estimate of the volumetric runoff coefficient and depression storage for the
watershed can be made, this approach coupled with accurate rainfall data may provide
one of the least expensive methods for determining total volume of flow from a
watershed on a storm-by-storm basis.
Estimates of Rainfall
If a nearby rainfall gauge is not available, rainfall at the monitoring site can be
approximated using available gauges that are located as close as possible and at similar
elevation. A network of gauges in an area can be analyzed to relate latitude, longitude,
and elevation to rainfall. The grid of gauges can be expanded and extrapolated to an area
lacking any gauges, provided that enough rainfall gauges exist.
Although raw rainfall data are often sufficient for monitoring needs, statistical evaluation
of the data is often more useful. For example, if rainfall is needed to estimate runoff,
most of the rainfall less than 0.1 inch will infiltrate into the ground and not produce any
runoff. These small events could be eliminated from the data set to allow for a more
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
67
-------
accurate account of actual runoff. Two statistical analysis tools used extensively in
separating and filtering continuous rainfall records, include:
SYNOP
SYNOP is a statistical rainfall analysis program that converts hourly data into descriptive
statistics for individual storm events and provides annual rainfall statistics. The program
takes an hourly precipitation record from a station, organizes the data into rainfall events,
and computes the statistics of the storm event parameters. When a complete hourly
record has been organized into a sequence of individual storm events, the mean and
standard deviation may be determined for each of the event parameters (EPA 1989).
SWMM
The SWMM model will conduct a complete statistical analysis almost identical to the
SYNOP tool. In most cases, SWMM is the preferred analysis tool as it is based on the
same basic approach as SYNOP and it lacks some minor bugs present in SYNOP.
3.2.2 Recommendation and Discussion of Monitoring Frequency
The number of storms to be monitored each year (i.e., monitoring frequency) is an
important consideration in planning your monitoring program. Budget and staff
constraints generally limit the number of storms, locations, and parameters to be
monitored. Program objectives should be weighed in light of available resources to
determine the best mix of monitoring frequency, locations, and parameters.
The cost of learning more (i.e., conducting more intensive monitoring) should be compared
to the cost implications of moving forward too far and implementing extensive controls
before having learned enough to guide planning, stormwater management commitments,
and/or negotiations with regulatory agencies. The cost of controlling unimportant pollutants
and/or unimportant sources, or implementing ineffective BMPs could easily exceed the cost
of monitoring to learn more about actual BMPs' performance under the conditions that
prevail in the system. Clearly, there is a need for balance here, because endless studies
should not be substituted for control actions.
In general, however, many measurements (i.e., many samples during many events) are
necessary to obtain enough data to be confident that actual BMP performance not just
"noisy data" (e.g., variability artifacts caused by external factors, equipment and operator
errors). Consequently, BMP effectiveness studies can be expensive and time-consuming.
3.2.2.1 Statistical Underpinnings of Study Design
Four factors influence the probability of identifying a significant temporal and/or spatial
change in water quality:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
68
-------
1) Overall variability in the water quality data.
2) Minimum detectable change in water quality (difference in mean concentration).
3) Number of samples collected.
4) Desired confidence level from which to draw conclusions.
Statistical analysis may be conducted to estimate how many events need to be monitored
to achieve a desired confidence in a conclusion (i.e., power analysis). Performing a power
analysis requires that the magnitude of detectable change, the confidence level, and the
statistical power or probability of detecting a difference are defined. Typically, the
confidence level and power are at least 95% and 80%, respectively, meaning that there is a
5% probability of drawing an incorrect conclusion from the analysis and a 20% probability
that a significant change will be overlooked.
The power analysis often shows that many samples are needed to yield a power of 80% to
90% (i.e., discern a small change). In fact, Loftis et al. (2001) report that achieving a
power of 80% requires double the data required for a power of 50%, and a power of 90%
requires triple the data required for a power of 50%. The exponential increase in data
required to achieve higher statistical power reinforces the need for careful consideration of
the minimum detectable change required (and amount of data required) to achieve project
objectives. In some cases, project objectives require quantification of small changes in
concentration (e.g., inefficient BMPs or BMPs receiving relatively clean influent), which
may call for larger power, but in many cases, less power (i.e., few samples) may be
sufficient. If available resources prohibit the frequent monitoring of all locations, then
reducing the number of locations or parameters tested may provide sufficient data to
resolve slight differences in concentration at a more reasonable cost. Statistical
confidence in the results of the monitoring program(collecting samples from a significant
number of events) should be assigned a higher importance than collecting information
from a larger number of locations or testing a multitude of water quality parameters.
3.2.2.2 Factors Affecting Study Design
Based on a review of existing studies, it is apparent that much BMP research in the past
has not considered several key factors. The most frequently overlooked factor is the
number of samples required to obtain a statistically valid assessment of water quality.
This section focuses on estimating the number of samples required prior to beginning
monitoring activities.
Number of Samples
Stormwater quality may vary dramatically from storm to storm. Therefore, monitoring a
large number of storms is required if the objective of the program is to obtain accurate
estimates of stormwater pollution in a given catchment (e.g., to determine whether water
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
69
-------
quality is changing over time or whether a given BMP is effective). However, staff and
budget constraints typically limit monitoring to either a limited sampling methodology
incorporating a smaller set of parameters for many storms, or a more detailed monitoring
approach including a larger set of parameters for a few storms.
Determining the Number of Observations Needed
Typically a large portion of the costs associated with conducting a BMP monitoring
program are related to collection and analysis of water quality samples. It is imperative
that samples are not only collected in a manner consistent with the guidelines, but also
that an adequate number of samples are collected for statistical validation. Estimates of
the number of samples required to yield statistically valid monitoring results are also
useful for making decisions about the nature and extent of monitoring efforts prior to
implementation. Often goals for a monitoring effort (e.g., to demonstrate that a specific
BMP is achieving a given level of removal of a constituent) may not be consistent with
fiscal limitations of the project. This section provides a method for estimating the
number of samples required for obtaining a statistically valid estimate of both the mean
event mean concentration at a single sampling station and the percent difference observed
at two stations.
As mentioned above, four factors affect predictions as to whether a sampling program
will collect an adequate number of samples to provide a useful estimate of the mean
station EMC:
1) Allowable level of error in estimates of mean (i.e., variance)
2) Level of statistical confidence in estimates of the mean
3) Number of samples collected
4) Variability in population trends
A variety of methods are available for estimating the number of observations required to
predict the range surrounding a sample mean that contains the population mean. EPA
(1993b) presents a nomograph relating the coefficient of variation (COV, defined as the
ratio of the sample standard deviation to the sample mean) to the allowable error in the
estimate of the population mean as a fraction of the sample mean. This nomograph is
given in Figure 3.1 for normally distributed data and a statistical confidence of 95%.
Figure 3.1 can be generated using Equation 3.2 below. The number of samples required
(n) is a function of the allowable error in the data mean (E) and the standard deviation (s),
(or in the case of Figure 3.1, the COV) (Cochran 1963).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
70
-------
where,
(Ef
n: number of samples
s: sample standard deviation
E: allowable error in the data mean
Equation 3.2
This approach is useful for estimating the number of samples required when sampling at
a single location where an acceptable upper bound for the error is known. However,
Equation 3.2 does not provide an estimate of the number of samples required to
determine if the mean concentrations from two sample sets are statistically significantly
different.
o
a
Figure 3.1:
as ซ
Nomograph relating coefficient of variation of a sample set to the
allowable error in the estimate of the population mean (Pitt 1979).
Consideration of the number of samples required to draw statistically significant
conclusions from data is often ignored until after monitoring work has been completed.
However, there is great benefit to performing this analysis before initiating a monitoring
program, particularly where the variability of the data is expected to be quite high
because resources may be better spent on control measures than verification of BMP
efficiency.
Appendix C expands the approach described in EPA (1993b) to the analysis of the
number of samples required to conclude that there is a statistically significant difference
between means calculated from sample data selected at random from two populations.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
71
April 25, 2002
-------
Appendix C provides a straightforward method for estimating the number of samples
required to determine, with some degree of confidence, that observed means (such as the
EMCs resulting from a BMP monitoring program) are statistically significant.
One assumption of the approach provided in Appendix C is that measured influent and
effluent concentrations are normally distributed having a mean equal to the EMC.
Collection of water quality sample data at the inflow and outflow of a structural BMP
allows for the determination of a mean EMC and the variance of the data (or log-mean
and log-variance for log-normally distributed data). The mean and variance (square of
the standard deviation) are the first and second moments of the distribution, respectively.
These moments completely describe a normal distribution; thus, using the mean and
variance of the distribution corresponding to any probability can be determined.
Additionally, probabilities are additive so that confidence intervals between any two
probabilities can be determined simply by calculating values of the distribution
corresponding to the upper and lower probabilities of the confidence interval (i.e.,
confidence limits). The most common application is to determine the range of values
surrounding the mean that falls within a specified 95% confidence interval (i.e.,
probabilities of 2.5% and 97.5%, which are the mean plus/minus 1.96 times the standard
deviation).
One test that can be used to evaluate whether the means of two data sets (e.g., influent
and effluent) are statistically different is a hypothesis test (e.g., student t-test), which is
basically a test that quantifies the overlap of two confidence intervals surrounding the
mean. The mean values will be considered different if there is little (as defined by the t-
statistic distribution) overlap between the confidence intervals. This document presents
hypothesis testing with the assumption that data sets are large (i.e., are composed of 30 or
more values). Given this assumption, the Z-statistic can be used in place of the t-statistic,
which eliminates the need to incorporate the degrees of freedom of a data set into
hypothesis analysis. However, for analysis of small data sets, users should use the t-
statistic in place of the Z-statistic (and refer to the student t-test in a standard statistics
text). An iterative solution is required to determine the number of samples needed if the t-
statistic, due to its dependence on the number of measurements, is used in place of the Z-
statistic (Gilbert 1987).
The confidence interval about the mean for normally distributed data is defined as:
Equation 3.3
C ~ Z -= C + Z ~
where,
C = mean concentration
a = standard deviation for the population of the concentrations
Zo/2 = Z-statistic obtained from a standard normal distribution table
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
72
-------
n = number of measurements
The confidence interval corresponds to a significance level (a), where (1 - a)xlOO% is
the probability that Cwill fall within the confidence interval. As a increases, the
confidence interval will become larger (all other variables remaining the same). If the
population standard deviation (a) is unknown, which is typically the case, then a can be
estimated using the sample standard deviation (s). Prior to the collection of field data, the
standard deviation typically is estimated from existing data sets either from local or
nationally published data on expected quality of stormwater runoff.
The confidence interval is often used to show the likely range containing the population
mean, and for comparing the means for two populations (i.e., influent and effluent).
However, the confidence interval formula contains the number of samples in the data
sets, and therefore the equation can be solved for the number of samples needed to
achieve a desired confidence interval for an expected difference in population means.
The derivation of this formula is provided in Appendix C. As Appendix C shows, the
resulting equation is (see the appendix for variable definitions):
n =
Za/ + Z,,/ xCOFx(2-%removal)
^removal
Equation 3.4
w = 2 [(Zi_a + ZI_P)/(HI -H2)]V Equation 3.5
This assumes that the sample sets have identical n, COV, Zo/2, and Zi_p. Assuming the
COVs of the sample sets are equal is a significant assumption because it mandates that
Sin/Sout equals Cin/Cout . This assumption allows for the generation of a simple nomograph
showing iso-sample number lines on a plot of COV versus percent difference in the
means (see Figure 3.2). If the influent and effluent COVs are not assumed to be equal, n
can be found from Equation 3.6 below:
Equation
%removal
Where COV is defined for influent and effluent data sets.
Zo/2 is a function of the desired level of certainty. For example, to determine a confidence
interval with 95% certainty (significance level a = 0.05), Za/2 equals 1.96. Values for
Zo/2 are tabulated in most statistical texts.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
73
-------
As an example of the application of the confidence interval, consider the case where the
researcher wants to determine if a mean influent concentration is greater than a mean
effluent concentration (assuming effluent concentrations are lower than influent
concentrations). To do this, the 95% confidence interval of the influent and effluent
EMCs are calculated. If the upper confidence limit (i.e., 97.5 percentile) of the effluent is
less than the lower confidence limit of the influent (2.5 percentile), then the mean influent
concentration is not equal to the mean effluent concentration, with 95% confidence.
As mentioned above, the Equation s derived in Appendix C allow for the solution of the
COV, percent removal, or n in terms of the other two variables. Solving for the required
COV for an estimated percent removal and n is shown in Figure 3.2 (for 95% confidence
limits and a power of 80%). The primary use of Figure 3.2 is to estimate the n required
to have 95% confidence in a hypothesis test given estimates of COV and percent
removal. It is recommended that Figure 3.2 be used to provide a reasonable estimate of
the number of samples (i.e., events) needed to quantify whether or not a BMP achieves an
anticipated level of performance (i.e., measured by percent removal). It can be seen from
Figure 3.2 that as the relative difference between influent and effluent mean event mean
concentrations becomes small, the number of required monitored events becomes quite
large.
Variations of the plot presented in Figure 3.2 are provided in Appendix B for a variety of
different confidence intervals, powers, and percent differences. These plots were
developed by Pitt and Farmer (1995).
Many commonly used statistical tests (e.g., parametric analysis of variance) are based on the
assumption that the data are sampled at random from a normally distributed population.
Thus, prior to applying the methods outlined in this section, the limitations imposed by
assumed normality of sample data sets should be fully understood. Several methods can
be used to determine the normality of a data set (or of data that is transformed to be
normally distributed). Some of these tests are the W-test, Probability Plot Correlation
Coefficient (PPCC), and graphical methods; all are useful for the analysis of stormwater
quality data.
As mentioned previously, researchers have found that stormwater quality data is
generally best fit by a log-normal distribution (EPA 1983; Driscoll et al. 1990;
Harremoes 1988; Van Buren et al. 1996) and theoretical justification for using a log-
normal distribution is provided by Chow (1954). Although, Van Buren et al. (1997) and
Watt et al. (1989) found that pond effluent and/or soluble constituents in stormwater may
be better fit using a normal distribution.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
74
-------
100
80 -
60 -
0)
C/3
g 40
20
Number of Sample Pairs Needed
(Power=80% Confidence=95%)
35
300
3000
0.75 1.00 1.25
Coefficient of Variation
Figure 3.2 Number of samples required using a paired sampling approach to observe a
statistically significant percent difference in mean concentration as a
function of the coefficient of variation (power of 80% and confidence of
95%)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
75
April 25, 2002
-------
The following are some properties of the lognormal distribution. If a sample (a data set of N
observations) is drawn from an underlying population that has a lognormal distribution, the
following apply:
1. The natural logarithm of log-normally distributed data is normally distributed with a
log-mean (ninfc) and log-standard deviation (
-------
potential pollutants. For example, an earlier study may have used outdated analytical
methods which had higher detection limits than current methods.
Beneficial uses of the receiving water. Information on water quality within a
stormwater drainage system often is used to indicate whether discharges from the
system are likely to adversely affect the receiving water body. For example, if a
stormwater system discharges to a lake, consider analyzing for nitrogen and
phosphorus because those constituents may promote eutrophication.
Overall program objectives and resources. The parameter list should be adjusted to
match resources (personnel, funds, time). If program objectives require assessing a
large number of parameters (based on a review of land uses, prior monitoring data,
etc.), consider a screening approach where samples collected during the first one or
two storms are analyzed for a broad range of parameters of potential concern.
Parameters that are not detected, or are measured at levels well below concern, can
then be dropped from some or all subsequent monitoring events. To increase the
probability of detecting the full range of pollutants, the initial screening samples
should be collected from storms that occur after prolonged dry periods.
A recommended list of constituents (along with recommended method detection limits
for comparing stormwater samples to water quality criteria) for BMP monitoring has
been developed and is presented in Table 3.1 below. Refer to Strecker (1994), Urbonas
and Stahre (1993), and the ASCE Database website (http://www.bmpdatabase.org/) for
more information on BMP monitoring parameters. The choice of which constituents to
include as standard parameters is subjective. The following factors were considered in
developing the recommended list of monitoring parameters:
The pollutant has been identified as prevalent in typical urban stormwater at
concentrations that could cause water quality impairment (NURP 1983; FHWA 1990;
and recent Municipal NPDES data).
The analytical result can be related back to potential water quality impairment.
Sampling methods for the pollutant are straightforward and reliable for a moderately
careful investigator.
Analysis of the pollutant is economical on a widespread basis.
Controlling the pollutant through practical BMPs, rather than trying to eliminate the
source of the pollutant (e.g., treating to remove pesticide downstream instead of
eliminating pesticide use).
Although not all of the pollutants recommended here fully meet all of the factors listed
above, the factors were considered in making the recommendations. When developing a list
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
77
-------
of parameters to monitor for a given BMP evaluation, it is important to consider the
upstream land uses and activities.
The base list represents the most basic arrangement of parameters. There may be
appropriate applications where other parameters should be included. For a discussion of
why some parameters were not included, see Strecker (1994).
Table 3.1: Typical urban stormwater runoff constituents and
recommended detection limits
Parameter
Conventional
PH
Turbidity
Total Suspended Solids
Total Hardness
Chloride
Bacteria
Fecal Coliform
Total Coliform
Enterococci
Nutrients
Orthophosphate
Phosphorus - Total
Total Kjeldahl Nitrogen (TKN)
Nitrate - N
Units
pH
mg/L
mg/L
mg/L
mg/L
MPN/ 100ml
MPN/lOOml
MPN/lOOml
mg/L
mg/L
mg/L
mg/L
Target Detection Limit
N/A
4
4
5
1
2
2
2
0.05
0.05
0.3
0.1
Metals-Total Recoverable
Total Recoverable Digestion
Cadmium
Copper
Lead
Zinc
Metals-Dissolved
Filtration/Digestion
Cadmium
Copper
Lead
Zinc
Organics
Organophosphate Pesticides (scan)
jig/L
0.2
1
1
5
0.2
1
1
5
0.05 - .2
Note: This list includes constituents found in typical urban stormwater runoff. Additional parameters may be needed to address site
specific concerns.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
78
-------
3.2.3.2 Dissolved vs. Total Metals
Different metal forms (species) show different levels of toxic effects. In general, metals
are most toxic in their dissolved, or free ionic form. Specifically, EPA developed revised
criteria for the following dissolved metals: arsenic, cadmium, chromium, copper, lead,
mercury (acute only), nickel, silver, and zinc. Chronic criteria for dissolved mercury
were not proposed because the criteria were developed based on mercury residuals in
aquatic organisms (food chain effects) rather than based on toxicity. For comparisons
with water quality criteria, it is advised that the dissolved metals fraction be determined.
If selenium or mercury is of concern, total concentrations should also be measured to
enable comparison with criteria based on bioaccumulation by organisms.
The distribution of pollutants between the dissolved and particulate phases will depend
on where in the system the sample is collected. Runoff collected in pipes with little
sediment will generally have a higher percentage of pollutants present in the dissolved
form. Runoff collected in receiving waters will generally have a higher percentage of
pollutants present in particulate form due to higher concentrations of suspended solids
that acts as adsorption sites for pollutants to attach to. It is difficult to determine how
much of the dissolved pollutants found in storm system pipes will remain in the dissolved
form when they are mixed with suspended sediments in receiving waters. As a result, it
is difficult to determine the ecological significance of moderate levels of dissolved
pollutants present within the conveyance system. In addition, hardness values for
receiving waters are often different than those for storm water. Hardness affects the bio-
availability of heavy metals, further complicating the ecological impact of dissolved
heavy metals.
If loads to the receiving waters are of concern (e.g., discharge to a lake known to be a
water quality limited water body) it may be desirable to determine total recoverable
metals in addition to dissolved metals to assess the relative load from different sources.
Finally, total recoverable metals data together with dissolved metals data can be used to
assess potential metals sediment issues.
3.2.3.3 Measurements of Sediment Concentration
A variety of methods have been employed in stormwater quality studies for quantifying
sediment concentration. The most frequently cited parameter is "TSS" or total suspended
solids. The "TSS" label is used, however, to refer to more than one sample collection
and sample analysis method. The "TSS" analytical method originated in wastewater
analysis as promulgated by the American Public Health Association.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
79
-------
The USGS employs the suspended-sediment concentration (SSC) method (ASTM 2000),
which was originally developed for the Federal Interagency Sedimentation Project
(USGS 2001). SSC data is often described as TSS data, when in many cases results from
the two methods may be significantly different. The difference between methods is
sample size - the SSC method analyzes the entire sample while the TSS method uses a
sub-sample. The process of collecting a representative sub-sample containing larger
sediment particles is problematic as large sediment particles (e.g., sand) often settle
quickly. Differences between the results obtained from SSC and TSS analytical methods
become apparent when sand-sized particles exceed 25% of the sample sediment mass
(Gray et al. 2000). Gray demonstrates that at similar flow rates, sediment discharge
values from SSC data can be more than an order of magnitude larger than those from TSS
data (USGS 2001) due primarily to larger particles that are often missed in the TSS
method. "The USGS policy on the collection and use of TSS data establishes that TSS
concentrations and resulting load calculations of suspended material in water samples
collected from open channel flow are not appropriate" (USGS 2001).
It is recommended that both TSS (for comparison to existing data sets) and SSC be
measured.
The discrepancies in sampling methodologies currently employed in the field highlight
the importance of particle size distribution (PSD) analysis as an essential component of
any BMP monitoring study. PSD data provide the information necessary to meaningfully
interpret the ability of a BMP to remove suspended materials. However, PSD methods
are varied and include (USGS 2001):
Dry sieve.
Wet sieve.
Visual accumulation tube (VA).
Bottom withdrawal tube.
Pipet.
Microscopy.
Coulter counter.
Sedigraph (x-ray sedimentation).
Brinkman particle pize analyzer.
Laser diffraction spectroscopy.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
80
-------
Light-based image analysis.
The investigator must select and use a consistent and appropriate method.
Specific gravity (SG) of sediments is also an important component in determining the
settleability of sediments and is recommended for sediment analysis by ASTM (1997).
For BMP studies where PSD data are being collected, SG provides additional useful
information about the ability of a particular BMP to remove sediment.
In addition, settling velocities of sediments are highly important and can be either
measured directly or calculated theoretically from SG and PSD data. Settling velocities
give the most useful information for quantifying BMP sediment removal efficiency.
The difficulty of collecting accurate sediment samples underscores the need to fully
understand the conditions under which sediment data were collected and analyzed.
Regardless of the analytical methods used, the sampling methodology often introduces
the largest bias to sediment data.
3.2.3.4 Analytical Methods
After the parameters have been selected, the analytical methods to be used to measure
them must be chosen. Select analytical methods that will provide results of sufficient
quality to support the intended uses of the data. To determine the quality of data
necessary for a program, consider the following:
Appropriate analytical levels. EPA guidance suggests tailoring the analytical level to
the intended use of the data. EPA has defined five analytical levels:
I. Field screening and analysis using portable instruments
II. Field analysis using more sophisticated portable analytical instruments,
possibly set up in a portable laboratory at the site
III. Analysis performed at an off-site analytical laboratory using EPA Contract
Laboratory Program (CLP) or equivalent methods, but without the
validation or documentation procedures required for CLP
IV. CLP routine analytical services and complete data reporting packages
V. Analysis by non-standard methods (to achieve very low detection limits or
measure a specific parameter not included in standard methods)
Stormwater samples are generally analyzed using Levels I, II, or III. Levels IV and V
are not used very often for stormwater projects because these levels are intended for
situations requiring low detection limits and high confidence, such as human or
ecological risk assessments or Superfund/MTCA investigations.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
81
-------
Appropriate methods should be selected for the chemicals of concern. These are the
most significant contributors to human health or environmental risk at the site.
Chenicals of concern are generally the most toxic, mobile, persistent, and/or
frequently occurring chemicals found at the site. Commonly occurring chemicals of
concern in stormwater runoff include metals (cadmium, copper, lead, and zinc),
polycyclic aromatic hydrocarbons (PAHs), and organo-phosphate insecticides (e.g.,
diazinon and chloropyrifos). The latter are included because recent studies in the San
Francisco Bay area found that diazinon accounted for much of the observed aquatic
toxicity in urban runoff (Cooke and Lee 1993). Other chemicals (e.g., organochlorine
pesticides and PCBs) should be included if there is reason to believe they are present.
Note that the potential toxicity of some metals in freshwater systems is affected by
the hardness of the water; thus, water quality standards for cadmium, copper,
chromium, lead, nickel, silver, and zinc are calculated based on water hardness. For
this reason, total hardness should be measured if metals are measured at sites where
fresh water quality standards may apply.
Level of concern. This term refers to the chemical concentration that is of concern.
Typically, state or federal water quality criteria for protection of aquatic life or human
health are used as the default level of concern for water sample results, and sediment
quality criteria are used as the level of concern for sediment sample results. For
pollutants that do not have state or federal water or sediment quality standards, the
Risk-based Concentration Table developed by EPA Region III (EPA 1994a,b) can be
used as levels of concern for water and soil sample results.
Required detection limit/practical quantitation limit. The level of concern directly
affects the data quality requirements because the sampling and analysis methods used
must be accurate at the level of concern. Sampling variability is often difficult to
control, especially in stormwater. The relative accuracy of most laboratory methods
decreases as concentrations approach the detection limits. For these reasons, the
practical quantitation limit (5 to 10 times the detection limit) should be below the
level of concern, if possible.
If the objective is to conduct a screening study to identify chemicals that appear to be
present at levels of concern, consider analyzing for a wide range of constituents using
analytical methods with low detection limits. An initial screening analysis can generally
reduce the number of chemicals analyzed in subsequent studies by eliminating those that
were detected below their corresponding levels of concern.
In cases where it is known that there is a high degree of correlation between the
concentration of the target pollutant(s) and some other parameter (e.g., fine particles, TSS,
total organic carbon), then it may be possible to use less costly monitoring approaches to
track the substitute, or "proxy" parameter(s). Although this approach can introduce some
uncertainty because it does not track the target pollutants, it is still worthy of consideration.
If the correlations are known to be strong and the cost differences pronounced, this strategy
may provide a way to obtain much more data (i.e., more frequent observations during more
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
82
-------
storm events and/or at more locations). Such improvements in data quantity could more
than offset the uncertainties introduced by imperfect correlations.
There are many precedents for using proxy parameters as indicators. For example, fecal
coliform are bacteria often used as proxies for pathogens and as an indicator of fecal
contamination. Total organic carbon and COD are sometimes used as proxies for BOD.
Turbidity is commonly used as a proxy for suspended solids, which in turn, is sometimes
used as a proxy for other pollutants of concern (e.g., metals, PAHs). The important
consideration is that other factors could also account for observed changes in the proxy
parameter relationship to other pollutants.
In many BMP monitoring programs, there are opportunities to obtain additional information
at little or no incremental cost (e.g., temperature or pH data). Such information may turn out
to be valuable to the overall stormwater program at some time in the future and/or to others
programs.
3.2.4 Recommendation and Discussion of Monitoring Equipment and
Methods
BMP monitoring can be done using a variety of equipment and methods. The type of
equipment and methods used often directly affect the usability of the data collected. Both
options and recommended approaches for monitoring are provided in this section.
3.2.4.1 Equipment
Equipment used to monitor BMPs includes a variety of data loggers, primary devices
(e.g., flumes, weirs, and nozzles), secondary devices (e.g., bubblers, pressure transducers,
and ultrasonic devices), automatic samplers, manual sampling devices, and rain gauges.
These devices and their uses are described below.
Data Loggers
Data loggers are used to monitor signals from various pieces of equipment and store the
impulses that they generate. When data loggers are combined with software to measure
and route signals between instruments and analyze data, they are referred to as "data
acquisition systems" and are often used as the execution center of a monitoring station.
Most data loggers have several input ports and can accommodate a variety of sensory
devices, such as a probe or transducer (e.g., flow meters, rain gauges, etc.). While
specific design characteristics vary between instruments, overall data logger design is
relatively standard. Some water quality samples have data loggers built into them;
however, they are usually more limited in capabilities (e.g., programmability,
communication options, etc.) than independent data loggers.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
83
-------
Data loggers suitable for stormwater monitoring applications are typically constructed of
weather-resistant materials capable of protecting their internal circuitry from water and
dust hazards. They are designed to operate at extreme temperatures, from as low as
-55ฐC to as high as 85ฐC (-67ฐF to 185ฐF). In addition, most models can be securely
mounted in remote locations, providing protection from wind and rain, wildlife, and
vandalism.
Figure 3.3: Data Logger with Weatherproof Housing (Handar)
Typical data loggers for field use consist of the following components: a weatherproof
external housing (a "case"), a central processing unit (CPU) or microprocessor, a quantity
of random-access memory (RAM) for recording data, one or several data input ports, a
data output port, at least one power source, and an internal telephone modem. In
addition, most data loggers have an input panel or keyboard and a display screen for field
programming. The CPU processes the input data for storage in RAM, which usually has
a backup power source (such as a lithium battery) to ensure that data are not lost in the
event of a failure of the primary power. Data stored in RAM may be retrieved by
downloading to a portable personal computer (PC), or to a host PC via modem.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
84
April 25, 2002
-------
Figure 3.4: Data Logger Without Housing (Campbell Scientific)
Data loggers vary in size from 0.2 to 9 kilograms (0.5 to 20 pounds) or more. Both
portable and fixed data-logging systems are available. For long-term, unattended
monitoring projects, a fixed instrument capable of serving as a remote transmitting unit
(RTU) may be preferable to a portable one. Manufacturers of data loggers suitable for
stormwater monitoring include: Campbell Scientific, Logan, Utah; Global Water
Instrumentation, Fair Oaks, California; Handar, Inc., Sunnyvale, California; In-Situ, Inc.,
Laramie, Wyoming; ISCO, Inc., Lincoln, Nebraska; Logic Beach, Inc., La Mesa,
California; and Sutron Corporation, Sterling, Virginia.
Programmability
Most data loggers can be programmed to record data at user-selected intervals. For
example, a particular model may be designed to permit a user to select a data recording
frequency from once every two seconds to once every 48 hours, with the choice of
frequencies varying by two-second intervals. The minimum and maximum intervals vary
from vendor to vendor, and often vary among models offered by the same vendor. In
addition, some data loggers have the ability to record event-related data, such as
minimum and maximum flow rates and event timing and duration. Data loggers can also
record data simultaneously for several different intervals (15 minutes, storm event, daily).
Most data loggers are field programmable, meaning that the software is equipped with an
interface that permits on-site manipulation. However, some less expensive models may
only be programmed at the factory. These models provide the advantage of cost savings
but provide limited versatility, especially if project requirements change over time.
In addition, most data loggers possess the capability of remote programming via
telephone modem. These models offer a significant advantage over factory programmed
and field programmable data loggers because they allow the user to manipulate the
program or monitor its effectiveness remotely. A network of data loggers used in a
multi-site monitoring effort can be reprogrammed more efficiently than by traveling from
site to site. An example where this functionality would be useful is if a predicted storm
rainfall depth changes after sites are set up, the sampling interval could be adjusted
remotely.
Although many vendors offer data loggers with the capability of remote manipulation via
modem and PC, the user-friendliness of the various models may vary greatly between
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
85
-------
vendors. Most vendors have developed software packages that are provided free of
charge with the purchase of their data logging systems. These software packages allow
for remote data logger programming, and provide for data manipulation, analysis, and
presentation at the host PC location. The interface environments used by these packages
varies from DOS-like command lines to menu-driven point-and-click environments.
Most data loggers that are provided with vendor-developed software packages require an
IBM-compatible PC with Windows to run the packages. Therefore, this additional cost
should be considered when evaluating a particular model. Another point of consideration
is the format in which a particular model logs the data it receives. Some models log data
in a format that can be converted from ASCII files to any of several commonly available
spreadsheet or word processing files, while others require the use of their particular
vendor-developed software for data analysis and manipulation.
Data Capacity
Memory type and capacity vary greatly between instruments. Standard capacity varies
between models and vendors from 8K or less, to more than 200K. In general, one data
point uses 2 bytes of information; therefore, a data logger with 64K of memory could be
expected to have a maximum data point capacity of 32,000 data points before data
downloading or additional memory would be required. However, some types of data
require as much as 4 bytes of memory per point. It should be noted that when recording
sets of data related to storm events, memory may be exhausted more quickly than
expected.
The type of memory used by a particular model is also an important consideration. Most
data loggers use non-volatile RAM, (i.e., memory that is not lost in case of a power
failure). Although this provides insurance that essential data will not be lost, the use of
non-volatile memory may not be necessary if the data logger is equipped with a backup
power source. A backup power source is automatically activated when the primary
power source is lost. Typically, backup power is supplied by a lithium battery, with
protection varying from 1 to 10 years.
Most models are programmed to stop recording data upon exhaustion of available
memory ("stop when full"). However, some models are equipped with wraparound or
rotary memory, which rewrites over the oldest data when available memory becomes
exhausted. When using rotary memory, it is important to realize that data may be lost if
it is not downloaded before it is written over.
Data loggers separate from water quality samplers increase the flexibility of the system
because of their increased programmability over those loggers on samplers. Memory
capacity is often an issue (even with the current inexpensive memory) and requires that
careful attention be paid to downloading data before it is overwritten.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
86
-------
Communications
Models vary in their ability to accept input from more than one source. Some data
loggers are designed with a single analog input channel, while others are designed with
up to 16 channels. In addition, some of the newer models accept digital input data. The
choice of a particular model should be based upon the number of sensors or probes from
which the instrument will be required to accept data.
Data loggers can accept information from many different types of sensors and
transducers. This allows for versatile use of most data logging systems. Some vendors
offer probes and transducers with built-in data loggers; however, these systems typically
cannot accept input data from other sensory devices, and their ability to communicate
output data is often limited.
With regard to output communications, all data loggers interface with the standard RS-
232 interface type, and some possess the capability to communicate with other interface
types. In most cases, data can be downloaded on-site to a laptop PC or a unit may be
transported to a lab or office so that the data can be downloaded to a desktop PC. As
indicated earlier, data loggers can be equipped with an internal modem for
telecommunications, allowing a user to download data from a remote host PC without
having to visit the field site.
In most cases, use of a telephone modem requires an IBM-compatible PC as the host and
the vendor's software. Typically, baud rates can be selected by the user. However, some
models are capable of only a few baud rates, a limitation that should be considered when
choosing a specific model. Some machines also possess the capability to transmit data
via line-of-sight, UHF/VHF, or satellite radio. These options also allow for remote
manipulation of programming and downloading of data.
Power Requirements
In general, data loggers are energy efficient devices. Most are powered by an internal
battery, with the option of using external electrical power, if available. Some can also be
equipped to use solar power.
Data loggers powered by internal batteries often offer a choice of cell type. Some models
offer the choice of rechargeable cells or standard 12 volt alkaline cells, while others offer
either alkaline or lithium batteries. The choice of power source and model selection,
depends upon several factors, including site accessibility, distance, and amount of data to
be recorded.
Alkaline cells are less expensive than lithium or rechargeable batteries, but they have a
shorter life and must be replaced more often. While alkaline cells offer a potential power
life of several months, lithium cells offer a potential power life of several years.
However, since lithium batteries are considered a hazardous material, data loggers using
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
87
-------
lithium batteries are subject to more stringent shipping requirements than models using
standard alkaline cells. In addition, since alkaline batteries must be replaced and
discarded frequently, the use of alkaline batteries may actually be more expensive than
using rechargeable batteries. Although rechargeable batteries offer less battery waste and
potential cost savings, the time and cost required to recharge the batteries should be
considered when evaluating power options.
Operating temperature range is another important factor to consider when choosing a
power supply. Lithium expands both the minimum and maximum temperatures at which
power can be used by the data logger. Under extreme conditions, it may not be feasible
to use a data logger powered by alkaline batteries.
Input
Output
Special Sensors
Temperature,
Conductivity, etc.
Control Module
Data Logger
Communications
Telephone Line Cellular RF(SCADA, etc.)
Desk Top PC
Figure 3.5: Data Logger Summary
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
88
April 25, 2002
-------
Flow
Natural channels, engineered open channels, and pipes are used as storm water conveyances. In
each case, hydraulic considerations dictate the mathematical relationships that can be used to
describe the flow rate at a given point in time. One of the primary hydraulic considerations is
whether the flow configuration represents an "open" or "closed" channel. Open channel flow
has a free water surface, and because the flow is driven by gravity, it varies with depth. Closed
channel flow, in which the flow fills a conduit, is caused by and increases with the hydraulic
pressure gradient. Some stormwater conveyance system pipes may function as open channels
during periods of low storm runoff and as closed channels when the runoff volume becomes
sufficiently large or when water is backed up due to downstream flow conditions (e.g., tide, river
flooding, etc).
In general, the flow rate in an open channel depends on the depth of flow and several other
factors (Chow 1959) including:
Geometric shape and changes in shape and slope along the length of the channel (affects
potential for development of turbulence and/or varied flow and therefore the choice of
methods and instruments used for measurement of flow).
Hydraulic roughness of the conveyance surface, whether natural or manmade (affects the
energy losses of the flow).
Rate at which the depth of flow changes over time (steady vs. unsteady flow).
Spatial scale over which the flow rate changes (uniform vs. varied flow).
The measurement of the flow rate in an open channel is more difficult to obtain than that of a full
pipe, because the free surface will change with respect to time.
Typically, stormwater flow through BMPs will fit the open channel flow configuration.
However, some BMPs are drained by pipe systems, which may be flowing, full at times.
Therefore, methods used for measuring flow in full pipes will also be discussed.
Table 3.2 summarizes available flow measurement methods, the requirements for their use,
typical BMP use, and required equipment. Each of these methods is discussed in more detail in
the following sections.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
89
-------
Table 3.2: Flow Measurement Methods
Method
Volume-Based
Stage-Based
Empirical
Equations
Stage-Based
Weir/Flume
Stage-Based
Variable Gate
Meter
Velocity-Based
Tracer Dilution
Pump-Di scharge
Major
Requirements
For Use
Low flow rates
Open flow
Known
channel/pipe
slope
Channel slope,
geometry,
roughness
consistent
upstream
Open flow
Constraint will
not cause
flooding
4-, 6-, or 8-inch
pipes only
None
Adequate
turbulence and
mixing length
All runoff into
one pond
Typical BMP
Use
Calibrating
equipment
Manual
sampling
Manual or
automatic
sampling
Manual or
automatic
sampling
Not typically
used for
BMPs
Automatic
sampling
Typically used
for calibrating
equipment
Not typically
used for BMPs
Required
Equipment
Container and
stopwatch
Depth Measurer
Weir/flume and
depth measurer
ISCO Variable
Gate Meter
Depth measurer
and velocity
meter
Tracer and
concentration
meter
Pump
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
90
April 25, 2002
-------
Volume-Based Methods
The concept behind volume-based flow measurement is simple: one collects all the flow over a
short period of time, measures the volume, and divides the collected volume by the length of the
time period:
Q = V/T Equation 3.7
where,
Q: flow, m3/s (ft3/s)
V: volume, m3 (ft3)
T: time, s
A stopwatch can be used to measure the period required to fill a receptacle of known quantity to
a predetermined level. The receptacle must be large enough that it requires some accurately
measurable period of time to fill. The receptacle could be a bucket, a drum, or a larger container
such as a catch basin, holding tank, or some other device that will hold water without leakage
until the measurement is made.
This method is easy to understand, requires relatively simple equipment, and can be very
accurate at low rates of flow. At higher rates of flow, collection of all the runoff from typical
BMP conveyances (an essential component of the method) will probably become infeasible.
This method is most useful for conducting limited research and for calibrating equipment.
Stage-Based Methods
Flow rate can be estimated from the depth of flow (i.e., water level or stage) using well-
understood, empirically derived mathematical relationships. That is, for a set hydraulic
configuration, the relationship between stage and flow is known. The most commonly used
empirical relationship, the Manning Equation, is appropriate for open channels in which flow is
steady-state (i.e., the flow rate does not vary rapidly over time) and uniform (the depth of flow
does not vary over the length of the channel) (Gupta 1989). In automated stormwater sampling
the Manning Equation is commonly used to estimate the flow rate of the flow stream.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
91
-------
Manning's Equation
The variables required for the Manning Equation (Equations 3.8 and 3.9) are the slope of the
energy grade line (usually assumed to be the slope of the channel bottom), the cross-sectional
area of the flow, the wetted perimeter, and an empirical roughness coefficient, which takes into
account channel material, age, and physical condition.
Equation 3.8
Q =
Equation 3.9
where,
Q: flow, m3/s
n: Manning roughness coefficient
(dimensionless)
A: cross sectional area, m2
R: hydraulic radius, m = A/(wetted
perimeter)
S: slope of the channel, m/m
where,
Q: flow, frVs
n: Manning roughness coefficient
(dimensionless)
A: cross sectional area, ft2
R: hydraulic radius, ft =A/(wetted perimeter)
S: slope of the channel, ft/ft
The Manning Equation truly applies only to steady and uniform flow but can provide a
fairly accurate estimate of flow rates if certain conditions are met. The channel slope and
cross-sectional geometry must be constant for some distance upstream of the site, the
exact distance varying with overall system hydraulics (a rule of thumb is a length of
twenty channel diameters upstream). Flow conditions at the site should not be affected
by downstream features (i.e., no backflow effects). The cross-sectional area and wetted
perimeter are both geometric functions of the channel shape and the depth of flow. The
"roughness" of the conveyance walls can be described by a roughness coefficient.
Additional information on applicability and values for Manning's roughness coefficients
for common channel types are provided in most hydraulics texts (Chow 1959; Gupta
1989).
Use of the Manning Equation assumes that the slope of the channel bottom is accurately
known. Monitoring studies using this technique to estimate flow rates often rely on as-
built drawings to determine channel slope. Because these drawings vary in accuracy,
direct measurement of the slope of the channel bottom and verification of hydraulic
conditions is recommended.
The flow rate of stormwater runoff tends to be unsteady. This is due to changes in the
intensity of precipitation and the dynamic nature of overland flow, which causes the flow
rate to vary with time, either gradually or rapidly. Depending on the frequency with
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
92
-------
which the depth of flow is measured, rapid fluctuations in flow rate will be missed and
the total runoff volume from a storm event will be miscalculated.
Other Empirical Stage-Flow Relationships
Another empirical relationship used to estimate flow is the Chezy Equation (Gupta
1989):
Equation 3. 10
where,
Q: flow, m3/s (ft3/s)
A: cross-sectional area, m (ft )
R: hydraulic radius, m (ft)
S: slope of the energy grade line, m/m (ft/ft)
C: flow coefficient, mir2/s (ft1/2/s)
Under open channel flow, the coefficient "C" can be defined as:
R1'6
C = - Equation 3.11
n
where,
n: Manning's Roughness Coefficient
When "C" is substituted into Chezy 's Equation, the resulting Equation is identical to the
Manning Equation.
A failure of both the Manning and Chezy Equations is that they imply that the Manning
"n" value is constant for a given channel. However, it is known that for natural channels
"n" may vary greatly with respect to flow (Ponce 1989). Therefore, when considering
applying these equations to a natural channel, one should first evaluate the alluvial
material in the channel and the magnitude of flows expected. It may be desirable to
select another flow measurement approach for natural channels with highly varied
surfaces and flow rates.
Stage Based Method Using Weirs and Flumes
The accuracy with which flow is estimated can be improved by using a weir or flume to
create an area of the channel where the hydraulics is controlled (control section). Each
type of weir or flume is calibrated (i.e., in the laboratory or by the manufacturer) such
that the stage at a predetermined point in the control section is related to the flow rate
using a known empirical equation (for examples, see Stevens 1991).
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
93 April 25, 2002
-------
Stage-Based Variable Gate Meters
A relatively new development in flow metering technology is ISCO Inc.'s (Lincoln, NE)
Variable Gate Metering Insert. Discharge flows through the insert and under a pivoting
gate, creating an elevated upstream level that is measured with a bubbler system. The
meter uses an empirical relationship to calculate the discharge rate based on the angle of
the gate and the depth of flow upstream of the gate. This approach can be used only
under conditions of open channel flow in circular pipes. Currently the system is only
available for pipe diameters of 10.16, 15.24, and 20.32 cm (4, 6, and 8 inches). The
Variable Gate Metering Insert was designed to measure the flow rate under fluctuating
flows and should be effective at both very high and very low flow rates. Its main
limitation is the size of the conveyance for which it is designed. The insert may be useful
for sampling very small catchment areas. Again, problems with debris accumulation can
occur.
Velocity-Based Methods
The continuity method is a velocity-based technique for estimating flow rate. Each
determination requires the simultaneous measurement of velocity and depth of flow.
Flow rate is calculated as the sum of the products of the velocity and the cross-sectional
area of the flow at various points across the width of the channel:
Q = A;*Vi Equation 3.12
where,
Q: flow, m3/s (ft3/s)
A;: cross-sectional area of the flow at section i, m2 (ft2)
V;: mean velocity of the flow at section i, m/s (ft/s)
The sections i = 1-n are planar segments of a cross-section of the flow where n is the
number of points across the width of the channel. In stormwater runoff applications, the
conveyance is small enough that a single cross-sectional area and estimate of average
velocity is typically used to estimate flow rate. That is, it is not necessary to segment the
cross-sectional area of the flow. The accuracy of this method is dependent on the ability
of a sensor to measure velocity over a range of flow.
Although this method is useful for calibrating equipment, it is more sophisticated and
expensive than the stage-flow relationships previously discussed. In addition, this
method is suitable only for conditions of steady flow. That is, water level must remain
essentially constant over the period required for obtaining velocity measurements. This
is not generally a problem in small conveyance systems when instruments that make
measurements rapidly are employed.
Additional relationships, developed for pipes that are flowing full, are the Darcy-
Weisbach equation and the Hazen-Williams equation. These equations are used in
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
94 April 25, 2002
-------
systems where pressurized flow (i.e., pipes flowing full; no free water surface) is present
and can be found in Gupta (1989).
Tracer Dilution Methods
Tracer dilution methods can be used where the flow stream turbulence and the mixing
length are sufficient to ensure that an injected tracer is completely mixed throughout the
flow stream (USGS 1980; Gupta 1989). Tracers are chosen so that they can be
distinguished from other substances in the flow. For example, chloride ion can be
injected into fresh water, and dyes or fluorescent material can be used if turbidity is not
too high.
Dilution studies are well suited for short-term measurements of turbulent flow in natural
channels and in many manmade structures such as pipes and canals. However, these
methods are better suited to equipment calibration than to continuous monitoring during a
storm event. Two dilution methods can be used to determine flow rate as described
below.
Constant Injection Rate Tracer Dilution Studies
A known concentration of tracer is injected at a constant rate into a channel. The
concentration of the tracer in the flow is measured at a downstream point over time.
After some time period has passed, the tracer becomes completely mixed in the flow so
that the downstream concentration reaches steady state. Flow is calculated from the
initial tracer concentration, the tracer injection rate, and the steady-state downstream
concentration.
Total Recovery Tracer Dilution Studies
A discrete "slug" of tracer is injected into the channel. Near-continuous measurements of
tracer concentration in the flow are taken at a downstream point until the plume has
entirely passed. Flow is calculated from the volume and concentration of injected tracer
and the total area under the concentration-time curve.
Pump Discharge Method
In some cases, the overall discharge rate for a catchment may be measured as the volume
of water that is pumped out of a basin per unit time while holding the water level in the
basin constant. This method can be applied at sites where flow runs into a natural or
manmade basin from several directions or as overland flow. If the pump is precalibrated,
the number of revolutions per minute, or the electrical energy needed to pump a given
volume, may be used as a surrogate for measuring the pumped volume during a
stormwater runoff event. Application of this method requires considerable knowledge of
the installed pump's performance. Because this setup (i.e., all of the runoff from a
catchment flows into one pond or basin which can be pumped out) is not usually
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
95 April 25, 2002
-------
encountered in the field as the only available monitoring method, pumps are not
discussed further in this manual.
3.2.4.2 Automatic Sampling Techniques
Selection of Primary Flow Measurement Device
This section provides an overview of the process of selecting a primary flow
measurement device.
Changes to surface hydrology due to urbanization result primarily from the increases in
impervious areas (roofs, streets, parking lots, etc.) and the increased hydraulic
conveyance of the flow channels. The naturally occurring channels are often
straightened, deepened, and lined in addition to the installation of storm sewers, drains,
and gutters. Without detention storage, the resulting hydrograph has a higher peak
discharge and shorter duration. This necessitates the ability of a primary flow device to
accurately measure large discharge rates for storm events with high precipitation
intensities. Due to the highly variable nature of storm events, low runoff rates will result
from the smaller storm events. Analysis of long-term rainfall records indicates that
smaller storm events generally account for the majority of stormwater runoff and
resulting pollutant loads. Therefore it is essential that the primary device selected is also
capable of accurately measuring the lower range of the expected flows. The potential for
a wide range of flow rates resulting from stormwater runoff makes the assessment of the
required range of discharge rates an important consideration for selection of a flow
measurement device.
Flow measurements are critical to monitoring stormwater BMPs. Accurate flow
measurements are necessary for accurate composting of samples used to characterize
storm runoff and for the estimation of volumes (including pollutant loads) treated in the
BMP. Many methods are available to estimate the flow in open channels: volume-based
methods, velocity-based methods, empirical equations, and tracer-dilution methods.
While these methods are all valid ways to measure the flow in open channels, they are
not potentially as accurate as the use of a primary flow measurement device. Researchers
monitoring flows pertaining to stormwater BMP effectiveness are encouraged to use
primary flow devices where possible.
Types of Primary Flow Measurement Devices
Primary flow measurement devices fall into the general categories of flumes and weirs.
Primary flow measurement devices allow for accurate measurement of discharge rates by
creating a channel geometry in which the hydraulics are controlled (control section).
Primary devices are calibrated (i.e., in the laboratory or by the manufacturer) to relate the
stage at a predetermined point in the control section to the discharge rate using a known
empirical equation (for examples, see Stevens 1998). These types of measurement
devices are called depth (or stage) based methods because the discharge through the
device is directly related to the depth (stage or head) of the flow. The relationship
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
96 April 25, 2002
-------
between the depth of flow and the discharge is called the rating. Tables referred to as
rating curves are available for all standard flumes and weirs.
Weirs
A weir is an obstruction (usually a vertical plane) built or placed across an open channel
(or within a pipe under open channel flow) so that water flows over the weir's top edge or
through a well-defined opening in the plane. Many types of weirs can be used to measure
discharge; the three most common are the rectangular, trapezoidal (or Cipolletti weir),
and triangular weirs. The weir opening (i.e., the rectangular, trapezoidal, or triangular
opening) is called the "notch." Each type of weir has a specific discharge equation for
determining the flow rate through the weir.
Weirs are generally low in cost, easy to install (relative to flumes), and can be quite
accurate when used correctly. A weir can be used to regulate flow in a natural channel
with irregular geometry, a situation where Manning's Equation, for example, would not
provide reliable estimates for the flow rate. However, a weir will back water up in
channels by creating a partial dam. Weirs are generally used for flow measurements with
relatively large head available to establish free-flow conditions over the weir. A weir is
intended to back up water by creating a partial dam. During large storm events, backed-
up water could cause or worsen flooding upstream, particularly in a closed conduit.
Some jurisdictions prohibit the use of weirs for this reason. When evaluating the
suitability of a monitoring site for a weir, it is important to determine whether the system
was "over designed." That is, will the conveyance be able to move the design capacity
after weir installation. In the case where the downstream depth of flow is greater than the
crest of the weir, a different stage-flow relationship for the weir will apply.
Sediments and debris that accumulate behind a weir can alter the hydraulic conditions,
changing the empirical relationship between flow depth and discharge rate. Weirs are
often not a good choice where representative suspended sediment samples are desired.
Weirs should be inspected regularly and accumulated sediment or debris removed. If
high amounts of sediment or debris occur in the flow, then use of a flume may be more
appropriate as they generally avoid sedimentation problems.
Flumes
A flume is a specially built reach of channel (sometimes a prefabricated insert) with a
converging entrance section, a throat section, and diverging exit section.
Because the velocity of water accelerates as it passes through a flume, the problem of
sedimentation associated with weirs (see below) is avoided; however, problems with
debris accumulation may still occur. Another benefit is that flumes introduce a lower
headloss than weirs, resulting in a reduced backwater effect. A flume may be more
expensive and difficult to install than a weir due to its more complex design; however,
where applicable, flumes can provide accurate results and significantly reduced
maintenance.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
97 April 25, 2002
-------
The most common types of flumes are the Parshall, the Palmer-Bowlus, the HS, H, and
HL flumes and the trapezoidal flume.
Figure 3.6: Parshall flume (Plati-Fab Inc.)
The area or slope (or both) of the flume is different from that of the channel, causing an
increase in water velocity and a change in the level of the water flowing through the
flume (Grant 1989). Stage-flow relationships have been established for a variety of
flume configurations (USGS 1980; Gupta 1989; Stevens 1991).
Figure 3.7: H-flume (Tracom Inc.)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
98
April 25, 2002
-------
Considerations for Selection of Primary Flow Measurement Device
Consideration should be given to the following items when selecting a primary flow
measurement device.
Range of Flows
Triangular thin-plate weirs have a large range in their ability to measure flows because of
the 2.5-power relationship between flow depth and flow rate. That is, relative to other
devices, flow increases quite rapidly as a function of head. The range of flow rates that
can be measured accurately can vary by a factor (ratio of largest flow to smallest flow
rate) of 200 for fully contracted weirs to around 600 for partially contracted 90ฐ notches
that can utilize the allowable range of head (ASTM 1995).
For rectangular thin-plate weirs, the range is typically about a factor of 90 and increases
to about 110 for full-width weirs. These ranges depend somewhat on the crest length to
channel width ratio. These results are based on a minimum head of 0.1 ft (0.03 m) and a
suggested (although not absolute) maximum head of 2 ft (0.6 m). However, the range-
ability of smaller rectangular weirs can be significantly less (ASTM 1995).
The range in flow measurement for Parshall flumes varies widely with size. The range of
Palmer-Bowlus and other long-throated flumes depends on the shape of the throat cross-
section and increases as the shape varies from rectangular toward triangular. For typical
Palmer-Bowlus flumes of trapezoidal section, the range of flow rates that can be
measured accurately generally varies by a factor of 30. The USGS has developed and
tested a modified Palmer-Bowlus flume (USGS 1985) for use in circular pipes that carry
highway stormwater runoff. This flow can occur under either open or pressurized
channel flow. This flume has been designed to measure the discharge under pressurized
flow by using two bubbler sensors, which detect the hydraulic pressure change between
upstream and downstream locations on the flume. This system was found to be one of
the most accurate after calibration is performed. However the range between low and
high flows that can be measured accurately using a Palmer-Bowlus flume is not as large
as some other types of devices.
In cases in which there is a need for measurement of extreme flow ranges along with
sediment transport capability, which is often the case for stormwater runoff, the H, HS, or
HL flumes should be considered. The range of flows that can be measured accurately
using H-type flumes can exceed three orders of magnitude; for example, a 3 ft H flume
can measure flows between 0.0347 cfs at 0.10 ft of head to 29.40 cfs at 2.95 feet of head.
For some cases when low flows are expected to occur for an extended period but will
ultimately be superseded by much larger flow rates, the interim use of removable small
flumes inserted inside larger flumes can provide a method for accurate measurement of
the range of flows.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
99 April 25, 2002
-------
Flow Rate
As mentioned in the beginning of the section, one of the most important factors
influencing the selection of a primary device is the flow capacity necessary to
accommodate runoff. Small and moderate flows are generally best measured with thin-
plate weirs, with the triangular notches most appropriate for the smallest flows (ASTM
1995). Small Parshall and Palmer-Bowlus flumes are also available to measure low
flows. The flumes do not have issues related to sediment passage and head loss as do
thin-plate weirs, but this comes at some sacrifice in potential accuracy (ASTM 1995).
Flumes and broad-crested weirs are generally the best choices for the measurement of
large discharges.
Accuracy
Weirs are generally recognized as more accurate than flumes (Grant and Dawson 1997).
A properly installed weir can typically achieve accuracies of 2 to 5% of the rate of flow,
while flumes can typically achieve accuracies of 3 to 10% (Spitzer 1996). The ASTM
cites lower errors for weirs ranging from about 1 to 3% and Parshall and Palmer-Bowlus
flumes with typical accuracies around 5%. However, the overall accuracy of the flow
measurement system is dependant on a number of factors, including proper installation,
proper location for head measurement, regular maintenance, the accuracy of the method
employed to measure the flow depth, approach velocities (weirs), and turbulence in the
flow channel (flumes). It should be noted, however, that the largest source of error in
flow measurement of stormwater results from inaccuracies related to low flow or
unsteady flow. Improper construction, installation, or lack of maintenance can result in
significant measurement errors. A silted weir or inaccurately constructed flume can have
associated errors of ฑ5 to 10% or more (Grant and Dawson 1997). Circumstances present
in many stormwater monitoring locations can result in errors well in excess of 100%.
Potential inaccuracies in the method used to measure the depth of flow will tend to
increase the error in flow measurement as the flow depth approaches the minimum head.
For primary devices operating near minimum head, even a modest error can have a
significant effect on the measured flow rate. Therefore, it is important to select sizes or
combinations of primary devices that avoid prolonged operation near minimum head
(Spitzer 1997).
Cost
The important factor of cost consideration should include manufacturing, installation, and
operational costs. Weirs are often considerably less expensive to fabricate than flumes
due to simpler design and material requirements (Grant and Dawson 1997). Weirs are
also usually easier and less expensive to install, although installation of flumes designed
for insertion into a pipe (e.g. Palmer-Bowlus and Leopold-Lagco) are generally
straightforward. Despite the higher initial costs of flumes, the relatively low maintenance
requirements may outweigh this with time (Grant and Dawson 1997). Consideration
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 00 April 25, 2002
-------
should be given to the expected sediment loads in the flow to be measured for likely
accumulation and maintenance requirements for weir installations.
Head Loss and Flow Characteristics
The head difference that is required for a weir or flume to operate properly may be an
important selection criterion. Examples include, when the elevation difference is not
adequate to maintain the required flow or when the upstream channel cannot contain the
backwater.
For the same flow conditions, thin plate weirs typically require the largest head
difference, Parshall flumes require an intermediate amount of head, and the long-throated
flumes require the least (ASTM 1995).
Weirs are typically gravity fed and must be operated within the available head of the
system. Flumes also require a certain head range in which the discharge liquid level is
low enough that it does not exert back pressure on the liquid in the throat of the flume,
otherwise the flume will be in a submerged condition, and two head measurements will
be required to determine the flow rate.
Operation of a weir is sensitive to the approach velocity, often necessitating a stilling
basin or pond upstream of the weir to reduce the fluid velocity. Operation of a flume is
sensitive to turbulence or waves upstream from the entrance to the flume, which can
require a section of straight channel upstream of the flume.
Sediment and Debris
Flumes tend to be self-cleaning because of the high flow velocity and the lack of any
obstruction across the channel (Spitzer 1997). A flume is therefore generally more suited
to flow channels carrying solids than is a weir.
Debris accumulation is likely to occur behind a weir especially due to the presence of a
stilling basin to reduce flow velocities to an acceptable rate. Debris accumulation behind
a weir can affect flow measurement. This requires periodic inspection and maintenance
to remove debris. To allow periodic removal of deposits, it is recommended that the weir
bulkhead be constructed with an opening beneath the notch to sluice accumulated
sediments (Spitzer 1997).
Flumes, while typically not susceptible to problems due to sedimentation, can have debris
accumulate in the throat portion of the flume and require periodic maintenance (although
generally less frequently than weirs).
Construction Requirements
The Parshall flume is usually the most difficult device to construct due to the relatively
complex shape and the possible need to excavate the channel floor to accommodate the
sharp downward slope of the throat. Because this flume is an empirical device it is
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
101 April 25, 2002
-------
necessary to closely follow the design specifications (ASTM 1995). The discharge
coefficients for long-throated flumes can be obtained theoretically which allows for some
departure from the prescribed dimensions. Many types of flumes are available in
prefabricated sizes up to several feet in width.
Weirs are generally easier to construct than flumes. The most difficult task is the
fabrication of the notch edges, which require a sharp edge so the nappe is free flowing.
Selection of Secondary Flow Measurement Device
A variety of instruments may be used to measure water depth. Because some techniques
are relatively cumbersome, they are more useful for calibrating equipment than for
routine or continuous data collection during storm events. The equipment required for
each technique and the associated advantages and disadvantages for sampling runoff at
BMP sites are described below. Table 3.3 summarizes available equipment for
measuring depth of flow, major requirements for use, and typical use within a BMP
monitoring program.
Table 3.3:
Equipment for measuring depth of flow
Method
Major Requirements For Use
Typical Use in a BMP Monitoring
Program
Visual Observations
Small number of sites and events to
be sampled.
No significant health and safety
concerns
Manual sampling
Float Gauge
Bubbler Tube
Stilling well required
Open channel flow.
No velocities greater than 5 ft/sec
Manual or automatic sampling
Automatic sampling
Pressure Transducer
Better if remains submerged
Automatic sampling
Ultrasonic Depth Sensor
Open channel flow.
No significant wind, loud noises,
turbulence, foam, steam, or floating
oil & grease
Automatic sampling
Ultrasonic Uplooking
No sediments or obstructions likely
to cause errors in measurement.
Automatic sampling
Radar/Microwave
Similar to Ultrasonic Depth Sensor
but can see through mist and foam
Automatic sampling
3-D Point Measurement
Highly controlled systems.
Typically not useful in the field
Automatic sampling
Pressure Probe
Open channel flow.
No organic solvents or inorganic
acids & bases
Automatic sampling
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 02 April 25, 2002
-------
Float Gauge
A float gauge consists of a float that is free to move up and down in response to the rising and
falling water surface in a channel. Prior to an actual stormwater sampling event, the site is
calibrated to establish an initial reference depth. During the storm, the float rises and falls with
changes in water surface elevation, and a device attached to the float records the magnitude of
these changes. The changes in water surface elevation are converted to depth of flow by the float
gauge. A data logger can record the depth of flow, and if capable of performing mathematical
equations, can also determine the flow rate. The data can also be used as input to appropriate
software to compute the flow rate.
In some applications, use of a float gauge requires a stilling well. A stilling well is a reservoir of
water connected to the side of the conveyance that isolates the float and counterweight from
turbulence in the main body of the flow. The need to retrofit an existing channel or conduit with
a stilling well, a potentially expensive and time-consuming process, is the principal drawback of
this technique. However, this method may be useful if sampling is conducted at a site where a
float gauge and stilling well have previously been installed.
Bubbler Tube
Bubbler tubes are used by some types of automated flow meters to measure the depth of flow.
Compressed air (or gas) is forced through a submerged tube attached to the channel invert (i.e.,
bottom of the channel). A pressure transducer measures the pressure needed to force a bubble
out of the tube. This pressure, in turn, is linearly related to the depth of the overlying water:
P = ph Equation 3.13
where:
9 9
P: hydrostatic pressure, N/m (Ib/ft )
p: specific weight of water, N/m3 (Ib/ft3)
h: depth of water, m (ft)
Bubbler tubes are commonly integrated with a flow meter, or a data logger that is capable of
performing mathematical calculations. This approach allows the measurement of depth to be
immediately converted to a flow. These real-time inputs along with a program that tracks
accumulated flow volumes can be used to trigger the collection of samples for flow-weighted
compositing by an automated sampler.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 03 April 25, 2002
-------
Figure 3.8: Bubbler flow meter (ISCO)
Bubbler tubes are simple to use and are not usually affected by wind, turbulence, foam, steam, or
air-temperature gradients. Accuracy is not lost under dry conditions in a conveyance between
runoff events (some other types of probes must remain submerged). Although they are generally
reliable, bubblers are susceptible to error under high velocity flow. That is, as flow velocity
increases to over 1.5-1.8 m/s (5-6 ft/s), a low pressure zone is induced around the mouth of the
bubbler tube, interpreted by the flow meter as a drop in flow rate. These instruments therefore,
should not be used in channels where the slope of the bottom exceeds 5-7 percent. Sediments
and organic material can also plug bubbler tubes. Some units are periodically purged with
compressed air or gas to prevent this problem, but visual inspection and periodic maintenance
are recommended for any unit installed in the field. Bubblers are commonly available in
integrated systems, such as those manufactured by ISCO and American Sigma, but are also sold
as independent devices.
Ultrasonic Depth Sensor
An ultrasonic depth sensor consists of a sonar-like device mounted above the surface of the
water at a known distance above the bottom of the channel. A transducer emits a sound wave
and measures the period of time taken for the wave to travel to the surface of the water and back
to a receiver. This time period is converted to a distance and then converted to a depth of flow,
based on measurements of the site configuration. As with bubbler tubes, an ultrasonic sensor can
be integrated into a flow meter or interfaced with a data logger. An ultrasonic depth sensor and
data logger can provide the real-time flow data necessary to trigger an automated sampler to
collect a stormwater sample for flow-weighted compositing.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 04 April 25, 2002
-------
Figure 3.9: Ultrasonic-depth sensor module (ISCO)
Some manufacturers have built redundancy into their ultrasonic depth-measuring instruments.
Redundancy helps to ensure that useful data will be collected even if some of the sensors in the
array become fouled with grease, surface-active materials, or organisms. Experience has shown
that this type of fouling can occur during storm events. Because an ultrasonic sensor is mounted
above the predicted surface of the water, it is not exposed to contaminants in the runoff (unless
the depth is greater than anticipated or installed in a pipe that reaches fully pressurized flow).
However, ultrasonic signals can be adversely affected by wind conditions, loud noises,
turbulence, foam, and steam, and they will require periodic inspection and maintenance.
Ultrasonic signals can also be affected by changes in density associated with air temperature
gradients; however, some manufacturers build a compensation routine into their instruments.
Background noise can interfere with a sensor's ability to accurately measure water depth. For
example, an ultrasonic sensor was used in Portland, Oregon to measure the depth of flow at an
urban stormwater sampling site located in a manhole, in which runoff from an arterial pipe
splashed down into the main conveyance. To dampen the effect of the interfering signal, the
ultrasonic sensor was retrofitted with a flexible noise guard.
Pressure Probe
A pressure probe consists of a transducer, mounted at the bottom of the channel, that measures
the hydrostatic pressure of the overlying water. This hydrostatic pressure is converted to a depth
of flow. Some pressure probes have a built-in thermometer to measure the temperature of the
water, allowing for temperature compensation in the depth of flow calculation. As with bubblers
and ultrasonic probes, the pressure probe can be integrated into a flow meter or interfaced with a
data logger to provide real-time inputs to an automated sampler. If the instrument is fitted with a
thermometer, the temperature data used for compensation can possibly also be input to memory
and retrieved as additional useful data.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
105 April 25, 2002
-------
Figure 3.10: Pressure transducers (In-Situ Inc.)
Submerged probes are not adversely affected by wind, turbulence, foam, steam, or air
temperature gradients. However, because contaminants in the water may interfere with or
damage the probe, periodic inspection and maintenance is recommended. Dry conditions
between storms can affect the accuracy of the probe, as can sudden changes in temperature.
Ultrasonic "Uplooking"
This depth of flow sensor is mounted at or near the bottom of the channel or pipe. It uses
ultrasonic signals to determine the depth of the flow. This sensor is very accurate unless
interference occurs. However, according to a vendor, this equipment is not recommended for
stormwater applications because the sensor is likely to become covered by sediments and debris.
This then interferes with the signal and does not allow the sensor to work properly.
Radar/Microwave
A variation of the ultrasonic method is a non-water contacting instrument that emits and
reprocesses electromagnetic waves in the radar/microwave spectrum. By altering the wavelength
of the electromagnetic signal, problems associated with foam, mist, and rapid changes in air
temperature and pressure are eliminated or significantly reduced.
A radar/microwave sensor is used in the same manner as an ultrasonic "downlooking" sensor for
measuring fluid levels in tanks. Based on experience, this device does not present a significant
advantage over other methods of level measurement, since foam and mist are not typically a
large concern during stormwater monitoring.
Radar/microwave sensors have not been extensively tested by manufacturers for this type of
application, and there is no existing literature that shows them being used for stormwater
monitoring.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 06 April 25, 2002
-------
Equipment for Measuring Velocity
Use of the continuity equation for measuring flow requires the estimation of average velocity as
well as depth. The velocity of flow may be measured using visual methods (i.e., the float-and-
stopwatch and the deflection, or drag-body methods), tracer studies, the use of instruments such
as rotating-element current meters and pressure, acoustic, ultrasonic (Doppler), and
electromagnetic sensors. Electromagnetic sensors have been found to be the most accurate.
Among these methods, many are more useful for the calibration of automated equipment than for
continuous data collection. Only the ultrasonic and electromagnetic methods are recommended
for measuring velocity during a storm. In the following text, velocity measurement methods
potentially suitable for calibration are described (more details are available in USGS 1980).
More extensive discussions, including advantages and disadvantages related to sampling
activities, are provided for the ultrasonic and electromagnetic sensors.
Methods Suitable for Calibration
The most important aspect of any calibration method is its ability to obtain accurate results with
a high degree of certainty and repeatability. A variety of methods have been employed in the
past. The most common methods are described in this section. Table 3.4 summarizes the
available methods.
Table 3.4: Velocity measurement methods suitable for calibration
Method
Tracer Studies
Rotating-Element Current Meters
Pressure Sensors
Acoustical Sensors
Float-and-Stopwatch
Deflection (or Drag-Body)
Method
Comments
Recommended Method Where applicable, one
of the best calibration methods. Requires
complete mixing of tracer with flows.
Useful for larger flows that do not rapidly vary
with time. Typically useful for large systems
with appreciable flows. Low flows are difficult to
monitor.
Not useful for velocities above 1.5-1.8 m/sec or
in pipes with steep slopes (>5%).
Not applicable to most monitoring locations.
Large flow rates are typically required. Base flow
required to observe complete storm hydrograph.
Typically applicable only to large channels.
Rarely accurate enough for calibration purposes.
Not recommended for most situations.
Rarely accurate enough for calibration purposes.
Not recommended for most situations.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 07 April 25, 2002
-------
Tracer Studies
Tracer methods have been developed to measure flow velocity under uniform flow (USGS
1980). As described in the flow measurement methods section, for Total Recovery Tracer
Dilution studies, a discrete slug of tracer is injected into the flow. Concentration-time curves are
constructed at two downstream locations. The time for the peak concentration of the dye plume
to pass the known distance between the two locations is used as an estimate of the mean velocity
of the flow. This method is not practical for continuous flow measurement, but is useful for site
calibration.
Rotating-Element Current Meters
A current meter or current meter array can be used to measure the velocity at various points
throughout a flow stream. The measured point velocities can be combined to estimate a mean
velocity for the flow. As with the deflection or drag-body method, if employed for longer
periods, a current meter inserted into the flow will accumulate debris causing it to malfunction
and possibly break away. This method should therefore only be used for short-term
measurements such as during equipment calibration or to develop a rating curve. Two types of
readily available instruments that meet USGS standards are the type AA Price and Pigmy current
meters.
Pressure Sensors
A pressure sensor or transducer measures the dynamic pressure head at a given point in the flow.
The dynamic pressure is a measure of the point velocity and can be used to estimate the mean
velocity of the flow. A common example of a pressure sensor is the pitot tube used on an
airplane or on some boat speedometers.
The same caution described for bubbler tubes must be applied to pressure sensors. That is, as the
velocity of the flow increases above 1.5-1.8 meter/second (5-6 feet/second), a low pressure zone
is induced across the sensor, interpreted by the flow meter as a drop in flow rate. These
instruments should not be used in channels where the slope of the bottom exceeds 5 to 7 %.
Acoustical Sensors
An acoustical sensor emits a sound wave under water across a channel and measures the time
required for the signal's return. Transit time is correlated with channel width. The relative
positions of the emitting and receiving sensors are used to estimate velocity. A minimum depth
of flow is required. This type of sensor can only be used at sites with sufficient base flow to
provide the medium in which the sound wave travels. If there is no base flow, the lower portions
of the rising and falling limbs of the hydrograph will be lost.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
108 April 25, 2002
-------
Float-and-Stopwatch Method
In this method, the time it takes for a float to move a known distance downstream is determined.
Velocity is calculated as the distance traversed divided by the travel time. The characteristics of
a good float are: an object that floats such that it is partially submerged, allowing some
averaging of velocity above and below the surface of the water; an object that is easily observed
and tracked; an object that is not easily affected by wind; and an object that does not cause
problems if not recovered. Citrus fruits such as oranges, limes, or lemons are commonly used as
floats. Ping-pong and styrofoam balls float well but are too light and are easily blown by the
wind (they may also pose environmental problems if not recovered).
In a variation of this method, a vertical float with a weighted end is used. The vertical float
provides a better measure of mean velocity over the depth of the water column than a float
moving primarily at the surface. In addition, it can be designed to minimize bias due to wind.
In most cases, this method is not accurate enough to be of significant utility in stormwater
monitoring studies and is particularly inaccurate for very deep systems and where there is a
significant difference in velocity across the water surface (e.g., in natural channels).
Deflection (or Drag-Body) Method
In this method, the deflection or drag induced by the current on a vane or sphere is used as a
measure of flow velocity. This method is only practical for short-term, real-time measurements,
such as equipment calibration, because an object of this size inserted into the flow will
accumulate debris, causing it to change the hydraulic form, provide inconsistent data, and
(possibly) break away.
Methods Most Suitable for Continuous Velocity Monitoring
Ultrasonic (Doppler) Sensors
An ultrasonic sensor applies the Doppler principle to estimate mean velocity. A sound wave,
emitted into the water, reflects off particles and air bubbles in the flow. The shift in frequency of
waves returning to the sensor is a measure of the velocity of the particles and bubbles in the flow
stream. The instrument computes an average from the reflected frequencies, which is then
converted to an estimate of the average velocity of the flow stream.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 09 April 25, 2002
-------
Figure 3.11: Area velocity sensors module (ISCO)
The sensor is mounted at the bottom of the channel. However, because the ultrasonic signal
bounces off suspended particles, the signal may be dampened (i.e., not able to reach portions of
the flow stream) when suspended solids concentrations are high. The sensor may also be
mounted on the side of the channel, slightly above the invert. Combined with the appropriate
hardware and software, the sensor can filter out background signals associated with turbulence in
the flow.
Ultrasonic Doppler sensors can be used under conditions of either open channel or pressurized
flow. When combined with the hardware and software required for real-time flow measurement,
data logging, and automated sampling, and when properly calibrated, this system is capable of
greater accuracy than one relying on a stage-flow (i.e., Manning's Equation) relationship. The
ultrasonic sensor-based system may be more expensive but the additional expense may be
justified by program objectives. Without routine maintenance, the accuracy of ultrasonic sensors
may decrease due to fouling by surface-active materials and organisms.
Electromagnetic Sensors
Electromagnetic sensors work under the principle stated in Faraday's Law of electromagnetic
induction; that is, a conductor (water) moving through an electromagnetic field generates a
voltage proportional to its velocity. This instrument, mounted at or near the channel bottom,
generates the electromagnetic field and measures the voltage inducted by the flow. Although
velocity is measured at only a single point, that measurement is used to estimate the average
velocity of the flow stream.
Electromagnetic sensors can be pre-calibrated for many types of site configurations. The sensor
is usually mounted at the channel invert but can be mounted on the side of a channel, slightly
above the invert, if high solids loadings are expected. A built-in conductivity probe senses when
there is no flow in the conveyance.
These types of instruments are not sensitive to air bubbles in the water or changing particle
concentrations, as is the ultrasonic sensor, but can be affected by extraneous electrical "noise."
As with the ultrasonic system, when an electromagnetic sensor is combined with the hardware
and software required for real-time flow measurement, data logging, and automated sampling,
and when properly calibrated, it may be capable of greater accuracy in specific circumstances
than a system relying on a stage-discharge relationship. On the other hand, the electromagnetic
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
110 April 25, 2002
-------
sensor-based system may also be more expensive, but the additional expense may be justified by
program objectives.
Acoustic Path
These sensors are used to determine the mean velocity of streams and rivers, and where they are
applicable, they have been found to be one of the most accurate flow measurement systems. The
method consists of an array of sensor elements that are installed at an even elevation across the
channel. The number of sensor elements used is dictated by the channel width (larger channels
require more sensors). Due to the sensor array's height above the channel bottom, its use is
generally limited to larger channels that have a base flow present. It is not practical for smaller
diameter conveyances with no base flow, which may be found at a BMP site. Additionally,
stormwater conduits for BMP runoff can be small enough that a single point measurement for
velocity provides a reasonable estimate for the average velocity. For these reasons, acoustic path
sensors are rarely applicable to BMP monitoring situations.
Water Quality Sample Collection Techniques
Grab Samples
The term "grab sample" refers to an individual sample collected within a short period of time at a
particular location. Analysis of a grab sample provides a "snapshot" of stormwater quality at a
single point in time. Grab samples are suitable for virtually all of the typical stormwater quality
parameters. In fact, grab samples are the only option for monitoring parameters that transform
rapidly (requiring special preservation) or adhere to containers, such as oil and grease, TPH, and
bacteria.
The results from a single grab sample generally are not sufficient to develop reliable estimates of
the event mean pollutant concentration or pollutant load because stormwater quality tends to
vary dramatically during a storm event. Nevertheless, grab sampling has an important role in
many stormwater monitoring programs for the following reasons:
A single grab sample collected during the first part of a storm can be used to characterize
pollutants associated with the "first flush." The first part of a storm often contains the
highest pollutant concentrations in a storm runoff event, especially in small catchment areas
with mostly impervious surfaces, and in storms with relatively constant rainfall. In such
cases, the first flush may carry pollutants that accumulated in the collection system and
paved surfaces during the dry period before the storm. Thus, the results from single grab
samples collected during the initial part of storm runoff may be useful for screening-level
programs designed to determine which pollutants, if any, are present at levels of concern.
However, this strategy may be less effective in areas subject to numerous low-intensity, long
duration storms with short inter-event times, because "first flush" effects are less obvious
under such weather conditions.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
111 April 25, 2002
-------
Some measurable parameters, such as temperature, pH, total residual chlorine, phenols,
volatile organic compounds (VOCs), and bacteria transform or degrade so rapidly that
compositing can introduce considerable bias. (Note: Grab sampling is the typical method
for VOCs because VOCs can be lost through evaporation if samples are exposed to air during
compositing. However, as discussed in Section 3.2.1 some automated samplers can be
configured to collect samples for VOC analysis with minimal losses due to volatilization).
Some pollutants, such as oil and grease and TPH, tend to adhere to sample container surfaces
so that transfer between sampling containers must be minimized (if program objectives
require characterization of the average oil and grease concentration over the duration of a
storm, obtain this information from a series of grabs analyzed individually).
To estimate event mean concentrations or pollutant loads, you could collect a series of grab
samples at short time intervals throughout the course of a storm event. There are several
different approaches for obtaining information from a series of grab samples. One approach
would be to analyze each grab sample individually. If the samples are analyzed individually, the
results can be used to assess the rise and fall of pollutant concentrations during a storm and to
estimate event mean concentrations of pollutants. This approach can be particularly useful if the
monitoring objective is to discern peak pollutant concentrations or peak loading rates for
assessing short-term water quality impacts. Analyzing each grab separately adds significantly to
laboratory costs; consequently, this approach is rarely used except when program objectives
require detailed information about changes in constituent concentrations over the course of a
storm.
Composite Samples
Another approach is to combine appropriate portions of each grab to form a single composite
sample for analysis, but this is generally impractical if there are more than a few stations to
monitor. Moreover, manual monitoring can be more costly than automated monitoring if your
program encompasses more than a few storm events. For these reasons, many monitoring
programs have found that the use of automated monitoring equipment and methods are more
appropriate for compiling composite samples than manual monitoring. If detecting peak
concentrations or loading rates is not essential, composite sampling can be a more cost-effective
approach for estimating event mean concentrations and pollutant loads. A composite sample is a
mixture of a number of individual sample "aliquots." The aliquots are collected at specific
intervals of time or flow during a storm event and combined to form a single sample for
laboratory analysis. Thus, the composite sample integrates the effects of many variations in
stormwater quality that occur during a storm event. Composite samples are suitable for most
typical stormwater quality parameters, but are unsuitable for parameters that transform rapidly
(e.g., fecal coliform, residual chlorine, pH, volatile organic compounds) or adhere to container
surfaces (e.g., oil and grease).
The two basic approaches for obtaining composite samples are referred to as time-proportional
and flow-proportional. A time-proportional composite sample is prepared by collecting
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
112 April 25, 2002
-------
individual sample "aliquots" of equal volume at equal increments of time (e.g., every 20
minutes) during a storm event, and mixing the aliquots to form a single sample for laboratory
analysis. Time-proportional samples do not account for variations in flow; pollutant
concentrations in sample aliquots collected during the portion of the storm with lower flows are
given the same "weight" as sample aliquots collected during higher flows. Consequently, time-
proportional composite samples generally do not provide reliable estimates of event mean
concentrations or pollutant loads, unless the interval between sample aliquots is very brief and
flow rates are relatively constant.
Flow-weighted composite samples are more suitable for estimating event mean concentrations
and pollutant loads. The event mean concentration is discussed in detail in Section 2.5.3. A
flow-weighted composite sample can be collected in several ways (EPA 1992):
1. Constant Time - Volume Proportional to Flow Rate - Sample aliquots are collected at
equal increments of time during a storm event, and varying amounts of each aliquot are
combined to form a single composite sample. The amount of water removed from each
aliquot is proportional to the flow rate at the time the aliquot was collected. This type of
composite sample can be collected using either manual or automated techniques.
2. Constant Time - Volume Proportional to Flow Volume Increment - Sample aliquots
are collected at equal increments of time during a storm event, and varying amounts from
each aliquot are combined to form a single composite sample. The amount of water
removed from each aliquot is proportional to the volume of flow since the preceding
aliquot was collected. This type of compositing is generally used in conjunction with an
automated monitoring system that includes a continuous flow measurement device. It
can be used with manual sampling in conjunction with a continuous flow measurement
device, but this combination is uncommon.
3. Constant Volume - Time Proportional to Flow Volume Increment - Sample aliquots of
equal volume are taken at equal increments of flow volume (regardless of time) and
combined to form a single composite sample. This type of compositing is generally used
in conjunction with an automated monitoring system that includes a continuous flow
measurement device.
Select the flow-weighted compositing method most suitable for your program based on the
monitoring technique (manual or automated) and equipment you plan to use. Compositing
Methods 2 and 3 are more accurate than Method 1 because Methods 2 and 3 use the total volume
of flow based on continuous flow measurement to scale the sample volume; in contrast, Method
1 uses a single instantaneous rate measurement to estimate the flow over the entire sampling
interval. However, if you intend to use manual methods, compositing Method 1 is generally the
most practical choice. If automated equipment is to be used, Method 3 is generally preferred
because it minimizes the need for measuring and splitting samples, activities that can increase
the chance for sample contamination. If you plan to use automated methods, review the
equipment manufacturer's specifications and instructions to select the compositing method most
appropriate for that particular make and model.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
113 April 25, 2002
-------
Storm events affect stream flows for variable lengths of time depending on the storm duration
and antecedent conditions and catchment characteristics. Runoff may persist for a period of a
few hours to one to two days. This suggests runoff rarely persists long enough to be considered
comparable to chronic exposure duration. Discrete sampling over the course of the storm event
will provide concentration information that can be used to determine how long water quality
criteria were exceeded during the storm. Alternatively, discrete samples can be composited on a
time-weighted basis over time scales comparable to the acute and chronic water quality criteria
exposure periods (one hour and four days) respectively. However, the latter would likely include
dry-weather flows since few storms last four days. For catchments which are relatively small (a
few acres), it is recommended one or more one-hour composite samples be collected during the
first few hours of flow by collecting and combining three or more grab samples.
Flow-weighted composite sampling can be used for comparison with water quality objectives
(for example, if flow-weighted composites are collected to measure loads). However, it should
be recognized that a flow-weighted sample would contain more water from peak flows than from
the initial part of the storm. Results from Santa Clara Valley Nonpoint Source Monitoring
Program indicated that for a large watershed with significant suspended sediment concentrations
(200 - 400 mg/L), peak total metals concentrations are generally 1.5 times the flow-weighted
composite concentrations (WCC 1993). Results from monitoring a smaller, highly impervious
industrial catchment with the lower suspended sediment concentrations were more variable, and
no conclusions could be drawn as to the relationship between flow-composite concentrations and
grab samples due to difficulties in grab sampling runoff that only occurred during precipitation.
Automatic Sampling
Automated monitoring involves sample collection using electronic or mechanical devices that do
not require an operator to be on-site during actual stormwater sample collection. It is the
preferred method for collecting flow-weighted composite samples. Automated monitoring is
generally a better choice than manual monitoring at locations where workers could be exposed to
inadequate oxygen, toxic or explosive gases, storm waves, and/or hazardous traffic conditions.
Also, automated methods are better than manual methods if you are unable to accurately predict
storm event starting times. Automated samplers can be set so that sampling operations are
triggered when a pre-determined flow rate of storm runoff is detected. Conversely, manual
monitoring relies on weather forecasts (and considerable judgment and good luck) to decide
when to send crews to their monitoring stations. It is very difficult to predict when stormwater
runoff is likely to begin; consequently, manual monitoring crews may arrive too early and spend
considerable time waiting for a storm that begins later than predicted, or they may arrive too late
and miss the "first flush" from a storm that began earlier than predicted. If the automated
equipment is set to collect flow-weighted composite samples using the constant volume-time
proportional to flow method, it reduces the need to measure samples for compositing.
If you have determined that field-measured "indicator" parameters (e.g., turbidity, conductivity,
dissolved oxygen, pH) are sufficient for your monitoring objectives, consider using electronic
sensors and data loggers. Using electronic sensors and data loggers, you can obtain near-
continuous measurements of indicator parameters at reasonable cost.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
114 April 25, 2002
-------
BMP monitoring can be an especially useful application for some automated systems (e.g.,
continuous flow recorders, auto samplers, continuous monitoring probes) for the following reasons:
Automated systems can provide data covering virtually the entire volume of runoff that passes
through the BMP (i.e., they are not likely to miss or leave out small events and the beginnings
and ends of other events).
Automated systems are well suited to providing data sets that are useful (recognizing that
performance evaluations are generally based on the differences between inlet and outlet
concentration data sets, both of which are inherently noisy).
The information obtained from good performance monitoring programs can be very valuable
by protecting against inappropriate BMP applications. Therefore, the cost of using
automated systems is often justifiable.
Automatic Sampling Equipment
An automated sampler is a programmable mechanical and electrical instrument capable of
drawing a single grab sample, a series of grab samples, or a composited sample, in-situ. The
basic components of an automated sampler are a programming unit capable of controlling
sampling functions, a sample intake port and intake line, a peristaltic or vacuum/compression
pump, a rotating controllable arm capable of delivering samples into sample containers and a
housing capable of withstanding moisture and some degree of shock. Commonly used brands
include: ISCO, Lincoln, Nebraska; American Sigma, Medina, New York; Manning, Round
Rock, Texas; and Epic/Stevens, Beaverton, Oregon.
An automated sampler can be programmed to collect a sample at a specific time, at a specific
time interval, or on receipt of a signal from a flow meter or other signal (e.g., depth of flow,
moisture, temperature). The sampler distributes individual samples into either a single bottle or
into separate bottles which can be analyzed individually or composited. Some automated
samplers offer multiple bottle configurations that can be tailored to program objectives.
Important features of automated samplers include:
Portability.
Refrigeration.
Volatile organic compound (VOC) sample collection (if needed).
Alternative power supplies.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
115 April 25, 2002
-------
Figure 3.12: Automatic sampler (American Sigma Inc.)
Portable samplers are smaller than those designed for fixed-site use, facilitating installation in
confined spaces. If a suitable confined space is not available or undesirable (e.g., because of
safety issues), the sampler can be housed in a secure shelter at the sampling site. Portable
samplers can use a 12V DC battery power supply, solar battery, or AC power.
Although none of the portable samplers currently available are refrigerated, ice may be added to
the housing of some units to preserve collected samples at a temperature as close to 4ฐC as
possible. The objective of this cooling is to inhibit pollutant transformation before the sample
can be analyzed. Refrigerated samplers hold samples at a constant temperature of 4ฐC.
However, their large size and requirement for a 120V AC power prohibit most field installations.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
116 April 25, 2002
-------
Figure 3.13: VOC sampler (ISCO)
An automated sampler designed for VOCs is currently available from ISCO. The bladder pump
used by this instrument minimizes physical disturbance of the sample (as opposed to the physical
disturbances imparted by peristaltic vacuum pumps), reducing the loss of volatile compounds.
The VOC sampler distributes the sample into sealed 40-ml sample bottles, as required by EPA
protocol. However, at present, the caps for the sample bottles are not compatible with automated
laboratory equipment, requiring more handling in the laboratory.
In typical installations for BMP sampling, for each of the types of samplers described above, an
intake line is bracketed to the channel bottom. The intake tubing should be mounted as
unobtrusively as possible, to minimize disturbance of the site hydraulics. Generally, the
optimum position for the intake is to the channel bottom. However, if high solids loadings are
expected and potential deposition could occur, the intake can be mounted slightly higher on one
side of the channel wall. Typically, a strainer is attached to the intake to prevent large particles
and debris from entering the tubing. The strainer is usually installed so that it faces upstream,
into the flow. This configuration minimizes the development of local turbulence that could
affect representative sampling of constituents in the particulate phase.
Two types of pumps are incorporated into automated samplers for typical water quality sampling
(i.e., not VOC sampling): peristaltic and vacuum/compressor. A peristaltic pump creates a
vacuum by compressing a flexible tube with a rotating roller, drawing a sample to the pump that
is then pushed out of the pump. Field experience with peristaltic pumps has shown that their
reliability in drawing a consistent sample volume is greatly reduced as the static suction head
(i.e., distance between the flow stream surface and the sampler) increases. It may be possible to
increase the efficiency of these samplers by placing the pump closer to the sample source,
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
117
April 25, 2002
-------
reducing the suction head. In general, the sampler itself should be installed no more than 6
meters (20 feet), and preferably less, above the channel bottom. If the sampler is to be installed
at greater than 20 feet above the channel invert, it may be necessary to use a remote pump that is
placed closer to the flow stream to ensure reliable sample collection.
The degree to which sampler lift affects the concentration of total suspended solids and other
pollutant parameters (especially coarser materials) is not well known. That is, the mean transport
velocity achieved by the peristaltic pump is sufficient to draw suspended solids; however, the
pulsed nature of the flow may allow suspended solids to settle back down through the pump
tubing during transport. In work performed by the USGS (FHWA 2001), it was found that
suspended solids concentrations did not vary with pumping height (0 to 24 feet); however,
sample volumes delivered to sample bottles did vary from sample to sample at high lift heights
for some of the older sampler models.
Another concern with peristaltic pumps is their incompatibility with Teflon-lined tubing in the
pump assembly. Compression of the intake tubing by the rollers tends to create stress cracks and
small recesses in the lining where particles can accumulate. Under these circumstances, some
pollutant concentrations could be underestimated and the cross-contamination of samples can
occur. Although Teflon-lined tubing is preferable because it reduces the potential loss of
pollutants through surface interactions, this advantage cannot be accommodated with a peristaltic
pump.
A vacuum/compressor pump draws a sample by creating a vacuum. This type of pump can
create a higher transport velocity in the intake tube and provide a more steady and uniform
discharge than a peristaltic pump. However, the higher intake velocity can scour sediments in
the channel near the sampler intake, resulting in disproportionately high concentrations of
suspended solids.
After a sampler is installed, it must be programmed to collect the desired sample size.
Calibration of peristaltic pumps is achieved by one of two methods: automatic or timed. In
automatic calibration, the actual volume of sample drawn is measured using a fluid sensor
located at the pump and the known pump speed. In timed calibration, the volume is determined
from the number of revolutions of the peristaltic pump and the time taken for the sample to travel
from its source to the sample container. Calibration by this latter method is site specific,
incorporating the pump speed, the head (vertical distance above the sample source), and the
length and diameter of the intake tubing. The Manning and Epic samplers, which employ
vacuum pumps, permit adjustment for specific sample volumes via a fluid level device in a
chamber. This chamber can cause sample cross-contamination, as it cannot be flushed as the
tubing can.
Overland Flow Sampler
An overland flow sampler is a non-automated sampler that can be used to take discrete grab
samples or a continuous sample over some duration. This type of sampler may be useful for
collecting stormwater samples for certain types of BMPs (upstream of catch basins). One
manufacturer's (Vortox, Claremont, California) unit within this class of samplers consists of an
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
118 April 25, 2002
-------
upper ball valve, a lower ball valve (through which runoff enters), and a sample container. The
upper valve can be adjusted to control the rate of intake, allowing continuous sampling of storm
events of different durations, provided depth of flow is not highly variable. The lower ball valve
seals and closes the intake when the water level reaches the top of the container.
Overland flow samplers (manufactured by Vortex) are available in two sizes: 3 liters (0.8
gallon) and 21 liters (5.5 gallons). They can be set into existing sumps or in the ground, but they
must be installed with the top of the sampler flush with the ground surface.
This instrument is inexpensive and simple to operate. Since the overland flow is not
concentrated, there are no other methods for collecting this flow. However, this sampler is not
capable of taking flow or time-weighted composites or of sampling the entire flow during a large
storm event. In fact, there is no way of knowing what part of the storm was actually sampled,
especially where flow depths are variable. Recently, the USGS developed and began testing an
automated overland flow sampler that may be capable of time-weighted composite sampling.
In-situ Water Quality Devices, Existing Technology
The concentration of most pollutants in stormwater runoff is likely to vary significantly over the
course of a given storm event. Some of this variability can be captured through the collection of
multiple samples. The ideal data set would contain not just multiple samples, but also a
continuous record of constituent concentrations throughout a storm, capturing both the timing
and magnitude of the variations in concentration. Given the availability of other continuous data,
this approach might allow better correlation with potential causative factors. Unfortunately, the
laboratory costs for even a near-continuous data set would be prohibitive. USGS determined that
between 12 and 16 individual samples resulted in a mean that was within 10 to 20 percent of the
actual event mean concentration (FHWA 2001). In-situ monitoring devices offer a possible
solution to obtaining a continuous record of water quality; however, at this time, they are only
practical for a limited set of parameters.
In-situ water quality probes have been adapted from equipment developed for the manufacturing
and water supply/wastewater industries. In-situ water quality monitors attempt to provide the
desirable near-continuous data set described above at a relatively low cost, eliminating (or
reducing) the need for analysis of samples in the laboratory.
In general, water quality monitors are electronic devices that measure the magnitude or
concentration of certain specific target constituents through various types of sensors. Discrete
measurements can be made at one minute or less intervals. Most monitors use probes that
provide a controlled environment in which a physical and/or electrochemical reaction can take
place. The rate of this reaction is typically driven by the concentration of the target constituent in
the flow. The rate of reaction, in turn, controls the magnitude of the electrical signal sent to the
display or a data-logging device.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
119 April 25, 2002
-------
Probes to detect and measure the following physical and chemical parameters are currently
available for practical use in the field:
Physical parameters
Temperature
Turbidity
Chemical parameters
pH
Oxidation-reduction potential (redox)
Conductivity
Dissolved oxygen
Salinity
Nitrate
Ammonia
Resistivity
Specific conductance
Ammonium
There are some potential probes for heavy metals, but given the complexities associated with
highly variable solids concentrations and other factors, studies have found that they are not
practical for field application (FHWA 2001). Instruments can be configured to measure the
concentrations of several of these parameters simultaneously (i.e., multi-parameter probes) and
provide data logging and PC compatibility. Manufacturers of this type of instrument include
YSI, Inc., Yellow Springs, Ohio; ELE International, England; Hydrolab, Austin, Texas; Solomat,
Norwalk, Connecticut; and Stevens, Beaverton, Oregon.
In many cases, the electrochemical reaction that drives a probe's response is sensitive to changes
in temperature, pH, or atmospheric pressure. Where appropriate, monitors are designed to
simultaneously measure these associated properties. Data on the target constituent are then
corrected through a mathematical routine built into the probe's microprocessor (e.g., dissolved
oxygen probes are compensated for temperature and atmospheric pressure, pH probes for
temperature and ammonia probes for pH), or are adjusted in a spreadsheet after downloading to a
personal computer.
Despite the advantage of these instruments for measuring near-continuous data, they require
frequent inspection and maintenance in the field to prevent loss of accuracy due to fouling by oil
and grease, adhesive organics, and bacterial and algal films. Therefore, these instruments should
always be cleaned and calibrated before use. Because water quality probes are designed to
operate while submerged in water, exposure of the electrochemically active probe surface to air
should be minimized.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
120 April 25, 2002
-------
In-situ Water Quality Devices, Future Technologies
There are several in-situ water quality devices that are used by industry but are not currently
applicable to stormwater monitoring. However, as the technology advances they may become
applicable and therefore are discussed in this section.
Ion-Selective Electrodes
An ion-selective electrode places a selectively permeable membrane between the flow and an
internal solution of known ionic strength. The voltage differential across the membrane is
proportional to the difference in ionic strength between the two solutions. Ion-selective probes
are currently available for the ionic forms of a number of parameters, including ammonia,
ammonium, copper, lead, nitrate, and nitrite.
An ion-selective electrode is specific to the targeted ion and will not measure other ions or other
complexed forms. For example, depending on the target parameter, a nitrate-selective electrode
will not measure the concentration of nitrite in the flow. However, these instruments are
sensitive to interference from other ions, volatile amines, acetates, surfactants, and various weak
acids. At present, the degree of interference can be judged only by comparing the performance
of the probe to that of one in a reference solution, a procedure likely to prove unwieldy in the
field. Consequently this type of probe is not typically used for stormwater monitoring.
On-Line Water Quality Analyzers
On-line water quality analyzers are spectrometers, similar to those used in analytical laboratories.
A light source that generates a known intensity of light over a range of wavelengths (i.e.,
ultraviolet or infrared) is transmitted through a sample introduced into a flow cell. The
instrument collects light absorbency information at multiple wavelengths and produces a light
absorbency signature (manufacturer's specifications, Biotronics Technologies, Inc., Waukesha,
Wisconsin, and Tytronics, Inc., Waltham, Massachusetts). The instrument is calibrated using 30
or more randomly varied mixtures of standards; the ultraviolet (UV) light-absorbency
characteristics of a sample are then compared to a baseline calibration file of known "UV
signatures."
On-line analyses are used in the water treatment and wastewater industries. Until recently, on-
line spectrometric analyzers were impractical for stormwater field use. The state of technology
of these systems was comparable to that in the field of computers 20 years ago: large machines
requiring a controlled laboratory environment were operated by highly trained specialists.
However, an increased demand for portability, the increased power and decreased cost of
microprocessor technology, the development of new statistical and mathematical analysis
software, and the availability of standardized control systems (i.e., communication interfaces,
actuators, and programmable controllers) have fostered the emergence of a new generation of
instruments.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
121 April 25, 2002
-------
Three types of spectrometers are currently available or under development for environmental
applications:
Ultraviolet-Array Spectroscopy (UVAS) employs a broad spectrum light generated by a
Xenon lamp and delivered to the sample through fiber optic cables. Light is transmitted
through the sample in specially designed optical probes. The light transmitted through the
sample is collected and returned to the analyzer where it is dispersed into wavelengths and
projected onto a photodiode detector array. Current applications are the detection of multiple
contaminants (metals, nitrates, organics, and aromatic hydrocarbons) in groundwater, the
detection of metals (chromium, zinc, and mercury) in industrial wastewater, and water
treatment quality parameters (copper, iron, molybdate, triazole, phosphorate) in industrial
processes and cooling waters.
Liquid Atomic Emission Spectrometry (LAES) employs a photodiode detector array similar
to that used in UVAS. A high-energy arc is discharged directly into the liquid as the source
of excitation and the resulting atomic light emission is analyzed by special pattern
recognition techniques. Qualitative analysis is derived from the detection of emission lines
and quantitative analysis is a function of intensity. Use of LAES has been demonstrated for
the analysis of metals, hydrogen, and sulfur.
Like UVAS, Near Infrared (NIR) analysis employs the transmission of light through a liquid.
This technology has been used extensively in the food processing industry and is under
evaluation for application elsewhere.
To date, portable on-line analyzers have not been tested extensively for use in stormwater or
BMP monitoring. The "ChemScan" analyzer, manufactured by Biotronics Technologies, Inc., is
reported to adjust automatically for changes in the turbidity of the flow and fouling of the optical
windows, features which suggest applicability to stormwater situations. According to the
manufacturer, routine maintenance is limited to a periodic baseline correction and occasional
chemical cleaning of the flow cell.
Particle Size Analyzers
There is a particle size analyzer available that can be installed in-situ. It employs laser
diffraction to determine the particle size distribution. However, the unit costs approximately
$30,000, is 3 feet long and 5 inches in diameter, and is required to be submerged. Currently it is
not applicable for stormwater monitoring.
Research is currently being conducted on applying ultrasonics for particle size analysis.
However, it is presently not available for stormwater application.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
122 April 25, 2002
-------
In-situ Filtration and Extraction System
Axys Environmental Systems, Ltd., British Columbia, Canada manufactures an in-situ filtration
and extraction system for monitoring trace organics, metals, and radionuclides in stormwater.
These systems retain the target pollutant on a resin filter as a portion of the flow passes through.
After the storm event, the filter is taken to the laboratory and the pollutant is removed through
solid phase extraction. The filtration system is comprised of a microprocessor, a pump, a flow
meter, and a DC power supply. A prefilter for suspended solids can be attached if levels high
enough to clog the resin filter are anticipated. Pollutants trapped in the prefilter can also be
extracted and analyzed.
These systems can be programmed so that samples of the flow pass through the filter at equal
time intervals, or so that signals from an external flow meter trigger flow- or time-weighted
composite sampling. As with other types of automated samplers, the sampling history is stored in
internal memory.
Filtration and extraction systems reduce the potential for contamination of a sample during
handling in the field and eliminate the need to transport large volumes of water to an analytical
laboratory. The detection limit of the samples depends on the amount of water flowing through.
Because large volumes of water can be passed through the system, even very small
concentrations of pollutants can be detected. On the other hand, where suspended sediment
concentrations are high, the prefilter may become clogged as a large volume of water passes
through it. Metals can be lost from the filter if the pH drops to 6.0 or lower, and resin filters are
available for only a limited number of pollutants. Due to the potential for clogging, this
methodology may not be useful for BMP monitoring sites.
Remote Communications with Automatic Equipment
The ability to remotely access the memory and programming functions of automated samplers is
a highly desirable feature for large stormwater sampling networks. Although this feature
increases the capital cost for a system, it can greatly reduce the expertise and training necessary
for field crews because many of the technical aspects of equipment set-up and shut-down can be
conducted by a system supervisor remotely.
Currently, modem communication is an available option to most commercially produced
automated samplers. However, there are several common drawbacks that may be encountered
with the communication systems currently offered by manufacturers:
Full access to all sampler programming features is limited. This means that trained field
crews may still be necessary to ensure sampler programming is correct.
For multiple instrument systems (i.e., separate flow meter and automated sampler)
communication and complete operation of both components through one modem system is
generally not available.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
123 April 25, 2002
-------
Remote communication for both samplers and flow meters is a rapidly advancing technology,
and companies like American Sigma and ISCO are developing systems that address the problems
described above.
Manual Sampling
Manual monitoring involves sample collection and flow measurement by personnel using hand-
operated equipment (e.g., bailer, bottle). For a monitoring program that is modest in scope (i.e.,
relatively few sampling sites and storm events), manual methods for obtaining grab and
composite samples may be preferable to those employing automated equipment. Also, if your
program requires monitoring large streams, you may need to use manual methods in order to
collect cross-section composites. The principal advantages to manual sampling are its relatively
low capital cost and high degree of flexibility. In addition to the capital outlay required for the
purchase of automated samplers, other costs, such as installation, training personnel to use the
samplers correctly, and field maintenance and operations (replacing batteries, interrogating data
loggers, retrieving and cleaning sample jars) can be substantial.
Manual sampling is usually preferred under the following circumstances:
When available resources for equipment purchase/installation (e.g., funds, personnel, time) are
very constrained and/or there is not the political will to invest in a program, despite the inherent
value of the resultant information.
When the target pollutants are ones that do not lend themselves to automated sampling or
analysis (e.g., oil and grease, volatile organic compounds, bacteria).
When the physical setting of the BMP does not allow the use of automated systems.
However, manual monitoring may not be feasible if:
Monitoring personnel are not available after normal working hours.
Monitoring personnel have strict job descriptions that do not include sampling.
The organization's insurance policy doesn't cover stormwater monitoring activities.
Managers and monitoring personnel are not able to deal with sick days, vacations, and
competing priorities.
Manual sampling is generally less practical than automated monitoring for large-scale programs
(e.g., monitoring programs involving large numbers of sites or sampling events over multiple
years). It is difficult to collect true flow-weighted composites using manual methods. Under
these circumstances, labor costs and logistical problems can far outstrip those associated with
automated equipment. For the same reason, manual sampling is seldom practiced if specific
program objectives require that samples be composited over the entire duration of a storm, which
is recommended for BMP monitoring.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
124 April 25, 2002
-------
Manual equipment can be used in collecting grab samples, composite samples, or both, as
described below.
Manual Grab Sampling Equipment
Manual sampling techniques and equipment have been reviewed in more detail by Stenstrom and
Strecker (1993). If site conditions allow, a grab sample can be collected by holding the
laboratory sample bottle directly under the lip of an outfall or by submerging the bottle in the
flow. A pole or rope may be used as an extension device if field personnel cannot safely or
conveniently approach the sampling point. Alternatively, a clean, high-density polyethylene
bucket may be used as a bailer and sample bottles may be filled from the bucket. Care should be
taken not to stir sediments at the bottom of the channel.
As described earlier, the concentrations of suspended constituents tend to stratify within the flow
stream depending on their specific gravity and the degree to which flow is mixed by turbulence.
Use of a discrete-depth sampler for multiple samples should be considered when constituents
lighter or heavier than water are targeted, or if the flow is too deep and/or not well mixed enough
to be sampled in its entirety (Martin et al. 1992). However, stormwater BMPs often drain
relatively small catchments and contain fairly shallow flows. Collection of depth-integrated
samples at these sites is not usually performed.
Given the extremely low detection limits that laboratory analytical instruments can achieve,
leaching of water quality constituents from the surface of a bailing device or sample bottle can
affect water quality results. Sample bottles of the appropriate composition for each parameter
are usually available from the analytical laboratory. Depending upon the pollutant to be
analyzed, bailers and discrete-depth samplers should be made of stainless steel, Teflon coated
plastic, or high-density polyethylene. When in doubt, a laboratory analyst should recommend an
appropriate material type for the collection device.
Manual Composite Sampling Equipment
If grab samples will be composited based on flow rate (i.e., grab samples collected during high
flow contribute more to the composited sample than those collected during low flow), some
receptacle for storing the individual grab samples prior to compositing will be required. The use
of polyethylene jugs, or the polyethylene cubes with screw-on caps manufactured for shipping
chemicals, is recommended. These can be shaken to remix the sample prior to pouring out the
required volume. The volume required from each receptacle can be measured in a graduated
cylinder and poured into a bucket for compositing. Both the cylinder and the bucket should be
made from a Teflon-coated plastic or high-density polyethylene and should be cleaned prior to
use.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
125 April 25, 2002
-------
3.2.4.3 Error Analysis and Measurement Accuracy
Every measurement has an unavoidable uncertainty due to the precision of the measuring tool,
the accuracy of the calibration, and the care with which the measurement is made. If all other
sources of error are minimized or removed, then the uncertainty in the measurement is generally
on the same order of the smallest numerical value that can be estimated with the measuring
instrument. The true value is typically contained in the range of values reflecting the
experimental uncertainty of the measurement. Calculating the mean of multiple measurements if
the measurement errors are random in nature and not systematic can provide a better estimate of
the true value.
Indeterminate (random) errors result from instrument precision, calibration, and inaccuracies in
the measuring process. The size and magnitude of indeterminate errors cannot be determined
(hence the name) and result in different values from a measuring process when the process is
repeated. There are several ways indeterminate errors can be introduced, including operator
error, variation in the conditions in which the measuring process is conducted, and the variability
of the measuring instrument.
Determinate (systematic) errors have an algebraic sign and magnitude and result from a specific
cause introducing the same error into every measurement. Determinate errors are more serious
than indeterminate errors because taking the average of multiple measurements cannot reduce
their effects. This is because determinate errors have the same sign and magnitude, which
prevents positive and negative errors from off setting each other. Causes of this type of error can
include operator bias, (consistent) operator error such as incorrect reading of the instrument, or
improper calibration of the measuring instrument.
Expressing Errors
Absolute and relative methods are the standard forms for expressing errors. Absolute error is
expressed as a range of values reflecting the uncertainty in the measurement and is reported in
the same units as the measurement. Measured values followed by the + sign express the absolute
error.
Relative (or fractional) error is expressed as the ratio of the uncertainty in the measurement to the
measurement itself. This is difficult to estimate, because it is a function of the true value of the
quantity being measured, which is unknown, otherwise the error estimate would be zero.
Typically this error estimate utilizes the measured value as the "true" value.
The type of measurement and instrumentation can provide an indication of the appropriate form
of expressing the error. For example, a pressure probe used to measure depth of flow is likely to
have the accuracy of the instrument expressed as a relative percent, while readings on a staff
gauge would have an absolute error related to the markings on the gauge. In these instances the
reported depth measurements would be expressed in the same manner as the precision of the
measuring instrument.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
126 April 25, 2002
-------
Propagation of Errors
Quite often, measurements taken of one or more variables are used in equations to calculate the
value of other variables. For example, to calculate the area of a rectangle, the length and width
are usually measured. For a cube, the length, width, and height are measured to calculate the
volume. Each measurement has a potential error associated with it and, as a result, the variable
calculated from the individual measurements will also contain some error. The magnitude of the
error in the calculated variable can be of a different order than the error associated with any one
of the measurements depending on the algorithm that describes their relationship.
A detailed discussion of the propagation of errors and methods for calculating estimates of errors
as a result of propagation are provided in Appendix A.
3.2.5 Recommendation and Discussion of Storm Criteria
The establishment and application of appropriate storm selection criteria can be a challenging aspect
of planning BMP monitoring programs. Ideally, one would want to obtain data from all phases of
all storms for as long a study period as possible, for the following reasons:
To know what the BMP does during periods of very low flow, normal flow, and very high
flows. Some BMPs' performance varies dramatically with throughput rate (some may even
release pollutants that had been previously trapped).
To estimate performance on the basis of differences of relatively noisy data sets (i.e., inlet
versus outlet data). This intensifies the value of large volumes of credible data (not just a few
samples from portions of a few storms).
To characterize the water quality of dry weather flows for some BMPs with significant wet
storage and/or base flows. This is particularly important when the wet volume of the BMP is
large relative to the storm event. The comparison of inflow to outflow during a storm event is
not valid because the outflow may have little or no relationship to the incoming storm. This
mistake has been made often in past studies.
Despite the desire for extensive and high quality data, there is still a need to tailor your methods to
be consistent with available resources. The types of storms to be monitored and optimal temporal
distribution of monitoring events also should be considered during project planning (Caltrans
1997).
3.2.5.1 Storm Characteristics
The application requirements for NPDES permits that require monitoring specify that
"representative" storms must be monitored. As defined in the regulations, a "representative"
storm must yield at least 0.1 inch of precipitation; must be preceded by at least 72 hours with less
than 0.1 inch of precipitation; and, if possible, the total precipitation and duration should be
within 50 percent of the average or median storm event for the area. Programs that are not part
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
127 April 25, 2002
-------
of the NPDES permit application process or in fulfillment of an NPDES permit may have other
requirements.
In general, it is desirable to monitor a broad range of storm conditions rather than just
"representative" storms as they are really not representative in many cases. For example, in the
Pacific Northwest, it is often difficult (and rare) to identify storms where there has been a 72-
hour dry period prior to the storm.
Because the initial objective of the monitoring is to consider a "worst-case" picture, it is
desirable to select storms with the highest pollutant concentrations rather than a representative
mix of storms. Worst-case conditions are likely to occur after long antecedent dry periods (72
hours to 14 days). Therefore, if feasible, storms should be selected with antecedent periods
greater than 72 hours. Few relationships between storm volume and water quality have been
observed. Lacking any basis for storm volume selection for worst-case conditions, and
acknowledging that storm characteristics are highly dependent on climatic region, the following
may be used as a starting point:
Rainfall Volume: 0.10 inch minimum
No fixed maximum
Rainfall Duration: No fixed maximum or minimum
Typical Range: 6 to 24 hours
Antecedent Dry Period: 24 hours minimum
Inter-event Dry Period: 6 hours
If these criteria prove inappropriate for your situation, you can develop site-specific storm event
criteria by analyzing long-term rainfall records using EPA's SYNOP or another appropriate
analytical program such as EPA's SWMM model (which incorporates the features of SYNOP).
It should be noted that biasing the storm selection to the "worst case" would not provide a
representative sample of the population of all types of storm events. The resulting data should be
used in screening mode and not to estimate statistically derived exceedance frequencies. The
level of effort required to sample all representative types and combinations of storm conditions
in order to generate reliable population statistics is beyond the resources of most agencies. For
this reason, it is recommended a "worst case" approach be taken. Often permits require that you
monitor "representative" storms that have been predefined. Operationally and practically, storm
event criteria may need to be further defined beyond the regulatory definition. The use of a
probability of rainfall above a certain magnitude, during a specific period, based on a
quantitative precipitation forecast (QPF) serves as a good indication of when and how to
mobilize for monitoring efforts. QPFs for a geographic area can be obtained from the National
Weather Service and site specific information can be obtained from private weather consultants.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
128 April 25, 2002
-------
3.2.6 Recommendation and Discussion of QA/QC
Prior to sample collection, you should prepare a Quality Assurance/Quality Control (QA/QC)
plan that describes the sample collection and laboratory analysis procedures. The first step in
preparing a QA/QC plan is to determine the data quality objectives (DQOs) appropriate to your
program. Ideally, the QA/QC plan should be prepared by someone with a good understanding of
chemical analytical methods, field sampling procedures, and data validation procedures. Select
an analytical laboratory that has been accredited to perform the analyses required for your
program. The analytical laboratory should provide its input to ensure the plan is realistic and
consistent with the laboratory's operating procedures.
It is recommended that the QA/QC plan should summarize the project organization, data quality
objectives, required parameters, field methods, and laboratory performance standards for the
measurements. A typical QA/QC plan for stormwater monitoring may include the following
sections:
1. Project Description
2. Project Organization and Responsibility
3. Data Quality Objectives
4. Field Methods
- sample collection methods
- field QA procedures such as equipment cleaning and blanks
- collection of field duplicate samples
- sample preservation methods
- type of bottles for subsampling
5. Laboratory Procedures
- constituents for analysis
- laboratory performance standards (e.g., detection limits, practical quantitation limits,
objectives for precision, accuracy, completeness)
-analysis method references
- frequency and type of laboratory QA samples (e.g., laboratory duplicates, matrix spikes
and spike duplicates, laboratory control samples, standard reference materials)
- data reporting requirements
- data validation procedures
- corrective actions
It is important that you develop your QA/QC plan in concert with your field personnel and your
analytical laboratory. If you have not already done so, you should visit the monitoring locations
to verify that the selected monitoring methods are feasible. Inform your managers of any
modifications to either the DQOs or laboratory performance standards due to field or laboratory
constraints.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
129 April 25, 2002
-------
Potential Sources of Error
This section describes some potential sources of error that can occur in the process of sampling
or transferring monitoring results to a database. These common errors can be specifically
addressed in the QA/QC plan to increase awareness and potentially reduce their occurrence.
In many cases error is introduced in the process of transferring or interpreting information from
the original data records. These errors most likely result from typographical errors or format and
organizational problems. In most cases, water quality data are returned from the lab in some
tabular format. Data are then entered into a database, typically with separate records for each
monitoring station and each storm event. The inconsistency of data formats between monitoring
events can considerably increase the potential for errors in entering data into the database and
subsequently interpreting and using the processed (digital) data.
Where errors in data are present in the processed information, format is often a causative factor.
In some circumstances interpretation of the data presented is not possible due to missing
explanations of the data format; in these cases, data should be excluded. It has been found that
missing records typically have to do with inadvertent skipping of a column or row of data. Errors
in data or parameter type, that were not typographical, typically resulted from misalignment of
rows or columns. Supporting information and useful summaries of parameters, such as
characteristics of the watershed, are often included as text in a general information column, or in
a report or record external to the water quality database. In addition to making the extraction of
this supporting information laborious, checking for errors in information not formatted succinctly
can also be quite cumbersome.
In addition to these "paper" errors, many other opportunities abound for introduction of other
errors, including errors in interpretation and reporting of supporting information (e.g.,
misreading of maps, poor estimates of design, watershed, and environmental parameters, etc.)
and reporting of information from previous studies that may have been originally incorrect.
In addition to potential reporting errors, all field collected and/or laboratory analyzed data on
flow and water quality are subject to random variations that cannot be completely eliminated.
These variations are defined as either "chance variations" or "assignable variations." Chance
variations are due to the random nature of the parameters measured; increased testing efforts and
accuracies cannot eliminate these variations. Although assignable variations cannot be
eliminated altogether, these variations can be reduced and the reliability of the data increased.
Assignable variations are those errors that result from measurement error, faulty machine
settings, dirty containers, etc. Increasing both the length of a study and/or the number of storms
sampled can reduce the assignable variations and increase the reliability of the data (Strecker
1992). Many monitoring studies take place over relatively short periods and have a small
number of monitored storms during those periods. Thus the resultant data sets are often
susceptible to both of these types of variations in addition to any reporting errors.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
130 April 25, 2002
-------
Prepare Health and Safety Plan
As part of the QA/QC plan, the health and safety of personnel involved in the monitoring
program should be considered. Aside from ensuring quality results and efficient implementation
of monitoring procedures, human health and safety are a priority.
The health and safety of field personnel should be considered throughout development of your
monitoring program. You should select monitoring locations and methods that have the lowest
potential for health and safety problems. You should then prepare a health and safety plan. The
first step is an assessment of the physical and chemical hazards likely to be associated with each
monitoring activity. Some of the potential considerations include:
Wet (and possibly cold) weather conditions.
Physical obstructions that complicate access to the site and sample collection point (e.g.,
steep slopes, dense blackberry bushes).
Traffic hazards.
Manholes (i.e., confined space entry, including toxic, explosive, or otherwise unsafe
conditions).
Flooding and fast moving water.
Dim lighting.
Slippery conditions.
Contact with water that could be harmful (e.g., caustic, pathogenic).
Lifting and carrying heavy and bulky pieces of equipment, including carboys and sample
bottles filled with water.
Based on the hazard assessment, identify the appropriate equipment and procedures to protect
field personnel from the potential hazards you have identified. Also, consider adjusting your
monitoring locations and/or methods if necessary to minimize the risk of health and safety
problems.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
131 April 25, 2002
-------
3.2.6.1 Sampling Methods
Proper sampling methods are essential in conducting a BMP monitoring program in order to
ensure resulting data are meaningful and representative of the water and other media being
processed by the BMP. Sampling methodologies and techniques that maintain and confirm the
integrity of the sample are discussed below.
Grab Sample Collection Techniques
During moderate flow events, grab samples can be collected at some stations simply by
approaching the water to be sampled and directly filling up the bottles, being careful not to loose
any preservative already contained in the bottle. It is important also to be aware of surface
conditions of the sampled water body, avoiding layers of algae and debris and areas of dense
vegetation if possible. The bottle cap should be handled carefully, making sure not to introduce
any extraneous dirt, water, debris or vegetation while filling the bottle; bottle caps should not be
placed on the ground facing downward.
Low flow events may not provide sufficient flows to allow filling of bottles directly. In this
case, sample collectors may be used to collect the low flow runoff and transfer the water into the
sample bottles. These sample collectors are typically cup to bucket sized containers with a wide
mouth and no neck, allowing the collector to be placed close to the bottom surface of the flow
path and then filled with the small depth of flow. Sample collectors must be compatible in
material with the sample bottles and the constituents to be analyzed. Sample collectors made of
stainless steel, teflon or glass could be considered after investigating the compatibility of these
materials with each constituent to be analyzed. After each sample bottle has been filled, and
before the next monitoring site is to be sampled, the sample collector should be rinsed
thoroughly with deionized water to prevent cross-contamination between sites. At least four
rinses with deionized water are necessary, followed by filling the sample collector several times
with new monitoring site runoff before finally using the collector to fill the sample bottles.
During high flow events, runoff may be unsafe to approach directly to collect the sample.
Modified sample collectors can be designed to allow remote sampling. Many stainless steel
buckets or cookware (asparagus cookers) have handles to which ropes may be tied at a length
that allows the sample collector to be lowered into the runoff and raised back up after filling with
water. These sample collectors with rope are ideal to use if sampling a creek from a bridge or
sampling an outfall from a creek bank. In addition, modified sample collectors will work well to
sample runoff in a manhole, eliminating the need to enter the confined space during higher
flows. The advantage of the rope and bucket device is that a significant length of rope can be
attached to the sample bucket to allow for sampling from great heights, yet the rope can be coiled
and stored compactly. If a sturdier sampling device is needed, sample collectors may be
attached to a pole using tape or rope and lowered into the runoff. Again, cross-contamination
between sample sites should be prevented by rinsing the sampling collector with deionized water
and new sample water several times.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
132 April 25, 2002
-------
Contamination/Blanks
Control over sample contamination is critical when attempting to measure concentrations of
compounds at the parts-per-billion level. Contamination can be introduced either during the
bottle/equipment preparation steps or during the sample collection, transport, or analysis steps.
Control over all of these steps can be achieved through the use of standardized equipment cleaning
procedures, clean sampling procedures, and clean laboratory reagents. The level of contamination
introduced during each of these steps is determined by analysis of different types of blank samples.
Each of these different types of blanks is described below:
Method Blanks are prepared by the laboratory by analysis of clean Type II reagent water.
They are used to determine the level of contamination introduced by the reagents and
laboratory processing.
Source Solution Blanks are determined by analysis of the deionized or Type II reagent
water used to prepare the other blanks. The source solution blank is used to account for
contamination introduced by the deionized water when evaluating the other blanks.
Bottle Blanks are prepared by filling a clean bottle with source solution water and
measuring the solution concentration. Bottle blanks include contamination introduced by
the source solution water and sample containers. By subtracting the source solution
blank result, the amount of contamination introduced by the sample containers can be
determined.
Travel Blanks are prepared by filling a sample container in the laboratory with Type II
reagent water and shipping the filled water along with the empty sample containers to the
site. The travel blank is shipped back with the samples and analyzed like a sample. The
bottle blank result can be subtracted from the travel blank to account for contamination
introduced during transport from the laboratory to the field and back to the laboratory.
Equipment Blanks are usually prepared in the laboratory after cleaning the sampling
equipment. These blanks can be used to account for sample contamination introduced by
the sampling equipment, if the bottle blank results are first subtracted.
Field Blanks account for all of the above sources of contamination. Field blanks are
prepared in the field after cleaning the equipment by sampling Type II reagent water with
the equipment. They include sources of contamination introduced by reagent water,
sampling equipment, containers, handling, preservation, and analysis. In general, field
blanks should be performed prior to or during the sample collection. Because the field blank
is an overall measure of all sources of contamination, it is used to determine if there are any
blank problems. If problems are encountered with the field blank, then the other
components of the sampling process should be evaluated by preparation of other blanks in
order to identify and eliminate the specific problem.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
133 April 25, 2002
-------
EPA's recent guidance on the use of clean and ultra-clean sampling procedures for the collection of
low-level metals samples (EPA 1993a,b) should be considered to ensure bottles and equipment are
cleaned properly and samples are collected with as little contamination as possible. While ultra-
clean techniques throughout are likely not necessary for stormwater runoff samples, some of the
laboratory procedures should be employed. For example, metals levels in highway runoff are
typically much greater than introduced errors associated with in-field clean sampling techniques.
These techniques are typically employed in receiving waters where their applicability is more
relevant.
Reconnaissance and Preparations
Reconnaissance and preparation is an important component of any field sampling program. Proper
reconnaissance will help field operations to go smoothly and ensure field personnel are familiar
with the sampling locations.
Site Visits
During the planning stage, a site visit should be performed by the field personnel, prior to
conducting sampling. The purpose of the site visit is to locate access points where a sample can be
taken and confirm that the sampling strategy is appropriate. Because of the transient nature of
meteorological events, it is possible sites may need to be sampled in the dark. For this reason, the
actual persons involved in the field sampling should visit the site during reconnaissance as a
complement to a training program for the monitoring effort.
The training program should include:
A discussion of what the programs goals are and why their efforts are important.
Familiarization with the site.
Training on the use and operation of the equipment.
Familiarization with field mobilization, sampling, and demobilization procedures.
Health and safety requirements.
QA/QC procedures.
Laboratory Coordination
Coordination with the laboratory is a critical step in the planning and sampling process. The
laboratory should be made aware of specific project requirements such as number of samples,
required laboratory performance objectives, approximate date and time of sampling (if known),
required QA/QC samples, reporting requirements, and if and when containers or ice chests will
be required. Laboratory personnel should be involved early in the process so they can provide
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
134 April 25, 2002
-------
feedback on methods and performance standards during the planning phase. Notifying the
laboratory that stormwater sampling is planned is also important to allow the laboratory to plan
for off-hours sample delivery and to set-up any analysis with short holding times.
Sample Containers/Preservation/ Holding Times
EPA recommends that samples be collected and stored in specific types of sample container
materials (e.g., plastic, glass, Teflon). For analysis of certain parameters, addition of specific
chemical preservatives is recommended to prolong the stability of the constituents during storage.
Federal Register 40 CFR 136.3 lists recommended sample containers, preservatives, and maximum
recommended holding times for constituents. Sample holding times should be compared to
recommended maximum holding times listed in the Federal Register. Laboratory quality control
sample data should be compared to target detection limits as well as precision and accuracy goals
and qualified according to EPA functional guidelines for data validation (EPA 1988).
If composite sampling procedures are to be used to collect one large sample that will be subsampled
into smaller containers, the composite sample bottle should be compatible with all of the
constituents to be subsampled. In general, the use of glass containers will allow subsampling for
most parameters (with the exception of fluoride).
Sample volumes necessary for the requested analysis should be confirmed with the laboratory prior
to sample collection. Extra sample volume should be collected for field and laboratory QA/QC
samples. As a general guide, if one station is to be used for both field and laboratory QA/QC
measurements, four times the normal volume of water should be collected.
Recommended Field OA/OC Procedures
Listed below are the recommended quality control samples and field procedures.
Field Blanks
Field blanks should be prepared at least once by each field sampling team to prevent or reduce
contamination introduced by the sampling process. It is recommended that field blanks be routinely
prepared and analyzed with each sampling event. In addition, it is desirable to prepare field blanks
prior to the actual sampling event as a check on procedures. This will ensure field-contaminated
samples are not analyzed. Additional field blanks should be prepared if sampling personnel,
equipment, or procedures change.
Field Duplicate Samples
Field duplicate samples should be collected at a frequency of 5% or a minimum of one per event,
whichever is greater. Field duplicate samples are used to provide a measure of the
representativeness of the sampling and analysis procedures. These types of duplicates are
recommended, but often not done due to expense.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
135 April 25, 2002
-------
Field Sample Volumes
Sufficient sample volumes need to be collected to enable the required laboratory QA/QC analysis to
be conducted. In general, one station should be targeted for extra sample volume collection and
identified on the chain-of-custody as the laboratory QA/QC station. If possible, this station should
be the one where the data quality is most critical.
Chain of Custody
All sample custody and transfer procedures should be based on EPA-recommended procedures.
These procedures emphasize careful documentation of sample collection, labeling, and transfer
procedures. Pre-formatted chain-of-custody forms should be used to document the transfer of
samples to the laboratory and the analysis to be conducted on each bottle.
Recommended Laboratory QA/QC Procedures
Method Blanks
For each batch of samples, method blanks should be run by the laboratory to determine the level of
contamination associated with laboratory reagents and glassware. Results of the method blank
analysis should be reported with the sample results.
Laboratory Duplicates
For each batch of samples, one site should be used as a laboratory duplicate. For the laboratory
duplicate analysis, one sample will be split into two portions and analyzed twice. The purpose of
the laboratory duplicate analysis is to assess the reproducibility of the analysis methods. Results of
the laboratory duplicate analysis should be reported with the sample results.
Matrix Spike and Spike Duplicates
Matrix spike and spike duplicates should be used to determine the accuracy and precision of the
analysis methods in the sample matrix. Matrix spike and spike duplicate samples are prepared by
adding a known amount of target compound to the sample. The spiked sample is analyzed to
determine the percent recovery of the target compound in the sample matrix. Results of the spike
and spike duplicate percent recovery are compared to determine the precision of the analysis.
Results of the matrix spike and spike duplicate samples should be reported with the sample results.
External Reference Standards
External reference standards are artificial standards prepared by an external agency. The
concentrations of analytes in the standards are certified within a given range of concentrations.
These are used as an external check on laboratory accuracy. One external reference standard
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
136 April 25, 2002
-------
appropriate to the sample matrix should be analyzed and reported at least quarterly by the
laboratory. If possible, one reference standard should be analyzed with each batch of samples.
3.2.7 Recommendations for Data Management
A monitoring program may generate a considerable amount of information in a wide variety of
forms. Before you begin monitoring, you should establish procedures for managing the data you
expect to generate and for presenting the results.
Data management is an important component of your overall stormwater quality program. You
need to be able to store, retrieve, and transfer the diverse hard copy and electronic information
generated by your monitoring program. Before you begin monitoring, you should establish:
A central file to accommodate the hard copy information your program is expected to
generate and practical dating and filing procedures to help ensure that superseded
information is not confused with current information.
A database to accommodate digital information such as results of laboratory analyses,
information recorded by data loggers (e.g., flow, precipitation, in-situ water quality
measurements), maps in CAD or GIS, spreadsheets, etc. It is recommended that data be
stored and reported according to the protocols described in Section 4 of this Manual.
In many cases, the laboratory can provide the analytical results in an electronic format (i.e., an
"Electronic Data Deliverable" or EDD) that you can input directly to your database. This can
save time and reduce the potential for data entry errors. You should work with the analytical
laboratory to determine if electronic data transfer makes sense for your program.
If you do not have one, you may want to consider instituting an electronic filing system to help
ensure that draft reports (including text, tables, and graphics) and unvalidated analytical data can
be easily distinguished from final reports and validated data.
After data from the field and/or laboratory have been received and the originals have been stored
in the project file, they may be routed to designated staff members who will perform one or more
of the activities. These activities include data validation, calculations and analysis, and data
presentation.
Data reports should be reviewed for completeness as soon as they are received from the laboratory.
Reports should be checked to ensure all requested analyses were performed and all required QA
data are reported for each sample batch. If problems with reporting or laboratory performance are
encountered, corrective actions (re-submittal of data sheets or sample re-analysis) should be
performed prior to final data reporting or data analysis.
3.2.7.1 Database Requirements
This section provides general guidance on storing data and is based on QA/QC procedures
developed for the ASCE/EPA National Stormwater Best Management Practices Database.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
137 April 25, 2002
-------
Databases provide a significant level of control over the types of data that are valid for a
particular field. These "rules" limit the format and structure of individual fields. For example
any field where a date is present should be entered in the mm/dd/yyyy format. In addition, drop
down boxes with lookup tables of relevant values can be used extensively in a database in order
to maintain consistency between records.
Additional fields can be included on forms in order to allow comments to be provided in each
data table. Water quality information can be entered in a tabular format where one row is used
for each sample and one column for each constituent. Macros can then be written to parse the
tabular format into a one-record-per-constituent format similar to that used in the National
Stormwater Best Management Practices Database (Database).
Analysis of Database Links
In creating a complex database, records are often linked between tables. Once all data have been
entered into a database, a check of the established links should be done between the tables
storing event data for flow, precipitation, and water quality. The start and stop date and time of
each water quality record can be checked against the date and time of the linked flow and
precipitation event. This can be conducted using a combination of database queries by
identifying dates that do not pair up. All dates that do not match should be flagged and the links
should be checked by hand. This process ensures internal consistency between the separate
tables in the database. Where any errors are encountered, the original document should be
consulted.
Analysis of Outlying Records
An analysis of the data contained in database tables can be done to identify outlying values that
resulted from typographical errors during data entry (e.g., wrong decimal place), unit errors (e.g.,
mg instead of |j,g), and incorrectly assigned STORET Codes. Two types of outlying records can
appear in the database: data entry errors (i.e., manifestations of the data extraction process) and
real outlying values (i.e., values present in a study's original documents). The efforts conducted
during outlier analysis seeks to identify and correct data entry errors. The assumption in looking
for outlying errors is that recorded water quality parameter values lie within an expected
reasonable range. Values that are outside of this range may be incorrectly entered into the
database and deserve close attention. This method is particularly useful for identifying errors in
units.
The usefulness of identification of outliers varies from constituent to constituent. For example
any mistyped entries are easily identified in pH or temperature data. If one digit is off in pH or
temperature data it is quite obvious, and, thus, there is a greater degree of confidence in the
quality of the data based on an outlier analysis of pH or temperature than for other water quality
parameters. Unfortunately, on the other end of the scale are other parameters such as Fecal
Coliform. Even an error in excess of two orders of magnitude is not readily identified in a series
of Fecal Coliform records, and thus an outlier analysis provides little or no additional
information about the quality of Fecal Coliform records.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
138 April 25, 2002
-------
Sample Comparisons Between Original Documents and Final Data Set
Finally, to better quantify the quality of the data stored in a data set, sample comparisons can be
made of the data set with the original source documents. A percentage of all records can be
checked in order to assess data quality. All errors identified in these documents should be
flagged and corrected. The sample comparisons conducted provide insight into overall quality of
the data entry process.
Digital Conversion of Data
In the event that data is provided in a digital format that is different from the designated
ASCE/EPA BMP Database format (see Section 3.4 of this Guidance), conversion of the data is
necessary. Data can be easily imported between database, spreadsheet, and word processing
software in more recent versions of most software. However, this data should be carefully
evaluated and checked for transition errors. Often, different programs will automatically round
numbers to a certain decimal and then truncate the remaining digits. Evaluation and comparison
between the original document or database and the converted data is recommended for all
records to ensure that the quality of the data is maintained.
Double Data Entry and Optical Character Recognition
Before data entry begins, both digital and hard copy data extraction/entry forms should be
created along with instructions for the data entry process. These forms should be based directly
on the database table structure. This methodology will allow the data collection and entry
process to take place in a consistent, uniform environment.
To improve the quality of data entry during any process that requires hand entry of large data
sets, it is typically necessary to implement a double entry procedure with automated flagging and
formal correction of all inconsistencies. This method should be considered as a potential part of
any data entry protocols. This is one of the few systematic methods for ensuring very small error
rates. In circumstances where significant understanding of the source of the data is required on
the part of the data entry personnel, the cost of this approach could be prohibitive.
In some cases, optical character recognition (OCR) can be used effectively to increase the speed
of data entry. In cases where OCR is used, all results should be hand checked to ensure data
quality. The data resulting from OCR typically contains a smaller number of errors compared to
hand entered data, depending on format of data.
3.3 Phase III - Implementation of Monitoring Plan
3.3.1 Training of Personnel
Each member of the monitoring team must receive whatever training is necessary to properly
perform his or her assigned roles. Generally, the first step is for each team member (including
back-up personnel) to review the monitoring plan and health and safety plan. Next, the team
members attend an initial orientation session that includes a "dry run" during which team
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
139 April 25, 2002
-------
members travel to their assigned stations and simulate monitoring, sample documentation,
packaging, etc., under the supervision of the instructor (usually the principal author of the
monitoring plan). Health and safety precautions should be reinforced during the dry run.
Periodic "refresher" orientation sessions should be conducted after long dry periods, or when the
monitoring team composition changes.
3.3.2 Installation of Equipment
If you plan to use manual monitoring techniques, equipment installation may be unnecessary. If
you plan to use automated monitoring methods, you must install the sampling and flow
measurement equipment at the monitoring locations. Equipment installation procedures vary
depending on the specific equipment and the configuration of the monitoring location. Follow
the equipment manufacturer's instructions for installation. Some general recommendations for
equipment installation are listed below:
Personnel must follow the health and safely plan when installing equipment. Some
monitoring locations may require use of protective clothing, traffic control, combustible gas
meters, and special training in confined space entry procedures.
Bubbler tubes, pressure transducers, and velocity sensors typically are mounted on the
bottom of the channel in the middle of the channel cross-section, facing upstream. Ultrasonic
depth sensors typically are mounted above the water surface.
In most cases, the automated sampler intake tube is mounted facing upstream and parallel to
the flow in order to reduce any flow distortion that could bias the sampling of suspended
solids. The intake often is covered with a strainer to prevent clogging.
Probes, sensors, and intake lines usually are anchored to the pipe or channel. The intake
tubing should be anchored throughout its length so that it will not bend, twist or crimp under
high flows.
Weirs and flumes must be secured to the bottom of the pipe or channel. If the monitoring
location is in a swale, the weir or flume cutoff walls must be buried in each bank so that the
structure extends all the way across the channel and all flow is directed through the weir or
flume.
If not installed inside a manhole vault, the flow meter and automated sampler should be
placed in a sturdy shelter to protect the equipment from vandalism and other damage.
If batteries are used as the power supply, install fresh batteries at the frequency
recommended by the manufacturer or before each anticipated storm monitoring event.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
140 April 25, 2002
-------
3.3.3 Testing and Calibrating Equipment
Water quality probes (e.g., pH, conductivity), automated samplers, and flow meters must be
periodically calibrated in order to ensure reliable operation and credible results. Typical
calibration procedures are summarized in this section; however, you should always follow the
manufacturer's instructions when calibrating a specific monitoring device.
Calibration of pH meters, conductivity meters, dissolved oxygen meters, and other water quality
instruments generally involves two steps:
1. Use the instrument to measure a known standard and determine how much the
instrument's measurement differs from the standard.
2. Adjust the instrument according to the manufacturer's instructions until it
provides an accurate measurement of the standard.
Automated sampling equipment should be calibrated after installation to ensure it pumps the
correct volume of sample. The condition of the sampler pump and intake tubing, the vertical
distance over which the sample must be lifted, and other factors can affect the volume drawn.
Therefore, you should test the sampler after installation and adjust the sampler programming if
necessary to be sure the system consistently draws the correct sample volume.
Flow meters can be affected by the hydraulic environment in which they are placed;
consequently, they should be calibrated after installation to ensure accuracy. Because sediments,
debris, and other materials carried by stormwater can damage or clog bubbler tubes and pressure
transducers used for depth measurements, they must be frequently inspected and calibrated by
checking the flow depth with a yard stick or staff gauge. Ultrasonic velocity sensors and other
instruments that measure flow rate must also be inspected and checked against velocity
measurements made using a current meter.
3.3.4 Conducting Monitoring
After you have completed the advance preparations described above, you are ready to begin
monitoring.
The general steps for automated monitoring are:
1. Perform routine inspection and maintenance to help ensure that the equipment will
function properly when a storm event occurs.
2. Keep track of precipitation. After each storm, check the local rainfall records (or
preferably a rain gauge at or near the center of the basin) to see if the amount of
precipitation and the antecedent dry period met your pre-determined criteria.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
141 April 25, 2002
-------
If the storm did not meet your criteria, remove the sample bottles from the sampler
and replace them with clean bottles. Empty the sample bottles and arrange for them
to be cleaned.
If the storm criteria were met, remove the sample bottles. Check them to be sure they
received the proper amount of sample. Check the sampling times against the storm
duration to see how much of the storm was sampled. If this meets your criterion,
complete the sample labels, chain-of-custody form and other field documentation,
then deliver the samples to the laboratory for analysis.
3. If the sampler overfilled or underfilled the sample bottles, refine the sampler
programming.
4. Reset the sampler and inspect all of its systems for possible damage or clogging so that it
will be ready to sample the next storm.
The general steps for manual monitoring are:
1. The monitoring team leader or another designated person tracks the weather forecasts.
2. When the weather forecasts indicate that a potentially acceptable storm is approaching,
the monitoring team leader contacts the monitoring team and the analytical laboratory. If
any of the primary team members are unavailable, the monitoring team leader arranges
for back-ups. The team members check their instructions, communications protocols,
monitoring equipment, and supplies to ensure they are ready.
3. The monitoring team leader contacts NOAA (or some other meteorological service, if
better information is available) to get updated forecasts as the storm approaches. When
the forecasts indicate that the storm is likely to start within the next few hours, and it still
appears likely to meet the storm selection criteria, the team leader directs the team
members to proceed to their assigned monitoring stations so that they arrive before the
predicted start time. The team leader also alerts the lab that samples are likely to be
delivered soon.
4. The team members travel to their assigned locations and start collecting samples and
taking flow measurements as soon as possible after stormwater runoff begins. They fill
out the sample labels, chain-of-custody forms, and other field documentation.
5. During monitoring, the team members may contact the team leader (usually by cellular
phone) to ask questions, notify him or her of changing conditions, receive direction, etc.
6. After samples have been collected, they are shipped or delivered to the analytical
laboratory.
7. If the lab is to prepare flow-weighted composite samples, the monitoring team members
must use the flow data they collected to determine the amount of each sample to be used
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
142 April 25, 2002
-------
to form the composite. Usually, the team will calculate the amounts using a spreadsheet
and fax the completed spreadsheet to the lab.
If you are using manual methods, you will need to maintain a vigilant "weather watch." This is
essential if you wish to monitor the initial runoff from a storm event. You need some advance
notice of an impending sampling event in order to have enough time to contact the monitoring
team, arrange for back-ups if the primary members are unavailable, notify the analytical
laboratory, work out communications protocols, pick up ice, and travel to the monitoring
locations. Also, if your are able to obtain reasonably accurate estimates of storm start times, you
can reduce the amount of stand-by time for your monitoring team. Finally, a close weather
watch can help reduce the risk of a "false start" which can occur when a predicted storm fails to
materialize or turns out to be a brief shower.
3.3.5 Coordinate Laboratory Analysis
Most stormwater monitoring programs involve laboratory analysis. Exceptions include (1) field
screening programs that rely solely on visual observations and field test kits, and (2) programs
that rely on "in-situ" monitoring of indicator parameters (e.g., pH, dissolved oxygen, turbidity)
using probes and data loggers.
It is a good idea to involve laboratory personnel in identifying the analytical methods
establishing communications protocols and QA/QC protocols. Typically, the laboratory will
provide the pre-cleaned sample bottles and distilled/deionized water used for monitoring.
Your mobilization protocols should include notifying the laboratory when a storm monitoring
event appears imminent. They should also include contacting the laboratory shortly after the
monitoring event to ensure that the samples were received in good condition and to answer any
questions the lab may have regarding the analyses to be conducted. Also, it is a good idea to
periodically contact the laboratory while the analyses are being conducted. Frequent
communication with the laboratory helps reduce the risk of incorrect analysis and other potential
unpleasant "surprises."
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
143 April 25, 2002
-------
3.4 Phase IV - Evaluation and Reporting of Results
3.4.1 Validate Data
You should evaluate the quality or adequacy of the laboratory analytical results before you
interpret the results. This evaluation is known as "data validation" or data quality review. The
basic steps are listed below.
1. Check that all requested analyses were performed and reported. Check that all requested
QA/QC samples were analyzed and reported.
2. Check sample holding times to ensure that all samples were extracted and analyzed within the
allowed sample holding times.
3. Check that the laboratory's performance objectives for accuracy and precision were achieved.
This includes a check of method blanks, detection limits, laboratory duplicates, matrix spikes and
matrix spike duplicates, laboratory control samples, and standard reference materials.
4. Check that field QA/QC was acceptable. This includes a check of equipment blanks, field
duplicates, and chain-of-custody procedures.
5. Check that surrogate recoveries were within laboratory control limits.
6. Assign data qualifiers as needed to alert potential users of any uncertainties that should be
considered during data interpretation.
If the laboratory and field performance objectives were achieved, further data validation is not
generally needed. Specifics of the instrument calibration, mass spectral information, and run
logs are not usually recommended for review unless there is a suspected problem or the data are
deemed critical. If performance objectives were not achieved (e.g., due to contaminated blanks,
matrix interference, or other specific problems in laboratory performance), the resulting data
should be qualified. EPA functional guidelines for data validation (EPA 1994a,b) should be used
as a guide for qualifying data.
3.4.2 Evaluate Results
After the chemical data have been validated, you should perform a preliminary data evaluation.
The main purpose of the preliminary evaluation is to determine whether you have obtained
enough information of sufficient quality to meet BMP assessment goals. If the answer is no, you
should continue monitoring until you have collected sufficient information. If the answer is yes,
you should proceed with the definitive evaluations that are best suited to your specific objectives.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
144 April 25, 2002
-------
3.4.2.1 Preliminary Data Evaluation
After the analytical results have been validated, consider graphing the flow and rainfall data vs.
time for each storm event in order to produce a storm hydrograph (flow rate versus storm
duration). It is often helpful to plot rainfall volume versus storm duration on the same graph. In
addition, you should denote the times when the grab or composite samples were collected. This
information can be very helpful in interpreting the chemical results.
Generally, stormwater quality variability is so high that statistical evaluation is not worthwhile
until you have monitored several events (at least four). You should conduct an initial statistical
analysis using the validated chemical data. This analysis will provide summary statistics that
indicate how well your sample results represent stormwater quality at a given site. Summary
statistics include sample mean, variance, standard deviation, coefficient of variation, coefficient
of skewness, median, and kurtosis. Stormwater quality typically exhibits a lognormal
distribution (EPA 1983; WCC 1989). Therefore, you should calculate these descriptive statistics
based on an assumed lognormal distribution. Non-detects should be included in calculating the
initial statistics using a maximum likelihood estimator approach.
The initial statistical analysis can help you determine whether it will be useful to statistically test
various hypotheses regarding the existing data set. For example, if the standard deviations are
several times larger than the means (i.e., the coefficient of variation is 3 or more), hypothesis
testing may not be worthwhile. You may need to conduct additional monitoring to compensate
for the observed variability and allow statistically significant differences to be discerned.
3.4.2.2 Definitive Evaluations
If your initial statistical analysis indicates that your samples are representative of water quality at
the site(s) in question, you should conduct additional statistical analyses (or perhaps modeling)
as needed to answer the key questions about your stormwater catchment area.
Consider the initial statistics when selecting the statistical procedure(s) you will use to answer
the key questions about your stormwater catchment area. For example, if the data set does not
appear to follow a normal or lognormal distribution, or if the data set contains a high proportion
(i.e., >15%) of non-detects, non-parametric tests may be more appropriate than parametric tests.
The results of your monitoring program may also serve as input to a water quality model.
Loadings can be calculated using SUNOM (previously the simple model, Schueler 1987), or one
of several dynamic models. The simple model estimates the mean pollutant loading from a
particular outfall or subbasin to a receiving water. A dynamic model takes into account the
variability inherent in stormwater discharge data including variations in concentration, flow rate,
and runoff volume. A dynamic model can therefore be used to calculate the entire frequency
distribution for the concentration of a pollutant and the theoretical frequency distribution (i.e.,
the probability distribution) for loadings from the outfall or subbasin. Thus, the modeler can
describe the effects of observed discharges on receiving water quality in terms of the frequency
by which water quality standards are likely to be exceeded. Dynamic models include EPA's
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
145 April 25, 2002
-------
Stormwater Management Model (SWMM) and Hydrologic Simulation Program Fortran (HSPF),
the U.S. Army Corps of Engineers' Storage, Treatment, Overflow, Runoff Model (STORM), and
Illinois State Water Survey's Model QILLUDAS (or Auto-QI) (EPA 1992).
3.4.3 Report Results
The results of your monitoring program should be presented in one or more reports. The
appropriate report frequency and content depends on your monitoring program objectives and
your audience. If you are monitoring to comply with a permit, the permit will generally specify
the minimum frequency and content of the reports.
Most monitoring programs involve two types of reports: status (or progress) reports and final
reports. To determine the appropriate frequency of status reports, consider your monitoring
frequency and objectives, particularly any permit requirements. Many programs produce status
reports on a quarterly or semi-annual basis. A typical status report may contain the following
information:
Summary of work accomplished during the reporting period
Summary of findings
Summaries of contacts with representatives of the local community, public interest
groups, or state federal agencies
Changes in key project personnel
Projected work for the next reporting period
You should prepare more comprehensive reports at the end of the monitoring program (for short-
term programs) or at the end of each year (for multi-year programs). Consider including the
above-listed information and the following information in your annual or final report:
Executive summary
Monitoring program background and objectives
Monitoring station descriptions, analytical parameters, analytical methods, and method
reporting limits
Summary descriptions of the conditions and stations, equipment inspections and
calibrations, etc.
Sample collection, precipitation, and flow measurement methods
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
146 April 25, 2002
-------
Flow, precipitation, and water quality results and data validation information
Qualitative and statistical data evaluations/hypothesis testing as required for your specific
program objectives (see Section 3.4.2 and Appendix I)
Summary and conclusions, including any caveats or qualifying statements that will help
the reader understand and use the reported information in the appropriate context
Recommendations regarding management actions (e.g., changes in monitoring program,
implementation of BMPs)
3.4.3.1 National Stormwater BMP Database Requirements
This section is designed to provide guidance for consistent reporting of results collected from
BMP monitoring studies. The protocols described are based on those specified in the National
Stormwater Best Management Practices Database, which has been developed by the Urban
Water Resources Research Council of ASCE under grant funding from EPA to serve as a tool for
data organization and reliable comparison of BMPs. Minimum requirements for acceptance in
the National Database are outlined in this section, and standard format examples that can be used
as templates for reporting results of Stormwater monitoring studies are provided.
The National Stormwater BMP Database was developed to provide a scientifically sound tool for
the determination of the effectiveness of BMPs under various conditions for a range of design
parameters. The data fields included in this database have undergone intensive review by many
experts and encompass a broad range of parameters including test site location, watershed
characteristics, climatic data, BMP design and layout, monitoring instrumentation, and
monitoring data for precipitation, flow and water quality. In order to effectively compare the
performance of different BMPs under a variety of conditions, a set of "required" database fields
were identified. These "required" fields are considered the minimum requisites for acceptance
into the National Stormwater BMP Database. The database requirements vary with the different
types of BMPs, and special requirements exist for unique hydraulic conditions. Database
requirement categories and fields are as follows:
1) Information required for all BMPs (Table 3.5)
General Test Site Information Precipitation Data
Watershed Information Flow Data
Monitoring Station Information Water Quality Data
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
147 April 25, 2002
-------
Table 3.5: National Stormwater BMP Database requirements for all BMPs
Data Element Description
General Test Site Information
BMP Test Site Name
Dity
State
l\p Code
ountry
Mtitude
Name that site is known by locally.
City closest to test site.
State where test was performed.
Zip code of the test site.
Country where the test site is located.
Altitude to nearest 100 ft or 30 m.
sponsoring and Monitoring Agencies for Test Site
\ddress
Includes monitoring and sponsoring agency name and contact
information.
Watershed Information
Subject Watershed Name
otaI Watershed Area
ฐercent (%) Impervious Area
Regional Climate Station (US)
.and Use Information
Name that watershed is referred to locally.
Topographically defined area drained by system.
Total percent of impervious surface in watershed.
Regional climate station in US that is most relevant to test site.
Description of land uses (only required for non-structural BMPs).
Monitoring Stations
Station
dentify Upstream BMP
dentify Relationship to
Jpstream BMP
dentify Downstream BMP
dentify Relationship to
Downstream BMP
User-defined name for subject monitoring station.
BMP upstream of the monitoring point (if any).
Identify the relationship of the monitoring station to the upstream
BMP (i.e. inflow, outflow or not applicable).
BMP downstream of the monitoring point (if any).
Identify the relationship of the monitoring station to the
downstream BMP (i.e. inflow, outflow or not applicable).
Monitoring Instrumentation
Monitoring Station Name
Select monitoring station where the instrument is located.
'recipitation Data
vlonitoring Station Name
Identify monitoring station where precipitation event was
monitored.
storm Runoff and Base Flow Data
Select monitoring station where flow event was monitored.
Base flow or stormwater runoff.
Month, day and 4-digit year (e.g. 01/01/1998).
Total runoff volume minus runoff volume influent to BMP.
vlonitoring Station Name
'ype of Flow
low Start Date
'otal Bypass Volume (if any)
'otal Storm Flow Volume into orTotal runoff volume minus the bypass volume.
rom BMP
3ry Weather Base Flow Rate Flow rate during dry-weather conditions.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
148
April 25, 2002
-------
Water Quality Sampling Event
Monitoring Station Name
Related Flow-Event
Date Water Quality Sample
ollected
What Medium Does the Instrument
Monitor
Water Quality Parameters
Value
Unit
Qualifier
Select monitoring station where samples were collected.
Select flow data corresponding to water quality data.
Month, day and 4-digit year the water quality sample was
collected.
e.g. Groundwater, surface runoff.
STORE! water quality parameters analyzed.
Value of measured constituent.
Units of measured constituent.
Select STORE! qualifier code.
2) Data required for structural BMPs (Table 3.6)
Table 3.6: National Stormwater BMP Database requirements for structural BMPs
Data Element Description
Structural BMP Information
Structural BMP Name
Structural BMP Type
Date Facility Was Put Into Service
Number of Separate Inflows
Describe the Type and Design of
Each BMP Outlet
s the BMP Designed to Bypass
When Full?
3MP Drawing
Common name by which BMP is referred to locally.
Select the type of BMP being monitored at the site (drop-down
list).
Month, day and 4-digit year facility became operational.
Number of inflows into the facility.
Description of the outlet configuration (i.e. Perforated riser).
Select "Overflow" or "bypass" characteristics of BMP.
Plan view and profile of BMP (in bitmap format for database).
3) Information required for non-structural BMPs (Table 3.7)
Table 3.7: National Stormwater BMP Database requirements for non-structural BMPs
Data Element
Description
Von-structural BMP Information
Non-structural BMP Type
Non-structural BMP Name
Type of non-structural BMP (e.g. educational, maintenance
practices, etc.).
The name by which the non-structural BMP is referred to
locally.
Date Test Began Month, day and 4-digit year.
Describe the Quantity or Measure of Measure of the educational, maintenance, recycling or source
he BMP Being Practiced control BMP.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
149
-------
4) Individual staictural BMP requirements (Table 3.8) for:
Detention Basins.
Grass Filter Strips.
Infiltration Basins.
Media Filters.
Porous Pavement.
Retention Ponds.
Percolation Trenches and Dry
Wells.
Wetland Channels and Swales.
Wetland Basins.
Hydrodynamic Devices.
Table 3.8: National Stormwater BMP Database requirements for individual structural BMPs
Data Element
Detention Basin Design Data
Water Quality Detention Volume
Water Quality Detention Area (when
full)
Water Quality Detention Basin
Length
Detention Basin Bottom Area
Brim-full Volume Emptying Time
Half Brim-full Volume Emptying
fime
Bottom Stage Volume (if any)
Bottom Stage Surface Area
Is there a Micro Pool?
Forebay Volume
Forebay Surface Area
Describe Vegetation Cover Within
Basin
Flood Control Volume (if any)
Design Flood Return Periods
Description
The volume of runoff that is captured and released overtime.
The area of water surface in basin at full water quality detention
volume.
Distance between inflow and outflow (average for multiple inflows).
Area of the bottom of the detention basin, including bottom stage
area.
Emptying time of water quality detention volume.
Emptying time of lower half of water quality detention volume.
The volume of the lower "bottom stage" of the detention basin.
The surface area of the lower "bottom stage" of the detention basin.
"Yes" or "No" indication of micropool.
Volume of the forebay portion of the detention basin.
Surface area of the forebay portion of the detention basin.
List and description of types of vegetation on the basin sides and
bottom.
Volume in excess of water quality detention volume.
Design return period if basin is designed for flood control.
Grass Filter Strip Design Data
Grass Strip Length
Grass Strip Slope
Flow Depth During 2- Year Storm
2- Year Peak Flow Velocity
Describe Grass Species and
Densities
Is Strip Irrigated?
Length of strip in the direction of flow.
Slope of the strip along the flow path.
Design depth of flow during the 2-year peak flow.
Design flow velocity during the 2-year peak flow.
List of grass species and their densities.
"Yes" or "no" indication of irrigation.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
150
-------
nfiltration Basin Design Data
Capture Volume of Basin
Surface Area of Capture Volume
When Full)
nfiltrating Surface Area
Depth to Seasonal High
Sroundwater Table
Depth to Impermeable Layer (if any)
.ist of Plant Species
Describe Granular Material on
nfiltrating Surface (if any)
The design runoff capture volume of the basin.
The area of the water surface in the infiltration basin, when full.
The plan area of the surface used to infiltrate the water quality
volume.
Depth from basin bottom to seasonal high groundwater table.
Depth from basin bottom to impermeable layer, if is present.
List of plant species and densities on infiltrating surface.
Description of granular material depth and porosity.
Media Filter Design Data
Dermanent Pool Volume, Upstream Volume of the permanent pool, if pool is part of filter basin.
)f Filter Media (if any)
Area of water surface of permanent pool.
Distance between inflow and outflow (average for multiple inflows).
The design water quality detention volume, including the volume
above the filter.
The surface area of the design water quality capture volume.
3ermanent Pool Surface Area
ฐermanent Pool Length
Surcharge Detention Volume
Surcharge Detention Volume
Surface Area
Surcharge Detention Volume's
Design Drain Time
Surcharge Detention Volume
Design Depth
vledia Filter Surface Area
\ngle of Sloping or Vertical Filter
slumber of Media Filter Layers
Describe Depth and Type of Each
:ilter Media Layer
The drain time (in hours) of the water quality capture volume.
Depth of water quality capture volume.
Surface area of the media filter.
Inclination of filter in degrees above the horizontal plane.
Number of layers of different filter material in BMP.
Description of the type and depth of media used in the filter.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
151
-------
'orous Pavement Design Data
3orous Pavement Surface Area
Depth to Seasonal High
3 round water Table
Depth to Impermeable Layer (if any)
nfiltration Rate
'ype of Granular or Other Material
Jsed Below Pavement
ฐorosity of Granular Material (%)
'otal Storage Volume Above
ฐavement (if any)
Estimated Drain Time of the
Storage Volume Above the
ฐavement (if any)
'otal Storage Volume Under
3avement (if any)
Estimated Drain Time of Storage
Volume Under Pavement
Does Porous Pavement Have
Jnderd rains?
Surface area of porous pavement.
Depth from pavement surface to seasonal high groundwater table.
Depth from pavement surface to impermeable layer, if present.
Rate of infiltration for site soils under saturated conditions.
Description of the type and depth of each granular material layer
under the porous pavement.
The volumetric portion of the filter material that is not occupied by
solid matter, expressed as a percent of the total filter volume.
The volume of water stored in depressions or as a result of
attenuation (if any) above the porous pavement surface.
Drain time of holding areas above pavement, if any.
Net available volume of pore spaces in the granular materials
beneath the porous pavement.
Total emptying time for water stored in granular materials.
"Yes" or "no" indication of presence of underdrains.
detention Pond Design Data
i/olume of Permanent Pool
^ermanent Pool Surface Area
^ermanent Pool Length
.ittoral Zone Surface Area
A/ater Quality Surcharge Detention
i/olume (when full)
A/ater Quality Surcharge Area
when full)
A/ater Quality Surcharge Basin
.ength
Brim-full Emptying Time for
Surcharge
Half Brim-full Emptying Time for
Surcharge
rorebay Volume
rorebay Surface Area
Describe Vegetation Cover Within
Basin
rlood Control Volume (if any)
.ist Design Flood Return Period (in
ears)
Volume of permanent pool in structure.
Area of water surface of permanent pool.
Length of the permanent pool measured along the axis between the
inflow and outflow. For more than one inflow, take an average.
The surface area of the bank above the permanent pool that is
periodically covered with water during a storm event.
Water quality detention volume above permanent pool.
The surface area (plan view) of the water quality surcharge
detention volume.
Length of the water quality surcharge pool measured along the axis
between the inflow and outflow. For more than one inflow, take an
average.
Emptying time of water quality detention volume down to the
permanent pool.
Emptying time of lower half of surcharge detention volume down to
the permanent pool.
Volume of the forebay portion of the detention basin.
Surface area of the forebay portion of the detention basin.
List and description of vegetation on basin sides and floor.
Volume in excess of the retention basin water quality surcharge
detention volume.
Design return periods if pond was designed for flood control.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
152 April 25,
:oo2
-------
'ercolation Trench and Dry Well Design
ฐercolation Trench/Well Surface
\rea
ฐercolation Trench/Well Length
ฐercolation Trench/Well Depth
Depth to Seasonal High
Sroundwater Table
The surface area of the top of the percolation trench/well.
Length of percolation trench or diameter of the well.
The depth of trench or well that is exposed to permeable soils.
Depth below the bottom of the trench or well to the seasonal high
groundwater table.
Depth to Impermeable Layer (if any) Depth below the bottom of the trench or well to impermeable layer, if
impermeable layer is present.
Description of the stratification and the depth of each layer of soils at
the BMP site.
Description of the type and depth of granular material used in the
trench or well.
"Yes" or "no" indication of geotextile use above granular fill.
Depth and Type of Each Soil Layer
\djacent to and Below Trench/Well
'ype of Gradation of Granular
Materials Used in Trench/Well
A/as Geotextile Used Above
3ranular Trench Fill?
A/as Geotextile Used on the Side of "Yes" or "no" indication of geotextile use on sides of granular fill.
3ranular Fill?
A/as Geotextile Used on the Bottom "Yes" or "no" indication of geotextile use below granular fill.
)f Granular Fill?
The volumetric portion of the granular material that is not occupied
by solid matter, expressed as a percent of the total volume.
Volume of available pore space in the trench or well.
3ive Porosity (%) of the Granular
'otal Storage Pore Volume in
'rench
Describe Type of Geotextile Used
Hydraulic Conductivity of Adjacent
Boil
jroundwater Flow Gradient
Description of types and locations of geotextile fabrics.
Hydraulic conductivity of the soils adjacent to the trench or well.
Slope of the local groundwater table without influence from the BMP.
Wetland Channel and Swale Design Data
.ength of Channel/Swale
.ongitudinal Slope of
Dhannel/Swale
Bottom Width of Channel
Swale
Bide Slope of Channel Swale
2-Year Flow Design Depth in
Dhannel/Swale
2-Year Peak Design Flow Velocity
'ype of Plant Species in Wetland
lone or Swale
Length of channel or swale from stormwater inflow to outflow point.
Measured slope between grade control structures in swale.
Average width between side slopes.
Average slope of swale sides.
Average depth of water in channel/swale during 2-yrflow.
Design velocity for 2-yr flow.
List and description of plant species, percent of cover and densities.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
153 April 25,
:oo2
-------
Wetland Basin Design Data
i/olume of Permanent Pool
3ermanent Pool Surface Area
ฐermanent Pool Length
A/ater Quality Surcharge Detention
i/olume (when full)
A/ater Quality Surcharge Area
when full)
A/ater Quality Surcharge Basin
.ength
Brim-full Emptying Time for
Surcharge
Half Brim-full Emptying Time for
Surcharge
rorebay Volume
rorebay Surface Area
Describe Vegetation Cover Within
Basin
rlood Control Volume (if any)
.1st Design Flood Return Period (in
ears)
A/etland Surface Area
Dercent of Wetland Pond with 12
nches Depth
3ercent of Wetland Pond with 12-24"
Depth
ฐercent of Wetland Pond with 24-
18" Depth
ฐercent of Wetland Pond with >48"
Depth
ฐercent of Wetland Basin's Area
hat is Meadow Wetland
.1st All Known Plant Species in the
A/etland
Volume of permanent pool in structure.
Surface area of permanent pool.
Length of the permanent pool of water, measured at the water
surface along the axis of the inflow and outflow (average for multiple
inflows).
Water quality detention volume above permanent pool.
The surface area of the water quality surcharge detention volume.
Water quality surcharge basin length, measured at the water surface
along the axis of the inflow and outflow (average for multiple
inflows).
Emptying time of water quality detention volume down to the
permanent pool.
Emptying time of lower half of surcharge detention volume down to
the permanent pool.
Volume of the forebay portion of the detention basin, when full.
Water surface area of the forebay portion of the detention basin.
Description of types of vegetative cover within the basin.
Volume in excess of the water quality detention volume.
Design return period if basin is designed for flood control.
The surface (plan view) area of the total wetland.
Percent of wetland surface area with less than 12 inches of standing
water.
Percent of wetland surface area with 12-24 inches of standing water.
Percent of wetland surface area with 24-48 inches of standing water.
Percent of wetland surface area with greater than 48 inches of
standing water.
Percent of wetland surface area with meadow wetlands (no standing
water).
List of plant species, percent of cover and densities.
iydrodynamic Devices
i/olume of Permanent Pool
3ermanent Pool Surface Area
^ermanent Pool Length
A/ater Quality Surcharge Detention
i/olume (when Full)
nlet Chamber Volume (if any)
Brim Full Emptying Time for
Surcharge
Half Brim Full Emptying Time for
Surcharge
Volume of permanent pool in structure.
Surface area of the permanent pool.
Distance between inflow and outflow (average for multiple inflows).
Water quality detention volume above permanent pool.
Volume of the inlet chamber portion of the hydrodynamic device.
Emptying time of water quality detention volume down to the
permanent pool.
Emptying time of lower half of surcharge detention volume down to
the permanent pool.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
154
-------
5) Requirements for non-structural and structural BMPs that are based on minimizing directly
connected impervious areas (Table 3.9).
Table 3.9: National Stormwater BMP Database requirements for non-structural BMPs and
structural BMPs that are based on minimizing directly connected impervious areas
Data Element Description
Watershed Information
'otal Length of Grass-Lined Total length of natural and grass-lined channels in watershed.
Channel
'otal Watershed Area Disturbed Total watershed area that is actively disturbed or under construction.
ฐercent Irrigated Lawn and/or Percent of lawn or agricultural areas that are irrigated.
\griculture in Watershed
ฐercent of Watershed Served by The percent of watershed served by storm sewers.
Storm Sewers
\verage Runoff Coefficient Based on area-weighted average.
Boil Type NRCS soil type.
'ype of Vegetation Type of vegetation predominant in pervious area.
loads and Parking Lots
'otal Paved Roadway Area Total area of paved roads, streets and alleys in watershed..
'otal Length of Curb/Gutter on Total length of curb/gutter on paved roads.
ฐaved Roads
'otal Unpaved Roadway Area Total unpaved roadway area.
'otal Length of Curb/Gutter on Total length of curb/gutter on unpaved roads.
Jnpaved Roads
ฐercent of Paved Roads Draining to Percent of paved roads draining to swales/ditches.
Srass Swales/Ditches
^ercent of Unpaved Roads Draining Percent of unpaved roads draining to swales/ditches.
o Grass Swales/Ditches
'ype of Pavement on Roads, Description of type of pavement (i.e. concrete, asphalt, etc.).
Streets and Alleys
'otal Paved Parking Lot Area Total area of paved parking lots in the watershed.
'otal Length Curb/Gutter on Paved Total length curb/gutter on paved lots.
.ots
'otal Unpaved Parking Lot Area Total unpaved parking lot area.
'otal Length Curb/Gutter on Total length of curb/gutter on unpaved lots.
Jnpaved Lots
ฐercent Paved Lot Area Draining to Percent of paved lot area draining to swales.
3rass Swales
ฐercent Unpaved Lot Area Draining Percent of unpaved lot area draining to swales.
o Grass Swales
ype of Pavement in Parking Lots Type of pavement in parking lots.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
155
-------
6) Requirements for structural BMPs that are based on minimizing directly connected
impervious areas (Table 3.10)
Table 3.10: National Stormwater BMP Database requirements for structural BMPs that are
based on minimizing directly connected impervious areas
Data Element | Description
Watershed Information
Storm Sewer Design Return Period Most common design return period for the storm sewers in the
watershed.
Average Watershed Slope Average unit less slope of the watershed (i.e. ft/ft, in/in).
NRCS Hydrologic Soil Group Dominant NRCS hydrologic soil group.
3.4.3.2 Standard Format Examples
The purpose of this section is to provide standard format examples that can serve as a guidance
tool for developing monitoring plans and promoting consistent reporting and documentation of
Stormwater monitoring studies. These forms include, but are not limited to, the required data
entry fields for the National Stormwater BMP Database. The database requirements were used
as a guideline for development and organization of forms because of its ability to aid in
consistently evaluating BMP effectiveness under different conditions. The following sections
provide standardized document formats that can be used as a template when performing a BMP
monitoring study. Each form is categorized based on the sub-sections presented in the National
Stormwater BMP Database.
General Test Site Information
The general test site information form provides data to aid in the identification of the testing
location. Location information is important because it enables identification of the general
climatic conditions under which a BMP was evaluated. Data reported on this form also provides
a cross-link with other national EPA databases. The general test site information form includes
data about the sponsoring and monitoring agencies conducting the study and georeferencing
information for exact identification of the site location. A detailed description of the data
element fields for the general test site information form is available in Table 3.11. The General
Test Site Information form, Form A, follows:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
156
-------
Table 3.11: General test site form data element descriptions
Data Element
Description
BMP Test Site Name1
City1
ounty
State1
Zip Code1
Country1
Time Zone
Name that the site is known by locally (e.g., Shop Creek, First Bank).The site
may contain more than one BMP, but ONLY if the watersheds tributary to
these BMPs are virtually identical.
City closest to the test site. The site does not have to be within the city limits.
County in which test site is located.
State where test was performed (2 characters).
Zip code of the test site.
Country where the test site is located (2 characters).
Time zone in which the BMP test site is located off-set in hours from
Greenwich Mean Time. For example, in the United States, Eastern Time is -5,
Central Time is -6, Mountain Time is -7 and Pacific Time is -8.
Georeferencing Information
USGS Quadrangle Map
Name
Principal Meridian
Range
Township
Section
Quarter-Quarter-Quarter
section
Latitude
Longitude
Altitude1
U.S. Geological Survey (USGS) 7.5-minute map on which the site can be
located. This information should be provided for U.S. sites only.
Local or international meridian from which the degrees of longitude locating
the BMP test site are measured.
Range identifies the site distance and direction (east or west) from the
selected principal meridian. For example, Range 60 West (R60W). This
information can be found on a U.S. Geological Survey quadrangle map (U.S.
sites only).
Townships are located by their distance and direction (north or south) from a
selected baseline. For example, Township 2 North (T2N) (U.S. sites only).
Section is a land area usually containing one square mile (640 acres) that can
be identified on a U.S. Geological Survey quadrangle map. There are 36
sections in a given township and range numbered from 1 to 36 (U.S. sites
only).
Quarter-Quarter-Quarter section should be provided to locate the BMP test
site on a U.S. Geological Survey quadrangle map. U.S. sites only.
Latitude is the North-South coordinate that locates the project to the nearest
second on the globe relative to the equator. The degree, minute and second
measures of the latitude can be obtained from a U.S. Geological Survey
Quadrangle Map.
The East-West coordinate that locates the project to the nearest second on
the globe relative to the selected principal meridian. The degree, minute and
second measures of the latitude can be obtained from a U.S. Geological
Survey (USGS) Quadrangle.
Elevation above mean sea level provided to the nearest 100 feet from a U.S.
Geological Survey quadrangle map or to the nearest 30 meters for studies
outside of the United States.
Sponsoring and Monitoring Agency Information
Agency Type
Address1
Agency type, such as city, county, state, industry, federal, special district,
council of governments, authority, consultant, or other.
Address information including agency name, department (if any), street or
post office address, city, state, zip code, country, phone, fax and e-mail.
1 - National Stormwater BMP Database requirement for all BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater Bh'IP Database Requirements
157
-------
Form A
1
Test Site Name
City
Zip Code
Geo referencing
Township
USGS Quadrangle Map
GENERAL TEST SITE INFORMATION
County
Country
Range
State
Time Zone
Principal Meridian
Altitude Section
Quarter Sections: Quarter
Latitude: Degrees
Longitude: Degrees
Sponsoring Agency
Sponsor's Name
Quarter-Quarter
Minute
Minute
Quarter-Qusrter-Quarter
Seconds
Seconds
Sponsoring Agency's Description
Address
Zip Code
Phone
Fax
Cfty
State
Country
E-Mail
Monitoring Agency
Monitoring Agency Name
Monitoring Agency Description
Address
Zip Code
Phone
Fax
City
State
Country
E-Mail
Comments
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
158
April 25, 2002
-------
Watershed Information
The watershed form contains important information about the physical and relational
characteristics of the watershed where the BMP was monitored. Watershed characteristics play a
significant role in the quantity and type of pollutants in stormwater runoff. The form includes
information on the physical characteristics of the watershed, parking lots and roads, streams and
land uses. This information plays a significant role in comparing BMP performance under
various watershed conditions. If multiple watersheds were examined at a single test site then
additional watershed information forms can be completed for each watershed. Table 3.12
provides descriptions of the watershed form data elements, and the watershed form is presented
as Form B.
Table 3.12: Watershed form data elements description
Data Elements Description
Subject Watershed Name Name by which the watershed is referred to locally.
Hydrologic Unit Code The U.S. Geological Survey (USGS) 8-digit hydrologic unit code (HUC) which
represents a geographic area containing part or all of a surface drainage basin
or distinct hydrologic feature.
EPA Reach Code EPA-designated RF1 or RF3 river reach with which the station is associated.
Sites will either have an RF1 code or an RF3 code, but not both.
Jnit System (S.I. or U.S. The unit system used for measurement for the study. The unit system should
standard) be consistent for all reported data.
'hysical Characteristics
'otal Watershed Area Topographically defined area drained by an urban system, channel, gulch,
stream, etc., such that all outflow is directed to a single point.
'otal Length of Watershed Length of the watershed along the main drainage path to the furthest point on
the watershed divide.
'otal Length of Grass- Total length of grass-lined and natural channels in the watershed. This is the
.ined Channel5 portion of the storm drainage network in the watershed that is not conveyed in
concrete channels, storm sewers or pipes.
'otal Watershed Area Total watershed area that is actively disturbed or under construction. This
Disturbed 5 parameter may be useful in indicating the types and levels of pollutant loads in
stormwater.
3ercent (%) Irrigated Lawn Percent of watershed area that is irrigated.
and/or Agriculture in
A/atershed 5
3ercent (%) Total The percent of the total watershed that is impervious can be determined as
mpervious Area in the total impervious area divided by the total area of the watershed. Common
A/atershed1 impervious surfaces include, but are not limited to, rooftops, walkways, patios,
driveways, parking lots, storage areas, concrete or asphalt paving, gravel
roads, packed earthen materials, and macadam or other surfaces that
similarly impede the natural infiltration of urban runoff. Rainfall on impervious
areas can cause rapid overland flow to drainage inlets.
3ercent (%) of Total Parameter calculated by dividing the hydraulically connected impervious area
mpervious Area (above) by the total impervious area. An example of hydraulically connected
hat is Hydraulically impervious area includes building rooftops that drain onto paved areas.
Connected
ฐercent (%) of Watershed The percentage of watershed area served by storm sewers (typically higher in
Served by Storm Sewers5 urbanized areas than in rural areas).
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
159
-------
Data Elements Description
Storm Sewer Design Most common design storm return period for the storm sewers in the
Return Period (yrs)6 watershed provided in years. For example, most storm sewers in the
watershed may be designed to handle flows generated by the 25-year storm.
Average Watershed Average unitless slope of the watershed (i.e., ft fall/ft run or m fall/m run-
Slope unitless). Slope for each linear reach can be determined as the elevation
difference for the reach divided by the length of the reach, and the average
slope for the watershed can be calculated as a weighted sum of the slopes of
individual reaches.
Average Runoff Coefficient Rational Method runoff coefficient. If data permits, calculate the average of
individual storm runoff coefficients using each storm's runoff volume divided
by its rainfall volume. Otherwise determine as area-weighted average for
watershed land uses.
NRCS Hydrologic Soil Dominant Natural Resource Conservation Service (NRCS-formerly Soil
Group6 Conservation Service) hydrologic soil group-A, B, C, or D.
Soil Type3 NRCS soil type-(c)lay (s)ilt, s(a)nd. Clay particles are smaller than 0.002
millimeters (mm) in diameter. Silt particles are between 0.002 and 0.05 mm in
diameter. Sand particles range from 0.05 mm to 2.0 mm.
Type of Vegetation5 Type of vegetation predominant in pervious areas (i.e. grass turf, dry land
grasses, etc.).
Roads
Total Paved Roadway Total area of paved roads, streets and alleys in the watershed. Associated
Area5 paved shoulders should be included in this area.
Total Length Curb/Gutter Total length of curb & gutter along paved roads, streets, and alleys.
on Paved Roads5
Total Unpaved Roadway Total area of unpaved roads, streets, and alleys in the watershed. Unpaved
Area5 shoulders should be included in this area.
Total Length Curb/Gutter Total length of curb & gutter along unpaved roads, streets, and alleys.
on Unpaved Roads5
% Paved Roads Draining, Parameter calculated by dividing the length of paved roads, etc., draining to
:o Grass Swales/Ditches grass swales and ditches by the total length of paved roads, streets and
alleyways in the watershed.
% Unpaved Roads Percentage of unpaved roads, street and alley areas draining to grass
Draining to Grass swales/ditches that can be calculated by dividing the length of unpaved roads,
Swales/Ditches5 etc., draining to grass swales and ditches by the length of unpaved roads,
streets and alleyways in the watershed.
Type of Pavement on Pavement Type. Can be (C)oncrete,(A)sphalt, or a Mix of (B)oth.
Roads, Streets and Alleys3
Parking Lots
Total Paved Parking LotD Total area of all paved parking lots within the watershed.
Area
Total Length Curb/Gutter Total length of curb & gutter along paved parking lots.
on Paved Lots5
Total Unpaved Parking Lot Total area of all unpaved parking lots within the watershed.
Area5
Total Length Curb/Gutter Total length of curb & gutter along unpaved parking lots.
on Unpaved Lots5
% Paved Lot Area Percentage of parking lot areas draining to grass swales or ditches. This can
aining to Grass Swales5 be calculated by dividing the total parking lot area draining to swales by the
total parking lot area.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 60 April 25, 2002
-------
Data Elements
% Unpaved Lot Area
Draining to Grass Swales5
Type of Pavement in
Parking Lots
Description
Percentage of unpaved parking lot areas draining to grass swales or ditches.
This can be calculated by dividing the total unpaved parking lot area draining
to swales by the total unpaved parking lot area.
Can be (C)oncrete,(A)sphalt, or a Mix of (B)oth. Additionally, provide the
percentages of porous concrete, porous asphalt and porous modular
pavement present relative to the total paved parking lot area.
Land Uses
Land Use Information"
Should be provided for each land use present in the watershed. The percent
of each land use in the watershed, categorized according to % Light Industrial,
% Heavy Industrial, % Multi-family Residential, % Office Commercial, %
Retail, % Restaurants, % Automotive Services, % Rangeland, % Orchard, %
Vegetable Farming, etc.
1 - National Stormwater BMP Database requirement for all BMPs
National Stormwater BMP Database requirement for all Non-Structural BMPs
National Stormwater BMP Database requirement for Non-Structural and Structural BMPs that are based on minimizing
directly connected impervious areas
National Stormwater BMP Database requirement for Structural BMPs that are based on minimizing directly connected
impervious areas
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
161 April 25, 2002
-------
FormB
WATERSHED INFORMATION
Watershed Name
Hydrotogic Unit Code (8-dlgit) EPA Reach Code {RF1 or RF3) _
Unit System (S.I. or U.S. Standard)
Physical Characterstics
Total Watershed Area Total Length of Watershed
Total Length of Grass-Lined Channels
Total Disturbed Area
% Irrigated Lawn and/or Agriculture % Total Impervious Area in Watershed
% of Total Impervious Area that is Hydraullcally Connected
% of Watershed Served by Storm Sewers Storm Sewer Design Return Period _
Average Watershed Slope Average Runoff Coefficient
Hydrologlc Soil Group Soil Type
Type of Vegetation
Roads
Total Paved Roadway Area Total Unpaved Roadway Ares
Total Length of Curb and Gutter on Paved Roads
Total Length of Curb/Gutter on Unpaved Roads
% Paved Roads Draining to Grass Swales/Ditches
% Unpaved Roads Draining to Grass Swales/Ditches
Type of Pavement on Roadways
Sheet 1 of 2
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 62 April 25, 2002
-------
I WATERSHED INFORMATION
Parking Lots
Total Paved Parking Lol Area Total Unpaved Parking Lot Area
Total Length of Curb and Gutter on Pawed Roads
Total Length of Curb/Gutter on Unpaved Parking Lois
% Paved Parking Lot Draining to Grass Swaies/Ditohes
% Unpaved Parking Lot Draining to Grass Swales/Ditches
Type of Pavement in Parking Lots
% Porous Concrete % Porous Asphalt
Land Uses
Land Use Type
% of Land Use in Watershed
Sheet 2 of 2
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormy ater BMP Database Requirements
1 63 April 25, 2002
-------
Structural BMP Information
The purpose of the structural BMP form is to provide general BMP information inherent to all
structural BMP types. Structural BMPs include constructed facilities or measures to help protect
receiving water quality and control stormwater quantity. Representative practices include
structures for storage, infiltration and filtration. Structural BMP information requested includes
items such as date of installation, various design parameters, design drawings, and rehabilitation
and maintenance frequencies. Structural BMP form data elements and the form are presented in
Table
3.13: and Form C, respectively.
Table 3.13: Structural BMP form data elements description
Data Element
Description
BMP Name"
Type of BMP Being
Tested2
What date was the BMP
:acility put into service?2
How many separate inflow
points does the facility
lave? 2
s the BMP designed to
bypass or overflow when
Full?2
Describe the type and
frequency of maintenance,
f any
The name by which the BMP is referred to locally.
The type of structural BMP being tested at the site. Major categories of
structural BMPs include detention basins, retention ponds, wetland channels
and swales, wetland basins, hydrodynamic devices, percolation trenchs and
dry wells, media filters, grass filter strips, porous pavement and infiltration
basins.
Month, day and 4-digit year (e.g., 04/05/1998) when BMP became
operational. If the exact day is unknown, use the first day of the month.
Number of separate inflow points. For example, a wet pond may receive flow
from two (2) storm sewers and one (1) natural drainage, for a total of three (3)
separate inflow points.
Identifies 'Bypass" or "Overflow" when full..
What was the last date
hat the facility was
rehabilitated, if any?
Describe the type of
rehabilitation, if any
Describe the type and
design of each BMP
outlet2
3MP Drawing ^
Type of frequency and maintenance. Practices include: Tree/Shrub/Invasive
Vegetation Control, Mowing, Algae Reduction, Sediment Removal/Dredging,
Litter/Debris Control, Erosion Control/Bank Stability, Inlet Cleaning, Outlet
Cleaning, Media Replacement/Regeneration, Pump Cleaning/Repair, Valve
Cleaning/Repair, Pipe Cleaning/Repair, General Maintenance, Odor Control,
Mosquito Control, Vector Control.
Month, day and 4-digit year (e.g., 04/05/1998) of most recent rehabilitation. If
the exact day is unknown, use the first day of the month.
Description of rehabilitation activities such as structural modification or major
repair.
Outlet configuration and design information.
Drawings of the BMP in plan, profile and layout view.
- National Stormwater BMP Database requirement for all Structural BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
164
April 25, 2002
-------
Form C
STRUCTURAL BMP INFORMATION
BMP Name
Date Facility Placed in Service
Type o< BMP Being Tested
Number of Inflow Points
BMP Designed to Bypass or Overflow
Maintenance Type and Frequency
Last Rehabilitation Date
Type of Rehabilitation
Description, Types, and Designs of Outlets
BMP Layout Drawing
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormy ater BMP Database Requirements
165 April 25, 2002
-------
Non-Structural BMP Information
The purpose of the non-structural BMP form is to provide general BMP information inherent to
all non-structural BMP types. A non-structural BMP can generally be described as a
preventative action to protect receiving water quality that does not require construction.
Nonstructural BMPs rely predominantly on behavioral changes in order to be effective. Major
categories of non-structural BMPs include education, recycling, maintenance practices and
source controls, as described below.
Educational BMPs: Include efforts to inform city employees, the public, and businesses
about the importance of using practices that protect stormwater from improper use, storage,
and disposal of pollutants, toxics, household products, etc. The ultimate goal of educational
BMPs is to cause behavioral changes.
Recycling BMPs: Include measures such as collecting and recycling automotive products,
household toxics, leaves, landscaping wastes, etc.
Maintenance practices: Include measures such as catch basin cleaning, parking lot sweeping,
road and street pavement repair, road salting and sanding, roadside ditch cleaning and
restoration, street sweeping, etc.
Source controls: Include preventing rainfall from contacting pollutant-laden surfaces and
preventing pollutant-laden runoff from leaving locations such as automobile maintenance,
salvage and service stations; commercial, restaurant and retail sites; construction sites;
farming and agricultural sites; industrial sites, etc.
The Non-structural BMP form data reports narrative/descriptive information on the type and
extent of the BMP being practiced, as well as cost data. The non-structural BMP form and the
form fields are described in Table 3.14: and Form D, respectively.
Table 3.14: Non-structural BMP form data elements description
Data Element Description
Non-structural BMP Type Categories of non-structural BMPs, such as education, recycling, maintenance
practices and source controls.
BMP Name for the subject BMP Name for the subject non-structural BMP (e.g., Erosion and Sediment
non-structural BMP3 Control Pamphlets).
Date Test Began" Date (month, day and 4-digit year) that the BMP test was begun (e.g.,
01/01/1998).
Educational BMPJ Measure of eductational BMP effectiveness/progress. Examples include: the
"measurements" number of brochures distributed per resident and employee in watershed per
year, number of radio ads, percent of stormwater inlets in watershed stenciled,
etc.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 66 April 25, 2002
-------
Data Element
Recycling BMP
"measurements" 3
Maintenance BMP
"measurements" 3
Source Control
"measurements" 3
Cost Information
Initial Costs
Annual Costs
Description
Measure of recycling BMP effectiveness/progress. Could include gallons of
used oil collected per resident in the watershed; pounds of household toxics
collected per resident in the watershed; tons of landscaping waste per
resident collected, etc.
Measure of maintenance BMP effectiveness/progress. Could include percent
of stormwater catch basins cleaned once each year, twice each year, etc.;
tons of materials removed per average inlet each year; lane miles of street
swept each year and tons of material removed per lane mile each year; etc.
Measure of source control BMP effectiveness/progress. Could include
percent of industrial storage area in watershed that is covered; etc.
Initial costs, including the time and measures necessary to design and
mplement a program.
Year-to-year costs once the initial program has been developed.
3 - National Stormwater BMP Database requirement for all Non-Structural BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
167 April 25, 2002
-------
FormD
NON-STRUCTURAL BMP INFORMATION"
BMP Name
Type of BMP Being Tested
Date Test Began
Description of Quantity or Measure of BMP
Cost Information
Initial Costs
Annual Costs
Urban Stornmater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
168 April 25, 2002
-------
Detention Basin Design Data
The primary purpose of the detention basin design data form is to provide structural BMP
information specific to detention basins. Detention basins are designed to collect stormwater
runoff and completely empty sometime after the end of the runoff event. Detention basins used
for water quality purposes differ from flood control basins only by their outlet structures.
Detention basin design characteristics are extremely important for comparing their performance
under various hydrological and environmental conditions. The detention basin form and the
form data elements are presented respectively in Form E and Table 3.15.
Table 3.15: Detention basin design form data elements list
Data Element Description
lA/ater Quality Detention The volume of storm runoff that is captured and slowly drained over a period
Volume4 of time (e.g., 12 to 48 hours).
lA/ater Quality Detention The area of the water surface in the detention basin at full water quality
Surface Area When Full4 detention volume.
i/Vater Quality Detention Length of the water quality detention basin, measured as the distance
Basin Length4 between inflow and outflow. If there is more that one inflow point, use the
average distance between the inflow points and the outflow weighted by the
tributary impervious area.
Detention Basin Bottom Area of the bottom of the entire detention basin, not including the side slopes
ea4 but including the bottom stage area.
Brim-full Volume Emptying Emptying time (in hours) of the water quality detention volume.
fime4
Half Brim-full Volume Emptying time (in hours) of the lower half of the water quality detention
Emptying Time4 volume.
Bottom Stage Volume, If The volume of the lower "bottom stage" portion (if applicable) of the detention
y4 basin.
Bottom Stage Surface The surface area of the lower "bottom stage" portion (if applicable) of the
^rea, If Any4 detention basin.
s There a Micro Pool?4 "Yes" or "No" indication of micropool.
rorebay Volume4 Volume of the forebay portion of the detention basin when filled to the point of
overflow into the rest of the basin.
rorebay Surface Area4 Surface area of water in the forebay at the level of overflow to the bottom
stage.
Describe Vegetation CoverDescribe the types of vegetation on the basin sides and floor.
i/Vithin Basin
rlood Control Volume, If The flood control detention volume in excess of the water quality detention
y4 basin volume (if any).
_ist Design Flood Return List the flood return period (in years) for which the flood control volume is
3eriods4 designed (e.g., 25-year).
Depth to Seasonal High The minimum depth from the basin bottom to the water table during the
i/Vater Table, If Known monitoring season.
Detention Basin Construction Cost Estimates
/ear of Cost Estimate Four-digit year (e.g., 1998) for which cost estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 69 April 25, 2002
-------
Data Element [Description
Structural Control Devices The estimated cost of establishing all structural control devices, such as inlet
and outlet structures, trash racks and energy dissipaters, including cost of
materials and construction.
Vegetation and The estimated cost of establishing vegetation for the BMP, including acquiring
Landscaping Costs landscape materials, establishing vegetation, and establishing the irrigation
infrastructure, if any.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
.and Costs or Values The estimated value of the land or the cost of acquiring the land.
Rehabilitative Costs: |
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the
Removal Costs detention basin.
Average Annual Estimated average annual cost to revegetate the sides and floor of the
Revegetation Costs detention basin.
- National Stomiwater BMP Database requirement for all Non-Structural BMPs
Urban Stomiwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stomiwater BMP Database Requirements
1 70 April 25, 2002
-------
Form E
DETENTION BASIN DESIGN DATA
Test Sile Name
Watershed Name
BMP Name
Design Information
Water Quality Detention Volume
Water Quality Detention Surface Area When Full
Water Quality Detention Basin Length
Detention Basin Bottom Area
Brim-full Volume Emptying Time
Bottom Stage Volume, If Any
Hatf Brim-lull Volume Emptying Tbite
_ Is there a micro pool?
Bottom Stage Surface Area, If Any
Forebay Volume
Forebay Surface Area
Vegetation Cover Within Basin
Rood Control Volume, If Any
Depth to Water Tabte
Design Rood Return Periods
Detention[BasinConstruction _Cost Estimates
Year of Cost Estimate
Construction Costs:
Excavation
Vegetation and Landscaping
Land Costs and Value
Rehabilitative Costs:
Average Annual Sediment Removal
Average Annual Revegetation
Structural Control Devices
Engineering and Overhead
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
111
April 25, 2002
-------
Retention Pond Design Data
The retention pond design data form reports BMP specific information for retention ponds.
Retention ponds are also commonly known as "wet ponds" because they have a permanent pool
of water, unlike detention basins, which dry out between storms. The permanent pool of water is
replaced in part, or in total, by stormwater during a storm event. The design is such that any
available surcharge capture volume is released over time. Retention of stormwater in the
permanent pool over time can provide biochemical treatment. A dry weather base flow, pond
liner and/or high groundwater table are required to maintain the permanent pool. The retention
pond form and the form data elements descriptions are shown in Form F and Table 3.16:
Table 3.16: Retention pond design form data elements list
Data Element
Description
Volume of the permanent pool of water.
Area of the water surface in the permanent pool.
Volume of permanent
pool4
3ermanent Pool Surface
Area4
Permanent Pool Length4 Length of the permanent pool of water, measured along the axis of the inflow
and outflow. If more that one inflow point, use the average distance between
the inflow points and the outflow weighted by the tributary impervious area.
Littoral Zone Surface Surface area of the littoral zone. The littoral zone refers to the area above the
Area4 level of the permanent pool that is periodically and temporarily covered by
captured storm runoff.
Littoral Zone Plant Species List plant species (by Latin name, if known), percent of cover and densities in
List the littoral zone.
Water Quality Surcharge Water quality detention volume above permanent pool, when full.
Detention Volume When
Full4
Water Quality Surcharge
Surface Area When Full4
Water Quality Surcharge
The surface area of water quality detention volume above the permanent pool
if applicable.
Length of the water quality detention volume, measured along the axis
between the inflow and outflow. If more that one inflow point, use the average
distance between the inflow points and the outflow weighted by the tributary
impervious area.
Time (in hours) required for the retention pond water quality surcharge
detention volume to be released to the permanent pool level.
Time (in hours) required for the lower half of the retention pond water quality
surcharge detention volume to be released to the permanent pool.
Volume of the forebay portion of the retention basin when it is filled to the
point of overflow into the lower part of the basin.
Surface area of water in the forebay when it is filled to the point of overflow
into the lower part of the basin.
Describe Vegetation CoverDescribe the types of vegetation (provide Latin names, if known) on the basin
Within Basin sides and floor.
Flood Control Volume, If The flood control detention volume in excess of the retention basin volume (if
any).
List the flood return period (in years) for which the flood control volume is
designed (e.g., 25-year).
Basin Length
Brim-full Emptying Time
For Surcharge4
Half Brim-full Emptying
Time For Surcharge4
Forebay Volume4
Forebay Surface Area4
Any4
List Design Flood Return
3eriods (in years)
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
172
April 25, 2002
-------
Data Element Description
[Retention Pond Construction Cost Estimates
Year of Cost Estimate Four-digit year (e.g., 1998) for which cost estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
Structural Materials Costs The estimated cost of materials used in constructing the retention pond,
excluding vegetation costs.
Basin Construction Costs The estimated cost for construction of the retention pond, including site survey
and construction activities.
Structural Control Devices The estimated cost of establishing all retention pond control devices, such as
osts inlet and outlet structures, spillways, and culverts. Includes the cost of
materials and construction.
Vegetation and The estimated cost of establishing vegetation for the BMP, including acquiring
Landscaping Costs landscape materials, etc.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative Costs:
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the
Removal Costs retention pond.
Average Annual Estimated average annual cost to revegetate and/or reseed the retention
Revegetation Costs pond.
4 - National Stormwater BMP Database requirement for all Retention Ponds
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 73 April 25, 2002
-------
FormF
RETENTION POND DESIGN DATA
Test SHe Name
Watershed Name
BMP Name
Design Information
Volume of Permanent Pool
Permanent Pool Surface Area
Littoral Zone Surface Area
Littoral Zone Plant Species
Permanent Pool Length
Water Quality Surcharge Detention Volume
Water Quality Surcharge Surface Area, When Full
Water Quality Surcharge Basin Length, When Full
Brim-full Emptying Time
Forebay Volume
Rood Control Volume
Half Brim-full Emptying Time
Forebay Surface Area
Design Flood Return Periods
Retention Pond Construction Cost Estimate
Year of Cost Estimate
Construction Costs:
Excavation
Basin Construction
Vegetation and Landscaping
Land Costs and Value
Structural Materials Cost
Structural Control Devices
Engineering and Overhead
Rehabilitative Costs:
Average Annual Sediment Removal
Average Annual RevegetaHon
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
174
April 25, 2002
-------
Percolation Trench and Dry Well Design Data
The percolation trench and dry well form contains essential design information for percolation
trenches and dry wells. Percolation or infiltration trenches can be generally described as
trenches or excavations filled with porous media designed to encourage rapid percolation of
runoff to the groundwater. A dry well is a drilled well, often drilled through impervious layers to
reach lower pervious layers. The percolation trench and dry well form and data elements are
presented in Table 3.17: and Form G.
Table 3.17: Percolation trench and dry well design form data elements list
Data Element
Description
Percolation Trench/Well Surface The top surface area of the percolation trench or well.
4
Percolation Trench/Well Length The length of the percolation trench, or the diameter of the well.
Percolation Trench/Well Depth4 The depth of the trench or the well that is exposed to permeable soils.
Depth to Seasonal High The minimum depth to the seasonal high groundwater table below the
Groundwater Below Bottom of trench or well.
Trench/Well4
Depth to Impermeable Layer The depth to the first impermeable layer below the trench or well.
Below Bottom of Trench/Well4
Depth and Type of Each Soil The order of stratification (from the surface downward) and the depth of
Layer Adjacent To and Below each layer of soils at the BMP site.
Trench/Well4
Type and Gradation of Granular Describe the type and depth of granular material used in the trench or
Materials Used in Trench/Well4 well.
Was Geotextile Used Above "Yes" or "no" indication of geotextile use above granular fill.
Granular Trench Fill? 4
Was Geotextile Used On the "Yes" or "no" indication of geotextile use on sides of granular fill.
Sides of Granular Fill?4
Was Geotextile Used On the "Yes" or "no" indication of geotextile use below granular fill.
Bottom of Granular Fill? 4
Give porosity (in percent) of the The volumetric portion of the granular material that is not occupied by
granular fill material4 solid matter (expressed as a percent).
Total Storage Pore Volume in The volume of the available pore space in the granular materials.
Trench4
Describe Type of Geotextile Describe the types and locations of the geotextile fabrics used in the
Used4 trench or well, if any. Include the effective pore opening of the fabrics.
Hydraulic Conductivity of The hydraulic conductivity of the soils adjacent to the trench or well
Adjacent Soils4 infiltration surfaces.
Groundwater Flow Gradient4 The flow gradient of groundwater below the infiltration basin (expressed
as unit length per unit length, e.g., feet/feet).
urpose of Trench or Well Describe the purpose of the percolation trench or well (e.g., water quality
treatment, reduction of surface runoff, groundwater recharge, etc.).
Jercolation Trench and Dry Well Construction Costs Estimates
Year of Cost Estimate Four-digit year (e.g., 1998) for which cost estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
Well Drilling The estimated cost of establishing the well, if this is a dry well.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
175 April 25, 2002
-------
Data Element
Description
Trench Construction Costs The estimated cost of establishing the trenches, if this is a percolation
trench.
Structural Control Devices Costs The estimated cost of establishing all percolation trench or dry well
control devices, such as inlet and outlet structures and culverts. Include
the cost of materials and construction.
Structural Materials Costs The estimated cost of materials used in the percolation trench, such as
granular fill and geotextiles.
Engineering and Overhead The estimated engineering and associated overhead costs, including
Costs site, structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of
acquiring this land.
Rehabilitative Costs:
Average Annual Sediment
Removal Costs
Estimated average annual cost to remove sediment accumulated in the
retention pond.
4 - National Stormwater BMP Database requirement for all Percolation Trenches and Dry Wells
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 76 April 25, 2002
-------
Form G
PERCOLATION TRENCH AND PRY WELL DESIGN DAT/
Test Site Name
Watershed Name BMP Name
Design Information
Percolation Trench/Well Surface Area
Percolation Trench/Well Length Percolation Trench/Wall Depth
Depth to Qroundwater Depth to Impermeable Layer
Depth and Type of Each Soil Layer Type and Gradation of Granular Materials Used
Was geotextlle fabric used above granular trench fill? Type of Geotextlle Used, H Any
Was Geotextile Used On the sides of granular fill?
Was Geotextlle Used On the Bottom of Granular Fill?
Porosity of Granular Material Total Storage Volume
Hydraulic Conductivity of Soils
Groundwater Flow Gradient
Purpose of Trench or Well
Percolation Trench and Dry Well Construction Cost Estimates
Construction Costs: Year of Cost Estimate
Excavation Well Drilling
Trench Construction Structural Control Devices
Structural MateriaJs Engineering and Overhead
Land Costs or Value
Rehabilitative Costs:
Average Annual Sediment Removal
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 77 April 25, 2002
-------
Media Filter Design Data
The media filter design data form contains design information related to the performance of
media filters. A Media Filter is a facility that uses some form of granular or membrane filter,
with or without a pre-settling basin, to remove a fraction of the constituents found in stormwater.
The most typical filter is sand, but other materials, including peat mixed with sand, compost with
sand, geotextiles, and absorption pads and beds are commonly used. The media filter form and
data elements are presented in Table 3.18 and Form H.
Table 3.18: Media filter design form data elements list
Data Element I Description
Permanent Pool Volume Volume of the permanent pool (if any) if the pool is part of the filter basin
Upstream of Filter Media, If installation and not a separate pretreatment retention pond or a detention
Any4 basin.
Permanent Pool Surface Area of the water surface in the permanent pool (if any).
Area of Sedimentation
Basin Preceding Filter, If
Any4
Permanent Pool Length of Length of the permanent pool (if any) measured as the distance from pool
Sedimentation Basin inflow to outflow. If more than one inflow point, use the average length.
Preceding Filter, If Any4
Surcharge Detention The design water quality capture volume, including the volume above the
Volume, Including Volume filter.
Above Filter Bed
Surcharge Detention The surface area of the design water quality captured runoff including the area
Volume Surface Area4 above the filter.
Surcharge Detention The length of the design captured runoff volume, including the portion above
Volume Length the filter, measured as the distance along the flow path. If more than one
inflow point, use the average length.
Surcharge Detention The design time for complete drawdown (in hours) of the water quality capture
Volume's Design Drain volume if the drain time is controlled by a flow regulating device such as an
Time, If Controlled and orifice. Leave blank if the drain rate is only a fraction of the filter's flow-
Known 4 through rate.
Surcharge Detention The design depth of water quality capture volume that can be stored above
Volume Design Depth4 the filter before overflow or runoff bypass occurs.
Media Filter Surface Area4 Surface area of the media filter (e.g., the sand bed orgeotextile filter) as a
whole orthogonal to the flow.
Angle of Sloping or Vertical Inclination of filter in degrees above the horizontal plane.
Filter4
Number of Media Layers in The number of layers of different filter materials in this BMP.
Filter4
Describe Depth and Type Describe the type of media used in the filter (Example: ASTM C-33 Sand with
of Each Filter Media Layer4d50=0.7 mm, 50% ASTM C-33 Sand with d50=0.6 mm and 50% Peat).
Media Filter Construction Cost Estimates
Year of Cost Estimate Four-digit year (e.g., 1998) for which the above estimates were made.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
178 April 25. 2002
-------
Data Element
Description
[Construction Costs:
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
| drilling and blasting, trenching and shoring.
iasin Construction Costs The estimated cost for construction of the media filter, including site survey
and construction activities.
liter Construction Costs The estimated cost of establishing the filter system itself, including filter
material and the underdrain system. Include costs of materials and
construction.
tructural Control Devices The estimated cost of establishing all BMP control devices, such as inlet
3osts devices, trash racks, energy dissipaters, and outlet structures. Include costs
of materials and construction.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
Dosts structural, and landscape design and engineering expenses.
.and Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative/
Maintenance Costs:
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the media
Removal and Media filter and replace the filter material.
Replacement Costs
National Stonnwater BMP Database requirement for all Media Filters
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
179
April 25, 2002
-------
FormH
MEDIA FILTER DESIGN DATA
Test Site Name
Watershed Name BMP Name
Design Information
Permanent Pool Volume Upstream of Media Fitter, If Any
Permanent Poors Surface Area Permanent Pool's Length
Surcharge Detention Volume, Including Volume Above Fitter Bed
Surcharge Detention Volume Surface Area, Including Volume Above Filter Bed
Surcharge Detention Volume's Length
Surcharge Detention Volume's Design Depth
Surcharge Detention Volume's Drain Time In Hours
Media Filter's Surface Area
Angle of sloping or vertical filter media In degrees (D to 90)
Number of Filter Layers
Type and Depth (or Thickness) ot Each Fitter Media Layer
Media Filter Construction Cost Estimates
Year ol Cost Estimate
Construction Costs:
Excavation Basin Construction
Filter Construction Structural Control Devices
Engineering and Overhead Land Costs and Value
Rehabilitative Costs:
Average Annual Sediment Removal and Media Filter Replacement Costs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
180 April 25, 2002
-------
Grass Filter Strip Design Data
The grass filter strip form provides design information specific to grass filter strips. Grass filter
strips, sometimes called buffer strips, are vegetated areas designed to accept sheet flow provided
by flow spreaders which accept flow from an upstream drainage area. Vegetation may take the
form of grasses, meadows, forests, etc. The primary mechanisms for pollutant removal are
filtration, infiltration, and settling. The grass filter strip form and data elements are shown in
Table 3.19 and Form I.
Table 3.19: Grass filter strip form data elements list
Data Element Description
Grass Strip Length Length of the grass strip in the direction of the flow path.
Grass Strip Slope4 The slope of the strip along the flow path expressed as unit length per unit
length (e.g., feet/feet).
Flow Depth during 2-Year The design depth of flow over the strip during the 2-year storm peak flow.
Storm4
2-Year Peak Flow The design flow velocity over the strip during the 2-year peak flow.
Velocity4
Describe Grass Species List of grass species and their densities.
and Densities4
s Strip Irrigated?4 "Yes" or "no" indication of irrigation.
Estimated Manning's n The estimated Manning's roughness factor, n, during the 2-year flow event.
During 2-Year Flow
Depth to Groundwater or Depth to the seasonal high groundwater table and/or the impermeable layer,
mpermeable Layer whichever is shallower.
Measured Saturated Rate of infiltration into the filter strip under saturated soil conditions.
nfiltration Rate, if Known
NRCS Hydrologic Soil The Natural Resource Conservation Service Hydrologic Soil Group (e.g., A, B,
Group C, or D) comprising the infiltrating surface.
Grass Filter Strip Construction Cost Estimates
of Cost Estimate Four-digit year (e.g., 1998) for which the above estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
Structural Control Devices The estimated cost of establishing all BMP control devices, such as slotted
Costs curbing or other flow spreading devices, and outflow collection and
conveyance systems. Include costs of materials and construction.
Vegetation and The estimated cost of establishing vegetation for the BMP, including acquiring
Landscaping Costs landscape materials, establishing vegetation, and establishing the irrigation
infrastructure, if any.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative Costs: |
Average Annual Sediment Estimated average annual cost to remove sediment accumulated on the grass
Removal Costs filter strip.
Average Annual Estimated average annual cost to revegetate and/or reseed the grass filter
Revegetation Costs strip.
- National Storm water BMP Database requirement for all Grass Filter Strips
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
181 April 25. 2002
-------
Form I
GRASS FILTER STRIP DESIGN DATA'
Test Site Name
Watershed Name BMP Name
Design Information
Grass Strip's Length Longitudinal Slope
Row Depth during 2-Year Storm 2-Year Peak Bow Velocity
Grass Species and Densities Is Strip Irrigated?
Manning's n During 2-year Row
Depth to Groundwater
Saturated Infiltration Rate
Soil Group
Percolation Trench and Dry Well Construction Cost Estimates
Year of Cost Estimate
Construction Costs:
Excavation Structural Control Devices
Vegetation and Landscaping Engineering and Overhead
Land Costs or Value
Rehabilitative Costs:
Average Annual Sediment Removal Costs
Average Annual Revegetatlon Costs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 82 April 25, 2002
-------
Wetland Channel and Swale Design Data
The purpose of the wetland channel and swale design form is to consistently collect and report
wetland channel and swale information. A wetland channel is a channel designed to flow very
slowly, probably less than two feet per second at the two-year flood peak flow rate. It has, or is
designed to develop, dense wetland vegetation on its bottom. A swale is a shallow grass-lined
channel designed for shallow flow near the source of storm runoff. The wetland channel and
swale form and data elements are provided in Table 3.20 and Form J.
Table 3.20: Wetland channel and swale form data elements list
Data Element
Average Longitudinal
Inflow Spacing
Length of Channel/Swale4
Longitudinal Slope of
Channel/Swale4
Bottom Width of
Channel/Swale4
Side Slope of
Channel/Swale4
2-Yr Flow Design Depth in
Channel/Swale4
2-Yr Peak Design Flow
Velocity4
2-Yr Manning's n
Type of Plant Species in
Wetland Zone or Swale4
Maximum Design Flow
Capacity Return Period of
Swale
Depth to High
Groundwater or
Impermeable Layer
Groundwater Hydraulic
Conductivity
Description
The average longitudinal spacing between all separate stormwater inflow
points.
The length of the wetland channel or swale, from the stormwater inflow to
outflow point.
The average longitudinal slope (in unit length per unit drop, e.g., feet per feet
or meter per meter) of the wetland channel or swale, as measured between
grade control structures.
The average width of the nearly flat bottom of the channel or swale between
its side slopes.
The average (in vertical unit length per horizontal unit length) of the channel or
swale's side slopes.
The average depth of water in the channel or swale during the two-year flood
peak flow.
The flow velocity in the channel or swale during the two-year flood peak flow.
The Manning's roughness factor, n, for the 2-year peak flow.
List the plant species, percent of cover and densities.
The flood return period that the channel has been designed to convey within
its banks in addition to the water quality design event. (Example: 2-year and
10-year flood).
The minimum depth to the water table during the high water table season, or
to the first impermeable layer.
The hydraulic conductivity of the groundwater below the channel or swale.
Wetland Channel and Swale Construction Cost Estimates
Year of Cost Estimate
Construction Costs:
Excavation Costs
Structural Control Devices
Costs
Four-digit year (e.g., 1998) for which cost estimates were made.
The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
The estimated cost of establishing all wetland channel or swale control
devices, such as inlet and outlet devices, trash racks, etc. Include the cost of
materials and construction.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
183 April 25. 2002
-------
Data Element
Description
Vegetation and The estimated cost of establishing vegetation for the BMP, including acquiring
Landscaping Costs landscape materials, establishing vegetation, and establishing the irrigation
infrastructure, if any.
Engineering and OverheadThe estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative Costs: |
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the
Removal Costs swale/wetland channel.
Average Annual Estimated average annual cost to revegetate the sides and floor of the
Revegetation Costs swale/wetland channel.
National Stormwater BMP Database requirement for all Wetland Channels/Swales
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BAfP Database Requirements
1 84 April 25, 2002
-------
Form J
WETLAND CHANNEL AND SWALE DESIGN DAT/
Test SMe Name
Watershed Name
Design Information
Length of Channel/Swale
BMP Name
Side Slope of Channel/Swale
Bottom Width of Channel/Swale
Longitudinal Slope of Channel/Swale
Average Longitudinal Inflow Spacing
2-Year Flow Design Depth In Channel/Swale
2-Year Peak Design Row Velocity
Depth to High Groundwater
Plant Species in Wetland Zone/Swale
2-Year Manning's n
Groundwater Hydraulic Conductivity
Design Flow Return Periods
Wetland Channel and Swale Construction Cost Estimates
Year of Cost Estimate
Construction Costa:
Excavation Costs
Vegetation
Land
Rehabilitative Costs:
Sediment Removal
Revegetation
Control Devices
Engineering and Overhead
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
185
April 25, 2002
-------
Porous Pavement Design Data
The porous pavement form provides design information particular to porous pavement BMPs.
There are two forms of porous pavement: modular block, which is made porous through its
structure, and poured-in-place concrete or asphalt which is porous due to the mix of the
materials. Modular block porous pavement consists of perforated concrete slab units underlain
with gravel. The surface perforations are filled with coarse sand or sandy turf. It is used in low
traffic areas to accommodate vehicles while facilitating stormwater runoff at the source. It
should be placed in a concrete grid that restricts horizontal movement of infiltrated water through
the underlying gravels. Poured-in-place porous concrete or asphalt is generally placed over a
substantial layer of granular base. The pavement is similar to conventional materials, except for
the elimination of sand and fines from the mix. If infiltration to groundwater is not desired, a
liner may be used below the porous media along with a perforated pipe and a flow regulator to
slowly drain the water stored in the media over a 6 to 12 hour period. The porous pavement
design form and data elements are given in Table 3.21 and Form K.
Table 3.21: Porous pavement form data elements
Data Element Description
Porous Pavement Surface Surface area of the porous pavement.
Area4
Depth to Seasonal High The minimum depth to the seasonal water table below the porous pavement.
Groundwater4
Depth to Impermeable The depth to the first impermeable layer below the BMP, if known.
Layer4
NRCS Hydrologic Soil The Natural Resource Conservation Service Hydrologic Soil Group (e.g., A, B
Group C, or D) comprising the infiltrating surface.
Infiltration Rate4 Rate of infiltration for site soils under saturated conditions.
Type of Granular or Other Describe the type and depth of each granular material layer under the porous
Materials Used in or Below pavement, if any. Include each layer of geotextile fabric used as though it was
Pavement4 a granular layer.
Porosity of Granular Porosity measures the volumetric portion of the filter material that is not
Materials, as a Percent4 occupied by solids. If the layer is geotextile fabric, give the effective pore size.
Is Grass Growing in "Yes" or "No" indication of grass growing in modular pores.
Modular Pores?
If Yes, is Grass Healthy? "Yes" or "No" indication of grass health, if applicable.
Describe Depth of Each The order of stratification (from the surface downward) and the depth of each
Soil Layer Below layer of soils below the porous pavement, to a depth of at least ten feet (3.05
Pavement, If Known meters).
Total Storage Volume The volume of water stored in depressions or as a result of attenuation (if any)
Above Pavement, If Any4 above the porous pavement surface.
Estimated Drain Time (hrs)The emptying time of the storage volume above the pavement.
of Storage Volume Above
Pavement, If Any4
Total Storage Volume The net available volume of the pore spaces in the granular materials under
Under Pavement, If Any4 the porous pavement, if any.
Estimated Drain Time of The total emptying time (in hours) for the storage detention volume under the
Storage Volume Under pavement.
Pavement, If Any4
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
186 April 25. 2002
-------
Data Element
Description
Groundwater Hydraulic
Conductivity
Groundwater Flow
Gradient
Does Porous Pavement
Have Underdrains?4
Describe Purpose of
Porous Pavement
The hydraulic conductivity of the groundwater underlying the BMP.
The flow gradient (in unit length per unit length, e.g. feet/feet) of groundwater
below the infiltration basin.
"Yes" or "No" indication of underdrains for the porous pavement.
Describe the purpose(s) of the porous pavement (examples: water quality
treatment, reduction in peak surface runoff rate and volume, groundwater
recharge, etc.)
Porous Pavement Construction Cost Estimates
Year of Cost Estimate Four-digit year (e.g., 1998) for which cost estimates were made.
Construction Costs:
I
Excavation Costs
The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
The estimated cost of establishing the structural and piping features of the
BMP, including modular blocks, retaining concrete, sub-base material, and
inlay material. Include costs of materials and construction.
The estimated cost of establishing the granular fill for the BMP, including sand
or gravel inlay materials, filter fabric, and perforated underdrain (if any).
Include costs of materials and construction.
If poured-in-place porous concrete or asphalt paving was used, this is the
estimated cost of establishing the paving. Include costs of materials and
construction.
The estimated cost of establishing curbs and gutters for the BMP. Include
costs of materials and construction.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
Costs structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Structural and Piping
Costs
Granular Fill Costs
Paving Costs
Curb and Gutter Costs
Rehabilitative/
Maintenance Costs:
Average Annual
Vegetation Replacement
and Granular Media
Replacement and
Maintenance Costs
Estimated average annual cost to revegetate void spaces in modular block
pavement. If poured-in-place porous pavement, report estimated average
annual cost to wash, vacuum, pressure wash, patch, gutter clean, etc. at a
frequency that ensures the continued function of the BMP.
National Stornrwater BMP Database requirement for all Porous Pavement
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
187
April 25, 2002
-------
FormK
1
Test Site Name
Watershed Name
POROUS PAVEMENT DESIGN DATA I
BMP Name
Design Information
Porous Pavement Surface Area Depth to Groundwater
Depth to Impermeable Layer NRCS Hydroiogic Soil Group
Infiltration Rate
Is grass growing In modular pores? if yes, is grass healthy?
Total Storage Volume Above Pavement, If Any
Estimated Drain Time of Storage Volume Above Pavement, If Any
Total Storage Volume in the Granular Media Below Pavement
Estimated Drain Time of Porous Media Volume
Groundwater Hydraulic Conductivity
Groundwater Flow Gradient
Does porous pavement have underdralns?
Depth of Each Soil Layer Below Pavement Purpose of Basin Above Pavement
Porous Pavement Construction Cost Estimates
Construction Costs: Year ol C08* Estimate
Excavation Granular Fill Paving
Structural and Piping Curb and Gutter
Lend Costs and Value Engineering and1 Overhead
Rehabilitative Costs:
Average Annual Vegetation Replacement and Granular Media Replacement Costs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
188 April 25, 2002
-------
Infiltration Basin Design Data
The infiltration basin form reports important design information for infiltration basins. An
infiltration basin is a basin that can capture a given stormwater runoff volume and infiltrate it
into the ground, transferring this volume from surface flow to groundwater flow. The infiltration
basin form and data elements are listed in Table 3.22 and Form L.
Table 3.22: Infiltration basin form data elements list
Data Element
Capture Volume of Basin4
Surface Area of Capture
Volume, When Full
Infiltrating Surface Area4
Basin Length
Depth to Seasonal High
Groundwater Below
Infiltrating Surface4
Depth to Impermeable
Layer Below Infiltrating
Surface4
NRCS Hydrologic Soil
Group
Depth and Type of Each
Layer of Soil
Field Measured Infiltration
Rate
List Plant Species on
Infiltrating Surface4
Describe Granular Material
on Infiltrating Surface, If
Any4
Hydraulic Conductivity of
Underlying Soils
Groundwater Flow
Gradient
Flood Control Volume
Above Water Quality
Detention Volume
List All Design Flood
Control Return Periods
Describe Purpose of Basin
Description
The design runoff capture volume of the basin.
The area of the water surface in the infiltration basin, when full.
The plan area of the surface used to infiltrate the water quality volume.
Length of the infiltration basin, measured as the distance between inflow and
outflow.
Depth to the seasonal high groundwater table.
Depth to the impermeable layer, if any.
The Natural Resource Conservation Service Hydrologic Soil Group (e.g., A, B,
C, or D) comprising the infiltrating surface.
The order of stratification (from the surface downward) and the depth of each
layer of soils at the infiltration basin site, to a depth of at least ten feet (3.05
meters).
The saturated soil infiltration rate, based on soil surveys, infiltrometer
measurements or observed draw down of a new basin.
List the plant species (by Latin names, if known) and densities of cover on the
bottom of the infiltration basin.
Describe the granular material and its depth and porosity (if any).
The hydraulic conductivity of the soils underlying the infiltration surface.
The flow gradient (in unit length per unit length, e.g. feet/feet) of groundwater
below the infiltration basin.
The volume of the flood control detention volume above the infiltration basin
volume.
List the flood return period (in years) for which the flood control volume is
designed (e.g., 25-year).
Describe the purpose of the infiltration basin (e.g., surface water quality only,
groundwater recharge, etc.).
Infiltration Basin Construction Cost Estimates
Year of Cost Estimate
Construction Costs:
Excavation Costs
Four-digit year (e.g., 1998) for which cost estimates were made.
I
The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
189
April 25, 2002
-------
Data Element Description
Structural Materials Costs The estimated cost of materials used in constructing the infiltration basin,
excluding vegetative cover.
Basin Construction Costs The estimated cost for construction of the infiltration basin, including site
survey and construction activities.
Structural Control Devices The estimated cost of establishing all BMP control devices, such as inlet
osts devices, trash racks, energy dissipators, and outlet structures. Include costs
of materials and construction.
Vegetation and The estimated cost of establishing vegetation for the infiltration basin,
Landscaping Costs including acquiring landscape materials, establishing vegetation, and
establishing the irrigation infrastructure, if any.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative/
Maintenance Costs:
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the
Removal Costs infiltration basin.
Average Annual Estimated average annual cost to revegetate the infiltration basin.
Revegetation Costs
4 - National Stormwater BMP Database requirement for all Infiltration Basins
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 90 April 25, 2002
-------
Form L
INFILTRATION BASIN DESIGN DATA
Test Site Name
Watershed Name
BMP Name
Design Information
Capture Volume of Basin
Basin Length
Surface Area of Capture Volume When Full _
Infiltrating Surface Area
Depth to Groundwater
Soil Group
Depth and Type of Each Soil Layer Below Basin
Depth to Impermeable Layer
Infiltration Rate
Row Gradient
Hydraulic Conductivity
of Underlying Soils
Plant Species on Infiltrating Surface
Granular Material on Infiltrating Surface
Design Flood Control Return Periods Purpose of Basin
Rood Control Volume above Water Quality Detention Volume:
Inftttration Basin Construction Cost Estimates
Construction Costs:
Excavation
Basin Construction
Vegetation and Landscaping
Land Costs and Value
Year of Cost Estimate
Structural Materials Cost
Structural Control Devices
Engineering and Overhead
Rehabilitative Costs:
Average Annual Sediment Removal
Average Annual RevegetatJon
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stonmvater BMP Database Requirements
191
April 25, 2002
-------
Hydrodynamic Device Design Data
The hydrodynamic device form provides important design criteria specific to hydrodynamic
devices. The hydrodynamic device BMP category includes BMPs such as oil-water separators,
sand interceptors, swirl-type concentrators, sedimentation vaults, and other prefabricated and
package-type treatment devices. The hydrodynamic device form and data elements are provided
in Table 3.23 and Form M.
Table 3.23: Hydrodynamic device form data elements
Data Element
Description
Volume of Permanent Volume of the permanent pool (dead pool) of water.
Pool4
Permanent Pool Surface Area of the water surface in the permanent pool (dead pool).
Permanent Pool Length4 Length of the permanent pool of water, measured as the distance between
inlet and outlet. If more than one inlet location, use the average distance
between the inlet location and the outlet location.
Water Quality Surcharge The surcharge detention volume above the permanent pool volume (device
Detention Volume When active storage volume).
Full4
nlet Chamber Volume, If Volume of the inlet chamber portion of the hydrodynamic device when it is
<\ny4 filled to the point of overflow into the lower (next) part of the device.
Brim-full Emptying Time Time (in hours) required for the hydrodynamic device water quality surcharge
For Surcharge4 detention volume to be released from the outlet discharge.
Half Brim-full Emptying Time (in hours) required for the lower half of the hydrodynamic device water
Time For Surcharge4 quality surcharge detention volume to be discharged from the outlet.
;omments. This field can be used for comments and other miscellaneous information
such as model type and related manufacturer's specifications for design.
Hydrodynamic Device Construction Cost Estimates
of Cost Estimate Four-digit year (e.g., 1998) for which cost estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring, and backfilling.
Structural Materials Costs The estimated cost of materials such as gravel, pavement and vegetation
necessary for the installation of the hydrodynamic device. These costs should
include installation costs but exclude the cost of the device itself.
Device Construction Costs The estimated cost for supply, construction, and installation of the
hydrodynamic device, including site survey and construction activities.
Structural Control Devices The estimated cost of establishing all hydrodynamic device control devices,
osts such as inlet and outlet structures (manholes), spillways, pipelines and
culverts. Include the cost of materials and construction.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative Costs: |
Average Annual Sediment Estimated average annual cost to remove oils, sediments, and trash
Removal Costs accumulated in the hydrodynamic device.
4 - National Stormwater BMP Database requirement for all Hydrodynamic Devices
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 92 April 25, 2002
-------
Form M
HYDRODYNAMIC DEVICE DESIGN DATA
Test Site Name
Watershed Name BMP Name
Design Information
Volume of Permanent Pool
Permanent Pool Surface Area Permanent Pool Length
Water Quality Surcharge Detention Volume When Full
Brim-full Emptying Time for Surcharge Detention Volume
Half Brim-full Emptying Time for Surcharge Detention Volume
Forebay Volume
Comments
Hydrodynamlc Device Construction Coซt Estimates
Year of Cost Estimate
Construction Costs:
Excavation Structural Materials Cost
Basin Construction Structural Control Devices
Engineering and Overhead Land Costs and Value
Rehabilitative Costa:
Average Annual Sediment Removal
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stonmvater BMP Database Requirements
1 93 April 25, 2002
-------
Wetland Basin Design Data
The wetlands basin form provides important design information specific to wetland basins. A
wetland basin is a BMP similar to a retention pond (with a permanent pool of water) with more
than 50% of its surface covered by emergent wetland vegetation, or similar to a detention basin
(no significant permanent pool of water) with most of its bottom covered with wetland
vegetation. The wetland basin data form and data elements list are shown in Table 3.24 and
Form N.
Table 3.24: Wetland basin form data elements list
Data Element
Description
Volume of permanent
DOOl4
ฐermanent Pool Surface
<\rea4
ฐermanent Pool Length4
lA/ater Quality Surcharge
Detention Volume When
-ull4
lA/ater Quality Surcharge
Surface Area When Full4
i/Vater Quality Surcharge
Volume of the permanent pool of water, if any.
Area of the water surface in the permanent pool, if any.
Length of the permanent pool of water, measured at the water surface along
the axis of the inflow and outflow. If more that one inflow point, use the
average distance between the inflow points and the outflow weighted by the
tributary impervious area.
The water quality surcharge detention volume above the permanent volume
(when full).
The surface area of any supplementary water quality detention volume above
the permanent pool, if applicable.
Length of the water quality detention volume, measured at the water surface
along the axis of the inflow and outflow. If more that one inflow point, use the
average distance between the inflow points and the outflow weighted by the
tributary impervious area.
Time (in hours) required for the wetland basins water quality surcharge
detention volume to be released to the permanent pool.
Time (in hours) required for the lower half of the water quality surcharge
detention volume to be released to the permanent pool.
Volume of the forebay portion of the wetland basin when it is filled to the point
of overflow into the rest of the basin.
Surface area of water in the forebay when it is filled to the point of overflow
into the rest of the basin.
Describe Vegetation CoverDescribe the types of vegetation on the basin sides and floor.
lA/ithin Basin 4
rlood Control Volume, If The volume of the flood control detention volume above the wetland basin
volume.
List the flood return period (in years) for which the flood control volume is
designed (e.g., 25-year).
Surface area of the wetland basin, including all pond areas and meadow
wetland areas. Use permanent pool surface area if no other wetland area
exists adjacent to the pool.
Basin Length
Brim-full Emptying Time
ror Surcharge 4
Half Brim-full Emptying
fime For Surcharge 4
rorebay Volume 4
rorebay Surface Area 4
Design Flood Return
ฐeriods 4
i/Vetland Surface Area
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
194
April 25, 2002
-------
Data Element [Description'
Percent of Wetland Pond Percent of the wetland basin's surface area typically having 12 inches (0.3 m)
with 12 inches (0.3 m) or less water depth.
Depth 4
Percent of Wetland Pond Percent of the wetland basin's surface area typically having 12 to 24 inches
with 12 - 24" (0.3 - 0.6 m) (0.3 - 0.6 m) water depth.
Depth 4
Percent of Wetland Pond Percent of the wetland basin's surface area typically having 24 to 48 inches
with 24 - 48" (0.6 -1.3m) (0.6 -1.3m) water depth.
Depth 4
Percent of Wetland Pond Percent of the wetland basin's surface area typically having greater than 48
with > 48" (> 1.3 m) Depth inches (> 1.3 m) water depth.
Percent of wetland basin's Percent of the wetland basin that is meadow area, that is, area without
area that is meadow standing water.
wetland 4
List All Known Plant Type and percent cover of the wetland basin by each wetland species, and
Species in the Wetland 4 densities.
Wetland Basin Construction Cost Estimates
Year of Cost Estimate Four-digit year (e.g., 1998) for which the above estimates were made.
Construction Costs: |
Excavation Costs The estimated cost of all excavation-related activities, including stripping,
drilling and blasting, trenching and shoring.
Structural Materials Costs The estimated cost of materials used in the wetland basin, such as imported
topsoil or fill.
Basin Construction Costs The estimated cost of establishing the wetland basin itself, not including
vegetation costs.
Structural Control Devices The estimated cost of establishing all wetland basin control devices, such as
Costs inlet and outlet devices, trash racks, etc. Include the cost of materials and
construction.
Vegetation and The estimated cost of establishing vegetation for the BMP, including acquiring
Landscaping Costs landscape materials, establishing vegetation, and establishing the irrigation
infrastructure, if any.
Engineering and Overhead The estimated engineering and associated overhead costs, including site,
osts structural, and landscape design and engineering expenses.
Land Costs or Values The estimated value of the land dedicated to this BMP or the cost of acquiring
this land.
Rehabilitative Costs:
Average Annual Sediment Estimated average annual cost to remove sediment accumulated in the
Removal Costs wetland basin.
Average Annual Estimated average annual cost to revegetate the sides and floor of the
Revegetation Costs wetland basin.
National Stormwater BMP Database requirement for all Wetland Basins
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
195 April 25, 2002
-------
Form N
WETLAND BASIN DESIGN DATA
Test Site Name
Watershed Name
BMP Name
De$lgn Information
Volume of Permanent Pool
Permanent Pool Surface Area
Permanent Pool Length
Water Quality Surcharge Detention Volume
Water Quality Surcharge Surface Area, _
Water Quality Surcharge Basin Length, When Full
Brim-full Emptying Time
Forebay Volume
Half Brim-full Emptying Time
Forebay Surface Area
Flood Control Volume
Wetland Surface Area
Design Flood Return Periods
% of Pond 12" (0.3m) Deep Depth % of Pond with 12" - 24" (0.3-0.6m) Depth
% of Pond with 24" to 48" (0.6-1.3 m) Depth % of Pond with >48" (1.3m) Depth
% of Wetland Basin Area Without Standing Water
Plant Species in the Wetland
Wetland Basin Construction Cost Estimates
Year of Cost Estimate
Construction Costs:
Excavation
Basin Construction
Vegetation and Landscaping
Structural Materials
Structural Control Devices
Engineering and Overhead
Sheet 1 of 2
Urban Stormwater BMP Performance Afoni taring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 96 April 25, 2002
-------
WETLAND BASIN DESIGN DATA
Land Costs and Value
Rehabilitative Costs:
Average Annual Sediment Removal
Average Annual Revegetatlon
Sheet 2 of 2
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
1 97 April 25, 2002
-------
Monitoring Station Information
Monitoring station information is requested for both structural and non-structural BMPs in a test
site. The monitoring station information form contains information on monitoring station
locations, instrumentation, and monitoring costs. More than one instrument may be present in a
monitoring station. For example, a monitoring station may contain a flow gauge and a water
quality sampler. A single form should be filled out for each individual monitoring station at the
site. The monitoring station form and data elements list are provided in Table 3.25 and Form O.
Table 3.25: Monitoring station form data elements
Data Element
Description
Monitoring Station Information
Monitoring Station Name
dentify Upstream BMP1
dentify Relationship to
Upstream BMP1
dentify Downstream BMP1
dentify Relationship to
Downstream BMP1
User-defined name for subject monitoring station.
BMP upstream of the monitoring point (if any).
Identify Relationship to Upstream BMP. These may include inflow, outflow,
bypass, intermediate or not applicable.
BMP downstream of the monitoring point (if any).
Identify Relationship to Downstream BMP. These may include inflow,
outflow, bypass, intermediate or not applicable.
Site Monitoring Instrumentation
Select monitoring station
where instrument is located1
What date was the
nstrument installed?
What type of instrument is in
place?
A monitoring station that contains the instrument must be selected or defined
before entering data on specific instruments.
Provide the date (month, day and 4-digit year) the instrument was installed
(e.g., 6/1/1998).
The instrument type at the monitoring station. These may include a Bubble
Gauge, Digital Recorder, Graphic Recorder, Land Line Telemetered, Radio
Telemetered, Satellite Relayed, ADHAS, Crest Stage Indicator, Tide Gauge,
Deflection Meter, Stilling Well, CR Type Recorder, Weighing Rain Gauge,
Tipping Bucket Rain Gauge, Acoustic Velocity Meter, or Electromagnetic
Flow Meter, Pressure Transducer, Unknown or Other.
The type of data collected by the instrument based on U.S. Geological
Survey (USGS) code. Data types may include: Tide, Water Flow/Stage
Continuous, Water Flow/Stage Intermittent, Water Quality Continuous, Water
Quality Grab, Precipitation Continuous, Precipitation Intermittent, Evaporation
Continuous, Evaporation Intermittent, Wind Velocity Continuous, Wind
Velocity Intermittent, Tide Stage Continuous, Tide Stage Intermittent, Water
Quality Probe Continuous, Water Quality Probe Intermittent, Unknown, or
Other.
What type of control structure Type of control structure in place at the monitoring station (i.e. 90-degree V-
s in place, if any? notched weir, etc.).
Additional Comments May be necessary to explain special features associated with the instrument
or other information deemed important to the user.
What type of monitoring is
conducted?
Site Monitoring Costs
Number of years in which
monitoring was conducted
The number of years over which the monitoring station was in operation
(Table continued on the following page)
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
198
Apnl 25, 2002
-------
Data Element
Description
lomments
Fixed Monitoring Station
osts
Temporary Monitoring
Station Costs
Vear of Cost Basis
Equipment Costs
Maintenance Costs
Sampling Costs
.aboratory Costs
May be needed to clarify unusual monitoring costs or other details as
deemed appropriate by the user.
Those costs associated with fixed monitoring instrumentation installed for
long-term use. For example, a shed may be constructed to house the
instrumentation. Year of cost basis, equipment, maintenance, sampling and
laboratory costs are requested for fixed monitoring stations.
Costs associated with temporary monitoring instruments not intended for
long-term use. Year of cost basis, equipment, sampling and laboratory costs
are requested for temporary monitoring stations.
Year that the monitoring activities were conducted or equipment purchased.
Costs of sampling and flow gauging equipment (rental or purchase) and
installation in U.S. currency.
Annual maintenance costs for equipment in U.S. currency.
Annual costs of sampling in U.S. currency.
Annual costs of sample analysis by a laboratory.
National Stonnwater BMP Database requirement for all BMPs
Urban Stonnwater BMP Performance A foni taring
A Guidance Manual for Meeting the National Stonnwater BMP Database Requirements
199
April 25. 2002
-------
FormO
MONITORING STATION INFORMATION
Site Name
BMP Name
Monitoring Station Information
Monitoring Station Name
Upstream BMP Name
Downstream BMP
Relationship to Upstream BMP
Relationship to Downstream BMP
Site Monitoring Instrumentation
Date of Installation
Instrument Type
Data Type
Type of Control Structure
Additional Comments
Site Monitoring Costs
Number of Years Monitoring Conducted
Year of Cost Baste (Fixed Station)
Equipment Costs
Maintenance Costs
Year of Cost Basis (Temporary Station)
Equipment Costs
Laboratory Cost
Comments
Sampling Costs
Laboratory Cost
Sampling Costs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
200
April 25, 2002
-------
Precipitation Data
The precipitation form contains important precipitation data, which can be used for evaluating
the performance of BMPs under various conditions. Precipitation information requested includes
data such as time and date that the event began and ended, total depth and one-hour peak
precipitation rate. The precipitation data form and data elements list are provided in Table 3.26
and Form P.
Table 3.26: Precipitation Form Data Elements
Data Element
Description
Event ID
Select Monitoring Station
For Event1
Start Date
Start Time
End Date
End Time
Total Storm Precipitation
Peak One Hour
Precipitation Rate
User provided name or identifier for the precipitation event.
Monitoring station name where the precipitation event was monitored.
Calendar date (month, day and 4-digit year) that storm started (e.g.,
01/01/1998).
Time that the storm started, e.g., 21:00. If only storm duration is available,
record 00:00 for start time and enter the storm duration for end time.
Calendar date (month, day and 4-digit year) that storm ended (e.g.,
01/01/1998). Use six hours as the separation criteria to define a new storm.
Time that the storm ended, e.g., 13:21. If only storm duration is available,
record 00:00 for start time and enter the storm duration for end time.
Amount of precipitation that occurred during the storm. For example, a total of
4 inches of rain fell during a 12-hour storm.
The most intense one-hour of rainfall for the storm. For storms with less than
one-hour duration, divide the storm rainfall depth by one hour.
National Stormwater BMP Database requirement for all BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
201
April 25, 2002
-------
Form?
WATER QUALITY INFORMATION
BMP Test Site
Watershed
Monitoring Station
Station Type
Sample Type
QA/QC Description
Number of Samples, II Composite
Comments
Sample ID Sample Date Sample Time Related STORET Parameter Value Qualifier Analysis
Flow Event Method
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
202
April 25, 2002
-------
Flow Data
The flow data form provides on-site stormwater runoff information. Accurate flow data coupled
with water quality information can be used to estimate removal efficiencies for BMPs, providing
a relative measure of a BMP's ability to remove certain pollutants. The flow data form contains
information on the date and time of the beginning and end of the flow event and total flow
volumes and peak flow rates for runoff and baseflow. Each flow event should have a related
precipitation event recorded on the precipitation form. The flow form and data elements list are
provided in Table 3.27 and Form Q.
Table 3.27: Flow form data elements
Data Element
Description
Monitoring Station Provide monitoring station where flow event was monitored.
Select the type of flow1 The type of flow: base flow or storm runoff.
f storm runoff, select the The start-date of the precipitation event associated with the current flow event.
elated precipitation event,
f available1
Flow Start Date1 Date (month, day and 4-digit year) that the measurement began being taken
(e.g., 01/01/1998).
Flow Start Time Time at beginning of measurement event, e.g., 23:30. If only flow duration is
provided, enter 00:00 for start time and enter the flow duration for end time.
Flow End Date Date (month, day and 4-digit year) that the measurement event ended (e.g.,
01/01/1998). The end of runoff event can be defined as that point in time
when the recession limb of the hydrograph is <2% of the peak or is within 10%
of the pre-storm base flow, whichever is greater.
Flow End Time Time at the end of the measurement event, e.g., 01:30. The end of runoff
event can be defined as that point in time when the recession limb of the
hydrograph is <2% of the peak or is within 10% of the pre-storm base flow,
whichever is greater.
Total Storm Flow Volume Total Runoff Volume minus the Bypass Volume.
nto or from BMP1
Peak Storm Flow Rate into Greatest rate of storm flow into or from the BMP.
or from BMP
Total Bypass Volume, if Total Runoff Volume minus the Runoff Volume Influent to the BMP.
any1
Peak Bypass Flow Rate, if Peak rate of flow measured for flows bypassing the BMP.
any
Dry Weather Base Flow Flow rate during dry-weather conditions. Base flow is collected during non-
Rate1 wet weather conditions.
1 - National Stormwater BMP Database requirement for all BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
203 April 25, 2002
-------
Form Q
FLOW INFORMATION |
BMP Name
Watershed Nam*
Monitoring Station Name
Row Start Flow Start Type of Flow: Related Flow End Flow End Total Flow Peak Flow Total Peek Dry Whether
Date Time Runoff or Base Precipitation Date Time Volume Rate Bypass Bypass Base Flow
Flow Event Flow Rate Rate
Urban Stornrwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Storrnwater BMP Database Requirements
204
April 25, 2002
-------
Water Quality Data
The water quality sampling event form provides the general information for a water quality
sampling event such as date, time, location, and QA/QC measures used for a study. Provided
water quality information must have associated flow and precipitation information recorded on
the precipitation and flow forms. The water quality data form and data elements list are provided
in Table 3.28 and Form R.
Table 3.28: Water quality form data elements
Data Element
Description
Sample ID
Select Monitoring Station
Where Data Collected1
Related Flow Event1
Date Water Quality
Sample Collected
Time Water Quality
Sample Collected
What medium does the
nstrument monitor? 1
User provided name or identifier for the water quality sample.
Monitoring station name where the data was collected.
Flow event associated with the water quality sampling event.
Date that the water quality sample began being collected.
Time that the water quality sample began being collected.
Groundwater, Surface Runoff/Flow, Soil, Dry Atmospheric Fallout, Wet
Atmospheric Fallout, Pond/Lake Water, Accumulated Bottom Sediment,
Biological, or Other.
collected?
What type of samples are The type of samples that the instrument collects, including: Flow Weighted
Composite EMCs (Event Mean Concentrations), Time Weighted Composite
EMCs, Unweighted (mixed) Composite EMCs, or Grab Sample.
The number of samples collected or mixed (if composite).
Provide the Number of
Samples, If Composite
Describe Quality
Assurance/Quality Control
Measures in Place for the
Sampling Event
Provide Additional
omments, If Needed
Water Quality Parameter
(STORET)1
Value1
Unit1
Qualifier1
Analysis Method
Describe the types of Quality Assurance/Quality Control (QA/QC) measures in
place for both laboratories and field activities.
Discuss special circumstances associated with the sampling event.
The STORET name for the U.S. Environmental Protection Agency's STORET
water quality database for streams and other waterbodies throughout the
United States.
Value of the measured constituent should be provided. If the value is below
detection limits, provide the reported detection limit with a "U" qualifier in the
qualifier field and place a minus sign in front of the value.
Unit of the measured constituent should be provided.
Numerical STORET qualifier associated with a data point.
Analysis Method should be provided for the constituent. For example EPA
8270 or Standard Method 513.
1 - National Stormwater BMP Database requirement for all BMPs
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
205
April 25, 2002
-------
FormR
WATER QUALITY INFORMATION
BMP Test Site
Watershed
Monitoring Station
Station Type
Sample Type
OA/QC Description
Number of Samples, If Composite
Comments
Sample Date Sample Time Related STORET Parameter Value Qualifier Analysis
Flow Event Method
Urban Stormwater BAfP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
206 April 25, 2002
-------
3.4.3.3 On-line Information
Forms and field descriptions can be printed from the world-wide-web at
www.bmpdatabase.org. Each set of forms is subcategorized into its subsequent BMP
type. Each folder contains all of the necessary forms and information needed for
monitoring and reporting for a particular BMP type. BMP categories include:
Non-Structural BMPs.
Detention Basins.
Retention Ponds.
Percolation Trenches and Dry
Wells.
Media Filters.
Grass Filter Strips.
Wetland Channels and Swales.
Porous Pavement.
Infiltration Basins.
Hydrodynamic Devices.
Wetland Basins.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BA<1P Database Requirements
207 April 25, 2002
-------
References
American Society for Testing and Materials. 1995. Standard Guide for Selection of Weirs
and Flumes for Open-Channel Flow Measurement of Water (D5640). American
Society for Testing and Materials, West Conshohocken, PA.
American Society for Testing and Materials. 1997. Standard Guide For Monitoring
Sediment in Watersheds. ASTM D 6145-97.
American Society for Testing and Materials. 2000. Standard Test Method for
Determining Sediment Concentration in Water Samples. ASTM Designation D
3977-97, 395- 400.
ASCE. 1999. National Stormwater Best Management Practices Database, Verisionl.
American Society of Civil Engineers, Urban Water Resource Research Council,
June.
ASCE. 2000. Data Evaluation Report, Task 3.4. National Stormwater Best Management
Practices Database Project. June.
Burton, A.G. and R. Pitt. 2001. Stomwater Effects Handbook: A Toolbox for Watershed
Managers, Scientists, and Engineers. Lewis Publishers, Boca Raton, FL.
Caltrans. 1997. Guidance Manual: Stormwater Monitoring Protocols. Prepared by Larry
Walker and Associates and Woodward-Clyde Consultants. August 15.
Caltrans. 2000. California Department of Transportation District 7 Litter Management
Pilot Study, Final Report, Caltrans Document No. CT-SW-RT-00-013, June 26.
Cave, K. A., and L.A. Roesner. 1994. Overview of Stormwater Monitoring Needs.
Proceedings of the Engineering Foundation Conference on Stormwater Monitoring.
August 7-12, Crested Butte, CO.
Chow, V.T. 1954. The lognormal distribution and its engineering applications. Proc.,
American Society of Civil Engineers. 110, 607-628.
Chow, V.T. 1959. Open-Channel Hydraulics. McGraw-Hill, New York, NY.
Coffey, S.W., and M.D. Smolen. 1990. The Nonpoint Source Manager's Guide to Water
Quality Monitoring - Draft. Developed under EPA Grant Number T-9010662.
U.S. Environmental Protection Agency, Water Management Division, Region 7,
Kansas City, MO.
Cochran, W.G. 1963. Sampling Techniques. Second Edition. John Wiley and Sons, Inc.
New York, New York.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
208 April 25, 2002
-------
Driscoll, E., G. Palhegyi, E. Strecker, and P. Shelley. 1989. Analysis of Storm Event
Characteristics for Selected Rainfall Gauges Throughout the United States. Draft
Rep. Prepared for U.S. Environmental Protection Agency, Woodward-Clyde
Consultants, Oakland, CA.
Driscoll, E.D., P.E. Shelly, and E.W. Strecker. 1990. Pollutant Loadings and Impacts from
Stormwater Runoff, Volume III: Analytical Investigation and Re search Report.
Federal Highway Administration Final Report FHWA-RD-88-008, 160 p.
EPA. 1983. Results of the Nationwide Urban Runoff Program, volume I Final Report.
U.S. Environmental Protection Agency, Water Planning Division, Washington
D.C.
EPA. 1989. Analysis of Storm Event Characteristics for Selected Gages throughout the
United States. Environmental Protection Agency. Washington D.C.
EPA. 1993 a. Guidance Specifying Management Measures for Sources of Nonpoint
Pollution in Coastal Waters. EPA-840-B-93- 00Ic. U.S. Environmental
Protection Agency, Office of Water, Washington D.C.
EPA. 1993b. Investigation of Inappropriate Entries into Storm Drainage Systems: A
User's Guide. EPA-600-R-92-238. U.S. Environmental Protection Agency,
Office of Research and Development, Washington D.C.
EPA, 1993c. Memo. Office of Water Policy and Technology. Guidance on Interpretation
and Implementation of Aquatic Life Metals Criteria, Washington, D.C.
EPA. 1994a. Laboratory Data Validation: Functional Guidelines for Evaluating
Inorganics Analyses. U.S. Environmental Protection Agency, Data Review
Workgroup, Washington, D.C.
EPA. 1994b. Laboratory Data Validation: Functional Guidelines for Evaluating
Organics Analyses. U.S. Environmental Protection Agency, Data Review
Workgroup, Washington, D.C.
EPA. 1994c. Guidance for the Data Quality Objectives Process. EPA-QA/G-4. U.S.
Environmental Protection Agency.
EPA. 1997. Monitoring Guidance for Determining the Effectiveness of Nonpoint Source
Controls. EPA 841-B-96-004. U.S. Environmental Protection Agency, Office of
Water, Washington, D.C. September.
FHWA. 2000. Stormwater Best Management Practices in an Ultra-Urban Setting:
Selection and Monitoring. FHWA-EP-00-002. U.S. Department of
Transportation. Federal Highway Administration. Washington, D.C., May.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
209 April 25, 2002
-------
FHWA. 2000. Guidance Manual for Monitoring Highway Runoff Water Quality.
Federal Highway Administration, U.S. Department of Transportation. Federal
Highway Administration. Washington, D.C.
Fischer, H. B., List, J. E., KOH, R. C. Y., Imberger, I, and N.H. Brooks. 1979. Mixing in
Inland and Coastal Waters. Academic Press Inc., San Diego, CA.
Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring. John
Wiley and Sons, Inc. New York, New York.
Grant, D.M. and B.D. Dawson. 1997. ISCO Open Channel Flow Measurement Handbook.
ISCO Environmental Division. Lincoln, NE.
Gray, J.R., G.D. Glysson, and L.M. Turcios. 2000. Comparability and reliability of total
suspended solids and suspended-sediment concentration data. U.S. Geological
Survey Water-Resources Investigations Report 00-4191. U.S. Geological Survey.
Green, D., T. Grizzard, and C. Randall. 1994. Monitoring of Wetlands, Wet Ponds, and
Grassed Swales. In Proceedings of the Engineering Foundation Conference on
Storm Water Monitoring. August 7-12, Crested Butte, CO.
Gupta, R.S. 1989. Hydrology and Hydraulic Systems. Prentice Hall. NJ.
Gwinn, W.R., and D.A. Parsons. 1976. Discharge Equations for HS, H and HL Flumes.
Journal of the Hydraulics Division American Society of Civil Engineers, Vol. 102,
No. HY1 pp. 73-89.
Harremoes, P. 1988. Stochastic models for estimations of extreme pollution from urban
runoff. Water Resource Bulletin, 22, 1017-1026.
Harrison, D. 1994. Policy and Institutional Issues of NPDES Monitoring: Local Municipal
Perspectives of Stormwater Monitoring. Proceedings of the Engineering Foundation
Conference on Storm Water Monitoring. August 7-12, Crested Butte, CO.
James, W. 2001. http://131.104.80.10/webfiles/james/KevModels.html.
Loftis, J.C., L.H. MacDonald, S. Streett, H.K. Iyer, and K. Bunte. 2001. Detecting
Cumulative Watershed Effects: The Power of Pairing. In Review Journal of
Hydrology.
Martin, E.H., and J.L. Smoot. 1986. Constituent-load changes in urban storm water
runoff routed through a detention pond-wetland system in central Florida. Water
Resources Investigation Report 85-4310. U.S. Geological Survey.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BA<1P Database Requirements
210 April 25, 2002
-------
Martin, G.R., J.L. Smoot, and K.D. White. 1992. A Comparison of Surface-Grab and
Cross Sectionally Integrated Stream-Water-Quality Sampling Methods. Water
Environment Research. November/December 1992.
Minton, G.R. 1998. Stormwater Treatment Northwest, Vol. 4, No. 3, August.
National Urban Runoff Program. 1983. Final Report. U.S. Environmental Protection
Agency Water Planning Division, Washington, D.C.
Oswald, G.E., and R. Mattison. 1994. Protocols for Monitoring the Effectiveness of
Structural Stormwater Treatment Devices. In Proceedings of the Engineering
Foundation Conference on Storm Water Monitoring. August 7-12, Crested Butte,
CO.
Pitt, R. 1979. Demonstration ofNonpoint Pollution Abatement Through Improved Street
Cleaning Practices, EPA-600/2-79-161, U.S. Environmental Protection Agency,
Cincinnati, Ohio.
Pitt, R. 2001. Personal Correspondence. Peer review meeting, August.
Pitt, R. and K. Farmer. 1995. Quality Assurance Project Plan (QAPP) for EPA
Sponsored Study on Control of Stormwater Toxicants. Department of Civil and
Environmental Engineering, University of Alabama at Birmingham.
Ponce, V.M. 1989. Engineering Hydrology Principles and Practices. Prentice Hall,
Englewood Cliffs, NJ.
Ruthven, D.M. 1971. "The Dispersion of a Decaying Effluent Discharged Continuously
into a Uniformly Flowing Stream," Water Research, 5(6) 343-352.
Quigley, M.M, and E.W. Strecker. 2000. Protocols and methods for evaluating the
performance of Stormwater source controls. NATO Advanced Research Workshop
on Source Control Measures for Stormwater Runoff, St. Marienthal, Ostritz,
Germany.
Saunders, T.G. 1983. Design of Networks for Monitoring Water Quality. Water
Resources Publications. December.
Schueler, T.R. 1987. Controlling Urban Runoff: A Practical Manual for Planning and
Designing Urban BMPs. Washington Metropolitan Water Resources Planning
Board, Washington, D.C.
Schueler, T. 1996. Irreducible Pollutant Concentrations Discharged from Urban BMPs.
Watershed Protection Techniques, 1(3): 100-111. Watershed Protection
Techniques 2(2): 361-363.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
211 April 25, 2002
-------
Schueler. 2000. National Pollutant Removal Performance Database: for Stormwater
Treatment Practices, Second Edition. Center for Watershed Protection. June.
Spitzer, DW (ed.) 1996. Flow Measurement Practical Guides for Measurement and
Control. Instrument Society of America. NC.
Strecker, E.W., J.M. Kersnar, E.D. Driscoll, and R.R. Horner. 1992. The use of wetlands
for controlling storm water pollution. The Terrene Institiute, Washington, D.C.
Strecker, E.W. 1994. Constituents and Methods for Assessing BMPs. In Proceedings of
the Engineering Foundation Conference on Storm Water Monitoring. August 7-
12, Crested Butte, CO.
Strecker, E.W., M.M. Quigley, B.R. Urbonas, I.E. Jones, and J.K. Clary. 2001.
Determining Urban Storm Water BMP Effectiveness. Journal of Water
Resources Planning and Management, 127(3), 144-149.
Stenstrom, M.K. and E.W. Strecker. 1993. Assessment of Storm Drain Sources of
Contaminants to Santa Monica Bay. Volume II. UCLA School of Engineering and
Applied Science. UCLAENG 93-63.
Stevens Water Resources Data Book. 1998. Leupold & Stevens, Inc. Beaverton, OR.
Taylor, J.R. 1997. An Introduction to Error Analysis. University Science Books,
Sausalito, CA.
Urbonas, B.R. 1994. Parameters to Report with BMP Monitoring Data. In Proceedings
of the Engineering Foundation Conference on Storm Water Monitoring. August
7-12, Crested Butte, CO.
Urbonas, B., and Stahre, P., 1993. Stormwater: Best management practices and
detention for water quality, drainage, and CSO management., Prentice Hall,
Englewood Cliffs, NJ.
USGS. 1980. National Handbook of Recommended Methods for Water-Data Acquisition.
U.S. Geological Survey. Reston, VA.
USGS. 1985. Development and Testing of Highway Storm-Sewer Flow Measurement
and Recording System. Water-Resources Investigations Report 85- 4111. U. S.
Geological Survey. Reston, VA.
USGS. 2001. A Synopsis of Technical Issues for Monitoring Sediment in Highway
and Urban Runoff. Open-File Report 00-497. U.S. Geological Survey,
Northborough, MA.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
212 April 25, 2002
-------
Van Buren M.A., W.E. Watt, and J. Marsalek. 1997. Applications of the Log-normal and
Normal Distributions to Stormwater Quality Parameters. Wat. Res., Vol. 31, No 1,
pp. 95-104.
Waschbusch, R., and D. Owens. 1998. Comparison of Flow Estimation Methodologies in
Storm Sewers. United States Geological Survey, January 16.
Watt, W.E., K.W. Lathem, C.R. Neill, T.L. Richards, and J. Rouselle, (Eds). 1989.
Hydrology of Floods in Canada: A Guide to Planning and Design. National
Research Council Canada, Ottawa, Ontario.
WCC. 1993. Data from storm monitored between May 1991 and January 1993. Final
Data Rep. Woodward-Clyde Consultants. Prepared for Bureau of Environmental
Services, City of Portland, Oregon; Portland, OR.
WCC. 1995. Stormwater Quality Monitoring Guidance Manual. Woodward-Clyde
Consultants. Prepared for Washington State Department of Ecology Water Quality
Program. November.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
213 April 25, 2002
-------
INDEX
A
Accuracy, 52, 61, 62, 64, 65, 82, 92, 93,
94, 100, 106, 110, 120, 126, 129, 135,
136, 141, 144, 1
Achievable efficiency, 32, 34, 35, 37
Agriculture, 155, 159
Analysis,
of variance, 26, 41, 70, 72, 74, 145
sample, 119, 125
standards violations, 11
Analytical, 12, 13, 45, 46, 48, 76, 77, 79,
80, 81, 82, 121, 123, 125, 128, 129,
137, 142, 143, 144, 145, 146, 209
ASCE, 33, 48, 77, 137, 139, 147, 208, 1
Assumptions, 23, 24, 26, 29, 31
Atmospheric deposition, 63
B
Bias, 81, 109, 112, 126, 140
BMP
effectiveness, 12, 40, 45, 52, 68, 96,
156, 166, 167
efficiency, 2, 4, 8, 12, 14, 15, 16, 17,
18,21,32,33,34,40,61,71
monitoring, 2, 4, 5, 6, 8, 18, 21, 37,
43, 46, 47, 48, 49, 52, 53, 55, 58,
61, 63, 70, 72, 77, 80, 83, 102, 111,
115, 122, 123, 124, 127, 132, 147,
156
non-structural, 149, 166
structural, 149, 161, 164, 166, 167,
170, 207
systems, 44
types, 13, 33, 164, 166
Calibration, 107, 118, 141
Central tendency, 41
Chi-square, 41
Clean Water Act, 4, 11
Climate, 16, 65, 148
Coefficient of variation, 22, 70, 71, 75,
76, 145
Completeness, 129, 137
Compliance, 6, 49, 51
Concentrations, 17, 18, 205, 211
Construction, 7, 100, 155, 159, 166, 170,
173, 176, 179, 181, 183, 187, 190,
192, 195
Cost, 6, 100, 167, 169, 173, 175, 178,
181, 183, 187, 189, 192, 195, 199
CZARA, 4, 5
Data,
water quality, 22, 58, 69, 130, 149,
205
Data Logger, 83, 84, 85, 88
Data management, 45, 48
Data Quality Objectives, 129, 209
Design flow, 181
Detention Basin, 150, 169, 207
Dry weather flow, 18
Effectiveness, 6, 10, 209, 211, 212
Efficiency, 13, 15, 18, 20, 21, 24, 25, 28,
29, 30, 32, 33, 34, 35, 37, 38, 39
Efficiency Ratio (ER), 20, 21, 24, 31,
32,33,34,35,38,39,41
Effluent, 18, 20, 33, 34, 35, 36, 40, 211
EPA, 4, 5, 7, 11, 12, 22, 31, 48, 49, 50,
65, 66, 68, 70, 71, 74, 79, 81, 82, 113,
117, 128, 134, 135, 136, 137, 139,
144, 145, 147, 156, 159, 205, 208,
209,211, 1
Erosion, 164, 166
Error, 28, 29, 126, 130, 212, 2, 5, 7
Evaluation, 2, 17, 37, 41, 51, 52, 67, 78,
122, 144, 145
Event mean concentration, 16, 17, 18,
21, 22, 23, 24, 36, 70, 72
FHWA, 77, 118, 119, 120, 209, 210
Field blanks, 133, 135
Field duplicate samples, 135
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
214 April 25, 2002
-------
Field operations, 134
Filter, 150, 151, 178, 179, 181,207
Flow Measurement, 90, 96, 99, 102, 208,
210,212
Flow meter, 141
Geology, 64
Grass filter strip, 181
Great Lakes, 64
H
Habitat, 6
Human Health, 50
Hydrodynamic device, 192
Hypothesis testing, 72, 145, 147
I
Implementation, 6, 139, 209
Infiltration Basin, 150, 151, 189, 190,
207
Inflow, 183
In-situ, 119, 121, 123
Irreducible concentration, 33
K
Kolmogorov-Smirnov, 41
Lakes, 64
Load, 10, 11, 18,31
M
Management, 2, 3, 10, 15, 16, 18, 48, 65,
66, 137, 138, 146, 147, 208, 209, 212
Mean concentration, 21
Methods
Effluent Probability, 18, 20, 34, 40
Reference Watershed, 43
tracer dilution, 95
Model
Non-Linear, 20, 40
Models, 20, 40, 87
Monitoring, 4, 5, 8, 9, 10, 11, 12, 14, 24,
43, 45, 46, 47, 49, 51, 53, 55, 56, 57,
59, 60, 61, 62, 68, 76, 83, 92, 102,
109, 114, 124, 139, 141, 146, 147,
148, 149, 157, 198, 199, 201, 203,
205, 208, 209, 210, 211, 212, 213
Monitoring station, 59, 60, 61, 62, 146,
198,201,205
N
National Stormwater BMP Database, 2,
147, 148, 149, 150, 155, 156, 157,
161, 164, 167, 170, 173, 176, 179,
181, 184, 187, 190, 192, 195, 199,
201, 203, 205
Nonpoint, 4, 114, 208, 209, 211
NURP, 12, 13, 77
O
Overland flow sampler, 119
Percent removal, 39
Pesticides, 17, 79
Ponds, 150, 173,207,210
Porous Pavement, 150, 152, 186, 187,
207
Q
Quality assurance and quality control,
129, 130, 131, 134, 135, 136, 137,
143, 144, 205
Quality assurance project plan, 211
Range, 99, 128, 157
Regression, 20, 21, 25, 27, 32
Regression of loads (ROL), 20, 21
Retention Pond, 150, 152, 172, 173, 207
Safety, 6, 131
Sample, 111, 113, 125, 132, 135, 136,
139, 146, 149, 205
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
215 April 25, 2002
-------
composite, 112
grab, 111
Sampling, 13, 17, 58, 63, 77, 82, 96,
114, 115, 124, 125, 132, 149, 199,
205,208,211
automatic, 90, 102
manual, 90, 102, 124, 125
methods, 13, 77
sediment, 63
Sampling Methods, 132, 211
Sensor
ultrasonic depth, 102, 104
Sensors
electromagnetic, 107, 110
Soils, 189
Stage-flow relationships, 98
Statistic
mean, 11, 17, 18, 20, 21, 29, 157, 205
variability, 70
Storm event, 114
Stormwater quality, 7, 8, 51, 69, 145
Stream, 65, 211
Summation of loads, 21
Surface Water, 49
Test, 45, 147, 148, 149, 156, 157, 166,
208
Toxicity, 6, 16
Training, 134, 139
TSS, 16, 17, 24, 25, 27, 28, 29, 32, 33,
34, 35, 38, 43, 64, 79, 80, 82
V
Variability, 70
VOC, 112, 115, 117
W
Water Quality, 11, 36, 41, 49, 50, 51, 64,
138, 141, 152, 154, 172,205
Wetland, ix, 150, 153, 154, 183, 184,
194, 195, 207
Wetland Basin, 150, 154, 194, 195, 207
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
216 April 25, 2002
-------
APPENDIX A
ERROR ANALYSIS
Estimating flow in a pipe or open channel is generally accomplished by measuring two or
more variables and relating them with an equation to calculate the flow. The continuity
equation relates flow to area and velocity:
Q = Axv (A.I)
where,
A: Area
v: Velocity
For a rectangular channel, the cross-sectional area can be calculated as the water depth
multiplied by the width of the channel.
A = Hxw (A.2)
where,
H: Depth
W: Width
Velocity can be directly measured with a mechanical current meter or Doppler
technology. Estimating flow in the rectangular channel requires three measured variables;
each will have an error associated with it:
Q = Hxwxv (A.3)
For depth and width measurements, the accuracy will usually be expressed as absolute
error governed by the tolerance of the measuring device (i.e. measured depth + X cm).
For velocity, the error in measurement will most likely be a relative error expressed as a
percent of the measured value (i.e. measured velocity + X %). The total error in the
calculated flow measurement will include all of the errors associated with the individual
measurements as illustrated in the following example:
Equipment tolerances provided by manufacturers generally are based on laboratory data
under ideal conditions (e.g. steady state, laminar flow), which may not be representative
of installed conditions. A recent USGS study compared several flow monitoring devices
designed specifically for stormwater application, and found the error in the observed
measurements ranged from 12 to 28 percent.
The actual error is most likely somewhat less than the maximum error and mathematical
formulas have been described by Taylor (1997), which describe how error propagates
when variables (with associated errors) are combined.
If variables x; (for 1=1 to n) are measurements with small but known uncertainties 5x; and
are used to calculate some quantity q, then 5x; cause uncertainty in q as follows.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A- 1 April 25, 2002
-------
If q is a function of one variable, q(xi), then
Sq =
If q is the sum and/or difference of x;s then
Sq =
dq
dx.
(A.4)
(for independent random errors) (A. 5)
Estimates of 5q from Equation A.2 are always less than or equal to:
5q = ^8x1
where x; are measured with small uncertainties 5x;.
If q is the product and quotient of x;s then
Sq =
" I fir
(for independent random errors) (A. 6)
Estimates of 5q from Equation A.6 are always less than or equal to:
(A.7)
This approach can be directly applied to the analysis of error propagation. Examples for
applying this method to flow measurement follow.
Relative Error in Flow Versus Relative Error in Head
Errors in flow measurements are most often caused by field conditions that are
inconsistent with the conditions under which rating curves for flow devices were
calibrated. However, even under ideal conditions, errors in flow measurement can be
significant. This section discusses calculations for estimating the theoretical error
associated with flow measurement equipment under ideal circumstances. It can be seen
that errors, particularly in low flow measurements, can be quite large.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-2
April 25, 2002
-------
Equations relating the head (H) measured in a primary device to discharge (Q) (i.e.,
Rating Equations) fall into four general forms:
1) Q = aHd
2) Q = a(H + c)d
3) Q = a(bH + c)d
4) Q = a + blH+b2H2 +b3H3 + --- + bnH"
The first rating equation is a straight forward application of error propagation for a power
function. This equation is
Flow and head can only be positive values and the power for Rating Equation 1 is always
positive (i.e., flow increases proportionally to head, not decreases), thus the absolute
value sign is omitted in the above equation. The relative error in flow equals the relative
error in head multiplied by the exponent d.
Rating Equations 2, 3, and 4 require an equation relating the error in flow to the
derivative of the flow equation and the error in the measured head, which is:
5Q =
dQ
dH
8H (A. 9)
Before applying this equation, the derivatives of Rating Equations 2, 3, and 4 are taken
with respect to H.
For Rating Equation 2:
^- = ad(H + c}d'1 (A. 10)
dH
For Rating Equation 3:
^ =abd(bH+c)d-1 (A. 11)
For Rating Equation 4:
^- = bl+2b2Hl + 3b3H2+--- + nbnH"-1 (A. 12)
dH
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-3 April 25, 2002
-------
Prior to applying the equation to the derivatives of Rating Equations 2, 3, and 4 the
equation is modified by dividing each side of the Equation by the flow (Q). This yields
an equation for the relative error in the flow on the left hand side.
SQ
o
dO
dH
5H_
Q
(A. 13)
Substituting flow Rating Equation 2 for Q and the derivative of Rating Equation 2 for
dQ/dH into the right hand side of the above equation, yields:
CXy i/ TT \ d-} Oti . . ,
^ = ad(H + c)dl (A. 14)
Q a(H + c)
which reduces to:
S d 8H
(A. 15)
Equation A. 1 1 relates the relative error in the flow to the relative error in the head.
A similar analysis for Rating Equation 3 yields:
S d 8H
(A. 16)
Determining an equation for the relative error for Rating Equation 4 is more
cumbersome, but is calculated the same way:
^- = b, + 2b7Hl +3b,H2 +--- + nb H"-1 - - (A. 17)
Q " a + bl 23 "
Rearranging yields:
S b+2bH2+3bH3+--- + nbH" 8H
n
a + bH + bH2+bH3 +--- + bH" H
(A.lo)
Equation A.4, A. 1 1, A. 12, and A. 14 relate the relative error in flow to the relative error in
head for four common equations describing flow through a primary device. While the
equations can be unwieldy, it is a relatively simple exercise to enter them into a
spreadsheet program to estimate the error in flow based on estimated error in head and
other variables. Most primary devices have a relatively simple flow equation that is
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-4 April 25, 2002
-------
sufficiently accurate throughout most of the flow range for the device, which allows for
the use of an error equation related to one of the Rating Equations.
The equations relating the relative error in the estimate of flow to the relative error in the
measurement of head can also be expressed in terms of absolute errors by multiplying
each side of the equations by Q. For example the flow Equation 3 becomes:
d 8H
0
-
m)
H
x a(bH + c]d = abd(bH + c)<" SH
(A. 19)
An Example of Error Analysis for a BMP
The following example illustrates how estimates of error propagation can be applied to
flow measurements. This example assumes a stormwater BMP has two separate sources
of inflow and one outflow. The flow measurement devices and errors are listed in Table
1.
Table A.I: Example of inputs for estimation of errors in flow measurement devices
Station
Inlet 1
Inlet 2
Outlet
Variable
Width
Depth
Velocity
Depth
Depth
Equipment
Tape Measure
Pressure Transducer
Doppler
Bubbler
0.457m (1.5')
Palmer-Bowlus
Flume
Pressure Transducer
45ฐ V notch weir
Measured Value or
formula
3 meters
1.2 meters
0.071 meters/sec
0.12 meters
Q(L/s) =
1076.4(H + 0.005715)L8977
0.70 meters
Q(L/s) = 571.4H2':'
Accuracy
+ 0.025 meters
+ 0.007 meters
+ 4%
+ 0.001 meters
+ 3%
+ 0.007 meters
+ 6%
For Inlet 1, the flow calculation is:
Qmlet_l= 0.2556
The error associated with this measurement can be calculated using the equation for error
of products and quotients (i.e., Equation A.6):
Assuming that the errors are independent and randomly distributed, the relative error in q
equals:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-5
April 25, 2002
-------
q
+(0.04)2
1.2 V '
(51 0.12m
0.12m J
8Q = 0.021m3 /sxO.ll = 0.0023In
Relative error for the flume itself also has to be included. Since the error is a function of
one variable, it can be calculated using Equation A.4:
Sq =
dq
dx
Sx = 0.03 x 0.021 m3/s = 0.00063 m3/s
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-6 April 25, 2002
-------
The total error is therefore the sum of errors associated with the measuring device
(Equation A.5).
Met-2(totan = A/0.00232+0.000632 = 0.0024 m3/s
Qrnlet-2= 0.0210 + 0.0024 M* / S
For the Outlet weir, the flow can be calculated using the following equation:
(9 = 571.4x0.7025 =234.251/5 = 0.234 m3/s
This is also a power function (Rating Equation 1) and the error can be calculated
similarly to the equation for the flume:
SQ= 2.5
0.007
0.70
0.234m3 ls = 0.059w3 Is
The error associated with the weir itself is a single variable as was the flume:
8q = 0.06x0.234w3 Is = 0.014m3 Is
The total error is the sum of the errors associated with the measuring device and is
calculated as follows:
-, = V0.0592+0.0142 = 0.061 m3/S
Results of this error analysis are provided below in Table A. 2.
Table A.2: Summary of examples demonstrating the propagation of errors in flow
measurement
Inlet- 1
Inlet-2
Outlet
Flow (mj/sec)
0.255
0.021
0.234
Total Error (mj/sec)
+ 0.011
+ 0.0024
+_0.061
Total Relative Error
(m3/sec)
4%
11%
26%
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
A-7
April 25, 2002
-------
APPENDIX B
NUMBER OF SAMPLES REQUIRED FOR VARIOUS POWERS, CONFIDENCE
INTERVALS, AND PERCENT DIFFERENCES
The figures in this Appendix are from: R. Pitt and K. Farmer. Quality Assurance Project
Plan (QAPP)for EPA Sponsored Study on Control of Stormwater Toxicants. Department
of Civil and Environmental Engineering, University of Alabama at Birmingham. 1995.
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) 1 April 25, 2002
-------
Number of Sample Pairs Needed
(Power = 0.5 Difference = 10%)
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.8 Difference = 10%)
0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) o April 25, 2002
-------
Number of Sample Pairs Needed
(Power = 0.9 Difference = 10%)
1.0
0.9
0.8
0.6
0.5
2WO
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.5 Difference = 25%)
0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) -y April 25, 2002
-------
Number of Sample Pairs Needed
(Power = 0.8 Difference = 25%)
1.0
0.9
0.8
O
0.7
0.6
0.5
A\\\ V\ V
\
\
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.9 Difference = 25%)
1.0
0.9
8 ฐ-8
I
1
O
ฃ 0.7
0.6
0.5
u
\
to
^"1280
\
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) A April 25, 2002
-------
0.5
Number of Sample Pairs Needed
(Power = 0.5 Difference = 50%)
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
1.0
0.9
0.8
0.6
0.5
Number of Sample Pairs Needed
(Power = 0.8 Difference = 50%)
\
\
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) c April 25, 2002
-------
Number of Sample Pairs Needed
(Power = 0.9 Difference = 50%)
0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.5 Difference = 75%)
1.0
0.9
0.8
o
O
5
-------
Number of Sample Pairs Needed
(Power = 0.8 Difference = 75%)
S
0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.9 Difference = 75%)
1.0
0.9
0.8
-------
Number of Sample Pairs Needed
(Power = 0.5 Difference = 95%)
1.0
0.9
0.8
H0.7
0.6
0.5
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 0.8 Difference = 95%)
1.0
0.9
0.8
ง
o
0.7
0.6
0.5
\
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
T) o April 25, 2002
-------
Number of Sample Pairs Needed
(Power = 0.9 Difference = 95%)
1.0
0.9
0.8
0.6
0.5
\
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Number of Sample Pairs Needed
(Power = 90% Confidence = 95%)
100
0)
en
-------
100
Number of Sample Pairs Needed
(Power = 50% Confidence = 95%)
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Coefficient of Variation
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
B-10
April 25, 2002
-------
APPENDIX C
DERIVATION OF THE NUMBER OF SAMPLES REQUIRED TO MEASURE A
STATISTICAL DIFFERENCE IN POPULATION MEANS
Define: COV = a / C
(r -r }/
% removal = v '" out'/^
/ ^ in
Setting the lower boundary of the influent confidence interval to the upper boundary of the
effluent confidence interval gives:
C'-j in /^ . '-y out
, ^ / '^= *~-,f ~r t-1 f* /.
The COV is substituted for the a in the above equation. While the a of a BMP effluent is almost
certainly less than the a of the BMP influent, the assumption that COV;n = COVout is a more
reasonable one. In most instances the COV of the BMP effluent would be less than the influent.
Ample data are available for estimating the COV for influent flows to stormwater BMPs, such as
the ASCE database; this is not the case for effluent flows. It is also assumed that n is the same
for the influent and effluent (n^ = nout). These assumptions simplify the equation.
Substituting oin = COV x C, and aol,t = COV x Coa , where COVin = COVollt yield:
- CQVxCm _- COVxCout
in *-V I ~ ^-'out + ^ a,
/
V n
rearranging:
C +C
'" ot"
Substituting for C^ =Q - C, (%removal) gives:
2 x C,,, - %removal x C.
Cm x ^removal = COVxZa,\
Dividing both sides by C.m and solving for n yields:
" Za/ x COV x(2- %removal)~
n =
%removal
The above approach considers the number of samples required for a power of 50%. For an
arbitrary power the equation becomes:
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
C-l
-------
n =
where,
:COVx(2-%removal)
%removal
Zp/2: false negative rate (l-(3 is the power. If used, a value of |3 of 0.2 is common, but i
is frequently ignored, corresponding to a (3 of 0.5.)
it
Urban Stormwater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
C-2
April 25, 2002
-------
APPENDIX D
RELATIONSHIPS OF LOG-NORMAL DISTRIBUTIONS
Table D.I
T =
M =
M =
CV
EXP(U)
= EXP(U + 0.5
= T* SQRT(1
= SQRT (EXP
*W2)
f CV2)
(W2) - 1)
s =
w =
u =
u =
M*CV
= SQRT (LN (1 +
CV2)
LN (M/EXP (O.5 * W2))
LN(M/SQRT(1
+ CV
2)
Parameter designations are defined as:
Arithmetic Logarithmic
MEAN M U
STD DEVIATION S W
COEF OF VARIATION CV
MEDIAN T
LN(x) designates the base e logarithm of the value x
SQRT(x) designates the square root of the value x
EXP(x) designates e to the power x
Urban Storrmvater BMP Performance Monitoring
A Guidance Manual for Meeting the National Stormwater BMP Database Requirements
April 25, 2002
D-l
------- |