EPA/625/R-03/003
February 2003
National Conference on
Urban Storm Water:
Enhancing Programs
at the Local Level
Proceedings
Chicago, IL
February 17-20, 2003
Technology Transfer and Support Division
National Risk Management Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
-------
NOTICE
Mention of trade names or commercial products does not
constitute endorsement or recommendation for use.
in
-------
FOREWORD
The U.S. Environmental Protection Agency is charged by Congress with
protecting the Nation's land, air, and water resources. Under a mandate of
national environmental laws, the Agency strives to formulate and imple-
ment actions leading to a compatible balance between human activities
and the ability of natural systems to support and nurture life. To meet this
mandate, ERA'S research program is providing data and technical support
for solving environmental problems today and building a science knowl-
edge base necessary to manage our ecological resources wisely, under-
stand how pollutants affect our health, and prevent or reduce risks in the
future.
The National Risk Management Research Laboratory is the Agency's
center for investigation of technological and management approaches for
reducing risks from threats to human health and the environment. The
focus of the Laboratory's research program is on methods for the preven-
tion and control of pollution to air, land, water, and subsurface resources;
protection of water quality in public water systems, remediation of contami-
nated sites and groundwater; and prevention and control of indoor air
pollution. The goal of this research is to catalyze development and imple-
mentation of innovative, cost-effective environmental technologies; de-
velop scientific and engineering information needed by EPA to support
regulatory and policy decisions; and provide technical support and infor-
mation transfer to ensure effective implementation of environmental regula-
tions and strategies.
This publication has been produced as part of the Laboratory's strategic long-
term research plan. It is published and made available by ERA'S Office of
Research and Development to assist the user community and to link research-
ers with their clients.
Hugh McKinnon, Acting Director
National Risk Management Research
Laboratory
IV
-------
TABLE OF CONTENTS
Source Area and Regional Storm Water Treatment Practices: Options for Achieving Phase II Retrofit
Requirements in Wisconsin 1
Management Strategies for Urban Stream Rehabilitation 20
Remediation of Stormwater Residuals Decant with HydrocotyleRanunculoides 29
Inappropriate Discharge Detection and Elimination: What Phase I Communities are Doing to Address the
Problem 37
Use of the Clean Water State Revolving Fund for Municipal Storm Water Management Programs 51
Planning and Assessment of Best Management Practices in the Rouge River Watershed 59
The Maryland Stormwater Management Program: A New Approach to Stormwater Design 76
Enhancing Storm Water Infiltration to Reduce Water Temperature Downstream 85
Local Solutions to Minimizing the Impact of Land Use Change 98
Thornton Transitional Reservoir Strom Water Management 106
Developing Split-Flow4™1 Stormwater Systems 114
Reforest the Bluegrass: Empowerment of the Citizen Watershed Manager 123
Storm Water Phase IMS4 Permitting: Writing More Effective, Measurable Permits 134
Conservation Design Tools for Stormwater Management 142
Using Technical Data and Marketing Research to Change Behavior 147
Evaluating Innovative Stormwater Treatment Technologies Under the Environmental Technology
Verification (ETV) Program 156
Using an Indicators Database to Measure Stormwater Program Effectiveness in Hampton Roads 167
Using Incentives and Other Actions to Reduce Watershed Impacts from Existing Development 181
-------
TABLE OF CONTENTS (continued)
Lessons Learned from In-Field Evaluations of Phase I Municipal Storm Water Programs 191
EcoRoofs (Greenroofs) - A More Sustainable Infrastructure 198
Storm Water Management in the City of Chicago 215
Regional Facility vs. On-Site Development Regulations: Increasing Flexibility and Effectiveness in
DevelopmentRegulationlmplementation 221
A Regional Approach to Phase II Permitting Encourages Cooperation and Reduces Cost 236
Fish Community Response in a Rapidly Suburbanizing Landscape 253
Elements of Successful Stormwater Outreach and Education 263
AConservation Plan for Three Watersheds within the Milwaukee Metropolitan Sewerage District
(MMSD) 272
Critical Components for Successful Planning, Design, Construction and Maintenance of Stormwater
Best Management Practices 291
Evaluation of NPDES Phase I Municipal Stormwater Monitoring Data 306
Funding Phase II Storm Water Programs 328
Protecting Water Resources with Higher Density Developments 340
Predicting the Impact of Urban Development on Stream Temperature Using a Thermal Urban Runoff
Model (TURM) 369
Rain Barrels—Truth or Consequences 390
The Stranger Amonst Us: UrbanRunoff, The Forgotten Local Water Resource 395
AProcess for Determining Appropriate Impact Indicators for Watershed Proj ects 409
Dual Function Growth Medium and Structural Soil for Use as Porous Pavement 416
The Wash Project—Thinking Outside the Culvert 427
vi
-------
TABLE OF CONTENTS (continued)
Necessity and Opportunity: Urban Stormwater Management in Rockville, Maryland 440
Re-Inventing Urban Hydrology in British Columbia: Runoff Volume Management for Watershed
Protection 453
Development of the San Diego Creek Natural Treatment System 470
" Sherlocks of Stormwater" Effective Investigation Techniques for Illicit Connection and Discharge
Detection 489
Low Impact Development Strategies for Rural Communities 497
An Assessment and Proposed Classification of Current Construction and Post Construction Structural
Best Management Practices (BMPs) 502
Overcoming Challenges in Establishing a Regional Public Education and Outreach Partnership 518
The Watershed Partners: An Education Collaboration that Works 526
Educating the Las Vegas Community about Storm Water Pollution 537
Illicit and Industrial Storm Water Controls: AMunicipal Perspective 546
AReassessment of the Expanded EPA/ASCE National BMP Database 555
Posters
EPAs Management Measures Guidance to Control Nonpoint Source Pollution from Urban Areas: It's
Time to Develop and Implement Your Storm Water Management Program.... Are You Ready? 574
Evaluating Innovative Stormwater Treatment Technologies Under the Environmental Technology
Verification (ETV) Program 580
Managing Storm Water in Wisconsin: ALocal Partnership Protects the Kinnickinnic River 582
Duluth Streams: Community Partnership for Understanding Urban Stormwater and Water Quality Issues
at the Head of the Great Lakes 600
Maximum Utility for Minimum Cost: Simple Structural Methods for Stormwater Quality Improvement 601
vii
-------
ABSTRACT
A wide array of effective storm water management and resource protection
tools have been developed for urban environments, but their implementation
continues to be hampered by a lack of technology transfer opportunities. At
the national conference Urban Storm Water: Enhancing Programs at the
Local Level, attendees learned about state-of-the-art technologies and imple-
mentation programs that have proven success in local communities.
The timing of this Conference coincided well with the implementation of U.S.
ERA's Phase II NPDES Storm Water Program. Participants learned about
the most effective tools and technologies for meeting these new NPDES
permit requirements. Attendees included staff and engineers representing
local municipalities, as well as water resource managers, conservation
groups, local officials, researchers, educators, and state agency personnel.
Conference sessions featured progressive scientists and researchers, along
with managers of successful projects from across the country. Two concur-
rent sessions—one focusing on tools and technology, the other focusing on
program implementation—allowed participants to tailor the Conference ex-
perience to fit their personal educational goals.
This Conference was the fifth in a popular series of water quality specialty
conferences sponsored by the U.S. Environmental Protection Agency's Re-
gion 5 Water Division office. The Chicago Botanic Garden, which is owned
by the Forest Preserve District of Cook County and managed by the Chicago
Horticultural Society, was pleased to coordinate the conference. Also co-
sponsoring the event were the U.S. Environmental Protection Agency's Of-
fice of Wastewater Management and Office of Research and Development,
as well as Tetra Tech, Inc. The conference was conducted in cooperation
with the Center for Watershed Protection. Approximately 350 attendees par-
ticipated.
Three pre-conference wprkshops were held on February 17. Smart Water-
sheds: Building Municipal Programs to Restore Urban Watersheds pro-
vided practical and useful advice on how to implement "smart" watershed
programs, which relate to a group of 17 municipal programs that can be inte-
grated together at the watershed level to improve the quality of runoff and
habitat in urban streams. The workshop was led by staff from the Center for
Watershed Protection. The second pre-conference workshop, Countdown
to the Phase II Implementation Deadline: Putting the Final Touches on
Your Storm Water Permit, presented details that Phase II municipal pro-
grams and construction site operators need to know in order to complete
their programs and storm water permit applications. Instructors for this work-
shop were staff from
Tetra Tech, Inc. The third pre-conference workshop, Certified Professional
in Storm Water Quality (CPSWQ) Exam Review Course, provided partici-
pants with an understanding the CPSWQ exam's cpntent and format. Instruc-
tors for this workshop were from Certified Professional in Erosion and Sedi-
ment Control, Inc.
Vlll
-------
This Conference Proceedings includes many of the papers presented dur-
ing the conference, and a copy has been provided to each attendee. All pa-
pers included were peer reviewed. Additional copies, in either paper or CD-
ROM format, are available free of charge from the U.S. Environmental Pro-
tection Agency's National Center for Environmental Publications: telephone
800/490-9198, or visit their Web site at .
Robert J. Kirschner
Conference Coordinator,
Chicago Botanic Garden
IX
-------
ACKNOWLEDGMENTS
The success of the conference and the preparation of this proceedings are due
largely to the efforts of the presenters as well as the following individuals:
Conference Planning Committee
Robert J. Kirschner, Conference Coordinator
Chicago Botanic Garden, Glencoe, IL
Thomas E. Davenport, Project Officer
U.S. Environmental Protection Agency - Region 5, Chicago, IL
Scott Minamyer, Proceedings Editor
U.S. Environmental Protection Agency, Cincinnati, OH
Ted Brown, Center for Watershed Protection, Ellicott City, MD
James H. Collins, TetraTech, Inc., Fairfax, VA
Alyssa Dodd, Penn State University, University Park, PA
Rod Frederick, U.S. Environmental Protection Agency, Washington, DC
Ralph Reznick, Michigan Department of Environmental Quality, Lansing, Ml
Eric W Strecker, GeoSyntec Consultants, Portland, OR
Peer Reviewers
Diana Allen, National Park Service, St. Louis, MO
Lisa Austin, GeoSyntec Consultants, Portland, OR
Brian Bohl, Hamilton Soil and Water Conservation District, Cincinnati, OH
Paul Braasch, Clermont County Office of Environmental Quality, Cincinnati, OH
Donald Brown, U.S. Environmental Protection Agency, ORD, Cincinnati, OH
Ted Brown, Center for Watershed Protection, Ellicott City, MD
Deb Caraco, Center for Watershed Protection, Ellicott City, MD
Thomas Davenport, U.S. Environmental Protection Agency, Region 5, Chicago, IL
Jennifer Deaton, Butler Soil and Water Conservation District, Hamilton, OH
Nancy Ellwood, Mill Creek Watershed Council, Cincinnati, OH
Chi-Yuan Fan, U.S. Environmental Protection Agency, ORD, Edison, NJ
Richard Field, U.S. Environmental Protection Agency, ORD, Edison, NJ
Rod Frederick, U.S. Environmental Protection Agency, OW, Washington, DC
Todd Hesse, GeoSyntec Consultants, Portland, OR
Jim Howell, GeoSyntec Consultants, Portland, OR
Sarah Lehmann, U.S. Environmental Protection Agency, Region 5, Chicago, IL
-------
Marc Leisenring, GeoSyntec Consultants, Portland, OR
Peter Manganella, GeoSyntec Consultants, Portland, OR
Nikki McClain, USDA, NRCS, Muncie, IN
John McManus, Clermont County Office of Environmental Quality, Cincinnati, OH
Scott Minamyer, U.S. Environmental Protection Agency, ORD, Cincinnati, OH
Nhien Pham, U.S. Environmental Protection Agency, Region 5, Chicago, IL
Marsha Rolph, Warren County Soil & Water Conservation District, Lebanon, OH
Joyce Perdek, U.S. Environmental Protection Agency, ORD, Edison, NJ
Asim Ray, U.S. Environmental Protection Agency, ORD, Edison, NJ
Steven Roy, GeoSyntec Consultants, Portland, OR
Michael Royer, U.S. Environmental Protection Agency, ORD, Edison, NJ
Marcus Quigley, GeoSyntec Consultants, Portland, OR
Trent Schade, U.S. Environmental Protection Agency, ORD, Cincinnati, OH
Tom Schueler, Center for Watershed Protection, Ellicott City, MD
William Shuster, U.S. Environmental Protection Agency, ORD, Cincinnati, OH
Eric Strecker, GeoSyntec Consultants, Portland, OR
Mary Stinson, U.S. Environmental Protection Agency, ORD, Edison, NJ
Anthony Tafuri, U.S. Environmental Protection Agency, ORD, Edison, NJ
Paul Thomas, U.S. Environmental Protection Agency, Region 5, Chicago, IL
Ariamalar Selvakumar, U.S. Environmental Protection Agency, ORD, Edison, NJ
Peter Swenson, U.S. Environmental Protection Agency, Region 5, Chicago, IL
Jennifer Zielinski, Center for Watershed Protection, Ellicott City, MD
Publishing Support
Jean Dye and Steve Wilson, U.S. Environmental Protection Agency, ORD, Cincinnati,
OH
XI
-------
Source Area and Regional Storm Water Treatment Practices: Options for Achieving
Phase II Retrofit Requirements in Wisconsin
Roger Bannerman, Wisconsin Department of Natural Resources, Madison, WI
Greg Fries, City of Madison, Madison, WI
Contributor: Judy Horwatich, U.S. Geological Survey; Middleton, WI
Abstract
A recently calibrated urban runoff model, the Source Loading and Management Model (SLAMM), is used
to compare the cost-effectiveness of using source area and regional stormwater treatment practices. The
demonstration is done for the totally urbanized Lake Wingra watershed in Madison, Wisconsin. The goal is
to retrofit practices that are able to reduce the annual total suspended solids load by 40%. Model results
indicate the parking lots and streets are the most important sources of total suspended solids. Practices
evaluated for the parking lots include the Delaware Perimeter Sand Filter, Stormceptor, Multi-Chamber
Treatment Tank, bioretention, porous pavement, and infiltration trenches. Individually they reduced the
solids load to Lake Wingra by 7 to 19%. High efficiency street sweeping is projected to reduce the annual
solids load by 17%.
Nine combinations of the source area practices are able to achieve the 40% reduction goal. For example, a
42% reduction in solids load to Lake Wingra is estimated for the combination of high efficiency street
sweeping on all the streets and Delaware Perimeter Sand Filters on all the parking lots. Alternatively, the
40% reduction is achieved by using regional detention ponds with a total of 20 acres of permanent pool
area. Many of the combinations of source area practices are more cost-effective than the regional practice.
Assuming a lifespan of 20 years the annual cost of the source area practices ranges from abut $573,000 to
$1,504,000, while the range for the detention ponds is $963,000 to $1,840,000. The least expensive
combination of source area practices would only increase the annual stormwater utility bill for the Madison
taxpayers by about $6, while the most likely detention pond alternative will increase the utility bills by
about $18. Cities should consider retrofitting source area practices as a cost-effective way to meet reduction
goals for total suspended solids.
Introduction
A new rule (NR151) to be administrated by the Wisconsin Department of Natural Resources (Department)
contains performance standards to reduce the impacts of stormwater for both developing and established
urban areas. Over 200 Wisconsin cities will be affected by the rules, because the performance standards
will be in their EPA Phase II permits. Standards for the developing areas address problems of construction
erosion, post-development suspended solids loads, and sustaining the natural hydrology of the watersheds.
These developing areas standards should reduce the risk of any future degradation to our lakes and streams.
The Department also hopes to enhance the quality of our degraded urban lakes and streams by requiring
some sediment reduction in established urban areas.
Performance standards for the established areas will require the cities to reduce the annual total suspended
solids (TSS) loads by 40%. The standard must be achieved by the year 2013. Since the Phase U permits
will be issued in 2003, the cities will have two permit cycles to achieve the standard. Ten years seems like a
long time, but the cities will need the time to implement the practices. It might take more than two years just
for cities to develop their management strategies
-------
The 40% reduction assumes no stormwater treatment practices (STPs) exist in the established urban areas.
A city will receive credit for any existing STPs. Since most cities rely on street sweeping and catch basin
cleaning for reducing solids loads in older neighborhoods, they will have to add more practices or
completely replace their old ones to achieve the 40%. Older style broom street sweepers and catch basin
cleaning is not expected to achieve more than a 20% reduction in annual suspended solids loads.
Cites will have the challenge of both determining the benefits of their existing STPs and deciding what
additional practices they will need to achieve the goal. At the same time they need to select STPs that have
the lowest possible capital and maintenance cost. To meet the challenge the cities will have to use urban
runoff models and the latest information available on the effectiveness and cost of STPs.
Our purpose is to demonstrate the types and cost of STPs that will achieve the 40% reduction in the Lake
Wingra watershed, which is an established urban area in Madison, Wisconsin. Of special interest to us is to
compare the benefits of using source area STPs, such as street sweeping and filtration devices, with regional
practices, such as detention ponds. An urban runoff model called Source Loading and Management Model
(SLAMM) is used along with literature values for practice effectiveness and cost.
A Description of the Lake Wingra Watershed
A lot of the information needed to complete a stormwater plan is already available for the Lake Wingra
watershed. Not only has there been a lot of research completed on the lake itself, but the watershed has also
been the object of two planning efforts (Univ. of WL, 1999; Dane County, 1992). Both of the plans identify
sedimentation as an important issue for the lake. Both plans say that stormwater is an important source of
the suspended solids load to the lake. The priority watershed plan suggests a 30 to 50% reduction in the
annual suspended solids load. Neither plan did a comprehensive analysis of the alternative stormwater
practices, which means they did not do a detailed comparison of source area and regional practices.
Lake Wingra is a small (325 acres), shallow, highly eutrophic lake, but its location in a highly populated
urban area makes it the focus of many recreational activities. Sedimentation problems are bad enough
around sewer outfalls to restrict access by boats - even canoes. Heavy weed growth in the lake also reduces
the area of the lake used by canoes, sailboats, and sail boarders. Water quality problems contribute to a
decline in attendance at the swimming beach, but there is still a lot of use of the beach.
The most recent landuse information is available from the City of Madison. The city has divided the
watershed into eight sub-watersheds (Figure 1). Five of the sub-watersheds are highly urbanized, while two
of the sub-watersheds (WI-05 and WI-08) are mostly in the University of Wisconsin arboretum. Most of
this land is forest and prairie preserve managed by the university. There is almost no new construction in
the watershed.
The watershed is about 3947 acres (6.2 square miles) in size (Table 1). This value does not include the area
of the lake, the 210 acres of wetlands and 48 acres of ponds in the watershed. Residential is the largest
landuse category in the watershed and most of it is medium density residential. Open space is the next
largest landuse category at 29%, which includes the University Arboretum, golf courses, city parks, and
cemeteries. About 62% of the open space is in the University Arboretum. Together the residential, open
space and commercial landuses account for 92% of all the land in the watershed. Most of the commercial
landuse is divided equally between shopping centers and office parks. The watershed also includes a
freeway, five schools, and some light industrial sites.
-------
Table 1. Landuse areas for the eight subwatershed in the Lake Wingra Watershed1
Landuse
Residential
Institutional
Commercial
Industrial
Open
Freeway
Total
Acres of landuse by subwatershed^
WI-01
418
0
256
0
88
53
815
WI-02
829
18
7
0
170
27
1051
WI-03
229
63
9
0
188
0
489
WI-04
31
0
0
0
0
0
31
WI-05
37
0
0
14
104
0
155
WI-06
43
0
0
0
13
0
56
WI-07
371
0
256
40
41
92
800
WI-08
11
0
0
0
539
0
550
Watershed Total
Ac
1968
81
528
54
1144
172
3947
%
50
2
13
1
29
5
100
1. Lake Wingra (325 ac), wetlands (210 ac) and ponds (48) are not included in landuse areas.
2. Most of WI-05 and WI-08 are in the University of Wisconsin Arboretum.
Explanation
/V Watersheds
C3 Lake Wingra ^— j Low_Density_Res
[771 Multi Family
' - J
I Pond
I Wetland
^^ | | Med Density Res
H Undeveloped institutional
L__J Park
CD Golf_Course Q-g Shopping_Center
CZI Cemetery ^j Office.Park
Beltline
A ZUSGS
I I Strip Commercial
| . •• J Shopping_Centei
^B Office Park
| | Light Industrial
Figure 1. Distribution of landuses in the Lake Wingra watershed.
For the purpose of the demonstration, we assumed no pre-exiting practices in the Lake Wingra watershed.
Consequently, our model runs do not include any pre-existing practices. In fact, the city does street
sweeping and there are seven detention ponds in the watershed. Six of the detention ponds are located on
-------
the University Arboretum property. These are seen as small blue dots in Figure 1. The remaining detention
pond is on the golf course in WI-01 [Figure 1]. The arboretum built the detention ponds to reduce the
erosive effects of the runoff and to protect their wetlands from sedimentation. These practices are helping
to reduce the suspended solid load to Lake Wingra. Otherwise much of the runoff from four of the more
urbanized sub-watersheds (WI-01, WI-02, WI-05, and WI-07) would flow unchecked down open channels
to Lake Wingra.
Also, we do not include sediment loads from bank erosion in our estimate of total sediment loads to Lake
Wingra. Severe bank erosion is occurring in several streams tributary to the lake. Bank stabilization
projects are necessary to control this source of sediment.
Six Steps to Finding the 40% Solution
Developing a stormwater plan that considers both source area and regional STPs will require more steps
than a plan that just considers regional practices. To include the source area practices, more work is needed
to identify the sources of the pollutants of concern, more types of STPs need to be evaluated, and more sites
in the drainage area must be identified. Although it takes more work to include source area practices, we
think a retrofit plan has a better chance of being implemented if it is not limited to regional practices. Source
area practices can be incorporated into places that regional practices will simply not fit and they are usually
less disruptive to the neighborhood. Previous experience in Wisconsin has demonstrated how unreceptive
people can be to being displaced from their parks and homes by regional stormwater treatment practices.
We think the following six steps should be part of any stormwater management plan that includes source
area practices. We used these steps to demonstrate the validity of using source area practices in the Lake
Wingra watershed. Since we are only trying to demonstrate the relative cost-effectiveness of source area
and regional practices, the steps do not include all the activities needed to actually install STPs in the Lake
Wingra watershed. For example, a more comprehensive stormwater plan should include collection of site
information, such as soil types and location of utilities, sizing of the STPs in each location, and the actual
cost of installation at each site.
1. Select and calibrate an urban runoff model.
2. Determine the annual suspended solids loads for each sub-watershed, landuse, and source area in the
watershed.
3. Select source area and regional practices to be evaluated for watershed.
4. Determine ability of each practice and combinations of practices to achieve pollutant reduction goal.
5. Identify unit capital and maintenance cost of each practice.
6. Determine cost of each management alternative that achieves pollutant reduction goal.
We think enough information is available now to complete all six steps for any watershed. Cost information
about each STP is the hardest to find. Fortunately we could find some conceptual cost data for each
practice. Information about the effectiveness of each practice is also very limited (Winer, 2000), but
ongoing monitoring efforts, such as the EPA's Environmental Technology Verification effort, should
greatly increase our database over the next few years. New monitoring sites are being added to the National
Stormwater Best Management Practices (BMP) Database all the time (EPA, 1999). We relied on an urban
runoff model to help identify the most important sources of the TSS.
-------
We selected the Source Loading and Management Model (SLAMM) to demonstrate the relative benefits of
regional and source area practices (Pitt, 2002). We considered other models, such as P8 and SIMTPM, but
only SLAMM is designed to easily produce a TSS load for each source area, such as streets and parking lots
(Sutherland, 1999 and Walker, 1990). All three models are capable of testing regional practices, but only
SLAMM is designed to specifically evaluate the effectiveness of practices on all the source areas.
Source areas are the building blocks for calculating runoff volumes and pollutant loads for the six landuses
addressed by SLAMM - residential, commercial, industrial, institutional, open space, and transportation
landuses. Examples of the source areas characteristic of each landuse are roofs, parking lots, driveways,
sidewalks, streets, small landscaping (lawns), large landscaping, playgrounds, isolated areas, undeveloped
areas, and unpaved parking lots. Pollutant loads and runoff volumes calculated for each source area are
added together to produce the estimates for each landuse.
Stormwater treatment practices can be applied to each source area, the conveyance system, and/or the end-
of-the-pipe. Some of the practices are only applied to source areas, such as street sweeping and porous
pavement. Others, such as catch basin cleaning and grass swales, are reserved for the conveyance system.
Many of the available practices in SLAMM, such as detention ponds and infiltration devices, are applied to
both source area and end-of-the-pipe solutions. A user may select multiple sites and practices or just decide
to apply one practice at one location. The model output summarizes the benefits of the practices by source
area and landuse.
To make the source area loads as valid as possible, we think it is very important to calibrate SLAMM for all
parts of the country. A minimum calibration requires the collection of event related flow and TSS
concentration data at the end of a stormsewer pipe. Although most people preparing stormwater plans will
not have enough data to calibrate a model, our efforts to calibrate SLAMM should make the model a
reasonable choice for preparing stormwater plans in the upper Midwest.
SLAMM Calibration
To help people prepare stormwater management plans in Wisconsin, we calibrated SLAMM using data
collected by the U.S. Geological Survey office in Madison, Wisconsin. Fortunately, they have recently
collected source area runoff volumes and TSS concentrations, rain depths for monitored storms, and runoff
volumes and TSS concentrations at the stormsewer outfall at six sewersheds in Wisconsin and one in
Michigan (Table 2).
Table 2. Comparison of measured and predicted TSS loads and runoff volume at eight stormwater study sites.
Site
Harper1
Monroe1
Canterbury '
Marquette
Superior
West Towne '
Syene1
Badger Road '
Landuse Type
Residential
Res/com
Res/com
Res/com
Commercial
Commercial
Light Industrial
Light Industrial
TSS
Number of
Events for
Calibration
23
32
14
71
21
-
82
18
Percent
Difference
11
-52
12
-29
-66
N/A
19
-40
Runoff Volume
Number of
Storms for
Calibration
55
75
55
64
91
66
108
40
Percent
Difference
-27
7
10
19
-4
31
-8
-4
1. Sites are near or in Lake Wingra Watershed.
-------
The mostly residential Monroe study site is in the Lake Wingra watershed and four of the study sites are
located very near the Wingra watershed (Bannerman and others, 1990 and Waschbusch and others, 1999).
These are the Harper, Canterbury, Syene, and Badger Road study sites. The Marquette site is in Michigan
(Steur and others, 1997) and the Superior site is northern Wisconsin (Steur and others, 1997). The median
number of storms collected for flow is 64, while the median value for the number of water quality storms is
23.
The following is a list of the files we calibrated in SLAMM and the name of the file we use in Wisconsin.
These and other files for the model are on the U.S. Geological web page with the URL of
http://wi.water.usgs.gov/slamm/index.html. Copies of SLAMM are available at WINSLAMM.com.
1. Runoff coefficient: .rsv (WISIOl.rsv)
2. Particulate Solids Concentration: psc (WIAVG01)
3. Pollutant Probability Distribution: .ppd (WIGEO01)
4. Particulate Residue Reduction: .ppr (WIPLV01)
5. Street Delivery Parameter: .std (WISTR01)
SLAMM did a good job of matching the total runoff volumes and TSS loads measured at the end of the
stormsewer pipe for each study site. The median difference between the predicted and measured runoff
volume is 8% and the median difference for the total suspended solids loads is about 29% (Table 2). We
are concerned about the differences of around 50% for suspended solids at Monroe, Superior, and Badger
Road sites. It appears the model is not accounting for some of the sediment collected by the automatic
samplers at these three sites during the largest rainfall events. Over half the difference between the
measured and estimated sediment load at the Superior site are caused by the model underestimating the load
for the largest rainfall. Estimated sediment loads would be ten percent higher without the effect of the
largest rainfall at the Badger Road site. Piles of soil observed at both sites could be the source of sediments
the model does not account for during larger events. Estimated and measured runoff volumes are very close
for those larger events, so the difference in loads is due to the difference in concentrations.
A 52% difference at Monroe seems to be explained by the unusual amount of deposited sediment observed
in the flat part of the storm sewer pipe. Six high intensity storms accounted for most of the error at Monroe
Street. The model is not designed to account for the re-suspension of sediment deposited at the bottom of a
storm sewer pipe.
Sources of Total Suspended Solids in the Lake Wingra Watershed
After we completed the calibration, we thought SLAMM was ready to help us identify the important
sources of TSS in the Wingra watershed. We first ran SLAMM on the eight sub-watersheds with the hope
of eliminating some of the sub-watersheds from the rest of the analysis. The city of Madison provided the
acres of each landuse in the subwatersheds and the development characteristics we needed for each landuse
were obtained from the average development characteristic files on the U.S. Geological Survey web page
(http://wi. water.usgs.gov/slamm/mdex.html). Examples of the development characteristics are the acres of
each source area, amount of connected impendousness, and street texture.
We used the average rainfall year file for the Madison area (MSN1981.ran) to run SLAMM for the eight
subwatersheds. Four of the sub-watersheds contribute about 92% of the annual suspended solids load for
the watershed (Table 3). In an average rainfall year sub-watersheds WI-01, WI-02, WI-03, WI-07
contribute about 457 tons of suspended solids to Lake Wingra. This is about the same as the average load
-------
(401 tons) estimated for the watershed when the principle landuse was agricultural (Corsi and others, 1997).
It is not a surprise that these four watersheds contribute most of the sediment, since they contain about 95%
of all the built-up landuses in the Wingra watershed.
Table 3. Annual TSS loads and runoff volume for each subwatershed
in the Lake Wingra Watershed
Subwatershed
WI-01
WI-02
WI-03
WI-04
WI-05
WI-06
WI-07
WI-08
Total
% of Total
Area
21
27
12
1
4
1
20
14
100
TSS (Ibs)
269,000
253,000
108,000
8,000
19,000
8,000
284,000
44,000
993,000
TSS,
%
27
26
11
1
2
1
28
4
100
Annual runoff
volume (ft3)
30,519,000
23,886,000
11,149,000
724,000
2,376,000
663,000
37,314,000
3,114,000
109,745,000
Percent runoff
volume
28
22
10
1
2
1
33
3
100
Regional or source area STPs should be implemented in these four critical subwatersheds. If regional STPs
were to be installed at the ends of the critical subwatersheds, they would need to have at least a 50%
removal efficiency in order to achieve the 40% reduction goal. The output from the model runs used to
identify the critical subwatershed can also be summarized to determine landuses with the highest TSS loads.
This is the next step in the identification of the most important source areas to control.
Commercial and residential landuses in the critical subwatersheds contribute about 82% of the annual TSS
loads (Figure 2). Residential loads are proportionate to the percent of the area they occupy, while percent of
the load contributed by the commercial is almost twice as high as the percent of the area it occupies. This
makes the commercial landuse an important target for our management efforts. On the other hand it is less
cost effective to treat the open space landuses, since 16% of the area produces only 5% of the load. We did
not add industrial landuse to our targeted landuse list, because they represent only 2% of the load. If we
assume the institutional and commercial landuses have similar source areas, we can add the 4% TSS load
from the institutional landuses to the commercial load for a total of 35%. Source areas within the
commercial, institutional, and residential landuses were expected to yield the highest percent of the annual
TSS load.
16%--T\
r£ ^
58%
% Total Suspended
Solids
5%
51%
% Runoff volume
47%
36%
2% 1%
Figure 2. Percent area, TSS load, and runoff volume for landuses in four Lake Wingra sub-watersheds.
-------
Parking lots and streets in the four sub-watersheds represent only 26% of the area, but contribute about 66%
of the annual suspended solids load (Figure 3). These two source areas are mostly in the commercial,
institutional, and residential landuses. Roofs and lawns are a less critical source of suspended solids,
because they represent 47% of the area and only produce about 12% of the load. The same is true for large
landscaped areas, which includes city parks and golf courses. To be cost-effective our practice selection has
to target the streets and parking lots as much as possible.
If we want to evaluate source area STPs that have a removal effectiveness for TSS of less than about 70%,
we have to include some of the other source areas in our analysis.
DRoof
D Paved Parking
D Street
D Large
Landscape
D Lawns
• Freeway
• Other
% Source Area
13%
34%
% Total Suspended Solids
12% 5%
\
. 26%
% Water Volume
12%
17%
20%
40%
Figure 3. Percent area, TSS load, and runoff volume for source areas in four Lake Wingra sub-watersheds.
A 70% control of parking lots and streets would just achieve the 40% (46% control TSS) reduction goal for
the Wingra watershed. This is partially because a 100% control of the two source areas results in TSS
reduction of 66% for the entire watershed. To give us more choices in our practice selection, we needed to
boost the total% of the TSS load we could control. We did this by including other source areas in our
analysis, especially freeways, lawns, and roofs.
Selection of Stormwater Treatment Practices
To achieve the goal of the demonstration, it was only necessary to select one regional practice. Several
types of source area practices are needed, however, to cover all the types of source areas. Selection of a
number of source area practices would allow us to include proprietary and non-proprietary practices with a
range of TSS removal values. These could represent a number of treatment processes, such as settling,
filtration, and infiltration. Our criteria for selecting regional and source area practices included the
availability of good data to verify their effectiveness, some cost information, and hopefully some experience
with the practice in Wisconsin.
Regional Practice
Detention ponds met all our criteria, so they were selected as the regional practice to compare to source area
practices. Settling is the main treatment process for the detention ponds. Many studies including one in
Wisconsin indicate detention pond can achieve an 80% reduction in annual suspended solids loads (House
and others, 1993, Winer, 2000). The regional practice had to have a TSS removal capability of at least 50%
-------
to achieve the 40% reduction goal for the watershed. By using a practice with a TSS removal of 80% the
regional practice could be located to serve less the whole drainage area and still achieve the 40% goal
(Table 4).
Table 4. TSS removal values reported for selected stormwater
treatment practices.
Stormwater
Treatment Practices
Multi Chamber
Treatment Tank
Stormceptor
Delaware Perimeter
Sand Filter
High Efficiency Street
Sweeping (city street)
High Efficiency Street
Sweeping (freeway)
Detention Ponds
Bioretention
Broom Street
Sweeper
Porous Pavement
Infiltration Trench
Rain Gardens
Description of Stormwater Treatment
Practices
Three chambers - grit chamber,
settling chamber, and sand/peat
filter media chamber with by -pass
Vertical single cylindrical chamber
using swirl action and settling with
Duilt in by-pass
Underground sand filter using
settling chamber followed by sand
filter chamber
Vacuum action pick-up assisted by
brooms and/or jets of air
Vacuum action pick-up assisted by
brooms and/or jets of air
Holes in the ground with permanent
pools designed to settle particles
Shallow depressed planted area
underlain by a layer of formulated
soil (mostly sand) over a layer of
gravel. Treatment includes
sedimentation, filtration, adsorption,
microbial decay, and plant uptake.
Broom action pick-up assisted by
conveyor belt
Porous asphalt or interlocking
paving blocks providing infiltration
A lined excavated trench backfilled
with gravel. Infiltration followed by
filtration in native soils
Shallow depression that's planted
with a variety of perennials.
Abbreviation of
Stormwater Treatment
Practices
MCTT
Stormceptor
Delaware Filter
High Sweep
High Sweep
Ponds
Bioretention
Broom Sweep
Pavement
Trench
Gardens
Reported TSS
removal, % (1)
80
33
83
602
453
80
75
202
95
NA4
755
1. Percent assumes all devices working at maximum efficiency.
2. Removal efficiency for city streets with sweeping once per week for 30 weeks.
3. Removal efficiency for freeways with sweeping once per week for 30 weeks.
4. TSS removal is probably very high, because reportedTP removal is 100%.
5. Assume same as reported bioretention.
Of course, many detention ponds have been installed in Wisconsin. With so many being installed in new
development sites, Wisconsin cities have accepted them as a good way to meet their goals for flood control
and reduce TSS loads. Very few of them, however, have been retrofitted into existing urban areas.
Refrofiting a detention pond in an existing urban area has the potential to cause a lot of disruption to people
living in the neighborhood. In most cases, this alternative will not be politically feasible, except when a
there is a lot of open land, such as the presence of the arboretum in the Lake Wingra watershed. A
stormwater plan prepared for the Lincoln Creek Watershed in the City Milwaukee was promptly rejected
-------
when the groups involved realized the only alternative being offered was to put detention ponds in many of
the public parks - 60 ponds altogether.
In estimating costs for ponds, it was assumed that either the land is available and must be purchased at a fair
market price or the land is available but the purchase price included the cost of existing buildings (Table 5).
Both alternatives assumed a cost for repositioning the existing storm sewer system (Southeastern Regional,
1991). Since the retrofit cost calculations are over ten years old, we applied an annual inflation factor of 3%
to building and maintenance of the ponds and we increased the land cost by 10% each year. Retrofit cost of
about one to two million dollars for each acre of permanent pool is prohibitive compared to the approximate
cost of $100,000 for each acre pond in a new development.
Table 5. Conceptual unit capital and maintenance cost for selected stormwater treatment practices.
Stormwater treatment practice
Unit capital cost, $
Annual maintenance cost, $
Source area practices
MCTT
Stormceptor
Delaware Filter
Bioretention
Trench
Pavement
Broom Sweep
High Sweep
Gardens
38,000 / acre of imper. '
15,000 / acre of imper.1
1 7,500 / acre of imper.1
20/ft2 of practice or
44,000/acre of imper.1
1 8/ft2 of practice or 88/ft
of trench
85,500/acre of practice
39/curb mile
41/curb mile
6/ft2 of practice
2,200/practice
500/practice
1 ,700/practice
2/ft
6/ft
290/ac of practice
Included in capital
Included in capital
0
Regional Practices
Ponds (with no land cost)
Ponds (with land cost)
Ponds (with land cost & buildings)
383,000/acre of pond
980,000/acre of pond
1 ,935,000/acre of pond
3,500/acre of pond
3,500/acre of pond
3,500/acre of pond
1. Imper. = connected imperviousness.
Source Area Practices
Nine source area practices were selected that best met our criteria (Table 4). The TSS reduction capabilities
of the practices have been verified by at least one monitoring study (Winer, 2000, Shoemaker and others,
2000; Bell and others, undated; Young, 1996; National Stormwater, 1999). The TSS removal values
include the losses of pollutant load if the practice has a bypass mechanism. Although most of the practices
do not have many test results, the available results indicate most of the practices can achieve a high level of
suspended solids reduction. All the proprietary and nonproprietary practices that are available should have
an efficiency that falls somewhere in the range of efficiencies we used in the demonstration.
The Stormceptor™ represents many of source area practices with a moderate level of suspended solids
reduction, while the multi- chamber treatment tank (MCTT) represents the practices with a high level of
suspended solids reduction. Test results indicate the Stormceptor™ should reduce the annual suspended
solids load by about 30% (Waschbusch, 1999). Many single chamber practices relying on settling will
10
-------
probably achieve similar levels of reduction. Many multi-chamber practices that include filtration have a
better chance of achieving the 80% reduction in annual suspended solids loads observed for the MCTT
(Corsi and others, 1999). Eighty percent is probably near the maximum annual load reduction we can
expect for a source area treatment practice, because the practices that have 98% removal efficiencies, such
as the MCTT, usually bypass some of the higher flows. It is assumed most devices are designed to bypass
some flows for rainfalls greater than about 1.25 inches in 24 hours.
Reported TSS reduction for the old style broom street sweeper is low at 20% (Bannerman, 1983,
Sutherland, 1999). Street sweeping has the potential to be a very effective practice, because the source
areas that can be swept (parking lots and streets) are the most important sources of TSS. Changes to
sweeping schedules and types of machines would be much less disruptive to the public than any other
source area practice. New types of street sweepers appear to be more effective (Sutherland, 1999). High
efficiency street sweepers should be able to reduce TSS loads from residential streets by at least 60%.
These numbers are based on estimates from a calibrated version of the SEVITPM model. The same type of
high efficiency street sweepers should be able to reduce the TSS loads from freeways by about 45%
(Martinelli, 2002).
The selected source area practices cover a range of treatment processes. Bioretention, MCTT, infiltration
trenches (trench), rain gardens (gardens), and the Delaware perimeter sand filter (Delaware filter) all use
settling and filtration to remove solids from stormwater. Infiltration also lowers loads by reducing runoff
volumes. Infiltration is a key element of trenches, bioretention, gardens, and porous pavement (pavement).
We have experience in Wisconsin with all of the selected source area practices except for bioretention and
Delaware sand filters. Personnel communications with cities supporting the source area practices indicate
they are mostly happy with their performance. Public works people in Osceola, Wisconsin are telling us
they are happy with the performance of their high efficiency street sweeper. Two MCTTs installed in
different cites seem to performing well. We are not aware of any complaints about the several Stormceptors
that are installed around Wisconsin. Most of the porous pavement installations seem to be in the form of
paver blocks. Some people have observed failures of infiltration trenches. These failures appear to have
been caused by clogging during the construction process. Homeowners have reported they are very satisfied
with the operation of their rain gardens.
At best, the available cost information can only be used for conceptual purposes (Shoemaker, 2000;
Southeastern Regional, 1991) (Table 5). Obviously, the cost will vary with each site depending on factors
such as obstacles to installing the practice, cost of the land, and how difficult it is to connect the practice to
existing conveyance systems. Existing utilities have already increased the cost of some of our retrofit
efforts in Milwaukee. A need to support truck traffic and the presence of underground pipes increased the
cost of installing a MCTT in a city maintenance facility. The cost of connecting the existing plumbing to
the practices was the major part of the construction cost of installing two source area controls at a freeway
site. Conceptual is good enough, though, for a demonstration.
Unit capital and maintenance cost calculation varies from practice to practice (Table 5). Some of the
literature provides the cost in terms of the amount of drainage area to the practice, while other cost are
determined from the size of the practice. When more than one cost value was available we always selected
the higher value. For older cost values we assumed an inflation of 3% each year. Some of the practices
share similar costs. For example, the MCTT and bioretention cost about $40,000 for each acre of
imperviousness in the drainage area. Surprisingly, the Delaware filter achieves about the same solids
11
-------
reduction as the MCTT and bioretention, but only costs about $17,500 for each acre of imperviousness.
This is one reason we included the Delaware filter in our demonstration.
Location and Sizing of the Practices
Before we could use SLAMM to determine the benefits of installing each type of source area STP, we had
to match each practice to the appropriate source area(s). Street sweeping is an obvious match for streets in
the three landuses contributing the largest amount of TSS (Table 6). All of the source area practices except
street sweeping and rain gardens are applied to parking lots in the commercial and institutional areas.
Practices like the MCTT and bioretention are recommended for relatively small drainage areas such as a
parking lot. Not enough information is available about treatment levels and cost to include street sweeping
Table 6. Sizing information for selected stormwater treatment practices.
Stormwater
Treatment Practice
Source area
treated
Dimensions each site
(ft)
Total area of practice or
area of connected
impervious draining to
practice (ac)
Estimated
number of
treatment sites
Residential
Rain Gardens
Bioretention
MCTT
Stormceptor
Delaware Filter
Broom Sweep
High Sweep
Lawn & roof
All
All
All
Driveway
Streets
Streets
10 x 17x0.33
15x30x4
1 site/2 ac. of imper.
1 site/2 ac. of imper
1 site/driveway
1 /week for 30 weeks
1 /week for 30 weeks
47.6
27.5
5631
5631
921
-
-
12,200
2,666
281
281
6,100
41 102
41 102
Commercial/Institutional
Infiltration Trench
nfiltration Trench
Bioretention
Porous Pavement
MCTT
Stormceptor
Delaware Filter
MCTT
Stormceptor
High sweep
Broom Sweep
Parking lots
Roofs
Parking lots
Parking lots
Parking lot
Parking lot
Parking lot
All
All
Streets
Streets
5 x 200 x 4
5 x 200 x 4
15x30x4
-
1 site/ 2 ac imper.
1 site/2 ac imper.
-
1 site /2ac imper.
1 site / 2 ac imper.
1 /week for 30 weeks
1 /week for 30 weeks
6.2
2.2
15.6
306
31 01
31 01
31 01
5301
5301
-
-
270
96
1,500
20
155
155
55
265
265
9902
9902
Freeway
Infiltration Trench
MCTT
Stormceptor
High sweep
All
All
All
Freeway
5 x 200 x 4
1 site / 2ac. imper.
1 site / 2 ac. imper.
1 /week for 30 weeks
1.74
91
91
-
75
45
45
141"
Regional
Ponds
All
8.5 ac.
34
4
1. Acres of connected imperviousness.
2. Total curb miles each year.
12
-------
as a parking lot practice. Together lawns and roofs produce enough of the TSS load (12%) to include them
in the analysis of source practices. Residential lawns and roofs are treated with rain gardens and
commercial roofs are treated with infiltration trenches.
To understand the maximum possible benefit of using an STP in the three landuses, some of the source area
practices are applied to all the source areas in each landuse. By installing MCTTs, Stormceptors, and
bioretention systems near or under the streets they should be in a position to treat the runoff coming from all
the source areas. It is assumed that some of the water is bypassed for these source area practices.
For example, we assumed 2,666 bioretention systems or 27.5 acres of treatment surface area is required to
treat all the source areas in residential landuses (Table 6). Each bioretention site would cover a surface area
of at least 15 feet wide and 40 feet long and the practice would be installed next to the street in the right of
way. It is assumed the people living on the street are responsible for the maintenance of the bioretention
plants.
In most cases it seems impractical to assume enough source area practices would be installed in a
subwatershed to act as a regional practice. But some examples already exist in this country where cities
have installed source area practices in the public right-of-way to control the amount and quality of runoff
from all the source areas. Rain gardens are already being installed along residential streets in the
Maplewood, Minnesota (Cavett, 2002). They are also being installed as part of street drainage system
during street reconstruction projects. Bioretention swales have been installed along a street in Seattle,
Washington (http://www.ci.seattle.wa.us/util/urbancreeks/SEAstreets/history.htm) to treat the runoff from
the two year return interval storm. They project that the addition of bioretention swales will not
significantly increase the cost of street reconstruction projects.
For the regional practice we assumed that there is one detention pond for each of the four subwatersheds.
Since this is a demonstration effort, it is not necessary to match the number of ponds to the number of
available sites. It is very likely the total number of ponds would exceed four, if a number of ponds is
needed in each subwatershed to overcome the constraints of each site.
Among the selected practices, SLAMM is able to predict the TSS reduction of street sweeping, porous
pavement, rain gardens, bioretention systems, infiltration trenches, and detention ponds. Iterations of the
model are used to determine the optimum size of rain gardens, porous pavement, bioretention systems and
infiltration trenches (Table 6). Reported TSS removal values for the other practices are inserted directly into
the model. The model accepts the reported values in the "other" option for source areas, the conveyance
system, and the outfall controls.
Total Suspended Solids Reductions Estimated for Individual Practices
Evaluation of the individual source area practices produced only two examples of a practice achieving about
a 40% reduction in annual TSS loads to Lake Wingra (Table 7). Bioretention systems and MCTTs located
to control all the residential source areas are those two practices. They worked because the residential
landuse represents about 50% of the TSS load to Lake Wingra and they have a TSS removal capability of
80%. The other applications of the source area practices are usually treating landuses or source areas that
start with less than 40% of the annual TSS load. One exception is streets with 40% of the annual TSS load,
but a practice applied to streets would need almost a 100% removal of TSS to achieve the goal. Source area
practices will have to be combined to offer more ways for source area practices to achieve a 40% reduction.
13
-------
Since the ponds are designed to achieve an 80% reduction it is not surprising that the regional practice
achieved the TSS reduction goal (Table 7).
Table 7. Reduction in annual TSS loads to Lake Wingra for stormwater treatment
practices applied to four subwatersheds
Practice
Broom Sweep
Delaware Filter
Gardens
Stormceptor
High Sweep
MCTT
Bioretention
Broom Sweep
Trench
High Sweep
Stormceptor
Stormceptor
Trench
Bioretention
MCTT
Delaware Filter
Pavement
MCTT
Stormceptor
High Sweep
MCTT
Trench
Ponds (with land cost)
Source area treated
Residential
Streets
Driveways
Lawn & roof
All
Streets
All
All
Commercial/Institutional
Streets
Roofs
Streets
Parking lot
All
Parking lot
Parking lot
Parking lot
Parking lot
Parking lot
All
Freeways
All
Freeway
All
All
Regional
All
Annual TSS reduction,
%1
4
7
9
16
17
38
41
1
2
5
7
11
12
13
17
19
19
27
1
4
5
6
74
1. Percent of load for all eight subwatersheds, i.e. entire load to Lake Wingra.
Their actual reduction is 74% because we divided the total suspended solids load reductions by the solids
loads for the entire watershed, not just the four sub-watersheds where they were applied. Detention ponds
could, therefore, be located to serve less of each subwatershed and still meet the TSS reduction goal for the
entire watershed.
Cost Comparisons Between Source Area and Regional Practices
To make a valid comparison between source area practices and regional practices it was important to select
configurations of the practices that achieved about a 40% reduction in annual TSS loads. From the analysis
of the individual source area practices we discovered it is necessary to try combinations of them to have
14
-------
more than a couple of alternatives that achieve the 40%. These alternatives could also be more reasonable
than applying a source area practice to all the source areas in a landuse, which is needed to achieve a 40%
reduction with the MCTT and bioretention. Since detention ponds were determined to achieve a 74% in
annual TSS loads to Lake Wingra, it is possible to achieve the 40% reduction by assuming less of each
subwatershed drains to each pond. This not only has the effect of reducing the TSS removal by the ponds,
but also reduces their costs.
Combinations of Source Area Practices Determined to Achieve 40% Reduction
To evaluate the benefits of combining the source area practices, the practices were arranged into about 80
combinations. One important consideration is to avoid redundant practices, such as using street sweeping
and the MCTT under the street in the same area. After eliminating all the combinations that were lower
than 40% or higher than a 45% reduction, we were left with a set of about 15 combinations. We dropped
about six more combinations for different reasons. For example, we eliminated all those combinations with
trenches on the parking lots because we thought this practice would be hard to implement due to the
potentially high cost of pretreatment. Porous pavement is not included because of the potential disruption
and cost associated with removing the existing pavement. Nine combinations of source area practices met
our criteria for percent TSS reduction and reasonableness (Table 8).
All of the combinations included at least one source area practice in the residential area. To make them
more reasonable, MCTTs and bioretention systems were applied to one-half the area. By treating one-half
the area the number of bioretention systems required drops from 2,666 to 1333. Rain gardens were designed
to treat one-half of the roof and lawn area. High efficiency sweeping is an important part of all the
combinations except one. The 40% could not be achieved for the combinations without some kind of
source area practice on the parking lots. In every case one of three source area practices (bioretention
systems, MCTTs, and Delaware Perimeter Sand Filter) was designed to treat the entire area for each parking
lot. Infiltration trenches along the freeway are the most effective freeway practice at a 6% TSS reduction, so
they are included in three of the combinations.
Selection of the Most Cost-effective Practices
The most cost-effective practices will achieve the 40% goal for the least amount of cost. To calculate the
cost the capital cost is added to the maintenance cost assuming the practices have a useful lifespan of 20
years. The twenty year cost for the source area practice combinations ranges from $11,000,000 to
$30,000,000 (Table 8). The next cheapest combination of source area practices is almost twice the cost of
the cheapest one. Five of the combinations have a very similar cost. Making a choice between the
combinations with similar cost is more a judgment of which ones are easiest to install.
All of the combinations of source area practices cost less than retrofitting detention ponds if you have to buy
the land and the buildings on the land. To create 40 acres (20 acres of permanent pool and 20 acres of space
around the pool) of open space in a developed area will probably mean buying some of the land that has
buildings on it. In a medium density residential area this is equivalent to about 136 homes. Even if the cost
of retrofitting the detention ponds is cheaper than the source area practices, it is unlikely the people living in
the neighborhoods would tolerate the condemning of 136 homes to build the detention ponds.
If the conceptual costs for the street sweeping and the Delaware filter are realistic than combining these two
practices is the most cost effective approach to reducing the TSS load to Lake Wingra by 40%. Improving
15
-------
the street sweeping program for all the streets and installing Delaware Perimeter Sand Filters on all the
parking lots seems like a reasonable goal for the city. To maximize the benefit of the enhanced sweeping
programs the city should also implement alternate side parking restrictions. The city should be able to meet
this goal by 2013 as required by NR 151. It will probably be more difficult to meet this time frame for
combinations using MCTT, rain gardens, and bioretention systems in the residential areas.
Table 8. Cost of combining stormwater treatment practices to achieve
a 40 to 45% reduction in annual TSS loads to Lake Wingra.1
Practice combinations
High sweep (AII)J + Delaware Filter
(Lots)
Bioretention (1/2 Res) + Delaware
Filter
(Lots) + High sweep (Com/I nst)
High sweep (Res) + MCTT (Lots) +
French (Freeway)
vlCTT (1/2 Res) + Delaware Filter
(Lots) + High sweep (Com/I nst)
Gardens (1/2 Res) + High sweep
(Res) +
Bioretention (Lots) + Trench
(Freeway)
Gardens (1/2 Res) + High sweep
(All) +
vlCTT (Lots)
Bioretention (1/2 Res) + MCTT
(Lots) + High sweep (Com/I nst)
vlCTT (1/2 Res+ Com Lots) + High
sweep (Com/I nst)
Bioretention (1/2 Res) + Trench
(Com/I nst roof) + Bioretention (Lots)
t- Trench (Freeway) 4
Detention Pond (treat 1/2 of area) 4
Detention Pond (treat 1/2 of area) 5
Total cost for
twenty
years1 ($)
11,460,000
20,420,000
19,860,000
21 ,540,000
25,240,000
26,020,000
27,940,000
29,060,000
30,080,000
19,260,000
36,800,000
Annual cost
($)
573,000
1,021,000
993,000
1,077,000
1,262,000
1,301,000
1,397,000
1,453,000
1,504,000
963,000
1,840,000
Additional utility fee
for households in
Madison,
$/household/year.2
6
10
10
10
12
13
14
14
14
9
18
Capital and maintenance cost included.
2 Annual cost divided by 46,553 household paying stormwater utility fee in City of Madison and multiplied by 45% to
adjust for percent of total utility revenues paid by homeowners.
3 Does not include freeways.
4 Includes cost of land.
5 Includes cost of land and buildings.
Although the annual cost of the cheapest combination of practices is only about $600,000, the impact of this
cost can only be measured in terms of how much it will cost each tax payer. We are able to do this for the
City of Madison because the city has created a stormwater utility district. Each household pays a utility fee
of about $36 a year. If we assume the utility district would use any additional fees to pay a bond back over
twenty years, we can calculate the amount of increase to this fee by dividing the annual cost of the practice
by the 46,553 households in the city and multiplying the result by 45%. In the City of Madison the
households are paying about 45% of the utility fee, while the commercial and institutional property owners
are paying the rest. To pay back the cost of the least expensive combination practice combinations would
16
-------
raise the annual fee to each household by $6 (Table 8). If the cost of the practices is assessed to just the
people living in the Lake Wingra watershed the annual cost of the practices for each household would be
approximately 6 times higher than the values in table 8.
The most expensive fee increase would be only $14 each year. All the source area fees are in the range of
the values for the regional practices. Only the taxpayers can answer the question if this too much money to
significantly reduce the pollutant load to Lake Wingra, but it seems like a reasonable fee to pay.
Conclusions
A six step process can be used to determine the most cost effective practices for achieving an annual TSS
load reduction of 40% in an established urban area. An important element of the process is the use of an
urban runoff model to determine the most important sources of the TSS and the levels of TSS reduction
achieved by each management alternative. The steps are valuable for demonstrating the most cost effective
management approach, but do not include the steps for selecting the sites, making final design decisions,
and determining the actual cost for installing the practices at each site.
The goal of reducing the annual suspended loads by 40% to Lake Wingra can be achieved at what seems to
be a reasonable cost to the Madison city taxpayers. A combination of source area practices, such as street
sweeping and Delaware Perimeter Sand Filters on parking lots, are the most cost effective practices. Given
the potentially high amount of disruption caused by the implementation of regional structural practices, a
combination of source area practices also appears to be a more feasible way to achieve the reduction goal.
Not only is a combination of source controls possibly more acceptable to the people living in the watershed,
but also the annual cost to each household could be as little as six dollars. This is much less than retrofitting
detention ponds at eighteen dollars for sites that include the cost of the buildings.
Although the retrofit performance standard in NR 151 is only for TSS, people in Wisconsin recognize there
are other problem pollutants in storm water. Levels of heavy metals, poly cyclic aromatic hydrocarbons
(PAHs), and bacteria in storm water frequently exceed water quality standards (Bannerman and others,
1996). Some of these pollutants will be reduced if the TSS performance standard is achieved. Since
SLAMM is designed to estimate loads for metals and PAHs, future reports will evaluate the sources and
levels of control possible for other problem pollutants.
Both source area and regional practices will take at least ten years to implement. The source area practices
because so many sites need to be installed and the regional practices because so much land must be secured.
Combinations of practices that include street sweeping and source area practices on the parking lots have
the best chance of meeting the retrofit deadline of 2013.
REFERENCES
Bannerman, Roger T., Baun, K., Bohn, M., Hughes, P.E. and Graczyk, D.A. (1983) Evaluation of Urban
NFS Pollution Management in Milwaukee County Wisconsin, Vol I. EPA, Water Planning Division, PB84-
114164.
Bannerman, Roger T., Legg, Andrew D., and Greb, Steven R. (1996) Quality of Wisconsin Stormwater,
1989-94, U.S. Geological Survey, Open-File Report 96-458.
17
-------
Bell, Warren; Stokes, Lucky; Gavan, Lawrence J. and Trong Ngu Nguyen. Undated, Assessment of the
Pollutant Removal Efficiencies of Delaware Sand Filter BMP S, Alexandria, VA: City of Alexandria,
Department of Transportation and Environmental Services.
Cavett, Christpher, 2002, Personal Communication, Maplewood, Minnesota.
Corsi, Steven R., Graczyk, David J., Owens, David W., and Bannerman, Roger T., 1997, Unit-Area Loads
of Suspended Sediment, Suspended Solids, and Total Phosphorus from Small Watersheds in Wisconsin,
U.S. Geological Survey, Fact Sheet FS-195-97
Corsi, Steven R., Greb, Steven; Bannerman, Roger T. and Robert E. Pitt, 1999, Evaluation of the
Multichambered Treatment Train, a Retrofit Water-Quality Management Practice. U.S. Geological Survey
Open-File Report 99-270. Middleton, WI: U.S. Geological Survey. 24p.
Dane County Regional Planning Commission. 1992. Yahara Monona Priority Watershed Project Plan.
Madison, WI: Dane County Regional Planning Commission. 141p.
EPA, 1999, National Stormwater Best Management Practices (BMP) Database. Version 1.0, 1999. Office
of Water, U.S. Environmental Protection Agency, Washington, D.C.
House, Leo B.; Waschbusch, Robert J. and Peter E. Hughes. 1993. Water Quality of an Urban Wet
Detention Pond in Madison, Wisconsin, 1987-88. U.S. Geological Survey Open-File Report 93-172.
Madison, WI: U.S. Geological Survey. 57p.
Martinelli, Thomas J., Waschbusch, R.J., Bannerman, R.T. and Wisher, Ann, 2002. Pollutant Loading to
Stormwater Runoff from Highways: The impact of Freeway Sweeping Program, Wisconsin Department of
Transportation, Research Project ID # 0092-4582.
Pitt, Robert and Voorhees, John. 2002. winslamm.com
Shoemaker, Leslie; Lahlou, Mohammed; Doll, Amy and Patricia Cazenas, 2000, Stormwater Best
Management Practices in an Ultra-Urban Setting: Selection and Monitoring. U.S. Department of
Transportation, Federal Highway Administration Publication No. FHWA-EP-00-002, Office of Natural
Environment, Federal Highway Administration, Washington, D.C. 287p.
Southeastern Regional Planning Commission, 1991. Costs of Urban Nonpoint Source Water Pollution
Control Measures. Technical Report 31. Waukesha, WI: Southwestern
Wisconsin Regional Planning Commission. 109p.
Steuer, Jeffrey; Selbig, William; Hornewer, Nancy and Jeff Prey, 1997, Sources of contamination in an
Urban Basin inMarquette, Michigan and an Analysis of Concentrations, Loads and Data Quality. Water
Resources Investigations Report 97-4242. Middleton, WI: U.S. Geological Survey. 25p.
Sutherland, Roger C. 1999 Water Quality Benefits of High Efficiency Sweeping. Pacific Water Resources,
Inc., Beaverton, Oregon.
18
-------
University of Wisconsin-Madison Water Resources Management 1999 Workshop, 1999, Lake Wingra
Watershed: A New Management Approach. Madison, WLUW-Madison Institute for Environmental Studies,
178p.
Walker, William W., 1990. P8 Urban Catchment Model- Program Documentation, Concord,
Massachusetts.
Waschbusch, R.J., Selbig, W. R. and R. T. Bannerman, 1999. Sources of Phosphorus in Stormwater and
Street Dirt from Two Urban Residential Basins in Madison, Wisconsin, 1994-1995. Water-Resources
Investigations Report 99-4021. Middleton, WI: U.S.
Geological Survey. 47p.
Waschbusch, Robert J. 1999. Evaluation of the Effectiveness of an Urban Stormwater Treatment Unit in
Madison, Wisconsin, 1996-1997. Water Resources Investigations Report 99-4195. Middleton, WI: U.S.
Geological Survey. 49p.
Winer, Rebecca, 2000, National Pollutant Removal Performance Database for Stormwater Treatment
Practices, Center for Watershed Protection Ellicott City, Maryland
Young, G. Kenneth; Stein, Stuart; Cole, Pamela; Kammer, Traci; Graziano, Frank and Fred Bank. 1996.
Evaluation and Management of Highway Runoff Water Quality. U.S. Department of Transportation. Federal
Highway Administration Publication No. FHWA-PD-96-032. Office of Environment and Planning. Federal
Highway Administration, Washington, D.C. 480p.
19
-------
MANAGEMENT STRATEGIES FOR URBAN STREAM REHABILITATION
By Derek B. Booth1'5, James R. Karr2'3, Sally Schauman4, Christopher P. Konrad1'5, Sarah A. Morley2, Marit
G. Larson1'5, and Stephen J. Burges1'5
1. Center for Water and Watershed Studies, 2. School of Aquatic and Fishery Sciences, 3. Department of
Zoology, 4. Department of Landscape Architecture, and 5. Department of Civil and Environmental
Engineering; all at University of Washington, Seattle, WA 98195 USA.
Abstract
Physical, hydrological, social, and biological conditions were evaluated at 45 stream sites in the Puget
Lowland of western Washington, with watersheds ranging in area between 5 and 69 km2 and having urban
development as their dominant human activity. Using the benthic index of biotic integrity (B-ffil) as our
biological indicator, we found a progressive decline in B-ffil with increasing watershed imperviousness but
with large site-to-site differences at any given level of imperviousness in the contributing watershed. This
variability is greatest at low to moderate levels of development; as development intensity increases, the
range of biological conditions narrows. No threshold effects are apparent. Instream biological condition
also varied directly with a new stream flow metric, showing significantly better correlations than with
imperviousness. We also found a wide range of landscape conditions, some very degrading, in the
backyards adjacent to these streams. These data do not suggest that the full range of hydrological and other
ecological conditions can be replaced in a now-degraded urban channel; thus key management tasks are to
identify those watersheds where low urbanization and associated high-quality stream conditions warrant
protection, and to develop a new set of management goals for those watersheds whose surrounding
development precludes complete ecosystem restoration but in which some recovery might be possible.
There is no rational basis to support a common strategy in all watersheds, developed and undeveloped alike.
Introduction
For decades, watershed urbanization has been known to harm aquatic systems. Although the problem has
been long articulated, solutions have proven elusive because of the complexity of the problem, the evolution
of still-imperfect analytical tools, and socio-economic and political forces with different and often
incompatible interests.
Recent Endangered Species Act (ESA) listings of Puget Sound chinook and bull trout, and the potential for
more salmonid listings, have brought new scrutiny to all aspects of the Pacific Northwest's watershed
protection and urbanization-mitigation efforts. Such increased attention is forcing a better articulation of the
goals, the means, and the justification for mitigating the effects of urban development. It also has
highlighted the failure of most stormwater mitigation efforts, not only in the Pacific Northwest but also
across the country, where well-publicized successes are overshadowed by progressive degradation of once-
healthy streams. This degradation has continued, despite sincere but ineffectual efforts via structural "Best
Management Practices" (BMP's), particularly detention ponds, buffer regulations, and rural zoning.
Several factors make Puget Sound ideal for this study. Streams within our study region share relatively
uniform soil, climate, and topography, allowing direct comparisons among streams. The region has a wide
range of watershed development intensities and ages within a circumscribed area, including minimally
20
-------
developed areas that serve as reference sites. All study watersheds have (or once had) diverse natural
biotas, including anadromous salmonids; some moderately developed watersheds still support regionally
valuable biological resources that merit protection and enhancement. Individuals and citizen groups support
protection of aquatic resources in general and salmon in particular, and these groups are the focus of a
variety of local agency efforts to improve public education and stewardship. Finally, major expenditures in
the region are expected over the next decade in the name of "stream enhancement." Improved knowledge
should help direct these outlays to activities most likely to protect the region's aquatic life (including its
iconic endangered salmonids), protect water quality, and thereby maintain cherished components of the
region's quality of life.
Study Sites and Methods
For this study, we focused on 45 sites selected from 16 second and third-order streams in King, Snohomish,
and Kitsap counties (Fig. 1) that share the following physical characteristics: (1) watershed area between 5
and 69 km2; (2) local channel gradients between 0.4 and 3.2 percent; (3) soils, elevation, and climate typical
of the central Puget Lowland; and (4) urban development as the dominant human activity (except in low-
disturbance reference sites).
Figure 1: Map of Puget Lowland showing location of study streams and watersheds.
21
-------
We explored the nature, and the causes, of change to aquatic -system health along a gradient of human
activity. We used common measures of land cover (road density and total impervious area percentages) to
characterize that "human activity." Benthic invertebrates were sampled at each site between 1997 and 1999
(Morley, 2000; Morley and Karr, 2002). Substrate data were collected at 19 of the sites, and hydrologic
analyses were made at the 18 sites located in close proximity to gauging stations without intervening
tributary input (Konrad, 2000). Hydrologic analyses for ten additional lowland streams of similar
characteristics, but some with watershed areas up to 171 km2, were also conducted. The social assessment
had three parts — a survey of stream professionals, an in-depth evaluation of the landscape conditions in
backyards adjacent to streams, and an evaluation of the values held by residents.
Although the hydrologic consequences of urban development are well documented at the scale of an
individual storm (e.g., Hollis, 1975), consequences over longer periods are less well known. Because we
expected the latter effects to be especially important to the biota of streams, we applied a hydrologic statistic
to represent the annual distribution of storm and baseflow patterns: namely, the fraction of a year that the
daily mean discharge exceeds the annual mean discharge (TQmean).
was calculated for each of the 18 streams by first determining the fraction of the year that the daily
mean discharge (Q daily) exceeded the annual mean discharge (Qmean) for each year of record for each stream.
TQmean was then calculated as the average annual fraction of a year that Qdaiiy exceeds Qmean, which averages
about 30 percent of the time across this range of Puget Lowland streams.
Results
Biological Condition at Multiple Land-Cover Scales
Relationships between land cover and biological conditions display several trends. As a group, our study
sites display a progressive decline in B-ffil (Karr, 1998) with increasing urban development, although large
site-to-site differences exist at any given level of imperviousness in the contributing watershed (Fig. 2). This
variability is particularly evident at low to moderate levels of development, where almost any degree of
biological condition may be associated with a given level of imperviousness (see also Karr and Chu, 2000).
As development intensity increases, the range of biological conditions narrows until, in the most urban of
our watersheds, conditions are uniformly poor.
Across all study sites, urban land cover (i.e. the combination of "intense," "grassy," and "forested" urban
categories) correlated approximately equally well with B-IBI at each of three spatial scales: subbasin (i.e.,
the entire watershed area upstream of the sample point; r = -0.73, p < 0.001), riparian (a 200-m-wide buffer
on each side of the stream extending the full length of the upstream drainage network; r = -0.75, p < 0.001),
and local (a 200-m-wide buffer on each side of the stream extending 1 km upstream; r = -0.71, p < 0.001)
(Morley and Karr, 2002). In our data set, riparian and subbasin land cover closely correlated with each
other (r = 0.98, p< 0.001).
22
-------
I £
c -a
I 2
ai =>
4= <"
OT S 20
E
10
10
20
60
30 40 50
Urban Development
(as measured by % total impervious area in contributing watershed)
70
Figure 2: Relationship between watershed urbanization and stream health (i.e. biological condition) for our study
streams as measured by total impervious area in the watershed upstream of benthic invertebrate sampling sites.
Stream health is measured using the benthic index of biological integrity (B-IBI); samples collected 1997, 1998,
and1999.
Hydrologic Changes
Hydrologic effects of urban development are evident, even amidst the variability generated by
physiographic differences among the basins in the Puget Lowland. In urban streams (road density >6
km/km2), the fraction of time that the mean discharge is exceeded (TQmean) generally is less than 30% (and
all < 32%), while in suburban streams (road density <6 km/km2), TQmean is generally greater than 30% (and
all but one > 32%; Fig. 3). For WY 1989 to 1998, the mean value of TQmean for 11 urban streams was
smaller (0.29) than for 12 suburban streams (0.34). The difference is statistically significant (p < 0.01 using
Student's t-test of samples with equal variance). Independent of urban development, however, larger
streams typically have more attenuated stream flow patterns than smaller streams and so higher values of
Tqmean (Konrad and Booth, 2002). Thus TQmean may only be a reliable indicator of urban development if
stream basins are similar in drainage area and other physiographic factors.
The biological conditions of streams varied directly with this stream flow metric (Fig. 4), with significantly
better correlations than for simple land-cover metrics (see Fig. 2). Variability in B-IBI is still significant,
however, because flow regime is only one factor controlling biotic integrity; for any value of TQmean, the B-
IBI range is about 10.
23
-------
c
ro
01
0.45
0.40 -
0.35 H
0.30 -
0.25
• Suburban
A Urban
A
A
A A
10
Road density (km/km )
Figure 3: Fraction of year that mean discharge rate is exceeded (TQmean) as a function of watershed road density.
eg
CD
50
40
30
20
10
0.25 0.30
0.35
I Qmean
0.40 0.45
Figure 4: Benthic index of biological integrity (B-IBI) plotted against fraction of time that daily mean discharge rate
exceeds annual mean discharge rate (TQmean) for Puget Lowland streams with biological and hydrologic data.
24
-------
Social Assessment
The social assessment yielded a rich array of results. The most insightful was finding a wide variation in
backyard conditions where streams were located. These subject properties ranged from those adjacent to
streams, located in watersheds having a county-funded steward who provided extensive public education, to
backyards in neighborhoods with little community awareness of the stream at all. In all locations the range
of conditions varied from benign neglect to severe, "ecopathic" destruction of the landscape adjacent to the
stream. Broad social measures do not explain these differences in behavior, but the influence of these
actions on stream health (whether benign or damaging) was locally very significant.
Discussion
Correlations between watershed development and aquatic-system conditions have been investigated for over
two decades. Klein (1979) published the first such study, where he reported a rapid decline in biotic
diversity where watershed imperviousness much exceeded 10 percent. Steedman (1988) believed that his
data showed the consequences of both urban land use and riparian condition on instream biological
conditions. Later studies, mainly unpublished but covering a large number of methods and researchers, was
compiled by Schueler (1994). Since that time, additional work on this subject has been made by a variety of
Pacific Northwest researchers, including May (1996), Booth and Jackson (1997), Karr (1998), and Morley
and Karr (2002)
These data have several overall implications:
• "Imperviousness," although an imperfect measure of human influence, is clearly associated with stream-
system decline. A wide range of stream conditions, however, can be associated with any given level of
imperviousness, particularly at lower levels of development.
• "Thresholds of effect," articulated in some of the earlier literature (e.g., Klein, 1979; Booth and Reinelt,
1993) exist largely as a function of measurement (im)precision, not an intrinsic characteristic of the
system being measured. Crude evaluation tools require that large changes accrue before they can be
detected, but lower levels of development may still have consequences that can be revealed by other,
more sensitive methods. In particular, biological indicators (e.g., Figure 2) demonstrate a continuum of
effects, not a threshold response, resulting from human disturbance (Karr and Chu, 2000).
• Although direct correlation of imperviousness with biological health is overly simplistic, imperviousness
is a useful index of human activity in a watershed because it provides a gross measure of the watershed
area appropriated by people, and thus it functions as a first-order indicator of human influence on
selected processes supporting stream ecosystems. Many of the changes that degrade streams are
progressively more likely to occur as human activity increases (Booth et al, 2002). The fraction of
impervious area is not a suitable surrogate of stream health, however, because this metric neither
captures nor diagnoses all major causes of stream degradation; neither does it provide an adequate guide
to effective solutions. In combination with other measures and analyses, however, it can enhance both
river protection and restoration.
25
-------
Management Implications
Development that minimizes the damage to aquatic resources cannot rely on structural BMP's, because
there is no evidence that they can mitigate any but the most egregious consequences of urbanization.
Instead, control of watershed land-cover changes, including limits to both imperviousness and clearing,
must be incorporated (see also Horner and May, 1999). We anticipate needing all of the following elements
to maintain the possibility of effective protection:
• clustered developments that protect half or more of the natural vegetative cover, preferentially in
headwater areas and around streams and wetlands to maintain intact riparian buffers;
• a maximum of 20% total impervious area, and substantially less effective impervious area through the
widespread reinfiltration of stormwater (Konrad and Burges, 2001);
• on-site detention, realistically designed to control flow durations (not just peak discharges);
• riparian buffer and wetland protection zones that minimize road and utility crossings as well as overall
clearing;
• no construction on steep or unstable slopes; and
• a program of landowner stewardship that recognizes the unique role of adj acent private property owners
in maintaining or degrading stream health.
Past experience suggests that each of these factors is important. However, we still lack empirical data on
the response of aquatic resources to such "well-designed" developments. Therefore, these recommendations
are based only on extrapolations, model results, and judgement; they have yet to be tested. Where
development has already occurred, these conditions clearly cannot be met and different management
objectives are inescapable: many, perhaps all, streams in already-urban areas cannot be truly protected or
restored, and a significant degree of probably irreversible stream degradation is unavoidable in these
settings.
Our detailed analysis of one feature, flow regime, demonstrates the importance of this particular aspect of
the aquatic system. Hydrologic alteration is ubiquitous in all urban watersheds, and flow regime is a key
determinant of ecological health and biological condition. Stream conditions are not solely determined by
flow regime, however, and flow regime is not solely determined by urban development—intrinsic watershed
characteristics (watershed geology, soil permeability and depth, topography, channel network, climate) are
also relevant. Thus no single watershed indicator can predict flow regime or the consequences of its change
on stream conditions, even a metric that provides ecologically useful measures of the variability of stream
flow. A new paradigm that systematically ignored water chemistry or the effects of alteration of stream
channels, for example, would be no more defensible than previous regulatory mandates that focused only on
these parameters.
We cannot find any basis to expect that the full range of hydrological and other ecological conditions can be
replaced in a now-degraded urban channel (Fig. 5). The key tasks facing watershed managers, and the
public that can support or impede their efforts, are therefore (1) to identify those watersheds where existing
low urbanization, and associated high-quality stream conditions, warrant the kinds of development
conditions that may protect much of the existing quality of these systems; and (2) to develop a new set of
management goals for those watersheds whose surrounding development precludes significant ecosystem
restoration but in which some recovery might be possible. Where urban development is virtually complete,
our results (and common sense) suggest that neither widespread riparian-corridor replanting nor extensive
26
-------
hydrologic rehabilitation of the watershed are feasible or could achieve great biological improvements.
Stream-enhancement efforts can still be important and worthwhile, for both in-stream biota and the people
that live in their watersheds. There is no rational basis to support a common strategy in all watersheds,
developed and undeveloped alike.
Biological Condition of Puget Lowland Streams
Management Strategies
O 40
eg
m so
UPPER LIMIT CF CBSERVED CATA
a-v,
a a
Q
Q J
o a
a o
O O'j
o o a
a
o o
10 20 30 40
Protection
Rehabilitation
CONDITIONS
NOT OBSERVED
IN THIS REGION
, Stewardship
Urban Development (TIA in watershed)
10 20 30 * 50 6
Urban Development (TIA in watershed)
Figure 5: Management strategies as suggested by the distribution of B-IBI data as a function of the % total
impervious area (TIA) in the contributing watersheds of our study. Although management goals are commonly
articulated for the upper right-hand corner of these graphs (i.e. high-quality streams in highly urbanized watersheds)
we find no evidence, and thus little hope, that this does or can occur.
Acknowledgments
This work was supported by the United States Environmental Protection Agency and National Science
Foundation Water and Watersheds Program, EPA Grant R82-5284-010, for which this paper is a partial
summary of the final report (Booth et al, 2001). Additional support for this project came from King County
and the Stormwater Technology Consortium of the Center for Urban Water Resources Management (for
DBB) and from the Consortium for Risk Evaluation (CRESP) by Department of Energy Cooperative
Agreements #DE-FC01-95EW55084 and #DE-FG26-OONT40938 (for JRK). We are grateful to
collaborative efforts from David Hartley, Rhett Jackson, Patricia Henshaw, Erin Nelson, and Jenna Leavitt;
Wease Bollman and Robert Wisseman for taxonomic consultation; Laura Reed and Kris Rein for field and
lab assistance; and Marina Alberti, Kristina Hill, and other members of the University of Washington
PRISM project for guidance in matters of GIS.
References
Booth, D. B., Hartley, D., and Jackson, C. R. 2002. Forest cover, impervious-surface area,
and the mitigation of Stormwater impacts. Journal of the American Water Resources Association 38: 835-
845.
Booth, D. B., and Jackson, C. R. 1997. Urbanization of aquatic systems—degradation thresholds, Stormwater
detention, and the limits of mitigation. Water Resources Bulletin 33: 1077-1090.
27
-------
Booth, D. B., Karr, J. R., Schauman, S., Konrad, C. P., Morley, S. A., Larson, M. G., Henshaw, P., Nelson, E.,
and Burges, S. J. 2001. Urban stream rehabilitation in the Pacific Northwest. Final report to U. S. EPA,
grant no. R82-5284-010, 78 p.
Booth, D. B., and Reinelt, L. E. 1993. Consequences of urbanization on aquatic systems—Measured effects,
degradation thresholds, and corrective strategies. Watersheds '93, Conference sponsored by U. S.
Environmental Protection Agency, Alexandria, VA, March 21-24, p. 545-550.
Hollis, G. E. 1975. The effect of urbanization on floods of different recurrence interval. Water Resources
Research 11:431-435.
Horner, R. R., and May, C. W. 1999. Regional study supports natural land cover protection as leading Best
Management Practice for maintaining stream ecological integrity. In Comprehensive Stormwater and
Aquatic Ecosystem Management, conference papers. New Zealand Water and Wastes Association, 22-
26 February 1999, p. 233-247.
Karr, J. R. 1998. Rivers as sentinels: Using the biology of rivers to guide landscape management. Pages
502-528 in R. J. Naiman and R. E. Bilby, eds. River Ecology and Management: Lessons from the
Pacific Coastal Ecosystems. Springer, NY.
Karr, J. R. and Chu. E. W., 2000. Sustaining living rivers. Hydrobiologia 422/423: 1-14.
Klein, R. 1979. Urbanization and stream quality impairment. Water Resources Bulletin 15: 948-963.
Konrad, C.P. 2000. The frequency and extent of hydrologic disturbances in stream in the Puget Lowland,
Washington. Ph. D. Dissertation, Department of Civil Engineering, University of Washington, Seattle,
Washington, USA.
Konrad, C. P., and Booth, D. B. 2002. Hydrologic trends resulting from urban development in western
Washington streams: U.S. Geological Survey Water-Resources Investigation Report, 02-4040.
Department of the Interior, Washington, DC, USA.
Konrad, C. P., and Burges, S. J. 2001. Hydrologic mitigation using on-site residential storm-water detention.
Journal of Water Resources Planning and Management 127: 99-107.
May, C. W. 1996. Assessment of cumulative effects of urbanization on small streams in the Puget Sound
Lowland ecoregion: implications for salmonid resource management. Seattle, University of Washington,
Department of Civil Engineering, Ph.D. dissertation, 383 p.
Morley, S. A. 2000. Effects of urbanization on the biological integrity of Puget Sound lowland streams:
restoration with a biological focus. Seattle, University of Washington, M.S. Thesis.
Morley, S. A., and Karr, J. R. 2002. Assessing the biological health of urban streams in the Puget Sound
Basin. Conservation Biology. In press.
Schueler, T. R. 1994. The importance of imperviousness. Watershed Protection Techniques 1: 100-111.
Steedman, R. J. 1988. Modification and assessment of an index of biotic integrity to quantify stream quality in
Southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences 45: 492-501.
28
-------
REMEDIATION OF STORMWATER RESIDUALS DECANT
WITH HYDROCOTYLE RANUNCULOIDES
Katie Bretsch
Program Development Assistance
Portland, Oregon
Abstract
A stormwater residuals decant treatment regime employing floating marsh pennywort, Hydrocotyle
ranunculoides, is apparently effective at remediating lead-contaminated suspended solids, 25 microns and
less, after one year's experience in Portland, Oregon.
Gravity settling provided by Portland's existing stormwater sediment dewatering facility does not give
sufficient pollutant removal, and Portland experienced occasional exceedances of local pretreatment limits
for lead. In March of 2001, Portland began a full-scale trial of stormwater residuals decant treatment using
marsh pennywort, or Hydrocotyle ranunculoides. This free-floating aquatic plant is locally acceptable for
aquatic landscaping and needs no special control.
First-year review found this project apparently successful and very inexpensive. Preliminary second-year
data continues to show promise and minimal cost.
Project Context
Portland, Oregon maintains a separate stormwater collection and treatment system, which includes over
15,000 sumps and sedimentation manholes that drain only curbed and guttered urban streets. Over 1,800
metric tons of stormwater residuals are recovered by vacuum eductor truck (Vactor®) from these facilities
annually. These residuals are contaminated with common urban stormwater pollutants, most prominently
TPH, lead and cPAHs. The contaminants are mostly fixed — adsorbed to the fine soils which dominate these
residuals (Bretsch, 2002). On average, fine particles 31.2 microns and less account for 22% of residual
solids particle counts.
The residuals are recovered along with substantial amounts of standing stormwater and injected chlorinated
tap water. They are discharged onto sloped pads at the City's Inverness Stormwater Sediment Dewatering
Facility from vacuum eductor trucks at about 90% water by weight, or pea soup consistency. After
dewatering to about 25% water by weight, or dry enough to pass a "paint filter test," the material is removed
for thermal remediation and recycling.
Decant off Portland's tennis court size Vactor® dumping pads flows through sloped channels with weirs of
wood and screen fabric intended to catch the large floatables, then through a system of ductile iron pipe and
shallow below-ground sedimentation manholes to a two-celled settling tank made from a section of the old
aeration settling basin of an abandoned wastewater treatment plant. An overflow stand pipe in the second
cell allows continuous discharge to the City's sanitary sewer system.
29
-------
This dewatering process yields about 684,000 liters of decant annually. The decant is pretreated prior to
discharge into the City's sanitary sewer system in order to protect the City's wastewater system biosolids
quality, a critical City objective.
The decant carries ultra-fine suspended solids which are negatively charged and resistant to settling by
gravity (Collins, 1999; Ghezzi, M., Collins, L, Moore, L, Bretsch, K., and Hunt, L., 2001). A $300,000
facility improvement provided additional gravity settling. But, gravity settling alone failed to provide
consistent enough pollutant removal at the desired levels of operation. In consequence, dewatering facility
decant occasionally exceeded local pretreatment limits for lead of 0.7 mg/L. The City's goal is to
consistently meet a 0.2 mg/L limit. In response, the City began plans for a second six-figure facility
expansion project to provide additional gravity settling capacity.
Working in cooperation with the Oregon Department of Transportation (ODOT) and the Oregon
Department of Environmental Quality (DEQ) under the auspices of the Federal Highway Administration
(FHWA) funded ODOT Roadwaste Research Project, Portland also explored methods for achieving better
removal of decant solids with the existing facility. Because Portland's stormwater Vactor© waste represents
the worst case for stormwater residuals quality in Oregon, finding a best value solution to Portland's
Vactor© waste decant pretreatment problem promised to be helpful to roadwaste management agencies
elsewhere, as well.
Portland conducted chemical flocculation trials as one alternative, and trial results are documented in the
Phase Two Report of the ODOT Roadwaste Research Project (Ghezzi, M., Collins, L, Moore, L, Bretsch,
K., and Hunt, L., 2001). Electroflocculation, as demonstrated by Dennis lurries, PE, of the Oregon DEQ
using stormwater with suspended fines from construction site erosion (lurries, 2000), was also considered.
These methods were found practicable, but the projected treatment costs of about US$0.38 per liter were
deemed prohibitive.
Reasoning that only a marginal increase in decant quality was required, that some of the stormwater
treatment value provided by plants in a constructed wetland might occur if a large enough planting could be
propagated and maintained in the decant tank, that the potential benefits were high and the cost of failure
was low, the author initiated a search for suitable aquatic plants.
Voluntary duckweed (Lemna) colonies had previously appeared in the tank, but were flushed through the
system during rain events. Pennywort was selected for trial because it is free-floating, easily contained, a
locally acceptable native, and available. Risks of escape were well considered. Because it propagates by
budding, seed distribution by wind or animal life is not a risk.
Implementation
A trial of phytoremediation was begun in May of 2001 by introducing a 19-liter starter bucket of the
floating marsh pennywort plant material into the first cell of the decant tank (Figure 1.). H. ranunculoides is
a native, free floating perennial found throughout the United States (PLANTS Database, 2002). The plant
material was gleaned from an ornamental pond maintained on the grounds of the City's Columbia
Boulevard Wastewater Treatment Plant.
30
-------
Figure 1 Photograph shows pennywort growing in the first cell of Portland's Vactor© waste decant tank. About three
months after its initial introduction into the tank, the pennywort has formed a dense colony about 2.5 m square.
Plastic roll screening material with 1.3 cm openings and non-woven filter fabric of the kind used in erosion
control were used to confine the plant material in the first cell of the tank.
The plant material thrived and filled out the cell by July of 2001 (Figure 2). So far, the plant material has
proven hardy in this implementation. Just as in an ornamental planting, it pales and slows its growth during
the winter months, but no substantial winter dieback has occurred. It also pales and slows its growth during
the warmest sunny summer months, when decant tank flow is warmed and reduced by evaporation.
Figure 2. Photograph shows a dense matt of the vigorous pennywort completely covering the surface of the first cell
of Portland's Vactor© waste decant tank in August, 2002.
-------
To further test the technology and compensate for variables such as weather and changes in Vactor®
cleaning program activity which couldn't be isolated in this trial, additional plant material was introduced
into the second cell of the decant tank starting in the Spring of 2002. A full second year review could be
conducted in June of 2003.
Operation
No appreciable additional operating needs or costs were presented by the introduction of plant material into
the decant treatment stream during the trial. Thinning of the plant colony may eventually be needed.
Replacement may be required if the very rare extended hard freeze that can occur in Portland proves fatal.
No additional nutrients or other treatments have been required for the health of the plants. As a public health
measure, the tank is treated with Bt (Bacillus thuringiensis) to inhibit mosquito hatching at appropriate
intervals during the warm season.
At about six month intervals, both cells of the tank are drained, and the mucky settled solids are cleaned
from the bottom by Vactorc extraction. The cleanings removed from the tank are placed back onto the
Vactor® dumping pad for dewatering, remediation and recycling.
H. ranunculoides plants are available locally in the Portland, Oregon area from commercial nurseries which
supply native aquatic plants at about US$1.00 per plant. The starting colony for one cell in this trial
probably consisted of the equivalent of 100 commercial plants.
If thinning or removal of the plant material is required, testing to assess pollutant concentrations in the
removed material should be performed. As with any phytoremediation project, disposal of plant materials
should be guided by the findings of appropriate testing.
Monitoring
Accurately measuring the fine, contaminated, negatively-charged colloidal soil particles found suspended in
stormwater Vactorc waste decant proved by itself to be a challenge. The standard pretreatment screening
test for total suspended solids (TSS) proved imperfect, because the filter used to capture solids was found to
have a 25-micron pore size. A particle size study found that over 90% of solids in the decant were under 25
microns.
We considered turbidity (NTU) as an alternative indicator, and rejected it because it also reflects other
factors which couldn't be controlled in this operational setting, such as color from dissolved substances and
non-target particles of organic matter. In the end, we chose total lead (EPA 200.8) as our primary
monitoring parameter. Lead is adsorbed preferentially to the fine solids (Collins, 1999); and lead is the
contaminant of concern for protection of the City's wastewater processes by decant pretreatment.
The progress of the plant colony was observed and photo-documented. Samples representative of the decant
discharges were tested for lead at routine intervals dictated by the City's pretreatment compliance
monitoring program. Older and younger plant material was removed from the tank for close visual
inspection.
32
-------
Discussion
Appropriate and Successful Plant
A perennial native species, H. ranunculoides (Figure 3) requires no special substrate or media. While
relatives of this plant have been identified as invasive pest species in Britain and elsewhere, H.
ranunculoides is listed as endangered in Illinois. In the maritime Pacific Northwest, it is considered a
desirable native species for ornamental propagation. It presents no obvious risk of escape in the setting
under trial. In Portland's trial, it quickly covered the surface area of the tank. It thrived for most of the year,
being somewhat discouraged in growth only during the warmest and coolest months. The test site near
Portland Airport did not experience a hard freeze during the trial period, however.
<
>
(
R i-. •T.-r.^ii V-af, •
Figure 3. H. ranunculoides plant material shown against graph paper to illustrate form and scale. Depth of highly
tangled root mass is about 10 cm. Height of mature stem and leaf is about 20 cm or more above root. Plant colonies
form a dense floating matt.
Volunteer blooms of duckweed (Lemma) had appeared previously in the decant tank, but had been flushed
out by rainfall events. H. ranunculoides is far more easily contained. In fact, it provides some containment
for duckweed, which appeared as a minor voluntary overgrowth in the second summer. Based on visual
observation as well as close handling of removed bucket samples, both the mass and immersed surface of
the pennywort, with its heavy, tangled and tough, almost woody root system, broad leaves and long stems,
dwarfed that of the duckweed in Portland's trial.
33
-------
Plant material has not yet been sampled to determine the amount, if any, of metal hyper-accumulation. From
an operational perspective, this testing will be critical to establish appropriate management of any plant
material wasted from the process
Apparently Successful Remediation
Operationally valuable improvement in decant lead results and visual observation appear to support the
finding that H. ranunculoides is effective at remediating wastewater contaminated with lead bound to ultra-
fine suspended solids in stormwater Vactor0 waste decant. Previously absent flocculation and settling is
observable in the tank and is the presumed method of remediation.
Portland Vac Waste Decant
3 5
2 S
2
1
0 5
0
A
o
8
D 0
o
Tl A'fi
'.
J
I
^A
... fl'7 ........... A< *•
A A ^4>]
A S """^P11
D
O
1 ° " j^ »
CD
I
A * i*j^
^A *A * A
0
a
o o
e ^
0 7
k AA
600
500
d
ft
400 E
so
•D
300 i
200
100
o
Aug-99
Mar-00
Oct-00
Apr-01
Mov-01
May-02
Dec-02
Jun-03
Date
Decant Lead
-Local Limit (a)
-Local Limit (b) + Propagation TS
10 per, Mov. Avg. (TS)
Figure 4. Graph illustrates Portland's Vactor© waste decant total lead and total solids results from January, 2000 to
December, 2002, in relation to local wastewater pretreatment limits and the dates pennywort was introduced into the
decant treatment regime, first in March 2001, and second in May 2002. A stable pattern of lower values has been
coincident with the presence of the pennywort.
Decant monitoring for total lead and total solids shows (Figure 4.) that the presence of pennywort in the
decant treatment stream has been coincident with an operationally significant and stable pattern of lower
values. Prior to the introduction of the pennywort, exceedances of a 0.7 mg/L local limit were a source of
concern. None have reoccurred since the introduction of the pennywort. No exceedances of the lower 0.2
mg/L limit have occurred since Fall of 2001.
34
-------
Minimal Cost
Because the plant material for this trial was obtained as surplus from an ornamental planting, and the plant
has proven both a vigorous grower, and to have no special operational needs in this implementation, the
treatment cost observed in this trial is estimated at less than US $0.01 per liter. Competing commercial
technologies would run about 40 times that, based on Portland's previous trials.
Unanswered Questions
As a field trial, this project was successful enough. However, as a scientific endeavor, this project leaves
many important questions unanswered.
Important variables such as changes in Vactor0 cleaning program activity and rainfall could not be isolated
in this full scale trial. How much remediation value is provided by plants alone in a controlled setting? Are
the author's beliefs about the primary remediation mechanism verifiable in the lab? How much filtration is
occurring? Fines may be adhering and then sloughing off the root surface; but, if so, this is not observable
with the naked eye. Do the plant roots carry a slight positive charge? Will waste plant mass require special
management? The author cannot say.
The data is good enough for operational purposes, but poor by scientific standards. The author has received
expressions of interest from individuals in the academic community to take these investigations further, and
hopes to see these questions answered in the future with their help. The author considers the field trial
results presented in this paper preliminary but promising.
Phytoflocculation ?
The American Heritage Dictionary defines phytoremediation as, "the use of plants and trees to remove or
neutralize contaminants, as in polluted soil or water (American Heritage Dictionary, 2003). In constructed
wetlands and other biologically based wastewater treatment regimes, plants are widely recognized to
provide treatment value via the natural phenomena of rhizofiltration, nutrient consumption and
hyperaccumulation.
The EPA defines flocculation as a "process by which clumps of solids in water or sewage aggregate through
biological or chemical action so they can be separated from water or sewage" (EPA, 2003). Based on field
observation, the author believes that the plant material provided remediation by flocculation of the lead
contaminated ultra-fine suspended solids in this trial. Although the exact mechanism of treatment has yet to
be clearly established in the lab, the author proposes to call this natural phenomenaphytoflocculation.
Thanks
The author would like to thank Linda Dartsch, PE, Manager of the City of Portland, Environmental
Services, Collection Systems Operations and Maintenance Division, for the opportunity to do this trial, Paul
Johnson of Portland's Maintenance Bureau for project assistance, Atina Casas of the City's Environmental
Services Investigations and Monitoring Division for technical assistance, and the staff and partners of the
ODOT Roadwaste Research Project for collegial support.
35
-------
References
Bretsch, K., July 2002. Oregon's Vactor© Waste Data Reviewed. Paper presented at the Transportation
Research Board A1F07 & A3C01 Committees 2002 Summer Workshop, Helena, MT. PDF available from
the author.
Collins, 1, 1999. Roadwaste: Issues and Options. FHWA-OR-99-05, ODOT, Salem. This report is
informally known as the ODOT Roadwaste Research Project Phase 1 Report. PDF available for download
from ODOT Research Division web site.
Ghezzi, M., Collins, 1, Moore, 1, Bretsch, K., and Hunt, L., 2001. ODOT Roadwaste Research Project:
Field Trials. Includes reports on filter box, chemical flocculant, and other field trials, and worker safely
plans. Available for download from the ODOT Research Division web site.
Houghton Mifflin, 2000. "Phytoremediation." The American Heritage® Dictionary of the English
Language, 4th ed. Boston: www.bartleby.com/61/.
lurries, D., 2000. Personal communication. lurries is a specialist in erosion control and the chemistry and
behavior of suspended sediments for Oregon DEQ. He has demonstrated electroflocculation as a method of
removing suspended sediments from construction site runoff. Contact information available via Oregon
DEQ's web site.
Moore, J., and Collins, J,, 2001. ODOT Roadwaste Research Project: Roadwaste: Guide for District Plans,
APWA Draft. ODOT, Salem. Includes recommended ODOT BMPs. Available from the ODOT Research
Division web site.
USD A, NRCS, 2002. The PLANTS Database, Version 3.5 (http://plants.usda.gov). National Plant Data
Center, Baton Rouge, LA 70874-4490 USA.
USEPA, ORD, 2002. "Terms of Environment", Update 30 Dec (http://www.epa.gov/OCEPAterms/ ).
36
-------
INAPPROPRIATE DISCHARGE DETECTION AND ELIMINATION
WHAT PHASE I COMMUNITIES ARE DOING TO ADDRESS THE PROBLEM
Jennifer Zielinski and Ted Brown
Center for Watershed Protection
Ellicott City, Maryland
Abstract
Inappropriate connections to storm drain systems account for significant annual pollutant loads from urban
areas. Inappropriate discharge detection and elimination (IDDE) are important elements of any effective
stormwater quality management program. Since 1990, under US EPA's National Pollutant Discharge
Elimination System (NPDES) Phase I Storm Water Program, cities and counties with populations of
100,000 or more that operate a municipal separate storm sewer system (MS4) were required to obtain
discharge permit coverage. An element of NPDES Phase I, Part I was that regulated MS4s were required to
perform discharge characterization by screening outfalls for inappropriate connections to MS4s. NPDES
Phase I, Part II required regulated MS4s to demonstrate adequate legal authority to control discharges,
prohibit inappropriate discharges, require compliance, and carry out inspections, surveillance and
monitoring (EPA, 1996). As a result, 173 cities and 47 counties (Glanton et a/., 1992) were required to
develop IDDE programs.
In 2001, the Center for Watershed Protection (CWP) and Dr. Robert Pitt from the University of Alabama
obtained a multi-year grant from US EPA to research the most cost-effective and efficient techniques that
can be employed to identify and correct inappropriate discharges, and write a "Users Guide" geared toward
use by NPDES Phase U communities and citizen volunteers. One element of the research is investigating
and compiling data and methods that have been employed in pursuit of IDDE by NPDES Phase I MS4s.
CWP conducted a survey of 24 NPDES MS4s representing various geographic and climatic regions in the
U.S. to research what these communities have been doing on the IDDE front. Surveys requested information
about: community characterization; system characterization; IDDE program characterization; legal
authority; system mapping; procedures used for inappropriate discharge identification, confirmation, source
identification and correction; education and outreach; and other programmatic features or references. This
paper presents the findings of the survey and provides inferences that can be drawn about the collected data.
Introduction
The Center for Watershed Protection (CWP) and Dr. Robert Pitt, University of Alabama, are working under
a multi-year grant from the US EPA to research the most cost effective and efficient techniques that can be
employed to identify and correct inappropriate discharges, and to develop a "Users Guide" for use by
National Pollutant Discharge Elimination System (NPDES) Phase n jurisdictions and citizen volunteers.
One element of the research is investigating and compiling data and methods that have been employed in
pursuit of inappropriate discharge detection and elimination (IDDE) by NPDES Phase I MS4s.
A survey was developed and submitted to over 50 local jurisdictions representing various geographic and
climatic regions in the United States that have implemented IDDE programs. The intent of the survey was to
37
-------
determine the current state of practices utilized by local governments, and to identify practical, low cost, and
effective techniques that have been implemented in the field and laboratory for inappropriate discharge
detection and elimination. The survey information will be used in the preparation and development of the
Users Guide. This paper summarizes the results of the survey.
Design of Survey
The survey was designed to elicit detailed information on existing IDDE programs and to gain insight on the
following topics: (A copy of the survey can be accessed from www.cwp.org)
1. Community Characterization
2. System Characterization
3. Inappropriate Discharge Detection Elimination (IDDE) Program Characterization and Cost
4. Legal Authority
5. System Mapping
6. Methods to Identify and Confirm Inappropriate Discharges
7. Inappropriate Discharge Corrections Program
8. Education, Outreach, and Pollution Prevention Programs
The target audience for the survey included jurisdictions that have implemented IDDE programs, primarily
those subject to NPDES Phase I requirements. Jurisdictions selected for the survey represent a variety of
geographic and climatic regions. The EPA stormwater coordinators for each region of the country were
contacted for recommendations on jurisdictions to include in the survey. A variety of jurisdiction sizes were
targeted on the basis of population, IDDE program service area, and land use. The ages and reputations of
the program were also considered. The survey was sent to 57 jurisdictions, with 24 jurisdictions (42%)
from 16 states completing the survey (Figure 1).
Surveys were supplemented by on-site interviews of IDDE program staff in seven jurisdictions: Baltimore
City, MD; Baltimore County, MD; Boston, MA; Cambridge, MA; Dayton, OH; Raleigh, NC; and Wayne
County, MI, witnessing field operations when possible.
Survey Results
Community Characterization
Of the 24 jurisdictions that completed the survey, 18 are NPDES Phase I jurisdictions, one was awaiting the
issuance of its Phase I permit, two are Phase II jurisdictions, two operate under a Stormwater General
Permit, and one is a Special Purpose District servicing both Phase I and Phase U jurisdictions (Table 1). Of
the 24 respondents, only 21 have fully implemented IDDE programs. Alexandria and Falls Church,
Virginia, are both currently developing programs as part of their NPDES Phase U requirements. Seattle,
Washington, currently addresses inappropriate connections via water quality complaints and a routine
business inspection program. Seattle's Phase I NPDES permit is currently being updated, and the next
permit cycle will require the implementation of a full inappropriate discharge reduction program. Even
38
-------
though these three jurisdictions have not fully implemented their programs, they have each implemented
some elements. Therefore, data reported throughout this paper reflects varying numbers of responses to
different survey questions.
Overall, the respondents included five counties, 18 cities, and one Special Purpose District. Land use was
varied, but tended towards ultra-urban, urban, and suburban. The population density ranged from 175 to
15,000 people per square mile, with a median of 2,600 people per square mile. The jurisdictions also vary
in service area, with ranges from 2 to 498 square miles, and a median of 70 square miles.
O Completed
© Returned Supporting Information Only
$ CWP Field Interview
Figure 1: Jurisdictions that Participated in the IDDE Survey
System Characterization
To help determine the relative scale of the programs, the survey requested information that would
characterize the jurisdictions drainage systems in addition to population density, service area, and land use.
Specifically, information on length of storm drain network, number of major outfalls, and the ratio of
outfalls to miles of storm drain were compiled (Table 1).
39
-------
Table 1: Characterization of Jurisdictions that Part
. ..... Form of NPDES
Jur,sd,ct,on Government Status
Ada County Highway
District (ACHD), ID
Albuquerque, NM
Alexandria, VA
Arlington Co., VA
Austin, TX
Baltimore City, MD
Boston, MA
Cambridge, MA
Clackamas Co., OR
Dayton, OH
Durham, NC
Falls Church, VA
Howard Co., MD
Knoxville, TN
Lakewood, CO
Montgomery Co., MD
Phoenix, AZ
Portland, OR
Raleigh, NC
Seattle, WA
Springfield, MO
Thousand Oaks, CA
Wayne Co., Ml
Worcester, MA
Median
Special
Purpose District
City
City
County
City
City
City
City
County
City
City
City
County
City
City
County
City
City
City
City
City
City
County
City
cipated in the IDDE Survey
Land Use (%)
Ultra-
Urban
Phase I, 12
Phase II
Phase I
Phase II 100
Phase
Phase
Phase
Phase
Phase I
Phase
Phase
Phase
Phase
Gen. Perr
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
10
1
-
-
85
10
20
4
10
nit 15
10
-
-
-
5
100
5
-
Gen. Permit 33
Phase I
10
Urban
23
90
-
9
25
71
85
15
15
50
20
50
25
20
30
30
-
20
-
50
6
25
Sub-
urban
28
-
-
47
54
-
-
-
60
10
43
39.5
53
55
N/R
12
60
-
40
-
30
33
41
N/R
41
Rural
11
-
-
-
20
-
-
-
5
5
5
-
6
5
30
10
-
10
-
-
10
13
10
Forest/
Undev'd
26
10
-
33
-
29
15
-
10
15
28
0.5
-
10
28
-
-
25
-
15
47
7
15
Population
Density
(people/mi2)
1,070
2,400
8,000
7,149
2,745
7,173
12,271
15,000
181
3,115
1,950
5,000
972
1,750
3,225
1,762
2,537
3,534
1,800
6,706
2,000
2,142
175
4,600
2,600
Service
Area (mi2)
69.73
181
15.75
20
238
92
48
6.25
22
52
92
2
255
100
44
496
473
47
120
84
70
58
498
37.6
70
Total Length of
Storm Drainage
Network (mi)
351
582
N/R
400.5
600
726
542
81
N/R
600
2,690
N/R
300
324
N/R
2,597
3,500
562
3,200
630
500
N/R
3,265
347
582
# of Major
Outfalls
65
6
N/R
100
250
345
94
11
22
300
890
N/R
365
1,004
204
7,165
322
110
1400
200
6
N/R
2,000
250
250
Outfall / Mile of
Drainage
Network
0.19
0.01
N/A
0.25
0.42
0.48
0.17
0.14
N/A
0.50
0.33
N/A
1.22
3.10
N/A
2.76
0.09
0.20
0.44
0.32
0.01
N/A
0.61
0.72
0.33
Notes: N/A = Not applicable; N/R = Not reported
40
-------
Program Characterization
Staff time dedicated to the IDDE programs surveyed ranged from 0.08 to 10 person-years, with a median of
1.5 person-years (Table 2). It was difficult for many of the jurisdictions to quantify actual staff time
dedicated to IDDE activities since the responsibilities are spread among many departments, or because the
staff who work on IDDE also perform other un-related tasks.
Table 2: Staff Time Dedicated to IDDE Program Annually
Jurisdiction
WL«II 1 IIIIC y|JdO\/l
Program
ii~yc«ioj i-rcui\*<
Annually (n =
Field Staff Office Staff'
Wayne Co., Ml
Baltimore City, MD
Phoenix, AZ
Knoxville, TN
BWSC, MA
Worcester, MA
Durham, NC
ACHD, ID
Montgomery Co., MD
Cambridge, MA
Albuquerque, NM
Austin, TX
Raleigh, NC
Thousand Oaks, CA
Springfield, MO
Howard County, MD
Portland, OR
Clackamas Co., OR
Dayton, OH
Arlington Co., VA
Lakewood, CO
Median
6
6
5
2
2
2
2.1
1
2
12
NoteS
1
1
0.9
0.5
N/R
0.22
0.1
0.1
0
0.04
1.0
4
2.25
2
1.5
1.25
1
0.5
1.5
0.5
0.50
1.5
0.35
0.3
0.3
0.5
0.6
0.11
0.1
0.05
0.1
0.04
0.5
HLCU iv ii_ri_ri_
21)
Total Staff
10
8.25
7
3.5
3.25
3
2.6
2.5
2.5
1.50
1.5
1.35
1.3
1.2
1.0
0.6
0.33
0.2
.15
0.1
0.08
1.5
Ratio of Field
to Total
60%
73%
71%
57%
62%
67%
81%
40%
80%
66%
N/A
74%
77%
75%
50%
N/A
67%
50%
67%
0%
50%
67%
Notes:
1. Includes administrative and professional office staff.
2. Additional 1.75 person-years spent by professional consultant performing sampling, inspection work.
3. Field monitoring subcontracted to a consultant.
For similar reasons, it was also difficult for jurisdictions to accurately report the full IDDE program budget,
as well as costs associated with different related activities (Table 3). Annual IDDE program expenditure
ranged from $3,500 to $613,561, with a median of $121,825.
41
-------
Table 3: Annual IDDE Program Expenditure
Jurisdiction
Wayne Co., Ml
Phoenix, AZ
Cambridge, MA
Baltimore City, MD
Albuquerque, NM
Worcester, MA
Montgomery Co., MD
BWSC, MA2
Durham, MA
ACHD, ID
Thousand Oaks, CA
Raleigh, NC
Springfield, MO
Austin, TX
Knoxville, TN
Portland, OR
Clackamas Co., OR
Arlington Co., VA
Lakewood, CO
Howard Co., MD
Median
Staff Total
($)
460,672
500,003
100,200
298,750
110,000
160,000
200,000
142,000
156,600
160,450
60,000
53,000
70,000
67,500
33,000
15,000
16,000
7,000
3,500
3,000
$85,100
(% of total)
75%
84%
25%
75%
28%
57%
97%
73%
89%
100%
72%
64%
84%
82%
55%
58%
100%
95%
57%
86%
75%
Office Computer /
Software
($)
3,760
-
1,000
-
-
-
200
2,500
-
-
5,000
5,000
1,000
1,000
-
-
-
300
-
$1,000
(% of total)
0.6%
-
0.2%
-
-
-
0.1%
1 .4%
-
-
6.0%
6.0%
1 .2%
1 .7%
-
-
-
4.9%
-
1%
Field Equipment
($)
319
15,665
3,000
10,000
14,000
-
5,500
1,000
3,500
-
10,000
6,000
5,000
4,000
500
-
-
50
1,600
-
$4,000
(% of total)
0.1%
2.6%
0.7%
2.5%
3.6%
-
2.7%
0.5%
2.0%
-
12.0%
7.2%
6.0%
4.8%
0.8%
-
-
0.7%
26.0%
-
3%
Lab Equipment /
Testing
($)
7,500
13,840
10,000
87,000
20,000
15,000
500
8,000
-
5,000
12,000
1,000
5,000
15,000
10,000
-
300
500
500
$8,000
(% of total)
1%
2%
2%
22%
5%
5%
0%
5%
-
6%
14%
1%
6%
25%
38%
-
4%
8%
14%
5%
Other1
($)
141,273
64,571
297,200
-
250,000
100,000
50,000
4,600
-
5,000
7,000
2,000
-
10,000
1,000
-
-
250
-
$10,000
(% of total)
23%
11%
73%
-
63%
36%
26%
3%
-
6%
8%
2%
-
17%
4%
-
-
4%
-
11%
Total
Annual
($)
613,561
593,134
406,400
395,750
394,000
280,000
205,500
193,700
175,000
160,450
83,200
83,000
83,000
82,500
59,500
26,000
16,000
7,350
6,150
3,500
$121,825
Notes:
1. Typical costs included in the "other" category include education, training, travel, consultants, and contractors.
2. The annual budget information provided by BWSC does not include the costs associated with corrections, nor the costs associated with special drainage
system studies.
42
-------
Legal Authority
Ninety-six percent of the surveyed jurisdictions have some type of regulation that prohibits inappropriate
discharges from entering the MS4. Discharge prohibitions typically come under at least one of three
regulations:
1) A stormwater ordinance that addresses inappropriate discharges to the storm sewer system or receiving
waters;
2) A plumbing code that addresses illegal connections to the storm sewer system; or
3) A health code that regulates the discharge of harmful substances to the storm sewer system or receiving
waters.
Most jurisdictions surveyed have the legal authority necessary to inspect private properties for illegal
discharges, but based on our interviews, few seem to have found it necessary to invoke that authority.
Communities noted that owners are usually cooperative with respect to property inspections by jurisdictions
investigating inappropriate discharges, and that achieving compliance is not usually problematic.
Mapping Capabilities
Over 80% of the jurisdictions surveyed utilize Geographic Information Systems (GIS) to track outfalls and
record site data. Despite the convenience and power of the digital maps, many communities still relied on
supplemental information provided on paper maps, particularly where information transfer to the GIS was
not complete or was unverified. Based on interviews with select jurisdictions, preferences for paper or
digital mapping varied. For instance, Baltimore City field crews expressed a preference for paper mapping,
which they felt to be easier to interpret than printouts from the digital mapping system. In addition, for areas
where sewer mapping either does not exist, they have often turned to historic topographical maps to
determine possible pre-development stream locations.
A primary use of mapping in an IDDE program is to prioritize areas for outfall screening or dye testing. In
addition, it is useful for tracking areas that have been investigated versus those that still need to be
investigated. Table 4 displays the IDDE program mapping elements that surveyed jurisdictions use.
Based on interviews, other key areas that are useful to map include:
• Certain industries by SIC code
• Historic complaints
• Sanitary and storm sewers in close or in common manholes
• "Gaps" in sanitary mapping
• Licensed businesses, SIC codes, industrial permittees
• Areas with businesses with night hours (e.g., bars and restaurants)
43
-------
Table 4: Common IDDE Program Mapping Elements
Elements Mapped by Jurisdictions
% of Jurisdictions Responding (n = 24)
Storm sewers
Waters of the US receiving discharges from outfalls
Outfalls
Open channels (conveyance channels)
Land use
Sanitary sewers
Industrial discharge permit holders
Building connections to storm sewers
Connections to adjacent systems / communities
Building connections to sanitary sewers
Watershed, outfall drainage area boundaries
Hotspot areas
96%
83%
79%
71%
67%
63%
33%
25%
25%
21%
13%
13%
Methods to Identify and Confirm Potential Inappropriate Discharges
Table 5 displays the procedures utilized by the surveyed jurisdictions to determine the presence of a
suspected inappropriate discharge. Most of the jurisdictions used several different methods and there was no
apparent trend based on geographical location. The top three procedures selected were: 1) pollution
reporting hotline (86%); 2) regular inspection of outfalls by jurisdiction (76%); and 3) water quality
monitoring of receiving waters (71%).
Some of the jurisdictions found that the initial outfall screening conducted was very successful at
identifying chronic problems, but that the following screening was less useful. For sporadic discharges,
jurisdictions are relying more heavily on telephone hotlines and cross-training inspection and maintenance
staff than on monitoring or field screening.
Table 5: Investigative Procedure(s) Used to Determine the Presence of a Suspected Inappropriate Discharge to a
MS4 or Receiving Water
Investigative Procedure
% of Respondents (n = 21)
Pollution reporting hotline for citizens to call
Regular inspection of outfalls by jurisdiction
Water quality monitoring of receiving waters
Regular inspection of storm sewers
Regular inspection of sanitary sewers
Dye- or smoke-testing of buildings in problem areas
Sporadic outfall inspection by watershed/citizen organization
Regular outfall inspection by watershed/citizen organization
86%
76%
71%
62%
48%
48%
38%
24%
44
-------
Sporadic inspection of outfalls by jurisdiction 24%
Dye- or smoke-testing of buildings at the time of sale 5%
Water quality monitoring of discharge waters 5%
Septic system inspection at time of sale 5%
Sources of Discharges Typically Found
Common sources of discharge found by jurisdictions responding to the survey are displayed in Table 6.
While certain sources are random and may occur anywhere, such as illegal dumping, other sources can often
be associated with specific factors within a community or subwatershed. These include:
• Land use (e.g., industrial discharges, restaurant grease, failing septic systems)
• Type and age of sewer system (e.g., pump station failures, inflow/ infiltration, SSOs)
• Historic plumbing codes (e.g., connection of floor drains to storm sewers)
• Recreational facilities (e.g., chlorine from swimming pool discharges, sewage from marina pumpouts)
No significant relationship was apparent relating sources of discharge to geographic location.
Table 6: Sources of Inappropriate Discharges Typically Found
Sources of Inappropriate Discharge % of Respondents (n = 21)
Illegal dumping practices 95%
Broken sanitary sewer line 81%
Cross-connections 71%
Connection of floor drains to storm sewer 62%
Sanitary sewer overflows 52%
Inflow/infiltration 48%
Straight pipe sewer discharge 38%
Failing septic systems 33%
Improper disposal of wastes from recreational vehicles 33%
P u mp stati o n fai I u re 14%
Outfall Monitoring
All but two of the jurisdictions surveyed conduct some sort of outfall monitoring program. Most conduct
outfall monitoring on a regular basis, per NPDES Phase I requirements.
Jurisdictions reported that beyond initial outfall screening, continued outfall monitoring was less useful in
finding intermittent or one-time discharges. For instance, Wayne County, MI, noted that outfall monitoring
is not the most effective method for identifying inappropriate connections due to the potential for dilution,
the periodic nature of some discharges, and the time delay between discharge into the system and discharge
from the outfall. This is supported by survey results that indicate the periodic nature of discharges is the
biggest impediment to identifying inappropriate discharges.
Jurisdictions seem to place a heavy reliance on physical indicators of discharges, as opposed to chemical
outfall screening, even in light of a 30% false positive identification rate (Lalor, 1993). The most common
approach to outfall screening involves conducting a visual inspection of the outfall and a qualitative
45
-------
assessment of any flow present, including observation of water color, odor, turbidity, floatables, and
sedimentation. In some cases, if the flow is suspected to be inappropriate, a follow-up grab sample is taken
for quantitative analysis. Many jurisdictions bypass the quantitative tests and immediately move upstream
to find the source of the discharge.
In-Stream Monitoring
Some jurisdictions utilize in-stream monitoring to enhance or supplement outfall monitoring. In-stream
monitoring is used to identify trends that may lead toward characterization of inappropriate discharges.
The City of Raleigh, NC has conducted baseline monitoring on nine streams for basic parameters, some of
which are used to detect sewer leaks including fluoride, fecal coliform, ammonia, sodium, and conductivity.
Deviation from the baseline for these parameters observed during regular in-stream monitoring prompts
further investigation of possible inappropriate discharges. Baltimore City conducts weekly screening of
receiving waters using a hydrolab or equivalent and field test kits for ammonia. When a threshold value is
exceeded, sampling continues upstream until the source is located. To address chronic problems, a monthly
sampling program is conducted using an extensive variety of laboratory-analyzed chemical parameters at
approximately 40 receiving water stations. When long-term medians exceed a certain percentile based on
the entire database, investigations are conducted by sampling further upstream in the storm drain network.
Citizen Hotlines
Citizen hotlines are a common method for indicating the presence of a suspected inappropriate discharge.
Nineteen (90 %) of the surveyed jurisdictions have pollution reporting hotlines, and 18 of these track the
number of complaints that have been received and corrected to help determine IDDE program success.
Montgomery County, MD, noted that the success of their IDDE program is directly related to their water
quality outreach, complaint, and enforcement system, not to their outfall-screening program. On average,
County staff identify and correct about six inappropriate discharges per year as a result of regular screening.
By contrast, over 185 inappropriate discharges are corrected each year as a direct result of citizen
complaints and calls to the hotline.
Public education and labeling of outfalls and other storm drain infrastructure is an important element of
establishing a successful citizen hotline. Boston Water and Sewer Commission (BWSC) has labeled
outfalls along the Charles River so that citizens can identify outfalls from the water. Dayton has labeled
outfalls along the City's popular riverfront, and recommends labeling catch basins and manhole covers.
Tracers and Methods Used
The majority of surveyed jurisdictions utilize tracers to confirm the presence of a suspected inappropriate
discharge (Table 7). Emphasis is on quick and simple tests that do not require extensive and time-
consuming laboratory analysis. Qualitative physical parameters are the most widely used tracers, including
color, odor, deposits and stains, temperature and presence of floatable matter. When chemical tracers are
used, communities tend to focus on a single parameter such as bacteria, ammonia, or detergents so that field
and lab equipment costs are controlled. However, using only one parameter as a tracer can leave
unanswered questions about other sources of inappropriate discharges. This uncertainty can be reduced
somewhat when sampling is conducted in conjunction with land use data analysis. In addition, there are
46
-------
certain situations where a single source is known to dominate the inappropriate discharges to a watershed
and a single tracer is warranted. For example, Baltimore, MD, has chronic sewage infrastructure problems
and makes the assumption that sewage is the likely dominant inappropriate discharge in many of its
subwatersheds. Consequently, Baltimore often uses ammonia as a sole tracer to track inappropriate
discharges.
Table 7: Tracer Parameters Used to Confirm the Existence of Inappropriate Discharges
Tracer Parameter
Color
Odor
Deposits and stains
Floatable matter
pH
Temperature
Chlorine
Turbidity
Changes in flow
Specific conductivity
Vegetation change
Ammonia / ammonium
Structural damage
Surfactants
Fecal coliform
Fluoride
Copper
Florescence
Phenols
Potassium
Detergents
Dissolved oxygen
Grease / oil
Hardness
Physical or Chemical
P
P
P
P
C1
P
C
P
P
C
P
C
P
C
C
C
C
C
C
C
C
C
P
C
% of Respondents (n = 21)
95%
95%
90%
86%
86%
86%
76%
76%
62%
62%
62%
52%
52%
48%
33%
33%
29%
24%
14%
14%
10%
10%
10%
10%
Some chemical parameters can be measured in the field with probes or test strips. These methods are often not as
sensitive as those that would be used in a laboratory analysis.
Inappropriate Discharge Corrections Program
Some jurisdictions simply bear the cost of inappropriate connection repairs and bill the owners after the
repairs have been completed. Ada County, ID and Raleigh, NC use this method as a last resort to gain
compliance. Worcester, MA pays half of repair costs and bills the owner for the remainder.
Most jurisdictions reported that diplomacy, trust, reasoning and education are the primary people skills
required to successfully perform their jobs effectively. Diplomacy and trust are important when trying to
gain access to private property for plumbing inspections and dye testing. Reasoning and education are
necessary when explaining to property owners that a problem exists on their property when trying to get the
owners to make required connections. The bottom line is that different tactics and approaches work to gain
compliance from different people. Wayne County, MI mentioned that the publicity surrounding the Rouge
River Project helped open doors for them, because property owners had heard enough about programs to
clean the river prior to having IDDE inspectors knock on their doors.
47
-------
Education, Outreach, and Pollution Prevention Programs
Nineteen of the IDDE programs surveyed include some type of education and outreach elements. Of these,
all target residents, 75% target the commercial sector, 63% target the industrial sector, and 50% target the
government sector. In some cases, educational messages relating to inappropriate discharges are
incorporated into campaigns developed for other departments or programs within the jurisdiction. Other
jurisdictions run very targeted IDDE education programs.
Resident Education
For jurisdictions that rely heavily on citizen hotlines as a means of identifying potential inappropriate
discharges, residential education is an important program component. Some common forms of residential
education identified through the surveys include storm drain stenciling or marking; signage at outfalls;
educational brochures or newsletters in utility bills; and promotion of citizen hotlines.
Schoolchildren Education
Some communities such as Dayton, OH and Phoenix, AZ have educational programs geared towards
schoolchildren. Dayton's inappropriate discharges education is part of a larger schoolchildren educational
effort that includes regular visits to schools and the "Children's Water Festival." This one-day event for
3,000 students from the 4th-6th grade levels offers a series of presentations, games, experiments, and
exhibits on groundwater, surface water, conservation, land use, and other water related topics. Phoenix
noted that the school presentations made to third and fourth graders are an effective part of their stormwater
program. City stormwater inspectors give presentations to the children and distribute Storm Drain Dan
coloring books, pencils, erasers, rulers (all bearing the City's stormwater logo and phone number) and
Storm Drain Dan dolls. They have found this to be particularly helpful in lower income neighborhoods
where school supplies are in high demand. The children are reported to be enthusiastic and motivated to
keep the environment clean.
Commercial and Industrial Education
In most cases, jurisdictions have developed targeted commercial or industrial education programs based on
specific local problems, land uses, or "hot spot" activities likely to contribute specific types of problems.
For example, several jurisdictions have developed educational programs regarding grease handling and
disposal at restaurants. Clackamas County, OR has developed educational brochures for contractors
regarding concrete and mortar management. Both land use mapping and a historical record of problems and
complaints help jurisdictions to identify areas to focus on in these types of educational campaigns, which
tend to be accomplished through one-on-one contact as opposed to mass distribution of educational
materials used for residential education.
48
-------
Public Employee Education
Several jurisdictions identified cross training of public employees as an important means of identifying
potential inappropriate discharges. For example, Wayne County, MI currently trains field crews of the
Division of Public Works, County Drains, and Recreation and Parks on inappropriate discharge detection to
increase both awareness and the number of "eyes" looking for problems. Effective training typically
includes presentations, videos, and problem-solving activities.
Conclusions
Several conclusions were developed from the surveys and interviews regarding IDDE program
development. Typically, 67% of program staff time is dedicated field staff. As program staffing increased,
this ratio stayed fairly consistent. Also, several program directors noted that experienced field staff are a
valuable asset, while several others noted that the lack of staff expertise and experience is a top problem in
identifying inappropriate discharges. Accurate mapping resources can improve the efficiency of a program
in the identification of outfalls and prioritization of problem areas. The wide range of program budgets can
be attributed to the methods used by the programs to identify potential inappropriate discharges. The five
programs with the highest annual expenditures dedicate significant portions of their budgets to support
intensive outfall screening, continuous in-stream monitoring, and targeted area investigations. Their budgets
support larger field staffs or consultants who conduct these investigations; the purchase of more
sophisticated lab and field equipment; and targeted educational programs. IDDE programs have invoked
legal authority using one or more of three mechanisms: 1) a stormwater ordinance that prohibits illicit
discharges to the drainage network; 2) a plumbing code that prohibits illegal connections to the drainage
network; or 3) a health code that regulates the discharge of harmful substances to the drainage network.
Drawing from these conclusions, there are several program development challenges that will likely be faced
by NPDES Phase n communities and potential ways to alleviate them. The range of responses with regard
to program characterization questions indicates a defined need for relatively simple guidance for performing
inappropriate discharge investigations. The guidance should provide programmatic recommendations as
well as recommendations for field methods and anticipated costs. A lack of staffing resources may prove to
be a significant hindrance to implementing a successful IDDE program. Phase I communities rely heavily
on the expertise of their field staff - expertise that has been largely developed as the programs were being
developed. Methods or approaches recommended for Phase II communities should be less dependent on
professional judgment. Many communities do not have current mapping. Focus should be placed on
mapping storm sewers, open drainage channels, waters of the US, outfalls, and land use. This will provide
field staff the minimum data necessary to conduct field investigations, and will serve as a basis for
prioritizing field investigations.
Outfall screening can require significant staff and equipment resources. An efficient approach that examines
a limited number of parameters at each outfall is necessary. In addition, more effective and reliable tracers
and associated analytical techniques are needed to reduce the uncertainty (i.e., number of false negatives and
false positives). When examining equipment needs, Phase n programs should communicate with other
jurisdictional programs that utilize the same types of field equipment and examine the possibility of sharing
49
-------
purchase expenses. Model ordinance language should be provided to Phase n communities to ensure that
all potential sources of inappropriate discharges are prohibited; and that the community is provided with the
necessary legal authority to inspect private properties and to enforce corrections. Effective IDDE programs
need to have a balanced approach involving field screening, hotspot targeting, hotlines, public education,
and municipal employee cross-training.
Acknowledgements
The authors would like to express our sincere appreciation to the communities and individuals that took the
time to complete the survey. Without the detailed feedback that we received, it would not be possible to
generate the necessary conclusions about the current state of the practice or develop a clear understanding
about the areas where uniform guidance is needed. In addition, several of the project team members were
instrumental in developing and compiling the survey, including: Deb Caraco, Stephanie Linebaugh, Dan
O'Leary and Tom Schueler of the Center for Watershed Protection and Dr. Bob Pitt of the University of
Alabama. Lastly, Mr. Bryan Rittenhouse of USEPA Office of Water has provided valuable project support
as project officer.
References
Glanton, T., M. Garret and B. Goloby. 1992. The Illicit Connection: EPA Storm Water Regulations Field
Screening Program and the City of Houston's Successful Screening System. Water Environment and
Technology (September 1992): 60-68.
Lalor, M. 1993. Assessment of Non-Stormwater Discharges to Storm Drainage Systems in Residential and
Commercial Land Use Areas. Ph.D. thesis. Department of Environmental and Water Resources
Engineering. Vanderbilt University. Nashville, TN.
US EPA. 1996. Overview of the Storm Water Program (EPA833-R-96-008). US EPA Office of Water.
Washington, DC. June 1996.
50
-------
USE OF THE CLEAN WATER STATE REVOLVING FUND FOR MUNICIPAL
STORM WATER MANAGEMENT PROGRAMS
Amy Butler and Jim Collins
Tetra Tech, Inc.
Fairfax, Va
Sandra Perrin
U.S. EPA
Washington, DC
The Need for Funding
Many municipalities have funded traditional storm water management activities through their general
revenue sources. Traditionally, storm water management was thought of as minimizing street flooding and
reducing property damage caused by peak runoff flows. Controlling the water quality aspects of urban
runoff is a much more recent addition to the perceived municipal storm water management responsibility.
With few exceptions, incorporating water quality controls in tandem with the traditional quantity
management has occurred through the regulatory process. Therefore, municipalities typically consider the
quality component of storm water management to be a new and separate mandate. Some municipalities
recognized the link between storm water quantity and quality and took the initiative to establish
comprehensive storm water management programs to address both issues. More often than not, however,
municipalities began managing storm water quality and quantity together in response to regulations
implementing the National Pollutant Discharge Elimination System (NPDES) permit program for storm
water.
Subsequent to the 1987 amendments to the Clean Water Act (CWA), EPA published regulations
establishing Phase I of the NPDES Storm Water Program in 1990. Under Phase I, EPA required NPDES
permit coverage for discharges of storm water associated with industrial activity, discharges of storm water
from construction sites greater than 5 acres in size, and storm water discharges from medium and large
municipal separate storm sewer systems (MS4s) located in incorporated places or counties that serve
populations of 100,000 or more. The Phase II Final Rule, also a result of the 1987 CWA Amendments, was
published in the Federal Register on December 8, 1999. The Phase II rule requires NPDES permit coverage
for storm water discharges from construction sites that disturb between 1 and 5 acres and from small MS4s,
defined as those systems serving areas populations less than 100,000 to a lower limit based on the U.S.
Census Bureau's definition of an urbanized area.
Costs of Municipal Storm Water Management Programs
Every four years, EPA conducts an assessment of the water quality and human health protection financial
needs for wastewater collection and treatment systems, storm water management programs, and nonpoint
source projects. This effort is the Clean Watersheds Needs Survey (CWNS), which is a joint effort between
states and EPA. During the Construction Grants Program the CWNS only included project-specific costs
51
-------
for traditional wastewater collection and treatment system needs. Over the last 10 years, however, the
survey has expanded to include nonpoint source, estuary management, and storm water management
projects. The storm water management projects typically included in the CWNS are the capital costs of
developing and implementing municipal storm water management programs under the NPDES. Very few
Phase I MS4s had provided sufficiently detailed planning information to serve as project-specific
documentation for their needs in the last two surveys, thus the assessment of storm water management
program costs and needs relied primarily on modeling. The modeling approach used in the 1996 CWNS for
estimating Phase IMS4 needs assumed the use of regionally-targeted best management practices (BMPs)
for the major program areas based on hydrologic regions and variation in soil characteristics. Beginning
with the 2000 Clean Watersheds Needs Survey (CWNS 2000), several states made significant progress in
obtaining documentation for eligible storm water management program (SWMP) elements from the
operation of MS4s.
EPA was not required to conduct an analysis of the estimated cost expected to be incurred by municipalities
when developing their SWMPs and otherwise implementing the 1990 Storm Water Phase I regulations. The
1996 CWNS estimate for municipal storm water management program elements (i.e., facilities) was $7.4
billion, but this value was recognized as an underestimation. Table 1.1 provides a list of cost estimates that
were identified in the Phase I storm water modeling for the 1996 CWNS. These costs largely represent
one-time costs such as the cost to develop ordinances or the cost for initial training of municipal staff.
Because such expenditures are generally discrete and predictable, as are structural BMPs, they are examples
of items ideally suited to being included in the CWNS.
Table 1.1. Cost Estimates used inthe Phase I storm water model ing for the 1996 CWNS.
Institutional Source Controls
Site Plan Review
Inspection and Enforcement of Sedimentand Erosion
Control Plans at Construction Sites
Proper Storage, Use and Disposal of Fertilizers,
Pesticides, and Herbicides
Used Oil Collection and Recycling Program
Solid Waste Management/Litter Control Ordinance
Pet Waste Removal/Pooper Scooper1 Ordinance
Non structural Source Controls
Enhanced Litter Control
Costs
$10,000 per municipality for initial training
$10,000 per municipality for initial training
$10,000 per municipality for initial training
$30,000 per municipality for an ordinance and
development of regulations
$15,000 per municipality to pass an ordinance
$15,000 per municipality to pass an ordinance
Costs
Cost to place additional trash receptacles - $100.00
each (must be multiplied by the number of acres served
by enhanced litter control)
Source: USEPA, 1997
EPA estimated costs to Phase II municipalities to be between $848 million and $981 million. The costs to
MS4s are based on an annual per household cost of compliance. The individual household cost was
52
-------
calculated based on two different approaches. First, EPA used a survey of Phase II storm water program
costs developed by the National Association of Flood and Stormwater Management Agencies (NAFSMA).
The NAFSMA Phase II Survey was sent to more than 1,500 communities potentially impacted by Phase II,
with 121 communities responding. The communities were asked to report actual costs to implement any of
the six minimum control measures (or equivalent) that they are currently implementing. Not all
communities responded to each measure, and public involvement costs were not included (however, EPA
believed that cities included public involvement costs with public education costs). Table 1.2 presents the
average and percentile costs for five Phase II minimum control measures as estimated by the NAFSMA
survey (USEPA, 1999).
Table 1.2. Average and Percentile Costs for Five Phase II Minimum Control Measures (Per Household Costs, 1998
Dollars)
Mean Cost
Minimum
25%
50%
75%
95%
Maximum
Public
Education/
Outreach
$0.91
$0
$0.08
$0.37
$1.01
$3.04
$5.97
Illicit
Discharges
$1.78
$0.03
$0.20
$0.75
$2.65
$5.61
$5.95
Erosion/
Sediment
Control
$1.84
$0.09
$0.30
$1.08
$2.10
$7.92
$13.10
Development
$2.64
$0.07
$0.37
$1.24
$2.79
$10.68
$17.47
Municipal
Runoff1
$1.75
$0.01
$0.14
$0.52
$1.63
$9.08
$12.19
Totals: All
Categories
$8.93
$0.19
$1.09
$3.96
$10.17
$36.34
$54.68
Source: USEPA, 1999
1 A single outlier was removed because it was 15 times the mean cost for all municipalities.
The NAFSMA survey found an average annual household cost for Phase II of $9.16 (the table above lists
$8.93, and the difference is due to the addition of administrative costs of the program, including
recordkeeping and reporting requirements of the rule).
EPA also looked at an alternative approach for estimating Phase II costs. Thirty-five Phase I MS4s were
evaluated, with 26 providing adequate cost data Smaller Phase I MS4s were selected in order to be
comparable to Phase II communities. The average annual household costs to implement a program similar
to the six minimum measures for these Phase I municipalities was $9.08.
With the continual expansion of water quality protection initiatives in storm water management,
municipalities are constantly faced with finding new and creative methods of funding projects.
Additionally, as more Phase II communities develop their storm water management programs, traditional
sources of funding will be less available, leaving storm water program managers with the need to find
alternative ways to fund multiple projects.
53
-------
Sources of Funding
Municipalities, counties, states, and private citizens have relied on a variety of sources of funding for storm
water management projects. Largely, these have included storm water utilities, tax revenue, grants, loans,
and fees. The Clean Water State Revolving Fund (CWSRF) program is one that is traditionally underutilized
for funding storm water management programs. The CWSRF program was established in the 1987
amendments to the CWA under title VI. In these amendments, Congress instructed EPA to replace the
Federal Construction Grant Program with the CWSRF program. Since its inception over ten years ago, all
fifty states and Puerto Rico use the CWSRF Program. Using a formula determined by Congress in the 1987
CWA amendments, EPA grants each state an allotment of funds; the states then match up to 20 percent of
the federal grant to set up their CWSRF program. The program acts as a revolving fund to provide
independent and permanent sources of low interest loans for all types of water pollution control activities. It
is a unique system that relies on the continuous awarding and repaying of the loans to provide a permanent
funding source for water quality protection projects (USEPA, 2001). Communities, non-profit
organizations, municipalities, counties, individuals, and citizens are all eligible to apply for CWSRF loans.
To date, it has awarded more than $34.3 billion, using more than 10,900 low interest loans (USEPA, 2002a).
Congress designed the CWSRF program to give each state the utmost flexibility in providing financial
assistance. States can choose the types of assistance programs (e.g., loans, refinancing, purchasing, or
guaranteeing local debt and purchasing bond insurance) and set the loan terms, interest rates, and repayment
methods (EPA, 2002b). In addition to giving each state the authority to determine how to distribute funds,
Congress awarded states complete flexibility in determining the types of projects eligible for funding. Over
the years CWSRF monies have funded nonpoint source projects, wetland and estuary protection, storm
water management programs, and traditional wastewater collection and treatment system projects. (USEPA,
2001).
Nationally, the CWSRF loan average interest is 2.4 percent (individual state loan interests vary), with
repayment terms up to 20 years. Projects using CWSRF loans at this interest rate are funded using 23
percent less money than projects using the current market rate (USEPA, 2002a). CWSRF loans can be used
to partially or wholly fund a project. To apply for a CWSRF loan, a public or private entity submits an
application with the state-required information about the project. Most applications require a description of
the problem and information about how the project will be implemented (e.g., specifics on the water quality
and public health benefits, usually expressed in dollars per unit, the start and completion dates, as well as the
cost disbursement plan). States use the application forms to rank the projects and create a list of priority
projects that are eligible for CWSRF loans. These lists typically are called the project priority lists (PPL) or
intended use plans (IUP). A state will fund the projects on the PPL or IUP as money is available.
Depending on a state's program, projects that are not funded in one year might be transferred to the next
year.
Typical Storm Water Management Projects Funded with CWSRF
Restrictions on the types of projects eligible for CWSRF money are determined by the state, however, as a
general rule, projects should have a water quality or public health benefit. CWSRF loans can be used for
funding the capital costs for developing and implementing municipal storm water programs as required by
54
-------
an NPDES permit. This can include the costs for design, construction, and implementation of erosion and
sediment control and storm water BMPs and development of a storm water management program; operation
and maintenance costs are not funded by the CWSRF.
Since the expansion of the CWSRF program to include storm water and NFS projects, the number of
projects funded with CWSRF loans has expanded. The increase was not apparent in the 1996 CWNS
because needs for SWMP were mostly derived from modeling; however, the CWNS 2000 reported the
increase because better data were available. Despite the increase, the number of loans for storm water
management is still considerably less than the number of traditional wastewater collection and treatment
loans. For example, the CWNS 2000 reports 20 states with municipal storm water management program
needs, where as all 48 participating states had wastewater collection and treatment system needs. The
projects that are submitted to the CWNS 2000 must be CWSRF eligible; the projects do not require funding
by CWSRF. Only 5 states appeared to have used CWSRF loans to meet their storm water management
program costs: Maryland, Florida, New Jersey, Colorado, and Nebraska. (USEPA, 2002c). The CWNS
2000 has strict data requirements that can prohibit some storm water management projects from being
classified as storm water management needs. Projects that have a storm water management component that
are not associated with an MS4 permit program are categorized as anonpoint source (NPS) project in the
CWNS 2000. Twenty-three states submitted needs for NPS projects; of these 23 states only 8 states (New
York, New Jersey, North Dakota, Florida, Connecticut, Colorado, Wisconsin, and Maryland) appeared to
have used CWSRF loans to meet their storm water management costs (USEPA, 2002c).
Below are examples of storm water management projects in the State of Maryland that were funded using
CWSRF loans.
Baltimore County, Maryland
In 2000 Baltimore County developed a watershed management plan to identify storm water pollutants and
storm water management retrofits for the three watersheds as part of their NPDES permit. The plan
identified storm water management retrofits for 9 areas. The projects were designed to help control
unmanaged storm water runoff in a fully developed watershed and to improve water quality. The County
submitted a CWSRF loan application to the state for assistance with financing these projects. The CWSRF
loan applications called for developing feasibility analyses, enhancing existing storm water facilities,
designing extended detention ponds with shallow marshes, restoring stream channels, enhancing aquatic and
riparian habitats, and retrofitting storm drain outfalls. Baltimore County applied for loans to cover
approximately two-thirds of the engineering and construction costs; the county would pay the remaining
one-third (USEPA, 2002d).
Howard County, Maryland
In 1999 Howard County conducted an assessment of all the publicly owned storm water management
facilities in the Patapsco River Watershed. The County's NPDES permit required the County to determine
the viability of its storm water management facilities. The study identified and ranked the facilities that
were candidates for retrofitting. The county used the results of the study to apply for CWSRF loan
55
-------
assistance with the retrofits. Six individual projects were identified and submitted as separate loan
applications. Each project requested funds for reconstructing of sediment ponds, redesigning ponds to
include shallow marshes and extended detention ponds, retrofitting ponds to include water quality
management in addition to quantity control, removing concrete channels, adding forebays, implementing
stream restoration projects, and planting riparian and aquatic vegetation. As with Baltimore County, the
requested CWSRF loans covered approximately two-thirds of the engineering and construction costs; the
county and other stakeholders (e.g., homeowners associations) covered the remaining one-third (USEPA,
2002d).
Below are several examples of storm water management projects that could have been funded partially or
wholly using CWSRF loans.
Suffolk County, New York
In Suffolk County, New York, several projects were developed to prevent and contain road runoff from
entering Long Island Sound. The county applied for 12 grants to construct several recharge basins and
sediment traps to receive highway runoff and remove pollutants. The basins were designed to contain the
10-year design storm and the sediment traps were designed to intercept the first flush of runoff. For each
grant, the county matched the amount of the state funds requested. In this case, if grant money was not
available or if the county could not match the grant fund, the county could have applied to the state CWSRF
program for a loan (USEPA, 2002d).
Malabar, Florida
The Town of Malabar is a Phase II community that is approximately 20 percent developed. Its storm water
management system consists of swales and ditches, storm water pipes, baffle boxes, drain gutters, and
outfall structures. In low lying areas the town experiences flooding of ditches, clogged drains, eroding
stream channels, and discharges of pollutants into the Indian River Lagoon. Storm water management needs
for this town include development and implementation of a Master Plan, construction of swales along
streets, retrofitting of outfall structures, and addition of outfall structures. Although the town has developed
a storm water utility fund, because the storm water system needs major upgrades, more funding will be
needed beyond what the utility can provide. In this case, the town can apply for loans for both planning and
engineering costs necessary to begin construction, in addition to the actual construction costs. The town has
approximately 2,500 people, which allows the town to qualify for CWSRF benefits associated with a small
community (USEPA, 2002d). For small communities, the state sets aside 15 percent of all the CWSRF loan
funds (FLDEP, 2002).
Guadalupe, Arizona
The town of Guadalupe, in Maricopa County, will be constructing several retention basins along a canal
and an outfall system to control storm water runoff. The canal has a history of ponding and flooding the
nearby homes. The storm water collection system upgrades will contain the storm water runoff, prevent
flooding, and remove pollutants. This is a good example of combining traditional flood control designs with
56
-------
water quality protection techniques in the arid west. Maricopa County will be funding this initiative using
tax money because the town of Guadalupe is not able to contribute financially. The CWSRF program could
have been a viable alternative because the town of Guadalupe could have applied for loans directly (USEPA,
2002d).
Missouri
Across the State of Missouri there are several urban NFS projects that involved storm water management to
prevent erosion and flooding. Examples of projects to be completed included, installation of rip-rap and/or
grouted rock, retaining walls, culverts, natural bank stabilization, berms, gabions, detention ponds, inlets,
and new storm sewers. The projects were submitted to the CWNS as needs for a particular sewershed.
These types of projects are all candidates for CWSRF loans for NFS pollution control. If the projects could
be directly linked to anMS4 storm water management program, then the CWSRF loans would fall under the
storm water management category (USEPA, 2002d).
Conclusion
Despite the fact that the CWSRF program has been available to fund storm water management programs at
the local level for more than ten years, it is still a highly underutilized source of funding for this pollution
source in most states. As storm water programs continue to evolve and communities, municipalities, and
states begin to focus on the water quality benefits of storm water BMPs, finding creative financing
mechanisms will become even more of a challenge. Using the CWSRF to fund part if not all of a project
has already been demonstrated to be a practical mechanism for investing in elements of Phase I SWMPs.
Phase I municipalities should continue to use the CWSRF loans as a viable source of funds as retrofits and
upgrades are required. Consideration of using this funding source more widely should be strongly
encouraged for Phase II municipalities. Additionally, communities that cannot show a link between a
specific storm water management project and their MS4 storm water management program, should also
consider the potential of CWSRF funding by describing their project as an NFS pollution control project.
Interested municipalities should investigate their state's PPL or lUPs for information about projects that are
most important in their state. These lists can serve as an example of the types of projects that the state
approves for CWSRF loans. It appears that in some instances, states are failing to adequately get the word
out about the availability of the revolving loan funds for storm water projects. However, in other states, the
impediments to using this funding sources for storm water projects is due more to competition from projects
that address other water pollution sources, which are in many cases traditional wastewater collection and
treatment systems.
References
Florida Department of Environmental Protection (FLDEP). 2002. Florida Department of Environmental
Protection, Water Facilities Funding. Last updated 02/20/02. URL:
http: //www .dep.state.fl.u s/water/wff/faq. htm.
57
-------
USEPA. 1999. Economic Analysis of the Final Phase II Storm Water Rule. U.S. Environmental Protection
Agency, Washington, D.C.
USEPA. 2001. Financing America's Clean Water Since 1987, A report of Progress and Innovation. EPA-
832-R-00-011. U.S. Environmental Protection Agency, Office of Water, Washington, D.C.
USEPA. 2002a. Clean Water State Revolving Fund. Last updated on Tuesday, July 9th, 2002. URL:
http://www.epa.gov/OWM/cwfmance/cwsrf/index.htm
USEPA. 2002b. Clean Water State Revolving Fund, How the Program Works, Last updated on
Wednesday, July 10th, 2002. URL: http://www.epa.gov/OWM/cwfinance/cwsrf/basics.htm
USEPA. 2002c. Draft Clean Watersheds Needs Survey 2000, Report to Congress. U.S. Environmental
Protection Agency, Washington, D.C. Unpublished report.
USEPA. 2002d. Supplemental Information Submitted in Conjunction with the CWNS 2000. U.S.
Environmental Protection Agency, Washington, D.C. Unpublished data.
58
-------
PLANNING AND ASSESSMENT OF BEST MANAGEMENT
PRACTICES IN THE ROUGE RIVER WATERSHED
Kelly A. Cave, Director, Division of Watershed Management
Wayne County Department of Environment
Detroit, Michigan
Carl R. Johnson, Senior Vice President
Camp Dresser & McKee
Detroit, Michigan
ABSTRACT
The Rouge River National Wet Weather Demonstration Project in Wayne County, Michigan, has developed
an approach to linking the performance of best management practices (BMPs) to receiving water impacts.
The approach considers the various stages of the entire BMP process, including design, implementation, and
a system of performance measurements at each stage.
INTRODUCTION
In the management of watersheds, measuring progress is an untamed frontier of professional practice.
Watersheds present us with situations that defy accurate measurement. Consider the following contrasts
between measurements for point source controls versus measurements for watershed management.
• While pollution controls for point sources typically involve large engineered facilities that can be
equipped with sophisticated systems for measuring the quality of influent and effluent, watershed
management entails numerous and geographically scattered projects making it more difficult to
measure influent and effluent cost-effectively.
• While point source controls provide accountability to one single unit of governmental or business
organization, watershed management often depends on the individual actions of tens or hundreds of
organizations, each working with an individual set of priorities and budget limitations.
• While point source controls involve one particular technology, such as secondary treatment, or a
bundled set of technologies, such as storage and treatment, watershed management may involve a
detention basin in one area, a wetland with nutrient uptake in another, a street sweeping effort in yet
another area. Each technology has its own set of measurement requirements and differing
hydrologic factors.
• While point source controls typically are implemented with the ability to enforce compliance,
watershed management involves numerous efforts for water quality protection that often are beyond
the bounds of regulation, and therefore rely on voluntary efforts. Voluntary efforts by local units of
government must compete with mandatory efforts for budgetary resources, and this makes it more
difficult to achieve standard design criteria.
59
-------
It is against this backdrop that the Rouge River National Wet Weather Demonstration Project (Rouge
Project) sets out to link the performance of best management practices (BMPs) for wet weather pollution
control to improvements in water quality in the Rouge River watershed. While there is abundant
information on the technical performance of many BMPs in controlled settings for scientific or engineering
performance analysis, there is much less information on the performance of BMPs in real urban watershed
applications. The Rouge Project is filling this information gap by constructing and measuring the
cumulative performance of BMPs in complex urban watershed settings.
In the context of this paper, the term "best management practices, or BMPs" is used as a generic term to
mean any technology - either structural or non-structural - for the control of flows or pollutants that
adversely impact a receiving stream. This paper examines the array of mechanisms that the Rouge Project
has created to link and measure the performance of BMPs to water quality and ecosystem health
improvements. The array of mechanisms considers all of the complex factors in watershed management
which complicate the measurement process - dispersed geographic distribution of BMPs, multiple project
owners, a wide variety of pollution control technologies, and the voluntary nature of many activities. The
linking mechanisms used in the Rouge Project take into account the whole process of BMP development,
from setting design criteria, to project implementation and post-construction monitoring, and watershed-
wide assessments of progress.
PROJECT BACKGROUND
The Rouge Project, initiated in 1992 by the Wayne County, Michigan Department of Environment, has
learned a great deal on what it takes to restore an urban waterway to its beneficial uses. The project is
partially funded by Congressional appropriations managed by the U.S. Environmental Protection Agency
(EPA). As an indicator of the project's success, continuous grants have been awarded to Wayne County
each year since 1993. Some of the project funding is spent on watershed-wide activities such as sampling
and monitoring, but the majority of the funding is passed to local communities and nonprofit groups for
watershed management activities such as design and construction of pollution controls.
The Rouge River Watershed is largely urbanized, spans approximately 438 square miles, and is home to
over 1.4 million people in 48 communities and 3 counties. The Rouge Project initially concentrated efforts
on the control of combined sewer overflows (CSOs). The early objective of the project singled out the
control of CSOs as a means to improve water quality in the river. However, as the project unfolded, the
monitoring showed that other sources of pollution needed to be controlled before full restoration of the river
would be achieved throughout the watershed. In fact, the data showed that even if all of the CSO discharges
were totally eliminated, the waters still would not meet water quality standards. Based upon what was
learned, the Rouge Project has taken a wide-angle lens view of pollution sources. The project now has a
holistic approach to consider the impacts from all sources of pollution and use impairments of receiving
waters. The project is therefore proceeding on parallel paths, controlling CSOs, while pursuing the
watershed approach to address storm water management, flow management, non point sources, failing on-
site sewage disposal systems, habitat and riparian restoration, and the development of new recreational
opportunities.
One of the primary goals of the Rouge Project is to guide state and federal regulatory policy in wet weather
pollution control. The chief way that the project guides policy is by demonstrating the implementation of
BMPs for an urban river system, and by demonstrating workable governmental processes that support the
implementation of watershed restoration. Critical to both the technology design and to the processes of
60
-------
government is the ability to measure individual BMP performance and to measure the cumulative beneficial
impacts of all efforts in the watershed.
The Rouge Project distinguishes itself among other watershed efforts by not relying on a single point of
institutional accountability. The federal, state, county, and municipal units of government are in agreement
that watershed management is the ultimate responsibility of each local municipality. The municipalities
collaborate with each other, and they have formed alliances in seven subwatershed groups that range in size
from about 20 square miles to over 80 square miles. The municipalities also support watershed-wide
activities for monitoring, geographic information systems (GIS), technical information sharing, public
involvement and grant administration. The Rouge Project has included a large number of voluntary
activities, particularly in the arena of storm water management, where mandatory federal regulations will
not take effect until 2002, and state policy has been through a voluntary General Permit since 1997.
THE SERIES OF STAGES
The Rouge Project uses a series of stages to link BMP performance to receiving water impacts. The project
has found that it is necessary to proactively build the links so that useful measurements and conclusions can
be obtained.
There are five stages that span the BMP process:
• Design criteria for BMPs,
• Assessment of water quality needs by subwatershed,
• Promotion of the implementation of the most effective BMPs in each subwatershed,
• Standard protocols for receiving water quality measurements, and
• Watershed wide monitoring program and data assessment.
Each of the stages has three principal components:
• A technical basis developed from engineering analysis;
• A basis of authority, which typically is a process of government, such as an ordinance,
adaptation of existing regulation, new regulation, or as simple as a peer-supported voluntary
guideline; and
• A physical measurement of the effectiveness of the stage, such as a performance monitoring
program, a watershed monitoring program, or other type of assessment.
All three components are necessary. The technical basis provides the functional fit of the BMP into the
engineered watershed ecosystem. An authority is needed to provide a reason and motivation for the BMP
to be implemented in the context of other public needs - education, safety, transportation, etc. The
measurement component is the way to test the success of implementation and assess the need for further
action.
The concept of looking at the entire BMP process is important, because of: 1) the relatively long time span
for BMP implementation; 2) the complexities of multiple parties responsible for implementation; and 3) the
evolving learning curve of watershed management technologies.
61
-------
The concept of a subwatershed is also important in the establishment of links between BMP performance
and receiving water impacts. Subwatersheds allow us to tackle the larger problems of a watershed in a
series of smaller bites. For example, a subwatershed that is a headwater area allows the suite of BMP
solutions to focus on headwater protection, which may not require dealing with the complications of CSO
controls typical in downstream areas of the Rouge watershed. The subwatershed provides a smaller
geographic area, a smaller range of technical solutions, a smaller list of objectives, and a small group of
stakeholders - overall, a more manageable problem to tackle. The delineation of subwatersheds may
therefore be an important step in the BMP process. A discussion of the locally controlled subwatershed
delineation process in the Rouge River watershed is given by Cave, et al., 1998.
DESIGN CRITERIA FOR BMPS
The first link between BMP performance and receiving water quality improvement comes at the beginning
of the staged BMP process - that being the design criteria of the project.
Technical Basis
The Rouge Project has developed design criteria, or facilitated the development thereof, for a number of
efforts to standardize design criteria for BMPs. Examples include:
• Development of a guide for planning and estimating costs for BMPs that is tailored to metropolitan
Detroit applications. This guide presents a "public works director" view of design criteria and cost
estimates for 23 categories of BMPs. Figure 1 shows an example entry from this guide. (Ferguson, et
al, 2001)
• New design standards for storm water management in Wayne County which establish peak discharge
rates, restrict activities in flood plains, and set forth provisions for operation and maintenance of storm
water facilities. (WCDOE, 2000)
• Development of design criteria for demonstration size CSO storage and treatment basins. These criteria
established a "demonstration" basin size to capture 0.17 inches of runoff compared to the state
regulatory agency presumptive size of 0.35 inches of runoff. (Alsaigh, 1994)
• Water quality models for evaluation of river impacts. These tools are primarily used in work with the
state regulatory agency (MDEQ) for CSO basin sizing and with performance evaluation of the basins
and storm water detention pond operation. The water quality models utilize the US EPA SWMM and
WASP models, and are configured for both dynamic and steady state simulations.
Wayne County has invested in a program of technology transfer to disseminate the design criteria that the
Rouge Project develops. The technology transfer program includes an educationally acclaimed website
(www.rougeriver.com), training programs and publications that are for audiences in the Rouge watershed
and in other watersheds. The Rouge Project also offers a technical extension service for communities in the
Rouge River watershed.
62
-------
Type:
Description:
Function:
Application:
Site Requirements:
Effectiveness:
Who Does It?
Design Requirements:
Basis for Cost:
Who Pays For It?
Cost ($)
Non-Structural, Urban Source Control BMP.
Periodic inspection of on-site sewage disposal systems (OSDS) and regular pumping of septic tanks
will prevent, detect and control spills, leaks, overflow and seepage from on-site sewage disposal
systems.
Prevents premature failure of on-site sewage disposal systems and detects problems that will
minimize pollution.
Maintenance practice.
Availability of a plan showing the location of the on-site sewage disposal systems.
Pumping of septic tanks on a regular basis and inspection of the on-site sewage disposal system can
prevent premature failure and detect problems so that repairs can be less costly. An inspection of the
on-site sewage disposal system is recommended every 5 years. Health Departments recommend a 3-
year cleaning cycle for septic tanks.
Can be done by municipal staff or by county health agency.
Risers on septic tanks make location, inspection and pumping easier.
Pumping must be done by a Licensed Septage Waste Servicer. A Registered Sanitarian should
perform inspections or a person certified as a septic system evaluator by the local health department
orNSF International.
Cost of regular inspections of on-site sewage disposal systems. Assumes 20 percent of a
community's septic tanks are inspected each year so that a five-year cycle is maintained. Time for
inspection usually takes 1 to 3 hours, but can take much longer if the location is not well defined.
Cost per septic tank for pumping and proper disposal of the contents
Paid for by municipality
Inspection: $100/hour, 3 hours per site including reporting and travel time. (This time can be
substantially more if the on-site sewage disposal system is difficult to locate.)
Pumping: $100-$150/septic tank including disposal
FIGURE 1 - SEPTIC SYSTEM MAINTENANCE
(Excerpt From "Cost Estimating Guidelines: Best Management Practices And Engineered Controls", Rouge River National
Wet Weather Demonstration Project)
Authority
Technical criteria need to have a basis of authority to assure that BMPs are implemented in accordance with
the technical standards. The Rouge Project has been successful in taking its design criteria and working
these into ordinances, regulations, model ordinances, etc. For example, the project implemented new storm
water management standards for Wayne County in October 2000 (WCDOE, 2000). Key features of these
standards include:
• Storm water outlet design, and sizing and location of the outlet with regard to stream capacity
• For drainage areas of 5 acres or more, the runoff rate must not exceed 0.15 cfs per acre for a 100-year
storm; for less than 5 acres, the runoff rate must not exceed 0.15 cfs per acre for a 10-year storm
63
-------
• Storm water runoff should conform to natural drainage patterns where feasible
• Storm water management systems should not generally be constructed within the 100-year flood plain;
work within the flood plain has restrictions and requires compensatory storage and riparian habitat
mitigation.
Another example of bringing technical criteria into law is the Slate of Michigan Wetlands Mitigation Bank.
The Rouge Project worked with the State of Michigan Department of Environmental Quality to develop a
wetlands banking system (State of Michigan, 1998). Units of government can apply for membership in the
bank, and Wayne County was successful in becoming a member. The program establishes criteria for
design, construction and maintenance of wetlands. At this time, over 10 acres of wetland are built or under
construction for the bank.
A final example of the authority for promoting design criteria is in the CSO control program for 157
overflow points in the Rouge River. The authority was based on a court-ordered compromise under the US
EPA and Michigan Department of Environmental Quality NPDES (National Pollutant Discharge
Elimination System) program. The compromise ordered a phased approach to CSO control. Phase I
required the elimination of raw sewage and the protection of public health for approximately 40 percent of
the combined sewer area. The Phase 1 control plan was based on the technical design criteria (capture 0.17
inches of runoff) developed by the Rouge Project noted earlier. Under Phase I, six communities separated
their sewers and eight communities constructed basins to evaluate varying sizes and control technologies of
CSO basins.
Measurement
The third component in the design criteria stage is that of measurement. Design criteria are first established
with computer models, engineering analyses, or results from other locations. The criteria need to be tested
and examined, and ultimately refined based on the actual implementation in the watershed. The Wayne
County Storm Water Management Program also requires post-construction monitoring, and we will learn
from these new data. The Michigan Wetlands Banking Program requires 5 years of biological and water
quality monitoring.
The CSO Phase 1 program has completed an extensive program of monitoring to determine if the
demonstration size basins had met the water quality standards. A work group of staff from the Michigan
Department of Environmental Quality, the NPDES permitted communities, and from the Rouge Project
evaluated 2-years of measurements of basin influent and effluent and receiving water quality data. The
Michigan Department of Environmental Quality has certified 6 of the 9 basins to date, and the design
criteria that were established are being used to plan the next phase of controls.
ASSESS WATER QUALITY NEEDS BY SUBWATERSHEDS
In the previous examples, CSO locations were known and locations for wetlands banking sites were
governed by land use opportunity. What happens when there is a watershed sector suffering from
eutrophication in an impoundment, stream bank erosion, and high wet weather bacteria?
This the second stage of the BMP process when the issue is not the design criteria, but the questions are:
what is the type of technical solution, and at what scale should it be applied? What are the most
appropriate BMPs for the specific environmental needs?
64
-------
Technical Basis
The technical works at this stage is to thoroughly and systematically analyze the needs of each part of the
watershed. In the Rouge Project, this stage was completed through a series of subwatershed management
plans. The subwatersheds can be classified in three categories: those in headwaters where issues involve
preservation, open space is relatively plentiful, and development ordinances can be useful; those at the most
downstream and developed reaches, where the land is fully developed, and the issues are restoration and
redevelopment; and those in growing suburban areas, which have a mix of issues from the other areas.
The seven subwatershed management plans for the Rouge watershed specify a series of BMPs to be
implemented over the next 5 years (Rouge Subwatershed Advisory Groups (7), 2001). General goals for the
period after 5 years were established, and these goals will be formulated into more specific BMP
implementation after the first 5 years of progress are complete. The BMPs have been identified through a
collaborative planning process involving the local units of government and Counties responsible for
performing the work, the general public, and the state regulatory agency. Over 900 BMPs have been
identified for implementation by 38 communities and agencies in the watershed.
Authority
The subwatershed management plans were developed and implemented as part of the Michigan Storm
Water General Permit of 1997 (State of Michigan, 1997). The US EPA has accepted the General Permit as
meeting criteria for EPA's national Phase n storm water program, which takes effect in 2002. In tailoring
the General Permit to the needs of the Rouge watershed, the Project has attempted to incorporate watershed
planning components from other of water resource management programs, including:
• TMDL Program: Various segments of the Rouge River are listed on the federal Clean Water Act
Section 303(d) list for various parameters. The Total Maximum Daily Loads (TMDLs) for these
segments are not scheduled for completion until approximately 2005. The river will require multiple
TMDLs (approximately 15) that may result in conflicting implementation strategies in the watershed as
a whole. Under the USEPA's proposed TMDL regulations, use of the watershed approach is
encouraged, an approach already being implemented in the Rouge Project.
• Water Quality Trading Program: The State of Michigan is in the process of completing its Water
Quality Trading Program rules. Through this program, the trading of nutrients in impaired water bodies
(for which TMDLs have not yet been developed) can only occur where an approved watershed
management plan has been developed. Unlike other "approvable" watershed plans, the watershed
management plan for the trading program must include a "cap" and allocations.
As described earlier, the seven subwatershed advisory groups in the Rouge Watershed have developed
watershed management plans as required under the Michigan General Permit. Obviously it is desirable to
develop only one "comprehensive watershed management plan" that will meet stakeholder goals and
objectives as well as all applicable program requirements any other programs that emerge. Therefore, the
Rouge Project subwatershed management plans have a goal of being comprehensive watershed management
plans that will meet objectives of multiple programs. By doing so, both the watershed communities and
regulatory agencies will save time, money and effort by having one plan that fulfills multiple objectives. In
65
-------
addition, these comprehensive plans will provide much needed certainty to the communities, counties and
other stakeholders in planning for watershed management activities and expenditures.
Measurement
The Michigan General Permit requires that each subwatershed management plan include a description of the
measures that will be used to gauge progress on meeting the goals of the plan. As Rouge Project
representatives met with the Michigan Department of Environmental Quality to examine the requirements
for measurement, we determined that the MDEQ would be satisfied with rather general forms of
measurement. As a result, the Rouge Project established an overall architecture for the measurement
program, and key elements of the program are noted below:
• The BMPs identified by the stakeholders should be designed to address all known causes of water
quality standards violations
• Each BMP is "scored" relative to its potential ability to improve major designated uses of the receiving
water, including fish propagation, partial human body contact, boating, and aesthetic enjoyment
• Measurements of the effectiveness will be made based on in-stream flow and water quality monitoring
stations, along with biological surveys
• The performance standards and budgeting assumptions for all the actions have been standardized
throughout the watershed to help assure that the implementation approach for various BMPs is
relatively standard
• At the end of the 5-year period, the water quality results achieved will be assessed, along with the costs
and other implementation issues
• A subsequent 5-year program of BMPs will be developed through the upcoming federal Phase n storm
water program
Now that subwatershed communities are planning local actions to improve Rouge River water quality, the
potential of these actions to solve condition and use problems are being evaluated. Figure 2 shows the
structure for developing an action score for each BMP. The effect rating for actions can be combined with
condition and use ratings, as shown below, to produce an overall "action score" which is location-specific.
Logically, the highest score should represent a case where the most appropriate action has the greatest
beneficial effect on the worst river condition and use problems. Rating values have been assigned
accordingly. Action scoring of this type is necessarily based on "expert opinion", not hard data; but the
score numbers should provide a useful scale for comparing the likely benefits of applying different actions
to different problems in different watershed situations.
The effectiveness of community actions is highly dependent on where and when actions occur, and how
well they address river quality problems. In general, the most beneficial actions are those, which have the
most direct effects. Other less beneficial actions have indirect or only potential effects. Some actions may
be highly effective in one location or season and ineffective in another. Moreover, an action may improve
one kind of river condition or use, and have no effect or even undesirable effects on others. In short, the
effectiveness (or cost-effectiveness) of community actions can be evaluated only in the context of local river
conditions and public uses.
The effects of community actions on Rouge quality can best be measured at monitoring stations where
historical conditions are known. Prior data for river quality indicators at these stations provide a yardstick
for monitoring future trends in condition or use quality. The data provide a basis for gauging the long-term
66
-------
Public Uses
River Conditions
Community Actions
ARE AFFECTED BY.
AFFECT.
Use Category
1. Fishing
2. Canoeing & Boating
3. Wading & Swimming
4. Aesthetics
Use Quality
• Full
• Limited
• Restricted
Rating Value
1
2
3
Condition Indicators
1.
2.
3.
4.
5.
DO
Flow
Bacteria
Aquatic Life
Stream Habitat
Condition Quality
• Good
• Fair
• Poor
Rating Value
1
2
3
Community Actions
1. BMPs
2. Etc.
Effect Quality
• Direct Effect
• Indirect effect
• Potential effect
• No effect
Rating Value
3
2
1
0
Use
Rating x
(1-3)
Condition
Rating x
(1-3)
Effect
Rating
(3-0)
Action
= Score
(0-27)
FIGURE 2 - ROUGE RIVER NATIONAL WET WEATHER DEMONSTRATION PROJECT:
BMP ACTION SCORING SYSTEM
effectiveness of community actions as well. Site-specific ratings of various actions can help communities to
design local programs, which yield the greatest returns for their money and effort.
PROMOTING THE IMPLEMENTATION OF THE MOST EFFECTIVE BMPS
As we come to the third stage of the whole BMP process, the design criteria have been established and the
plan is in place for what BMPs are needed, where, and at what scale. The next challenge is implementation
67
-------
— how do we implement the plan and build the projects that best fit the environmental needs and meet the
design criteria?
Successful implementation is difficult in watershed management because there is seldom one agency with
funding and authority to perform all the work. In addition, implementation often relies partially on
voluntary efforts. Consequently, there are no guarantees that design criteria will be used or that BMPs will
be implemented in accordance with a desired schedule. The Rouge Project has relied again on its three-part
formula of a sound technical basis, an authority, and a measurement system to make progress with
implementation.
Technical Basis
The Rouge Project has developed a program management approach to promote the implementation or
construction of BMPs that meet the design criteria and are in accordance with the plans. The most powerful
tool that the Rouge Project has for implementation is a source of funding. The US EPA demonstration
grant funds are primarily used for sponsoring projects by stakeholders in the watershed. Over 93% of all
the grant funding received has been given as "subgrants" to communities for the design and construction of
CSO, storm water, and non point source BMPs.
The subgrants are offered on a competitive process to communities, agencies and non-governmental
organizations in the Rouge watershed that meet minimum qualifications. Since October 1997, the project
has issued "Notices of Grant Availability" at approximately six-month intervals. The regularity of these
grant notices is designed to facilitate the funding of projects by the grantee communities and agencies. The
funding is a maximum of 50% on a reimbursement basis, so each grant recipient needs to encumber local
matching funds for their projects, which can take six or more months.
The Notices of Grant Availability specify requirements for proposals from communities and establish a date
for submittal and project evaluation criteria. The Notices also identify the types of activities that will be
eligible for funding, and these activities have included:
• wetlands creation or restoration
• habitat and recreational opportunities
• storm water management
• on-site sewage disposal system management
• illicit discharge elimination
• public education on storm water
• geographic information system implementation
• other projects that implement the subwatershed management plans.
Figure 3 shows the evaluation criteria that have been used in recent competitive proposal selection. A
technical review team comprised of representatives of the County and other independent agencies performs
the proposal evaluation.
-------
CRITERIA
1.
2.
3.
4.
5.
6.
7.
Consistency with the watershed management goals of the subwatershed
management plan and the Rouge River restoration and its national
demonstration goals. Higher scores will be given to those projects that most
directly improve water quality.
Consistency with the community's or agency's Certificate of Coverage for the
Storm Water General Permit and subsequent subwatershed management
plan and storm water pollution prevention initiative
Availability of other funding sources. If other sources are available, scoring
will be lower.
Performance of the community in timely execution and progress and
expense reporting of projects under previous interagency agreements. .
Cooperative approaches with other communities or agencies.
Cost-effectiveness and timely schedule of the proposed project.
Clarity of the proposal and conformance to the submittal requirements.
WEIGHT
30
15
10
20
10
10
5
FIGURE 3 - TYPICAL CRITERIA FOR PROPOSAL EVALUATION, ROUGE RIVER NATIONAL WET
WEATHER DEMONSTRATION PROJECT
Authority
In this stage, the authority for the implementation effort rests with the Steering Committee of the Rouge
River Watershed. This is a group representative of the counties, municipalities, subwatersheds, regulatory
agencies and other parties with oversight over the project. It is a group of peer communities that governs
by consensus. The Steering Committee reviews the notices of grant availability and the evaluation criteria,
and then reviews and ratifies the selection process. The Steering Committee is an ad hoc group without legal
authority, but is operates on a consensus basis. In 2002, the communities of the Rouge watershed began
planning discussions to form a Local Management Assembly to replace the Steering Committee with a more
formal organization having limited legal authority through inter-governmental agreements.
Measurement
In this stage of the whole BMP process, the most useful measurement is BMP implementability. Such
measures should address any barriers to implementation, what would be done differently next time, and
what lessons were learned. The project is seeking practical advice that is in the language of the local
community public works department director.
Each subgrantee is required to submit a report that summarizes the implementation of the BMP project.
The following are examples of reporting on the BMP implementation:
• Erosion Controls at Construction Sites - compared fabrics, fences, and hay; found hay most versatile
• Catch Basin Cleaning - found 3-year frequency optimal in terms of cost and effectiveness in
maintaining catch basin functionality
69
-------
• Stream Bank Stabilization - improved designs for bioengineered stabilization, as well as traditional
stone bank protection; developed training for municipalities in stabilization design and construction
practice
• Public Education Projects - resulted in surveys that measured public opinion (Powell, et al, 2000)
STANDARD PROTOCOLS FOR RECEIVING WATER MEASUREMENTS
The next stage in the whole BMP process is the use of standard protocols for field measurement. Once
there are BMPs built according to design criteria and fulfilling watershed protection needs, then uniformity
in measuring receiving water measurements is required.
Technical Basis
The Rouge Project has spent considerable effort in analyzing ecosystem health and receiving water quality,
and then determining the key parameters to be measured.
Historically, the Rouge River has been damaged by industrialization and suburban expansion. The river's
name reflects the inherent problem of erosion of the river's red clay soil banks even from the early days of
French settlers 300 years ago. Since industrialization, public health agencies measured oils and greases and
toxics such as mercury and PCBs in the sediments. The Rouge Project began a major annual monitoring
program in 1993. Those surveys have shown the following pollutants to be the main problems in the
Rouge:
• Dissolved oxygen deficits, particularly downstream of combined sewer overflows, but also
upstream in impoundments and reaches of the river affected by sanitary discharges
• Extremes of flow - either due to increasing impervious areas and flash flooding, or due to
extremely low flow
• Pathogens from combined sewers, leaking septic systems, sanitary sewer overflows, and
illicit connections to storm drains
• Nutrients from lawn fertilizers and sanitary discharges
Metals and toxics have generally not been a problem, except in the sediments of the most downstream
portion of the river. There are also some hot spots of sediment contamination, and one lake that had been
contaminated with PCB in the sediments. This lake was dredged in 1997 and 1998, and it is an example of
an easily measured BMP. The removal of the contamination could be measured, the bottom dredged deeper
and fish stocked. Water quality measurements have confirmed the viability of the new fishery and new
recreational uses of the lake. There is now more recreation, fishing, boating, and a triathlon celebrating its
second year in 2001.
70
-------
Authority
The Rouge Project has established definitive standards for measurement. Because it is a federally funded
demonstration project, the protocols for all measurements are established in accordance with quality
assurance and control standards established by the US Environmental Protection Agency. The US EPA
provides grant funding for a portion of the sampling cost. The project has demonstrated the effectiveness
of a variety of sampling and modeling techniques and has published the information on the Rougeriver.com
web site. By using the web site, communities that need to develop less extensive sampling programs can
benefit from the experience of the Rouge project.
A Field Sampling Plan (FSP) Preparation Guide has been developed. This guidance document serves as a
template for the preparation of site-specific FSPs. The FSP Preparation Guide also serves as a review
checklist for quality control reviews to ensure that the appropriate level of detail is provided in the FSP.
Activities that are undertaken routinely in a consistent manner are documented in Standard Operating
Procedures (SOPs). SOPs are available for laboratory methods (e.g., the 5-day Biochemical Oxygen
Demand Determination) and field sampling (e.g., sediment coring) techniques.
Each laboratory under contract to Wayne County is responsible for implementing a quality assurance
program specifically designed for laboratory activities. As part of this program, laboratories must document
and update SOPs regularly in their Quality Assurance Program Plans (QAPP). The Rouge Project maintains
on file current copies of all subcontract laboratory QAPPs. Only EPA approved analytical methods are
used for analyses of samples collected as part of the Rouge Project. For those activities, which require
modification of existing SOPs or development of new SOPs, internal review and approval will be sought
from EPA prior to their use.
Measurement
An example of the detail that the program has achieved is given by the evaluation of the Cedar Lake
detention pond shown in Figure 4. In this example, rainfall, influent and effluent data were analyzed
concurrently as part of the detailed examination of the wet detention pond.
WATERSHEDWIDE MONITORING PROGRAM AND DATA ASSESSMENT
The preceding stage of the entire BMP process yields an important end product —a comprehensive means
of measuring the collective contribution of many BMPs to the progress of water quality improvement. The
Rouge Project has successfully monitored the watershed since 1994 through a system of 7 continuous flow
and dissolved oxygen gages and dozens of dry weather grab sampling sites. Special studies have been
conducted on an annual basis to develop more information on phosphorus loadings from fertilizer, sediment
oxygen demands, time of travel, impoundment reaeration, and total residual chlorine, among other issues.
As an example of a low cost method of evaluating ecosystem health, frog and toad surveys have been
conducted for the last three years in headwater areas. These surveys, which have brought out an increasing
number of public volunteers each year, provide useful information with the added benefit of bringing people
to the resource which will hopefully assist with pollution prevention through increased awareness.
71
-------
Figure #B1
EXISTING
EXTENDED DETENTION PDND BMP #1 SUMMARY DATA: STORM #1
Site Informetion Storm Preaptaflon Summary
Pi«k¥8te IDs: n: PA/0 1009311 Out P'QDIOO&t! "Mai Brte Pr«n3rtrt^D^;
TV^ * B.n«n3 B,^**,*^ C^^to
•»*,«**«
I tilt
I
nsmEroa
Site, HyMra^H
I
*
i \/H
1 v \
^L>— - — = — •
OD
ES
00 £
in &
m mm
FIGURE 4- MEASUREMENT OF CEDAR LAKE BMP PERFORMANCE
Through its annual surveys, the Rouge Project has been able to document a continuing improvement in
dissolved oxygen downstream of the now controlled CSO discharges. The annual surveys also provide a
basis for further investigation and correction of other pollution sources. Among the benchmarks that future
annual surveys will consider are the following:
Flow variability
• Restrict peak flow rates at critical points
• Do not allow critical reach to meet the peak more than 10% of the time
12
-------
Nutrients
• Phosphorus limited not more than 0.05 mg/1 total phosphorus
Soil Erosion and Sedimentation
• Settleable solids or suspended solids not present in concentrations that interfere with designated uses
Dry Weather Total Suspended Solids
• Based on achieving desired aesthetic use, maintain or achieve TSS below 80 mg/1 in dry weather
Loss of Natural Features
• Benchmark compared to status in year 2000
Passive and Active Recreation
• Dissolved oxygen standard 4 mg/1 or 5 mg/1, depending on the location
• Bacteria standards
SUMMARY
The annual assessment of water quality completes the stages of the whole BMP process that the Rouge
Project uses to measure the performance of BMPs with respect receiving water impacts. In the year 2000,
the annual assessment showed that the Rouge River met the dissolved oxygen standards 94% of the time in
its most downstream reaches. Only six years ago, the river was only meeting the dissolved oxygen
standards in these reaches about 30% of the time, or less. Wildlife are responding, with ever increasing
numbers and varieties offish, birds, macroinverterbrates, and other species.
The staged approach to BMP performance allows the Rouge Project to measure, and continually improve
each step of the watershed management process. This approach has allowed the Rouge Project to meet its
two main goals; first, to make great progress in restoration in the Rouge watershed; and second, to share
practical and transferable results with other watersheds and demonstrate the implementation of wet weather
pollution control policy.
ACKNOWLEDGMENTS
This paper represents a summary of select elements from the ongoing efforts of many individuals and
organizations who are involved in the restoration of the Rouge River. The authors also gratefully
acknowledge the assistance of Ms. Charlotte Nichols for her assistance in the preparation of this manuscript.
The Rouge River National Wet Weather Demonstration Project is funded, in part by the United States
Environmental Protection Agency (EPA) Grants #X995743-02 to 07. The views expressed by individual
authors are their own and do not necessarily reflect those of the EPA. Mention of trade names, products or
services does not convey, and should not be interpreted as conveying, official EPA approval, endorsement,
73
-------
or recommendation.
REFERENCES
Alsaigh, R., 1994. CSO Facilities Design Parameter Report. Rouge River National Wet Weather
Demonstration Project, Report CSO-TR02.00.
Cave, K. A. and J. D. Bails, October 1998. Implementing a Model Watershed Approach Through A State
General Storm Water NPDES Permit, Proceedings of WEFTEC 98, Water Environment Federation.
Ferguson, T., R. Gignac, M. Stoffan, A. Ibrahim, and J. Aldrich, May 1997. Cost Estimating Guidelines:
Best Management Practices and Engineering Controls. Rouge River National Wet Weather Demonstration
Project, Supplemental Report NPS-SR10.00. Updated as NFS TR25.00, May 2001.
Lower 1 Rouge River Sub watershed Advisory Group, May 2001. Lower 1 Rouge River Sub watershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Lower 2 Rouge River Subwatershed Advisory Group, May 2001. Lower 2 Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Main 1-2 Rouge River Subwatershed Advisory Group, May 2001. Main 1-2 Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Main 3-4 Rouge River Subwatershed Advisory Group, May 2001. Main 3-4 Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Middle 1 Rouge River Subwatershed Advisory Group, May 2001. Middle 1 Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Middle 3 Rouge River Subwatershed Advisory Group, May 2001. Middle 3 Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
Powell, J, and J. Bails, February 2000. Measuring the Soft Stuff - Evaluating Public Involvement in Urban
Watershed. Proceedings WEF Watershed 2000.
Rouge River National Wet Weather Demonstration Project, May 2001. Common Appendix for Rouge
Subwatershed Management Plans Submitted in fulfillment of the MDEQ Stormwater General Permit,
Rouge River National Wet Weather Demonstration Project.
State of Michigan, Department of Environmental Quality, July 30, 1997. National Pollutant Discharge
Elimination System, General Wastewater Discharge Permit, Storm Water Discharges from Separate Storm
Water Drainage Systems, Permit No. MIG610000.
State of Michigan, Department of Environmental Quality, 1998. Wetlands Banking Regulations.
Upper Rouge River Subwatershed Advisory Group, May 2001. Upper Rouge River Subwatershed
Management Plan, Rouge River National Wet Weather Demonstration Project.
74
-------
Wayne County, Michigan, Department of Environment, October 19, 2000. Storm Water Management
Program (Version 1.0).
KEY WORDS
BMP (Best Management Practice)
Stormwater
Watershed
Receiving Waters
Wetlands
Water Quality
Wet Weather
Environmental Regulations
75
-------
THE MARYLAND STORMWATER MANAGEMENT PROGRAM
A NEW APPROACH TO STORMWATER DESIGN
Stewart R. Comstock, P.E. & Charles Wallis, P.E.
Maryland Department of the Environment
Baltimore, Maryland
Maryland's original stormwater management program was developed as part of the Chesapeake Bay
Initiatives in 1984. At that time, the prevailing attitude was that controlling flooding caused by increases in
new development would maintain the quality of receiving streams. Thus, the original Code of Maryland
Regulations (COMAR) specifying stormwater management was slanted towards flood control. Much
experience has been gained in years since Maryland implemented the original program.
Recently, additional emphasis has been directed on controlling the quality of runoff from land use changed
by urbanization and the quantity of this runoff to reduce stream channel erosion. Recognizing that the
State's stormwater management program had not changed in over a decade, the Maryland Department of the
Environment (MDE) proposed modifications to the COMAR in July 2000. The primary goals of the
proposed regulations were to refocus the overall objectives for controlling runoff from new development
and promote environmentally sustainable techniques. To that end, MDE developed the 2000 Maryland
Stormwater Design Manual, Volumes I & n (MDE, 2000) to establish stormwater design criteria and
provide specific procedures for local jurisdictional use in improving existing programs for nonpoint source
pollution control within the Chesapeake Bay and its tributaries as well as coastal bays. As such, the Design
Manual would serve as the primary source of stormwater management information for the development
community and regulatory agencies throughout the State.
In the beginning, MDE developed the Design Manual to address three goals to: (1) protect the waters of the
State from the adverse impacts urban stormwater, (2) provide design guidance on effective structural and
nonstructural best management practices (BMPs) for new development sites, and (3) improve the quality of
BMPs that are constructed in the State. While drafting the Design Manual, MDE recognized that the project
was evolving into a more comprehensive approach to stormwater design. Included in this approach was
better guidance for total site design and incentives for environmentally sustainable or "green" development
techniques. The projected outcome of this new approach would be site designs that more closely mimic
natural processes and reduce reliance on the use of structural management techniques to treat stormwater
runoff.
As a final product, the Design Manual shows great promise in accomplishing the goals and objectives
established in the beginning and during this project. The adopted manual serves as a primary source of
stormwater design information for the development community and regulatory agencies in both Maryland
and in many other areas.
76
-------
1. Introduction
Maryland's current stormwater management program was established in 1984 when the prevailing attitude
was that if the flooding caused by increases in runoff volume from new development was controlled, the
quality of receiving streams could be sustained. Hence, the original Code of Maryland Regulations
(COMAR) specifying stormwater management design requirements were slanted toward flood control.
Specifically, new development was required to reduce post-construction flows of the two and ten-year
design storms to pre-development levels. This policy, known as peak management, was thought to address
stream channel erosion concerns as well as provide adequate flood control in receiving waters. Although a
general definition of water quality management was included in the original regulations, specific guidelines
and design criteria were absent from the State's original stormwater management program.
More recently, more emphasis has been placed on controlling the quality of runoff from land use changed
by urbanization and the quantity of this runoff to prevent stream channel erosion. Recognizing that
Maryland's stormwater management program had not changed since its inception, the Maryland
Department of the Environment (MDE) proposed modifications to COMAR in 1993 to refocus the overall
objectives of Maryland's efforts toward controlling new development runoff. The goals of these
modifications included the control of more frequent storm events, prevention of stream channel erosion,
limiting the number of stormwater management waivers, and providing incentives to developers to design
projects in an environmentally friendly way. MDE solicited and received an enormous amount of
recommendations from numerous organizations and individuals including State and local government
officials, developers, design engineers, and environmental groups. While there was general agreement that
the State's stormwater management program needed revision, there was a huge disparity in the comments
regarding how the program ought to be revised. One common suggestion was that COMAR should set
general policy and that specific design requirements should be compiled in a single, separate guidance
document. Consequently, MDE commenced work on the development of a stormwater management design
manual in 1995.
Maryland's stormwater management program has been considered one of the more advanced of its kind.
However, the original program's focus on flood control and its reliance on a preference list for best
management practice (BMP) selection hampered MDE's goals to more effectively control nonpoint source
pollution, reduce stream channel erosion, and promote innovative stormwater design. The 2000 Maryland
Stormwater Design Manual, Volumes I & n was developed with three distinct goals to; 1) protect the
waters of the State from adverse impacts of urban stormwater runoff, 2) provide design guidance on the
most effective structural and non-structural BMPs for development sites, and 3) improve the quality of
BMPs that are constructed in the State, specifically with respect to their performance, longevity, safety,
maintenance, community acceptance, and environmental benefit. On October 2, 2000, the Maryland
Department of the Environment (MDE) adopted new stormwater regulations including the Design Manual.
Recognizing the demand for environmentally sustainable or "green" design, these regulations represent a
more comprehensive approach to stormwater design. Included in this approach are better guidance for total
site design and incentives for nonstructural BMPs. The anticipated outcomes of this program are projects
designed to more closely mimic natural processes.
While going a long way in promoting sustainable development, the State's stormwater management
program is not the only set of rules that govern development. There are several State and local programs
(e.g., Critical Areas, Forest Conservation, Wetlands Protection) that promote natural resource conservation.
There are also local zoning regulations that govern land development. Although the goal of these diverse
77
-------
programs is to protect the environment, there are instances where green development practices are
discouraged and older, less sustainable standards are required.
It is difficult to accommodate the requirements of the full spectrum of resource protection programs.
However, the Design Manual recognizes the importance of each and encourages these principles during
project design. Accordingly, the State's approach to stormwater design may be summarized as a three-step
process: avoidance, minimization, and mitigation. The first step, avoidance, is not just resource protection,
but also includes avoiding development practices such as large-scale clearing and mass grading, structural
fill, and suburban sprawl that have negative impacts on local hydrology. Any reduction in imperviousness
or a site's footprint significantly reduces the amount of stormwater runoff. The second step is minimization.
After all options for avoiding impacts are expended, the designer should incorporate practices that either
replace or disconnect impervious surfaces. For example, using green roof technology, permeable
pavements, or promoting sheet flow will also reduce runoff. After all options to avoid or minimize have
been exhausted, the remaining runoff must be treated using structural practices to mitigate water quality and
channel stability impacts.
2. The 2000 Maryland Stormwater Design Manual
2.1 Volume I
The first volume of the design manual presents the basic technical information for designing stormwater
management in Maryland. Its five chapters present background material on the importance of controlling
stormwater runoff, general performance standards for stormwater management, basic stormwater design
objectives, minimum design criteria for BMP design, guidance for selecting and locating BMPs, and an
innovative system of "credits" for environmentally sensitive design techniques. The information contained
in these chapters provides for meeting the three goals of the design manual.
2.1.1 Chapter 1 - Introduction
A basic understanding of the impacts of stormwater runoff on watersheds is critical before any stormwater
design criteria can be established. Chapter 1 provides fundamental information on the effects of stormwater
runoff on water quality, groundwater recharge, stream channel habitat, overbank flooding, and flood plain
expansion. This information is critical if innovative stormwater designs are to be successful.
Chapter 1 also establishes twelve general performance standards for stormwater design and provides
guidance on how to use the manual. The chapter concludes with a brief description of new stormwater
design requirements and a list of all symbols and acronyms used within the manual.
2.1.2 Chapter 2 - Basic Stormwater Design Criteria
The first goal of the stormwater design manual is to protect the waters of the State from adverse impacts
associated with urban runoff. Chapter 2 presents a unified approach to sizing stormwater BMPs for meeting
this goal. This approach consists of five criteria (see Table 1) that are designed to meet pollutant removal
goals, maintain groundwater recharge, reduce channel erosion, prevent overbank flooding, and pass extreme
floods. Of these criteria, the water quality (WQV), recharge (Rev) and channel protection (Cpv) volumes are
determined by soils, amount of imperviousness, proposed design and/or layout, and implementation of
nonstructural practices. This simplifies calculations, reduces error and/or abuse, and provides direct
incentives to reduce impervious areas.
78
-------
Another important feature of these three volumetric criteria is the relation to natural hydrologic processes.
Explicitly, the Rev criterion is designed to promote groundwater recharge and interflow. Likewise, the
rationale for the Cpv criterion is that runoff will be stored and released in such a gradual manner that critical
erosive velocities during bankful and near bankful events will seldom be exceeded in downstream channels.
While the WQV is the storage volume needed to capture and treat the runoff from 90% of the average annual
rainfall, it also provides management at a critical level (1/3 bankfull elevation) within stream channels.
When considered together, these three criteria capture and treat the runoff from at least 95% of the average
annual rainfall (see Figure 1) and mimic natural recharge and channel forming processes.
Chapter 2 also introduces five groups of structural BMPs and a group of non-structural BMPs that may be
used to meet pollutant removal and groundwater recharge goals. Lastly, this chapter designates certain land
uses as "stormwater hotspots" which may restrict the use of certain BMPs and may require pollution
prevention plans.
Table 1. Summary of Unified Stormwater Sizing Criteria
Sizing Criteria Description
Water Quality Volume WQV = [(P)(RV)(A)]/12
(WQV) (acre-feet) P = 1.0" in Eastern Zone and 0.9" in Western Zone
Rv = 0.05 + 0.009(1) where I is percent impervious cover
A = Area in acres
Recharge Volume Rev = [(S)(RV)(A)]/12
(Rev) (acre-feet) S = Soil Specific Recharge Factor
Rev is a sub-volume of WCv
Channel Protection Cpv = 24 hour extended-detention of the post-developed one-year 24 hour storm
Storage Volume event.
(Cpv)
Cpv is not required on the Eastern Shore of Maryland
Overbank Flood Local review authorities may require that the peak discharge from the ten-year storm
Protection Volume event be controlled to the pre-development rate (QP10). No control of the two-year
(Qpx) storm event (QP2) is required.
For Eastern Shore, provide peak discharge control for the two-year storm event (QP2).
No control of the ten-year storm event (Qpio) is required.
Extreme Flood Consult with the appropriate local reviewing authority. Normally no control is needed
Volume (Qf) if development is excluded from the 100-year flood plain and downstream
conveyance is adequate.
2.1.2.1. Unified Stormwater Sizing Criteria — Water Quality Volume (WQV)
The Water Quality Volume (denoted as the WQV) is the storage needed to capture and treat the runoff from
90% of the average annual rainfall (COMAR 26.17.02). In numerical terms, it is equivalent to an inch of
rainfall multiplied by the volumetric runoff coefficient (Rv) and site area. Treatment of the WQV shall be
provided at all developments where stormwater management is required. A minimum WQV of 0.2 inches
per acre shall be met at sites or drainage areas that have less than 15% impervious cover. Drainage areas
having no impervious cover and no proposed disturbance during development may be excluded from the
WQV calculations.
79
-------
2.1.2.2. Unified Stormwater Sizing Criteria - Recharge Volume Requirements (Rev)
The criteria for maintaining recharge is based on the average annual recharge rate of the hydrologic soil
group(s) present at a site as determined from the United States Department of Agriculture (USD A), Natural
Resources Conservation Service (NRCS) Soil Surveys or from detailed soil investigations. More
specifically, each specific recharge factor (S) is based on the USDA average annual recharge volume per
soil type divided by the annual rainfall in Maryland (42 inches per year) and multiplied by 90% (Table 2).
This keeps the recharge volume calculation consistent with the WQV methodology.
Table 2. Soil Specific Recharge Factors (S)
Hydrologic Soil Group USDA Average Annual
Recharge Volume*
Soil Specific Recharge
Factor (S)
A
B
C
D
*Rawls, Brakensiek & Saxton, 1982
18 inches/year
12 inches/year
6 inches/year
3 inches/year
0.38
0.26
0.13
0.07
The recharge volume is considered part of the total WQV that must be addressed at a site and can be
achieved either by nonstructural techniques (e.g., buffers, disconnection of runoff), structural practices (e.g.,
infiltration, bioretention), or a combination of both. Like WQV, drainage areas having no impervious cover
and proposed disturbance may be excluded from recharge calculations. Rev and WQV are inclusive. If Rev
is treated upstream of WQV, then Rev may be subtracted from the WQV when sizing water quality treatment.
The intent of the recharge requirement is to maintain existing groundwater recharge at development sites.
This helps to preserve water table elevations thereby maintaining the hydrology of streams and wetlands
4.5
3.5
Approx. Range for
Channel-Forming Storms
Range for Channel Protection (Cpv
0% 10% 20% 30% 40% 50% 60%
Percentile of Rainfall (707 rainfall events)
70%
80%
90%
100%
Figure 1. Rainfall events captured and treated by the recharge (Rev), water quality (WQV) and channel
protection (Cpv) volumes using 1980 to 1990 rainfall frequency records for Baltimore City
80
-------
during dry weather. The volume of recharge that occurs on a site depends on slope, soil type, vegetative
cover, precipitation, and evapo-transpiration. Sites with natural ground cover such as forest or meadow
have higher recharge rates, less runoff, and greater transpiration losses under most conditions. Because
development increases impervious surfaces, a net decrease in recharge is inevitable.
2.1.2.3. Unified Sizing Criteria - Channel Protection Volume (Cpv)
The primary purpose of the Channel Protection Storage Volume (Cpv) requirement is to protect stream
channels from excessive erosion caused by the increase in runoff from new development. The rationale for
this criterion is that runoff from the one year design storm will be stored and released in such a gradual
manner that critical erosive velocities during bankfull and near-bankfull events will rarely be exceeded in
downstream channels. The method for determining the Cpv requirement is based on the "Design Procedures
for Stormwater Management Extended Detention Structures" (MDE, 1987) and is detailed in Appendix
D. 11 of the Design Manual. The Cpv requirement does not apply to direct discharges to tidal waters or
developments located on Maryland's Eastern Shore.
2.1.3. Chapter 3 - Performance Criteria for Urban BMP DesignThe secondary and tertiary goals of the
design manual are to provide design guidance and improve the quality of BMPs that are constructed in the
State. Chapter 3 promotes these goals by outlining performance criteria for five groups of structural
stormwater BMPs for water quality treatment (see Figure 2). These performance criteria are designed to
ensure that each BMP group is capable of meeting the State's goal of an 80% reduction of total suspended
solids (TSS) from urban stormwater runoff. This allows prospective designers to choose from a variety of
BMPs that best fit individual site needs and still meet the State's pollutant removal goals. Each set of BMP
performance criteria is based on six factors that address general feasibility, conveyance criteria,
pretreatment needs, BMP geometry, environmental and landscaping requirements, and maintenance
concerns.
Stormwater Filtering Systems
• Surface Sand Filters
• Underground Sand Filters
• Perimeter Sand Filters
• Organic Filters
• Pocket Sand Filters
• Bioretention
Open Channel Systems
• Dry Swale
• Wet Swale
Stormwater Ponds
• Micropool Extended-Detention (ED) Ponds
• Wet Ponds
• Wet ED Ponds
• Multiple Pond Systems
• "Pocket" Ponds
Stormwater Wetlands
• Shallow Wetland
• ED Shallow Wetland
• Pond/Wetland System
• "Pocket" Wetland
Stormwater Infiltration
• Infiltration Trench
• Infiltration Basin
Figure 2. Structural BMPs that may be used for "stand alone" water quality treatment in Maryland
2.1.3. Chapter 4 -Selecting and Locating the Most Effective BMP System
In conjunction with the previous chapter, Chapter 4 promotes the secondary and tertiary goals of the manual
by outlining a process for selecting the best BMP or group of BMPs for a development site. The chapter
81
-------
also provides guidance on factors to consider when locating BMPs at a given site. This process is used to
filter those BMPs that can meet the pollutant removal targets for WQV and guides designers through six
steps that screen for watershed factors, terrain factors, stormwater treatment suitability, physical feasibility
factors, community and environmental factors, and locational / permitting factors. These factors, when used
progressively, allow designers to select BMPs that are most suitable for the various physiographic regions
within the State as well as for specific site and design characteristics such as land use or wildlife habitat
enhancement.
Natural Area
Conservation
Disconnection of
Rooftop Runoff
2.1.5. Chapter 5 - Stormwater Credits
One of the major programmatic changes promoted by the Design Manual is the notion that stormwater
management should not rely solely on the use of structural BMPs but should integrate stormwater into the
overall site design process. Chapter 5 supports this philosophical change by advancing a series of
nonstructural design practices that can reduce the generation runoff from a site thereby reducing the size and
cost of structural BMPs. Additionally, these practices provide partial removal of many pollutants. To
promote greater use, these non-structural practices have been classified into six sub-groups (see Table 3.)
with an associated "credit" provided for designers utilizing these progressive techniques.
Table 3. Stormwater Credits for Innovative Site Design
Stormwater Credit Description
Conservation of natural areas such as forest, non-tidal wetlands, or other sensitive areas
in a protected easement thereby retaining their pre-development hydrologic and water
quality characteristics. Using this credit, a designer may subtract conservation areas
from total site area when computing WQV Additionally, the post-development curve
number (CN) for these areas may be assumed to be forest in good condition.
Credit is given when rooftop runoff is disconnected and then directed over a pervious
area where it may either infiltrate into the soil or filter over it. Credit is typically obtained
by grading the site to promote overland flow or by providing bioretention on single-family
residential lots. If a rooftop area is adequately disconnected, the impervious area may be
deducted from the total impervious cover. Additionally, the post-development CNs for
disconnected rooftop areas may be assumed to be forest in good condition.
Credit is given for practices that disconnect surface impervious cover by directing it to
pervious areas where it is either infiltrated or filtered though the soil. As with rooftop
runoff, the impervious area may be deducted from the total impervious cover thereby
reducing the required WQV.
Credit is given when a stream buffer effectively treats stormwater runoff. Effective
treatment constitutes capturing runoff from pervious and impervious areas adjacent to the
buffer and treating the runoff through overland flow across a grass or forested area.
Areas treated in this manner may be deducted from total site area in calculating WCv and
may contribute to meeting requirements for groundwater recharge.
Credit may be given when open grass channels are used to reduce the volume of runoff
and pollutants during smaller storms. Use of grass channels will automatically meet the
minimum groundwater recharge requirement. If designed according to listed criteria,
these channels may meet water quality criteria for certain types of residential
development.
Credit is given when a group of environmental site design techniques are applied to low
density or rural residential development. This credit eliminates the need for structural
practices to treat both Rev and WQV. The designer must still address Cpv and Qpx
requirements for all roadway and connected impervious surfaces.
Disconnection of
Non-Rooftop
Runoff.
Stream Buffer
Credit
Grass Channel
(Open Section
Roads)
Environmentally
Sensitive Rural
Development
82
-------
2.2 Volume II - Technical Appendices
The second volume of the design manual was crafted to support the technical requirements of the first
without duplicating information that is readily available from other resources. This paring of support
information was necessary to prevent the design manual from becoming unusable because of repetitive
information. The decision to include information in this volume was based primarily on availability in
existing documents, or the relevance to information within Volume I. After sifting through the massive
amount of support information related to stormwater design, four appendices were drafted that contain the
minimum information required for the design manual to be self sufficient yet not overly large. These
appendices contain information such as landscaping guidance (App. A) and BMP construction
specifications (App. B.), as well as step-by-step design examples for each structural BMP group (App. C)
and an assortment of tools (App. D) that assist in the design of various stormwater systems. This collection
of information is either unavailable in outside sources or intrinsically valuable to the proper design of
stormwater management.
3. Conclusions
The Environment Article Title 4, Subtitle 2, Annotated Code of Maryland states that "...the management of
stormwater runoff is necessary to reduce stream channel erosion, pollution, siltation and sedimentation, and
local flooding, all of which have adverse impacts on the water and land resources of Maryland." The
program designed in the early 1980's to address this finding of the General Assembly concentrated
primarily on controlling runoff increases associated with new development. Over the last 18 years, tens of
thousands of BMPs have been constructed in order to curb flooding caused by urbanization. Although
implementation has not changed, our stormwater management knowledge and experience has continued to
evolve since Maryland enacted its stormwater statute. With the experience gained comes the identification
of improvements that are needed to fulfill the original intent of this essential water pollution control
program.
Conventional development and construction processes are increasingly identified as destructive to the
environment, encroaching upon natural areas such as wetlands, stream systems, and forests. These activities
also alter local hydrology. Trees and meadow grasses that intercept and absorb rainfall are removed and
natural depressions that temporarily pond water are graded to a uniform slope. Cleared and graded sites are
often compacted, contributing to the rapid conversion of rainfall into runoff. Impervious surfaces impede
groundwater recharge. Pollutants accumulated on these surfaces quickly wash off and are delivered to
receiving waters. While stormwater runoff from developed areas adversely impacts water quality, channel
stability, and disrupts aquatic life, using environmentally sustainable site design techniques may reduce
these impacts.
On October 2, 2000, the Maryland Department of the Environment (MDE) adopted stormwater regulations
including the 2000 Maryland Stormwater Design Manual, Vol. I & II (the Design Manual). Recognizing
the demand for environmentally sustainable or "green" development, these regulations represent a more
comprehensive approach to stormwater design. Included in this approach are better guidance for total site
design and incentives for nonstructural BMPs. The projected outcome of this new program is hoped to be
designs that more closely mimic existing hydrology.
While going a long way in promoting sustainable development, the State's stormwater management
program is not the only set of rules that govern development. There are several State and local programs
83
-------
(e.g., Critical Areas, Forest Conservation, Wetlands Protection) that promote natural resource conservation.
There is also the local zoning regulations that govern land development. Although the goal of these diverse
programs is to protect the environment, there are instances where green development practices are
discouraged and older, less sustainable standards are required.
It is difficult to accommodate the requirements of the full spectrum of resource protection programs.
However, the Design Manual recognizes the importance of each and encourages these principles during
project design. Accordingly, the State's approach to stormwater design may be summarized as a three-step
process: avoidance, minimization, and mitigation. The first step, avoidance, is not just resource protection,
but also includes avoiding development practices such as large-scale clearing and mass grading, structural
fill, and suburban sprawl that have negative impacts on local hydrology. Any reduction in imperviousness
or a site's footprint significantly reduces the amount of stormwater runoff. The second step is minimization.
After all options for avoiding impacts are expended, the designer should incorporate practices that either
replace or disconnect impervious surfaces. For example, using green roof technology, permeable
pavements, or promoting sheet flow will also reduce runoff. After all options to avoid or minimize have
been exhausted, the remaining runoff must be treated using structural practices to mitigate water quality and
channel stability impacts.
Maryland's stormwater management program is one of many State and local programs that regulate land
development. However, the three-step philosophy inherent in the Design Manual incorporates many of
these other programs in its approach. This philosophy refocuses design from the structural management of
runoff as an afterthought to the mimicking of natural processes as part of a total site design.
The Design Manual could never have been produced without the talents, experience, and hard work of the
many people involved in the project. The Maryland Department of the Environment, Water Management
Administration would like to acknowledge those individuals who helped in this process. In particular, Tom
Schueler, Richard Claytor and the staff of the Center for Watershed Protection as well as their project team
partners, Environmental Quality Resources, Inc. and Loiederman Associates, Inc. for their dedication and
efforts. Thanks are also extended to the members of the Stormwater Management Regulations Committee
whose insightful comments and local perspective were helpful in improving the manual. Finally, the staff
of MDE/WMA's Nonpoint Source Program for the patience and support necessary to complete the project
successfully.
4. References
MDE, 1987
Design Procedures for Stormwater Management Extended Detention Structures
Report to Water Resources Administration, 1987
Rawls, W.J., Brakensiek, D.L., and Saxton, K.E., 1982
"Estimation of Soil Properties"
Transactions of the American Society of Agricultural Engineers, Vol. 25, No. 5, pp.1316-1320, 1982
Schueler, T., Claytor, R. et al, 2000.
2000 Maryland Stormwater Design Manual, Volumes I & n
Maryland Department of the Environment, April 2000.
84
-------
ENHANCING STORM WATER INFILTRATION TO REDUCE WATER
TEMPERATURE DOWNSTREAM
Joseph M. Dorava, Vierbicher Associates, Inc.
Aircardo Roa Espinosa, Dane County Land Conservation Department
Ken Johnson, Wisconsin Department of Natural Resources
Daryl Severson, City of Sun Prairie
A substantial storm water management project was recently completed in the city of Sun Prairie, about ten
miles east of Madison, Wisconsin. The primary goal of this project was to protect the water quality of
Token Creek, one of the last remaining cold-water trout streams in south central Wisconsin.
Reducing the downstream movement of sediment and preventing excessive heating of the runoff were two
challenges faced in this project. A team of engineers and scientists from the city of Sun Prairie, the
Wisconsin Department of Natural Resources, Dane County's Land Conservation Department, and
Vierbicher Associates, Inc., joined forces to meet the goals of this project. This team designed and built a
series of stone-filled gabion weirs to filter sediment, and they engineered a stone-lined channel to infiltrate
runoff into the ground.
State funding earmarked for the reduction of non-point source pollution supported this project. The
outcome being a system which treats storm water runoff from more than 492 acres of new residential
development. Enhanced infiltration provided by the stone-lined channel is designed to reduce stream water
temperatures by moving the surface runoff under ground. The gabion weirs are designed to remove
sediment from the streamflow by trapping large particles and filtering smaller ones. The capability of this
storm water treatment system to reduce stream temperature was designed with a site-specific thermal model.
The substantial accumulation of sediment upstream from the gabions indicates the systems ability to treat
storm water runoff. The system's design and functionality, along with its aesthetic appearance in a densely
developed subdivision, demonstrate its success in suburban Sun Prairie. Because infiltration is becoming
more important as a storm water management practice, this treatment strategy may have applications
wherever development occurs.
85
-------
Introduction
The Token Creek Watershed is a 27 square-mile sub basin of the Yahara-Lake Mendota Priority Watershed
in south central Wisconsin, on the northeast side of Madison, immediately adjacent to the city of Sun Prairie
(Figure 1). This watershed supported a native brook trout fishery prior to European settlement (Sorge,
1996). Today, natural springs, which discharge more than 4000 gallons per minute of 50-degree Fahrenheit
water to Token Creek, continue to support a cold-water fishery (University of Wisconsin, 1997, Wisconsin
Department of Natural Resources Unpublished Data). Development around the city of Madison and
especially the outlying areas near Sun Prairie is increasing. The result is increased pressure to build near
wildlife habitat areas and watersheds that support such fisheries. The challenge then is to create
development that is compatible with the surrounding environment and to develop in ways that minimize
degradation of natural resources.
Dane County
Figure 1. Location of Dane County, Wisconsin, and Token Creek in the Yahara River and Lake Mendota Priority
Watershed. The proximity of Token Creek to the growing cities of Madison and Sun Prairie increases the demand for
development in the watershed. The importance of the cold-water fishery in Token Creek and the priority designation
of downstream lakes create a regulatory agency emphasis on protecting water quality. Map modified from Dane
County, http://www.co.dane.wi.us/landcopnservation/pwshed.htm.
Background
Regulatory agencies realize the importance of the natural resources and they understand the value of
limiting sediment inflow and water temperature increases to an urbanizing stream that also supports a cold-
water fishery. As a result, proposed developments in the Token Creek Watershed are closely scrutinized for
their contributions of non-point source pollution. In addition, there is a regulatory emphasis placed on
managing water temperature increases and there are no concise compliance standards, documented best
management practices (BMPs), or design manuals to rely upon or use as targets. Therefore, biologists and
engineers commonly use professional judgement, personal experience, and modeling to predict the outcome
of various management practices.
86
-------
In the case of Token Creek, where there were benefits to protecting the creek for, the participants were quite
cooperative. For example, the developer for the residential subdivision generously donated land along
Token Creek tributary drainageways to the city of Sun Prairie so it could be managed in the public interest.
The developer realized benefits from protecting Token Creek if home site and property values are higher as
a result of the attractive storm water management features in the dedicated public lands and a viable cold-
water fishery downstream. Furthermore, the city of Sun Prairie will benefit from an increased tax base of
the higher home values. Regulatory agencies also benefit because enhanced protection of the natural
resources is one of their primary directives.
Purpose and Scope
Token Creek is part of the Yahara-Mendota designated Priority Watershed Project, which aims to reduce
sediment and nutrient flows into Lake Mendota. This designation and the creek's high value as a cold-water
fishery, prompted the State of Wisconsin's Department of Natural Resources, (WiDNR) to award a Non-
Point Source Pollution Abatement Program cost-share grant to the city of Sun Prairie to design and install
BMPs in the Token Creek Watershed. In support of the Priority Watershed Program, Dane County's Land
Conservation Department is working with the agricultural industry to ensure that agricultural BMPs are
installed throughout the watershed to reduce sediment inflows to the lake. The Land Conservation
Department is also developing a model to predict the effects of land-use change on water temperature and to
predict the change in water temperature derived from various land-management practices. The resources at
Dane County and the WiDNR assisted the city of Sun Prairie and their engineering consultant, Vierbicher
Associates Inc., with the design of BMPs to reduce the movement of sediment and heated runoff to Token
Creek.
The cost-share grant from WiDNR supported design and construction of BMPs in a proposed 492-acre
residential subdivision along a tributary to Token Creek. Dane County provided design recommendations
based on their experience with agricultural practices in the area and results of detailed temperature
modeling. Vierbicher Associates provided engineering design and construction plans. The city of Sun
Prairie supervised design and construction of the project and the WiDNR and Dane County provided
regulatory agency oversight. The two primary goals of the project:
• To protect the water quality of Token Creek (primarily by controlling sediment inflow
and water-temperature increases)
• To provide BMPs that are attractive and improve property values
The purpose of this paper is to describe the Token Creek Water Quality Control Project and the design
process used to select BMPs for this project. Primarily because the project provides an introduction to
relatively new storm water management techniques (rock-filled gabion dams and rock-lined channel storm
water infiltration), and a new engineering tool (water temperature modeling). These new techniques and
tools provide protection against non-point source pollution, in addition to mitigating thermal impacts from
storm water runoff. Both the project's design process, including the new engineering techniques and tools
and the project's unique BMPs, should have broad applications in urban storm water management. The
project also provides valuable examples of cooperation between adjacent city governments, regulatory and
funding agencies, and design professionals.
87
-------
BMP Selection
The proposed 492-acre single-family development was planned to include about 15 acres of green space
along the tributary drainageways to Token Creek and the remaining land converted to approximately 0.25-
acre residential lots (Figure 2). Each lot was planned to contain a 3-bedroom home (2,500 square feet) and a
2-car garage (480 square feet). This lot configuration results in about 4,400 square feet of total impervious
surface if an allowance of 900 square feet is made for roads and 520 square feet is allowed for driveways,
and sidewalks. The result of this development is an alteration of land use from 100 percent open-pastureland
and forest to about 34 percent impervious area.
.25 Miles to Token Creek
? ' I
___
-"
3
^
- I -
-
f £
s
t
N
I^ot to Scale
Green space dedicated
to the public
Figure 2. Proposed single-family residential development in the watershed of a tributary to Token Creek. Of the
approximately 492-acres proposed for development, 15-acres will be dedicated to the public as green space and the
remaining land will be subdivided into approximate 0.25-acre lots.
The result of this type of land-use conversion typically is an increase in runoff and a substantial increase in
peak discharge, severe streambank erosion, and degradation of water quality including elevated water
temperatures. Common BMPs available to address these concerns would include storm water detention
ponds, streambank reinforcement, and created wetlands. Principal concerns with these common BMPs as a
result of a cold-water fishery less than 0.25 miles downstream include storing and ponding water that would
potentially increase the water temperature and unsightly wetland areas that might attract mosquitoes. The
city of Sun Prairie as the supervisor of design and construction and the regulatory agencies within their
review capacity both understood the need to closely coordinate this project. Early in the design process
consultations with regulatory agency staff resulted in considerable efficiencies in the design. For example,
in headwater areas where wetlands prevail along the drainageway, consideration of the need to infiltrate
runoff and preserve wetlands resulted in agreement on selection of an erosion control mat for stream bank
stabilization instead of rock lining. In addition, the agreement between engineers and regulators to place
rock dams near planned or existing roadway and bike path embankments minimized the disturbance to the
88
-------
site by concentrating fill materials and provided for easy maintenance of the storm water management
system. The common acceptance of the need to mitigate water temperature increases in Token Creek
among designers and reviewers brought together a team of engineers and scientists that otherwise would be
working independently. The design process, techniques, and tools this team used to complete this one
project are now complementary items in new county wide storm water management and erosion control
ordinances, Statewide model ordinances, and the daily practice of the individual engineers and scientists
involved in the project. One of the most important new engineering tools is the application of a temperature
model developed during the project.
Temperature Modeling
A Temperature Urban Runoff Model (TURM) was developed and tested in Dane County, Wisconsin, to
predict the thermal impact of proposed development projects (Arrington et al., 2002, Roa-Espinosa, 2003).
A number of sample model runs are presented here to help understand how several variables interact to
result in the stream temperatures predicted by the model. Three of the important variables that determine
stream temperature as a result of a storm are:
• the percentage of impervious area of the parcel,
• the parcel area and
• the baseflow of the stream that the parcel drains into.
Percentage of Impervious Area and Water Temperature
Impervious surfaces, such as pavement or asphalt, increase stream temperature for two reasons. First,
impervious surfaces absorb solar radiation, which raises their surface temperature. When it storms, some of
this heat is transferred to the water that falls on these surfaces as precipitation. Second, impervious surfaces
reduce infiltration, which increases the runoff volume from these surfaces. (Pervious surfaces, like grass or
other vegetation, allow some of the precipitation that falls on them to infiltrate into the soil.) As the
percentage of impervious area of a parcel increases, more of the total runoff from the parcel comes from the
heated runoff contributed by the impervious surfaces. Therefore, as percentage impervious area increases,
the temperature of the water runoff from the parcel increases and the temperature of the stream that the
runoff enters increases as well.
Because there are some significant seasonal variations in storms and their effect on water temperature the
model uses a typical summer rainstorm event in Dane County to predict water temperature changes. The
assumed storm and local environmental conditions accompanying the storm event are described in Table 1.
89
-------
Table 1. Typical storm and environmental conditions assumed for mid-summer storms in Dane County, Wisconsin
from TURM predictions.
rainfall depth
rainfall duration
0.5 [inches
4 [hours
hour of day rain start (between 1 and 24 hours) | 14 |
Time of concentration (Tc)
wind speed
rain temperature (during storm)
Initial temp, of impervious surface
Air temperature
Relative humidity
I 0.100 Ihours
I 10.2 Ift/s
I 73.7 | FQ
I I F
i^^^~^^^~i °
\ 80.0 I F
I 80.0% I
80
50 <
45
40
Trout Lethal Zone
Trout Optimum Temperature Range
t
20 40 60 80
Percentage of Impervious Area
100
• Baseflow = 15cfs, Parcel Area = 50
acres
Trout Lower Optimum Temperature
Limit
'Trout Upper Optimum Temperature
Limit
•Trout Lethal Temperature Limit
Figure 3. There is an increasing trend in stream temperature with increasing percentage impervious area for a given
parcel area and baseflow. Baseflow is given in cubic feet per second (cfs).
Parcel Area and Water Temperature
In general, at a given percentage of imperviousness, the larger the parcel area, the more runoff it contributes
to the stream. More heated runoff means greater stream temperature increases resulting from a storm.
90
-------
Trout Optimum Temperature Range
20 40 60 80
Percentage Impervious Area of Development Parcel
100
— * — Parcel Area = 1 acre
• Parcel Area = 5 acres
Parcel Area = 10 acres
Parcel Area = 50 acres
* Parcel Area = 100 acres
— • — Parcel Area = 500 acres
Lower Optimum Temperature
Limit
Limit
Upper Optimum Temperature
Lethal Temperature Limit
Figure 4. For a given percentage of impervious area and a given baseflow, the greater the parcel area, the greater
the stream temperature.
Baseflow and Water Temperature
Baseflow is the flow rate (volume of water per unit time) of a stream before a storm. Typically small
baseflow is found on small streams and tributaries, whereas large baseflow is found on larger streams.
Stream temperature resulting from a storm is a mixture of the initial stream temperature and the runoff
temperature. At a given volume of heated runoff (determined from the parcel area and the percentage
imperviousness) there is a greater stream temperature increase in a stream with a small baseflow than a
stream with a large baseflow. This is because the runoff volume is a greater proportion of the stream
volume in a small baseflow stream than a large baseflow stream.
91
-------
Trout Optimum Temperature Range
•Baseflow = 5 cfs
"Baseflow = 35 cfs
"Trout Lower Optimum
Temperature Limit
•Trout Upper Optimum
Temperature Limit
•Trout Lethal Temperature Limit
20 40 60 80
Percentage of Impervious Area
100
Figure 5. For a given parcel area and a given percentage of imperviousness, higher stream temperatures are found
in streams with smaller baseflow and lower stream temperatures are found in streams with larger baseflow. Baseflow
is given in cubic feet per second (cfs).
Watershed Characteristics and Water Temperature
Understanding the inter-relation between watershed characteristics and water temperature elucidates
opportunities to manage development or mitigate the effects of development in a watershed (Figure 6). For
this developing tributary watershed to Token Creek, which generally has a larger parcel area (492-acres)
and a lower base flow about (9 cubic feet per second), mitigating increases in stream temperature and
reducing the movement of sediment to the creek were common goals of the developer, the city, and the
regulatory and funding agencies. Because additional single-family housing is in high demand in this area
mitigating the potential harmful effects of development was more desirable than reducing the size or
number of housing units developed.
92
-------
• Baseflow = 5cfs, Parcel Area =
5 acres
• Baseflow = 5 cfs, Parcel Area =
50 acres
Baseflow = 5 cfs, Parcel Area =
500 acres
Baseflow = 40 cfs, Parcel Area
= 5 acres
Baseflow = 40 cfs, Parcel Area
= 50 acres
• Baseflow = 40 cfs, Parcel Area
= 500 acres
•Trout Lower Optimum
Temperature Limit
'Trout Upper Optimum
Temperature Limit
•Trout Lethal Temperature Limit
20 40 60 80
Percentage Impervious Area of Parcel
100
Figure 6. The relative trends of how stream temperature varies with percentage impervious area for different
combinations of parcel areas and baseflow. For small parcels and large baseflow, there is little thermal impact to the
stream, regardless of the percentage of impervious area. On the other hand, large parcels that drain into a stream
with a small baseflow cause a substantial stream temperature increase, even at relatively low percentages of
imperviousness. Baseflow is given in cubic feet per second (cfs).
A 21.6-degree F increase in stream temperature is predicted to result from the proposed development in this
Token Creek Tributary watershed by the TURM (Table 2). The resulting water temperature of 71.6 degrees
F is above the stress zone for trout and, thus, is undesirable. Therefore some temperature mitigating
management practices are necessary.
93
-------
Table 2. For a given rainfall event a temperature increase of 21.6 degree F is predicted to result from the proposed
development in this tributary to Token Creek.
Temperature Urban Runoff Model
POST-OEVELOPMENTT
Units:) en9lish
Required Inputs:
% Connected imperviousness in watershed | 34% |
VAfetershed area I 492.00 lacres
Base flow in stream I 9.0 Icfs
Existing stream temp. | 50.0 | °F
Outputs:
Temp, of runoff from development
93.8 I F
Difference between runoff and stream term 43.8 | F
Temp, of stream after development I 71.6 I F
Increase in stream temp. | 21.6 | F
The model runs described here represent the relative thermal impact of various development scenarios if
heated runoff has little opportunity to cool before entering a stream. The combinations of percentage of
impervious area, parcel area, and baseflow do not necessarily have the impact shown above if temperature
reduction practices are used to mitigate the thermal impacts of development. The two basic principles
behind thermal reduction practices are to slow down heated runoff on its way to the stream (to give it time
to cool) and to increase infiltration of heated runoff (to reduce the volume of heated water that reaches the
stream). Some useful temperature reduction practices include rock cribs, thermal swales, and
retention/infiltration area.
In this development a treatment train was proposed where storm water runoff was collected in the streets
and developed lots and directed to the existing drainageway. In the most headwater areas where the
drainageway was poorly defined, an erosion mat was used to stabilize the channel and rock-check dams
slowed the water and enhanced infiltration (Figure 7).
94
-------
Figure 7. Erosion control matting and a rock check dam combine to reduce stream channel erosion and enhance
storm water infiltration in developed headwater areas.
In areas where runoff is concentrated into a defined channel, a rock lining was used in the channel to protect
the streambank from erosion, to dissipate heat by contact, and to more rapidly infiltrate the runoff below the
surface (Figure 8). Rock-filled gabion dams were installed along the drainageway where flow was restricted
by a roadway or bike path embankment. These rock dam sites were also used for maintenance access as
considerable debris and sediment accumulated upstream from these structures (Figure 8).
Figure 8. A rock lined channel provided rapid infiltration of runoff, substantial heat dissipation, and near complete
control of channel erosion. Rock-filled gabion dams located near channel restrictions provided easily accessible
maintenance sites. Sediment and debris that accumulated upstream from the dams could be readily removed in these
areas. The rock dams filtered large sediment and debris, slowed the flow of water, and dissipated heat.
95
-------
The treatment train strategy implemented in the Token Creek Water Quality Project included a total of
3,055 feet of channel reinforcement and five gabion-dam structures (Figure 9). TURM predicted an increase
in water temperature of only 10.7 degrees F as a result of the planned development following installation of
the storm water BMPs (Table 3).
0.25 Miles to Token Creek
t
N
otto Scale
Gabibrt. 4
-- - V\ \ > - ,--
low
Figure 9. A storm water treatment train that included five gabion dams and 3,055 feet of channel reinforcement was
installed along this tributary to Token Creek to mitigate water temperature increases and reduce stream bank erosion.
Table 3. TURM predicted the water temperature in Token Creek would be 60.7 degrees F following the development
of a 492-acre single-family residential area once BMP's designed to mitigate for water temperature increases were in
place. The increase in water temperature of 10.7 degrees F relates back to the 9-cfs baseflow from the springs that
had a temperature of 50 degrees F.
Temperature Urban Runoff Model
POST-DEVELOPMENT
Temperature Reduction Practices:
F1 stone bed/basin 40000 cubic feet, 6 inch stone
Temperature outletting practices: I 66.5 | F
Temperature of stream after practices
I 60.7 I3 F
Increase in stream temp, after practices | 10.7 | F
96
-------
Conclusions
The Token Creek Water Quality Control Project positively affected the water quality of the creek. The
project also demonstrated the success of close working relationships among designers, regulatory and
funding agencies, and contractors. Everyone, including the developer, supported the project's emphasis on
mitigating thermal impacts and controlling the downstream movement of sediment. The rapid and profitable
sales of homes in the subdivision demonstrate the project's acceptance by the public. The lack of
streambank erosion and the accumulation of debris and sediment upstream from the rock-filled gabion dams
indicate adequate performance of the project's erosion control features. Although not supported by a post-
construction monitoring program at this site, a healthy cold-water fishery downstream in Token Creek
indicates the relatively new TURM may be providing useful guidance to designers. Although specifically
developed for Dane County Wisconsin, this temperature model, the temperature mitigating BMP's, and the
design process used on this project may have applications much wider than the local area. More details of
the TURM are also presented in these proceedings (Roa-Espinosa, 2003). Additional documentation of the
TURM and guidance for its use can be found on the Dane County WWW page at
"http://www.co.dane.wi.us/landco^ Additional examples of similar BMP's
and projects are also available by contacting any of the authors.
References Cited
Arlington, K.E., Roa, A., and Norman, 1, 2002. Understanding Thermal Impact, A GIS Web Tool For
Users (unpublished www document) http://www.co.dane.wi.us/landconservation/thmodelpg.htm
Dane County Land Conservation, 2002. (Unpublished www document)
http://www.co.dane.wi.us/landcopnservation/pwshed.htm
Roa-Espinosa, Aircardo, 2003. Predicting the Impact of Urban Development on Stream Temperature Using
TURM (Temperature Urban Runoff Model): A National Conference, Urban Storm Water: Enhancing
Programs at the Local Level, Chicago, Illinois, 2003, Proceedings, Environmental Protection Agency
Sorge, M. 1996. Lake Mendota Priority Watershed Resource Appraisal Report, Wisconsin Department of
Natural Resources
University of Wisconsin, 1997. Water Resources Atlas Token Creek, A Water Resources Management
Study: Institute for Environmental Studies, University of Wisconsin, Madison, Wisconsin, 140 p.
97
-------
LOCAL SOLUTIONS TO MINIMIZING THE IMPACT OF LAND USE CHANGE
Kyle Dreyfuss-Wells
Chagrin River Watershed Partners, Inc.
Willoughby, Ohio
Abstract
This presentation introduces the Chagrin River Watershed Partners, Inc., (CRWP) discusses why local decision
makers joined the organization, and presents recommendations for minimizing the impact of land use change. The
paper concludes with a discussion of CRWP's implementation of these recommendations through two active
program areas - assisting member communities with zoning regulations for riparian setbacks and compliance with
USEPA's Phase IINPDES Storm Water Program. The paper provides examples of how local communities in the
Chagrin River watershed have implemented CRWP's recommendations and are using their required compliance
with Phase II as an opportunity to address issues of local importance.
Chagrin River Watershed Partners
Formation & Membership
CRWP is a non-profit educational and technical organization formed by watershed communities to address
concerns over flooding, erosion, and water quality problems. Since its formation in 1996, CRWP has grown to
represent 30 townships, counties, cities, and park districts, approximately 80% of the Chagrin River watershed.
Each community selects a trustee to CRWP, either a council member, mayor, or township trustee. These individuals
form our Board of Trustees and direct our member services and watershed studies. With its unique structure,
CRWP works directly with elected officials and their engineers, law directors, and other professional advisors.
Communities joined CRWP due to concerns over rising infrastructure costs and threats to public and private
property created by the loss of natural resource functions and subsequent increases in flooding, erosion, and water
quality problems.
CRWP's structure enables the organization to work directly with communities to update comprehensive plans,
zoning ordinances, and other programs guiding land development, and to introduce innovative practices that prevent
or minimize flooding, erosion, and water quality problems. Building on its relationships with communities, CRWP is
also uniquely positioned to assist members with their NPDES Phase n compliance.
The Watershed
The Chagrin River watershed drains approximately 265 square miles northeast of Cleveland, Ohio. The Chagrin
watershed, like most of Northeast Ohio, was shaped by glacial activity. Many areas of the watershed, particularly
along its steep hillsides and steam banks contain loose sand and gravel that naturally erode at a high rate. Other
areas of the watershed have clay soils that do not easily absorb water, allowing much of the rainfall and snowmelt to
98
-------
runoff quickly. As a result of this glacial past, the Chagrin River watershed has varied topography and naturally high
rates of both flooding and erosion.
The Chagrin, a Lake Erie tributary, is recognized statewide as a high quality resource with State Scenic River
designation from the Ohio Department of Natural Resources (ODNR) on all of its five (5) branches. Several of the
Chagrin's tributary streams support Coldwater Habitat (CWH) aquatic life use designations from the Ohio
Environmental Protection Agency (Ohio EPA). CWH use designation applies to waters that support assemblages of
coldwater organisms. The portions of the Chagrin supporting CWH are primarily small tributaries of the Main Stem
and the Aurora Branch, several of which support breeding populations of Ohio Brook Trout (Salvelinus
fontinalis). CWH is considered among the highest quality aquatic habitat in Ohio and the Chagrin River watershed
is unique for the extent of this high quality habitat so close to a major urban area. (Ohio EPA, 1997)
Other portions of the Chagrin River are designated as Warmwater Habitat (WWH). WWH use designation defines
a typical warmwater assemblage of aquatic organisms. WWH is the principal restoration target for the majority of
water resource management efforts in Ohio and waters with this designation are considered to be in generally good
health. (Ohio EPA, 1997)
Ohio EPA's most recent sampling data on the Chagrin in 1994, 1995, and 1996 places the River on the Agency's
303(d) list of impaired streams (Ohio EPA, 2002). A Total Maximum Daily Load (TMDL) study of the Chagrin is
scheduled for 2006. This sampling data also indicates that many reaches of the Chagrin are not meeting their CWH
or WWH aquatic life use designations. The principal causes of impairment and non-attainment in the Chagrin are
hydromodification, sedimentation, and pollution from urban storm water runoff; nutrient enrichment from failing
home sewage treatment systems and suburban lawn care; sedimentation from streambank erosion and poorly
controlled construction sites; riparian encroachment from land use changes, and the filling and draining of wetlands.
In 2002 CRWP completed a study of wetland loss in the watershed, estimating both historic and current wetland
acreage using available digital data. Our initial estimates place wetland loss at approximately 80%. Adequate
restoration and mitigation for the assimilative capacity of these lost wetlands has not been completed within the
watershed.
Problems Causing Local Decision Makers to Act
Land use and the problems associated with unmanaged development form the common theme among the watershed
problems highlight above. Development increases both the flow and velocity of storm water runoff and, with the
exception of nutrient pollution due to home sewage treatment systems, the water quality problems of the Chagrin
River watershed are due to increases in water quantity. The current land use practices in the Chagrin have caused a
variety of flooding, erosion, and water quality problems. These concerns are seen in Ohio EPA's sampling data as
well as in watershed wide and localized flooding and erosion. These problems cost local governments and residents
as they must clean up from flooding, rebuild threatened or damage roads and bridges, and protect homes and
infrastructure from flooding and eroding streams.
Current land use practices cause flooding, erosion, and water quality problems in two ways, both of which are
linked to increases in water quantity. Traditional land use planning, the guide for a community's long-term
development, does not account for the amount and functions of floodplains, wetlands, and open spaces that
naturally control water quality and quantity. As a result, communities and developers are not aware of these
99
-------
resources and they are lost when land is developed. Traditional land use practices then compound this loss of
natural resource functions by increasing impervious cover. Impervious cover includes roads, rooftops, driveways,
lawns, and other surfaces that do not absorb storm water, and impervious cover increases both the volume and
velocity of storm water runoff. The result of these two impacts of current land use practices is that as the cause of
the flooding, erosion, and water quality problems - impervious cover - grows, the ability of floodplains, wetlands,
and open spaces to control these problems declines.
Our Recommendations for Solving Problems in the Watershed
Faced with a high quality natural resource experiencing the stresses of land use change but not yet in need of
significant remediation, the communities in the Chagrin River watershed have a unique opportunity to implement
innovative, prevention focused solutions to minimize the impacts of development. To assist member communities in
capturing this opportunity, CRWP has developed a series of recommendations on minimizing the impacts of
development. These recommendations are based on the following three (3) principles:
1. Natural resources provide services: Wetlands, riparian areas, and other natural resources provide flood
control, erosion control, and water quality protection services. Table 1 summarizes the services provided by
wetlands and riparian areas.
Table 1: Health & Safety Benefits of Wetlands and Riparian Areas.
Wetlands
Reduce peak flood flows: by storing flood waters and
maintaining stream flow patterns.
Minimize streambank erosion: by reducing runoff
volume and velocity.
Protect ground water quality: by filtering pollutants
from storm water runoff.
Recharge groundwater reserves: by allowing water to
filter into the ground.
Maintain surface water quality: by minimizing
sediment pollution from streambank erosion, and
trapping sediments, chemicals, salts, and other
pollutants from flood waters and storm water runoff.
Riparian Areas
Reduce flood impacts: by absorbing peak flows,
slowing the velocity of flood waters, and regulating base
flow.
Stabilize stream banks: to reduce bank erosion and the
downstream transport of sediments eroded from stream
banks.
Reduce pollutants in watercourses: by filtering,
settling, and transforming pollutants in runoff before
they enter watercourses.
2. Prevention is cheaper than remediation: Preventive steps to maintain the services of natural resources
cost less than remedial actions to recreate these services.
3. Local governments have a role: Actions to maintain these services are matters of public health and safety
and are within local government authorities.
100
-------
Building on these principles, we recommend that each member community have the following:
Comprehensiveplanning: Regular planning that incorporates natural resource management and catalogs
natural resource health and safety benefits.
Riparian and wetland setbacks: Limits on soil-disturbing activities around wetlands and streams. To support
the implementation of this recommendation we have model ordinances for wetland setbacks and riparian setbacks.
Erosion and sediment control: Regulations to minimize erosion on construction sites with strong inspection,
enforcement, and maintenance requirements. To support the implementation of this recommendation, we worked
with the local soil and water conservation districts to develop a model erosion and sediment control ordinance.
Storm water management: Policies and ordinances that require and provide incentives for nonstructural
practices. To support the implementation of this recommendation, we have developed a model storm water
management ordinance that encourages the use of nonstructural storm water management activities.
Options and incentives: Programs to encourage alternative site designs to reduce impervious cover and the
creation of storm water runoff.
Assistance and acquisition: Provide tools to interested landowners on natural resource management and
acquisition of critical areas.
In reviewing these recommendations, it is important to note that the specific tools used by a community to prevent
or solve natural resource management problems vary with a community's level of development. Less developed
communities have a wider range of preventive measures, such as wetland and riparian setbacks, available to them
than communities in more developed areas of the Chagrin River watershed. As the amount of impervious cover
increases in a community, solving problems requires more costly retrofit solutions. In areas where land use is
intense, communities can expect to spend hundreds of thousands of dollars to solve flooding and erosion problems
and to restore the services of natural resources.
Much of our work is focused on assisting members to implement the above recommendations. To date, these
recommendations have been implemented as follows:
Comprehensive planning: The Village of Moreland Hills, Russell Township, and the City of Aurora have
included natural resource inventories in their comprehensive planning efforts.
Riparian and wetland setbacks: The Cities of Aurora and Kirtland, the Villages of Hunting Valley and
Chagrin Falls, and Russell Township have adopted riparian and wetland setback zoning regulations. The Village of
Gates Mills, Bainbridge Township, and Lake County are considering such regulations.
Erosion and sediment control: The City of Kirtland and Lake County have adopted CRWP's model for
improved erosion and sediment control.
101
-------
Storm water management: Russell Township and the Village of South Russell have adopted alternative site
design criteria including limitations on impervious cover and provisions for natural landscaping in common open
spaces.
Acquisition: The Villages of Chagrin Falls, Hunting Valley, Gates Mils, the Cities of Eastiake and Kirtland, and
the Townships of Bainbridge and Russell have active land acquisition programs for permanent open space.
The remainder of this paper details our efforts to promote one of these recommendations, riparian setbacks, and
highlights the linkages between our recommendations and compliance with the Six Minimum Control Measures of
the NPDES Phase H Storm Water Regulation.
Riparian Setbacks
Riparian refers to the streams!de area, or the floodplain, of a watercourse. If appropriately sized, riparian areas can
provide flood control, erosion control, and water quality protection services. These services come from the ability
of riparian areas to slow storm water flow, and slowly release this flow to watercourses. The protection of riparian
areas is important to maintain these services. There are several ways that communities can maintain riparian areas,
including:
Direct landowner assistance: Working with interested landowners on the proper maintenance of their
backyard streams is important to maintaining riparian functions on developed parcels. The Chagrin River watershed
is fortunate to be served by excellent soil and water conservation districts as well as various state agencies available
to assist interested landowners. This approach, however, only reaches interested landowners and does not provide
communities with a mechanism to ensure riparian functions are maintained.
Land acquisition: As mentioned above, many Chagrin River watershed communities have chosen to acquire,
either through conservation easements or direct purchases, critical riparian lands. The Chagrin River watershed
benefits from the highly sophisticated work of the Chagrin River Land Conservancy to assist communities with land
acquisition. While this approach provides direct community control over riparian functions, it is neither realistic nor
desirable for a community to keep all land as open space.
Zoning: Communities may also maintain riparian area functions through land use controls in their zoning codes that
limit development within certain distances of watercourses. CRWP has focused its efforts in this area and developed
a model riparian setback ordinance tailored to the specific concerns of member communities. The details of this
model are presented below.
Model Riparian Setback Ordinance
Riparian protection has historically been a contentious issue in Ohio, raising concerns over impacts on private
property rights. CRWP addressed these concerns in the components of the model ordinance, including:
Whereas clauses: The whereas clauses of an ordinance establish the rational for a community's adoption of a
zoning control. The whereas clauses of the riparian setback model emphasize the public health and safety rational
for riparian protection including the flood control and erosion control services of riparian areas. The whereas
102
-------
clauses also highlight the technical nature of the specific setback widths and their link to the best professional
judgment of natural resource management professionals.
Minimum setback widths: Working with professional staff from Ohio EPA, ODNR, and other agencies, as
well as reviewing national literature on riparian widths, CRWP developed minimum setback widths based on
drainage area. These widths range from 300 feet on either side of a watercourse to 25 feet on either side and are
expanded for the 100-year floodplain as well as riparian wetlands.
Variances: The riparian setback model ordinance contains variance language specific to riparian areas. Most
important in the variance language is the guidance to communities to implement riparian setbacks while ensuring, to
the extent possible, that lots remain buildable and that subdivision lot yields are maintained. This is done by granting
a community's planning commission the flexibility to adjust all setbacks on a parcel - front yard, side yard, rear
yard, and riparian - to enable a landowner to build while staying as far as possible from a watercourse. A
community's ability to require these type of negotiations would be limited without the riparian setback as part of its
zoning code.
Riparian Setbacks in Northeast Ohio
With the development and refinement of the model riparian setback ordinance, CRWP has been successful in
working with member communities to implement the model. As summarized above, five (5) member communities
have riparian protection and two (2) others are considering adoption. CRWP has also assisted communities outside
the watershed as our model ordinance is increasingly seen as the state standard. This assistance resulted in the first
countywide application of riparian setbacks in Summit County, Ohio.
NPDES Phase II Member Assistance Program
The majority of CRWP's member communities are in the Urbanized Area of Cleveland, Ohio and designated under
the Phase n Storm Water Regulations. These communities must develop a Storm Water Management Program by
March 10, 2003. The Phase n rule highlights Six Minimum Control Measures that communities must address in
their Storm Water Management Programs, including public education and outreach on storm water impacts; public
involvement and participation; illicit discharge detection and elimination; construction site storm water runoff control;
post construction storm water management on new development and redevelopment; and pollution prevention for
community operations.
The minimum control measures of Phase n, particularly requirements for post construction storm water control, are
consistent with and closely parallel CRWP's recommendations to member communities for minimizing the impacts
of development. As a result, Phase n represents a unique opportunity for CRWP to provide direct member
technical assistance while promoting our recommendations. In response to this member need, CRWP developed its
Phase II Member Assistance Program. Under this program we are providing services to designated members both
in developing and implementing their Storm Water Management Programs. These services are summarized in Table
2.
103
-------
Table 2: CRWP Phase II Member Assistance Program
Developing a Storm Water Management
Program
Implementing a Storm Water Management
Program
Ohio EPA updates and resolution of member
concerns: CRWP updates members on the latest
developments in Ohio EPA's implementation of Phase
n and works with the Agency to address member
questions and concerns.
Workshops and Training: Since its formation,
CRWP has been a leader in the watershed by
providing educational workshops on the latest
developments in storm water management. CRWP
will continue this focus during the first Phase n permit
term with workshops addressing different aspects of
implementing structural and nonstructural storm water
management practices in Ohio.
Coordination of Phase II service providers: Soil
and water conservation districts, health departments,
and solid waste management authorities currently
provide services, or have the expertise to provide
services, necessary for Phase II designated
communities to implement successful Storm Water
Management Programs. CRWP works with these
service providers to determine what specific services
these organizations will offer and how they will be
delivered to communities.
Model Ordinances: Several of the Phase II
Minimum Control Measures require communities to
implement regulatory mechanisms. CRWP will
provide members with model ordinances compliant
with Ohio EPA's requirements under each of these
measures and will assist in tailoring these to specific
member concerns. As mentioned above, we already
have models for minimum control measures 4 and 5
with models for erosion and sediment control and
riparian and wetland setbacks.
Assistance in drafting Storm Water Management
Programs: CRWP assists communities in drafting
their Storm Water Management Programs in several
ways. We have developed a series of worksheets to
help communities inventory their current programs and
areas where additional activities may be necessary for
Phase II. We have also developed a Storm Water
Management Program outline and a list of
recommended best management practices. Finally, we
developed a draft Storm Water Management Program
based on Ohio EPA's General NPDES Phase II
permit. This draft program provides an easily tailored
format for members.
Educational Services: CRWP will work with other
service providers to offer print ready copy for
newsletters, web sites, and other outlets on various
aspects of watershed and storm water management.
Our staff will also be available to participate in
community meetings on storm water topics.
CRWP has been uniquely positioned to assist members in complying with Phase n. Since its formation, CRWP has
worked to increase understanding about the impact of impervious cover on both storm water quantity and quality.
Our recommendations to member communities emphasize the central theme that it is more cost effective to minimize
the creation of storm water through innovative land use practices, than to attempt to solve storm water problems
104
-------
once they are created. Phase n, while seen by many communities as a burdensome regulation, is being tailored by
our member communities to address their concerns of flooding, erosion, and water quality problems caused by
increases in storm water flow.
References
Ohio Environmental Protection Agency, 2002. Ohio 2002 Integrated Water Quality Monitoring and Assessment
Report Prepared to Fulfill the Requirements of Sections 305(b) and 303(d) of the Clean Water Act. Division of
Surface Water, Columbus, Ohio.
Ohio Environmental Protection Agency, 1997. 1995 Biological and Water Quality Study of the Chagrin River and
Selected Tributaries - Cuyahoga, Geauga, Lake and Portage Counties, Ohio. Ohio EPA Technical Report
MAS/1996-12-6. Division of Surface Water, Monitoring and Assessment Section, Columbus, Ohio.
For More Information on the Chagrin River Watershed Partners, Inc. please contact
Kyle Dreyfuss-Wells
Executive Director
P.O. Box 229
Willoughby, Ohio 44096-0229
(440)975-3870
kdw@crwp.org
105
-------
THORNTON TRANSITIONAL RESERVOIR STORM WATER MANAGEMENT
Didi G. Duma1 and G. Nicholas Textor2
Consoer Townsend Envirodyne Engineers, Inc.
Chicago, Illinois
INTRODUCTION
Consoer Townsend Envirodyne Engineers, Inc (CTE) has completed the design of a multidisciplinary
project for the Metropolitan Water Reclamation District of Greater Chicago (MWRDGC). The project will
divert more than 80% of the 100-yr peak discharge of Thorn Creek (i.e. 6200 cfs) into an existing quarry
located south of I-80/I-294 between Halsted Street and Indiana Avenue, in Thornton, Illinois (Figure 1).
The project will be used in connection with the Tunnel and Reservoir Han (TARP), shown schematically in
Figure 2, one of the most important flood control and water pollution prevention projects in the Chicago
Metropolitan area. The major goals of TARP are:
• Prevent flooding in Chicago Metropolitan area and the backflows into Lake Michigan
• Reduce or eliminate pollution of the various waterways in the area caused by combined sewer overflow
• Comply with the Federal and State environmental laws
• Accomplish results in the most cost effective manner
STAGE II
COMPOSITE
RESERVOIR
167LJI ST.
STAGE I
TTlANSniOMAL
NRCS
nf.SL-IWOin
RELOCATED
VINCENNES AVE.
-KI-KMANI-NI
TUNNEL
22'IXA,
LOCATION MAP
PROJECT ITEMS
VINCENNES AVENUE RELOCATION
THORN CRFFK DJVFHSION
nCWATTRINC TUNNFl
wrr,T L nnr TRANSI
NURl H LOBE fcXPANiilON
Figure 1. Thornton Reservoir Project (in final phase)
1 Didi Duma, Ph.D., Senior Project Manager, Consoer Townsend Envirodyne Engineers, Inc
Nick Textor, P.E., M.S., V-P, Head of Environmental Resources Department, Consoer Townsend Envirodyne Engineers, Inc.
106
-------
Tunnel And Reservoir Ran (TARP)
O'Hare
Upper Dei Plalncs
System
6,6 Ml.
Total
Mainstream
System
40,5 ML
OwPtalnes
Syttem Total
25.6 Ml.
Me Cook Aria calumet Systsm
Reservoir Tom)ice *„«. Log
8.1 Ml.
Ufltitdi
Completed Tunnel
Tunnsl UiKl« CaMtructCon
TuortBl ProposBd
Storags R«nrvolr Prrass U/Gup
(By U.S.A.C.E.)
Calumet System
Lltlle Cal. Leg
7.7 ML
Thornton
Reservoir
Starts* Retarvstr Phm It/Cuip
Uftdnr COKtfUBtton (By
Total Calumet
Syitorn
36.3 Ml.
Storsse RBserMQJt Ptoaw II Cup
Wafcr fladamatlon Ptorrt
(On-Um)
O Puwplnj 8«««OB (UirtuMtaf)
Figure 2. General schematics of Tunnel and Reservoir Plan (TARP) for stormwater management
and water quality improvement in Chicago metropolitan area. Thornton Reservoir is the southern
component of the TARP system
107
-------
The TARP system consists mainly of two principal components:
a. The tunnels, which are associated primarily with water pollution control since they will convey
the water stored in various reservoirs to the Water Reclamation Plants for cleaning and water
quality improvement.
b. The reservoirs, which are associated primarily with the flood control in the Chicago
Metropolitan area since they will store significant stormwater volumes during major flood
events that will be slowly released after the flood peaks will recede.
The Thornton Transitional Reservoir is a first stage of the Thornton Composite Reservoir since it will use
only the West Lobe of the Thornton Quarry. After the mining of North Lobe of the Quarry will be closed,
the project will include and the North Lobe as part of the Thornton Composite Reservoir, the most southern
component of the TARP system.
Thornton Transitional Reservoir will provide flood control in the Little Calumet River Watershed and will
detain only stormwater. The project consists of several important components:
• The diversion structure that will divert over 80% of the 100-year peak discharge of Thorn Creek into a
connection tunnel with variable width.
• The connection tunnel will convey the diverted water to an approximately 300 feet deep drop shaft, with
a 24 foot diameter, that has at the lower end a deaeration chamber (L = 200 ft, W = 32 ft and H = 60 ft).
• The deaeration chamber that is connected to the 22 foot diameter diversion tunnel along I-80/I-294.
• The 22 foot diameter diversion tunnel that will convey the diverted water to the West Lobe of the
Thornton Quarry, which will act as a storage reservoir during big flood events.
• The 8 foot diameter dewatering tunnel that will convey by gravity, the water stored in the quarry to the
Calumet Water Reclamation Plant (CWRP) via the existing Calumet tunnel.
The design of these complex-function structures was accomplished using sophisticated 2-D hydraulic
computation models, and advanced structural design methods. Details of this project and its overall positive
effect on water quality are given in this paper.
DIVERSION STRUCTURE AT THORN CREEK
The existing flow conditions on Thorn Creek are mainly influenced by the water levels at its confluence
with the Little Calumet River. Flow conditions along the channel reach in the area of the proposed
diversion structure are characterized by relatively flat slopes and low flow velocities. In the proposed
conditions, more than 80% of the 100-yr peak discharge of Thorn Creek will be diverted into the diversion
structure. Significant flow regime changes on Thorn Creek would occur during a 100-yr flood event, as
compared to the existing flow conditions, that mainly would consist of:
a. Decrease of water surface elevations of about 6.3 to 6.6 feet at the diversion structure, due to
the reduction of the 100-yr peak discharge from 7400 cfs for existing conditions, to 1500 cfs
under project conditions.
b. Increase of the longitudinal water surface slope along Thorn Creek, upstream of the diversion
structure intake, from an average of 0.027% in existing conditions, to about 4.13% for the
project conditions, with a peak diverted discharge of 6200 cfs.
108
-------
Due to these changes of flow conditions, the flow velocity during the 100-yr flood event, along the Thorn
Creek reach upstream of the diversion structure, would increase from approximately 3.0 fl/s under existing
conditions, to approximately 12 - 14 ft/s under proposed conditions. In these conditions, some channel
erosion could develop, in time, along the upstream reach of the creek, the extent of which would depend on
the sediment characteristics of the channel bed.
The diversion structure at Thorn Creek was designed using a sophisticated 2-D computation model
(CCHE2D) developed at the University of Mississippi [1]. The model was used to determine the optimum
configuration of the diversion intake and the connection tunnel (Figures 3a and 3b) in order to convey the
diverted storm water to the 300 feet deep drop shaft. The CCHE2D computation model is a depth-
integrated two-dimensional hydrodynamic model that can be used for numerical simulation of steady and
unsteady flows in rivers, basins and estuaries. This advanced computation model can accept, a "cold start"
(i.e. zero flow velocity field) as well as a "hot start" (i.e. with flow velocity field obtained from previous
calculations) as initial conditions. It also accepts a "dry bed" condition for starting computations, which is
an advanced feature as compared to other similar computation models.
a.
b.
Figure 3. Initial (a) and final (b) configuration of the intake/diversion structure and the connection
tunnel to the drop shaft.
The diversion intake is designed to convey discharges up to 6700 cfs, which is 500 cfs more than the
required design discharge of 6200 cfs. A sediment barrier wall (weir) of 83 feet in length, with top
elevation above the normal water elevation in the creek, of 585.50 feet relative to National Geodesic
Vertical Datum (NGVD) or 6.00 feet relative to Chicago City Datum (CCD)3, in order to prevent potentially
heavy bedload sediment from Thorn Creek entering into the structure.
Elevation "0" CCD was approximated as 579.50 feet NGVD. The exact value is 579.48 NGVD.
109
-------
The intake bay area, downstream of the entrance weir is at elevation 574.90 feet (NGVD) or - 4.6 feet
(CCD) with a 2% bottom slope toward the gates. Three 12' x 12' sluice gates will control the flow into the
connection tunnel to the drop shaft. The gates will be operated manually from the gatehouse, located on the
top of structure, or remotely from the Calumet Water Reclamation Treatment Plant. The gatehouse floor
elevation is above the existing condition 100-yr flood elevation. Due to the steep rise of the creek bank at
this location, it was possible to locate the diversion structure in such way that most of the structure is
underground; hence the natural esthetics of the area will not be adversely impacted. The intake structure is
equipped with stop log supports to isolate the gates for routine maintenance and repair.
Since the structure is located in a Forest Preserve the stormwater could carry floating debris during floods.
In order to prevent such debris from entering the structure, a curved alignment of 12" diameter pipes, at 3
feet center apart, was provided in front of the structure. This protection screen follows the existing
curvature of the bank, so that the natural configuration of the channel will not be adversely impacted. In
order to prevent intentional or accidental access into the diversion structure, a grate with 6"openings was
provided at the entrance of the intake bay, on the top of the sediment barrier wall (weir). This feature
prevents also pedestrians or animals from falling into the structure. An access road with a wider parking
and maneuvering area ensures the access for service vehicles to clean up the collected debris in front of the
structure and for periodic maintenance.
THE CONNECTION TUNNEL AND THE DROP SHAFT
As previously mentioned, the diverted water from Thorn Creek is conveyed through a connection tunnel
with variable width into a 22 foot diameter drop shaft, approximately 300 feet deep, that has at the lower
end a huge deaeration chamber (200 ft x 32 ft x 60 ft), connected to the 22 foot diversion tunnel that ends in
the West Lobe of Thornton Quarry.
The CCHE2D computation model was used to analyze the flow pattern inside of diversion structure and the
connection tunnel, and to design the optimum configuration of the entire structure. Based on the CCHE2D
numerical modeling, the connection tunnel will be 12 feet in height with a tapered width, of 48 feet at the
control gates to 24 feet at the drop shaft entrance. The longitudinal slope of the tunnel is 7% on a length of
about 110 feet downstream of the gates. The downstream end of the tunnel, at the junction with the drop
shaft, is rounded in order to ensure a proper hydraulic transition. As recommended by the U. S. Army -
Corps of Engineers, the radius of rounding should be 1.5 Ht (where Ht - is the tunnel height). Therefore, for
Ht = 12 feet, a rounded transition with a radius of 20 feet was designed at the downstream edge of the
connection tunnel.
The maximum discharge capacity of the connection tunnel is 6700 cfs for a free flow condition The flow in
the connection channel is supercritical (i.e. Fr > 1.0), with flow velocities ranging from 15 ft/s, just
downstream of the control gates, to 30 - 40 ft/s at the downstream end of the tunnel. The nappe for the
design discharge, at the downstream end of the connection tunnel to the drop shaft will hit the drop shaft
wall at an angle of about 25 to 29 degrees, hence no special construction measures were needed to protect
the wall. The water impact point would be approximately elevation 527.00 feet NGVD (i.e. elevation - 52.5
feet CCD) which is 20 feet below the downstream edge of the connection tunnel.
110
-------
THE DIVERSION TUNNEL
The deaeration chamber at the lower end of the drop shaft (that will prevent the air entertained in the drop
shaft from entering into the tunnel) is connected to a 22 feet diameter diversion tunnel machine bored in
rock, along interstate I-80/1-294, approximately 300 feet below the surface of ground. In the first stage of
the project, the diverted Thorn Creek stormwater will be conveyed to the West Lobe of the quarry. The
diversion tunnel has a double function: diversion of Thorn Creek stormwater into the quarry, and draining
the reservoir to the Calumet Water Reclamation Treatment Plant (CWRP) through the Calumet (TARP)
tunnel. To accomplish the dewatering, an 8 foot diameter drain tunnel, connected to the main diversion
tunnel just east of Vincennes Avenue, will convey gravitationally the water stored from the West Lobe
reservoir to the CWRP. The dewatering tunnel empties to a valve shaft with two 42" hydraulically
operated cone valves to regulate the discharge of water to the CWRP and to prevent back flow of combined
sanitary and stormwater flow from the Calumet TARP System.
THE RESERVOIR AND WATER QUALITY ENHANCEMENT
The Thornton Transitional Reservoir will occupy only the West Lobe of the quarry, as a first stage of the
final project of Thornton Composite Reservoir. The reservoir will provide flood storage of the 3.1 billion
gallons of water from Thorn Creek during floods. After the peak flood stages in Thorn Creek and Calumet
River will recede, the reservoir will be gravitationally dewatered through he Calumet TARP System to the
Calumet Water Reclamation Plant (CWRP). After dewatering, sediment and other debris that were settled
in the reservoir will be disposed to off-site. Therefore, the Thornton Transitional Reservoir project has a
double role: flood protection and water quality improvement for an area of approximately 300 square miles,
which includes parts of the City of Chicago and its southern suburbs.
SEDIMENT TRANSPORT AND BANK PROTECTION
The flow regime and the sediment transport on Thorn Creek during floods exceeding 1500 cfs would be
significantly impacted by the operation of the diversion structure. A sediment transport analysis for the
Thorn Creek reaches adjacent to the proposed diversion structure was performed using the results of the
CCHE2D hydraulic computation model (Figure 4), and the results of the grain size analysis of the sediment
samples collected from the channel.
The total sediment load (gs) on Thorn Creek for the proposed conditions was estimated using the relation
proposed by Grade - Albertson [2], which appears to give the most reasonable results:
s = (1.36V4n3)/{[v(105)]3(D50)3/2H}
where V - is the flow velocity
DSQ - is the median sediment size
H - is the water depth
n - is the roughness coefficient
v - is the settlement velocity for the sediment
111
-------
The numerical simulations performed using the CCHE2D computation model were compared with
analytical calculations. The results showed good agreement. Figure 4 presents the flow velocity
distribution in the Thorn Creek reach influenced by the diversion structure operation
Based upon the results of the analysis, the sediment transport on Thorn Creek could be significantly
influenced during the operation of the diversion structure. However, the sediment transport analysis was
done considering that the design discharge lasts until equilibrium conditions for sediment transport occur.
Since the time duration for the entire 100-yr flood on Thorn Creek is generally only two days, the
equilibrium sediment transport conditions will be reached only for a very short time interval. Therefore, the
sediment transport on Thorn Creek could be less affected than predicted by the sediment transport analysis.
However, a program to monitor channel stability and sediment transport upstream and downstream of the
diversion structure will be implemented after completion of project.
Figure 4. Flow velocity distribution in Thorn Creek (CCHE2D numerical simulation)
In addition, erosion control measures for bank protection upstream and downstream of the proposed
diversion structure were provided.
112
-------
CONCLUSIONS
The project is now under construction (Figure 5), and will be completed at the beginning of 2003. As part
of TARP system, Thornton Transitional Reservoir will contribute to mitigation of the flooding potential, and
will improve the water quality of the natural waterways in the Chicago Metropolitan area.
Figure 5. Diversion structure at Thorn Creek during construction
REFERENCES
1. Yafei Jia and Sam S.Y. Wang: "CCHE2D - A Two-Dimemional Hydrodynamic and Sediment
Transport Model for Unsteady Open Flows Over Loose Bed\ National Center for Computational
Hydroscience and Engineering, Technical Report No. CCHE-TR-97-2, September 1997
2. Garde, R. J., Albertson, M. L.: "The Total Sediment Load of Streams"., Journal of the Hydraulics
Division, ASCE, Paper 1856, November 1958, pp 59-79
113
-------
DEVELOPING SPLIT-FLOW"11 STORMWATER SYSTEMS
Stuart Patton Echols, PhD
Assistant Professor, Division of Resource Management
West Virginia University, Morgantown, WV 26506-6108
Abstract
There are significant problems with urban stormwater management practices using current detention,
infiltration and bioretention methods. The main problem with current detention methods is that they do not
meet current environmental protection goals because they fail to adequately address stormwater volume and
quality. The main problem with current infiltration and bioretention methods is that they do not meet flood
control goals because they fail to adequately address stormwater peak flow rates when rainfall events
occur in which the peak flow rate does not correlate with the specific design storm. What is needed
is a site-based urban stormwater management strategy that will meet both our environmental and flood
control goals. This paper introduces a newly developed stormwater management strategy that provides a
practical, comprehensive and integrated approach to preserving predevelopment stormwater flow rates,
quality, volumes, frequency, and duration This new strategy is based on site-based systems that treat non-
point pollution and split runoff into relative portions based on existing hydrological conditions.
Introduction
In the past, different stormwater management systems have been designed to reduce downstream flooding,
reduce non-point source pollution, recharge groundwater, and prevent stream degradation. The split-flow
strategy is one system designed to do all these things by preserving the predevelopment site hydrology. The
result is a management strategy that separates out and retains or infiltrates precisely the runoff volume
created by development while the natural runoff that existed before development is cleaned and discharged
downstream. As flash flows are maintained at predevelopment levels and first flush is captured on site, the
reduction in downstream degradation should be quite substantial. A complete explanation of the
development, design and application of the split-flowstormwater management strategy can be found in
Split-Flow Method: Introduction of a New Stormwater Strategy., in Stormwater, July/Aug., Echols, S.
(2002) or online at http://www.forester.net/sw 0207 splithtml.
This paper will summarize:
1. What are distributed split-flowsystems?
2. What are the benefits to be gained through their application?
3. When can distributed split-flowsystems be best utilized?
4. What are the hydrological calculations needed to design these systems?
5. How can these systems be used to meet current stormwater regulations?
6. What are the best methods for integrating these systems into site design?
7. How can these systems help guide evolving stormwater policy?
What are distributed split-flow systems?
The basic premise of split-flowstormwater systems is that rainfall can be divided into three portions specific
to any given design storm based on existing conditions for evapotranspiration, infiltration and natural runoff
volumes and that these portions can be filtered, distributed and redirected respectively into bioretention,
recharge and downstream discharge. The traditional objective of stormwater management systems has been
to control the peak flow rate for specific design storms. However, the primary objective of split-flow
systems is preserving the predevelopment hydrological conditions by retaining and or infiltrating the total
114
-------
volume difference created by development and thereby controlling peak flow rates for all design storms.
The first two objectives are to lengthen the time of concentration and control the first flush by emulating the
reduction in runoff adsorbed in the predevelopment initial abstraction. This reduction in runoff is most
easily emulated using existing bioretention techniques sized to capture the first flush. The basic methods of
designing bioretention systems as a water quality practice using plants and soils to remove stormwater
pollutants are outlined in the Prince George's County Government published the Design Manual for Use of
Bioretention in Stormwater Management prepared by Engineering Technologies Associates, Inc., and
Biohabitats, Inc., and subsequent publication explaining Low Impact Development methods including the
Low-Impact Development Manual (2000) developed by Prince Georges County, Maryland Department of
Environmental Resources under the direction of Larry Coffman. In Split-Flow systems, runoff is first
directed to a bioretention facility where the designated first flush volume of contaminated stormwater is
retained by mulch, soil and plant material. Such bioretention facilities can be designed as separate off-line
facilities to assure that the first flush pollutants is not re-suspended and released downstream. Excess runoff
greater than the designated first flush is filtered through the bioretention facility and directed into
proportional splitters where it is divided into diversion and bypass volumes based on specific
predevelopment infiltration and runoff rates. The double weir splits the runoff so that the portion of post
development hydrograph created by buildings and impervious surfaces is diverted into distributed
infiltration facilities and the pre-existing runoff flows are routed downstream. This method most closely
recreates the pre-development hydrograph for the design storm as shown in figure 1.
VOLUME FROM DEVELOPMENT
TIME
Figure 1 - Runoff volume caused by development above pre-development peak flows.
To infiltrate the total difference in volume for all design storms using a double weir and distributed
infiltration facilities, one weir would be designed to emulate the predevelopment runoff while the second
weir would be designed to emulate increase in runoff caused by site development. This concept is easily
conceptualized as a level curb with two Vee-notch weirs sized for the bypass and diversion flow rates as
shown in figure 2.
115
-------
r
FQ HT -D B/El QBJ ENT
PRE-D E'fcl OFW EUT
Openings Cut with Saw
Typical Curb
Figure 2 - Level roadside curb with two Vee-notch cuts of different size corresponding to conceptual
hydrographs for small and large flows.
As water backs up against the curb, it is split into two volumes proportional to the weir openings as it passes
through the curb. The proportional flow splitter apparatus can also be comprised of a drop-inlet or other
water conveyance device with two Vee-notch weirs designed in specific proportions to the predevelopment
rates of stormwater infiltration and runoff. The diversion volume is directed into distributed infiltration
facilities and the bypass volume is cleaned and directed to an existing drainage outlet.
What are the benefits to be gained through the application of distributed split-flow systems?
Stormwater management, as it is often practiced, satisfies the single purpose of storing runoff and releasing
it at flow rates that do not exceed the pre-development peak flow rates. This is generally intended as a local
flood control practice. The process is most often accomplished by detention structures designed to hold the
increase in runoff, and outfall structures designed to release water at specified discharge rates. This
practice, however, fails to address issues such as: (1) downstream flooding from combined detained flows;
(2) groundwater and stream base flow depletion; (3) decreased wildlife habitat; and (4) non-point source
pollution. This current concept of stormwater management by delayed discharge is flawed because the
combined effect of different detention facilities often causes downstream flooding while simultaneously
depleting groundwater and stream base flow. Stormwater management strategies that include some form of
infiltration can satisfy the goals of mitigating effects of impervious surfaces and maintaining pre-
development runoff characteristics. As a result, on-site infiltration currently offers the greatest opportunity
for solving our urban runoff and non-point source pollution problems.
The most logical and practical system of responsible stormwater management is to sustain the natural flow
rate and volume of stormwater runoff by duplicating pre-development runoff hydrographs in post-
development conditions. In theory, pre-development runoff conditions can be duplicated after development
using existing infiltration based Best Management Practices (BMPs) such as porous pavement, dry wells,
infiltration trenches, basins, etc. However, on-site infiltration is not widely accepted in current practice as a
viable stormwater management concept because of short-sighted past infiltration practices. Therefore,
urban runoff problems continue to be addressed by designing stormwater detention systems. Adaptations of
these traditional stormwater management strategies have had limited success in protecting aquatic
116
-------
environments, because they are simple modifications of techniques intended to control peak flow rates and
are not intended to address issues of ecological protection. An alternative stormwater management strategy
is needed that will approach stormwater as an environmental resource and be compatible with land
development practices.
There are multiple stormwater management benefits to be gained through the application of such an
alternative stormwater management strategy including:
1. reducing on-site and downstream flooding
2. reducing flooding caused by combining detained runoff
3. reducing site and regional stormwater systems cost
4. reducing duration of peak storm flows
5. reducing soil erosion, downstream scouring and silting
6. reducing non-point source and thermal pollution
7. replenishing groundwater
8. restoring downstream base flow and wildlife habitats
9. enhancing esthetics and recreational opportunities
10. improving safety by elimination of detention basins
When can distributed split-flowsystems be best utilized?
Preliminary studies still under way show that split-flowsystems can be designed to fit on sites with an
impervious surface coverage of up to 80%. These systems can often be designed to fit within the space used
for existing detention basins. This would, however, not meet the goal of distributing recharge throughout a
site. The more distributed a system is, the more it costs because of increased piping to convey bypass flow
to a discharge point and less efficient use of infiltration facilities compared to clustering them in one
location. This highlights a need for design standards to help assure that split-flow systems will be used to
preserve a site's natural hydrology and not simply used to create more land for building on each site. Sites
using split-flowsystems need to incorporate open space immediately down slope from impervious areas.
These sites should also be designed with open space distributed throughout the development. Ideally,
developments can be designed such that most paved surfaces are built with porous material and the split-
flow systems are only needed to control runoff from buildings. The split-flowstrategy's decentralized
design also creates additional design flexibility, as suitable locations for large stormwater facilities become
a low priority. An additional advantage of the split-flow strategy is that once calculations are complete,
split-flowsystems are simple to design because each impervious area can be designed separately. There is
no need to run routing models commonly used to size detention systems as long as the split-flow facilities
do not overflow into each other. Providing an overflow drainage system to existing discharge outlets
prevents the potential for the facilities to overflow into each other. This ability to design each stormwater
facility separately allows simple revisions if development plans are changed or phased. Even years later as
residents add more impervious areas such as additions, out buildings, or surfaces, split-flowfacilities can be
added to maintain the predevelopment hydrology. Simple regulations need to be written that specify the
size of split-flowfacilities based on square footage of new impervious areas created by landowners. This
would even allow easy retrofits to restore a site's natural hydrology years after a development is completed.
What are the hydrological calculations needed to design these systems?
The bypass weir is sized for pre-development peak flow rate and the diversion weir is sized for the
difference in pre and post development peak flow rate. Using a chart such as the Vee-notch weir
nomograph shown in figure 3, each weir can be sized based on identical head and different flow rates.
117
-------
0.5
1.5
2.5
Head above notch vertex in feet
Figure 3 - Vee-notch weir nomograph showing flow rate, hydraulic head, and corresponding Vee-notch weir
angles.
For example, if the pre-development peak runoff rate is 5.6cfs and the post-development peak runoff rate is
8.5cfs, the bypass weir would be sized for 5.6cfs and the diversion weir would be sized for 2.9cfs. Using
the Vee-notch weir nomograph, the bypass weir angle could be 120 degrees and the diversion weir angle
could be 90 degrees as long as the weirs are constructed at the same elevation.
The total volume difference between pre- and post-development design storms can be calculated with the
equation:
(post Qp x ToC x 80.1) - (pre Qp x ToC x 80.1)
while the total volume for the bypass can be calculated with the equation:
pre Qp x ToC x 80.1.
However, the key to success with a stormwater management system based on this strategy is to install
proportional flow splitters for each impervious surface and distribute the flow from the diversion weir into
individual infiltration facilities. This requires that the flow splitters be designed to divide the runoff from
each of these surfaces into portions that emulate the predevelopment runoff flows and the difference in
predevelopment and post development flow for each individual surface which will not be the same as the
ratios for the entire drainage area. This is done by sizing each individual pair of Vee-notch weir angles for
the proportional flow splitters based on the predevelopment runoff and the increase in runoff caused by each
individual impervious surface. The volume of runoff that needs to be infiltrated for each individual
impervious surface can be calculated with the equation:
Volume = individual impervious surface area x ((post Qp x ToC x 80.1) - (pre Qp x ToC x 80.1)) /
total on-site impervious surface area)
This volume should be based on the largest design storm chosen according to the acceptable level of flood
risk for the site design. This allows the stormwater management system for each impervious area to be
designed independently based on unique site conditions.
How can these systems be used to meet current stormwater regulations?
Traditional stormwater management regulations require peak flow rates be maintained at predevelopment
levels. New regulations also regulate total maximum daily loads for non-point source water pollution. A
118
-------
few regulations address some level of runoff volume reduction but do not require runoff volumes be
maintained at predevelopment levels. Split-flowsystems, however, are based on the premise that we can
recreate predevelopment runoff rates, volume and quality in urban development and that preserving the
existing hydrology is a better way to manage stormwater. This is a change from traditional stormwater
management practices designed to accommodate development by disposing of runoff as quickly as feasible.
Many stormwater regulations currently place runoff in the category of flood hazard planning based on the
view that stormwater is a useless and unwanted byproduct of development that should be collected and
removed as quickly as possible. This is accomplished through systems of inlets, pipes, and basins that
decrease infiltration, stream baseflow, groundwater recharge, and degrade water quality. However,
stormwater can also be viewed as a renewable natural resource that sustains our streams, replenishes our
lakes, and recharges our ground water supplies. This renewable public resource is owned by all of us, a
result of a natural process, used as an economic resource, and has an enormous impact on the quality of
other ecosystems. As a public resource, it's positive and negative economic externalities need to be
acknowledged. If sites are properly designed, this resource can be managed to prevent flooding as well as
safeguard our lakes, streams and groundwater. If site are not properly designed, this resource will flood
downstream properties and destroy aquatic ecosystems. Hence, a basic goal of this alternative stormwater
management strategy is to meet our environmental goals and work within our land development needs by:
(1) not increasing down stream flow rates, (2) reducing non-point source water pollution, (3) recharging at
predevelopment rates, and (4) not polluting our ground water. In theory, if runoff volumes were maintained
throughout the site at predevelopment levels, peak flow rates would also remain at predevelopment levels.
It could, however, be difficult at this time to convince local stormwater regulators that controlling runoff
volume will control peak runoff rates. Further studies using in ground testing will be needed to show how
these systems will perform under actual development conditions.
What are the best methods for integrating these systems into site design?
The crucial element for success with the split-flow stormwater management strategy is to install small flow
splitters for individual paved surface and distribute the runoff into multiple small-distributed infiltration
facilities. This is best done by sizing each proportional flow splitter on the increase in runoff caused by
each impervious surface. For example, a building erected on land with a runoff coefficient of 70 would
require the weir angles designed for 7 cfs and 4 cfs. This would result in one weir having a 90° Vee-notch
angle while the other weir would have a 60° Vee-notch angle. These flow splitters can then be distributed
throughout the site in existing open space or landscape islands as shown in figure 4.
To Downstream
Figure 4 - Example split-flowfacility: depressed landscape island in parking lot with bioretention area, raised
drop-inlet flow-splitter, underground infiltration chamber for diversion flow and bypass to downstream
outlet.
119
-------
This ratio could be used in all the flow splitters used for impervious surfaces on site to control the peak flow
rates for the entire development. Similar ratios can be derived for other runoff coefficients or other runoff
methods. An advantage of the split-flowstrategy is that the volume to be infiltrated is precisely the same as
the excess runoff created by the development and not any larger as in other infiltration and bioretention
methods. This is especially important on sites with clay soils where very little water recharges naturally.
The proportional flow splitter would assure that the same volume and no more would need to be infiltrated
into the ground after development in order to control the peak flow rates. A second advantage of this
strategy is that the volume to be infiltrated is adjusted by the flow splitters for each storm and not based on a
specific design storm. However, without adequate distribution on site the system will not work because
there must be sufficient soil area for the diversion volume to able to infiltrate in a reasonable time.
Therefore, many small split-flow facilities need to be placed throughout a site as shown in figure 5.
Figure 5 - Example plan with location of Split-Flow facilities. Impervious surfaces are outlined in blue. The
underground infiltration chambers are shown as small blue rectangles while above ground bioretention
facilities are shown in green. Thin blue lines show which impervious areas and buildings are directed to
which split-flow facilities.
This concept will succeed in controlling peak flow rates where other infiltration and bioretention strategies
have not because the amount of stormwater to be infiltrated in each facility is carefully controlled and it is
never concentrated in large quantities. The stormwater management system will still control the peak flow
rates by distributing and infiltrating the difference in volume over the entire site.
How can these systems help guide evolving stormwater policy?
Many communities have implemented stormwater utilities to pay for building storm sewers and runoff
treatment facilities. Some communities base their fees on impervious surface areas for each property.
120
-------
Many of these communities also allow reasonable reduction in fees based on reduction in volume, which
will hopefully encourage more environmentally responsible stormwater management practices. If a builder
installs a system to control the runoff rate and volume and can demonstrate there is no change in the
existing hydrology, the fee could be waived. This can provide an incentive for developers to install
environmentally responsible stormwater management systems if the costs are reasonable. A preliminary
study shows that split-flowsystems would likely cost the same or less to build than detention systems.
Split-flowsystems would provide non-point source pollution and flood control benefits to the community,
as well as lower the owner's annual operation cost by eliminating the annual stormwater utility fees. As a
result, the split-flowstrategy can provide a reasonable financial alternative to existing detention practices,
which could become a financial incentive for developers to install more environmentally responsible
stormwater management systems. Maintenance costs should be the same as existing bioretention systems,
however, further research is needed.
The split-flowstrategy intends to preserve the predevelopment site hydrology by duplicating year-round
natural infiltration volumes. Water balance studies indicate that spring flooding results when the ground is
saturated from winter precipitation stored in the soil and the soil's water absorption capacity is greatly
reduced causing increased runoff. The split-flowstrategy would emulate these conditions and therefore
likely infiltrate less precipitation during the spring flooding season. Detention systems, on the other hand,
are not designed for, or affected by, soil infiltration capacity, which changes during the year. In effect,
split-flowsystems could reintroduce local stream flooding that may have been prevented with detention. As
a result, a question arises regarding the conflict between the wisdom of restoring natural processes, which
could include local spring flooding, versus installing detention systems that could artificially control local
spring flooding but destroy aquatic ecosystems. Conversely, development has also been shown to cause
increased year-round flooding and multiple detention systems can combine and elevate these floods
depending on how the basins' outflows combine downstream. As stated, the split-flowstrategy is based on
the premise that preserving the natural hydrology is a better way to manage stormwater. However, the land
development industry has historically operated under the strategy that we should modify natural systems to
accommodate development rather than modify development practices to accommodate natural systems.
Changing these basic beliefs and operation procedures will likely require numerous long-term
demonstration studies.
Conclusion
The goal of this paper is not to claim excellence of one stormwater management method over another but
rather to contribute an additional management option that hopefully can start to change our stormwater
management expectations. The intent is to demonstrate that a viable stormwater management strategy can
be derived from the premise that preserving the natural hydrology is a better way to manage stormwater and
that modifying land development practices to accommodate natural systems can be more effective than
modifying natural systems to accommodate land development practices.
The split-flowstrategy, however, is still a theory that needs in-ground testing to discover what problems will
result in the design and construction processes. For example, including construction erosion and sediment
control measures on sites with split-flowsystems will create addition design challenges. Current design and
construction practices incorporate temporary sediment basins in the location of future detention facilities.
These temporary sediment basins are then converted to detention basins when construction is completed.
However, split-flowsystems do not need detention basins. Therefore, other erosion and sediment control
solutions will be needed during construction. Possible solutions include: use alternative prevention and
control methods that do not require sediment basins, build temporary sediment basins that can be converted
121
-------
into bioretention facilities when construction is completed, or build temporary sediment basins elsewhere on
site that can be removed after construction is completed. Regardless of what methods are used for erosion
and sediment control, the split-flow systems should not be activated until the site is completely stabilized.
Additional research will be needed as other site design and construction implications arise.
Preliminary research shows that split-flow systems can be comparable in construction cost to detention
systems depending on the complexity of the stormwater designs. Findings show that split-flow infiltration
practices can often be used to lower the cost of on-site stormwater management and provide a higher level
of environmental protection. Findings also indicate that non-point source water pollution reduction
objectives currently achieved by other infiltration and bioretention strategies could be more cost effective
construction using the split-flowstrategy. Notable implications that need to be addressed with further
development of the split-flowstrategy include: stormwater policy, site design and construction practices,
runoff modeling and environmental concerns.
Coffman, L. (2000). Low-impact development manual. Prince Georges County, Maryland Department of
Environmental Resources.
Echols, S.P. 2002. Split-flow method: Introduction of a new stormwater strategy. Stormwater -The Journal
for Surface Water Quality Professionals, 3(5): 16-32.
122
-------
Deforest tfce (Bfuegrass
Empowerment of the Citizen Watershed Manager
H. David Gabbard, P.E.
Division of Engineering
Lexington-Fayette Urban County Government
Lexington, Kentucky 40507-1310
Angela Poe
Bluegrass PRIDE
Lexington, Kentucky 40591
Abstract
In 1992, the National Pollutant Discharge Elimination System permitting program of the Clean Water Act
sought to address non-point source pollution from stormwater discharges. Lexington, Kentucky, was a
Phase I city that was required to file for a permit under this program. The permit required the Lexington-
Fayette Urban County Government (LFUCG) to assess the environmental damage to its water resources and
develop urban stormwater pollution prevention programs using best management practices (BMPs) to the
maximum extent practicable (MEP).
The assessments showed that aquatic life had been greatly affected by the alteration of stream corridors.
From the filling of floodplains and the alteration of stream morphology to the clearing of streambanks of
unwanted vegetation, human activities had greatly diminished optimal habitat conditions. It was determined
that one of the most effective BMPs to reverse the affects of these activities was to restore riparian forest
cover to the stream channels. However, two centuries of agricultural uses of the land has left an aesthetic,
"The Bluegrass Aesthetic," in which citizens expect creeks to be seen and heard. Rolling hills are covered
with carefully mowed non-native bluegrass and fescue; streambanks are mowed down to the water's edge;
and trees dot the landscape in various places - but do line fencerows and driveways.
Because of the "Bluegrass Aesthetic," citizens regard urban streams as mostly open, stormwater ditches and
that it is the government's responsibility to keep them clean. Most property owners have applied the
"Bluegrass Aesthetic" to every lawn - mowing or weed-eating down to the water's edge. Furthermore, the
little remaining Fayette County riparian forests contain a dense understory of invasive bush honeysuckle.
Because of community concerns regarding the concealing of illicit activity, many forested stands with bush
honeysuckle have been removed.
The final constraint was that the LFUCG Division of Engineering did not have a stormwater budget that
would allow for large public works projects to address major riparian planting programs. However, even if
the DOE had the budget for such projects, it would have to overcome negative public perceptions regarding
early successional growth.
123
-------
The solution was to create the (Reforest the (Bluegrassprogram in the spring of 1999. This is a Public Works
program that empowers citizens to protect their own water resources. By using citizens to plant the forests,
there is a sense of ownership of the project and that support is critical in the early stages of forest growth -
when the project looks "weedy" and contrary to the "Bluegrass Aesthetic." Furthermore, it educates and
trains citizens why to plant trees to protect their properties along streams or "ditches" (there are 560 miles of
blueline streams in Fayette County).
The success of the (Reforest the (Bluegrass program has been phenomenal! Since April 1999, 3,975
volunteers have been trained as urban watershed managers in eight different events. They have planted over
108,000 trees in 140 floodplain acres. The LFUCG has spent approximately $85,000 of local taxpayer
dollars and other $50,000 has been raised via donations or grants. If the project had been contracted out (as
some first suggested), the project would have cost over $650,000!
Introduction
The creation for the (Reforest the (Bluegrass program is founded in the need for the LFUCG to comply with
various components of the Clean Water Act. The LFUCG has been monitoring the conditions of the waters
of Fayette County since it was first required to apply for a stormwater discharge permit in 1992. This
permit serves the purposes of qualifying and quantifying urban sources of non-point source pollution
conveyed by stormwater runoff. Other non-point sources in Fayette County are comprised of agricultural
sources from tobacco farming, cattle grazing, and the equine industry.
(Reforest the (Bluegrass addresses three important goals facing large urban communities:
• An NPDES municipal stormwater discharge permit to control urban, non-point source pollution;
• Restoration of streams listed on the 303(d) lists of each state; and
• Changing the landscaping habits of citizens to protect water resources and value riparian forests.
Goal: Urban Non-point Source Pollution Control - Municipal Stormwater Permit
A stormwater discharge permit is required as part of the Water Quality Act of 1987. Medium sized cities
with populations greater than 100,000 and less than 250,000 which had municipal separate storm sewer
systems (MS4s) were required to apply for permits as a phased approach to the management of water
quality within the United States. Earlier legislation and programs (1972 Clean Water Act and the National
Pollutant Discharge Elimination System (NPDES)) focused on removing point sources of water pollution.
The 1987 stormwater permitting requirements were designed to manage non-point source water pollution
from various industrial and municipal activities.
The NPDES Stormwater Discharge Permit for the LFUCG required an assessment of the environmental
damage to Fayette County water resources and develop urban stormwater pollution prevention programs
using best management practices (BMPs) to the maximum extent practicable (MEP). During the assessment
of the urban watershed, the following problems were identified:
124
-------
1. Floodplains have been filled and developed utilizing past engineering designs that forced more flow
through narrower channels thus altering and reducing the benthic macroinvertebrate habitat;
2. Tree canopy over the streams has been either eliminated, consisted of invasive bush honeysuckle; or was
comprised of ornamental shrubs and trees in single rows;
3. There have been problems associated with dense communities of algae dominating the streams.
Because of the high phosphorus content of the soil, the concentrations of phosphorus in the storm water
runoff quickly trigger algal growth (background phosphorus concentrations are 0.2-0.3 mg/L). Where
there is full sunlight, in most places, algal mats form quickly and in abundance. However, anywhere
there is tree canopy, the stream is void of algae;
4. Lexington is situated on a hill. Six 11-digit HUCs (watersheds) drain from the central part of the city
out to the county line. Because all urban streams are small headwater streams, the impacts of thermal
pollution, heavy metals, and dissolved oxygen-robbing algal mats have resulted in frequent fish kills and
poor aquatic insect communities; and
5. The destabilized streambanks and shallow soil depths (to bedrock) have resulted in streams eroding and
widening their bank widths.
In creating a watershed management program, the LFUCG would have to:
• Apply Best Management Practices (BMPs) to the Maximum Extent Practicable (MEP);
• Seek intra and inter-governmental cooperation;
• Involve public education and involvement; and
• Seek ways to reduce the use of lawn care chemicals and their impacts.
As a stormwater management program, (Reforest the
-------
• Presence of fecal coliform in streams and storm sewer outfalls;
• Fair to poor aquatic communities; and
• High nutrient and organic enrichment.
Dry and wet weather water chemistry samples indicate high levels of fecal coliforms in most streams.
Biological community monitoring indicates that streams in the urban service area generally do not fully
support aquatic life. Habitat evaluations indicate inadequate instream and riparian habitat to support aquatic
life at some sites; at other sites, habitat is adequate but aquatic life is still poor.
As previously mentioned, nutrient enrichment is a problem because of the high phosphorus concentrations
that occur naturally in the central Kentucky region. Only 7% of the streams of the United States are
limestone-based systems. And of those, central Kentucky is an oddity because the upper limestone layer has
a high phosphorus content. Groundwater in the area has a phosphorus concentration of 0.2-0.3 mg/L, two
to three times higher than the 0.1 mg/L concentration that triggers algal blooms elsewhere in the country.
303(d) List of Waters for TMDL Development
For the initial selection of reforestation sites, the 1998 303(d) listed streams were examined for Fayette
County:
First Priority (Does not support one or more designated uses, KDEP 1998):
Unnamed Tributary to Baughman's Fork
Impaired Use
Aquatic Life
Pollutants of Concern
Organic Enrichment/Low DO
Nutrients
Cane Run
Town Branch
Wolf Run
Aquatic Life
Swimmable
Aquatic Life
Swimmable
Swimmable
Organic Enrichment/Low DO
Pathogens
Organic Enrichment/Low DO
Nutrients
Pathogens
Pathogens
Second Priority (Partially supports designated use)
West Hickman
Impaired Use
Aquatic Life
Pollutants of Concern
Habitat Alteration
Siltation
126
-------
Goal: Alteration of Human Habitat Habits
Over 200 hundred years ago, the central Kentucky plateau region was a savannah covered in mostly buffalo
clover and cane breaks. However, along the stream corridors were dense hardwood forests -oak-hickory
forests. With the settlement of the area, the cane breaks and dense riparian forests were cleared for livestock
grazing and cropland. Furthermore, it was discovered that the rich soils from the weathering of limestone
layers prevalent in this region resulted in exceptional land upon which to graze and raise thoroughbred
racing horses. With these types of agricultural uses for the land, trees were relegated to fencerows and
driveways. Also, forests were left in hard-to-reach or unfarmable areas. After over a hundred years of this
change in land cover, the "Bluegrass Aesthetic" was born - rolling hills, mowed fields of non- native
Bluegrass, and a few trees dotting the landscape.
Almost all modern property owners have applied the "Bluegrass Aesthetic" to their lawns - mowing or
weed-eating down to the water's edge with a few trees here and there. Citizens have viewed urban streams
as open ditches and that it is the government's responsibility to keep them clean. Furthermore, the limited
existing Fayette County riparian forests contain a dense understory of invasive, non- native bush
honeysuckle. These areas have been used for concealing illicit activity and the Parks and Recreation
Department and neighborhood associations have previously thinned out these areas to make them safer and
more aesthetically pleasing. Therefore, any education in regards to the use of riparian buffers must address
the impacts of the modern, chemically- addicted lawn.
It should also be noted that as part of any NPDES stormwater discharge permit, the permittee is required by
the Clean Water Act to create educational programming to alter the lawn care practices of the urban area to
reduce the use of fertilizers and pesticides. Therefore, riparian buffer education and implementation by
citizens is a positive way to affect meaningful change without a lot of effort put into informing citizens what
they are doing wrong.
^Reforest tfie ^Bfuegrass - The Early Years
199 9
With the consideration of all the aforementioned goals, discussions began with the LFUCG Division of
Parks and Recreation as to a suitable area to begin work. The reason the Parks department was approached
was that it was the only division of the LFUCG that owned and maintained long stretches of stream
corridors. It was determined that the first year's event would be performed along a "ribbon park" which had
been donated to the LFUCG as part of a commercial development. The University of Kentucky was
converting agricultural land, Cold Stream Research Farm, into a commercial "research park" along three
miles of Cane Run Creek in northeastern Lexington. Because the floodplains were undevelopable, the
University gave the floodplain areas to the LFUCG as greenspace with the condition that it is for passive
recreation: trails, meadow areas, riparian forests, etc.
A local landscape architecture firm, John Carman and Associates, was hired by Parks and Recreation to
create the design. Even though the design showed a riparian buffer strip along the three miles of stream
corridors, no one had ever planted that many trees before and Parks did not have a budget to purchase the
trees. Up until that time, all trees which were planted on Park property were saplings or greater in size.
Therefore it was considered impracticable to plant forests of large caliper trees - but it was nice to look at
127
-------
on paper. It was decided that the Cold Steam Park would be a good proving ground for the project. After
all, this land had not been open to the public previously so if our project failed, no one would really notice.
Even though early planning was chaotic and there was still an on-going discussion as to whether or not to
involve the public (the Urban Forester wanted to hire migrant workers to plant the seedlings - the
Environmental Engineer wanted the "public outreach" component for his stormwater management
program), the project became an overnight success! During two weeks in April 1999, over 1,200 volunteers
assisted in the installation of 45 acres of floodplain forests. 25,000 tree seedlings were planted along three
miles of First Priority streams in Lexington's effort to systematic restore riparian forests along all 560 miles
of streams within its borders.
200O e£ ^T(B 2O01
During the following two years, another 45,000 trees were planted by training over 2,000 citizens to plant
riparian forests. The site for these projects were in Masterson Station Park which has two tributaries to the
Town Branch, another First Priority watershed in Fayette County. The park is the largest in Fayette County,
770 acres of rolling hills and denuded streambanks.
Deforest THe ®Cuegrass 2O02- "I think we got it right this time..."
OffiB 2002, April 6, was by far the best event yet - crystal blue skies (high of 49° F); well-trained group
leaders; over 900 volunteers (planters and staff) showed up to plant 15,975 trees; and there was plenty of t-
shirts, food, and supplies. The event also took place down inside two, large regional detention basins that
were installed as a part of a commercial development. The detention basins and the land surrounding them
were deeded over to the LFUCG as park area. Therefore, the connection between the creation of an urban
forest and the control of stormwater pollution was clear for the first time.
Project Design
(Reforest the (B/ue^rassuses the wealth of knowledge and experience gained by the use of riparian,
streamside, buffer systems. This "system" is nothing more than examining and mimicking the beneficial
controls applied by nature to protect and preserve stream corridors. The buffer system approach uses the
beneficial qualities of native vegetation to achieve desired goals of resource management.
In Figure 1, the buffer system consists of using three different kinds of vegetation to achieve the desired
results. For bank stability and aquatic habitat enhancement, tree or shrub species that can tolerate a moist
environment are selected. These are planted along the stream and in the floodplain. For nutrient control,
optimal wildlife habitat, and slope stability, tree and shrub species are selected that prefer average to dry soil
conditions. Finally, to control nutrients even more, a zone of wildflowers or native grasses (or both) are
planted along the outer perimeter of the forested zone.
With this information, species are selected that will enhance the biota of the localized planting. Also,
species selected are strongly influenced by their availability through the National Tree Trust. As the largest
sponsor of the (Reforest tfte
-------
disregarded. However, the National Tree Trust has been the largest supplier of trees. Additional tree
species that are not on the list of the Tree Trust are ordered through the Kentucky Department of Forestry.
Water Quality Benefits of Green ways and Creekside Buffers
Planting (Vegetation) Zones:
Zone 1-
Zone 2-
Zone 3-
Flood plain species
Upland species
Native grasses
or wildflowers
Stream
Benefits of Creekside (Riparian) Planting Zones:
Bank Stability
Aquatic Habitat
Sediment Control
Nutrient Control to Limit Algae
Wildlife Habitat
Minimum (black)
Maximum (white)
•i i i | i i i i | i i i i | i i i
0 25 50 75 100
Recommended riparian zone widths (feet)
Figure 1. Riparian Buffer Management System, RiMS (Source: Schultz, NREM Dept. Iowa State University)
Table 1 contains a current list of tree species used (although some were not available for this past year's
event). The Recommended Planting Zones refers to the previous discussion on buffer systems and the
appropriate zones for different species. "W" means "wet" and these trees are suitable to plant in areas
where the ground may be inundated for extended periods during the year. "1" is for trees that are suitable to
be planted in "Zone 1," the floodplain zone. These trees will experience somewhat frequent flooding and
the soils are generally moist to wet. "2" is for trees that are suitable for "Zone 2." Zone 2 trees do not
tolerate root systems that are inundated with water. They prefer average to dry soil moisture conditions.
"3" refers to trees suitable for "Zone 3." Zone 3 are areas that can become dry; tops of hills, south facing
slopes, next to parking areas or commercial zones, etc.
Project Implementation
Project Coordination
Reforest the Bluegrass is a cooperative effort of the LFUCG Divisions of Engineering (Stormwater), Parks
and Recreation, and Planning (Urban Forestry). The project also uses engineering, forestry, and ecological
experts from academia and natural resources agencies to design and layout the project. Tree seedlings of
various species, native to the inner Bluegrass physiographic region, are donated by the National Tree Trust.
Seedlings are mixed together in bags that are sorted by planting design areas. Dots are spray painted on the
restoration site at a recommended spacing and bags of trees are col or-coordinated with the dots on the
129
-------
ground (green dots for floodplain species, pink dots for upland species, etc). During the planning stages of
the event, many community organizations and businesses assist with project organization and
implementation. These organizations include Bluegrass PRIDE, Fayette County Conservation District and
Extension Office, First Link of the Bluegrass, Inc., Kentucky-American Water Company, Kentucky
Division of Forestry, Kentucky Utilities, Kentucky Waterways Alliance, League of Women Voters,
Lexmark International, and the University of Kentucky Department of Forestry. (Each year, Kentucky-
American Water Company has donated $5,000 to the project. Sponsors who contribute $5,000 or more per
year are called, "Friends of the Forest."}
Table 1. Current tree species selected for the Reforest tfie (BCuegrass project.
Common Name
Scientific Name
Recommended
Planting
Zones
No. of trees
ordered
Allegheny Serviceberry
Bald Cypress
Blackgum
Black Locust
Black Walnut
Bur Oak
Buttonbush
Eastern Redbud
Green Ash
Hackberry
Kentucky Coffee Tree
Northern Red Oak
Paw Paw
Pecan
Persimmon
Red Maple
Sassafras
Shagbark Hickory
Shellbark Hickory
Shingle Oak
Shumard Oak
Silky Dogwood
Spicebush
Superior Cottonwood
Sugar Maple
Sweet Gum
Sycamore
Tulip Poplar
White Ash
White Oak
Wild Plum
Amelanchier laevis
Taxodium distichum
Nyssa sylvatica
Robinia pseudoacacia
Juglans nigra
Quercus macrocarpa
Cephalanthus occidentalis
Cercis canadensis
Fraxinus pennsylvanica
Celtis occidentalis
Gymnocladus dioica
Quercus velutina
Asimina triloba
Carya illinoensis
Diospyros virginiana
Acer rubrum
Sassafras albidum
Carya ovata
Carya lacinosa
Quercus imbricaria
Quercus shumardii
Cornus amomum
Lindera denzoin
Populus deltiodes
Acer saccharum
Liquidambar styraciflua
Plantanus occidentalis
Liriodendron tulipifera
Fraxinus americanus
Quercus alba
Prunus americana
1-2
W
W~l
2-3
1
2-3
W-l
2-3
W-l
2-3
2-3
2
2-3
1-2
2-3
1-2
2-3
1
1
2-3
2-3
2
1-2
1
1-2
W-l
W-l
1-2
2-3
1-2
2
900
1,000
1,000
500
1,000
1,500
1,500
1,000
1,000
1,300
1,000
800
1,000
1,000
2,000
2,000
1,000
800
1,000
1,500
1,500
1,000
1,000
1,000
130
-------
Volunteer Coordination
On the day of the planting event, volunteers are escorted into the field by Group Leaders who teach each
citizen about the value of riparian forests in urban water pollution removal, the reduction of greenhouse-
gases and the urban heat-island effect, and the enhancement of wildlife diversity. The volunteers are taught
to use dibble bars to plant seedlings and then protect them from competitive vegetation using the tree mats.
Once the group is finished planting the trees, about 20 per person, the volunteers are treated to a free t-shirt,
pizza lunch, musical entertainment, the building of bird houses, and educational displays by various
community organizations. Once the planting has occurred at each site, the areas are deemed as "no mow"
zones, surveyed for specie survival rates, and monitored and controlled for animal browsing and impacts by
invasive species.
Volunteer Education
deforest the (Bfuegrass cannot be considered successful, no matter how many trees are put into the ground,
unless there is a successful educational component. (Reforest the (Bfuegrass'vt, the perfect situation in which
to foster an understanding of environmental issues that will lead to long-term positive environmental
behavior. Through (Reforest the (BluegrassfasxQ is an opportunity to expand the action and awareness
components inherent in a reforestation project to a deeper understanding of watershed management on both
a personal and community level. At the event, volunteers are treated not only to entertainment and food, but
they have many opportunities to learn more about why they are participating in the event and what a
difference their time and efforts are going to make for Lexington's future.
Themes
Communities have different environmental perspectives that should be taken into account when identifying
educational themes for an event. For (Reforest the
-------
It is important to utilize local TV and radio stations as well as local papers. Traditionally, one month prior
to the event, a press conference is held. The Mayor, Vice-May or, major sponsors, lead organizing agencies,
and other groups of importance are invited to participate. The one year that a press conference was not held,
the week before (Reforest tfie (Bluegrass, an insert is put in the local paper. This insert contains the event
location and time, a rain date, registration information, and suggestions on what to wear and bring. It also
contains basic information on watersheds, stormwater management, and riparian areas, all applied to local
waterways. It would be advisable to both hold the press conference and print the insert. Various TV and
radio interviews are given in the weeks leading up to the event, in which information similar to that in the
insert is shared.
At the Event
On the day of the planting event, volunteers are escorted into the field by Group Leaders who teach each
participant basic information on the value of riparian forests in urban water pollution removal; the reduction
of greenhouse gases; the urban heat-island effect; and the enhancement of wildlife diversity. The volunteers
are taught to use dibble bars to plant seedlings and then protect them from competitive vegetation using the
tree mats. The newly reforested area is not a pretty site. In fact, it looks like a field full of weeds and litter.
Therefore, it is important to help volunteers appreciate the need for the forest successional process in order
to create a population that is willing to tolerate, even defend, this young forest.
Once the group is finished planting trees, they are directed to a common area where there is food,
entertainment, and educational booths. Various organizations from throughout the central Kentucky are
asked to participate by bringing displays that will allow people to learn more about protecting and restoring
our environment, particularly waterways. Groups that regularly participate in the OffiBoutreach area
include: Bluegrass PRIDE, the Fayette County Conservation District, Wild Ones, Tree Guide, and
environmental groups from the University of Kentucky. Each year the list expands. There is traditionally
an erosion demonstration, a display board on riparian forests, an exhibit that labels and discusses the
properties of the Reforest the Bluegrass tree species, and information on wildlife habitat. For RTB 2001
and 2002, there was a booth that offered children the opportunity to build birdhouses. During RTB 2002,
one of the booths passed out grocery sacks so volunteers could pick up the litter that was prevalent on the
site. Over 200 bags worth of litter was collected. This cleanup offered young children another activity in
which they could participate.
In the outreach area, it is important to inform the adults, but it is also important to have booths targeting age
groups that are too young to plant seedlings. Many families participate in 3&B, so it is imperative to
involve the entire family. If it is a successful family outing, it is likely that families will become annual
participants.
A highlight of the 2002 event was the ceremonial planting of a Princeton elm, celebrating the planting of
100,000 OffiBtrees. The Mayor, Vice-Mayor, major sponsors, and other important local figures were
invited to participate in this planting, which was covered by the local media. The tree is labeled with a
plaque that explains its purpose and lists the "Friends of the Forest."
Post-Event
Now that an engaged population of volunteers has been empowered, it is important to encourage them to
remember the lessons of the day and to present them with opportunities in which they can continue to be
good stewards of their local environment. As people leave (Reforest the (Bluegrass, they are given a tulip
132
-------
poplar, the state tree, to take home and plant. The participants are given a dichotomous tree key that aids in
the identification of all species planted that day to encourage them to revisit the site. The volunteers also
leave with a pamphlet that contains basic watershed and nonpoint source pollution information and details
ways that they can continue their involvement in improving local waterways through adopting a stream,
testing water quality, planting more trees, or labeling storm drains. Making a reforestation program an
annual event is also a wonderful follow up. Many of the (Reforest tfie
-------
STORM WATER PHASE I MS4 PERMITTING: WRITING MORE EFFECTIVE,
MEASURABLE PERMITS
Laura Gentile and John linger
U.S. EPA Region IX
San Francisco, CA
John Kosco, Wes Ganter, and James Collins
Tetra Tech, Inc.
Fairfax, VA
Abstract
Approximately 1,000 municipal separate storm sewer systems (MS4s) are permitted under Phase I of EPA's
storm water program. These Phase IMS4 permits require MS4s to reduce the discharge of pollutants to the
maximum extent practicable and prohibit illicit discharges to the MS4. Permit writers have discretion to
write permits specific to each MS4, or group of MS4s, resulting in a wide variety of permit requirements.
When these permit requirements are not specific, determining compliance with the permit can become
difficult.
The storm water Phase U program requires Phase U MS4s to include "measurable goals" in their program
for each BMP. Phase I storm water MS4 permits are beginning to include these measurable goals allowing
the permitting authority to assess whether each permitttee is in compliance. Specific examples of MS4
permits with 'enforceable' permit language are presented and discussed.
Introduction
On November 16, 1990, the U.S. Environmental Protection Agency (EPA) published regulations (the 'Phase
I rule') requiring National Pollutant Discharge Elimination System (NPDES) permits for certain industrial,
construction and municipal sources of storm water runoff and fundamentally changing the way storm water
runoff is regulated at the state and federal levels. Approximately 1,000 MS4s ('municipal separate storm
sewer systems'), consisting primarily of city and county government agencies responsible for storm water,
have been permitted under the Phase I regulations. The Phase IMS4 regulations generally require MS4s to
reduce discharges of pollutants to the maximum extent practicable and to prohibit illicit discharges to the
MS4. Specific elements in a Phase I Municipal Storm Water Management Program include public
education, public agency or municipal maintenance activities, new development, construction,
industrial/commercial facilities, illicit discharges and improper disposal, monitoring and reporting.
Most Phase IMS4 permits have been individual NPDES permits, often issued to multiple co-permittees.
Individual permits are written specifically to address the activities, pollutant sources, and discharges of the
covered co-permittees.
Phase II of the storm water program, established in 1999, extends NPDES storm water permit coverage to
include municipalities within urbanized areas. Phase U permits, to be issued beginning in March 2003, will
in most cases be general permits issued to a broad range of permittees.
134
-------
Storm Water Phase I Regulations
The Phase I storm water rule defines "municipal separate storm sewer" at 40 CFR 122.26(b)(8) to include
any conveyance or system of conveyances that is owned or operated by a state or local government entity
and is designed for collecting and conveying storm water which is not part of a Publicly Owned Treatment
Works (i.e., not a combined sewer). The Phase IMS4 regulations apply to MS4s serving populations of
100,000 or more. Some MS4s with populations under 100,000 can be designated for Phase I permit
coverage. In addition to larger cities and counties, many state Departments of Transportation were also
permitted under Phase I.
Phase IMS4 permits are required to establish controls to the maximum extent practicable (MEP) and
effectively prohibit non-storm water discharges to the MS4. MEP has not been defined by EPA, but is
intended to be flexible to allow the development of site-specific permit conditions based on the best
professional judgment of the permit writer.
The Phase I regulations required a two-part application process for Large and Medium MS4s (40 CFR
122.26(d)). The regulations only specified application requirements, not permit requirements. Therefore,
permitting authorities have various interpretations as to what should be required in an MS4 permit.
The Part 1 application required information regarding existing programs and the means available to the MS4
to control pollutants in its storm water discharges. In addition, Part 1 required field screening of major
outfalls to detect illicit connections. Part 2 of the permit application required a limited amount of
representative quantitative data and a description of proposed storm water management plans. The purpose
of the two-part application process was to develop information that would build successful MS4 storm water
programs and allow the permit writer to make informed decisions with regard to developing permit
conditions.
State and EPA permit writers used the information contained in these Part 1 and Part 2 permit applications
to write the individual NPDES permits. NPDES permits are issued for 5-year permit terms, with most of the
first round MS4 permits containing fairly general requirements. In many cases, these permits simply require
the permittees to implement the storm water management plan contained in the Part 2 application.
Subsequent MS4 permits, particularly many implemented in California, are more specific and include more
detailed requirements.
Permit examples: Unenforceable language
NPDES permitting authorities must be able to determine compliance with individual permits. In traditional
wastewater NPDES permits, this is a relatively simple process of verifying wastewater sampling results with
permit discharge limits. MS4 permits are BMP-based, therefore determining compliance with the MS4
permit is more difficult. The examples presented below illustrate MS4 permit language that is vague and
therefore difficult for an NPES permitting authority to determine compliance. Without specific, measurable
elements, almost any activity an MS4 takes could be deemed to be in compliance with the permit.
The permittee and permitting authority names have been removed, and the specific problems associated
with determining compliance with this permit language are discussed.
135
-------
Example 1
Permit Language:
The permittee shall demonstrate compliance with this Order through the timely implementation of
control measures and other actions to reduce pollutants in discharges to the maximum extent
practicable in accordance with their SWMP..."
This permit does not define what "timely implementation" is, allowing the permittee to determine what is
timely. Timely implementation could be up to 5 years in the view of the permittee, or within 6 months in the
view of the permitting authority. In addition, "other actions" are mentioned in the permit, but never
described. If the permit is going to require "other actions," then these actions should be specifically
described in the permit.
Example 2
Permit Language:
"Structural controls for water quality improvements are considered for inclusion in site drainage
plans, storm drain projects, and flood control projects where applicable."
A permit should not require the permittee to "consider" an action; it should require the permittee to take an
action. Also, "where applicable" leads to additional interpretation problems. If there are only certain
circumstances where this permit provision should be applied, then those circumstances should be spelled out
in the permit.
Example 3
Minimum best management practices (BMPs) include: standard plans and specifications,
maintenance of storm drain systems, street sweeping, litter control, spill response, and hazardous
material disposal.
This permit language lists a series of BMPs, but doesn't specify where, how much, or how often the BMPs
must be employed. For example, how often should the MS4 conduct street sweeping and how many miles
need to be swept in order to be in compliance with the permit? The permit language above does not specify
this.
Example 4
The permittee shall control pollutants in storm water discharges to the maximum extent practicable,
and to demonstrate compliance with this requirement, the permittee shall implement in its entirety
the proposed storm water management program (SWMP) described in ...
This permit requirement repeats the regulation language to control discharges to the "maximum extent
practicable" without specifying exactly how that will be achieved. Implementation of a storm water
management program (again, unspecified in the permit) is assumed to meet this standard. Unless the SWMP
describes the activities and set specific performance expectations for those activities, compliance will be
difficult to determine.
Permit Examples: Enforceable permit language
The most difficult aspect of writing MS4 storm water permits is drafting permit language whereby
compliance can be easily determined.
136
-------
The following sections provide examples of permit language that provides more measurable permit
language where compliance can be more easily determined.
Construction Inspections Example:
From the Orange County Municipal Storm Water NPDES Permit: (Board Order No. R8-2002-0010, NPDES
Permit No. CAS618030)
Each permittee shall conduct construction site inspections for compliance with its ordinances
(grading, Water Quality Management Plans, etc.) and local permits (construction, grading, etc.).
Inspections shall include a review of erosion control and BMP implementation plans and an
evaluation of the effectiveness and maintenance of the BMPs identified. Inspection frequency will,
at a minimum, include the following:
a. During the wet season (i.e., October 1 through April 30 of each year), all high priority sites are
to be inspected, in their entirety, once a month. All medium priority sites are to be inspected
at least twice during the wet season. All low priority sites are to be inspected at least once
during the wet season. When BMPs or BMP maintenance is deemed inadequate or out of
compliance, an inspection frequency of once every week will be maintained until BMPs and
BMP maintenance are brought into compliance. During the 2001-2002 wet season, prior to
the development of the inventory database, all construction sites must be visited at least
twice. If a site is deemed out of compliance, an inspection frequency adequate to bring the
site into compliance must be maintained;
b. During the dry season (i.e., May 1 through September 30 of each year), all construction sites
shall be inspected at a frequency sufficient to ensure that sediment and other pollutants are
properly controlled and that unauthorized, non-storm water discharges are prevented; and,
c. Information including, at a minimum, inspection dates, inspectors present and the results of the
inspection, must be maintained in the database identified in Section WI. 1 or must be linked
to that database. A copy of this database must be provided to the Regional Board with each
annual report.
This permit language describes what needs to be conducted (inspections), when (October 1 through April
30) and how often (once a month). This ensures that both the permitting authority and the permittee
understand what needs to happen to ensure compliance.
Construction Training Example:
From the Municipality of Anchorage and the Alaska Department of Transportation and Public Facilities
NPDES permit: (NPDES permit No. AKS 05255-8)
"Permittee shall develop a training program for construction site operators and developers.. .within
24 months of the effective date of this permit. Permittee shall ensure that such training is provided
at a minimum of once per year..."
This permit language specifies the action (a training program), a deadline for achieving the action (within 24
months), and a frequency for continuing performance (once a year).
Illicit Discharge Example:
From the City of Long Beach Municipal Storm Water NPDES Permit" (Board Order No. 99-060, NPDES
Permit No. CAS004003)
137
-------
"The Permittee shall inspect those portions of the storm drain system consisting of storm drain pipes
36 inches in diameter or greater, for illicit connections within 5 years after the permit is adopted."
This permit provision specifies the minimum pipe size expected to be inspected and specifies that the
permittee has up to five years to complete this task. Interim deadlines could also have been set here by, for
example, requiring that at least 50% of these pipe are inspected within 3 years.
Public Education Example:
From the City of Stockton and County of San Joaquin Municipal Storm Water NPDES Permit: (Board
Order No. R5-2002-0181, NPDES Permit No. CAS083470)
At least three times during the life of the permit, Permittees shall send information on problems
caused by storm water runoff and potential solutions to each household within the service area.
Both a timeframe (life of the permit, or 5 years) and a target number (each household within the service are)
are specified along with a quantity (three times) in this public education example.
Industrial storm water inspection example:
From the Orange County Municipal Storm Water NPDES Permit: (Board Order No. R8-2002-0010, NPDES
Permit No. CAS618030)
"After July 1, 2003, all high priority sites are to be inspected at least once a year; all medium priority
sites are to be inspected at least once every two years; and all low priority sites are to be inspected at
least once per permit cycle."
This permit language sets specific inspection frequencies for high, medium and low priority industrial
facilities. In order to be effective, the permit must also specify, or provide a clear expectation, of the types
of facilities that should fall into each priority category.
Municipal Maintenance Example:
From the City of Long Beach Municipal Storm Water NPDES Permit: (Board Order No. 99-060, NPDES
Permit No. CAS004003)
Catch basin maintenance, under Permittee's jurisdiction, shall include:
a. All catch basins will be cleaned out and inspected one time between May 1 and September
30 of each year; and,
b. All catch basins that are at least 40% full of trash and debris between October 1 and April
30, shall be cleaned-out.
This permit provision sets the amount expected (all catch basins), the time frame (May 1 to September 30),
and the frequency (each year). It also establishes a performance expectation for when a catch basin should
be cleaned.
New Development - Maintenance example:
From the Los Angeles Region Municipal Storm Water NPDES Permit: (Board Order No. 01-182, NPDES
Permit No. CAS004001)
"Maintenance Agreement and Transfer
Each Permittee shall require that all developments subject to SUSMP and site specific plan
requirements provide verification of maintenance provisions for Structural and Treatment Control
138
-------
BMPs, including but not limited to legal agreements, covenants, CEQA mitigation requirements, and
or conditional use permits. Verification at a minimum shall include:
a) The developer's signed statement accepting responsibility for maintenance until the
responsibility is legally transferred; and either
b) A signed statement from the public entity assuming responsibility for Structural or
Treatment Control BMP maintenance and that it meets all local agency design standards;
or
c) Written conditions in the sales or lease agreement, which requires the recipient to assume
responsibility for maintenance and conduct a maintenance inspection at least once a year;
or
d) Written text in project conditions, covenants and restrictions (CCRs) for residential
properties assigning maintenance responsibilities to the Home Owners Association for
maintenance of the Structural and Treatment Control BMPs; or
e) Any other legally enforceable agreement that assigns responsibility for the maintenance of
post-construction Structural or Treatment Control BMPs."
In this example, SUSMP stands for Standard Urban Storm Water Mitigation Plan and is a relatively new
requirement in California MS4 permits to address post-construction storm water impacts. CEQA is the
California Environmental Quality Act that requires environmental review of certain projects.
These permits provide more specifics, including set frequencies, deadlines, and detailed expectations for the
permittees. This allows both the permittees and the permitting authority to determine compliance.
Effective MS4 Permit Writing
NPDES MS4 permits and MS4 stormwater management programs must contain quantifiable, measurable
elements so that compliance can be determined. Storm water permits vary significantly in their level of
detail. For example, some third-term permits issued in California contain very specific, measurable
elements which are clear for permittees to implement and relatively straightforward for the state to
determine compliance. For nonspecific permits that simply require the MS4 to "implement a storm water
management plan," compliance becomes more difficult. More importantly, the permit does not specify, or
measure, the level of effort expected, so MS4s do not have a clear target to achieve.
The storm water Phase n regulations require small MS4s to develop "measurable goals" for each BMP in
their programs. These measurable goals are intended to provide quantifiable targets for the MS4s to achieve
in the implementation of BMPs. Although a similar requirement does not specifically exist for Phase I,
permits and programs developed under Phase I should also contain these measurable goals. This provides a
level of certainty to the MS4 that they are successfully implementing the permit and allows the state to more
easily evaluate compliance.
Some MS4 permits in California include specific, measurable requirements that make determining
compliance easier. Also, the City and County of Sacramento have developed stormwater plans that are
clear, well-written, and begin to address the issue of measurable goals which are called 'minimum
performance standards' and 'performance and effectiveness measures', respectively, in each plan (City of
Sacramento, 2000 and County of Sacramento, 2000).
139
-------
In order to be measurable, each permit requirement should specify:
• What needs to happen
• Who needs to do it
• How much they need to do
• When they need to get it done
• Where it is to be done
For each permit requirement, "what" is usually the BMP or activity required, "who " in most cases is
implied as all the permittees (although in some cases the permitting authority may need to specify exactly
who the require applies to), "how much' is the performance standard the permittee is expected to meet (how
many inspections), "when" is a specific time (or a set frequency) when the BMP or activity should be
complete, and "where" is the specific location or area (if necessary). Without these specifics, it is almost
impossible for the permitting authority to determine compliance with a vague MS4 permit.
Writing more specific, measurable permits will take more time and resources than writing less specific ones.
For Phase IMS4 permits, which are in some cases entering their 3rd round of MS4 permits, these more
specific permits are becoming a necessity. States are finding that both the regulated community and the
public are demanding more accountability, which the specific, measurable permits provide.
Conclusions
With over 1,000 large cities, counties, and other governmental organizations under storm water Phase IMS4
permits, a significant amount of money is being spent implementing these programs. Unless the permits are
written with specific, measurable requirements, determining compliance with permits is often difficult, if
not impossible.
Permit writers can develop these specific, measurable permit requirements by building upon existing storm
water permit programs and ensuring that permit elements address:
• What needs to happen
• Who needs to do it
• How much they need to do
• When they need to get it done
• Where it is to be done
As Phase n MS4s begin the process of identifying measurable goals for each of the BMPs in their program,
permits issued to the larger, more mature Phase IMS4 programs should include these same measurable
elements.
References
City of Long Beach Municipal Storm Water NPDES Permit (Board Order No. 99-060, NPDES Permit No.
CAS004003) issued July 1999 by the California Regional Water Quality Control Board, Region 4.
City of Sacramento, 2000. Stormwater Quality Improvement Plan, Draft.
140
-------
City of Stockton and County of San Joaquin Municipal Storm Water NPDES Permit (Board Order No. R5-
2002-0181, NPDES Permit No. CAS083470) issued October 2002 by the California Regional Water Quality
Control Board, Region 5.
County of Sacramento, 2000. Stormwater Quality Improvement Plan for County of Sacramento and Cities
of Citrus Heights, Elk Grove, Folsom, and Gait. Draft.
Los Angeles Region Municipal Storm Water NPDES Permit (Board Order No. 01-182, NPDES Permit No.
CAS004001) issued December 2001 by the California Regional Water Quality Control Board, Region 4.
Municipality of Anchorage and Alaska Department of Transportation and Public Facilities (NPDES permit
No. AKS 05255-8) issued January 1999 by the U.S. Environmental Protection Agency, Region X.
Orange County Municipal Storm Water NPDES Permit (Board Order No. R8-2002-0010, NPDES Permit
No. CAS618030) issued January 2002 by the California Regional Water Quality Control Board, Region 8.
141
-------
CONSERVATION DESIGN TOOLS FOR
STORMWATER MANAGEMENT
Randell K. Greer, P.E.
Delaware Department of Natural Resources and Environmental Control
Dover, DE
Abstract
The release of Delaware's "Conservation Design for Stormwater Management" document in 1997 provided
guidance to land use planners and civil site design consultants in the application of conservation design principles to
meet regulatory stormwater management requirements. Proof of concept in this document relied on traditional
techniques based onNRCS methodology, such as "Technical Release No. 55", to verify the results. However,
this was a cumbersome approach, since these methods do not easily model Best Management Practices (BMPs)
such as biofiltration swales, bioretention practices and riparian buffers. It became apparent that new tools would
have to be developed to model these practices so that prospective developers were given full credit for their
implementation. As a result, the Delaware Department of Natural Resources and Environmental Control (DNREC)
with assistance from outside contractors, have developed two design tools for use with this so-called "Green
Technology" approach to stormwater management. The Delaware Urban Run-off Management Model (DURMM)
accounts for both disconnection of impervious area as well as the "run-on" process to derive both the volume and
rate of run-off from a given site. A decision tool is also being developed based on USDA's Riparian Ecosystem
Management Model (REMM) for designing riparian buffers in an urban environment for both quantity and quality
control of stormwater runoff. This decision tool is still under development. Therefore, this paper will focus on the
development of DURMM and how it will be used to fulfill the Delaware regulatory requirements for stormwater
management. It is also felt that both these tools have application outside the State of Delaware, with the caveat that
the local regulatory authority conducts proper testing and verification.
Background
The State of Delaware has had a Sediment & Stormwater Law in effect since 1990. While the law and subsequent
regulations were instrumental in mitigating many of the negative impacts associated with urbanization, it soon became
clear that traditional approaches were leading to an over dependence on structural practices. If this trend were to
continue, the operation and maintenance requirements for these structural practices would become a tremendous
burden for the entities responsible for them. In 1996, the Delaware Department of Natural Resources and
Environmental Control partnered with the Brandywine Conservancy to develop a manual for a new approach to
stormwater management. The goal would be to mimic the natural hydrology of a site as much as possible without
relying on structural practices. This new approach to stormwater management was referred to as "Conservation
Design".
The "Conservation Design for Stormwater Management" document was released in September, 1997. It provided
background information on the hydrologic impacts associated with urbanization and explains how making better use
of the existing physical features of a site can minimize the increases in stormwater runoff that often accompanies land
development. This can be accomplished by altering the building program, minimizing impervious surfaces and
142
-------
disconnecting those impervious surfaces wherever possible. Where additional management is required to meet
regulatory requirements, the emphasis is on non-structural measures, or "Green Technology BMPs", such as
vegetated swales, biofiltration practices, terraforming, riparian buffers, etc.
Proof of concept for the Conservation Design approach was provided through six case studies of actual
development projects throughout the State. The traditional development plans were conceptually redesigned
utilizing the Conservation Design principles, while maintaining the original density and unit counts. Stormwater
management computations were also completed to ensure full compliance with the existing regulations. These
computations were based on traditional NRCS methodology. Although the results confirmed the benefits, it proved
to be a rather tedious process. It was clear that an improved methodology would be necessary to take full
advantage of this approach. With the assistance of several outside contractors, the DNREC has developed two
design tools, the Delaware Urban Runoff Management Model (DURMM) and the Urban Riparian Buffer Design
Decision Tool, that will hopefully fill this need.
Delaware Urban Runoff Management Model (DURMM)
Traditional structural BMPs such as stormwater ponds and wetlands can be effective in controlling peak flows from
a site. However, current regulatory requirements in the State of Delaware do not address the frequent storms that
erode stream banks, and do little or nothing to promote recharge. Furthermore, structural BMPs can contribute to
downstream flooding when discharges from separate on-site structural BMPs overlap. Structural BMPs can be
effective in pollutant removal; but since they generally omit recharge, consume space, and require extensive
maintenance, they are less appropriate for the task. There is an emerging body of research indicating that these
BMPs contribute to elevated stream temperatures, and discharge algae laden effluent, which can substantially
degrade the benthic community in the receiving stream [Delaware Department of Natural Resources and
Environmental Control and B. Lucas, 2002].
As a result, many progressive agencies are promoting a less structural approach, designed to intercept runoff from
rooftops, parking lots and roads as close as possible to its source, and direct it into recharge/filtration facilities
incorporated into the overall site design and runoff conveyance system. Nonstructural BMPs thus include impervious
area disconnection, conveyance of runoff through swales and biofiltration swales, filter strips, terraces, bioretention
facilities, and infiltration facilities. However, while these BMPs may seem less significant than structural BMPs, the
procedures for their proper design require the same hydrologic and hydraulic methods used in designing structural
BMPs. Otherwise, realistic estimates of effectiveness are difficult to quantify. These so-called "Green Technology
BMPs", form the basis of DURMM at the site engineering level.
The BMPs addressed in DURMM and pertinent aspects of their design and performance are briefly summarized
below:
Source Area Disconnection- Disconnecting flow from impervious surfaces so it discharges onto adjacent
pervious areas provides additional infiltration and potential for some pollutant removal.
Filter Strips - This BMP provides for runoff to be spread uniformly over a filtering surface of vegetation,
which can provide substantial treatment if not overloaded by sediment and runoff
143
-------
BioFiltration Swales/Grassed Swales- Research shows that overland conveyance through properly
designed swales can be very effective in removing Total Suspended Solids (TSS) and adsorbed metals,
although less effective in terms of nutrients. While swales are not thought to be capable of quantity
management, designs incorporating check dams can provide substantial attenuation of peak flows.
Terraces- Terraces are essentially swales extending across slopes to intercept runoff and increase the
potential for infiltration. Terraces are similar to swales in terms of runoff responses and pollutant removal with
the exception that flow exfiltrates laterally.
Bioretention Structures- These landscaped pocket depressions incorporated into the urban landscape can
provide substantial filtering and nutrient transformations before runoff is discharged into the conveyance
system. Ongoing research suggests that this BMP can be designed to have substantial nitrogen removal
capabilities, unlike most other BMPs. [Delaware Department of Natural Resources and Environmental
Control and B. Lucas, 2002].
Infiltration Practices- Most non-structural BMPs incorporate infiltration as part of the treatment process. Specific
infiltration facilities include trenches, basins and dry wells. Infiltration trenches located in swales provide additional
wetted surface area and storage volume, and often they can be designed to penetrate shallow impermeable soil
profiles to recharge deeper soil horizons.
Unfortunately, while there is great interest in using nonstructural BMPs, there are few rigorous procedures available
for the engineering and regulatory community to utilize in designing them. Many regulatory programs use a
straightforward runoff volume approach, in which the increase in small storm runoff volume due to land development
is to be treated and/or retained on site. However, this approach typically assumes a constant runoff volume in
proportion to rainfall amount, and does not route runoff through nonstructural BMPs. Instead, simplified
volume/outflow equations are specified, without knowing precisely how they work during storm events. When this
approach leads to overdesign, it may be beneficial if the original reduction targets are inadequate, otherwise it
causes unnecessary expense. Where it leads to underdesign, the hydrological impacts are not adequately mitigated.
DNREC has partnered with a private consultant, Mr. William Lucas of Integrated Land Management, Inc., to
create DURMM to provide a more rigorous hydrological design tool for nonstructural BMPs. A spreadsheet
program is provided that incorporates modified TR-20 storm hydrology to project the hydrological response from
contributing source areas. It segregates directly connected runoff from that which flows overland. It provides
routines that account for the reductions in peak flow due to overland conveyance. In this way, it is possible to more
precisely determine the actual volume and peak rate reductions over the duration of a 24 hour storm event, and
through the following days. This is particularly important for calculating total infiltration, and designing proper stream
bank erosion controls. Furthermore, since the design community is already familiar with TR-20 input variables, the
same input data parameters required for design of flood controls can be used for design of quality treatment,
streambank protection, and conveyance runoff events.
The process of BMP design involves a spreadsheet file for each source subarea and its BMP. Discrete
combinations of hydrological soil group and land cover are averaged to generate composite Curve Numbers (CN)
144
-------
for the pervious and impervious portions of each source area. Impervious areas are calculated separately, and
routed according to the extent of their linkage with adjacent pervious surfaces. The resulting runoff hydrograph from
the source area worksheet is imported into the BMP hydraulic design worksheet. Pollutant loading is calculated by
applying typical event mean concentrations (EMCs) to the runoff volume allocated to each type of pervious and
impervious surfaces.
Site design parameters of infiltration rates, surface and subsurface stage/storage, and outflow controls are entered
into the BMP worksheet. The worksheet routes the source area hydrograph through the BMP based upon the input
parameters. The resulting output displays peak flows, flow duration and infiltration volume for each storm event.
By segregating subarea loads according to the type and extent of land cover, the discrete source area approach
used in the hydrologic calculations refines accuracy in estimating total pollutant loads. Pollutant removal by the BMP
is based upon physical parameters such as slope, pretreatment volume, hydraulic residence time, surface/volume
ratio, filter media type, and underlying infiltration characteristics. Given these factors, pollutant load reduction is
calculated by algorithms relating input concentrations and decay transformations to estimated mass removal for each
pollutant of concern.
The reported pollutant removal effectiveness of BMPs can be highly variable. However, by incorporating hydrologic
and hydraulic parameters in runoff routing, and addressing the various removal processes as discrete algorithms
within a BMP, more accurate estimates of removal rates are possible. Some variability in projected removal rates is
acceptable in any event, since hydrological changes are recognized as perhaps the primary impact of runoff.
Furthermore, polluted runoff from the most frequent storms that causes the greatest stress can often be eliminated by
the infiltration components of nonstructural BMPs.
Conclusions
The Delaware Urban Runoff Management Model (DURMM) was developed to facilitate the adoption of so-called
"Green Technology BMPs" in the land development process. This tool is based on rigourous, physically-based
methodologies. Yet at the same time, it has advantages in ease of use over the traditional models now being used
for stormwater management analysis. It is hoped that the additional development of the riparian buffer decision tool
based on the REMM will provide designers with two powerful, quantitative tools that will further encourage the use
of Conservation Design techniques.
The DNREC is currently embarking on an extensive outreach and education effort with the design community to
introduce this tool and familiarize them with its mechanics. It is anticipated that this effort will allow designers to
become proficient with its use within a year's time.
References
Delaware Department of Natural Resources and Environmental Control and The Brandywine Conservancy, 1997.
"Conservation Design for Stormwater Management". Dover. DE.
145
-------
Delaware Department of Natural Resources and Environmental Control and B. Lucas, 2002. "Delaware Urban
Runoff Management Model: A Technical Manual for Designing Green Technology BMPs to Minimize Stormwater
Impacts From Land Development". Dover, DE
Delaware Department of Natural Resources and Environmental Control and B. Lucas, 2002. "Delaware Urban
Runoff Management Model: A Users Manual for Designing Green Technology BMPs to Minimize Stormwater
Impacts From Land Development". Dover, DE
146
-------
USING TECHNICAL DATA AND MARKETING RESEARCH
TO CHANGE BEHAVIOR
Stephen Groner
S. Groner Associates, Inc.
Long Beach, CA
Abstract
The City of Los Angeles is faced with the task of educating over three million residents regarding the
various pollutants effecting water quality. With limited resources, the City is challenged with effectively
reaching and influencing the greatest number of residents who have the greatest impact on improving water
quality.
To develop this program, S. Groner Associates, Inc. (SGA) was hired by the City to develop and implement
a strategic social marketing plan based on technical data and marketing research. The goal of the plan was to
target audiences who have the greatest impact on water quality. With those key groups in mind, outreach
efforts were developed based on the specific audiences' attitudes, styles, and behaviors. This would focus
resources most cost effectively on efforts with the greatest chance to influence behavior change and thus
prevent pollution.
In developing the plan, we used technical data analysis and existing market research information to
determine the following:
* activities posing the greatest threat to water quality
* activities/behaviors most influenced by public education
* audiences engaged in those activities
* psychographics of the audience (i.e., attitudes, characteristics and styles of the audiences)
* methods to reach our audiences to increase the influence of the outreach
These key points served as the foundation for developing outreach efforts as well as the emphasis, style, and
tone of our communication pieces.
This presentation reviews the role of market research and data analysis in developing a social marketing
plan, in addition to designing marketing materials and implementing the outreach efforts. We will also
illustrate how incorporating new marketing data helped gage the outreach's successes and areas for further
refinement.
Building a Foundation for the Marketing Plan
Effective outreach requires developing a solid information base about behaviors you want to change. The
information ascertained assists in determining how and who to target in order to maximize the impact on
improved water quality.
This information is highly effective when developing a social marketing plan for stormwater pollution.
Building the marketing plan's foundation, however, is still very difficult because of the complexity of
stormwater. By its very definition, stormwater pollution or non-point pollution is not one single source, but
a complex collection of problems to target. In turn, developing a marketing plan for stormwater pollution is
147
-------
not just about developing one plan to address one issue. It requires a multi-pronged campaign that attacks
the various causes of stormwater pollution. The plan must be composed of several focused marketing
strategy's each addressing the most problematic pollutants/behaviors.
Because of the issue's complexity, understanding the limits of your resources and strategically focusing
your outreach to maximize impact is essential. In determining our outreach, we used both technical data and
existing market research information to strategically lay out our direction for public education. This process
included the following steps:
• Determine the pollutants/activities posing the greatest threat to water quality
• Determine what corresponding activities/behavior are best influenced through public education
• Identify the audiences engaged in those activities/behaviors
• Understand the psychographics of the audiences (i.e., attitudes, characteristics and styles of the
audiences)
• Understand the motivators that will best influence our audiences
Any effective marketing plan or outreach effort must be designed with an intimate understanding of the
audience that you are targeting. Many times, there is a misconception that because the issue is important
people will automatically listen to it. But the message is competing for attention with thousands of other
messages that bombard residents everyday; everything from ads selling cars and beer, to other social
marketing ads like recycling campaigns, anti-smoking campaigns, or drug prevention campaigns. In the end,
if the outreach piece is generic and does not in someway connect with a specific audience and compel them
to listen, they won't.
This paper lays out the methodology used to develop a solid social marketing plan and introduces the City
of Los Angeles' public education program as an example of this type of strategic planning's success.
Prioritizing Pollutants
The first step in targeting outreach is determining the pollutants that pose the greatest threat to water quality.
This effort requires an analysis of water quality data and reconciling this information across watersheds if
the jurisdiction covers more than one watershed.
In conducting this technical research for the City of Los Angeles, we worked with GeoSyntec Consultants,
Inc. to evaluate and analyze water quality data from the City and County of Los Angeles. The City of Los
Angeles lies within three primary watersheds and a multitude of subwatersheds. The three primary
watersheds all have broad similarities of a mostly urban environment. However, at the subwatershed level,
there were vast differences in the environment.
After results were evaluated, five pollutants were selected for the campaign:
Bacteria/Pathogens
- Pesticides
- PAH's
Nutrients
Trash and Debris
148
-------
Prioritizing Activities to Target
Prioritizing activities requires balancing technical information about pollutants with an understanding of
which pollutants are most effectively targeted through public education.
Based on this, we looked at activities that produce pollutants and prioritized which activities could most
effectively be targeted. The following criteria served as a guideline for prioritization:
How pervasive the activity is across the target area
How active or passive is the polluting activity
How effective behavioral BMPs are vs. structural BMPs
How complex or simple the solutions are to implement and
Where possible evaluate the proportion of pollution the activity contributes to the total pollutant load
Ideally, this process begins with analyzing pollutant source data. However, given the nature of
stormwater/non-point source pollution, this information may not available for most jurisdictions, so there is
a need for best judgment.
In addition, the area and process of evaluation, be it individual watershed, across a jurisdiction, or across a
regional area, must be determined. This issue is important for obvious environmental science reasons as
well as strategic marketing reasons. For example, a pollutant or activity in one watershed may be prioritized
differently if evaluated in different areas because of the watershed's maximum sustainable load. But from a
marketing perspective, this determination will be critical in determining what outreach methods are most
effective and available in the area to reach the target audience. For example in some areas billboard
advertisements or newspaper advertisement may target the area and activity well, while in other areas an
activity may be best targeted through point of purchase advertising.
Applying the above criteria helped prioritize the activities as the following:
Bacteria/Pathogens
o Leaks from sewer systems - low
o Improper BMPs at Restaurants — medium
o Owners picking up after their dogs — high
o Improper BMPs by horse owners - low region wide - high in certain subwatersheds
o Proper maintenance by septic system owners - low region wide - medium in certain
subwatersheds
- Pesticides
o Residential users - high
o Commercial users - low
o Government users — medium
149
-------
- PAH's
o Vehicle leaks - low
o Improper BMPs at auto repair shops/gas stations - medium
o Improper disposal of vehicle fluids by residents - high
Nutrients
o Fertilizer application by residents - high
o Proper maintenance by septic system owners — low region wide - high in certain
subwatersheds
Trash/Debris
o Active littering by residents - high
o Litter from uncovered trash containers — low
o Litter from uncovered commercial vehicles - low
Identifying Audiences
In selecting a target audience, the program's developmental focus shifts from the technical field to the
marketing arena. Marketing research is key to identifying which audiences, or in marketing terms "segments
of the population," are engaged in the problematic behavior. The next step is discovering common
characteristics among the audience and developing a focused message that is tailored to their interest and
motivations.
The best way of collecting this information is to conduct surveys of residents. The survey would incorporate
questions to ascertain what types of residents are engaged in the improper behavior. Cross referencing the
results with psychographic information (i.e., attitudes, behaviors, lifestyles, which "segments of the
populations"), helps target how to best address the issue and change behavior.
This approach, however, can be expensive. Depending on the campaign's size, less expensive and simplified
research can yield similar information. One effective method is matching up behavior with a consumer
market. For example, when targeting people who improperly dispose of their oil, you can target people who
buy oil and identify them as consumers at auto parts stores. While this will not narrow your audience down
to only those who are illegally dumping their oil, it serves as a solid starting point for further refinement.
Later, a simple intercept survey conducted at auto parts stores can help better assess the audience and hone
strategies to target the audience.
Another cost-effective way to understand your audience is through the use of the US Census Bureau's Web
site. The Census Bureau's site gives demographic and socio-economic information broken down by city, zip
code, and census tract. The site allows you to import the data to spreadsheets or even use a Web based GIS
software program to map the data. This information is extremely valuable in targeting an activity that may
focus in on a specific area. One example would be targeting homeowners in a specific area. From the
website, you could identify homeownership rates and then correlate that to other demographic and socio-
economic information such as income levels, languages most commonly spoken, ethnic background,
employment rates, etc.
Understanding Psychographics
Once the target audience is identified, the next step is understanding the "psychographics" of the audience
(their attitudes, interests, and styles). This information provides insight into the audience's thoughts and is
150
-------
an essential step in designing effective outreach. Without outreach efforts/messages that connect directly to
a group's sensibilities, interests, or concerns, changing a habit is almost impossible.
Obtaining an audience's psychographic information is more difficult than merely identifying the audience.
To gather information regarding attitudes, marketing surveys are critical. The surveys gather relevant
information by correlating residents' interests and priorities with their activities and behaviors. The resulting
information helps isolate key issues and motivators relevant to the audience.
Depending on survey results regarding a target audience's priorities and motivators, an issue could be
positioned in various ways. For example, the issue of pesticide use could be presented with three different
focuses depending on the audiences' psychographics:
1) as an environmental issue (chemicals impact on the watershed),
2) as a "dollars and cents" issue (addressing the source is cheaper than treating the problem), or
3) as a family/child safety issue (safety concerns of children playing on a lawn with chemicals).
Understanding the psychographics of the target audience, simplifies choosing the most meaningful and
effective message.
If creating and/or performing a survey is not possible, relevant information based on a similar issues or
audiences can be frequently found in marketing surveys completed by other organizations. In researching
segmentation information for the City of Los Angeles, SGA based its information on three previously
completed marketing research surveys that could be analyzed for information relevant to the City's
demographics characteristics: two were conducted by the County of Los Angeles (one on stormwater issues
and one focused on do-it-yourselfers) and one conducted by the State (on residential used oil recycling).
The resulting information gave SGA a full picture of various target audiences and helped differentiate our
messages based on each particular audience. Based on the results from our marketing research and technical
data, we identified and prioritized our three main target audiences:
1) Neat Neighbors -
Description - Younger families with children who want to do the right thing but needed a little
coaxing
Motivators to change - Concern about children, concerned about the neighborhood, interested
in doing what's good for the environment
Activities/Behaviors - Picking up after pets, pesticide and fertilizer use
2) Fix-it Foul-ups -
Description - Middle class homeowners who are do-it-yourselfers
Motivators to change - Put family first, want to follow rules, not interested in the environment
Activities/Behaviors - Pesticide and fertilizer use
3) Rubbish Rebels -
Description - Younger males who are just getting out on their own
Motivators to change - Concern with their image and peer's perception of them; following
rules is not "cool"
151
-------
Activities/Behaviors - Littering and used oil disposal
Strategic Outreach
The last element is determining the best outreach efforts to effectively reach the target audiences. The goal
is to identify outreach efforts that strategically delivers the message and increases the message's influence
on their behavior. SGA looked at three elements in evaluating the outreach strategy:
1) How timely is the message in relation to the activity
2) How well placed is the message to reach the target audience
3) How well delivered is the message to catch the attention of the audience
These three elements help compare potential outreach methods in terms of the ability to reach and influence
the audience. The first element addresses the issue that people receive information all the time, but unless it
is delivered at a relevant time, the audience may not focus their attention and note the information. A good
example of this is giving out information on pet care when one gets a pet. The timing is perfect because the
owner is excited about the pet and is open to learning about them. Delivering the pet message at this time
also increases the likelihood of changing behavior because the owner has not developed bad habits yet.
The second element addresses the quality and focus of outreach aimed at the target audience. For example,
an ad in a newspaper regarding pesticide use may reach a large number of residents but may not be
strategically placed, and therefore, does not effectively reach the target audience. However, an ad placed in
the weekly "Home and Gardening" section of the newspaper would be far more effective because it was
strategically placed in an area relevant to the specific readers of that section.
The third element addresses how effectively the outreach method catches the audience's attention. For
example, a small logo placed on a banner for an event may not be noticed next to a dozen other logos.
However, a well placed booth at an event with a staff member actively approaching the target audience (as
opposed to waiting for them to approach the booth) can be far more effective.
Examples of how these evaluations helped in developing
strategic outreach methods to address high priority
activities for the City are the following:
Picking Up After Your Pet
Material placement at animal shelters and
inclusion with pet adoption materials
- Participation a pet adoption events held by
animal shelters
Point of purchase displays at pet stores
Material placement at veterinary clinics
In this effort, we partnered with animal shelters. This
allowed us to deliver our message to residents who were
planning on adopting new pets. We accomplished this by
placing our information in animal shelters' adoption
package- an item given to all new pet owners. To reach the
same audience, we also set up a booth and distributed
152
up after yonr dog
to Wp curl) pollution.
-------
information at pet adoption events. This outreach strategically accomplished three goals:
1) ensured our information would be received by the new pet owner,
2) allowed us to get our information to the owner at a point in time when they are most interested in
learning about the new pet and
3) delivered the information to the owner before they developed bad habits regarding their pet's care.
In addition to animal shelters and pet events, pet stores and veterinary clinics were utilized as key venues for
outreach. Materials were strategically placed in immediate view of our target audience at a time when they
were thinking about their pet (i.e., shopping for their pet or bringing the pet in for medical attention).
Pesticide and Fertilizer Use
Partnership with home improvement stores
Develop point of purchase
o
o
displays
Conduct staff training to
enable employees to answer
questions
- Placed radio ads on a local
weekend gardening show
This effort entailed partnering with the
major home improvement chains (Home
Depot, Lowe's, OSH) and obtaining pro
bono placement of materials on the shelves
where pesticide and fertilizer products are
sold. This put the information in the
audience's direct view at the point in time
when they were deciding what product
(toxic or non-toxic) to purchase. We then
trained store staff on the issues concerning
urban runoff and pollution prevention
issues. The result gave us credible advocates
for our message right on the "frontlines."
Along with that effort, we sponsored a local
weekend talk show about gardening. This
effort was strategic in two ways:
1) it was targeted directly at residents
most likely to use fertilizers and
pesticides and,
2) the program aired on the weekend, a time when residents are likely to be engage in their gardening
or lawn care activities.
153
-------
Used Oil Recycling
- Partnerships with car clubs
Strategic radio advertising
Partnerships with auto parts stores
In this effort, we faced the challenge of reaching an audience that did not want to be reached; younger males
who didn't care about the environment or
recycling their oil.
YOUR
DOIl'I LEI TIP TRRSH IT,
RECYCLE YOUR USED OIL and FILTERS,
for the nearest location call I (888) CIEAH LA
To reach this audience, we focused less on the
message and more on the campaign's image as
well as the person delivering the message. In
reaching the audience, the messages were
delivered through peers and at familiar
venues. For example, booths were set up at
lowrider car shows, but rather than staffing the
booth ourselves, SGA teamed up with
lowrider car clubs. The car clubs then brought
their cars and distributed our message. This
gave the campaign credibility with our
audience, helped build a brand image for the
campaign and made the campaign "peer to
peer."
Another part of this effort was placing ads on radio stations our audience identified with. Based on the
marketing research, we were able to identify the radio stations our audience listened to most. We then
placed ads on their weekend program, which allowed us to air our ads around the time when do-it-
yourselfers change their oil.
The last effort was to place materials in auto parts stores. SGA did this by placing floor graphics in front of
the oil products shelf, posters in storefront windows, and counter cards by the cash register. This effort
ensured that our message reached those buying new oil and hit them when they were focused on their
vehicle.
Results/Evaluation
Determining the effectiveness of outreach is a critical element in any public education campaign. Obtaining
results and feedback allows you to refine your outreach efforts and tweak your strategy to improve your
efforts. However, in tackling outreach on stormwater pollution, certain targets may be extremely difficult to
obtain good information on, while other activities may be straightforward. The key is to set up several
feedback points. While none may be perfect, the goal is to collect enough data to determine a trend and give
a sense of the program's effectiveness.
This was exactly the case in the City's campaign. In evaluating our outreach on used oil, SGA had two solid
methods of obtaining feedback. The methods were 1) surveying auto parts stores regarding how much used
154
-------
oil they were collecting and 2) tracking the number of calls to the environmental hotline asking for
information on used oil recycling. In both cases, our numbers went up.
1) Used oil collection - 9% increase over the previous year
2) Call to the hotline - 120% increase over the previous year
In evaluating pesticide/fertilizer and pet outreach, we could only rely on indirect methods.
For outreach targeting pesticides and fertilizers, we evaluated participation at household hazardous waste
collection events and conducted qualitative quizzes during employee training classes. The results from
collection events showed over a 10% increase in volume collected, however the training classes showed
only an adequate retention of information (based on trainer's judgment no actual data collected). The
feedback on the training classes, while not positive, proved helpful. SGA concluded from the information
that shorter periods between training classes are needed to address employee attrition and bring new
employees up to speed on the program.
In our outreach to target pets, we focused our evaluation on surveying pet owners at adoption events. At this
point, we have no clear feedback yet. Our goal is to build up a database of information regarding the habits
of pet owners and then determine if habits change over time. Currently, we are still developing our baseline.
Conclusion
Overall, there are a multitude of outreach efforts that can be implemented, however, most programs have
very limited resources as well as the tough challenge of trying to change someone's behavior. Many
consumer marketing campaigns have huge budgets completely dedicated to marketing and advertising a
simple message such as switching brands. Our challenge is marketing an issue and in sighting a behavioral
change that may be inconvenient. This challenge is increased when combined with a lack of resources.
Therefore, developing a smart social marketing plan is imperative to successfully implement outreach.
Additionally, developing a social marketing plan helps guide and direct a strategic public education
campaign. Given the complexity of stormwater pollution, it is an invaluable tool in analyzing all the
potential options/directions for the campaign. To effectively maximize limited resources, strategic planning
using technical data to target the activities combined with the use of marketing research, is critical. These
two pieces of information (the data and research) help ensure that the outreach is targeted at the highest
priorities and that limited efforts can be as effective as possible. In the end, the more information gathered
about pollution, behaviors, and your audience, the better your chances of success.
155
-------
EVALUATING INNOVATIVE STORMWATER TREATMENT TECHNOLOGIES
UNDER THE ENVIRONMENTAL TECHNOLOGY VERIFICATION (ETV)
PROGRAM
Author: Donna B. Hackett, NSF International
Co-Authors: John Schenk, NSF International
and Mary Stinson, USEPA/NRMRL
NSF International and USEPA/NRMRL
Ann Arbor, MI and Edison, NJ respectively
Abstract
Assessing, controlling, and treating combined-sewer overflows (CSO), sanitary sewer overflows (SSO), and
urban stormwater runoff have become priorities for communities. Improved and cost effective treatment
technologies are needed to reduce the adverse impacts that wet weather flows can have on surface water
quality.
In October of 1995, the U.S. Environmental Protection Agency (EPA) created a program to facilitate the
deployment of such innovative technologies through performance verification and information
dissemination. The goal of the Environmental Technology Verification (ETV) Program is to further
environmental protection by substantially accelerating the acceptance and use of innovative commercially
available treatment technologies. The ETV Program is intended to assist and inform the stakeholders
involved in the design, distribution, permitting and purchase of environmental technologies.
Since potential adverse effects on surface water quality from wet weather flow sources has been targeted as
a major environmental concern, the Wet-Weather Flow (WWF) Technologies Pilot was created as one of
the 12 pilots formed under this ETV Program. Through a cooperative agreement, US EPA and NSF
International have partnered to conduct this Pilot. Objective, quality-assured performance data will be made
available to all parties in the WWF technology marketplace in the form of a Verification Report and
Statement. These will be published on the Web sites, http://www.nsf.org/etv and http://www.epa.gov/etv.
This paper will focus on one of the five areas selected as a high priority within the WWF pilot, stormwater
treatment. The stormwater treatment devices or systems being evaluated are designed to intercept and
thereby reduce pollutants before they can adversely affect surface water quality. Their function is to serve as
an effective Best Management Practice (BMP) to assist end users in complying with meeting NPDES Phase
U stormwater compliance permits and other regulatory requirements for protecting surface runoff quality.
Based on their operating principles, there are three basic types of BMP devices that are being verified: in-
line filtration devices, hydrodynamic separators, and in-drain filtration devices.
An overview of the generic protocol prepared for use as a template for site-specific test plan preparation will
be presented. The names of applied vendors, the names and operating principles of their devices,
performance measures included in their test plans, and test site locations will be presented. The field-testing
organization that developed the test plan and performed the testing for each device will also be identified.
In conclusion, the testing process and available data will be discussed.
156
-------
PROTOCOL OVERVIEW
As an initial step in the verification process, the process of developing a protocol was embarked upon with
the guidance of a six-member technology panel of experts in this field. The chairman of this panel is Roger
Bannerman from the Wisconsin DNR. Other members include Michael Bloom from PBS&J, Stan Ciuba
from WA Dept. of Ecology, Jeff Dennis from Maine DEP, Tom Maguire from MA DEP, and Rod Frederick
from the EPA, Office of Water. The protocol was prepared under contract with Earth Tech, Inc., and peer
reviewed by Dale Scherger of Scherger and Associates. This protocol serves as a generic template for
preparation of site-specific test plans.
Both the technology panel and Dale's review deemed the protocol to be generally acceptable, with
expectations that modifications and improvements would be made as test plans are drafted.
The latest version of the protocol for stormwater source area treatment devices is Draft 4.1, March 2002,
and is available on both the NSF International and EPA ETV web sites. This document has evolved from
several earlier versions of the original protocol.
The main elements of the protocol are as follows:
• Minimum 15 qualified sampling events required
• Automatic composite sampling (except HC/micro) - Minimum 5 subsamples
• Pollutant list based on vendors claims - Core list by pollutant category:
- Solids (TSS, TDS, Settleable Solids)
- Nutrients (P, TKN, Nitrate.Nitrite, Ammonium)
- Heavy metals (Zn, Pb, Cu, Cd)
- Petroleum/ Hydrocarbons (TPH, PAH series)
- Microbiological/Bacteria (Fecal Coliform, E.coli)
Technology panel recommendations that were added in this latest version (Draft 4.1) after a technology
panel meeting in November of 2001 include:
• Adding a requirement of suspended sediment concentration as a measure of solids load
• in addition to TSS, including sand/silt split
• Provision of additional guidelines on proper use of automated samplers and sample splitting
• Permitting, but not mandating, analysis of captured sediment/pollutants
• Improving guidance on sampling and lab Quality Assurance
Additional technology panel recommendations still under discussion from November's meeting include:
• Providing guidelines for characterizing trash & debris removal, but not establishing
• removal efficiency quantification procedures
• Adding language about collaborating with other protocol developers, and sharing 3rd
• party credible data generated
• Revisiting some target detection limits and comparing them to other protocols.
157
-------
Stormwater BMP Vendor Applications
The stormwater treatment devices being evaluated under the ETV program are designed to reduce the level
of one or more constituents of concern in stormwater drainage from a site. These parameters include
sediment or particulates, nutrients, heavy metals, petroleum hydrocarbons, and bacteria. The test plan
created for a specific device being verified at a given location contains the manufacturers' removal claims
relative to any number of these constituents.
To date, twelve vendors have applied for verification of their devices. These devices can be divided into
three categories based on their operating principles: In-line Filtration Devices, Hydrodynamic Separators,
and In-drain Filtration Devices.
In-line Filtration Systems
As the name of this category implies, these types of BMP devices employ some type of filtration media as
the mechanism for removal of stormwater constituents in an in-line device. There are three vendors who
have signed up for verification under this category:
1. Zeta Technology, Inc. (Arkal Filtration System)
2. Stormwater Management Inc. (StormGate, StormFilter, StormScreen, and Catch Basin StormFilter)
3. Aquashield, Inc.(Aqua-Swirl Concentrator and Aqua-Filter)
Arkal Filtration System
The Arkal Filtration System manufactured by Zeta Technology, Inc. is a pressurized stormwater filtration
system that was tested at St. Mary's Hos
2002 after fifteen events were captured.
system that was tested at St. Mary's Hospital in Green Bay, WI. Testing was completed September 17th,
This system consists of two filtration systems. The first filtration process consists of four "towers" of
commercial disk filters, each disk filter containing a set of grooved rings. The size of the grooves
determines the particle size that will be removed from the stormwater down to a 25-micron minimum size.
Disk size for testing purposes was set up with 50-micron rings. Automatic backwash occurs when the
pressure differential across the filter rings exceeds a pre-set level. The redundant system allows for
simultaneous filtration with three towers, while the fourth tower is in a backwash mode. This allows for
uninterrupted filtration. The backwash water is temporarily stored in a backwash tank and then discharged
to a sanitary sewer at the end of the runoff period. The filtered stormwater is sent to a second filtration
stage.
This second stage consists of a series of five sealed sand filter tanks that receive the water filtered from the
disk filters through a manifold distribution system. The sand filter tanks have an automatic backwash cycle
when the pressure differential across the sand filter exceeds a pre-set level. Like the first filtration system,
this second system is also redundant. The tanks are sealed to maintain a pressurized flow system. Overflow
from the back wash tank discharges back into the holding tank, and at the end of a runoff event, the
backwash tank is discharged to a sanitary sewer. The filtered stormwater is discharged to the storm sewer
system.
158
-------
The sand filter is designed to remove 90% of particles greater than a 5-micron size. Because of this specific
claim, particle size analysis was performed in addition to suspended solids analysis. Sample locations
included the influent and effluent and the by-pass, which occurs during larger runoff events.
Other pollutant constituents were selected in addition to manufacturers' claims. These were selected to give
watershed managers information to solve water quality problems in their area. These include but are not
limited to all the parameters included in the ETV Stormwater Protocol. Additional parameters include COD
and a nutrient series.
The field-testing organizations involved at this site included Earth Tech Inc., U.S. Geological Survey
(USGS), and the Wisconsin DNR.
Stormwater Management, Inc. System
Stormwater Management, Inc. (SMI, Inc.) has a system being tested under the ETV program in Griffin, Ga.
with Integrated Science and Engineering, Inc. as the Field Testing Organization (FTO). This system
consists of a StormGate, a StormFilter, and a StormScreen.
StormGate
A diversion baffle or hydraulic transistor called "the StormGate" by SMI, Inc. is incorporated into this
system. It is designed to divert a certain amount of flow to either the StormFilter or the StormScreen, the
other two components of the SMI, Inc. system. Stormwater on the east side of Fifth Street at the test site
will flow through a StormGate to divert 10 cfs to the StormScreen device. The StormGate will divert any
flows exceeding 10 cfs. The StormGate located on the west side of Fifth Street will divert 0.79 cfs to the
StormFilter. Flows exceeding 0.79cfs will be diverted back to the storm drain line.
StormFilter
The StormFilter portion of the system is composed of filter cartridges housed in a steel vault at a St. Clair
Shores, MI ETV test site. This system uses perlite filter media in the filter cartridges. The filter systems are
installed inline with the storm drain lines. The system works by percolating Stormwater through the perlite
filter media. This filter media is designed to trap particulates and adsorb materials such as suspended solids,
petroleum hydrocarbons, and paniculate bound removal such as paniculate bound phosphorus, nitrogen, and
metals.
The typical unit configuration consists of an inlet bay, flow spreader, cartridge bay, an overflow baffle and
outlet bay. The outlet bay serves as a grit chamber and provides for flow transition into the cartridge bay.
The flow spreader provides for the trapping of floatables, oils and surface scum. Water enters the cartridge
bay through the flow spreader and starts to pond. When the water ponds, it infiltrates through the filtration
media and into the center tube, and begins to raise the float. Once the ponding submerges the cartridges, the
float will pull loose from the lower float seal and generate a siphon effect, which greatly increases the flow
potential across the filter media. The siphon effect continues until the water is drawn down to the scrubbing
regulator portion of the hood, at which time air bubbles are entrained and the siphon is lost. As the bubbles
are entrained across the surface of the cartridge, scouring of the solids deposited on the outer screen of the
159
-------
filter occurs, which acts as a self-cleaning mechanism. Water will continue to drain gravitationally until the
float reseats itself and resets the system.
The anticipated removal efficiencies of the StormFilter are between 50 to 70% of TSS, 40 to 45% of Total
Phosphorus, and little to no change in Dissolved Phosphorus. Also anticipated are 30% removal of Total
Kjeldhal Nitrogen, 40% removal of Total Zinc, and 20 to 40% removal of Dissolved Zinc and Dissolved
Copper. All parameters listed in the Stormwater protocol will be tested for in the influent to the device and
the effluent from the device.
StormScreen
The StormScreen portion of the system is a device that incorporates screening technology with patented,
self-cleaning, siphon-actuated, radial flow cartridges. This system is designed to treat high flow rates
through fine screening of the influent, and is intended to target trash and debris and larger suspended solids.
The system configuration consists of 20 cartridges which are activated by buoyant forces lifting an internal
float and opening the lower float seal that draws polluted influent via a siphon, ensuring a constant operating
flow rate as well as even flow distribution over the entire cartridge surface. Polluted stormwater is treated
by settling as water enters the vault and by being drawn through the small openings of the StormScreen
cartridges.
This system was installed in Griffin, GA in August of 2002 with ISE, Inc. as the FTO.
Catch Basin StormFilter
The Catch Basin StormFilter is manufactured by Stormwater Management, Inc. (SMI, Inc.), and is a
passive, flow-through stormwater filtration system. It is engineered to replace the standard catch basin, and
consists of a concrete or steel vault that houses rechargeable cartridges filled with a variety of filtration
media. In the Catch Basin StormFilter, polluted runoff enters the system through a traffic-bearing grate into
the primary settling chamber where heavier solids drop to a sump. The runoff water containing the lighter
solids and dissolved pollutants is then directed under a baffle into the cartridge chamber where the
StormFilter cartridges are housed. The StormFilter works by passing this water through the media-filled
cartridges, which are intended to trap particulates and adsorb pollutants such as dissolved metals, nutrients,
and hydrocarbons. This catch basin device can be customized to site-specific conditions by using different
filter media to remove the desired levels of sediments, soluble phosphorus, nitrates, soluble metals, and oil
and grease.
A Catch Basin StormFilter unit designed using CSF® leaf media is being tested under the ETV program.
To create this media, Stormwater Management composts leaves into mature stable humus. This humus is
then processed into organic granular media created used to remove TSS, oil and grease, and soluble media.
CSF (Compost Stormwater Filter), a registered trademark type of media from SMI, Inc., that is a specific
gradation of media. It is a level of media retained by a certain sieve size.
The manufacturer states that there are three primary pollutant removal mechanisms performed by the media:
1. Mechanical filtration to remove sediments and associated total phosphorus
2. Chemical processes to remove soluble metals including lead, copper, and zinc
3. Adsorption processes to remove oil and grease
160
-------
This Catch Basin StormFilter comprised of four cartridges housed in a steel vault was installed in August at
a site in St. Clair Shores, MI. Environmental Consulting and Technology, Inc. (ECT, INC.), the selected
FTO, will evaluate this unit.
The performance claims from SMI, Inc. literature indicate that suspended solids removal during testing may
reach 95%, depending on particle size distribution and influent concentration. Heavy metals removal rates
from 65% to 95% may also be anticipated due to the cation exchange mechanism provided by the humic
substances in the CSF leaf media. The high organic content of this CSF media facilitates removal of oil and
grease as well as some other organic compounds. The system is optimized for oil and grease removal when
loadings are less than 25mg/l. Under these conditions, removal rates may be expected to reach 85%.
Aqua-Filter Stormwater System
The final vendor that has applied for ETV verification of a filtration device is Aquashield, Inc. Their
filtration device submitted for verification is known as the "Aqua-Filter Stormwater Filtration System." It is
an in-line Stormwater filtration system capable of treating large flow rates. Each Aqua-Filter system is
custom engineered for the site and utilizes a unique "treatment train" approach which includes a Swirl
Concentrator designed for pre-treatment followed by a filtration chamber designed to remove fine
sediments, water-borne hydrocarbons, and nutrients such as phosphorus and nitrogen. The Swirl
Concentrator portion of the system is a hydrodynamic separator designed to remove TSS (coarse/fine
sediment) and free floating oil and debris.
The filtration chamber that follows the Swirl Concentrator in the treatment train contains a cellulose filter
media designed for polishing of the Stormwater before discharge. There are no moving parts in the system.
The manufacturer claims that previous test results indicate a 90-95% removal rate of dissolved petroleum
and oils. The patented filter media changes from tan to black when it needs to be removed. High Density
Polypropylene is used in lieu of concrete, making the Aqua-Filter System relatively lightweight and
chemically resistant.
Field-testing of this unit under the ETV program has not been initiated to date.
Hydrodynamic Separators
A second classification of Stormwater treatment devices is generally referred to as "hydrodynamic
separators." Basically, a hydrodynamic separator is some type of cylindrical vessel in which a flow stream
is introduced tangentially to induce a swirling flow pattern. This causes settleable solids to be accumulated
and stored in a manner and a location that will prevent re-suspension of previously captured particulates.
There are five vendors that have applied whose operating principles fit this hydrodynamic separation
classification. These include: Baysaver, Inc. with the Baysaver, Practical Best Management (PBM) with the
Crystal Stream Oil/Grit Separator, Vortechnics, Inc. with the Vortechs System, CDS Technologies, Inc.
with the Continuous Deflection Separator (CDS) device, and Hydro International with the Downstream
Defender.
161
-------
Baysaver
The Baysaver Separation System is designed for use as an in-line separation system for the removal of
sediments and floatable particles. Separation within the unit occurs as a result of density differences
between stormwater and materials being carried by the stormwater. Materials with a specific gravity greater
than one are removed as a result of sedimentation, while materials with a specific gravity less than one are
removed by floatation. Molecules such as hydrocarbons adsorb to particles that separate out in both the
primary and storage manholes. Flow through the BaySaver unit is controlled by the use of a trapezoidal
weir that allows the Baysaver Separation System to dictate the volume of water being treated in the storage
manhole.
The Baysaver Separation system is comprised of two precast manholes and a High Density Polyethylene
Baysaver Separator Unit. The primary manhole is set in-line with the storm drainpipe, and the storage
manhole is offset to either side. According to the manufacturer, the two manholes, which must be
watertight, provide the retention time and storage capacity necessary to remove the target pollutants from
the influent water. The Baysaver Separator Unit is designed to act as a flow control, diverting the influent
water to the flow path that will result in the most efficient pollutant removal.
The primary manhole is designed to remove coarse sediments from the influent water and retain them in an
eight-foot deep sump. A portion of the influent flow is skimmed from the surface of the primary manhole
by the Baysaver Separator Unit and conveyed to the storage manhole. This water enters the off-line storage
manhole at an elevation below the water surface and above the floor of the structure, allowing both flotation
and sedimentation to occur. The fine sediments and floatables that are entrained in this water remain
retained in the manhole.
The Baysaver Separator Unit is designed to limit the flow through the storage manhole by allowing excess
water to pass directly from the primary manhole to the outfall. During high intensity storms, the Baysaver
Separator Unit Draws water from the center of the primary manhole, approximately four feet below the
water surface, and discharges it to the outfall. Simultaneously, it continues to skim the surface water and
treat it through the storage manhole. Extremely high flows are conveyed by the separator unit to the bypass,
and bypass the storage manhole completely.
The storage manhole is designed to store oils, fine sediments, and floatables off-line; the internal bypass is
designed to minimize the risk of resuspension and discharge of contaminants. The system is also designed
to minimize the volume of water that must be removed during routine maintenance, resulting in lower
disposal fee.
Baysaver, Inc. reported that their Baysaver Separation System will provide a net removal efficiency ranging
between 60 to 80% removal of Total Suspended Solids and will also remove a significant portion of free oils
that enter the system.
The Baysaver, Inc. System was installed at a site in Griffin, GA in August of 2002, and testing is on-going.
162
-------
Crystal Stream Oil/Grit Separator
Practical Best Management (PBM) of Georgia, Inc manufactures the Crystal Stream Oil/Grit Separator. It is
a limited space BMP device that utilizes settlement as the primary constituent removal method; as velocity
slows, sediment and grit carried by the stormwater collect in the bottom of the device. It contains a separate
oil chamber designed such that motor oils and other fluids that float on water are skimmed and captured in
this reservoir for recovery. A trash rack on the top of the device is intended to capture Styrofoam cups and
cigarette butts. The unit is purported by PBM, Inc. to capture over 99% of petroleum products and nearly
95% of silt and grit, also entraining many chemicals and heavy metals.
This device was installed and is in the process of being tested at a site in Griffin, GAISE, Inc. is serving as
the FTO.
Vortechs System
Vortechnics, Inc manufactures the Vortechs System. It is a design that combines swirl-concentrator and
flow -control technologies to ensure effective capture of sediment and oils, and prevent resuspension of
trapped pollutants even at flow rates up to 25 cfs.
The Vortechs System consists of a Grit Chamber, an Oil Chamber and Baffle Wall, and Flow Control
Chamber. In the grit chamber, a swirling motion created by the tangential inlet directs settleable solids
toward the center of the chamber. Sediment is captured in the flow path and settles back into the chamber
after a storm event is over. The Oil Chamber has a center baffle that is designed to trap floatables in the oil
chamber even during cleanout. In the flow control chamber, the weir and orifice flow controls raise the
level and volume in the system as the flow rate increases, and gradually drains the system as the flow rate
subsides.
The Vortechs System is being tested at a site in Milwaukee, WI. EarthTech, Inc. in conjunction with the WI
DNR and USGS is serving as the FTO.
Downstream Defender
Hydrolnternational manufactures the Downstream Defender. The Downstream Defender is a dynamic
separator designed to remove floatables, sediment and free oil from stormwater runoff. Raw liquid is
introduced tangentially into the side of the of the cylinder and spirals down the perimeter allowing heavier
particles to settle out by gravity and the drag forces on the wall and base of the vessel.
The base of the unit is at a 30 Degree angle. As the flow rotates about the vertical axis, solids are directed at
the base of the facility where they are stored in the collection facility. The internal components are designed
to direct the main flow away from the perimeter and back up the middle of the vessel as a narrower spiraling
column rotating at a slower velocity than the outer downward flow. A dip plate is suspended from the
underside of a component support frame. The dip plate locates [better word?] the shear zone and establishes
a zone between it and the outer wall for floatables, oil, and grease. According to the manufacturer, the flow
that reaches the top of the vessel should be virtually free of solids and is discharged through the outlet pipe.
A sump vac procedure is used to remove floatables and solids.
Testing has not begun to date on the Downstream Defender; the test site has yet to be determined.
163
-------
CDS Technology
CDS Technology markets the CDS device that is designed to divert the portion of the stormwater containing
the majority pollutants (i.e. first flush) into the screen chamber. This water is treated and then returned to
the stormwater system. Flows in excess of the CDS treatment flow bypass the screen chamber. Captured
solids are permanently retained within the CDS screen and sump. Floating solids are kept in continuous
motion on the water surface while heavier materials go into the sump. CDS units use a continuously
cleaning screen. The screen is designed to remove neutrally buoyant particles that are captured by typical
baffled systems.
A test site for the CDS unit is yet to be determined.
In-drain Filtration Systems
In-drain filtration systems are catch basin inserts designed to remove various pollutants by means of some
type of filtration media. There are five different catch-basin inserts that we are verifying in the ETV
program. These are the Ultra Urban Filter with Smart Sponge from AbTech Industries, Inc., the Ultra-Drain
Guard Oil and Sediment Plus from UltraTech International, Inc., the Hydro-Kleen™ Filtration System from
Hydrocompliance Management, Inc., Drain Pac from DrainWorks, Inc., and the Flo-Gard Plus
manufactured by Kristar Enterprises, Inc.
UltraUrban Filter
AbTech Industries, Inc. manufactures this BMP Device. The Ultra Urban Filter with Smart Sponge is an in-
drain insert designed to remove sediment, hydrocarbons, and debris from stormwater. The Ultra Urban
Filter Series DI2020 is made of high strength corrugated plastic designed to "drop-in" existing stormwater
catch basins. It is used in storm drains that experience oil and grease pollution accompanied by sediment
and debris.
The filter is designed such that trash and sediment accumulate in the internal basket while oil and grease are
captured in the filtration media. According to the manufacturer, oil is bonded with the SmartSponge so that
it will not leach back into the environment.
It was installed in August of 2002 at a test site in Griffin, GA and is being evaluated by ISE, Inc. as the FTO.
Ultra-Drain Guard Oil and Sediment Plus
UltraTech International, Inc. manufactures this "Catch Basin Insert" device. It is designed to capture oil,
grease, trash, and sediment from stormwater runoff before it enters the storm drain system. It is installed in a
catch basin and is suspended by the grate itself. Stormwater runoff enters the Ultra-Drain Guard Oil and
Sediment Plus and is directed toward the pouch by a skirt made of a nom-woven [?] polypropylene, needle-
punched, geotextile material. The fabric itself is designed to filter pollutants as the runoff passes over and
flows through the material. In addition, each Ultra-Drain Guard Oil and Sediment Plus is equipped with
several "filter strips" made of "X-Tex," a unique filter material made of recycled synthetic fibers. The
manufacturer claims that this material is extremely effective in the capture and removal of hydrocarbons and
other pollutants from stormwater. These filter strips are intended to maximize oil and hydrocarbon removal.
164
-------
The Ultra-Drain Guard Oil and Sediment Plus are designed with "ByPass Ports" to prevent flooding and
ponding from occurring. One unit is said to be capable of filtering out and containing a minimum of forty
pounds of oil, sediment, debris, and floatables.
The Ultra-Drain Guard Oil and Sediment Plus was installed in a site in Griffin, GA. in August of 2002. ISE,
Inc. is serving as the FTO, and testing is ongoing.
HydroKleen
Hydro Compliance Management, Inc., of Ann Arbor, Michigan (Hydro Compliance), manufactures and
markets the Hydro-Kleen™ Filtration System. The Hydro-Kleen™ is a stormwater catch basin insert
designed to trap hydrocarbons, metals, sediments, and other contaminants contained in stormwater and other
surface runoff. The Hydro-Kleen™ contains a multi-chamber system that combines pre-settling sediment
removal with dual media filtration. The system is designed to filter hydrocarbons and other contaminants
while alleviating concerns with water flow. The Hydro-Kleen™ Filtration System is promoted as a
structural BMP to assist end users in complying with meeting NPDES Phase n stormwater compliance
permit and other regulatory requirements for protecting surface water runoff quality.
The Hydro-Kleen™ Filtration System is a patented multi-media filtration design combined with pre-settling
sedimentation containment and overflow by-pass protection for 'hot spot' applications. Each unit is custom
manufactured for retrofit or specification to fit a specific catch basin or drain invert size. Units are placed
into drains by removing the grate/cover, inserting the unit onto the grate lip, and replacing the cover. Water
flow enters the unit and is directed into a pre-settling sedimentation chamber that collects heavy sediments
and debris passing through the grate. Water then passes through transition inlets at the top of the sediment
chamber into the filtration chamber. The primary media, Sorb-44, is intended to remove hydrocarbons
through adsorption. The secondary media is a blend of activated carbon (AC-10) that is intended to remove
any remaining hydrocarbons, as well as a variety of other organics, metals, and other contaminants from the
runoff. Water then passes through the of the bottom treatment chamber into the catch basin.
Units are designed to trap contaminants contained in the 'first flush' from storm events while allowing
overflow protection to eliminate flooding during heavy wet weather events. To accomplish this, the
filtration chamber is designed to handle 40 - 50 gpm through the media chamber, effectively handling up to
l/2 in. of rain per hour in a properly designed drain. Higher flows from high intensity wet weather events are
diverted to by-pass outlets that are designed to move whatever flows the drain is designed to handle. This is
intended to prevent flooding or ponding on the surface while capturing contaminant loadings from
impervious surfaces.
The Hydro-Kleen System is being tested under two different protocols. Laboratory testing is being done
under the protocol for in-drain devices developed under the Source Water Protection Pilot in Ann Arbor at
NSF International. Field-testing is being conducted at a site in St. Clair Shores, MI under tie Stormwater
Source Area Treatment Device Protocol.
DrainPac
DrainWorks, Inc manufactures DrainPac. It basically consists of three types of parts: a metal support
bracket, flexible polymer support structure, and a replaceable bag filter. DrainPac is designed to trap or
collect sediment, oil and debris from drain inlets.
165
-------
A test site and testing organization has yet to be determined.
Flo-Gard Plus
Flo-Gard Plus is manufactured by Kristar Enterprises, Inc. It is a catch basin filtration system designed to
be effective in the removal of sediment, trash and debris. It features a stainless steel outer basket, a filter
liner, and an HDPE adapter ring to allow for use in a wide range of design applications. It also offers a dual
bypass feature, an initial "filtering" high flow bypass and an "ultimate" high flow bypass. In both bypass
modes, pollutants remain trapped in the system.
This device is not being tested yet, since the site has yet to be determined.
Summary
This is a snapshot of the Stormwater Technology Area of the ETV Program, as it exists in September of
2002. Twelve vendors have applied for verification with thirteen different devices submitted for verification
testing. Testing of the Arkal Filtration System has been recently completed in Green Bay, WI.
Testing is on-going for the Vortechs System in Milwaukee, WI under the direction of EarthTech, Inc. as the
FTO, and in conjunction with the Wisconsin DNR and US Geological Survey (USGS). Testing is also
underway for the Hydro-Kleen Filtration System and the Catch Basin StormFilter from SMI, Inc. in St.
Clair Shores, MI with ECT, Inc. as the FTO. In Griffin, GA, with ISE, Inc. as the FTO, verification testing
is on-going for the Crystal Stream Oil/Grit Separator from PBM of GA, the StormGate, StormFilter, and
StormScreen from the Stormwater Management Inc. (SMI), and the Baysaver Separation System from
Baysaver, Inc. Ultra-Urban Filter from AbTech Industries, and the Ultra-Drain Guard Oil and Sediment
Plus unit from UltraTech International, Inc. are also being tested in Griffin with ISE, Inc. as the FTO. Five
devices that have not begun testing include: the FloGuard Plus from Kristar, Aqua-Filter Stormwater
System from Aquashield, the CDS Device from CDS Technologies, the DrainPac from DrainWorks, and the
Downstream Defender from Hydrolnternational, Inc.
As mentioned, our protocol is constantly evolving as test plans are developed and finalized. A current copy
of the protocol can be found either on the EPA or NSF ETV web sites, http://www.nsf.org/etv and
http ://www. epa.gov/etv. Also, verification results in the form of Verification Reports and Statements for the
testing that has been completed to date can be found on these web sites.
166
-------
USING AN INDICATORS DATABASE TO MEASURE STORMWATER
PROGRAM EFFECTIVENESS IN HAMPTON ROADS
Julia B. Hillegass
Hampton Roads Planning District Commission
Chesapeake, Virginia
Abstract
The Hampton Roads Planning District Commission (HRPDC) has been working with the region's
sixteen localities to develop a regional stormwater management program since 1996. The program focuses
on activities that support the permit compliance efforts of the six communities with Virgina Pollutant
Discharge Elimination System (VPDES) Stormwater System Permits, technical assitance to the region's
non-permitted communities and regional education and training to support all of the communities. A set of
regional stormwater management goals that guide the regional program has been developed. Adopted by
the HRPDC, they are:
• Manage stormwater quantity and quality to the maximum extent practicable (MEP)
—Implement Best Management Practices (BMPs) and retrofit flood control projects
to provide water quality benefits.
—Support site planning and plan review activities.
—Manage pesticide, herbicide and fertilizer applications.
• Implement public information activities to increase citizen awareness and support for the
program.
• Meet the following needs of citizens:
—Address flooding and drainage problems.
—Maintain the stormwater infrastructure.
—Protect waterways.
—Provide the appropriate funding for the program.
• Implement cost-effective and flexible program components.
• Satisfy VPDES stormwater permit requirements:
—Enhance erosion and sedimentation control.
—Manage illicit discharges, spill response and remediation.
The Regional Stormwater Management Committee determined that a major technical study should
be undertaken cooperatively to support the stormwater programs of the six permitted localities and should
include the following components:
1. Analyze stormwater discharge sampling data to develop event mean concentrations (EMC)
by city and by land use.
2. Develop stormwater pollutant loads for watersheds in the six cities based on the EMC using a
geographic information system.
3. Develop a consolidated regional monitoring program for the six cities for consideration by
the Department of Environmental Quality (DEQ) in the VPDES stormwater permit
reapplication process. Develop recommendations on indicators of stormwater management
program effectiveness.
167
-------
The Regional Loading Study recommended the use of a series of Program Effectiveness Indicators,
rather than continued traditional chemical water quality monitoring. The HRPDC staff developed a
proposed modification to the monitoring component of each locality's municipal separate storm sewer
system (MS4) Permit, outlining the Regional Stormwater Management Program Goals that are to be met
through the local stormwater programs and how the Indicators would be used to measure progress toward
those goals. Ten indicators were developed to measure the overall success of local programs. The proposed
Permit Modification was submitted by each of the permitted localities and was incorporated by DEQ into
the reissued VPDES Stormwater Permits.
Background
During their first separate storm sewer system (MS4) Virginia Pollutant Discharge Elimination System
(VPDES) permit term, the Cities of Chesapeake, Hampton, Newport News, Norfolk, Portsmouth, and
Virginia Beach were required to monitor the chemical constituents from selected outfalls. Based on the
collected monitoring data, the local governments were required to calculate Event Mean Concentrations
(EMCs) of pollutants discharged from their monitored stormwater outfalls. A study was commissioned by
the affected local governments to determine the efficacy of this method of monitoring. A map of the study
area with major watersheds is included as Figure 1. The consultant on the project was charged with the
following:
1. Analyze stormwater discharge sampling data to develop event mean concentrations (EMC)
by city and by land use.
2. Develop stormwater pollutant loads for watersheds in the six cities based on the EMC using a
geographic information system.
3. Develop a consolidated regional monitoring program for the six cities for consideration by
DEQ in the VPDES stormwater permit reapplication process. Develop recommendations on
indicators of stormwater management program effectiveness.
168
-------
(Q
£
3
ET
Q.
>
-------
Process and Objectives
The process for developing the regional stormwater program and effectiveness indicators is shown in Figure
2 and is described below:
• The consultant conducted a literature search of regional monitoring programs and alternative
program effectiveness indicators.
• The consultant facilitated discussion of the development of regionally consistent stormwater
monitoring program goals, prioritizing potential indicators to be used in a regional program,
either to complement or replace the required chemical monitoring under the then existing
VPDES permits. The goal setting and prioritization was conducted over a series of
workshops from October 1998 to February 1999.
• The consultant performed an analysis of existing VPDES permit data to determine:
a Whether chemical monitoring can be replaced by other effectiveness indicators, by
comparing local data to the Nationwide Urban Runoff Program (NURP) data.
a If monitoring cannot be replaced, determine whether monitoring sites and land use
types can be consolidated based on representative data across cities and land use as
compared with NURP data.
An important objective of the new program was to effectively communicate the successes of the municipal
stormwater programs to the public and elected officials, with greater emphasis on social and programmatic
indicators. A second objective was to develop a more cost-effective approach to stormwater monitoring in
the Hampton Roads region that will both satisfy the permit requirements and measure the effectiveness of
local stormwater programs.
170
-------
Identify Program Goals
and Components
Identify Indicator Tools
Integrate Program
Goals and Tools
I
Stakeholder Input
Program
Recommendations
Figure 2: The Process
When compared to EMCs from other urban areas studied during the Nationwide Urban Runoff Program
(NURP), calculations indicated that the level of pollutants carried by stormwater in Hampton Roads is
typical of other urban areas and, in many cases, lower.
The Stormwater Management Program Effectiveness Indicator Tracking Program was developed to help the
region's local governments assess their achievement of common stormwater management goals developed
by the Hampton Roads Regional Stormwater Management Program. These goals are:
• Manage stormwater quantity and quality to the maximum extent practicable (MEP).
a Implement BMPs and retrofit flood control projects to provide water quality benefits
a Support site planning and plan review activities.
a Manage pesticide, herbicide, and fertilizer applications.
• Implement public information activities to increase citizen awareness and support for the program.
• Meet the following needs of citizens:
a Address flooding and drainage problems.
a Maintain stormwater infrastructure.
a Protect waterways.
a Provide appropriate funding for the program.
• Implement cost-effective and flexible program components.
• Satisfy VPDES stormwater permit requirements.
a Enhance erosion and sedimentation control.
a Manage illicit discharges, spill response, and remediation.
171
-------
The Indicators Program
A variety of program effectiveness indicators were selected during the series of workshops. These
indicators encompass all aspects of local stormwater programs in Hampton Roads and were selected based
upon technical, practical and programmatic considerations. To capture data representative of the activities
in stormwater programs, the indicators were divided into strategic indicator groups. An indicator was
defined as a measurable feature that provides managerially and scientifically useful evidence of stormwater
and ecosystem quality or reliable evidence of trends in stormwater quality and program effectiveness. The
Tracking Program stores the indicator data in a Microsoft Access database. The indicators that are recorded
in the database can be grouped into one of four categories as illustrated in Table 1 below:
Table 1: Database Indicators
Indicator Group Indicator
Water Quality Pollutant Loadings
Physical & Hydrological Greenlands Program
Programmatic Investigative Monitoring
BMP Implementation
Flooding and Drainage Control
Flooding and Drainage Projects
Erosion and Sediment Control
Permitting and Compliance
Operations and Maintenance
Socioeconomic Public Information Programs
Environmental Knowledge
Website visits
Publications Distributed
Media
Restoration Activities
Cleanup Activities
While the chemical monitoring program was useful in determining that the stormwater runoff in Hampton
Roads is comparable to other urban areas, it was not useful in communicating the effectiveness of local
stormwater management programs. The high variability of the data, due to natural factors such as rainfall,
makes it very difficult to detect any actual increasing or decreasing trends in pollutant levels carried by
stormwater runoff. In addition, the chemical monitoring program could not account for actions taken by
local stormwater programs to reduce flooding and drainage problems. Due to these shortcomings, the
permitted local governments of Hampton Roads proposed modifying their MS4 VPDES permits to replace
the chemical monitoring requirement with a Stormwater Management Program Effectiveness Indicator
Tracking Program for the second permit term. Initial data collection began in 2000 to provide examples of
the types of data that would be collected in future years, should the Tracking Program be allowed in the
permit renewal process. Data can be queried and illustrated by locality and regionally, in the form of
172
-------
summary tables and graphs. The Virginia Department of Environmental Quality accepted the proposed
Tracking Program in lieu of chemical monitoring and modified the MS4 VPDES permits accordingly when
they were reissued in April 2001.
Description of Indicators
Water Quality Nutrient Loadings
CH2MH111 estimated Stormwater pollutant loads for each of the local governments in Hampton Roads
permitted through the Virginia Pollutant Discharge Elimination System Program. The estimated pollutant
loads are documented in a series of Technical Memoranda contained in each locality's annual report.
Greenlands
Greenlands are lands that are permanently protected from development or lands that are restored to a more
natural state during redevelopment. They provide a water quality benefit by reducing the imperviousness of
the watershed. Such lands may include parklands, refuges, wetlands, and lands protected by conservation
easement. The database is structured to maintain the number of acres of greenlands to assess progress
toward reducing the potential watershed imperviousness and nonpoint source pollution loads.
BMP Implementation
Stormwater best management practices (BMPs) help to minimize flooding and water quality impacts
associated with development. Experience has shown that over time, lack of maintenance has caused BMPs
to lose their effectiveness. In addition, older developed areas lack BMPs or the designs of the BMPs that
have been installed do not include water quality protection measures. To measure the success of BMPs in
flood and water quality protection, the database is structured to include information on:
• The number and types of BMPs installed or retrofitted for water quality
• The number of developed acres served by BMPs, grouped by land use
• Inspection and maintenance activities
This information will eventually allow the estimation of pollutant removal by BMPs and the ascertainment
of whether BMPs are functioning properly.
Erosion and Sediment Control
Every local government in the Commonwealth of Virginia is required to administer an Erosion and
Sediment Control Program. The Erosion and Sediment Control Law requires that land disturbing activities
exceeding 10,000 square feet submit an Erosion and Sediment Control Plan and meet minimum standards.
Under the Chesapeake Bay Preservation Act, the threshold is decreased to 2,500 square feet in a Chesapeake
Bay Preservation Area. The minimum standards specify practices that reduce the amount of sediment
leaving a construction site and minimize downstream flooding and streambank erosion. The level of
enforcement and compliance limits the effectiveness of local erosion and sediment control programs. To
monitor the extent of land-disturbing activities, the database is designed to include information on the
number of approved erosion and sediment control plans and disturbed acreage. The number of inspections
and enforcement actions are also included to evaluate enforcement and the level of compliance with the
local erosion and sediment control regulations.
Flooding and Drainage Responses
Calls and complaints received from citizens can be an indicator of the performance of a Stormwater
program. Responsiveness of a Stormwater program, in the form of inspections and resulting maintenance
173
-------
activities, to citizen inquiries can also be an indicator of effective administration of the stormwater program.
The database is structured to collect data on the number of citizen calls and responses.
Flooding and Drainage Projects
An important function of a local stormwater program is to correct flooding and water quality problems.
Projects to address these needs may be included in local Capital Improvement Projects. Corrective actions
may involve retrofitting areas, installing BMPs, or restoration activities. To help determine whether a
stormwater program is actively performing this important function, the database is designed to include the
number and cost of flooding and drainage projects.
Investigative Monitoring
Hazardous material spills, wastewater cross connections, and other illicit discharges can represent a
significant source of pollution. Implementing an effective illicit discharge/connection management program
to control these sources can result in considerable improvements to water quality. The database is
structured to allow the collection of information on investigative and corrective actions, to assess whether an
illicit discharge/connection program is being effectively implemented. These actions include screening
inspections and measures taken to locate and eliminate illicit discharges/connections.
Operations and Maintenance
Operation and maintenance activities are crucial to a stormwater conveyance system's ability to reduce
flooding and minimize the amount of pollutants that are discharged into the region's waterways. Operation
and maintenance activities include street sweeping and cleaning and repairing both catch basins and
drainage facilities. By monitoring these activities, the proper functioning of the stormwater system can be
assessed, and the amount of sediment that was prevented from being discharged by the stormwater system
can be estimated.
Permitting and Compliance
Development increases the amount of runoff and pollution in a watershed. In an effort to monitor
development activity, the number of approved site and subdivision plans, and their associated developed or
redeveloped acres are maintained in the database.
Public Information Programs
Informing individuals about stormwater issues and measures they can take to reduce pollution is important
to gaining public support of a stormwater program. It also helps protect water quality. The database
maintains information on public education and outreach activities to help assess whether a stormwater
program is adequately carrying out this function. The parameters that are examined include: number of
publications produced and distributed, public outreach activities, media campaigns, riparian restoration
activities by citizens, stream cleanup activities, and web site hits. Where appropriate, citizens are surveyed
regarding their knowledge levels before and after an informational effort.
174
-------
The Database
The Main Menu
The database opens up to the Main Menu with several selection options. The upper portion of the menu lists
each of the effectiveness indicators. When an indicator is selected, a data entry form for that particular
indicator is displayed.
The bottom portion of the menu consists of administrative functions. The "Edit Lookup Tables" button
opens a form that allows the input of additional Activity Types, BMP Types, Green Areas, Municipalities,
Pollutants, Spot Types, Topics and Watersheds. The "Import/Export Data" button opens a form that will
allow each of the indicators to be exported in a text or Excel format, as well as import an indicator that has
already been exported in a text format by using this tool. The Main Menu is illustrated in Figure 3.
File Edit View Insert Format Records lools Window Help
s / N r i ,,", i .'i' U*'* ••:"{'*!> i ''.'
/ I » -;,.,,;'.•>__ ,'-•.,../","•'.< >•.•
_f~^—^^gfj^.^-. gr:;; •;,;-_ :v v^r^'^"" r ,;"T
j
Regional Stormwater Prosram
IIIIIIIIIH^
I Pollutant Concentrations
Greenlands Program
Investigative Monitoring
BMP Implementation
Flooding / Drainage Problems
| Permitting and Complian
View Watershed Map
Exit the Database
INUM i
Figure 3: Database Main Menu
Indicator Tools Menus
Data entry forms are set up for each indicator to facilitate the data-gathering task. Few of the permitted
localities have all of the tracked information in one department. The Tracking Program allows data entry to
be conducted by several departments, compiled by the respective locality, and then compiled for the region.
Many localities are able to use the data gathered in reporting on other related program efforts such as
Erosion and Sediment Control and the Chesapeake Bay Preservation Act.
Features unique to the Tracking Program include the ability to query for reporting by region, watershed or
locality. Data can also be entered in the datasheet view, which allows for full functionality of all of the
175
-------
associated pull-down menus. The Tracking Program also allows for different time intervals of data
collection, such as monthly, quarterly or annually, ensuring flexibility for the different local programs.
Localities can also customize specific reporting areas to more accurately capture local program efforts by
utilizing the Edit Lookup Tables function of the database. Existing lookup values can be added, deleted or
modified based on local program needs.
An Import/Export Data function allows electronic compilation and transfer of data between and among local
departments, as well as to and from the HRPDC staff. The data can be exported and manipulated in Excel
or exported to text to send a final version. Filenames are automatically assigned by concatenating the
municipality with the table name and current date. When importing data, automatic integrity checks will be
activated which prevent duplicate reporting, while allowing the user to upload the remaining records.
Sample Reports
Figure 4 and Figure 5 show examples of reports for Pollutant Concentrations (EMCs) and Pollutant Loading
data.
Virginia Beach
Total Phosphorus (TP) (mg.'L)
ling Location
Sam pllng
Figure 4: Pollutant Concentration Data for Virginia Beach
176
-------
Total Phosphorus (TP) Load (310,616 Ib/yr)
Distribution Between Major Watersheds in Study Area
For the Yean 1999
Norti Landing Rlwr
13%
raver / Llttts Crss h
13%
ch ra w r i1 P:i quo i on ra '.'* r
Ui in al Swam p
4%
Blabs in Rwr i'h6mpitun Roadi
Figure 5: Total Phosphorus Load Distribution by Major Study Area Watersheds
177
-------
A variety of reports can be generated from the myriad of data collected. Data can be sorted by locality,
watershed, activity type, watershed within a specific locality, or summarized for the entire Hampton Roads
region. Some examples of those tables and charts follow:
Acres of Greenlands in Hampton Roads
mnnnn
1 UUUUU
QOOOO
snnnn
ouuuu
(/>
o> ynnnn
o
< finnnn
M_
o c;nnnn
^ ouuuu
a)
n /innnn
3 Qnnnn
onnnn
mnnn
1 UUUU
89163.42
16053
8654.4
828
5979 „_.
2729
I
D Chesapeake
• Hampton
D Newport News
n Norfolk
• Portsmouth
n Va Beach
Acres by Locality
Figure 6: Acres of Greenland Areas in Hampton Roads
Flooding & Drainage Responses
FY 00-01
FY01-02
Locality
Figure 7: Flooding and Drainage Responses by Fiscal Year per Locality
178
-------
Table 2: Miles of Drainage Facilities
Serviced
Table 3: Street Sweeping Miles and Tons
Recovered
Miles of Drainage Facilities Serviced
Chesapeake
Hampton
Newport News
Norfolk
Portsmouth
Va Beach
FY 00-01
933
405
13880
11.14
109
9
FY01-02
97.41
325
242
199.9
504
92
Street Sweeping FY 01-02
Chesapeake
Hampton
Newport News
Norfolk
Portsmouth
Va Beach
Miles
6218.85
715
12004
50700
17073
10350
Tons
870
2663
9378
7245
653
15646
These various indicator groups, while not complete unto themselves, can together give a better indication of
the success of an overall storm water management program. The data is also helpful to local governments
in evaluating annual budgets; compiling long-term budget and program priorities for permit renewal; and
having hard data to share with citizens and elected officials. A challenge of the tracking program has been
keeping the data input consistent between and among localities, as often several staff members will be
responsible for entering various pieces of the data for their locality. The goal of the reports is not to
compare program weaknesses between localities, but rather to more effectively gauge local efforts and
spending in relation to program accomplishments.
Conclusion
Trial data was submitted to DEQ prior to formal permit renewal applications being submitted. During that
time, work sessions were also held with the committee to gauge the usefulness and efficiency of the
Indicator Tracking Program and to look at data management areas that needed enhancements or
refinements. Local government and HRPDC staff responsible for technical and educational efforts
participated in these sessions. Since inception, the tracking program has undergone several updates. This
will be the first full permitted program year for reporting the data gathered by the Tracking Program for the
Phase I communities.
In the recently enacted federal Phase II Stormwater Regulations, the U.S. Environmental Protection Agency
recognizes the shortcomings of chemical monitoring. Rather than conduct a chemical monitoring program,
Phase II communities are required to track the implementation of Stormwater management measures. These
management measures include public education and outreach, public involvement, illicit discharge detection
and elimination, construction site runoff, post-construction runoff, and pollution prevention/good
housekeeping activities. The Phase II Regulations recognize that this kind of tracking system provides a
better measure of program effectiveness than chemical monitoring of Stormwater outfalls. This is great
justification of what was proposed for Phase I communities.
The Stormwater Management Program Effectiveness Indicator Tracking Program is similar to the tracking
system required by the Phase II Stormwater Regulations. It is expected that the Stormwater Management
Program Effectiveness Indicator Tracking Program will also be used by the local governments of Hampton
179
-------
Roads affected by the Phase II Regulations to satisfy their permit requirements. This may require further
enhancement of the program to assist smaller localities with data gathering tasks.
We anticipate further update to the database, as well as a series of training sessions for local users. While
the tracking program allows the HRPDC to generate consist reports for all participating localities,
challenges remain in getting data input that is consistent between and among localities.
In addition, the basic Tracking Program has been submitted as a suggested beginning model for discussions
regarding consolidated tracking and reporting tasks that are typically required by various state agencies to
meet program requirements.
References
CH2MH111, 1999. Regional Stormwater Loading Study, Proposed Regional Monitoring Program and
Program Effectiveness Indicators.
CH2MHill, 2000. Proposed Regional Monitoring Program, Stormwater Effectiveness Indicators, Database
User's Guide Version 1.
Hampton Roads Planning District Commission, 2002. Indicators of Stormwater Effectiveness Fiscal Year
2000-2001.
180
-------
USING INCENTIVES AND OTHER ACTIONS TO
REDUCE WATERSHED IMPACTS FROM EXISTING DEVELOPMENT
Dawn Hottenroth, RS, CPESC
City of Portland Bureau of Environmental Services
www. cleanrivers- pdx. org
Portland, Oregon
Abstract
Local jurisdictions must find new ways to mitigate impacts from urban development. Urban development
creates a variety of negative impacts within watersheds. Impacts relating to the flow rate, volume and water
quality of urban stormwater runoff are varied and sometimes difficult to remediate. While most local
communities are beginning to implement post-development stormwater management requirements, many
communities struggle to address impacts from existing development. Many local communities can have 80-
90% of their land area already built out, which limits the overall effectiveness of new and redevelopment
stormwater management requirements. Local businesses and citizens can either harm or help keep local
waterways clean. They can also mitigate impacts from existing development. A combination of
educational, technical assistance, and incentive programs can be used to change the behavior of businesses
and citizens. Whether it is saving money, protecting the environment for future generations, gaining
recognition or some other motivator for change, local jurisdictions need to create a menu of programs and
incentives to gain the participation of citizens in protecting the environment. Portland, Oregon has made
great strides at limiting impacts to local watersheds through creative programs such as Downspout
Disconnection, Stewardship Grants, and Clean River Incentive and Discount Programs. These and other
programs are leading the way to addressing and hopefully minimizing negative impacts from existing urban
development.
Background
Portland, Oregon is located on the northern border of the state, at the confluence of the Columbia and
Willamette Rivers. Portland is home to 510,000 citizens in an area of approximately 130 square miles.
There are approximately 4,000 miles of street that are drained by 800 miles of combined sewer, 400 miles
of storm sewer, 129 miles of drainage ditch and over 9,000 public drainage sumps. The City of Portland,
Bureau of Environmental Services (BES) operates and maintains these storm drainage systems, two sewer
treatment plants, and implements water quality improvement / watershed health program efforts. The City
of Portland has a Phase 1 NPDES Municipal Stormwater Permit and was recognized in 1996 as the best
stormwater permit program in the nation.
Portland is located at the bottom of the Willamette River watershed - one of the few south-to-north draining
rivers in the United States. The Portland urban services boundary contains four major sub-watershed
drainage systems and a large number of smaller drainageways that discharge directly to the Willamette
River. Almost all of those drainages are listed by the State as not meeting their designated beneficial uses.
There is an EPA designated Superfund site in the Willamette channel at Portland Harbor - between river
miles 9 and 4 south of the Columbia River confluence.
Portland is home to a variety of state- and federal-listed threatened and endangered species. Perhaps the
most significant are the three species of salmonids that have been listed in the Willamette watershed over
the last three years. These fish species are directly impacted by citizen behavior and the runoff from
existing development1.
181
-------
Impervious surfaces from existing development account for approximately 33% of Portland's total land area
or just over 43 square miles of paved and other hard surfaces (see Figure No.l below for land coverage
breakdown). Of the 33% of the urban area that is impervious, 22% is paved areas that support car usage.
Pervious housing areas only account for 37% and 7% pervious industrial and commercial areas. Open
space and rural land use areas make up the remaining 23% of total land coverage2.
Urban Cover Land Uses
Commerc i a I
Large
Street
Surfaces
18%
Multi Family
Residential
Pervious
8%
Single Family
Residential Pervious
I
(Figure No. 1) City of Portland, Environmental Services GIS Zoning Layer. Information built from local
zoning ordinances and the Metropolitan Service District 2040 Urban Growth Boundary Framework Plan.
Problem
Impervious surfaces have a variety of negative impacts on local watersheds. Besides significantly altering
the natural water cycle, some of the most recognized specific impacts are:
• Decreased vegetative cover and stream shading. Damaged riparian zones provide minimal habitat and
stormwater management functions.
• Increased stormwater volume and flow rate that contributes to streambank erosion, stream
channelization, and flooding.
• Heat absorption by stormwater runoff that flows over impervious surfaces, resulting in increased surface
water temperatures.
182
-------
• Pollutant and sediment conveyance from impervious surfaces into surface water bodies, impairing water
quality, fish habitat and spawning grounds.
• Low summer stream flows from lack of infiltration into groundwater recharge areas.
Multiple studies from across the nation, endorsed by the National Marine Fisheries Service (NMFS) and
U.S. Fish and Wildlife Service (USFS), conclude that watershed degradation begins to occur when
impervious surfaces exceed 10% of the area within a drainage basin. The goal of any stormwater
management program aimed at addressing impacts from existing development should be to mitigate impacts
to those that could be expected from a 10% impervious area coverage watershed. By instituting actions to
transform a high impervious area drainage basin into a basin with only 10% effective impervious area,
watershed degradation can be kept to a minimum. Effective impervious area is a term used to describe the
portion of a site that discharges directly to a receiving system without any mitigation of impacts from
interception, filtration, infiltration or other site practices.
So how does a local jurisdiction effectively reduce impacts to those of a 10% impervious area coverage
basin? A mixture of education, technical assistance and incentive programs can make great strides to
reaching this goal.
Portland's Program
Portland has had an active watershed planning and education program in place since 1991. The first step of
any program to address existing development impacts should be education. Many local jurisdictions
already have in place foundational components to support educational programs. The City of Portland has
multiple outreach and educational programs that strive to attain the following goals:
• To educate residents and businesses of the City that they are part of a natural watershed. All
programs and outreach in the City are announced under their specific watershed areas - Johnson
Creek, Tryon Creek, Fanno Creek, the Columbia Slough or the Willamette River.
• To educate residents and business about the final destination their stormwater runoff and
sanitary drainage flows. In the City approximately one third of the urban services area discharges
stormwater to each of the following locations: the combined sewer to the treatment plant; to the
separate sewer, which mostly drains directly to local stream systems; and into underground aquifers
through public sumps and private drywells.
• To educate citizens and businesses about how their every day behaviors impact the environment
and what changes in behavior they can make to lessen those impacts. Usually programs include
tips on changing behavior and/or onsite actions that help citizens protect clean rivers. Examples
include washing cars over lawns to limit runoff of pollutants, planting trees to intercept rainfall and
limit runoff, and use of native plant specific to limit horticultural chemical use and the potential for
resulting polluted runoff.
• To create active citizenry and advocates for stormwater improvements within the City. These
advocates then take on neighborhood projects, support program implementation or help fight for
program funding. Many times stormwater program advocates come from related environmental
programs such as the Audubon Society, Sierra Club, and watershed councils.
183
-------
Portland's educational programs include a variety of standard activities such as brochures, billing inserts,
and speakers bureaus, as well as a few unique programs. Many of Portland's programs are developed and
implemented in partnership with other local agencies, such as:
Environmental Services Educational Program - The City has two staff people dedicated to presenting
programs for Portland school students K-12. Their curriculum includes education to support the goals
above through humorous and entertaining assembly programs and classroom presentations. These
educators also partner with schools to have students implement hands-on activities such as tree planting,
stormwater management facility construction, or monitoring projects on or near school grounds. This
program is funded through stormwater utility fees and reaches approximately 27,000 students every year.
Regional Coalition for Clean Rivers and Streams - This regional awareness programs strives to present
basic messages in the tri-county area in Portland. Working with nine other local stormwater agencies, the
program runs multi-media campaigns throughout the region encouraging all regional citizens not to pollute.
This program is funded through stormwater utility fees and reaches 1.4 million people a year.
Natut-escaping for Clean Rivers -This program was developed and is implemented by the City of Portland
and the East Multnomah Soil and Water Conservation District. Targeting lawn and yard water, pesticide
and fertilizer use, this program offers free workshops for local residents about the benefits and ease of using
native plants in their landscape. What is especially unique about this program is the advice of a landscape
architect in addressing specific property design questions. This program is funded through stormwater
utility fees and reaches around 400 people a year.
Most local jurisdictions know that more than education is needed to motivate people to make behavior
changes. What else is needed to motivate people to change? Primarily two things - giving citizens enough
information to know what to do and making doing the right thing easy and/or financially beneficial.
Explaining What to Do - Technical Assistance
Portland has a complex menu of options on what we want people to do to lessen their impacts on local
watersheds. Most actions fall into two broad categories - changing behaviors, like driving a car less, and
retrofitting a site for onsite stormwater management, through planting a tree or disconnecting downspouts.
Usually the behavior changes that the City promotes to lessen impacts on the local watershed also meet
objectives of other programs. For instance, driving your car less reduces the amount of oil drips, car
exhaust deposits, brake and tire wear particulates that end up on street surfaces, and are ultimately
discharged to local waterways during storms. Having fewer cars on the road can also help limit air
pollution and congestion on local roadways. Most suggested behavioral changes either limit the amount of
pollution or the total volume and/or flow rate of stormwater runoff. Water quality related actions primarily
focus on preventing or limiting pollution coming in contact with stormwater runoff. Volume control actions
focus on infiltrating stormwater onsite or otherwise mimicking the natural flow regime for the watershed
area. Because behavior changes that reduce volume or pollutants in stormwater have multiple benefits,
there are great opportunities to partner with other agency programs on these multi-objective pollution
prevention messages. BES looks to our educational programs to suggest behavior changes and make
referrals to other agency programs for specific implementation details.
Onsite stormwater management changes are a bit more complex. It can be very difficult to present solutions
in a way that can convince the average person to institute change. Many local programs simply suggest a
concept or idea but fail to provide enough implementation information to make a site retrofit possible. For
184
-------
example, the simplest stormwater retrofit programs (including Portland's) suggest that people "just plant a
tree." Seems simple enough, but what type of tree should they plant and where? Many city codes dictate
the "what" and "where" of an action. Property owners may be unaware of these regulations or have
troublesome site-specific constraints that seem to be barriers to implementing retrofits. Most citizens,
whether at their home or business, need additional help in mitigating impacts or changing behaviors.
Ideally, there would be a city staff person available to answer any request at any time and assist owners
through every step of the retrofit process. Yet, realistically, face-to-face assistance is not usually possible
due to limited staff and financial resources. So we look to surrogates - whether through detailed instruction
materials, in-depth workshops or short onsite visits.
One of the best places to look for detailed guidance on site retrofits is the new and redevelopment
stormwater facility requirements manual. Even though existing development is not likely to be required to
retrofit, they should still strive to manage stormwater to the same level as new and redeveloping properties.
In reality, specific site constraints usually limit the extent of area available for retrofits, thus limiting the
extent of onsite stormwater management.
Portland's Stormwater Management Manual (SWMM) provides a great deal of guidance on facility
selection, facility sizing, plant selection, and maintenance activities. There are a number of City programs
encouraging on-site retrofits that make great use of information in the SWMM. One particular element of
the SWMM that is especially useful is the sizing form. During the last 2 years, the SWMM has undergone
its second revision with the specific goal of making stormwater facility design as easy as possible. One
element of that effort was the creating of a sizing matrix for simple facility design. The matrix - SEVI form
(Figure No. 2) - from this new development manual can be used as a great guidance document for sizing
retrofit facilities for existing development. The SWMM is available at
www.cleanriverspdx.org/tech resources72002 swmm.htm in its entirety.
185
-------
I Form S(M: Simplified Approach for Stormwater Management
en precis Facilite^ syxd wth Sh?» fonn are prEsirned to cempV wth storrrMesSsr cpaSily ans HOT corbel
New or Redeveloped Impervious Site Area I
Column 1
iBox 1
Cciurnn 2
Celuran 3
INSTRUCTIONS
1 Enter square fociaeje cf new or
icsc t al the lap ctf Km fonn
2 SeSecl iTrpersiixjs area reducSm
eehniyuefi from rows 1 -3 So rts^jce
msmaerriEnt requrmsenL Tree -:redis
Ean fas? cslcyiaSed UEF>^ '5he tree £pe*il
xvorfeshee! an the nstl pag*!.
3 Se4ecl di?si!ed ^CTirMiaJai
rroia^ement ladlfejea fesm r^MB 4-12.
n Caymn 1, erfisr t^ square fostage
srf ini^moiis H^a Ih2§ sac!i feiHy
4. ^Wlipt')!' c^ch s(npefV9CM& ^ea %3tn
Cclurn 1 fctf the D3ire^andhg BKJTig
adcr m Cakrrn 2. and srier ihe
festjl in Cd^wm 3. T?IK is the faciity
surface ama neecterf ID rrariaae
5. Total Cduim 1 (Rcws 1-12iar«
^aragEd* in Bc« 2.
§. Sytlra^ Ba» 2 IJCITI Bos. 1 a"id
arter St?e re^dl in Box 3, If Sto
nufnfc«r j£ less ^hari 90Q sqiiare fee!,
leqjnamerts have been mel. SJamfc
!h»s tern wih She applsaficn iar
peintl.
7. If B&y. 3 is gieaEes than 500 sqisre
bet a^Ki Kiuaie totage ef faclrlse«
lo Cdutm 1 and nsc^cdale & use
addfesrsl faciliies tetn QTapJer '3.0
3f Ihe Sla-rrw*3^r ^fenaoemert
^auisl ID nranage BlE*n»ralef fnxn
Imper'/tous Area
Reduction Technique
1 'i EoB-Roo* / Roof Garten
Impervious
Area Managed =
faalrty Surface Area
el
2) Contansd Planter Bas si
3) Tree &'«Jt l-See NexJ Page) si
Note. Porous Pavsfnefit jeas do nol need to be included in Box 1
Stormwater
Management
Facility
4) InliHralicn Planter Box
5) Fiow-Threugp Banter 8m
6) Vegswsd Swale
7) Graa%- &va(s
8) VegttsSed Filter Stnp
S) VegetECBtt Infll. Basin
ID) Said Filter
11) East Side Soakage Traidi
12) Wteat Sida Saatege Tsench
Total Impervious Area
Managed
Box 1 - Box 2
Imperious Facility
Area Sizing Surface
Managed Facto Area
El x 0.06 = |
si x 0.06 = |
et * 0.06 = |
si x G.Q6 = |
si x aoe = I 1
| |Box2
I |Box3
Uri
at
sf
ef
sf
ef
ef
at
sf
sf
(Figure No. 2) Simplified Sizing Form from the 2002 Revision of the City of Portland Stormwater Management
Manual.
This sizing form is unique because it incorporates not only sizing for water quality treatment but also sizing
for flow control and detention as well. When seeking to retrofit existing development - guidance pieces
from your new and redevelopment Stormwater facility requirements manual can be very useful.
Portland has a number of programs geared toward assisting property owners to retrofit their sites to do
onsite Stormwater management. The majority of homes in Portland are currently piped into a combined or
separate storm sewer. Onsite Stormwater management facilities can help mitigate a site's effective
impervious area and better mimic the natural hydrologic water cycle. Here's a highlight of some of
Portland's most successful programs:
Downspout Disconnection (for residential properties) - Driven by the need to remove water from
the combined sewer system to reduce overflows, in 1996 the City of Portland created the Downspout
186
-------
Disconnection Program. This program targets properties in north, northeast and southeast Portland to
disconnect roof downspouts onto lawns and flowerbeds. Property owners may also use onsite stormwater
management facilities such as drywells and soakage trenches. This program is very unique in its approach.
BES developed an interagency agreement with the City's Plumbing division to work directly with
homeowners to disconnect downspouts without the homeowner having to get a plumbing permit for the
alterations to their building's drainage system. BES staff developed safety criteria for allowable
disconnections and set up a monitoring and inspection program to assure disconnections were completed
safely. To implement, a target area of Combined Sewer Overflow (CSO) basins is selected and
Disconnection Program staff go to work. An aggressive marketing and door-to-door canvassing campaign
begins, to get voluntary agreement from property owners to complete the disconnection. Owners then elect
to complete the disconnection themselves and receive a $53 per downspout incentive, or to have the City
complete the disconnection for them free of charge. The City disconnections are completed either by
volunteer groups (such as scouting troops, neighborhood groups, and students) or by emerging or minority
small business contractors. Volunteer groups receive a stipend for each downspout they disconnect.
Contractors are chosen through a City bid process. The City then inspects the work of the volunteers, City
contractors, homeowner or plumber the homeowner may have hired, to assure disconnections are made
safely. If the goal for the target amount of roof area removed is not met in a basin, a mandatory version of
the program can be implemented. Other stormwater management messages are delivered under this
program - such as planting trees for homeowners who have disconnected. The City has disconnected
downspouts at almost 17,000 homes over the last six and a half years, and has collected data on prior
disconnections at an additional 20,000 homes. The program is funded primarily by a mixture of capital and
operating funds due to this ability to remove enough stormwater from the CSO system, that collection pipes
may be able to be downsized providing significant pipe construction cost savings.
(Figure No. 3) Typical Residential Downspout Disconnection
187
-------
Sustainable Site Development- This program grew from the early pilot project efforts to comply with
the NPDES Municipal Stormwater Permit. The program offers technical assistance and design guidance for
retrofit and developing properties. This free assistance might sway a property owner to use a swale instead
of a pipe to convey parking lot runoff. City staff usually makes the initial contact from a referral of the
watershed planning staff or through a land use or building plan review. This program has had some good
initial success due to early contact and the ability to provide some design details to developers. The
program primarily involves investment of staff time only and supports approximately 20 projects a year.
Stewardship Program - This is a joint program of the City of Portland, Portland State University and
Americorps program. The Stewardship program staff members assist individual property owners with
revegetation and onsite stormwater management projects. Students assist property owners in developing
site designs, identifying and applying for appropriate local, state and federal permits, and identifying
volunteers or other resources to implement the project. Students are assigned to specific watershed
programs within the City and often coordinate and complete projects with local watershed councils.
Stewardship Program staff and grants are funded through stormwater utility fees and work with about 10
projects per year.
Providing Motivation - Recognition and Incentives
Although we have taught the citizens of Portland about their impacts on local watershed and given them
some guidance and technical assistance on how to change behavior and retrofit their properties, most people
still need more motivation to make a change. Individual motivations can be varied across a broad spectrum
- but two common motivations are recognition and money. The City has developed a number of programs
that rely on recognition and/or other incentives to drive change in our citizenry. Here are some program
highlights:
Ecological Business Program - Interviews of local NE Portland Businesses in 1995 demonstrated a
desire of business owners not to be characterized as the "environmental bad guys." Many business owners
strive to do the most environmentally friendly thing, and the number of bad actors from an environmental
standpoint is usually a small percentage of the businesses out there. So, rather than relying on the few
business horror stories as the only case studies reported by the media, businesses asked the City to develop a
program to highlight "environmentally friendly" businesses. The City took this request to heart. The City
already had a partnership with six other local and state agencies to produce educational materials in a
coordinated matter. The partnership, called the Pollution Prevention Outreach (P2O) Team, already
produced successful used oil disposal and paint waste outreach materials that were helpful to businesses. So
the P2O team developed the Ecological Business Program. "Eco-biz" was the first multi-media and multi-
jurisdictional business recognition program in the nation. Local regulatory staff with air, water quality,
wastewater, hazardous waste, solid waste, stormwater, energy and water-use backgrounds developed a
certification and recognition program to highlight environmentally friendly businesses in the Portland
region. The program is business sector-based. Eco-biz started with automotive service shops and is now
working on a Landscape Contractor program. Along the way, Dental and Print Shop programs similar to
Eco-biz have been developed in the region. Eco-biz partners work with local business trade groups to
develop environmentally friendly best management practices, a program certification checklist and
recognition materials for program participants. After a certification visit, participating shops receive a shop
display package, press coverage, listing on the program web site (www.ecobiz.org), and general promotion
on the radio and at public events. This program is funded by several agencies through grants, agency staff
time and minimal advertising and printing budgets (< $10,000). Over 40 automotive shops are certified
188
-------
since the program launched in September of 1999. Those shops on average have implemented 89% of all
the recommended environmental actions - including stormwater improvements from redirecting wash
waters away from storm systems and providing secondary containment for liquid storage and working areas.
An evaluation report completed in September of 2000 found the average ecobiz shop generated 5 cubic feet
less cardboard and paper, 2.5 cubic feet less of metal scrap and 4 less batteries to the solid waste system per
month.
(Figure No. 4) Ecological Business Automotive Services Program Logo
Stewardship Grants - One aspect of the Stewardship Program is the Stewardship Grants Program. BES
funds a small number of low cost grants (<$5,000) for community-based projects in the City. Grants have
been used to pay for streambank restoration projects, downspout disconnections, stormwater facility
retrofits and naturescaping. Applicants can be either public or private entities and a number of the grants
have gone to school projects - including one native plant greenhouse. Grants are awarded every May and
must be completed by the following summer. Applications stressing partnership with other community
groups or showing inclusion of other investment or funding sources are prioritized for grant award. In the
grant year of 2001, $46,374 was awarded yielding $242,683 worth of project investment. Projects resulted
in planting over 10,000 trees and restoration of over 8,800 lineal feet of streambank. Projects are
recognized each year in an annual report prepared by BES. Grants are funded by Stormwater utility fees.
Clean River Incentive and Discount Program (CRID) - This incentive program will provide
financial incentives to property owners who manage stormwater on their site. The program is currently
delayed due to the installation of a failing billing system. Once the billing system is repaired, the program
should be instituted. The main goal of the GRID is to drive property owners to retrofit through provision of
a discount on their monthly stormwater utility charge. The GRID was developed in the summer and fall of
2000 as a method of rate reform for the citizens of Portland. City sewer rates are rising at approximately
9% a year to fund the billion-dollar CSO program. The GRID actually alters the breakdown of the
stormwater utility rate. Previously, properties paid one rate based on the amount of impervious surface on
their property. In January 2001, the Portland City Council instituted a two-part rate -35% of the charge for
providing drainage services to the property and 65% of the charge to provide drainage services to the public
right of way that served the property. Not only did the charge breakdown reinforce that street drainage is an
issue the City must deal with, it also allowed a portion of the rate to be discounted for properties providing
189
-------
onsite stormwater management. So with 35% of the stormwater rate up for a potential discount, some
properties could be incented to make retrofit changes. The GRID has a simplified discount program for
residential properties based on volume control, and a more complex commercial property program that
requires water quality and flow control for the full discount. Surface vegetated facilities were ranked higher
than subsurface facilities for the eligible portion of the discount. BES was working on a prorated discount
funding program to help pay for the initial capital outlay when the City's new water and sewer billing
system started to fail.
Conclusions
The City of Portland has successfully developed a number of educational, technical assistance, recognition
and incentive based programs to encourage our citizens to help limit their impacts on local watersheds.
While these efforts may be noteworthy, they are not sufficient to address existing development watershed
impacts all by themselves. Some tasks for mitigating urban area impacts are the City's alone. So the City
will continue to build regional stormwater management facilities, improve our operations and maintenance
practices on City streets and sewers and protect and enhance riparian resource areas. But we will be looking
to develop additional programs to enlist the aid of Portland's citizens to limit our impacts on local
watersheds. While new programs may have staffing and other limited resources available for
implementation, there will be no lack in drive from City staff and our local environmental advocates to
reach for that 10% effective impervious area target.
References
1. City of Portland Bureau of Environmental Services Draft Biological Assessment of the Effects of
Stormwater Management Activities on Thirteen ESUs of Chinook, Sockeye, and Chum Salmon, Steelhead,
and Cutthroat Trout, Beak & CH2M Hill, July 2000.
2. City of Portland Zoning Codes and Metro Service District 2040 Land Use Plan Framework, City of
Portland Bureau of Environmental Services Graphical Information System Zoning Layer, October 2002
(updated quarterly)
3. Program Evaluation: Automotive Ecological Business Program, Valle & Nielson (BES Intern Program),
September 2000
190
-------
LESSONS LEARNED FROM IN-FIELD EVALUATIONS OF PHASE I MUNICIPAL
STORM WATER PROGRAMS
John Kosco, Wes Ganter, and James Collins
Tetra Tech, Inc.
Fairfax, VA
Laura Gentile and John linger
U.S. EPA Region IX
San Francisco, CA
Abstract
Tetra Tech is assisting EPA in the evaluation of a number of storm water Phase IMS4 permit programs in
California and selected other States. These evaluations consist of two components: a programmatic review
of individual city and county programs implementing permit requirements and an on-site/in-field
verification of these program elements. This in-field verification allows EPA and the State to assess
whether a program is actually being implemented as described 'on paper.' The overall goals of these
evaluations are to complete a baseline assessment of each program area, determine compliance with permit
requirements and the stormwater management plan, collect information for permit reissuance, and determine
how municipalities measure program effectiveness. In addition, the 'lessons learned' from these evaluations
can be directly applied by many of the Phase U jurisdictions, which will begin permit coverage in March
2003.
Introduction
On November 16, 1990, the U.S. Environmental Protection Agency (EPA) published regulations (the 'Phase
I rule') requiring National Pollutant Discharge Elimination System (NPDES) permits for certain industrial,
construction and municipal sources of storm water runoff fundamentally changing the way storm water
runoff is regulated at the State and Federal levels. Approximately 1,000 MS4s ('municipal separate storm
sewer systems'), consisting primarily of City and County government agencies responsible for storm water,
have been permitted under the Phase I regulations. The Phase IMS4 regulations generally require MS4s to
reduce discharges of pollutants to the maximum extent practicable and to prohibit illicit discharges to the
MS4. Specific elements in a Phase I Municipal Storm Water Management Program include public
education, public agency or municipal maintenance activities, new development, construction,
industrial/commercial facilities, illicit discharges and improper disposal, monitoring and reporting.
Phase U of the storm water program, established in 1999, extends the coverage to include municipalities
within urbanized areas and all construction disturbing at least one acre. Permits for these Phase U sources,
which will include over 5,000 additional MS4s, are scheduled to become effective on March 10, 2003.
Phase U Municipal Storm Water Management Programs are required to address public education, public
involvement, illicit discharges, construction, new development, and municipal operations.
Although many Phase IMS4 permits are in their second or third permit cycle, EPA has not yet completed a
comprehensive compliance assessment of these MS4 permits. A General Accounting Office report
published in June 2001 (GAO, 2001) found that neither the overall costs of implementing the storm water
191
-------
program nor the program's effectiveness had been determined. This GAO report followed an EPA report on
the Phase I storm water regulations (EPA, 2000) that found many effective Phase I program components,
but admitted that EPA did not have a system in place to comprehensively measure the success of the Phase I
program on a national scale.
Storm Water Phase I MS4 Evaluations
EPA Region IX hired Tetra Tech, Inc. in 2001 to begin a series of MS4 evaluations in the State of California
to assess the compliance status of individual storm water Phase IMS4 permittees. In order to assess on-the-
ground implementation of the programs, these program evaluations are conducted on-site. The on-site
evaluation consists of two components: a programmatic review of individual MS4 programs implementing
permit requirements and an in-field verification of these program elements. This in-field verification allows
EPA and the State to assess whether a program is actually being implemented as described 'on paper.'
The project goals of the on-site MS4 evaluations include obtaining an overall picture of MS4 compliance,
documenting effective elements of existing Phase I programs, identifying methods to improve MS4 program
reporting, and developing a guidance document to assist State and/or EPA inspectors in conducting future
MS4 evaluations.
Determining compliance with MS4 permits is in many cases subjective. Unlike some other environmental
programs such as the pretreatment program, there is no checklist, list of BMPs, or objective criteria that all
MS4s need to meet. In addition, EPA has not defined 'maximum extent practicable' or MEP which is the
regulatory standard that MS4s must meet. This leaves it up to individual permit writers to define for each
MS4 permit. Therefore, the MS4 inspectors have been using their best professional judgment and
experience to identify program elements that are 'effective' or 'deficient.'
The MS4 on-site evaluations conducted to date have typically consisted of a 3-4 day on-site review. This
on-site review has been conducted on a single MS4, and has also included multiple co-permittee MS4s
evaluated with up to three investigators. For each of the MS4s evaluated, a number of staff from multiple
departments were typically involved. Typical departments involved in the MS4 evaluations included public
works, transportation, planning, development, and parks/recreation. As of December 2002, 14 MS4
evaluations have been conducted in EPA Region IX, covering 41 separate permittees.
The MS4 inspectors typically do not review or make recommendations on financial resources. Where a
program element is clearly not being implemented to the maximum extent practicable - for example, when
compliance with local construction erosion and sediment control requirements is poor due to lack of
inspections - that will be noted as a deficiency. The MS4 inspectors will suggest improvements to the
program so resources can be used for effectively, but responding to those suggestions or how to resolve the
identified deficiencies is up to each individual MS4.
A wide variety of storm water permits, storm water management programs, and compliance with those
permits and programs were found during the evaluations. However, some common trends were observed as
indicated in the following sections. The trends and evaluation findings are grouped into the broad
categories of program management/planning, implementation, and evaluation.
192
-------
Program Management/Planning Findings
A clear, well-written permit and plan are critical for successful implementation of a storm water
management program. This requires the permitting authority to describe the required actions clearly in a
permit and the permittee to clearly articulate how it will meet these requirements in a storm water plan. The
Phase IMS4 evaluations conducted by Tetra Tech have found that the more advanced storm water programs
generally have more detailed, well-written permits and plans. Several findings common to most of the
programs evaluated are described below.
NPDES MS4 permits and MS4 stormwater management programs need to contain quantifiable,
measurable elements so compliance can be determined.
Storm water permits vary significantly in their level of detail. Some third-term permits issued in California
contain very specific, measurable elements which are clear for the permittee to implement and relatively
straightforward for the State to determine compliance. For nonspecific permits that simply require the MS4
to "implement a storm water management plan," determining compliance becomes more difficult. More
importantly, the permit does not specify, or measure, the level of effort expected, so MS4s do not have a
clear target to achieve.
The storm water Phase II regulations require small MS4s to develop "measurable goals" for each BMP in
their program. These measurable goals are intended to provide a quantifiable target for the MS4s to achieve
in the implementation of that BMP. Although a similar requirement does not specifically exist for Phase I,
permits and programs developed under Phase I should begin to include these measurable goals. For
example, the permit and program should specify the number of industrial inspections expected per year and
the number of catch basins that should be inspected and cleaned. This provides a level of certainty to the
MS4 that they are successfully implementing the permit and allows the State to more easily evaluate
compliance.
Some MS4 permits in California are including specific, measurable requirements that make determining
compliance easier. Also, the City and County of Sacramento have developed stormwater plans that are
clear, well-written, and begin to address the issue of measurable goals which are called 'minimum
performance standards' and 'performance and effectiveness measures', respectively, in each plan (City of
Sacramento, 2000 and County of Sacramento, 2000).
Programs are not designed to specifically address pollutants of concern.
The primary goal of programs under the Clean Water Act is to achieve fishable, swimmable waters by
meeting water quality standards. Many MS4 programs are not designed to address the specific pollutants of
concern already identified in their watershed. Where pollutants of concern have been identified, MS4
programs should be modified to include BMPs and programs that specifically target a reduction in these
pollutants.
Some Phase I programs in California are developing plans to address identified pollutants of concern in their
community, including those pollutants identified on the State's Section 303(d) list. Pollutants of concern can
also be identified from local studies or watershed research. Several programs, including programs in
Alameda County and Sacramento County, have developed strategies to more specifically target and reduce
pollutants of concern. For example, Sacramento County is developing a series of Target Pollutant
Reduction strategies to focus some program resources on pollutants that cause or are likely to cause
impairments in local receiving waters. Target pollutants for the Sacramento area include diazinon,
193
-------
chlorpyrifos, coliform/pathogens, copper, and lead. Sacramento County still implements baseline activities,
but uses the target pollutant reduction strategies to ensure activities are developed to address specific
pollutants.
Combining resources and expertise into a committee can save MS4s time and money.
Many MS4s that have been permitted together have joined resources in a committee structure. This sharing
of resources and experience can help all participating MS4s by more efficiently developing public education
materials, guidance, standard forms and other materials for all of the MS4s to use. Also, for smaller MS4s
with more limited budgets, the committee structure provides assistance these MS4s may not have been able
to otherwise obtain, such as use of a centralized database for entering and managing reporting information.
Examples of storm water management committees can be found in several California counties, including
Alameda, Sacramento, Ventura, San Diego, and Los Angeles.
Implementation Findings
As the stormwater Phase I program is implemented and matures, Phase I MS4s are continuing to struggle
with the implementation of several common aspects of the program. On-the-ground activities such as
inspections of construction sites and industrial facilities appear to be a common problem, while other
programs like public education and municipal maintenance are often more advanced. Below are several of
the common findings associated with implementation of the storm water Phase I program.
Compliance with local construction site erosion and sediment controls is a challenge for all MS4s.
Storm water Phase I regulations require MS4s to develop a local program to control construction site runoff.
Many MS4s, however, find this program a challenge to implement. The frequency of inspections at
construction sites required to ensure proper installation and maintenance of erosion and sediment control
BMPs is often lacking. Some MS4s count all inspector visits to construction sites, even inspectors who
have nothing to do with erosion and sediment controls. Also, some MS4s have different requirements for
public and private construction sites. All of these factors can contribute to a program that is ineffective in
preventing erosion and sediment control problems at construction sites.
Tetra Tech has found that successful programs often have dedicated erosion and sediment control inspectors
for local construction projects. These inspectors are involved in not only inspections, but also participate in
the plan review process so they are aware of what erosion and sediment controls and post-construction
BMPs the construction sites are required to implement. Also, these inspectors have adequate enforcement
mechanisms such as stop work authority or the ability to fine contractors to ensure compliance.
Local MS4 industrial and construction inspectors are often unaware of State permit requirements.
The State of California, like all states, has issued statewide general permits for controlling storm water
runoff from industrial facilities and construction activity. Within Phase I areas, however, industrial facilities
and construction operators also need to comply with the local MS4 program to address industrial or
construction runoff. Many local inspectors, although they are trained in the local requirements, are often
unaware of the requirements contained in the statewide permit. In some cases this is intentional, as the MS4
does not want the responsibility of enforcing the statewide permit requirements. However, MS4s can
provide a valuable service to their local construction and industrial facilities by explaining the difference
between the two sets of requirements, and what these facilities need to do to comply with the statewide
requirements.
194
-------
Some programs avoid this problem by simply adopting the statewide permit requirement for a stormwater
pollution prevention plan (SWPPP) as their own requirement. This ensures that local construction operators
only need to develop one plan to comply with both local and state stormwater requirements, and local
construction inspectors only need to know one set of requirements.
Pretreatment inspectors, if available, can efficiently conduct industrial stormwater inspections.
The pretreatment program is a well-established program with existing staff trained in water quality practices
and enforcement techniques. Some MS4s have expanded the role of pretreatment inspectors to also conduct
industrial stormwater inspections. Many of these industrial facilities are already included in the
pretreatment program, therefore the on-site inspector simply needs to also include several stormwater
elements in their inspections. For MS4s with an existing pretreatment program, this expansion of
pretreatment inspector duties to include stormwater inspections effectively implements the program without
creating a separate inspection program. Of course, this approach may not be as effective in areas where the
sanitary sewer system does not fully coincide with the storm drainage system (e.g., areas on septic systems).
Many MS4s fail to identify and eliminate dry weather discharges.
A separate storm drain system is designed to carry only storm water runoff. Dry weather, therefore,
presents MS4s an excellent opportunity to identify and eliminate non-storm water discharges to their storm
drain system. The evaluations have found that many MS4s, however, fail to identify and eliminate dry
weather discharges. These MS4s either fail to look for any discharges during dry weather, or assume that all
dry weather discharges are attributable to landscape irrigation, groundwater infiltration, or some other
uncontaminated source.
Municipal maintenance and spill response programs are often more advanced than other program
areas.
Due to the need to minimize episodes of flooding, MS4s often have effective maintenance programs of their
storm drain systems. The municipal maintenance staff are often well trained, equipped, and have detailed
records of their maintenance activities. Also, other related programs such as street sweeping, which are
often initiated for different reasons (e.g., aesthetics), also have significant stormwater benefits. In addition,
for obvious public safety reasons, many MS4s have effective spill response programs.
Many MS4s have extensive public education programs.
Public education programs are often an 'easy' and 'fun' program for MS4s to implement. Many MS4s have
been very innovative in finding new methods to reach target audiences. This includes websites, classroom
educational programs, radio and TV commercials, mascots, and public involvement programs such as storm
drain stenciling programs. Some MS4s have also taken surveys of their residents to determine the overall
level of awareness and effectiveness of their public education programs.
Evaluation Findings
As EPA found with its 2000 Report to Congress (EPA, 2000), evaluating the effectiveness of the stormwater
program is a difficult task. However, successful programs are developing local measures by which progress
or effectiveness can be evaluated, including the use of environmental indicators. Tetra Tech found that
many programs share common problems in terms of program evaluation, as described in the findings below.
195
-------
MS4 programs are not evaluating their data and are therefore not modifying programs in response to
trends in this data.
EPA envisioned the storm water program to be an iterative process. Storm water permits, and programs,
should evaluate what is working and be able to make modifications in response to changing conditions.
Many programs, however, are not collecting the data, such as monitoring or other performance and
effectiveness data, necessary to determine needed changes.
At a minimum, programs should complete a comprehensive outcome evaluation at the end of each permit
term, and should complete an annual process evaluation at the end of each year with the submittal of the
annual report. This will ensure that programs are responsive to changing priorities and needs.
MS4 programs should develop different methods to evaluate the effectiveness of their programs.
All Phase I MS4s collect monitoring data, but few programs are collecting enough water quality data to
show statistically significant changes. Other evaluation techniques, such as environmental indicators,
should be considered by these programs as a way to characterize water quality conditions and provide a
benchmark for evaluating the success of the stormwater management program. These indicators (Claytor
and Brown, 1996) should include a mixture of programmatic indicators, physical and hydrological
indicators, biological indictors, social indicators, programmatic indicators and site indicators. Examples
include toxicity testing as a water quality indicator and the number of illicit connections identified/corrected
as a programmatic indicator. These indicators are important due to the difficulty and expense in
documenting water quality improvements solely from water quality monitoring data. Environmental
indicators can also be used to ascertain that high quality waters are being maintained or provide an early
warning of when their beneficial uses are at risk of being degraded.
Annual reports provide useful information, but are not always good indicators of program
effectiveness.
The on-site evaluations have revealed that, although annual reports can indicate the success of a program,
poor programs can hide behind well-written annual reports and some aspects of effective programs can be
hidden or missing from annual reports. Because there is not a standardized reporting process for all Phase I
MS4s, this allows each MS4 to choose the type of information it wants to present. A knowledgeable report
writer can selectively report certain information, such as the total number of municipal inspectors visiting a
construction site instead of the number of inspectors specifically evaluating stormwater controls.
The absence of a standardized report could become especially important as the 5,000+ stormwater Phase n
MS4s begin to submit annual reports. A consistent reporting format will allow states to compare
information collected from MS4s and will also allow EPA to compare reporting results across states.
Compliance with a permit may not always indicate that a program is successful in protecting water
quality.
There is a significant variability in the requirements within the Phase IMS4 permits, even within the State
of California. This variability, along with the iterative nature of stormwater permitting, allows MS4s to
operate under different guidelines, and implement different programs. A programs success should be tied
not only to meeting permit requirements, but also to meeting water quality goals.
196
-------
Conclusions
Before the storm water Phase I program, most municipal storm water programs were primarily designed to
address water quantity issues (e.g, minimize flooding). The storm water Phase I program is beginning to
mature and learn from mistakes in the past, however a significant amount of work remains in developing
guidance or programs to document these lessons. Improved reporting, monitoring, and evaluation
techniques are needed, but will likely only be implemented in many programs through changes in NPDES
permit requirements.
References
U.S. General Accounting Office (GAO), 2001. Water Quality: Better Data and Evaluation of Urban Runoff
Programs Needed to Assess Effectiveness. GAO-01-679. Washington, DC.
U.S. Environmental Protection Agency, 2000. Report to Congress on the Phase I Storm Water Regulations.
EPA 833-R-00-001. Washington, DC.
Claytor, R. and W. Brown, 1996. Environmental Indicators to Assess Stormwater Control Programs and
Practices, Center for Watershed Protection, Ellicott City, Maryland.
City of Sacramento, 2000. Stormwater Quality Improvement Plan, Draft Version.
County of Sacramento, 2000. Stormwater Quality Improvement Plan for County of Sacramento and Cities
of Citrus Heights, Elk Grove, Folsom, and Gait. Draft.
197
-------
ECOROOFS (GREENROOFS) - A MORE SUSTAINABLE INFRASTRUCTURE
Tom Liptan1 and Eric Strecker2
of Portland, Bureau of Environmental Services, 1120 SW 5th Ave., Portland, Oregon 97204 Ph.503
823 7267 TomL@BES.ci.port! and, or.us
2GeoSyntec Consultants, 838 SW First Avenue, Suite 430, Portland, Oregon 97204, Ph. 504-222-9518,
estrecker@geosyntec. com
Abstract
All cities have two primary impervious elements; rooftops and pavement. These usually represent an
extensive network of imperviousness and make up about 45% of the surface area of a city at full build out.
The results of this imperviousness have been documented in a number of papers, but the main
environmental effects include increased destabilization of streams and increased runoff pollutant loadings
and concentrations. To address stormwater concerns and to provide other environmental benefits, the City
of Portland has developed a program to encourage the use of EcoRoofs (vegetated roofs). This paper will
present the overall City program, including a discussion of the incentives and assistance the City provides to
encourage development projects to employ green roofs. The paper will review some of the installations
that have occurred and discuss some of the practical lessons that have been learned regarding green roofs.
The City has also been monitoring runoff from several EcoRoofs in an attempt to ascertain the water
quantity and quality performance of the roofs in slowing down or eliminating runoff, as well as associated
pollutant loads and concentrations. The monitoring has included the installation of rooftop rain gages and
flow measurement devices. Water quality samples are also collected. One roof has had two different
depths of soil layers (2" and 4") employed with separate flow monitoring gages for each. The paper
presents hydrological results for selected storm events on a seasonal basis, as well as initial water quality
results.
Introduction
The elements of urban development are similar throughout the United States. Homes, apartments,
commercial and industrial sites and the supporting transportation systems cover the land in varying
densities. Large areas of impervious surface in the form of rooftops and pavement have been placed on the
land, wetlands, and even creeks. However, the ideal conditions for salmon, and other wildlife of the Pacific
Northwest are predominately an evergreen (coniferous) forest and its associated functions with clean cool
rivers and streams. The results of this imperviousness have been documented in a number of papers, but the
main environmental effects include increased destabilization of streams and increased runoff pollutant
loadings and concentrations (May et. al., 1997). Since these impervious urban elements are essential to
human communities, what can be done to mitigate their negative impacts? In Portland, we are implementing
new design techniques, which include EcoRoofs (living vegetated roof ecosystems), pervious pavements,
landscape planters and swales, infiltration gardens, watergardens, vertical landscaping, and trees. The
techniques are applicable to new and re-development, and to retrofitting existing development. The focus of
this paper is on the 'EcoRoof and its potential for reducing the impacts of urbanization.
-------
What is an EcoRoof?
An EcoRoof is a living vegetated ecosystem of lightweight soil and self-sustaining vegetation. It is
biologically 'alive' and as such provides a protective cover on the building by using the natural elements of
sun, wind, and rain to sustain itself. This protective cover allows the waterproof membrane to last for as
long as 30-40 years or more. The EcoRoof requires little maintenance and provides an aesthetic alternative,
with economic and ecological attributes not found in a conventional roof. The main components include a
waterproof membrane or material that prevents water from entering the building; drainage material such as
geotextile webbing that allows water to flow to the drains when the substrate is saturated; and soil or
substrate (growing medium) as light as 6 pounds per square foot (psf). To date in Portland, the lightest
weight substrate used is at Hamilton Apartments at 10 psf saturated, at a 3-inch depth. Selection of
vegetation or plant materials can range from mosses, lichens and ferns, to sedums and other succulents, to
grasses, herbs and ground covers. Irrigation requirements are very much affected by the plants selection.
Sedums and succulents appear to be the mainstay of least water dependent plants, based on experience in
Portland. Figure 1 shows the Hamilton Apartments EcoRoof.
Figure 1. Aerial view of a vegetated EcoRoof on Hamilton Apartments in downtown Portland.
A traditional Roof Garden (see example in Figure 2, left photo) by comparison usually requires more
substantial structural building upgrade and is made up of heavy soils and vegetation, often including trees,
and requires significantly more irrigation and maintenance. Roof gardens may cover only a small
percentage of the roof surface and usually have paved terraces for people to use. Although they do provide
some benefits not found with the use of conventional roofs, they do not provide the benefits as an EcoRoof.
They also are generally much more expensive to build and maintain than conventional roofs. EcoRoofs are
more comparable in cost to standard roofs.
Another type of vegetated roof is an ag-roof (see example in Figure 2, right photo). Some building owners
are finding it advantageous to grow crops on their rooftops. One such Portlander harvests hundreds of
pounds of tomatoes each week.
The City of Portland decided to use the term 'EcoRoof to describe their "green" roof program for several
reasons. First, the western United States including most of Oregon and Washington has dry warmer seasons
and may not receive precipitation for many months. Native plants although more self-sustaining often do
-------
not remain "green." Thus a "not green" or brown roof does not imply that the vegetation has died, thus the
prefix eco (for ecosystem) was chosen as being more descriptive of what the roofs are intended to achieve. .
Another reason was the many references to the economic value, especially the longer life, thus eco also
refers to the economic benefits.
Figure 2. An aerial view of a typical Roof Garden in downtown Portland and Doug Christie and Cameron Hyde atop
Doug's ag-roof in Portland with crops shown.
What Do EcoRoofs Do?
Based upon an evaluation of hydrological, energy, and other principals and monitoring data produced thus
far in Portland, EcoRoofs appear to be able to address many environmental and economic issues. The City's
original interest was stormwater management, but has since broadened to consideration of other EcoRoof
attributes. Precipitation that lands on an EcoRoof acts in the following ways. Portions of it are intercepted
by vegetation and then evaporate; portions are absorbed in the soil; portions in the soil are taken into the
vegetation and then transpire; some water evaporates from the soil; and excess amounts flow through the
soil and become runoff. These characteristics are highly affected by seasonal conditions. Interception,
evaporation, and transpiration act to prevent runoff and can be lumped into one term, rainfall retention. This
portion of the rain never turns into runoff. One of the primary objectives of the monitoring program has
been to assess the effectiveness of EcoRoofs in reducing tie volume of runoff. Some water quality
monitoring has also been performed to assess the potential for reduced as well as added pollutants in the
runoff that does occur. Finally, the hydrology and water quality results have been employed to assess
potential reductions in pollutant loads. Table 1 provides a comparison of EcoRoofs environmental and
other characteristics to conventional roofs. Note that the conventional roof is often the cause of the problem
being addressed.
The Portland EcoRoof Experience
Portland has a total area of 135 square miles. Although rooftops constitute only one type of surfacing, they
represent about 40% of all impervious surfaces in the city. At full build-out based on current zoning,
rooftops are likely to cover more than 25 square miles of the city. However, if zoning densities increase
over the coming decades the city roof area could be much larger.
-------
Table 1. A Comparison of Environmental and Other Characteristics of EcoRoofs and Conventional roofs.
Subject
Stormwater
Volume retention
Peak flow mitigation
Temperature
mitigation
Improved water
quality
Air quality
Energy Conservation
Vegetation
Green Space
Zoning floor area
bonus
City Drainage fee
reduction
Approved as
Stormwater
management
Habitat
Livability
Costs
Cost off-sets
Durability
EcoRoof
10-35% during wet season, 65-100% during dry season
All storms reduced runoff peaks
All storms
Retains atmospheric deposition and retards roof material
degradation, reduced volumes reduce pollutant loadings
Filters air, prevents temperature increases, stores carbon
Insulates buildings, reduces Urban Heat Island impacts
Allows seasonal evapotranspiration; provides
photosynthesis, oxygen, carbon water balance
Replaces green space lost to building footprint:, although
not equal to a forest
3 ft2 added floor area ratio (FAR) for each EcoRoof ft2
when building cover over 60%
To be determined, may be up to 45%
For all current city requirements
For insects and birds
Buffers noise, eliminates glare, alternative aesthetic, offers
passive recreation
Highly variable from $5-$12 ft2 new construction and $7 -
$20 ft2 retrofits
Reduced Stormwater facilities, energy savings, higher rental
value, increase property values, reduced need for insulation
materials, reduces waste to landfill, creates jobs and
industry
Waterproof membrane protected from solar and
temperature exposure lasts more than 36 years, membrane
protected from O&M staff damage
Conventional
Roof
None
None
None
No
None
None
None
None
None
None
No
None
None
Highly variable
from $2-$10 ft2
new construction
and $4 -$15 ft2
retrofits
None
Little protection,
exposure to
elements, lasts
less than 20 years
In 1991 the city of Portland was required by US Environmental Protection Agency and the Oregon
Department of Environmental Quality to begin more aggressive programs to reduce pollutants in
-------
stormwater discharges and abate combined sewer overflows. Both of these problems have as their common
cause urbanization. The traditional solutions that Portland began to implement, included the use of end-of-
pipe treatment such as ponds to treat stormwater flows and large pipes and underground storage systems to
address combined sewerage overflows. The City did embark on a program to "Start at the Source" using
such techniques as roof drain disconnect programs in combined areas of the City. Portland first began to
consider EcoRoofs in 1995 . The technique seemed to fit the concept of creating something that would be
more like nature, absorbing and then slowly releasing moisture through evapotranspiration and low flows,
thus providing precipitation retention and stormwater management. The City began to ask if this could be a
way to reduce or control CSOs and reduce the erosive scouring forces of runoff in streams. Many people in
the city were intrigued with the possibilities and investigative efforts began in earnest.
Milestones in Portland's EcoRoof Program
The following presents a brief summary of the milestones in the development and implementation of the
City's EcoRoof program:
• 1996 -- First EcoRoof installed on a residential garage, stormwater monitoring was conducted for 27
months from 1997-1999
• 1997 - Bureau of Environmental Services (BES) and Portland General Electric (PGE) assisted Portland
State University planning students with a study on roof gardens. A report was produced.
• 1997 - BES built a small EcoRoof shelter at the Portland Home and Garden Show. Survey of over 600
visitors was 75% favorable.
• 1998 -- BES and PGE provide grant funding of a 300 ft2 EcoRoof installation on an apartment building.
This would be the first use of BES Community Watershed Stewardship Program grants for an EcoRoof.
• 1998 -- BES begins to offer limited technical assistance to developers who consider EcoRoofs.
• 1999 -- A city worker is interviewed on the NRP 'Living on Earth' show and receives encouraging
phone calls from around the country.
• 1999 -- Almost simultaneously two projects, with different owners, request BES assistance to install
EcoRoofs.
• July, 1999 — The EcoRoof is officially recognized as a stormwater management technique and is
included in the city's Stormwater Management Manual.
• September, 1999 -- Hamilton Apartments EcoRoofs are completed.
• March 2000 — Buckman Terrace mixed-use building EcoRoofs are completed.
• Early 2001 — BES began measuring precipitation and runoff at the Hamilton. However the efforts were
plagued with technical problems. In December, 2001, problems are corrected. Subsequent flow data not
only supports the monitoring results of the garage data, but also shows better performance.
-------
• 2001 — BES begins work on a drainage fee discount for installation of EcoRoofs or other green
approaches. (This work has been delayed and the discount is not expected to be available till 2004).
• 2001 — Two small EcoRoof shelters are completed at nature areas.
• March, 2001 — The city zoning code is amended to include EcoRoofs as a floor area bonus option.
Property owners can add up to 3 ft2 of floor area for every ft2 of EcoRoof if the EcoRoof covers at least
60% of the rooftop. Less area is granted if the % coverage is less than 60%.
• 2001 -- BES offers potentially $30,000 grants for EcoRoofs (or other green techniques) in a portion of
the combined sewer area. Two roof retrofits were considered and one is approved for funding.
• 2001 — Mosaic Condominiums apply for EcoRoof bonus and get enough ft2 to add six additional
condominiums to the building.
• September, 2001 — Ecotrust building EcoRoof completed.
• October, 2001-- BES and the City's Office of Sustainable Development convene a City EcoRoof Forum.
An overwhelming majority of attendees supported the EcoRoofs concept. Three major issues are
identified: need more cost-comparative information, need incentives at the early stages, and need
technical assistance.
• December, 2001 — BES installs an EcoRoof on a portion of the it's wastewater treatment plant.
• 2002 — BES completes an EcoRoof Question and Answer brochure and posts it on its web site.
• July 2002 -- Fire Station #12 EcoRoof is constructed.
• 2002 -- Mosaic condos begin construction.
Portland EcoRoof Monitoring
The City of Portland has been active in implementing monitoring programs to assess the effectiveness of the
EcoRoof in reducing impacts to downstream receiving waters as well as reducing CSO impacts. This
section presents a brief overview of the two monitoring projects that the City has been conducting.
Residential Garage EcoRoof Monitoring
An EcoRoof was installed on a structure shown (Figure 3) in October 1996 . The building structure was
upgraded and a waterproof membrane was applied over the existing composite rollout shingles. Two to
three inches of topsoil and compost mix were applied and planted with seven species of sedum. Grass has
also grown on its own with what appears to be four predominate species. The EcoRoof is 180 ft2 and has
about a 7% slope toward the east. About half of the roof has full solar exposure and the other half is
partially shaded. Figure 3 shows the EcoRoof in late spring 2002.
-------
Figure 3. View of a residential garage EcoRoof that was monitored for a two-year period.
BES monitored rainfall retention of the garage EcoRoof from August of 1997 until October of 1999. A rain
gage was installed on the EcoRoof and the roof downspout was connected to two tanks with a total capacity
of 78 gallons. A spreadsheet was created to record the rainfall, runoff and retention. The rain gage and tanks
were checked every morning and evening during storm events. Any flow in the tanks represented runoff and
the difference between rainfall and runoff was the retention. Figure 4 shows the precipitation retention for
the 27-month period. Figure 5 show the results of a rainfall simulation test to identify how peak flows might
be attenuated. In the test, a large volume of water was applied to the roof and then the recorded runoff was
compared to this volume. Water was applied with a garden hose and before each application the flow from
the hose was measured and recorded.
Percentage of rainfall retained on ecoroof
Average over 27 month period = 30% retention
relative to total rainfall volume retained, (not monthly averages)
r-i
r-i
^ a.
3 VI
DC
rn
Oct-97
Nov-97
-
£ £
a 1
Feb-98
r-i
00
ON
i.
es
S
n n
i iUa^ eE&,+;j >• « ^ •«
Mar-99
1-1
-
1-1
—
^ ^ ^ ^ &• &• &•
^ iU B -3 M D. -g
^ | ^ ^ ^ !B O
Figure 4. Chart showing the month-by-month percentage of rainfall retained on a residential garage roof in Portland.
-------
QL
54
u.
O
gs
<
2
TIME (minutes)
1 rain gallons _ discharge gallons
reg. roofl
Figure 5. Chart showing gallons of water introduced to a residential EcoRoof and the much lower number of gallons
that ran off, as well as much lower peak flows.
The percentage of retention on this roof on a monthly basis during the study period, has ranged from <10%
in Jan 1999, with 11 in. of rainfall and up to 100% in the dry season months. For the rainfall volume for the
two-year period, the average annual retention was about 28%. Rainfall during this two-year period was 99
in. or 33% more than the average two-year total of 74 in. Higher than average rainfall and the fact that the
EcoRoof is partially shaded in spring, fall and winter would have reduced evaporation and thus reduce the
retention performance. The simulated storm demonstrated how the EcoRoof could attenuate a large storm
under dry season conditions. The most sensitive stream conditions often occur when a larger warm weather
storm occurs.
Hamilton Apartments EcoRoofs Monitoring
The Hamilton Apartment Building (Figure 1) in downtown Portland is the site of a more comprehensive
monitoring effort by the City. The Housing Authority of Portland, in cooperation with the City of
Portland's Bureau of Environmental Services (BES), built the Hamilton Apartments Building EcoRoofs in
the autumn of 1999. Over 75 species of plants were installed in an identical arrangement on each side of the
building. Three different mechanisms were used to plant the vegetation, plugs, hydroseed, and mats. The
idea was to gain some understanding of which plants would do the best and what type of planting would
provide the best growth and coverage. An irrigation system was installed. BES is testing to determine
-------
characteristics of planting methods, measurement of runoff flows and precipitation, and viability of soil and
vegetation. Insects and birds are also being monitored to a very limited extent. Garland Co. waterproof
membrane and planting design was used on this project. Figure 6 shows various views of the EcoRoofs.
Figure 6. Photographs of the Hamilton EcoRoof, including an aerial photo from above, a close-up of vegetation and
a ladybug, and two pictures of the roof from the roof; one showing the area of the roof where access is restricted and
the other including the patio area behind a fence.
There are two drains on the building: an east drain has a 3,848 ft2 catchment with 2,620 ft2 of EcoRoof area
and the west drain has a 3,690 ft2 catchment with 2,520 ft2 of EcoRoof area. All other surfaces are vents,
parapet walls, gravel, and terrace paved areas. All monitoring is relative to these other surface contributions
and implies that a 100% EcoRoof would have improved precipitation management. The conventional roof
runoff has been disconnected from the EcoRoofs, but the terrace areas drain to each of the EcoRoof drains
through the substrate. In both cases, the catchments are about 75% EcoRoof and 25 % hard surfaces.
The drainage from the EcoRoof was split in half for research purposes. The west half has a four to five inch
soil which weighs 20-25 lbs/ft2 and the east half has a two-three inch soil weighing 7-10 lbs/ft2 when
saturated. The east soil is composed of digested fiber, encapsulated styrofoam, perlite, peat moss and
compost. The west soil consists of digested fiber, compost, perlite and topsoil. Figure 7 shows the
chemical composition of the two substrates utilized. In general, the Westside soil mixture appears to have
higher concentrations of heavy metals and nutrients. As rain falls and soaks into the soil it flows to the roof
drains located at each end of the building where a monitoring station collects flow data prior to entering the
piped system. There is an additional roof drain with a two-inch collar in case the monitoring equipment or
the main roof drain was to become plugged.
-------
Figure 7 Hamilton EcoRoof Substrate (soil) Composition
Parameter
Total As
Total Cu
Total Pb
Total Zn
Extractable As
Extractable Cu
Extractable Pb
Extractable Zn
Extractable NO3-N
Extractable NH4-N
TKN
Total Phosphorus
Extractable PO4-P
Extractant
DTPA
DTPA
DTPA
DTPA
1NKC1
1NKC1
0.5 N NaHC03
Method Number
EPA 200.9
EPA 200.7
EPA 200.9
EPA 200.7
EPA 200.9
EPA 200.7
EPA 200.9
EPA 200.7
SM 4500-N03 F
SM 4500-NH4
EPA 35 1.4
EPA 200.7
SM 4500 P E
Unit
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
mg/kg
Eastside
4.54
17.5
5.57
48.2
0.01
1.25
0.26
4.9
253.6
2.7
1897
958
100
Westside
2.19
30.3
64.9
146.1
0.09
6.08
2.43
64.8
798.3
28.6
12802
2508
325
Ratio
0.5
1.7
11.7
3.0
9.0
4.9
9.3
13.2
3.1
10.6
6.7
2.6
3.3
Equipment
Flow monitoring equipment includes a small 60-degree, V-trapezoidal Plasti-Fab flume, and a hydraulic
bubbler-type flow meter, which measures the water level in the flume as shown in Figure 8 . The flumes
were custom made to attach to the two main drainage points. This data is instantaneously transmitted to the
BES Lab where it is converted and stored on the BES computer network.
BES has been testing another type of flow monitoring equipment. It is a small mobile Sigma flow meter,
Model 950, configured with a bubbler-type level sensor. It appears very small flow levels can be captured
with this type of meter. Data is stored in a mobile data logger. Figure 8 shows a BES staff installing the
added equipment. Figure 8 also shows the flume with the bubbler tubes, one connected to the data logger
and the other connected to a transducer that telemetrically sends flow and rainfall data directly to the lab.
A rain gauge was installed on the building to ensure that accurate rain data is collected for the site. This
data is collected and stored, then accessed via computer on the city network.
.1
'if
Figure 8 . Two photographs of one of the monitoring stations. One shows the data logger and bubbler with the flume
and the second is a close-up of the flume.
-------
Flow Monitoring
Initially BBS had a lot of problems with the flow meters, but has since corrected these problems and added
the two new meters. Currently each drain has two meters and both are showing comparable results. All data
collected since December, 2001 is considered good. The graphs represent 75 % EcoRoof and 25 %
impervious. Figure 9 shows the runoff associated with the long duration winter storm event and the slow
release water that cannot be retained in the saturated substrate. Note the mitigation of the peak intensities of
the event. Figure 10 shows the almost complete retention of a typical Portland summer storm. The
estimated runoff from a conventional roof surface would be very similar to the rainfall lines as the rainfall
would almost immediately turn to runoff as the rain occurs. An almost % in. storm was mostly retained on a
roof with 4 in. of soil.
Hamilton Apartments Ecoroof, westside flows
Storm event Jan. 4 - 8, 2002 (2.8")
0.0
1 103 205 307 409 511 613 715 817 919 1021 1123
time, 5 min. intervals
Figure 9. A chart of measured Hamilton Apartment EcoRoof Westside (2 in. soils) rainfall versus runoff, in units of
cubic feet per 5 minutes versus 5-minute time increments showing the reduction in runoff volumes and peaks for a
winter storm.
Water Quality Monitoring
To date, five storms have been monitored for water quality. The results are encouraging, but also show how
attention to substrate chemical composition may be needed (see Figure 7) depending on the receiving water
system.
Sampling Procedures
BBS Field Operations staff performed sample collection and field parameter readings. The BBS Laboratory
section performed the analytical testing. The minimum storm criteria for water quality analysis for this
project was 0.25 inches of rain in 24 hours to ensure runoff volumes are sufficient. Grab samples were
collected at the middle to latter part of the storms. The water quality grab samples were collected at the
termination of the flumes using a decontaminated stainless steel bailer or the sample container directly.
-------
Hamilton Ecoroof westside rainfall and runoff
June 28-29, 2002 storm event 0.73"
100
80
60
= 40
Time - hours
Figure 10 . Measured Hamilton Apartment EcoRoof Westside (4 in. soils) rainfall versus runoff in units of gallons-
per-hour versus time in hourly increments, showing the significant reduction in runoff volumes and peaks for a
summer storm.
Analytical Parameters
Samples were analyzed in the field for dissolved oxygen, pH, specific conductance, and temperature using
portable field meters. Samples were submitted to the laboratory for analysis of ammonia-nitrogen,
biochemical oxygen demand, chemical oxygen demand, color, total and dissolved metals (arsenic,
cadmium, copper, lead, silver, and zinc), Escherichia coli, orthophosphate-phosphorus, total phosphorus,
and total suspended and dissolved solids.
Figures 11 and 12 show constituents such as Total Phosphorous and Ortho-phosphorous at concentrations
above receiving water standards. Note the difference between Eastside and Westside flow concentrations
and the substrate chemical composition shown in Figure 7. It appeared that over time phosphorus levels
might be coming down, but there was a spike in one of the samples in the last storm. We believe that the
phosphorus issue can be corrected by being careful to specify a substrate, which would not allow excessive
amounts of TP to release from the soil or in fact one that might tend to retain phosphorus. Another issue is
the contribution of certain constituents from the terrace area. Numerous activities occur with lots of food,
drinks, fireworks, dogs and many other pollutant sources. Obviously these sources may affect some of the
characteristics of the water quality due to human and other impacts. One important lesson to date is that
these sources should be addressed in monitoring studies, either by conducting studies where they do not
exist or by education efforts. This is the only EcoRoof the City is monitoring for water quality at this time;
others will be monitored in the future.
Another important characteristic is the EcoRoof affect on loadings. As shown above, many storm events,
especially the warm season storms, significantly reduce flow volumes, thus reducing loadings. And in
many cases the flow is zero with zero concentrations, particularly during the drier times of year.
-------
Figure 13 shows dissolved copper concentrations which, based on water hardness, are usually below in-
stream standards. Again attention to substrate ingredients and materials to be used on the roof can affect
these parameters. For example, the roofing industry uses lots of galvanized metals, copper and lead. It is
unknown whether the wood was treated with copper, a potential source for copper on the Hamilton building
was treated lumber the landscape contractor used for edging material. However, as pointed out above, the
copper loadings would be much reduced as compared to a traditional roof. One option that should be
evaluated in reducing pollution from all roofs is the types of roofing materials that are allowed. Several
projects in Southern California (Crystal Cove, Newport Beach for example) have restrictions on copper and
zinc containing materials being used for roofs, gutters, and downspouts.
Hamilton Ecoroofs
1.2
Total Phosphorus mg/L
0.8
0.6
0.4
0.2
1.11
J0.7
10.47
0.35
0.29
[0.24
2/22/2001
3/25/2001
4/23/2001
least • west
5/14/2001
2/7/2002
Figure 11. A chart of showing the total phosphorus concentrations measured in roof runoff from both the east and
west roof areas. There is a deceasing trend in phosphorus concentrations with the exception of the west roofs last
sample.
What else have we learned?
Almost an inch of soil was lost to wind erosion, especially on the east side. The initial planting did not
provide good vegetative cover in all areas, which could have protected against this erosion. Depending on
the initial planting scheme, cover crops such as common clover may provide excellent soil coverage. Water
from air conditioning condensate is a possible source of free, non-potable water for irrigation. Condensate
flows were significant during the hottest part of the summer, with flows measured at 12 oz.-per minute in
the afternoon and 6 oz.-per minute in the late evening. This might prove to be a free source of irrigation
water, if considered during the design phase. Mosses have populated certain areas of exposed soil and
helped reduce wind and soil erosion. Lightweight soils must be fully covered to prevent erosion. The
eastside is now only about 2 in. thick and the west side is about 4 in. thick. A small colony of ladybugs has
been observed in the south half of the eastside and numerous other insects. Hummingbirds, blue jays, crows,
swallows, pigeons, sparrows, and signs of hawks or owls have been observed.
-------
Hamilton Ecoroofs
orthophosphate phosphorus mg/L
1.2
0.6
0.4
0.2
In fi?
10.43
10.32
2/22/2001
3/25/2001
4/23/2001
5/14/2001
2/7/2002
I east west I
Figure 12. A chart of showing the orthophosphate concentrations measured in roof runoff from both the east and
west roof areas. There is a deceasing trend in phosphorus concentrations with the exception of the west roofs last
sample.
Hamilton Ecoroofs
Dissolved Copper ug/L
12.00
10.00
8.00
6.00
4.00
2.00
0.00
Acute Standard at 250 mg/l
=31.9ug/l
^10.4
10.5
Acute standard at Hardness of 50 mg/l
/7.40
«8-
^6.43
2/22/2001
3/25/2001
4/23/2001
east*west
5/14/2001
2/7/2002
Figure 13. A chart of showing the dissolved copper concentrations measured in roof runoff from both the east and
west roof areas. Most samples (8 of 10) were below acute water quality criteria at a hardness of 50 mg/l.
-------
Soil and Vegetation Monitoring
As a part of our monitoring the city is photo documenting the EcoRoof vegetation on a regular basis in
addition to documenting changes and problems with the soil. The vegetation has gone through seasonal
changes yet has continued to grow and cover the soil. Some problems have included volunteer grasses,
plants and clover. The volunteer plants alone are not a large problem since in most cases they will not live
through the summer without irrigation.
Air Temperature
In addition to stormwater monitoring discussed above, some energy-related measurements have been
conducted. For example, the City has been comparing inside and outside temperatures of the garage
EcoRoof and found that EcoRoofs appear to provide cooling benefits. There is no insulation on the garage
except for the EcoRoof.
Demonstrations and Incentives
The BES has provided incremental funding for three projects to date, but not the residential garage. Funds
are obtained from BES sewer and stormwater revenues. The rationale for public funds being used is that
these projects will help the City determine the stormwater and CSO management values of EcoRoofs. In
addition, the City now allows builders to exceed building height restrictions with the implementation of a
EcoRoof. In addition, there is a stewardship grant program which, to date, has provided funding for four
projects. In the future, credits on stormwater utility fees will also likely be put in place. Finally, the
EcoRoof can be used to meet or partially meet stormwater treatment requirements.
Other Lessons - Buckman Terrace
Buckman Terrace is a redevelopment project by Prendergast Associates. The project was designed in 1998
and opened in 2000. This is a 0.8-acre site with 150 apartment units, with all below-building parking and a
1,500 ft2 commercial section in a 4-story structure. The building also has car sharing and numerous other
environmental attributes.
The entire building has a roof area of approximately 25,000 ft2 and is constructed with sufficient structural
capacity to hold an EcoRoof. As a test, EcoRoofs were placed on two sections. Figure 14 shows the main
EcoRoof, which comprises over 1,500 ft2 of commercial space that has full solar exposure. An additional
750 ft2 of impervious roof area drains onto this south facing EcoRoof. Figure 15 shows the entrance
EcoRoof, which is also planted with sword fern, licorice fern and white stonecrop. It is on the eastside and
is in the shade of a north-facing wall. Both were planted in March 2000. The main EcoRoof was planted
with two species of Oregon sedum, various wildflowers, native grasses and a few licorice ferns. Grasses and
wildflowers were planted from seed and mulch was hand broadcast to protect against wind erosion. An
irrigation system has not been installed for either EcoRoof. The soil profile is 4 in. deep and 20 Ibs ft2 when
saturated. American Hydrotech waterproof membrane and reservoir drain system was used. BES staff
specified the soil mix and vegetation.
While the grasses and wildflowers achieved a graceful, flowing appearance, they are reminiscent of an
Eastern Oregon or Midwestern American prairie. Since residents who would rather have a "greener look,"
-------
for the EcoRoof, the roof is going to be replanted. The Fire Department was also concerned about the dry
grasses, which is an important issue for EcoRoofs without irrigation systems.
During the warm season, storm event runoff was visually observed to be very low or non-existent. The
EcoRoof has capacity to hold much of the additional flow from the other roofs. During winter storms,
runoff occurs often, but it is detained and released slowly. Many of the plants survived or re-seeded
themselves with only one hand-watering. Although no maintenance was conducted this last year, it appears
the grasses will need to be mowed at least once a year.
Figure 14. Two photographs of the Buckman Terrace EcoRoof showing uses of grasses and wildflowers.
Figure 15. Buckman Terrace EcoRoof at the building entry with protection from north facing wall
-------
Summary
In initial sampling, EcoRoofs have been shown to significantly reduce runoff volumes, especially in spring,
summer, and fall. They also help to slow runoff during winter periods.
In addition it appears that water quality could be significantly improved via loadings (volume) reduction as
well as pollutant removal/avoidance. Additional monitoring data on EcoRoof water quality will be
conducted by the city to assess the benefits of concentration reductions, and the loading reductions from
reducing runoff amounts. There is a need to be strategic about the selection of soils/growing media to use
on EcoRoofs as some soils may contain higher levels of pollutants. In addition other roof materials, such as
treated woods need to be avoided.
Developers in Portland are gaining confidence in the value of EcoRoofs, as more and more builders gain
experience with EcoRoof design and construction. The City allows developers to meet or partially meet
their stormwater treatment requirements with an EcoRoof. In dense urban situations, this has become more
and more attractive to developers. In addition the City allows taller buildings as an incentive. In the future,
there will be a potential reduction in stormwater fees via a reduced fee for those sites with EcoRoofs. One
of the primary reasons that developers are embracing the program is the City's technical and permitting
assistance provided by the Bureau of Environmental Services.
As with any stormwater management measure, good design and maintenance are keys to their success. It is
expected that, due to virtual elimination of sun energy on roof surfaces and resulting degradation of roof
materials, that EcoRoofs will be likely found to last much longer than many traditional roof materials. As
with any roof, good construction techniques are important. The City is undertaking economic analyses of
life cycle costs and benefits of EcoRoofs to be able to further demonstrate their value and effectiveness to
developers and the community at large.
Reference
May, C.R., R. Horner, J. Karr, B. Mar, and E. Welsh. 1997. Effects of Urbanization on Small Streams In
the Puget Sound Lowland Ecoregion. In: Watershed Protection Techniques, 2(4): 483-494.
-------
Storm Water Management in the City of Chicago
Suzanne Malec
City of Chicago Department of Environment
Chicago, Illinois
Abstract
The City of Chicago owes its very existence to its location at the confluence of the Chicago River and Lake
Michigan. Lake Michigan provides the City with an abundant water supply while the Chicago River serves
as a highway to move goods and services critical to the City's growth. Chicago has built a historic legacy in
protecting these valuable water resources. To protect its water supply, engineers in the 1900s constructed the
Chicago Sanitary and Ship Canal to reverse the Chicago River's natural flow from eastward to westward,
steering human and industrial waste away from Lake Michigan. In 1972, Chicago pioneered the use of deep
tunnels to capture, convey, and store combined sewage during storms for later treatment.
Today, Chicago is taking a new comprehensive approach toward further improving the quality of its surface
waters. Rather than through large scale engineering projects, the approach centers on simple storm water
Best Management Practices (BMPs) at the source level to reduce the negative impacts of storm water runoff.
Through various model projects, the City aims to demonstrate the efficacy of various BMP approaches,
promote public acceptance and usage, and encourage modification of local ordinances to allow wide-spread
usage of BMPs.
History of Storm Water Management
In 1885 a severe rainstorm caused sewage-contaminated river water to flow into Lake Michigan,
contaminating the City's drinking water. This disaster led to a cholera and typhoid outbreak that killed over
90,000 people. Repeated outbreaks of epidemic diseases compelled the City to find a way to stop the flow
of polluted water into Lake Michigan. The Metropolitan Sanitary District of Greater Chicago was created in
1889 to safeguard the city's drinking water and determine an acceptable way to dispose of waste.
In 1900, the sewer overflow problem was solved by a massive engineering effort. Engineers constructed the
Chicago Sanitary and Ship Canal to reverse the Chicago River's natural flow from eastward to westward,
thereby steering human and industrial waste away from Lake Michigan. Now the river flows into the
DesPlaines River, the Mississippi River and, eventually, the Gulf of Mexico. Locks regulate the elevation
of the river and prevent Lake Michigan from draining freely (City of Chicago, 2000).
While this solution protected the Lake, it did not reduce the pollution level in the Chicago River. Rainfalls
of as little as 1/3 inch overloaded local sewer systems and caused combined sewer overflows (CSOs) - a
mixture of storm water runoff and raw sewage, into the waterway. Hundreds of CSOs are located along the
waterway. CSOs still polluted the waterways and, with the heaviest rainstorms, raised flood stages to levels
resulting in river backflows into Lake Michigan, causing beach closures. Major underlying causes of the
problem were lack of an adequate floodwater outlet and increasing urban growth.
In 1972, the Metropolitan Water Reclamation District of Greater Chicago (formerly Metropolitan Sanitary
215
-------
District of Greater Chicago) started construction of a large scale, multi-purpose Tunnel and Reservoir
Program (TARP), comprised of deep rock tunnels and surface reservoirs that capture, convey, and store
combined sewage during storms until it can be transferred to existing treatment plants when capacity
becomes available.
In 1974, prior to TARP, only 10 fish species were found in the Calumet and Chicago River systems. With
improvements in wastewater treatment technology, the species count rose to 33 by the early 1980s. In 1984,
the first TARP tunnel projects came online, reducing the frequency and volume of combined sewer
overflows. Subsequently, the species count rose gradually to 54 by 1990, and had reached 63 by 2000. This
steady climb over the years is due in part to additional segments of the TARP tunnels coming online, further
improvements in treatment plant performance, and supplemental aeration of the waterways (EPA Region V,
2002).
Today, increased residential and commercial development is ocurring along the banks of Chicago
waterways. The waterways are no longer considered just navigational canals, but are seen to be amenities or
center pieces of urban life. The public's interest in the river has grown, as evidenced by the increasing
numbers of paddlers, walkers, bikers, and even jet skiers on the river. Fishing on the river has also grown in
popularity. Fish consumption advisories still remain in place, however, and large portions of the rivers are
not safe for full body contact. Additional work remains to be done.
Current Storm Water Management Approach
The City is taking a new comprehensive approach toward further improving the quality of its surface waters.
Rather than through large scale engineering projects, the approach centers on implementing and promoting
demonstration projects that utilize simple storm water Best Management Practices (BMPs) at the source
level. The goals of these BMPs are to reduce the quantity and improve the quality of urban storm water
runoff.
Common Storm Water BMP Techniques
Storm water pollutants includes such substances as solids, metals, oil and greases, and road salt. BMPs
commonly employed in Chicago's model projects to treat storm water runoff include vegetated swales,
infiltration trenches or basins, detention basins, mechanical filtration/sediment and oil grease traps, rooftop
gardens, and cisterns that capture runoff for gray water use. A brief description of some of these BMPs are
described below.
Vegetated Swales - In vegetated swale designs, storm water is conveyed through a vegetated swale instead
of a storm sewer. Swales increase storm water infiltration potential and storage. Swales also remove
pollutants via settling, vegetative filtering, and to some extent infiltration through the soil. Sediments need
to be periodically removed from vegetated swales, and the vegetation mowed and replanted as needed
(NIPC 1995).
Infiltration Trench or Basin - In an infiltration trench or basin, storm water runs through a swale or into a
basin that has a porous bottom (sand or gravel), causing storm water to infiltrate into the ground. As the
storm water percolates through the ground, contaminated particles are trapped within the soil and the
216
-------
resulting treated water migrates to the groundwater. Water quality benefits are derived from the removal of
contaminants that are sorbed onto soil particles and decreased flows into the river. Sediment will tend to
clog systems unless the systems are routinely maintained. The condition of the trench should be
periodically checked and the accumulated sediment removed. After years of operation, the stone in the
trench may need to be removed and cleaned and the filter fabric replaced (NIPC 1995).
Detention Basin - In a detention basin, storm water enters a basin that has a structure to control outflow.
The water quality benefits result from attenuation of flows by slowing the velocity of water and removal of
solids by settling due to lower water velocities. Effectiveness is greatest for suspended sediments such as
heavy metals. Lower effectiveness is expected for soluble constituents and nutrients. Oil and grease
typically pass through, unless the detention basin is planted with vegetation in a manner that leaves no open
water flow paths from one end to the other. Sediments need to be removed periodically, and vegetation
should be mowed and replanted periodically (NIPC 1995).
Sediment and Oil and Grease Traps - In sediment and oil and grease traps, storm water runs through a
structural device that has a chamber that traps oil, grease, and sediment. The solids need to be removed
periodically. The advantage of this design is that oil, grease, and sediment are trapped at a location that is
easily accessible to maintenance crews. Water entering the chamber could pass over and under a series of
baffles. Baffles at the bottom of the chamber could trap sediment, and baffles at the top could trap oil and
grease.
Rain Gardens (bioretention cells) - Rain gardens have native plant amenity features and provide for 1he
infiltration of excess rain water from impervious surfaces. Native plants have root systems that are deeper
than typical turf grasses, and provide greater absorptive capacity not only into the plant but also into the soil.
Rain gardens are not meant to treat heavily polluted runoff, nor are they designed to absorb maximum
rainfall. Instead, they are designed to mitigate local and downstream flooding problems by providing space
for excess runoff to be absorbed into the soil or to slow the velocity of the runoff as it passes through the
remainder of the storm sewer infrastructure.
Model Projects
Working together, City departments have conducted specific model projects at the municipal, residential,
commercial/industrial, and public infrastructure levels. Each project utilizes one or more of the
aforementioned BMP techniques. Through these model projects, the City aims to demonstrate the efficacy
of various BMP approaches, promote public acceptance and usage, and encourage modification of local
ordinances to allow wide-spread usage. Some examples of model projects conducted by Chicago are
described below.
Municipal Facility Projects
City Hall Rooftop Garden - The City Hall rooftop garden encompasses 20,000 square feet of planted area
and includes more than 150 species of native plants. The roof system was designed to carry 1-inch of
precipitation. Aside from the storm water benefits, green roofs lower ambient air temperatures in the
summer, provide better insulation which reduces energy demands, and provides animal or insect habitat.
The project was selected for a pilot to study the benefits of green roof systems. The project also includes the
217
-------
development of prototypical guidelines and specifications that can be used elsewhere, and conduction of a
study quantifying the environmental benefits of green roof systems. Lessons learned from the project were
incorporated into the City's A Guide to Rooftop Gardens booklet. The booklet is targeted to the general
public to promote construction of green roofs in the City.
Chicago Center for Green Technology - This city building was renovated to serve as a model for an energy-
efficient and environmentally friendly design. The City expects to receive a Platinum Certification under
the U.S. Green Building Council's Leadership in Energy and Environmental Design (LEED) Program.
Storm water BMPs employed at 1he site include a functional green roof system, cisterns (capturing up to
12,000 gallons of roof runoff), sheetflow of parking lot runoff to vegetated swales, and a storm water
detention area.
Residential Projects
Downspout Disconnection Campaign - Individual residents are being encouraged to disconnect their
downspouts, blocking their sewer connection and redirecting the rainwater from their roofs to adjacent
landscaped areas. This reduces runoff flow to the combined sewer system, promote groundwater recharge
while supporting local green spaces. During summer 2002, the City canvassed flood prone areas of the city,
distributing door hangers and brochures to houses which were considered appropriate for downspout
disconnection. The City will be promoting the use of rain barrels in conjunction with the downspout
disconnection campaign. Gutters could be drained into rain barrels, storing rain water for later irrigation
use.
Model Rain Gardens - Model rain gardens are being built in City parkways to absorb additional rainwater
during heavy rain periods. Including French drains installed below ground level and plants that can
withstand extreme wet and dry conditions, twelve such gardens have been installed in a flood-prone area.
These rain gardens were installed to receive runoff from sidewalks and roof areas. Large rain gardens are
being planned for the future that will be connected to curb cuts to absorb additional capacity from roads.
Commercial/Industrial Projects
Ford Centerpoint Supplier Park- Ford Motor Company operates a car-manufacturing plant in the Calumet
area. Ford is currently finalizing plans to build a supplier park adjacent to their existing facility. This
development, which will eventually consist of 1.7 million square feet on 150 acres of land, has the potential
to exemplify how industry and environment can co-exist. The purpose of the development is to reduce
transportation costs and pollution from long ground delivery distances, and provide a just-in-time
manufacturing source of materials for the plant.
A range of innovative, conservation-minded options will be implemented to improve water quality, decrease
heavy runoff to the creek, and prevent pollution. First, the development will utilize a separate storm water
and sanitary sewer system. All storm water runoff from rooftops and parking lots will be routed into
vegetated swales. Swales will contain native vegetation that filters the water as it is conveyed. Storm water
runoff from public streets that will be constructed to accommodate the development will drain into roadside
swales through curb cuts. Although the swales will be privately owned, a drainage easement will be granted
to the City.
218
-------
The swales will empty into vegetated detention basins for treatment, then be conveyed to a wetlands area
and finally into Indian Creek. This design will slow the pace of movement of water into the creek,
removing harmful contaminants and decreasing the erosion often caused by major storm events. The entire
campus will be planted with shortgrass prairie, tallgrass prairie, and native trees.
Public Infrastructure Projects
130th_ and Torrence Intersection - The City is reconstructing the intersection of 130th and Torrence Avenue.
As part of this project, both streets will be depressed. Storm water from a rain event will be collected in an
underground chamber and then pumped to the Calumet River. The City is considering a variety of treatment
options for the storm water before its discharge to the river. These options involve selecting the right
combination of BMPs in series that will treat the runoff most effectively and at the least cost. The options
include a treatment train of sediment, oil, and grease traps, followed by vegetated swales, infiltration
trenches, and a wetland detention basin. The most efficient system is expected to remove 98% of total
solids, 88 % of oils and greases, and 40% of the road salt from the runoff (Tetra Tech 2002).
South Lake Shore Drive Project - South Lake Shore Drive is an important part of the City's transportation
system. It is an essential commuter link between the downtown area and the City's south side. Heavy
traffic and seasonal weather contrasts have led to crumbling road conditions on the drive. The City of
Chicago, Illinois Department of Transportation, and the Federal Highway Administration are investing $162
million to reconstruct more than 6 miles of the roadway. More than 14 acres of green space enhancements
will be included in the reconstruction efforts, including new median landscaping, trees, shrubs, perennials,
and ornamental grasses.
City engineers also looked at better management of storm water runoff from the drive to protect the water
quality in Lake Michigan. Prior to the reconstruction, storm water from the road was directly discharged to
the lake. In contrast, all the storm water runoff in the newer North Lake Shore Drive is directed to the City's
sewer system. Unfortunately, this sometimes overwhelms the system, causing sewage to backup onto the
drive.
As an alternative, City engineers are utilizing a system that directs only the first flush of the South Lake
Shore Drive runoff to the sewers. Remaining flow, which will be generally cleaner, will be discharged to
the lake. Diversion of the first flush helps reduce the flow into the City's combined sewer system and
thereby improve the quality of the runoff discharged into the lake. Once the reconstruction is completed, the
City will monitor water quality in the outfalls to see if modifications to the system are needed.
Infiltration Alley - In the Fall of 2001, the City reconstructed an asphalt alley using a permeable system.
The new alley has eliminated formerly chronic local flooding without using the sewer system and reduced
the "heat island" effect by eliminating dark, heat-absorbing surfaces.
The City used Gravelpave2™, a porous gravel structure, manufactured by Invisible Structures, that contains
gravel and provides heavy load bearing support, unlimited traffic volume, and indefinite parking duration.
In one 40 in. x 40 in. section of the structure, there are 144 rings made of highly durable plastic, each 2
inches in diameter and 1 inch high and held together underneath by a geo-fabric layer. The section below is
a 10-inch thick, compacted aggregate base course consisting of a 2/3 stone and 1/3 sand mixture. The new
system can handle up to 3" of rainfall per hour, allowing rainwater to soak into the ground and thereby
219
-------
reducing polluted run-off and flooding. The system is suitable for traffic, including residential and service
vehicles.
Rain Blacker Program - Rain Blocker is Chicago's program of installing "vortex" type restrictors in sewer
inlets to regulate the rate of storm water runoff entering the sewer system. The system is designed to keep
sewers flowing at capacity without backing up. The excess water remains on the street longer instead of
backing up in basements or causing CSOs.
Summary
Of course, no one project provides all of the answers. Rather, a combination of the above model projects,
implemented on a City-wide and case by case basis, could reverse current trends of urban infrastructure, and
thereby dramatically improve water quality.
Next Steps
In the coming year, the City will continue to implement model projects that demonstrate effective
management of storm water without requiring additional cost over more traditional methods. The City is
also working with the Northeastern Illinois Planning Commission in preparing an urban BMP booklet
designed specifically to educate and engage landowners in thoughtful, proactive storm water management
approaches. A variety of educational and regulatory programs are also being considered, in addition to
monitoring programs to assess the efficiency and replicability of our model projects.
References
City of Chicago website "http://www.cityofchicago.org/Environment/Rivertour/"
Environmental Protection Agency Region V, September 2002 "Chicago Area Rivers Restoration Initiative"
EPA proposal to the Urban Rivers Initiative.
Northeastern Illinois Planning Commission, August 1995 "Best Management Practice Guidebook for Urban
Development.
Tetra Tech EMI June 2002 "Calumet Region Green Infrastructure NFS Demonstration Project" Section
319(h) Grant Proposal.
220
-------
REGIONAL FACILITY VS. ON-SITE DEVELOPMENT REGULATIONS:
INCREASING FLEXIBILITY AND EFFECTIVENESS IN
DEVELOPMENT REGULATION IMPLEMENTATION
Miranda Maupin
and Theresa Wagner1
City of Seattle
Seattle, Washington
Abstract
Development regulations can sometimes be challenging to implement in ultra urban environments due to
limited space, high land value, and the expense of retrofitting existing infrastructure. In addition,
development patterns may not always correspond to high priority surface water management zones.
Development-driven basin planning combined with regional detention and water quality facilities can be
tools for locating surface water management investments strategically to protect aquatic resources while
creating livable communities. This presentation highlights policy, legal, finance and technical issues and
opportunities associated with a Seattle case study. The case study will help prompt discussion regarding the
effectiveness of this strategy as a tool for surface water managers in urban jurisdictions to meet multiple
interests and put limited stormwater management dollars to effective use.
A. Introduction
For purposes of discussion, this paper defines an off-site mitigation program as a program offered by a
municipality that allows developers to meet on-site development requirements relating to stormwater by
compensating the municipality to provide equivalent mitigation in an off-site public facility. Under this
scenario, the municipality clearly assumes additional risk and responsibilities, and even perhaps additional
costs, so why would a municipality consider such a program? Municipalities might consider offering an off-
site mitigation program if:
• The municipality has planning, capital or performance stormwater management obligations, as well as
authority to regulate development, and
• On-site stormwater management is required for new development or redevelopment projects, and
• Cost, environmental performance or community benefits can be gained by meeting the on-site
requirements off site.
A survey of 26 local jurisdictions in Washington State revealed that jurisdictions are quite interested in
understanding how to implement a program, and 9 jurisdictions have even implemented elements of a
program. However, no jurisdiction had as yet developed a systematic, programmatic approach that
addresses the key issues. This paper presents a discussion of the following issues organized around three
areas of responsibility: municipal drainage management, NPDES permit compliance, and development
regulation authority.
1 Ms. Maupin is a Senior Planner with Seattle Public Utilities of the City of Seattle. Ms. Wagner is a Senior Assistant City
Attorney with the Seattle City Attorney's Office. April Mills, as an intern with Seattle Public Utilities, contributed research and
analysis that assisted in the development of this paper. This paper represents solely the views of the authors and not of the City of
Seattle or any of its elected officials or departments.
221
-------
Table 1 - Key Issues Associated with Implementing an Off-Site Mitigation Program
Issues
Municipal Drainage Management
^ On-site vs. Off-Site
•S Development vs. Retrofit
•S Funding Options and Authority
•S Off- Site Mitigation Fee Structure
NPDES Permit Compliance
•S NPDES Permit Requirements and
Regulatory Authority
•S Point of Compliance
•S Environmental Protection
•S Timing
Development Regulation Authority
•S Applicability
Key Question(s)
When could a municipality consider offering an off-site mitigation program for on-site development
requirements?
What are the technical trade-offs for a municipality between on-site mitigation and off-site mitigation
of development impacts to stormwater?
Why might a municipality consider offering an off-site mitigation program for on-site development
requirements?
Would municipally-constructed facilities address only mitigation triggered by development, or would
the facility address existing runoff?
What are the funding option(s) and associated authority necessary?
How would a fee for off-site mitigation be calculated? How important is it for a municipality to
recover the full cost of the facility through fees?
Does the jurisdiction's NPDES municipal stormwater permit require the jurisdiction to regulate
development to mitigate stormwater impacts? Does the jurisdiction have legal authority, and
leeway under its NPDES permit, to allow off-site mitigation?
What legal risks should be evaluated when considering an off-site mitigation program?
How is the municipality's point of compliance determined for evaluating performance?
How is the regional facility determined equally or more protective than on-site projects?
What is the timing of development and regional facility construction? What if the development
occurs before the regional facility is constructed— leaving a window of time during which runoff is
uncontrolled?
How is applicability established for the program? To which developments is an off-site option made
available? How are developments handled that are not upstream of a planned or constructed
facility?
In the next section, this paper will provide a Seattle context, including the regulatory background, some
local drivers that invite further examination of off-site mitigation in Seattle, and a case study overview. The
following section of the paper will provide discussion of the key issues associated with off-site mitigation,
using the Seattle case study as an example to walk through the policy and legal implications of the issues
identified. Finally, the paper concludes with some thoughts on when regional off-site mitigation makes
sense and ideas for how these opportunities fit into the traditional basin planning framework.
222
-------
B. Background, Context and Case Study
Seattle Context
The Greater Seattle Area is Washington's largest urban center covering 60 square miles and a population
over 3 million and growing. Over the past 30 years, the region has grown nearly twice as fast as the national
average. The City of Seattle, itself, is just over 500,000 and fully developed with very few remaining parcels
that have not yet been developed. Known as the 'Emerald City,' Seattle is surrounded by water and
mountains on all sides. Functioning almost like an island, Seattle drains to the Puget Sound to the West,
Lake Washington to the East, the Duwamish River to the South, and Lake Union in the middle.
As a local government, the City of Seattle is multifaceted. In addition to possessing local police powers and
regulatory authority for land use and development, the City includes utility departments: Seattle Public
Utilities (providing drainage, wastewater, drinking water, and solid waste utility services) and Seattle City
Light (providing electric service). Seattle is characterized by a complex drainage infrastructure,
administered by Seattle Public Utilities. Nearly 1/3 of the City is the traditional combined system
conveying both stormwater and wastewater to the regional wastewater treatment facility operated by the
County, with the City's combined sewer overflows regulated by Washington State under a CSO NPDES
permit. The remainder of the City is regulated under the municipal separate storm sewer system ("MS4")
NPDES permit draining to the surrounding water bodies through more than 200 drainage basins. These
basins range in size up to 7,000 acres, though half of the basins are less than 100 acres in size and drain
through piped infrastructure directly to large receiving water bodies. About one-third of the jurisdiction
drains via informal "ditch and culvert" conveyance system to creeks and then to the surrounding water
bodies.
Politically, Seattle has generally tried to encourage development within the City particularly in downtown
and the urban villages designated for additional growth under the City's comprehensive planning. This
development is with few exceptions redevelopment—that is replacing existing impervious surface with
greater density. As the city densifies, demands have increased for public transit, affordable housing, and
pedestrian oriented retail with a number of civic scale projects in planning, design or construction. Seattle's
urban character is strongly influenced by its neighborhoods with a priority in recent years to coordinate City
improvements, including infrastructure, open space and pedestrian amenities, around neighborhood plans.
Seattle residents tend to support environmental values, with a particular interest in protecting and enhancing
the urban creeks, as demonstrated through several community-initiated watershed action plans.
Regulatory Context
Since 1995, six Washington entities have been covered by watershed-based general NPDES Phase IMS4
permits issued by the Washington State Department of Ecology ("Ecology"): City of Seattle (with one co-
permittee), City of Tacoma, King County, Pierce County, Snohomish County, and the Washington State
Department of Transportation; Clark County's permit differs slightly.
The 1995 MS4 permits required each municipality to create a stormwater management program ("SWMP")
which had to be approved by Ecology by a certain date during the permit term. The permits required
adoption of development regulations, source control efforts, enforcement of Stormwater Code pollutant
prohibitions, coordination with other jurisdictions, education, planning and reporting. The permits also
required compliance with state water quality standards but provided that "development and implementation
of approved stormwater management programs represent ongoing efforts towards meeting those standards
223
-------
on an approved compliance schedule .. .." The permits required each Phase I local jurisdiction to adopt a
set of ordinances regulating the stormwater impacts of new development and redevelopment, during and
after construction. Less typically, the SWMPs and ordinances were required to be approved by Ecology as
being "equivalent" to the 1992 state stormwater management manual guidance issued by the state. The
manual addresses both flow and quality of stormwater discharges from developed sites. Municipalities have
had varying experiences obtaining timely Ecology approval of the SWMPs and of development ordinances.
Ecology staff expressed frustration at the staff time required for individual municipal review, and
municipalities chafed at the mandate to use local regulatory powers subject to Ecology approval.
Ecology's 1995 MS4 Phase I permits still cover the seven jurisdictions, and Ecology has set the reissuance
effort aside for the time being in favor of other stormwater priorities. The state has not yet determined how
it would permit ports, drainage districts, or other entities that may fit the Phase I description, and Phase n
jurisdictions have not yet come under permit. Therefore, a patchwork of mandatory stormwater
development regulation exists in Washington State, with only the largest local jurisdictions currently
required by NPDES MS4 permits to regulate development in a certain manner.
In addition to Clean Water Act regulation, western Washington has been challenged since 1999 with
responding to threatened species listings of the Puget Sound chinook and of bull trout. The listings have
prompted independent action by the City and other local governments to preserve these aquatic species.
Ecology has voiced both a desire to tighten its regulation of MS4s and a fear of liability under the
Endangered Species Act for failing to regulate strictly enough.
The next Phase IMS4 permit may test the boundaries of regulation for municipal stormwater. Issues will
likely include whether the permit will require (1) compliance with water quality standards at MS4 outfalls or
at private development sites, (2) restoration of water quality or habitat within a defined period of time, (3)
stormwater planning with specified products which could form the basis for future permits, (4) land use
planning according to stormwater priorities, or (5) more rigorous local regulation and enforcement, possibly
requiring retrofitting or requiring municipalities to ensure compliance by private parties.
Seattle's on-site Stormwater, Grading and Drainage Control Code ("Code") development requirements are
found in the Seattle Municipal Code, Chapters 22.800-22.808, enacted by the City Council and in associated
rules adopted by City departments under administrative authority. (See
http://www.citvofseattle.net/dclu/codes/sgdccode.htm) In 2000, the City successfully and amicably
negotiated to obtain Ecology's approval of certain required elements, including on-site detention for sites
with 5,000 square feet of new and replaced impervious surface and on-site water quality treatment for sites
with 5,000 square feet of new, or one acre of new and replaced pollution generating impervious surface.2
Ecology has approved three options in the Code or rules for approving an alternative to on-site
requirements—each with provisions to demonstrate that a proposed alternative is equally protective of the
environment. Ecology agrees that the City may change its development requirements generally through
basin planning, "provided the level of protection for human health, safety and welfare, the environment, and
public or private property will equal or exceed that which would otherwise be achieved." Ecology has also
approved the City's process of granting an exception to a stormwater requirement on a project-by-project
basis "if the [City] determines that it is likely to be equally protective of public health, safety and welfare,
2 Pollution generating impervious surface includes areas subject to vehicular use, roofs that include zinc material, and landscaped
areas.
224
-------
the environment, and public and private property as the requirement from which an exception is sought."
And finally, Ecology approved the option to meet on-site water quality requirements off-site if there is a
City-approved integrated drainage control plan, which is "a drainage control plan that substitutes water
quality treatment from one or more projects through the design of and installation of offsite facilities within
a basin draining to the same receiving water body," accompanied by specific applicant contributions and a
construction start date within five years. The City has not yet asked Ecology to approve the option of off-
site flow control through an integrated drainage control plan.
Case Study Overview: Urban Center Re-development in Creek Watershed
A number of proposed civic-scale developments in Seattle, including large low-income housing projects,
several major transportation projects, and a few urban center developments, are worth considering for an
integrated drainage plan approach with off-site drainage facilities. One of the case studies being considered
is an urban center located in Seattle's largest creek watershed, (7,000-acres, 11 sq. miles) which drains to
Lake Washington. The watershed fabric consists primarily of single-family neighborhoods (with over
75,000 residents) intersected by several commercial arterials and a major interstate highway. The creek
demonstrates characteristically urban hydrologic patterns, with flashy uncontrolled storm flows and low
summer base flows. Flowing primarily through residential backyards, existing development is more often
within the 100 foot riparian corridor than not, and the banks are often reinforced to protect these buildings.
Despite encroachment and relatively poor benthic health, the creek hosts native vegetation and several fish
species, and the community has expressed interest in protecting and enhancing the creek by organizing a
community-initiated watershed action plan process among other efforts. The development regulations
described earlier are one tool for improving creek health. However, development patterns tend to be slow
and dispersed throughout the watershed save for a few areas, such as the urban center, expected to
experience more intense growth. For example, over a three-year period, 86 development permits were
issued in the watershed. Only 16 of these projects were large enough to trigger Ecology thresholds for
development requirements and totaled 4 acres out of the 7,000 acre watershed.
Although the urban center is currently fully developed, the center is expected to redevelop dramatically over
the next ten to twenty years with several civic projects, a large retail development and a major transit hub.
The community has developed a neighborhood plan expressing a vision of additional quality open space,
pedestrian-oriented streets, and civic center amenities including a library and community center. Much of
the area was developed prior to the current stormwater development requirements and thus drainage flows
directly to the creek without treatment or flow control.
In anticipation of this growth, the City is considering developing an integrated drainage plan to address the
drainage issues associated with the projected development at a sub-basin scale rather than a project-by-
project approach. The plan could help identify one or more sites to locate City-owned and City-operated
regional stormwater detention and treatment facilities within the sub-basin. Preliminary technical analysis
indicates a 2.5-acre site could potentially manage over 30 acres of drainage. The facilities could provide
management for both existing runoff from impervious areas not expected to redevelop, and runoff that will
be subject to development requirements. Thus, this project could be designed to accommodate future
partners that may use the facility to meet their stormwater treatment requirements. A partnership approach
could replace the need for numerous small, underground facilities with one larger facility that could provide
additional public amenities, such as landscaped open space with a trail extending the current creek trail
system and native landscaping.
225
-------
C. Discussion of Key Issues
The discussion of key issues is organized around three areas of jurisdictional responsibility: Municipal
drainage management, NPDES permit compliance, and development regulation authority. In addition, the
issues have been organized around a series of questions in the order a municipality might face them if
considering whether to offer an off-site mitigation program.
Municipal Drainage Management
S On-site vs. Off-Site
When could a municipality consider offering an off-site mitigation program for on-site development
requirements?
To successfully implement an off-site mitigation program, a municipality must possess both (1) sufficient
police power authority to plan for and regulate development -- typical of a local government -- and (2)
authority and responsibility for the quality and quantity of storm drainage, including compliance with any
NPDES municipal stormwater permit -- typical of a drainage or stormwater utility. Seattle has this
confluence of authority and responsibility, but this is not the case in many other local jurisdictions, where
local regulatory authority and drainage system authority are split between entities. Furthermore, options for
building and financing regional facilities are typically determined by state law, which may also constrain the
options for a municipality to receive funds in connection with approving construction or development.
Jurisdictions that lack complete authority may consider working with other jurisdictions by agreement,
undertaking joint projects, or seeking legislation to enhance authority.
What are the technical trade-offs for a municipality between on-site mitigation and off-site mitigation
of development impacts to stormwater?
The technical advantages and disadvantages of off-site mitigation vary under different situations. The table
below outlines a general checklist of pros and cons.
226
-------
Table 2 - Pros and Cons of an Off-Site Mitigation Program
Advantages
Disadvantages
Performance
Off-site location may allow more space intensive, but
superior performing technologies such as constructed
wetlands or bioswales.
If soil permits, infiltration technologies can perform
best if decentralized throughout the basin-
performance relies on sound maintenance practices.
Planning
Municipality has an opportunity to strategically locate
investments to address priority water body or known
water quality issues
The municipality must take on the responsibility of
determining where to site a facility based on priorities
and opportunities. Large regional facilities may be
difficult to site in urban areas.
Funding
Partnering may open up additional revenue sources
to fund more effective regional facility.
Partnering may complicate facility financing and not
fully fund the facility.
Maintenance
The municipality allocates staff to maintenance of a
few public facilities, rather than to review, inspection
and enforcement of multiple private facilities.
Increased assurance of maintenance over time.
Maintenance responsibilities are shifted to the
municipality, including disposal of hazardous waste
material.
Liability
The municipality takes on the responsibility for
managing the risk associated with changing the
location and party responsible for implementing water
quality requirements. Innovative local regulation or
funding may draw legal challenge or present permit
compliance issues.
Community
In facility siting and design, municipality can assist in
implementing community development plans for open
space, aquatic health and urban centers.
Community disagreement about use of public
resources and siting.
Why might a municipality consider offering an off-site mitigation program for on-site development
requirements?
Given the trade-offs outlined above, regional off-site mitigation is not advantageous in all circumstances.
Under what circumstances should a utility consider an off-site program?
In general if the off-site program can offer environmental, cost or community benefits that outweigh the
disadvantages, then an off-site approach should be considered.
Environmental—If analysis suggests that stormwater investments would be more effective located more
strategically -- either to address a more critical water quality issue, or to protect a higher priority water body.
In addition to flexibility in location, a municipality may have the opportunity to use a more effective
technology such as a biologically-oriented system that enhances treatment through plants and micro-
organisms.
Cost—Seattle, for example, has responsibility under its NPDES MS4 permit for reviewing, permitting,
inspecting and enforcing maintenance practices for privately developed stormwater facilities. These
responsibilities require staff time and associated resources and are likely to increase under future MS4
227
-------
permits. Municipalities might consider consolidating these costs in an off-site mitigation program if the
programmatic costs of administering on-site requirements over time outweigh the costs of the design,
construction and maintenance of a publicly owned structural facility. In some cases the municipality
already owns land for potential facilities that could substantially influence cost evaluations.
Community Goals—More often municipalities are being asked to play a role in the shaping of communities.
Growth management plans or other long-term development plans typically specify areas targeted for future
higher density development and other areas designated as green space to provide parks and protect
environmental resources. Municipalities can play a role in directing stormwater improvement, by
transferring investments from areas targeted for density to areas specified through regulation or community
goals for higher levels of environmental protection. In addition, municipalities can often integrate open
space goals into facility design to meet multiple goals in limited space.
In the Seattle case study, an off-site approach could fulfill both environmental and community goals. A
regional facility would be expected to provide better technology, target more critical flows and ensure better
maintenance over time. If no off-site program were available, high land value in the area would likely
drive developers to use multiple underground vaults to address stormwater requirements on site. In contrast,
a regional facility could offer constructed wetland technology with a downstream bioswale on a site located
at the mouth of the drainage basin discharging to the creek. In addition to a superior technology, a
municipality could have more confidence in the ability of its staff to maintain a single public facility, than in
the municipality's ability effectively to enforce maintenance practices on multiple private underground
facilities. The site's location, at the mouth of the basin just prior to discharging to the creek, provides
maximum flexibility in determining what area might be routed to the facility for treatment, thus allowing the
municipality to prioritize and mitigate drainage areas with higher pollutant potential.
Community goals can be served by integrating open space amenities with existing creek trail systems and
providing greater flexibility to implement desired development projects within the confines of limited space.
Cost is a determining factor, and it will vary greatly from site to site. A regional facility can be funded in
several ways, depending on the options available to a municipality or utility under state law. A regional
facility should not be expected to be funded entirely by private development, even if it provides some
service to redevelopment. This is true because, as in Seattle's case, the facility will likely address some
existing flows in addition to the developed sites. Also, municipal staff resources would be spent on design,
construction and maintenance.
•S Development vs. Retrofit
Would municipally-constructed facilities address only mitigation triggered by development, or would
the facility address existing runoff?
This decision will vary for each scenario and may be influenced by the following factors:
• size of the site in relation to the drainage area,
• the water quality characteristics of the drainage area,
• the relative ease of directing flows to the site, and
• how the site fits in the municipality's priorities for retrofitting.
228
-------
If the site is large enough to accommodate additional flows, and the drainage is relatively easy to direct to
the site, the municipality might consider combining off-site mitigation with mitigation of existing
development. Much of the cost of capital facilities is in the design, permitting and grading— and increasing
the size of one facility is often much less expensive than creating a separate facility. The municipality may
also have an interest in demonstrating a broader general public drainage benefit of the facility is funded in
part by drainage rates.
In the Seattle case study, some portion of the facility would likely address existing runoff providing public
benefits beyond enhanced development mitigation. The appropriate portion will vary by project and be
determined through technical analysis at the sub-basin level.
•S Funding Options and Authority
What are the funding option(s) and associated authority necessary?
Several options may be available for funding an off-site regional drainage facility. The available options
will depend on existing municipal or utility authority. In some cases, funding options may be combined.
Legal advice is essential in planning municipal action, and sorting through the range of legal authority
available to a municipality can present a significant challenge.
A municipality might choose to build and fund a regional facility using general municipal revenue or
drainage-specific funds:
• Use general municipal revenue, not associated with drainage rates or development options.
• Use general drainage utility rates. Costs could be spread over a larger service base.
• Create differential drainage utility rates reflecting the drainage service provided in geographical
areas. Increases could be targeted to areas receiving or needing more intensive service.
• Create drainage utility connection fees for users of a new facility. After a facility is built using
municipal authority and funds, drainage utility fees are charged to new users of the regional facility.
Each of these regional facility funding choices would leave legal and policy questions for a municipality
such as Seattle that currently requires on-site drainage facilities for redevelopment, as a result of its MS4
permit:
• Must developers still build on-site facilities, as required by the local development ordinance and the
NPDES MS4 permit issued to the City?
• If not, is it fair or legal to impose a general fee increase to build facilities that in part benefit private
development, without charging extra to the benefited properties?
• For funding, what difference does it make whether or not a development's actual drainage is
managed at a regional facility?
• If on- site detention/treatment requirements for new development will be fulfilled off site by using
capacity at a regional facility, can the local on-site drainage requirements be lifted? If so, how ?
What can or should the developers be charged for off-site regional drainage service?
• What legal authority is present, both to create a different fee for a developer (which could be a
drainage rate question) and to allow a developer to meet its drainage regulation obligation off-site
rather than on-site (which could relate to municipal responsibilities as a regulator of development
and an NPDES MS4 permittee)?
229
-------
An appealing option for funding at least part of a regional facility might be to create a fee for off-site
mitigation that developers could pay to fund off-site municipally-owned regional drainage service, instead
of requiring the developers to build on-site detention or treatment structures.
• Create a development-related alternative to pay a fee to obtain drainage service at the regional
facility rather than on site.
Utility rates or general utility funds could be used to build over-sized regional facilities. A municipality
could make excess capacity available to developers for a fee, to satisfy developers' on-site requirements.
Arrangements might be voluntary or mandatory, for a determined geographical area. Legal authority must
be established. In such a case, state law may explicitly permit developers to contribute to the cost of a
regional municipal facility, on a mandatory or voluntary basis. On the other hand, state law may limit or
prohibit this arrangement, or its mandatory nature.
In some limited cases, there may also be an opportunity for developers to agree among themselves to build a
privately-funded off-site facility.
• An agreement among parties to provide service off site, independent of municipal rates or fees.
In issuing development permits, the municipality as a regulator would have to determine whether the on-site
facility requirement would be met by the regional facility. The facility might be independently operated, or
the municipality might later choose to acquire the facility.
S Off-Site Mitigation Fee Structure
How would a fee for off-site mitigation be calculated? How important is it for a municipality to
recover the full cost of the facility through fees?
What are the options for structuring fees paid to a municipality for providing off-site mitigation at a
municipally-owned regional facility? Again, legal authority may determine the calculation methods
available for utility fees or development-related fees, but here are some options to consider in setting a fee:
• Based on cost of off- site facility:
• Pro-rata portion of the actual off-site facility cost based on capacity
— based on estimated runoff
— based on acreage or square footage of impervious surface
• Standardized fee per unit runoff reflecting average current cost of off- site facility construction
• Based on estimated cost of building facility on-site.
In some cases it may be wise to balance the on-site costs against the off-site costs, considering the options
available to a developer. For instance, if participation in a regional facility is an option to providing on-site
detention or treatment, the fee structure may affect the willingness of developers to participate in an off-site
option. A municipality should recognize that the full cost of a regional facility is unlikely to be recovered
from development-related contributions.
230
-------
Environmental Permit Compliance
•S NPDES Permit Requirements and Regulatory Authority
Does the jurisdiction's NPDES municipal stormwater permit require the jurisdiction to regulate
development to mitigate stormwater impacts? Does the jurisdiction have legal authority, and leeway
under its NPDES permit, to allow off-site mitigation?
The degree of legal authority municipalities have to mitigate stormwater development requirements off site
may range from explicit direction to explicit prohibitions. Each municipality should consider not only its
police power, utility and other state law authority, but also any requirements of its NPDES MS4 permit.
Each municipality will have to evaluate the appropriate level of authority and permit obligations, and the
associated level of risk, as well as the likely perspective of the NPDES permit issuing authority. The
following scenarios provide an example of the range of authority level and associated risks:
• Explicitly authorized
• Generally authorized
• Not Addressed
• Explicitly not permitted
In the Seattle case study, the City's NPDES MS4 permit requires the City to impose on-site detention and
treatment requirements for certain new development and redevelopment. The City's Code was required to
be, and was, approved by Ecology as equivalent to Ecology's guidance. Ecology's model of regulation is
site by site, but there is some leeway for modifying on-site requirements with sufficient justification. Both
Ecology's manual of model development regulations and the City's Code identify basin planning as a means
for jurisdictions to alter development requirements within the basin, but neither specifically mentions off-
site mitigation. Ecology has authorized the City to make off-site accommodations for treatment
requirements based on a City-approved integrated drainage control plan for construction that begins in five
years, but this has not yet been extended to detention. The City will need to determine what is necessary
and sufficient for basin planning and will need to justify an off-site mitigation program in a way that is
consistent with both the MS4 permit and the City's authority and needs.
What legal risks should be evaluated when considering an off-site mitigation program?
An off-site mitigation program can be legally risky or unexpectedly expensive. A municipality's authority
to implement the program may be questioned. A municipality may incur liability if it agrees to construct a
regional facility but is eventually unable to construct it, due to permitting or other complications. If the
facility was intended to replace on-site drainage control, then stormwater that would have been detained or
treated on site could go entirely unmanaged, and the developers' potential contribution to regional
stormwater control could be lost. Depending on NPDES MS4 permit conditions, the municipality might be
obligated to site the facility elsewhere or might be out of compliance. Under some funding mechanisms and
state law, the municipality might be obligated to refund monies not used within a certain time, losing the
financial means to complete the project. For instance, given permit constraints, funding uncertainties, and
changing priorities, even five years can be an ambitious timeframe for public facility construction.
231
-------
-S Point of Compliance
How is the municipality's point of compliance determined for evaluating performance?
For purposes of this discussion, point of compliance is the point at which the development requirement must
be met through equivalent mitigation. Theoretically point of compliance could be any of the following
scenarios, but these scenarios differ in risk level and relationship to the regulated drainage area.
• Site discharge point
• Point between site and discharge to receiving water body
• Discharge point to the receiving water body
• Receiving water body
A municipality must define "receiving water body" for this purpose. If "receiving water bod
water of the state, including a small creek, then off-site mitigation locations upstream of a discharge are
limited. If, on the other hand, "receiving water body" means only specified larger streams, rivers, or lakes,
then a greater number of off-site locations may be available.
One option is to evaluate performance at the receiving water body, or at the discharge point to the receiving
water body. Ecology has approved the option in Seattle to meet on-site water quality treatment requirements
from one or more development projects through off-site facilities within a basin draining to the same
receiving water body. This language defines point of compliance as the receiving water body. This approach
is more suitable for addressing water quality in major water bodies, than for addressing flow control in
creeks. For example, if off-site flow control is provided in a separate basin draining to a creek at a point
lower in the system than the basin with the development project, then technically an opportunity to improve
the flow regime in the reach between the sub-basins has been missed. Locating a regional facility
downstream of a participating development site would result in missed protection of the portion of the
stream between the development site and the regional facility. This makes a case for evaluating performance
for creeks at the basin's discharge point to the water body, not in the water body itself.
A municipality will likely want to retain maximum flexibility for siting regional facilities, to site facilities at
points of opportunity and where they will have the greatest impact. To this end, an important consideration
for funding, development regulation, and permit compliance is whether or not the off-site facility will
provide drainage service for the exact same stormwater that would have been managed on site under local
development regulations. If the same water will managed, it will be simpler and less risky to link
development requirements and funding from partners to an off-site municipal facility. Funding options that
do not rely on development-related fees or partnering present even less risk.
Available legal authority will determine to what extent funds related to a development site can be used for
an off-site mitigation facility that does not detain or treat the same stormwater. For instance, it may be that
connection charges are authorized only for developments directly served by a facility; in such a case, access
to the facility capacity would need to be consistent with authority. A fee could spur a legal challenge if it is
seen, on one hand, as opportunistically charging development for general municipal services provided
elsewhere or, on the other hand, giving benefit to development at unfair public expense.
As to permit compliance, the NPDES permitting agency will likely have an opinion about whether detention
or treatment services should be moved from the site of new development, and whether flow from the
development should be allowed to go unmanaged. The agency may support municipal spending on regional
facilities but hesitate to approve transferring drainage management from one subbasin to another.
Depending on the permit's terms and the agency's involvement with local regulations, the agency may even
232
-------
view an off-site mitigation program as noncompliance, so a municipality should work proactively with the
agency to smooth out disagreements.
Even if the permitting agency agrees that off-site mitigation meets the MS4 permit obligations, the
municipality should consider whether it is willing in the long term to take on detention or treatment
functions regionally that would otherwise be the obligation of site developers. Typically, municipal
regulation holds site operators responsible for discharge from their sites. If a problem is detected
downstream in the MS4, upstream dischargers can be held accountable. An off-site mitigation program
could alter this dynamic. If an MS4 permit requires that municipal stormwater complies with water quality
standards before discharge to waters of the state, an off-site mitigation program could shift to the public,
part of a private site-related water quality obligation.
•S Environmental Protection
How is the regional facility determined equally or more protective than on-site projects?
There are several options for evaluating the equivalency of on-site and off-site approaches, which is a key
inquiry to justify off-site vs. on-site detention or treatment in basin planning or in issuing a development
permit.
• Equivalent impervious surface (or pollution-generating surface)
• Equivalent volume of water
In addition to these one-to-one evaluations, greater effectiveness can be achieved by using a superior
technology than would be used on-site, and by treating areas contributing higher pollutant levels within the
sub-basin. Although prior to development the effectiveness of these two scenarios cannot be measured, a
simple model using information from previous research studies can be used to estimate the proposed
reductions under the two scenarios.
In general consolidating maintenance and providing bio-filtration features can be more protective of the
environment than multiple underground vaults because the effectiveness of WQ facilities is very dependent
on the frequency and quality of maintenance. By leveraging development and rate investments to treat both
existing runoff and runoff from a development, a regional project can be more protective.
-S Timing
What is the timing of development and regional facility construction? What if the development occurs
before the regional facility is constructed—leaving a window of time that during which runoff is
uncontrolled?
The least risk and most environmentally protective option is for the jurisdiction to first build the facility and
then offer off-site credit for future development projects. However, there may be partnership opportunities
where development occurs before a facility is identified or built; if those potential partners need
development permits before the option of regional stormwater management becomes available, opportunity
may be lost as partners opt for on-site facilities.
233
-------
On the other hand if the municipality sizes and constructs a facility "on speculation," and the future
development does not occur, or developers choose not to buy excess capacity in the facility under a
voluntary arrangement, then this capacity is an avoidable ratepayer cost.
There may be regulatory risk as well. ANPDES permit issuing agency may generally support off-site
mitigation in theory, recognizing the greater efficiency that may be possible. However, the permit issuer
and the municipality may have different perspectives if an off-site mitigation plan involves a delay in
providing detention or treatment for an area, as compared to what would be provided at the time of new
development under local on-site requirements. Such a delay may also create complications in issuing
development permits, where the on-site conditions cannot be fulfilled off site in the same time frame.
Municipalities may need to negotiate with the NPDES permitting authority to retain maximum flexibility in
timing. Local law may need to explicitly allow a developer a calculated delay in detention or treatment, if it
there is a firm commitment to provide the same off site.
Development Regulation Authority
•S Applicability
How is applicability established for the program? To which developments is an off-site option made
available? How are developments handled that are not upstream of a planned or constructed facility?
Typical development regulation criteria include:
• project size— Municipalities may only want to administer projects above a certain size threshold
where there will be more mitigation per transaction. On the other hand, municipalities may decide
that they can save administration costs by consolidating the review, inspection and enforcement of
smaller facilities into a single regional facility. In this case project size may not be a criteria.
• amount of pollution-generating surfaces— Municipalities may want to target land uses that are
known to contribute higher pollutant levels. On the other hand, municipalities may want to target
"cleaner" development projects to transfer the investment to areas contributing higher pollutant
levels. (For example, trading on-site residential development mitigation for a high turn-over
commercial parking lot that is currently un-treated.)
• drainage destination (to a creek or specific water body)— Depending on the utility's regulatory
flexibility and sophistication in prioritizing water bodies, the municipality may want to trade all
mitigation in one basin for treatment in another. However, depending on the specific situation, this
approach can undermine the development regulation by raising questions regarding the direct impact
of the requirement.
Additional application criteria for a municipally-administered program may include whether project is
located:
• within a priority drainage basin— The municipality may have designated specific basins for program
implementation, and only development in these basins would be applicable for the program. Basins
may be chosen through a prioritization process, through a growth management planning process, or a
combination of both.
• upstream of planned or constructed facilities— Development projects may be in the designated
basin, but not directly upstream of a planned or constructed facility. In this case, the municipality
must decide whether the drainage from the development project must flow through the facility to
234
-------
meet off-site mitigation, or whether an equivalent amount and quality of stormwater can be
mitigated within the basin prior to discharging to the receiving water body. Associated issues are
raised in the discussion of point of compliance, above.
Finally, the jurisdiction must decide how much capacity to provide and whether applicability will need to be
capped at a specific threshold and perhaps a timeframe. Capping the facility capacity ensures the
municipality will not have to site, design and build another facility if development continues beyond
projections. Ideally a facility would be sited and designed to compliment the development plan for the area.
The program should outline a template that ensures consistency, but allows for unique opportunities based
on the project location, circumstances and management goals for receiving water body.
The legal issues in determining applicability are similar to those discussed with in relation to the point of
compliance. Legal authority may limit the geographical boundaries for an off-site mitigation program. For
some funding mechanisms, it may be essential that flow from the development actually be detained or
treated by the regional facility in order to support a fee. In order to remove on-site detention or treatment
requirements, it may be necessary to justify that the alternative is equally protective of public health, safety,
and welfare, the environment, and public and private property. This may be a challenge if a regional facility
provides benefits at a location far away. In other cases, using fees for off-site mitigation not directly related
to a site can complicate development regulation in the future. For instance, if a development requirement is
lifted upon payment of a fee but flow from that specific site is not detained or treated, what happens if the
property is redeveloped later? A municipality should consider its overall strategy for off-site mitigation and
deal with as many issues as possible when the program is esatblished, to provide a predictable basis for
future development.
D. Conclusion
Off-site mitigation programs have the potential to shift development-required investments to address high
surface water priorities identified through basin planning. However, this type of program is not applicable
or appropriate to all municipalities, and even in appropriate situations, the approach shifts responsibility and
liability to the municipality. This paper has attempted to outline the municipal drainage management,
NPDES permit compliance and development regulations issues associated with offering an off-site
mitigation program. This paper is intended to prompt discussion regarding the effectiveness of this strategy
as a tool for surface water managers in urban jurisdictions to meet multiple interests and put limited
stormwater management dollars to effective use.
235
-------
A REGIONAL APPROACH TO PHASE II PERMITTING
ENCOURAGES COOPERATION AND REDUCES COST
Steve McKinley, PE
Water Resources Program Manager FMSM Engineers
John F. Damico, MBA President
Environmental Rate Consultants, Inc.
Patrick T. Karney, PE, DEE Director
Metropolitan Sewer District of Greater Cincinnati
Abstract
The communities in Hamilton County, Ohio are working together to integrate the EPA Phase U Storm
Water Permit by developing a financial plan and a legal organization (Ohio Revised Code [ORC] 6117) to
manage storm water on a regional basis. This approach will lead to an efficient and effective permit process;
encourage regional cooperation; and lower costs through the economies of scale. In many cases
communities are not able to afford the additional financial burden of the permit nor do they have the
resources to perform the requirements of the storm water permits. The villages and small townships have
expressed that they do not have the resources to develop and implement the permit requirements. If there is
no regional authority many of these small communities will be in violation of the NPDES Phase U Storm
Water Regulations.
This paper will describe a successful consensus building process used by a number of diverse municipalities
working together to address and develop solutions to the water resource problems. They are not alone;
hundreds of communities throughout Ohio and the United States are struggling to deal with these very same
problems. This has been a complex effort of more than a year of data gathering, consensus building, policy
development and regional decision making. There is too much data and information to describe all of the
tasks and events that have taken place in this effort. Therefore we will focus on the process used to achieve
regional cooperation and how it effected the NPDES Phase U Permit development. We will also look at
how regional groups working together can use economies of scale and provide a cost savings to many
communities in the region.
Introduction
Hamilton County is located in southwestern Ohio and consists of 49 communities including the City of
Cincinnati (also a Phase U community). Its suburbs, townships, and villages are all contained within three
major watersheds: the Great Miami, Little Miami, & Mill Creek. All but one community (a small township)
must comply with the Phase U Storm Water Regulations.
This paper will describe the following.
• Creation of a legal organization within the guidelines of an ORC 6117 to guide the Hamilton County
Regional Program.
• The financial aspects of funding such an organization.
236
-------
• Economy of scale cost savings as a result of joining a regional organization and participating in a
regional NPDES Phase II Permit.
More Than a Queen City
The City of Cincinnati has traditionally been referred to as the Queen City, a truly midwestern city located
in southwest Ohio. But while this is the way this area is known, there is much more to southwest Ohio than
just the City of Cincinnati. There is Hamilton County, home to a population of more than 845,000 people.
Hamilton County is situated in the extreme southwestern corner of the State of Ohio and covers an area of
414 square miles. Within the County are 49 municipalities, including 21 cities, 16 villages and
12 townships. Hamilton County is the third largest in the State in terms of population.
!»»«} _H lSiuwfcL^ •*"
'
1
t '
HAMILTON COUNTY
Figure No. 1. Hamilton County Map of Municipalities Including Cities, Villages and Townships
Three major watersheds that encompasses rural, suburban, and intense urban land uses cover the County.
These include the Little Miami River Watershed, Great Miami River Watershed, and the Mill Creek
Watershed. The Mill Creek Watershed is the smallest of the three watersheds and, except for a small area in
neighboring Butler County, is entirely contained within Hamilton County. Its drainage area contains the
most intense urban development (.Hedeen, S., 1994. The Mill Creek - An Unnatural History of an Urban
Stream).
Most of the urban and suburban communities are located in this watershed along with the area's industrial
complex. There are also more than 160-combined sewer overflows (CSO's) in this watershed. Because of
237
-------
the complexity of these problems the Rivers Unlimited Mill Creek Restoration Project (RUMCRP) and the
Mill Creek Watershed Council were formed to educate and address the water quality concerns of Mill
Creek. Both groups have been a part of the Steering Committee and have provided great contributions to
this process.
Each of the communities is important because they represent the growth and the vitality taking place in the
Cincinnati Metropolitan area. They also contribute to the problems of flooding and water pollution. While
this alone should be the reason for cooperation, EPA has provided another reason for communities to work
together through the NPDES Phase n Storm Water Permit.
The Storm Water Study
The Hamilton County Storm Water Study was initiated by the Board of County Commissioners (through the
Metropolitan Sewer District of Greater Cincinnati) to address storm water quantity (flooding) and quality
(Phase n NPDES) concerns. The specific purpose of this "Study" is to assist local governments throughout
Hamilton County, either individually or collectively, to address both the storm water quantity and the
NPDES Phase n water quality permit issues and regulations. These USEPA storm water regulations will
require all but one of the Hamilton County governments to obtain an NPDES Phase II permit by March 10,
2003. These permits require that each local government develop a Storm Water Management Plan (SWMP)
to address six (6) minimum control measures. Implementation of these minimum control measures is
intended to improve the quality of the region's rivers and streams.
Today, one of the most serious problems facing Hamilton County elected officials is storm water
management (Mill Creek Watershed Council, Summer 2002. "Voice of the Mill Creek"). Every local
government in Hamilton County has experienced varying degrees of storm water problems such as street
and basement flooding, street closures, stream bank erosion, clogged storm drains, sewer backups and un-
maintained detention basins, to name a few. Less frequent, but in many cases more severe, are extreme
rainfall events that wash out roads, flood homes and businesses, and in some cases result in injury or death.
Rescue workers along Sycamore Creek
Ally 18, 2001
238
-------
Erosion Dam age along Polk Run
from storm of July 17 — 18, 2001
Now, all of the Hamilton County elected officials must address the requirements of the unfunded USEPA
NPDES Phase II Storm Water Permit Program. The NPDES Phase n Storm Water Permit Program will
require most local governments to take action to improve water quality in rivers and streams in their areas.
Communities will also be required to reduce the pollution load coming from their storm sewers and drainage
ditches.
In July 2001, as a result of the torrential rainfall and ensuing devastating flood, the Storm Water Study
shifted its primary focus from establishing a regional Phase U permit application to include a means of
addressing some of the regional flooding and erosion control problems identified after that flood (Mill
Creek Watershed Council, Spring 2002. "Voice of the Mill Creek"). During the course of this Study, nearly
500 "Areas of Concerns" were identified. These "Areas of Concern" included flooding problems, erosion
problems, drainage problems and water quality problems. Additionally, The Hamilton County Department
of Public Works identified over 2,900 buildings that were located in floodplain areas, within the
unincorporated area of the County. A very preliminary estimate of the potential capital requirements would
exceed $500 million, including:
• $250 million for capital projects to address the local government's "Areas of Concern".
• $50 million as the local share of the potential costs to remove or mitigate structures in the 100-year
floodplain.
• $200 million as the local cost for the flood control component of the Mill Creek Tunnel Project.
As a result of these mandated water quality regulations and on-going water quantity problems, the Hamilton
County Board of County Commissioners have begun to "encourage regional cooperation" by initiating a
regional watershed based approach that will formulate and develop solutions for solving these problems. In
order to address these complex issues and begin the process of solving the water resource problems in
Hamilton County, a plan for regional cooperation was developed that included all of the communities. The
plan involved a series of community interactions that educate, inform, and provide a forum for interaction
239
-------
and consensus building. The plan also involved the formulation of a mission and series of goals that serve as
a foundation for regional cooperation.
Encourage Regional Cooperation
A kick-off meeting for the study was held on March 29, 2001. A Steering Committee was established with
representatives from County Departments, Local Governments, Regional Agencies and area Universities.
The Steering Committee has met monthly since April 2001, with an average of 40 to 45 people attending
each meeting. The purpose of this Steering Committee was to develop issues and policies for the Executive
Committee and to be a technical advisor to the Consulting Team. The Executive Committee was a small
group of elected officials that crafted the regional organization and set policy. The Consulting Team
developed and presented a series of "Issue Papers" to assist the Steering Committee in evaluating
alternatives and developing solutions to the quality and quantity problems facing the region.
There are many elements that go into the encouragement and development of regional cooperation. For this
project, a combination of planned and unexpected elements has come together to build the success we have
enjoyed to this point. The following discussion is a brief summary of the following critical success
elements:
• Planned Interactions
• Champions
• Mission and Goals
• Building Consensus
Planned Interactions
It was clear from the very beginning that good community relationships and trust would be needed to
develop regional cooperation. To accomplish this trust and relationship, four distinct types of community
interaction were planned to get as much interaction with community staff, management, and elected officials
as possible. These four types of interaction are as follows:
• Individual Interviews with Local Governments
• Steering Committee
• Executive Committee
• Regional Workshops
Individual Interviews with Local Governments
The Project Team met individually with each local government (a total of more than 50 meetings)
throughout Hamilton County. The purpose of these face-to-face meetings was two-fold. The first goal was
240
-------
to determine a current level of service for storm water in the regional service area. T he Level of Service are
those activities and functions that a community performs to address the storm water issues in a community.
For this study, the Level of Service for storm water includes the Administrative, Engineering & Technical,
Environmental & Regulatory, Operation & Maintenance, and Capital Improvement functions that support a
community's storm water management. We accomplished this in Hamilton County by identifying each local
government's problem areas (Areas of Concern) and obtaining copies of any existing ordinances,
regulations, and other pertinent information. Secondly, these meetings provided the Project Team with an
opportunity to begin building a relationship with communities located throughout the County, and to convey
the process and purpose of the project. This also helped the Project Team to provide each community with a
consistent message concerning impending NPDES Phase n Permit Program. Each community was invited
to participate in the Steering Committee process that will build relationships and trust throughout the region,
and most importantly provide a means for making decisions about how the County will address the NPDES
Phase II Permit Program.
Steering Committee
The Steering Committee consisted of a wide range of financial and non-financial stakeholder groups
including: community staff and management, several elected officials, county department representatives,
watershed and environmental groups, university representatives and others. The Steering Committee has
met each month since April 2001. All communities were invited to participate in the Steering Committee
process but not all of the communities attended the meetings. There has been a regular attendance of 40 to
45 at each of the monthly meetings. The purpose of this group was to discuss the details of each of the
issues of regional cooperation, continue the consensus building process started during the individual face-to-
face local meetings, and to conclude with recommendations that would be carried forward to the Executive
Committee comprised of elected officials. Issues such as the following were addressed by the Steering
Committee:
• Is there a need for a regional group?
• What is an NPDES Phase II Permit?
• What legal authority is available to form a regional district?
• What is the level of service?
• How much will a regional district cost?
241
-------
Storm Water Study Steering Committee
Executive Committee
The Executive Committee is a much smaller group of 10 members comprised of elected officials from
selected communities, the Township Trustees Association, the Municipal League, the Board of County
Commissioners, and the Metropolitan Sewer District of Greater Cincinnati. The purpose of the Executive
Committee is to consider the recommendations from the Steering Committee, create a legal organization
that will encourage regional cooperation, finalize and establish policy, define the storm water level of
service, and set rates and charges. The Executive Committee will make final decisions based on local
ratepayer interests.
Regional Workshops
Regional Workshops are an attempt to bring together as many of the community leaders (elected officials)
as possible to build consensus for the policies developed in by the Steering Committee and by the Executive
Committee. Thus far, only one workshop has been conducted. It was an important workshop because it
fueled the consensus to develop a small regional district to address the NPDES Phase n Permit.
Champions
Regional cooperation cannot occur without leadership. The Hamilton County Regional Storm Water
Program is no exception to that rule. The success that we have experienced to date has come largely from
the leadership of a group of concerned and passionate people. There are a number of people who could be
singled out from the Steering Committee and Executive Committee, and there are also those who have
paved the way (i.e., the City of Cincinnati Storm Water Utility, the City of Forest Park Storm Water Utility,
and the Mill Creek Watershed Council) for this project. There are however, those whose exceptional
leadership grants them the title of "Champion." Hamilton County Commissioner John Dowlin; Mr. Pat
242
-------
Karney, Director of the Metropolitan Sewer District of Greater Cincinnati; and Mr. Bill Brashaw, County
Engineer for Hamilton County, Ohio; have given their time, talent, and passion without reservation to the
pursuit of regional cooperation. Without their influence and support there would be no regional project. A
Champion is not created or named as a part of some defined process; they arise as a result of the
understanding of the vision and the sense of mission that can be accomplished by an effort. The Champions
in Hamilton County saw the vision of a regional district and responded with passion to provide the
leadership necessary to develop the Hamilton County Regional Storm Water Program.
Mission and Goals
Every successful endeavor must be planned with an understanding of the direction and destination of the
effort. In our initial meetings with the Steering Committee, a mission statement along with a series of goals
was developed to establish a foundation and guide for our entire process. The mission statement and goals
developed by the Steering Committee are listed below.
Mission Statement
Determine the most effective organizational / management / legal structure available in the State of Ohio, to
position Hamilton County and the local governments within the County, to address the NPDES Phase n
Storm Water permit regulations, and efficiently and effectively manage storm water on a watershed basis.
Goals
Water Quality
Develop a water quality program that will initially meet the requirements of the EPA NPDES Phase n
Storm Water Program and over the first five years of the program assist communities to move to
comprehensive water quality improvements throughout the district boundary.
Water Quantity
Develop a water program that will initially complement the EPA NPDES Phase n Storm Water Permit
requirements and over the first five years of the program move to a comprehensive floodplain and drainage
program.
Institutional / Organization
Create a legal organization to manage storm water on a regional basis utilizing Ohio Revised Code 6117 or
Ohio Revised Code 6119.
Environmental
Develop an environmental program that meets the requirements of the EPA NPDES Phase II Storm Water
Program and over the first five years of the program move to a comprehensive environmental program that
recognizes storm water as a valued community natural resource that needs to be preserved and protected.
243
-------
Finance
Establish a district-wide dedicated source of funding that supports the institutional goals of the program, is
fair and equitable, and creates both a short-term and long-term rate structure.
Public Involvement /Education
Create a Public Involvement / Education program that meets the requirements of the EPA NPDES Phase n
Storm Water Program and over the first five years of the program move to a comprehensive Public
Involvement / Education that includes all stakeholders and takes a watershed approach to help citizens
preserve and protect the environment.
Watershed
Implement a watershed approach throughout the district boundaries. (Note that the district boundary is
Hamilton County, Ohio, but the there are portions of three watersheds within Hamilton County and the
communities want to take a "Watershed Approach" to the management of the district).
Building Consensus
Building and achieving consensus with a large group was a real challenge. Some of these challenges
included: keeping the members' interests high, to motivate them to return to future meetings, to achieve
consensus, to communicate complex issues at a level that everyone comprehends, and to address personal
and political agendas. Techniques that were implemented and used for this process are as follows:
• Define Consensus - The group ultimately defined consensus as - "I can accept and live with this action
or solution." This definition does not necessarily provide the optimum solution for all members but
does provide a solution that everyone can live with as a region.
• Mission and Goals - We referred back to this foundational building block many times throughout the
process, which kept us on track and on target with our overall agenda.
• Agendas - An agenda was sent out before every meeting so everyone could attend the meeting and have
meaningful input in the process and topic of the day. We also sent meeting summaries to each
community after each meeting.
• Issue Papers - Key issues, policies, and topics were written in a "white paper" format called issue
papers. This contained important research, history, or regulatory information as well as alternatives and
recommendations.
• E-Mail & Internet - Communication with this many people is critical. We were able to use e-mail
(almost everyone had e-mail and internet access) for day-to-day communication and a project web site
was created on the Metropolitan Sewer District's Internet site. All of the presentations, issue papers,
agendas, meeting summaries and maps were placed on this web page.
• Variety of Materials and Presentation Methods - There was an attempt to make every meeting
interesting and informative by using a diversity of materials and techniques to present the meeting
244
-------
material. PowerPoint, presentation boards, Arc-View GIS demos and facilitated interaction were all used
in the meetings. In one of the meetings a written survey was used to gather information and opinion.
• Sergeant-at-Arms was selected from among the Steering Committee to keep order and focus.
Efficient and Effective Permit Process
Today, there is a new emphasis on dealing with storm water quality. Since enactment of the Clean Water
Act by Congress in 1972, local governments and industries in Ohio have spent hundreds of millions of
dollars to upgrade, expand or rebuild their wastewater treatment plants. The net result of this massive capital
program has been significantly improved effluents from wastewater plants with corresponding
improvements in the quality of receiving streams. As these treatment plants have improved however, it has
become apparent that there are other sources of pollutants to our rivers and streams that are adversely
affecting their quality and impacting aquatic life. These sources include agricultural runoff (fertilizers,
pesticides), hydro modification (channelization, stream maintenance), mining, urban runoff, land disposal,
construction site runoff and failing septic systems.
To address these sources of pollution, USEPA initiated the National Pollution Discharge Elimination
System (NPDES) storm water programs. The Phase I program required that major cities with populations
greater than 100,000, which had separate storm sewer systems (does not include combined sanitary sewer
and/or sanitary sewer systems) must obtain a permit from Ohio EPA by May 1993. In Ohio, only
Columbus, Akron, Dayton and Toledo were required to obtain a Phase I permit. The other major cities
meeting the population criteria were excluded from these regulations and fall under separate but related
combined sewer system regulations.
On December 8, 1999 USEPA adopted regulations that will require many of the remaining cities, villages,
urban townships and counties to obtain NPDES Phase n storm water permits. Currently Ohio EPA
estimates over 480 local governments across Ohio will be required to obtain a Phase n storm water permit.
All affected entities must obtain permit coverage by March 10, 2003. These local governments will be
required to develop a storm water management program (the permit is a storm water quality plan for the
community) that implements six minimum control measures. The following is a brief description of the Six
Minimum Control Measures.
Six Minimum Control Measures
1. Public Education and Outreach
Distributing educational materials and performing outreach to inform citizens about the impacts polluted
storm water runoff discharges can have on water quality.
2. Public Involvement / Participation
Providing opportunities for citizens to participate in program development and implementation, including
effectively publicizing public hearings and/or encouraging citizen representatives on a storm water
management panel.
245
-------
3. Illicit Discharge Detection and Elimination
Developing and implementing a plan to detect and eliminate illicit discharges to the storm sewer system
(includes developing a storm water system map and informing the community about the hazards associated
with illegal discharges and improper disposal of wastes).
4. Construction Site Runoff
Developing, implementing and enforcing an erosion and sediment control program for construction
activities that disturb one or more acres of land.
5. Post-Construction Management
Develop, implement and enforce a program to address the discharges of post construction storm water
runoff from new development. Controls could include protection of sensitive areas (wetlands), or the use of
structural Best Management Practices (BMP's).
6. Pollution Prevention /Good House Keeping
Develop and implement a program to prevent or reduce pollutant runoff from municipal operations. The
program must include municipal staff training on pollution prevention measures and techniques (e.g.,
regular street sweeping, reduction in the use of pesticides or street salt, or frequent catch basin cleaning).
Hamilton County Phase II Storm Water Permit
Hamilton County, Ohio is addressing the EPA Phase n Storm Water Permit as a regional multi-community
permit. This means that each community will be a co-permittee to a regional permit that is summitted by
the ORC 6117 Regional Sewer District. This Regional Sewer District will perform the "regional tasks" as
defined by the permit. The local communities will perform the "local tasks" as defined by the permit. The
District will also monitor, develop, and submit the permit document as well as the required annual reports.
A copy of one of the interm permit implementation plans is a separate document attached to the end of this
paper.
The first part of the study included the development of the permit through the facilitated Steering
Committee process. Various permit tasks and levels of service were reviewed by the Steering Committee
and a draft permit implementation plan was crafted by the Steering Committee.
The second part of the Study involved the preparation of many of the items required under EPA's Six
Minimum Control Measures. Items such as brochures, ordinances, and manuals are being developed in draft
form. The District will implement these items. However, the individual communities that participated in the
development process can use these materials even if they do not join the District. The products that are
being developed are shown in the table below:
246
-------
Table No. 1. NPDES Phase II Storm Water Permit Products for the Hamilton County Storm Water Study
1. Public Education 4. Construction Site Runoff
Brochures & Fact Sheets • Erosion & Sediment Control Ordinance
PowerPoint Slide Presentation • Enforcement Plan
Library of Materials • Site Review Procedures
Educational Programs • Proposed Sanctions
Press Information
5. Post Construction/Runoff Control
2. Public Participation • Model Storm Water Ordinance
• Speakers Materials • Draft BMP Manual
• Citizen Watch Group • Inspection Program
• Information Council
• Hotline 6. Pollution Prevention / Good Housekeeping
• Model Management Plan
3. Illicit Discharge Detection & Elimination • Facility Management Plan
• System Map
• Illicit Discharge Ordinance
• Detection Plan
Funding Legal Organizations
As previously discussed, Hamilton County will use a regional organization to cooperate in the development
of a regional NPDES Phase n Storm Water Permit and reduce the cost of development and implementation
to the communities. In order for this to be accomplished a legal framework must be available to create the
regional district. Two years ago the Ohio State Legislature crafted and passed House Bill 549 that modified
ORC 6117 to include Storm Water (along with Water and Sanitary Sewer) and to allow for the collection of
fees and charges to operate and maintain the storm water system. This is important because it allows
counties in the State of Ohio and all of the communities within the counties to form a regional district that
can assess and collect fees and charges to manage storm water similar to an incorporated city.
The Steering Committee made the decision to designate ORC 6117 "County Sewer District" to be the most
appropriate legal management structure to address regional storm water management issues throughout
Hamilton County. Once this decision was made, the process of selecting the appropriate size and scope of a
regional storm water organization was considered.
This process was accomplished by reviewing four "example" programs with different levels of service and
the related level of responsibilities for a given cost of service that the new organization would provide. For
example, the "small" regional storm water organization will only address the NPDES Phase n permit
requirements for each of the member communities. No other storm water services will be performed by the
small organization. Each local jurisdiction will remain in complete control of managing their respective
storm water programs including water quantity. They would also be responsible for local aspects of the
Phase n permit such as construction site sediment control, street sweeping, etc.
The "medium " regional storm water organization will address the NPDES Phase n permit requirements for
each of the member communities (the small organization service level) as described above, as well as a
capital improvement program that will address flooding and drainage issues on a regional watershed basis.
Staff will coordinate the planning, design, and management of regional capital projects. Capital projects
would only be constructed for regional areas of concern. The district will not perform maintenance. The
capital program will be designed to address flooding concerns.
247
-------
The "large/comprehensive" storm water program is an all-inclusive regional and local water quality, water
quantity, and floodplain management organization. This metropolitan storm water district would operate,
maintain, provide capital construction, and regulate all storm water activities for the district service area. For
the most part member communities would give up control of storm water activities. It should be noted that
the limits of local control would be based on the terms of the district's plan of operation and/or agreement
with local communities. The district would perform all planning, design, construction management, plan
review, administration, customer service, and billing services.
The program examples met the mission and goals developed by the Steering Committee. Even the low level
of service will meet the initial goals of the program. For example, the low level of service option will
develop the NPDES Permit and Implementation Plan for the regional district. No other storm water
activities will be performed as a part of this level of service. While this "low-end" program meets the
mission and goals established by the Steering Committee the extended time-dependent (5-years) portion of
the goals are not addressed by this level of service. This does not mean that this level of service will not
accomplish the program mission and goals; however, it does mean that the program will be limited to a
minimal level of service for a reasonable cost of service.
After careful consideration by the Steering Committee, consensus was achieved and a decision was made to
create a small organization with the purpose of administering and coordinating the regional permit and will
perform all roles responsibilities and activities associated with the NPDES Phase n program as will be
organized as follows:
• Five employees (senior engineer, planner, engineer, GIS specialist and public information specialist)
• Overhead charge of $12,000 annually
• A 6.2% administrative overhead charge to the County's general fund
• Mapping performed by District in the amount of $600,000 annually
• At the end of the first five-year permit term, additional staff would be hired for erosion and sediment
control and illicit discharge enforcement
• Inflationary cost factors of 2.75% for salaries and 2.90% for benefits
• Other expense cost escalation factors (3%)
• Any known costs that may be experienced by the District over the next five years
The following is a five-year average of the annual costs for the regional organization that will comply with
the NPDES Phase II regulations:
248
-------
Table No. 2. Hamilton County Regional Storm Water Program Five-Year Cash Flow Analysis (McKinley, S.
(FMSM), Damico J. (ERC), and J. Rozelle (FMSM), 2001-2002. Hamilton County Storm Water Program
Issue Papers No. 1-8).
5 Yr Ave.
Salaries and Fringe Benefits:
Salaries: $264,100
Fringe Benefits: $100,700
Total Salaries and Fringes * : $364.800
Other Expenditures:
Rent: $76,500
Furnishings & Office Equipment: $21,200
Overhead: $12,700
Accounting Payroll/ General Fund Chg: $100,900
Supplies/Materials: $21,200
NPDES Phase II Permit Costs: $10,600
Public Education Outreach: $114,100
MSD Startup Cost Annual Payment: $204,000
Print Brochures: $10,000
Develop and Maintain Website: $6,000
Storm Drain Labeling: $10,000
Watershed Signage: $5,000
Hotline: $10,000
Household Septic System Mgmt: $30,000
Sensitive Areas Plan: $20,000
Pilot BMP Program: $30,000
Dry Weather Screening: $15,000
Mapping: $637,100
Total Other Expenditures * : $1.334.300
Total Expenditures * : $1.699.100
* rounded to the nearest $100
The final cost associated with the small organization and level of service using a five-year average as
defined above, will be in the amount of $1,699,100. This figure equates into approximately $4.20 per parcel
(per household) per year, which meets the financial goal of this regional group to not exceed an initial cost
of $5.00 per household per year for each individual ratepayer developed as part of the strategic planning
process. It should be noted that inspection and maintenance issues are the responsibility of the local
communities. There is an option for the inspection and maintenance as well as other activities to be added to
the district in the future. This increase in level of service must also include an increase in cost of service and
the storm water fee.
249
-------
Lower Costs Through Regional Cooperation
The National Association of Storm Water Management Agencies (NAFSMA) conducted a survey of
communities required to obtain an NPDES Phase I permit (NAFSMA, 1999 "Survey of Storm Water Phase
n Communities). The survey determined that members had expended, on the average, $ 650,000 per
community for the permit application process alone. These costs are based on all Phase I communities
complying with the regulations on their own.
Examples of several Phase I communities that have already initiated programs to comply with the NPDES
Storm Water Regulations as follows:
Table No. 3. Examples of Phase I Communities with NPDES Storm Water Regulation Compliance Programs
(NAFSMA -1996. "Survey of Local Storm Water Utilities").
City
Dayton OH
Louisville, KY
Akron, OH
Toledo, OH
Annual Cost
$3.3M
$5.0M
$5.0M
$3.2M
Cost / Capita
$19.86
$ 7.21
$ 23.04
$10.20
USEPA estimates (based on the NAFSMA Study - "Survey of Storm Water Phase H Communities".) that
the annual cost to administer the Phase n program will be cost $1,525 per municipality for annual reporting
and an additional $9.16 per household per year for all other variable costs. Using this methodology, if all
communities within Hamilton County comply individually and ignore a regional approach, it would cost
approximately $3,041,975 ($74,725 annual reporting + $2,967,250 variable costs) annually. This compares
to the five-year average discussed above, where, if all of the communities join together and develop
regionally, the costs to comply with the permit are estimated to in the amount of approximately $1,699,100
annually, and $4.20 per parcel (per household) per year. This equates into a cost reduction and economies
of scale savings in the amount of approximately $1,399,300 per year for the entire region and a cost savings
to the individual ratepayer of approximately 44% per parcel (per household) per year when compared to the
EPA cost of complying estimates. The cost savings assumes that the individual communities have at least
minimal storm water programs for quantity and quality and that the local share of the program can be
implemented with little or no additional cost. Within Hamilton County there are programs that meet and
exceed these minimum requirements and those that do not meet these minimum requirements.
The cost savings can best be expressed using several examples. The first example that is already being
implemented is the labeling or marking of storm water catch basins and inlets. If purchased in small
numbers (> 20,000 markers) the cost is as much as $10.00 for each marker. The Mill Creek Watershed
Council (with the cooperation of the communities) through the regional efforts is able to purchase markers
in large amounts at a little over $2.00 per marker. The Regional District is planning to provide funding to
groups like the Mill Creek Watershed Council to manage programs like the Storm Drain Marking effort.
The second example involves the development of the three ordinances that are required. It is estimated that
the cost to develop one of these ordinances is approximately $10,000, assuming only a moderate amount of
250
-------
public input and revision. The cost of ordinance development for all 50 Hamilton County would be
$500,000 if each community did it on their own. Another way to look at this is, even if it costs twice as
much ($20,000) to develop an ordinance, the cost per community (if all fifty were to join the District) would
be $400 per community.
The last example is difficult to estimate cost savings at this time. The NPDES Phase n Permit requires all
permitted communities to map their storm water system and outfalls. This is one of the most difficult and
expensive portions of the permit. For many of the small villages, townships, and cities the development of a
storm water map is out of the question, they cannot afford to prepare the map. Their only hope of complying
with this part of the regulation is to share the cost of mapping with other communities through the regional
district.
Next Steps
IV- Steering Committee Recommendations ( McKinley, S. (FMSM), Damico J. (ERC), and J. Rozelle
(FMSM), 2001-2002. Hamilton County Storm Water Program Issue Papers No. 1-8.)
The Steering Committee has developed the following recommendations to the Executive Committee:
1) A County-wide Storm Water District should be established to administer the NPDES Phase n
Permit.
2) The District should initially be staffed with five FTE's including a Senior Storm Water Engineer,
Engineer, Public Information Specialist, Planner and GIS Technician.
3) The BMP's proposed in the amended Implementation Plan Matrix, including the mapping
component, should be used as the basis for the preparation of the Storm Water Management Plan
(SWMP).
4) Consider implementing a two-tiered rate for mapping costs/requirements to be determined based
on standards.
5) Initially, the goal should be to establish a storm water fee that does not exceed $5.00 per household
per year, excluding billing and collection costs.
6) For those local governments that wish to pass on the storm water fees to individual property
owners, an agreement between the County and the local government should so state; and the costs
of billing services and fee, including the cost of collection., will have to be added to the storm water
fee.
References
1) Hedeen, S., and the Rivers Unlimited Mill Creek Restoration Project, 1994. The Mill Creek - An
Unnatural History of an Urban Stream.
2) McKinley, S. (FMSM), Damico J. (ERC), and J. Rozelle (FMSM), 2001-2002. Hamilton County Storm
Water Program Issue Papers No. 1-8.
251
-------
3) Mil Creek Watershed Council, Spring 2002. "Voice of the Mil Creek".
4) Mil Creek Watershed Council, Summer 2002. "Voice of the Mil Creek".
5) NAFSMA - The National Association of Flood and Stormwater Management Agencies, July 1999.
"Survey of Stormwater Phase U Communities".
6) NAFSMA - The National Association of Flood and Stormwater Management Agencies, 1996. "Survey
of Local Storm Water Utilities".
252
-------
Fish Community Response in a Rapidly Suburbanizing Landscape
Robert J. Mltner
Ecological Assessment
Ohio Environmental Protection Agency
Groveport, Ohio 43125
Dale White
Ecological Modeling
Ohio Environmental Protection Agency
Columbus, Ohio 43216-1049
Chris Yoder
Midwest Biodiversity Institute and
Center for Applied Bioassessment and Biocriteria
Columbus, Ohio 43221-0561
Abstract
Stream biotic integrity in Ohio shows measurable declines when the amount of urban land use, measured as
impervious surfaces, first exceeds 5.3%, and declines below basic Clean Water Act goals when urban land use
exceeds 25%. Declining biological integrity was noted in Rocky Fork of Big Walnut, a stream with a rapidly
urbanizing watershed in the Columbus metropolitan area, at levels of total urban land use as low as 4%, suggesting
that poorly regulated construction practices constitute the first step toward declining stream health in suburbanizing
landscapes. The pervasiveness of this finding was evaluated in several streams in the periphery of the Columbus
metropolitan area by comparing measures of stream health sampled in 1996 and again in 2002. No declines in
biological integrity or numbers of sensitive species were noted between time periods. The rate of urbanization in
the surrounding watersheds was less in these streams than in Rocky Fork, and construction site environmental
practices were more noticeable than in Rocky Fork. This paper discusses the implications of these findings with
respect to current storm water and construction best management practices.
Introduction
Biological integrity in Ohio streams declines along a gradient of urban land use, measured as impervious cover
(Yoder et al. 2000, Miltner et al. in review). This finding is from IBI scores for streams draining urban and
suburban landscapes in the major metropolitan areas of Ohio paired with an estimate of the percent impervious land
cover in the watershed upstream from a sampling point. Yoder et al. (2000) observed in these data that both the
number of sensitive species and IBI scores declined with increasing amounts of impervious surfaces; however,
declines in the number of sensitive fish species were detectable at lower levels of impervious cover than IBI scores.
253
-------
Initial declines in the number of sensitive fish species were detectible when the amount of impervious cover
exceeded 5.3%, and overall biotic integrity declined below Clean Water Act goals when impervious cover
exceeded 27.1%. Overall loss of biological integrity, as measured by the Index of Biotic Integrity (IBI, Karr
1981), is characterized by shifts in community structure relative to the fish community expected for a given stream
size and location.
The results for Ohio are similar to other studies from around North America. The typical result being that the
quality of any given stream is negatively correlated with the amount of urbanization in its surrounding watershed
(Steedman 1988; Schuler 1994; Wang et al. 1997; Karr and Chu 2000; Wang 2001). Urban runoff carries toxic
contaminants (metals, polynuclear aromatic hydrocarbons [Yaun et al. 2001]), nutrients and sediment (Young et al.
1996), pathogens and debris. Impervious surfaces also result in hydrologic and geomorphic alterations to low
order streams: increased variance in stream flow, increased stream temperatures, and destabilization of the channel
(Bledsoe 2002). Collectively these stressors act to grossly impair biological communities when the range of
impervious cover within a watershed reaches 8 to 20 percent (Karr and Chu 2000, Schuler 1994), and become
irreparably damaged in the range of 25 to 60 percent (Karr and Chu 2000). Here "grossly impaired" and
"irreparably damaged" are in reference to minimum water quality standards (e.g., state narrative or numeric
standards for warm-water habitat), and do not necessarily capture the more subtle, but highly consequential, effects
evident at low levels of anthropogenic disturbance (Scott and Helfman 2001, Jones et al. 1999). The reason these
ranges vary exponentially is that the severity of impairment in urban areas is dependant on the number and type of
allied stressors (e.g.., combined sewer overflows [CSOs], wastewater discharges, landfills, accidental spills,
intentional dumping, and stream channel dredging and filling) associated with urbanization beyond the retinue of
hydrological and water quality consequences effected by imperviousness alone (Yoder and Rankin 1996).
Recently, declining biotic integrity was noted in Rocky Fork of Big Walnut (Miltner et al. in review), a stream
located in the rapidly suburbanizing Columbus, Ohio metropolitan area. The IBI scores for Rocky Fork fish
communities over time are provided in Figure 1. The declining biotic integrity observed in Rocky Fork mirrored
what was observed in the static state-wide urban gradient data set as describe above. These declines were
attributed to new home and allied infrastructure construction, and likely hastened by the rapid pace of development.
Portions of the watershed that were rural in 1990 had been decidedly urbanized by 2000. Conditions were also
aggravated due to a lack of meaningful environmental controls on construction sites, and suggest that land
disturbance is the initial cause of declining biotic integrity in a suburbanizing landscape.
We wanted to test for declining biotic integrity in several streams on the periphery of the Columbus Metropolitan
area that have suburbanizing watersheds to examine whether conditions observed in Rocky Fork could be
generalized among similar sized area streams. The streams chosen had all been sampled between 1996 and 1997,
and so offered the opportunity to observe whether measurable differences could be detected within five years, and
at rates of development modest compared to that observed in the Rocky Fork watershed. This paper discusses
our current findings in light of previous findings for urban streams (Yoder et al. 2000, Miltner et al. in review) and
potential directions for land-use policies.
254
-------
uu
50
40
IBI
30
20
10
1:
H~ •• ,
™P '•
VK. i
I -/* %.
«
Oj- - - ^r ^,/fi
+ '"'r1Ty--->^v ri'T K' X^ ''
: --0--1991 *
r x 1992
: t 1993
i i i 1 i i i 1 i i i 1
2 10 8 6
^s<~ -'
^^ tr
v 1994
---Q--- 1 996
• 2000
i i i 1 i i i 1 i i i
420
14
o>12
S10
Q_
W 8
if—
0 6
i_
<" /i
J2 4
3 2
0
-A--1991
--X-— 1993
-o— 1994
--D--1996
—•—2000
,.->*
- -ts ~.-~ o
12 10
River Mile
864
River Mile
Figure 1. Trends in IBI scores (left panel) and the number of sensitive fish species sampled in
Rocky Fork, 1991 -2000. The shaded bar in the left plot shows the minimum range for
acceptable IBI scores for small warm-water Ohio streams.
Methods
Fish communities were sampled at eight locations in seven streams (Figure 2; Table 1) using generator-powered,
pulsed D.C. electrofishing units and a standardized methodology (Yoder and Smith 1999). Fish community
attributes were quantified with the Index of Biotic Integrity (IBI; Karr 1981; Karr et al. 1985), as modified for
Ohio streams and rivers (Ohio EPA 1987,Yoder and Rankin 1995). Habitat was assessed at all fish sampling
locations using the Qualitative Habitat Evaluation Index (QHEI; Rankinl995). The QHEI is a qualitative, visual
assessment of the functional aspects of stream macrohabitats, and includes rankings for such things as amount and
type of cover, substrate quality and condition, riparian quality and width, siltation, and channel morphology.
An estimate of urbanization between 1990 and 2000 was made for each sampling location by comparing data from
census blocks immediately surrounding and upstream from a sampling location and using housing density as a
surrogate for urban land-use. The number of sensitive species and IBI scores sampled at the same locations and
for each time period were compared using a two sample t-test. Sample distributions were checked for normality
using a normal probability plot. Sample variances between time periods for both IBI scores and number of
sensitive fish species were compared using a two-tailed variance ratio test (Zar 1999) and found equal (F 1.a/2; 9,9 =
4.03, > ratio of variances for IBI scores and number of sensitive fish species was 52.778/42.000 and 4.528/2.444,
respectively).
255
-------
Rocky Fork
Figure 2. Study area and locations sampled in 2002.
Rocky Fork is located for reference.
256
-------
Table 1. Change in housing density (units* mi ) in census blocks surrounding and upstream from stream sampling locations.
Stream Name
Location
Drain
Area (mi2)
Housing
Density
1990
Housing
Density
2000
Percent
Change
mi
1996
mi
2002
QHEI
Clear Creek
Poplar Creek 2
Poplar Creek 1
Muddy Prairie Creek
Sycamore Creek
Big Run
George Creek
Blacklick trib 10.36
Rocky Fork 3.1*
Dst US 22, Amanda Twp. 19.7 25.80 29.61 15 50 38 58.5
Poplar Cr. Rd., Liberty Twp. 8.1 48.32 55.73 15 58 56 76.0
Bish Rd., Liberty Twp. 17.5 48.32 55.73 15 42 48 79.5
Amanda-Northern Rd., Amanda Twp. 3.8 25.80 29.61 15 52 42 41.5
Busey Rd., Violet Twp. 21.6 176.67 301.40 71 44 44 78.5
Hayes Rd., Madison Twp. 6.3 95.78 172.38 80 46 38 56.0
Groveport Rd., Madison Twp. 15.4 95.78 172.38 80 40 44 61.0
SR 256, Violet Twp. 2.9 153.19 281.76 84 44 50 71.0
Clark Rd., Jefferson Twp. 22.4 57.10 202.50 254 30 NA 66.0
* Rocky Fork was not sampled in 2002.
257
-------
Results and Discussion
In contrast to what was observed in Rocky Fork (Figure 1), no differences (P > 0.05) were found in either the
number of sensitive species at a given site, nor for IBI scores at the eight study sites (Figure 2; Table 1), most
notably at the two sites that had the greatest rate of increase in housing density between 1990 and 2000, Blacklick
trib 10.36 and Sycamore Creek. One explanation for this observation is that the level of urban land use in each of
the eight study sites is estimated at less than 5%, except for Blacklick trib 10.36 where the level of urban land-use
from the 1994 Landsat Thematic Mapper Data was 7%. Also, the rate of change in housing density in all cases is
less than that observed in Rocky Fork (Table 1). Another difference, though not directly quantified, is that proper
construction site environmental practices were observed in Fairfield County where six of the eight samples were
collected (Figure 2). Fairfield County has storm water and construction site regulations requiring environmental
measures, and performs regular inspections for compliance through the local Soil and Water Conservation District
(Fairfield County SWCD, personal communication, Chad Lucht). Environmental measures to mitigate construction
site impacts were rarely observed in the Rocky Fork watershed (Figure 3).
Water resources can be impacted by land development. Whether that is because existing regulations are under-
enforced or are under-protective is an open question. Regulations vary widely between political jurisdictions. In
Ohio, a general storm water construction permit that is applicable state-wide requires best management practices
(BMPs) to minimize sediment loads. Temporary stabilization is one such BMP wherein disturbed areas that will lie
dormant for at least 45 days must be stabilized with fast growing grasses and straw mulch within seven days, or
within two days if within 50 feet of a stream. Other required BMPs include sediment ponds, silt fences,
construction entrances, inlet protection, and permanent stabilization. This basic level of protection is augmented by
stricter regulations and enforcement in some Ohio counties, such as Fairfield County.
00
Cf\
ou
40
30
20
1 n
IBI Scores
IT.
i
I -T-
l ^
1.1.
1996 2002
Year
Number of Sensitive Species
1996 2002
Year
Figure 3.. Distributions of IBI scores (left panel) and number of
sensitive fish species (right panel) sampled at the same
location in 1996 and 2002 in seven streams located in the
periphery of the Columbus, Ohio metropolitan area.
258
-------
Other states have been more aggressive in regulating nonpoint pollution. Storm water protection in the State of
Maryland is administered is through local governance with state oversight. For example, Baltimore County has a
stream protection ordinance that calls for a forested buffer to extend on both sides of a stream and to include the
adjacent floodplain, slopes, and wetlands. And wherever development may adversely affect water quality, the
buffer can be extended to protect steep slopes, erodible soils and other sensitive areas. This is in addition to the
fourteen general performance standards for storm water management applicable throughout Maryland (Maryland
Department of the Environment 2000, and available at
http://www.mde.state.md.us/programs/waterprograms/sedimentandstormwater/stormwater design/index.aspY These
performance standards go beyond simply minimizing the amount of sediment from construction sites by striving to
maintain the pre-disturbance hydrology of the watershed including groundwater recharge, stream channel stability,
and peak discharge volume. Compliance with local storm water regulations is encouraged through performance
bonds. A performance bond is bond issued to a contractor or other responsible party conducting land
development, forfeiture of which is risked if the party does not comply with the terms of the bond (i.e..,
performance standards) Wisconsin has recently enacted sweeping state-wide regulations governing both urban
and agricultural nonpoint pollution.
The realization of environmental consequences from land development has brought environmental considerations to
the fore as evidenced by model "smart growth" legislation proposed by the American Planning Association (2002),
and as enacted in Maryland and Wisconsin. Aggressive regulation and follow-up enforcement is needed to address
water quality impacts associated with land development, but finite limits on development must also be an integral
component of any future land use planning and regulatory framework. Significant numbers of sensitive species are
lost at relatively low levels of impervious cover, suggesting that the upper limit of urban land use for the highest
quality watersheds is about 5%. This argues strongly for no net gains in impervious cover in some watersheds.
However, for less sensitive waterbodies, aggressive regulations that protect riparian buffers and preserve much of
the pre-development hydrology may be effective at maintaining aquatic life uses consistent with basic Clean Water
Act goals at comparatively high levels of urban land use. Such regulations should include performance standards
analogous to those for Maryland. More specifically, they should minimize the loss of pervious cover, manage and
treat stormwater runoff to remove pollutants, retain stormwater and promote infiltration to maintain groundwater
recharge and stream base-flow, and pre- and post development peak discharge should remain similar to protect
stream channels. The level of urban land-use that can be reached and stream biotic integrity maintained under a
regimen of aggressive protection is currently unknown, but may go as high 50%. For example, from our previous
study of state-wide urban gradient sites (Yoder et al. 2000), sites that maintain good ffil scores at impervious
cover greater than 30% have either intact riparian zones and undeveloped floodplains, or have high sustained base-
flows relative to their drainage area. Also, Steedman (1988) found that an intact riparian zone of 20 m width was
important in mitigating effects of urban land use on aquatic life in Toronto area streams.
In summary, the cause and effect relationship between increasing land development and decreasing stream quality is
clear and abundantly demonstrated. For future land development to be sustainable, finite, watershed-specific limits
to development must be defined, land use planning must consider the ecological aspects of the landscape and
allocate development accordingly, and state and local governments must adopt rigorously protective environmental
regulations governing land development.
259
-------
. - JmgaBt^gS-:-^-*." -~-
^j^^ii^^.^'^v."^^^'.:.-^'"!:::'^'_'t»-! ".•„„—,„ ?"
silt
feSW;3i-**iEg; ivs-• Ti-;JV" - v - -2- • -"r;f::.. -'
fcstf.Vj -.=-" . = :-j3*;-!*!'*-,~™™.* ;'?,"* -T.."" -- ,~ --""J." -•-."• -. V • •-*•
Figure 4. Construction sites observed in the rapidly suburbanizing Columbus, Ohio metropolitan area.
Upper left, a construction site in the Rocky Fork watershed; the exposed soil is supposed to be stabilized
with straw and seeded with grass. Upper right, another tributary bulldozed for new construction. Lower
picture, a construction site in Fairfield County instituting proper environmental controls including silt fencing
and a settling pond.
260
-------
Acknowledgments
The authors thank all the staff of the Ecological Assessment Section of the Ohio Environmental Protection Agency.
The comments of two anonymous reviewers greatly improved this manuscript.
References
American Planning Association. 2002. Growing Smart Legislative Guidebook: Model Statutes for Planning and
the Management of Change. Edited by Stuart Meek. American Planning Association, Washington, D.C.
Bledsoe, B. P. 2001. Relationships of stream responses to hydrologic changes. In Linking Stormwater BMP
Designs and Performance to Receiving Water Impact Mitigation: Proceedings of an Engineering Foundation
Conference, August 19-24, 2001, Snowmass Village, Colorodo, pp. 127-144. Edited by Ben R Urbonas. United
Engineering Foundation, Environmental and Water Resources Institue of ASCE.
Jones, E. B. D., HI, Helfman, G. S., Harper, J. O. and Bolstad, P. V. 1999. The effects of riparian deforestation
on fish assemblages in southern Appalachian streams. Conservation Biology 13: 1454-1465.
Karr J. R. 1981. Assessment of biotic integrity using fish communities. Fisheries 6: 21-27.
Karr J. R., Heidinger R. C. & Helmer E. H. (1985) Effects of chlorine and ammonia from wastewater treatment
facilities on biotic integrity. Journal of the Water Pollution Control Federation, 57, 912-915.
Karr J. R, and Chu, E. W. 2000. Sustaining living rivers. Hydrobiologia 422/423: 1-14.
Maryland Department of the Environment. 2000. 2000 Maryland storm water design manual, Volumes I and U.
Center for Watershed Protection and the Maryland Department of the Environment. Water Management
Administration, Baltimore.
Miltner, R. J., White, D., and Chris O. Yoder. In review. The biotic integrity of streams in urban and
suburbanizing landscapes. Landscape and Urban Planning.
Ohio Environmental Protection Agency. 1987. Users Manual for Biological Field Assessment of Ohio Surface
Waters, Vol U. Division of Surface Water, Columbus, Ohio.
Rankin E. T. 1995. Habitat indices in water resource quality assessments. In Biological Assessment and Criteria,
Tools for Water Resource Planning, pp. 181-208. Edited by W. S. Davis and T. P. Simon. Lewis Publishers,
Boca Raton, Florida.
Schuler, T. R. 1994. The importance of imperviousness. Watershed Protection Techniques 1: 100-111.
Scott, M. C., and G. S. Helfman. 2001. Native invasions, homogenization, and the mismeasure of integrity offish
assemblages. Fisheries 26: 6-15.
261
-------
Steedman, R. J. 1988. Modification and assessment of and Index of Biotic Integrity to quantify stream quality in
southern Ontario. Canadian Journal of Fisheries and Aquatic Sciences 45: 492-501.
Wang, L., Lyons, J., Kanehl, P. and Gatti, R. 1997. Influences of watershed land use on habitat quality and biotic
integrity in Wisconsin streams. Fisheries 22: 6-12.
Wang, X. 2001. Integrating water-quality management and land-use planning in a watershed context. Journal of
Environmental Management 61: 25-36.
Yoder C. O. & Rankin E. T. 1995. Biological criteria program development and implementation in Ohio. In
Biological Assessment and Criteria, Tools for Water Resource Planning, pp. 109-144. Edited by W. S. Davis and
T. P. Simon. Lewis Publishers, Boca Raton, Florida.
Yoder, C. O., and Smith, M. A. 1999. Using fish assemblages in a state biological assessment and criteria
program: Essential concepts and considerations. In Assessing the Sustainability and Biological Integrity of Water
Resources Using Fish Communities, pp. 17-56. Edited by T. P. Simon. CRC Press, Boca Raton, FL.
Yoder, C. O., Miltner, R. J., and D. White. 2000. Using biological criteria to assess and classify urban streams
and develop improved landscape indicators, pp. 32 - 44. in S. Minamyer, J. Dye, and S. Wilson (eds.), National
Conference on Tools for Urban Water Resource Management and Protection. U.S. EPA, Office of Research and
Development, Cincinnati, Ohio. EPA/625/R-00/001.
Young, W. J., Marston, F. M., Davis, J. R. 1996. Nutrient exports and land use in Australian catchments.
Journal of Environmental Management 47: 165-183.
Yuan, Y., Hall, K., and Oldham, C. 2001. A preliminary model for predicting heavy metal contaminant loading
from an urban catchment. The Science of the Total Environment 226: 299-307.
Zar, J. H. 1999. Biostatistical Analysis. Prentice Hall, Upper Saddle River, New Jersey.
262
-------
ELEMENTS OF SUCCESSFUL STORMWATER OUTREACH AND EDUCATION
Catherine Neiswender
University of Wisconsin-Extension
Madison, WI
Robin Shepard
Wisconsin Extension Water Quality Coordinator
University of Wisconsin
Madison, WI
Abstract
Population growth, residential and industrial development, and the resulting increase in impervious surfaces
have led to stormwater quality and quantity concerns and related habitat and fiscal issues. To effectively
manage such issues, stormwater professionals are finding it necessary to develop community support
through implementation of education strategies. This need arises not only from the regulatory requirements
of EPA Phase n Stormwater rules, but also from the recognition that local decision makers, citizens and
elected officials will require more than a rudimentary grasp of stormwater pollution concerns in order to
make effective decisions that will have a positive impact on stormwater issues.
Throughout EPA Region 5, the University Cooperative Extension System is playing a strong role in
developing effective, outcomes-based stormwater education and outreach programs that not only meet the
federal requirements, but also the needs of the communities they serve. This paper will highlight some of
the successful stormwater education and outreach programs that Cooperative Extension is involved in and
describe its role in building the capacity of decision-makers. Elements of successful stormwater education
programs will also be highlighted.
Situation Statement
Like many regions in the country, states in the Midwest are experiencing some areas with rapidly growing
populations and accompanying development pressures. Population growth has spurred industrial,
commercial and residential development not only around the major metropolitan areas, but also in the
surrounding agricultural landscapes as well. For example, Ohio, which ranks as the 5th most populated
state nationally, is experiencing land development rates (in acres) 4.7 times faster than its population
increase (Lawrence, 2002). The resulting increase in impervious surface has led to stormwater quality and
quantity concerns and related habitat and fiscal issues.
To effectively manage such issues, stormwater professionals are finding it necessary to develop community
support through implementation of education strategies. The need to develop a knowledge base arises not
only from the regulatory requirements in EPA Phase n Stormwater Rules, but also from the recognition that
elected and appointed officials may have little incentive to prevent stormwater problems from escalating
unless they have a rudimentary understanding of stormwater concerns and solutions.
Two critical elements of Stormwater Management plans are the development and implementation of an
educational plan and public participation.
263
-------
In the 6 states in EPA Region 5 (Minnesota, Wisconsin, Illinois, Indiana, Michigan and Ohio), over 1,800
communities are required by the Phase n Stormwater Rule to obtain a stormwater permit and develop a
stormwater management plan (Federal Register, 1999). Developing educational and public participation
strategies for all of these communities requires creative partnerships to not only meet the stormwater
requirements, but more importantly to ensure that stormwater programs are effective in reducing pollution
and improving water quality. The U.S. Environmental Protection Agency (EPA) has identified the
importance of informing and educating municipalities, the construction trades, professional service
providers, and citizens about storm water pollution. Control of stormwater pollution is most effectively
implemented when people and organizations understand the impact of stormwater pollution, its sources, and
the actions they can take to control it (Dane County, 2003).
University Extension Systems in many of the Midwestern states are involved in and taking a lead role in
developing education programs to address stormwater and urban water quality issues. These programs are
conducted at several scales including regional, statewide, local or watershed, and metropolitan area. These
programs encompass several key elements for successful educational programming. Programs from three
states, and successful educational program elements, will be highlighted below.
Highlighted State Programs
Ohio
Ohio's statewide program goes well beyond efforts required by stormwater regulations and finds its
foundations in long-term watershed work that has occurred over the past decades in the state. Ohio's
statewide Nonpoint Source Education for Municipal Officials (NEMO) program encompasses a broad
partnership of agencies, with educational efforts led by the Ohio State University Extension (OSUE). The
Ohio NEMO program attacks a broad range of land use related water issues including stormwater, source
water and general natural resources based land use planning. Modeled after the National NEMO program,
the Ohio version is a non-regulatory research based educational program that addresses NFS pollution and
its link to different land uses, particularly impervious surfaces and, transport and concentration of pollutants
in stormwater. The Ohio NEMO program is a multi-level education program that involves 5 OSU Extension
Watershed Agents and several partner agencies for statewide delivery of educational programs that meet the
needs of agency staff, watershed groups, and local officials who are facing rapid urban expansion into
traditional agricultural areas.
The NEMO program also works to continue delivery of education as the constant turnover of local township
trustees, county commissioners and zoning board officials highlight the need to keep these decision makers
aware of the ramifications of land use impacts on water quality. The goals of the program, which expand
beyond stormwater education needs, are to increase public participation in water resources decision making
processes, and increase collaborative efforts of citizens and local decision makers in both development and
implementation of watershed action plans and source water protection plans.
OSUE faculty have several roles in the NEMO program. In addition to providing overall coordination and
leadership, OSUE augments local education efforts with materials, slide shows and more importantly,
educators that have knowledge expertise in stormwater and natural resources planning as well as skills in
facilitation and teaching strategies.
Successful elements of the NEMO program which lend themselves to effective outreach and education
programs include the systematic approach to address the turnover of local decision makers and the
264
-------
interconnection of stormwater with other natural resources planning efforts. For example, since the
beginning of the program, 80 professional staff have participated in training sessions on the use of NEMO
materials with local officials and watershed groups. These staff are then available to provide ongoing
training to new decision makers when turnover occurs (Lawrence, 2002).
Wisconsin
University of Wisconsin Extension (UWEX) faculty are involved in several stormwater education initiatives
throughout the state. In Dane County, 19 communities came together to develop a joint Information and
Education Plan and hire a half time education specialist to implement the plan. UWEX faculty provided
information to communities on why education is important and how to develop an education plan. This
work built community support and led to the development of an agreement to set aside funding to support
development of a Plan, hire the stormwater educator and provide $ 10,000 of annual funding for program
implementation. UWEX also facilitated the process of developing the Information and Education Plan with
a committee of representatives from the 19 communities and Department of Natural Resources. The
stakeholder committee first developed educational goals and UWEX was able to bring their expertise in
proven outcomes-based educational strategies to bear on these goals. This included identifying and ranking
target audiences and subsequently prioritizing educational objectives for each of the specific audiences.
UWEX also played a significant role in writing the final Plan document.
Successful elements of this approach include the identification of what the educational program efforts are
to achieve (i.e. the goals) and the target audience. This approach prevents the scatter-shot effect of random
educational efforts that are difficult to prove whether they have had an impact or not. Another successful
element of this effort includes a significant evaluation component funded by a separate grant. A pre-
assessment survey will be delivered to 500 residents in the communities to assess perceptions, behavior and
willingness to change behavior. After five years, a post-assessment survey will be administered to evaluate
the effectiveness of the stormwater program. Additionally, each major educational programming effort will
be evaluated to ensure that it is having the desired affect on changing people's behavior (Wade, 2002).
A related effort in Dane County was the development of a public participation process for their stormwater
ordinances. The UWEX role included working with specialists and engineers to develop the ordinance,
then providing outreach to local government units about the ordinance, and providing technical workshops
for engineers and consulting firms. A key UWEX role was to involve a wide variety of stakeholders early
in the ordinance development process and ensure their time and skills were well utilized. They enabled the
ordinance information to be re-packaged for the various audiences they were targeting. They also
encouraged public participation prior to ordinance development so that concerns were brought out early in
the process (Habecker, 2002).
A third educational initiative in Wisconsin occurred in the Fox Valley in the northeast region of the state.
This more traditional educational initiative included regional stormwater conferences and workshops on a
variety of regulatory and technical stormwater topics; a county-based stormwater management plan
development process; and a high school youth based stormwater monitoring project. UWEX faculty and
staff play key leadership roles in developing and implementing these programs. These three nested
initiatives focused on targeting the various audiences, while linking education with technical expertise to
ensure audiences were able to understand the complex nature of stormwater management alternatives to
make the best decisions (Koles and Neiswender, 2002).
265
-------
Minnesota
The University of Minnesota Extension (UMNEX) is extensively involved in the Metro Water Quality
Education Program in the Twin Cities (St. Paul and Minneapolis) metro area. This program, which is a
partnership of several agencies, targets educational programs to citizens, industry and local decision makers.
Several deliverable programs focus on lawn care, volunteer stream monitoring, wetland evaluation, NEMO
and Phase n Stormwater Education. Since Metro area water quality education involves a host of other
organizations, departments and agencies, the UMNEX plays a lead role in coordinating educational efforts
of these entities to create both efficient and effective educational programs. UMNEX also helps the groups
enhance their efforts by pooling financial and institutional resources leading to less expensive educational
programming, more consistent information and greater educational impacts.
A new initiative in the Metro area will focus on lawn, garden and home practices that improve urban
Stormwater quality. This new educational program will target homeowners and public property managers
and have an accompanying evaluation plan that will evaluate short and medium-term outcomes of the
educational initiatives (Struss, 2002).
Role of Cooperative Extension
This sampling of education initiatives throughout the Great Lakes region emphasizes the value of a
proactive approach to building education into the development of Stormwater management programs. The
University Extension System has played key lead roles in these examples, which are ultimately all highly
collaborative with other partners. These programs elevate the importance of education to the same level of
importance as the engineering, modeling and monitoring work that must also go into development of a
Stormwater plan. Many of our clean water goals will only be met through the individual actions of citizens,
construction crews, and local decision makers - actions that require targeted educational programs to
change these behaviors.
University Extension faculty have the education and process skills that lend themselves well to Stormwater
programs. In these examples Extension faculty have acted as educators for a variety of audiences including
local government decision makers; facilitators of meetings and processes that lead to the development of
educational strategies and sound decisions; specialists in outcomes-based educational program
development; authors of educational plans; and conveners of broad collaborative groups during various
stages of Stormwater plan development.
Successful Education Elements
There is some feeling that regulation and enforcement should be the main tools to accomplish clean water
goals, instead of education. However, past programs that relied solely on enforcement or monetary
incentives have not been successful. Research in Milwaukee, Wisconsin showed that a strong education
program must complement other means - especially when enforcement is spotty, penalties are light and the
audience is vast and widespread. Education programs can often be under funded or eliminated as an
element of a comprehensive Stormwater management program. Therefore it is critical that anytime an
education program is developed, it must be effective and justify the resources and time used to implement
the program (Dane County, 2003).
Several elements of success are presented here to help communities, educators and program managers build
effective education programs. These elements are drawn from several Stormwater and urban water quality
education programs throughout the upper Midwest that have leadership by or involvement of the University
266
-------
of Extension System in the state. The definition of success will vary from program to program, but
generally speaking, a successful education program is one that targets its audience and achieves the desired
outcomes and behavior changes. Likewise, success also includes elements of efficiency and sustainability.
These elements are:
»«» Going beyond 'awareness' - using outcomes-based educational principles
»»» Audience targeting - particularly decision-makers
*»* Partnering educators with technical expertise
»«» Incorporating stormwater into other natural resources and land use planning efforts
*»* Using public participation effectively
*»* Coordination of multi-jurisdictional efforts to effectively use education dollars
*»* Evaluation strategies
Outcomes-Based Education
A large body of research describes education principles, communication science and current learning theory
and their application to environmental and community-based projects (see for example Rice and Atkin,
2001; Rogers, 1995). Addressing complex environmental issues, such as stormwater management, requires
a combination of technical programs, best management practices and a vigorous and targeted education
strategy. Without effective education programs, best management implementation is often only done by the
early adopters. Effective education programs are ones that apply the outcomes-based principles of situation
analysis, audience targeting, and a focus on the desired behavior changes, not the 'products' of a typical
outreach or public relations program. Social marketing theory and research points to flaws in traditional
single-media educational campaigns and their inability to target key audiences (Earle, 2000; Shepard, 1999;
Hill, 1996). However, this research has not been incorporated enough into development of outreach
programs for environmental programs. For this reason, these outreach programs become little more than
public relations efforts relying too much on mass media, and as a consequence too often fail to achieve
meaningful behavior changes.
The University Extension System has long practiced outcome-based education in its programming efforts
(Seevers, et al. 1997). These methods rely on developing locally driven programs with the audience in
mind, integrating research and knowledge to improve understanding and decision making, and focusing on
desired outcomes (Scarborough et al., 1997; Van den Ban and Hawkins, 1996). These principles are
regularly applied to a wide array of Extension programming and can be successfully applied to stormwater
programs as well. See Figure 1 for a diagram of Program Development and Evaluation method that is based
on outcomes-based education principles.
267
-------
Program Development in UW-Extension
PLANNING - IMPLEMENTATION - EVALUATION
SITUATION
ANALYSIS
Symptoms vtnsufi
Stakeholdor
PRIORITY
SETTING
Inputs
PROGRAM ACTION- Logic Mode!
Outputs Outcomes - Impact
Carol 4»r-
What we smelt: Vi'hji WB ila. whi
Vision
Mandate
Resources
Lacal dynamic
Cclataratara
:n::i -I..I : ||. ::i ir>
. i> 'i. ShDiworni ModiLffl-tenii Long-term
! > I -
. -.-i - Lcarriinq: Antian Cfirelihorts
•• Satisfaction AUilLde
x • Ztuts
Opr*—
OecStons I
-'••jl -:v Ci.ic
Unlv«f«iiy if WiMfilin-Exlennon Coop«f«Urt, Eittni.an Program Dovohaptnant & Evaluation O SMS
Figure 1. UW-Extension Program Development Model
Using outcomes-based education principles means focusing on desired outcomes of your educational
program (i.e., behavior change), not just the immediate outputs (i.e., factsheets, workshops and billboards).
Programs must go beyond making people aware of the problem and rather should focus on changing critical
behaviors. Though glossy publications are attractive, do they really lead to the behavior changes needed to
meet the water quality goals of the stormwater plan? Outcomes based education uses several social
marketing concepts to be successful, including 1) asking for a commitment from the audience, 2) placing
specific behavior prompts near behavior, 3) communicating the norm, and 4) removing barriers to desired
behavior (Dane County, 2003). An example of outcomes-based education is illustrated by Ohio's NEMO
program. A desired outcome of the educational initiative was the adoption of stormwater principles into
regulations and policies - an important behavior change by local officials that ultimately leads toward the
improvement of environmental quality. The program highlights several communities that adopted
stormwater management principles, due in part as a result of the Ohio NEMO educational programming
they participated in.
In Wisconsin, 19 communities in and around Dane County formed a committee to develop a joint
Information and Education Plan for their stormwater permit application. Specific behaviors that would
affect water quality change were identified and prioritized based on their potential impact to change water
quality. For example, controlling construction erosion in this rapidly developing area was identified as a
key issue; desired behavior changes included implementing specific Best Management Practices. The 19
communities deliberately worked to develop and prioritize strategies that will focus on these outcomes.
Also in Wisconsin, in the rapidly developing Fox Valley a county and regional education strategy was
developed. It focused on the desired behavior change of local decision makers to develop policy and
effectively apply tools and technologies to their stormwater programs. Positive outcomes of this
268
-------
educational strategy included the adoption of new stormwater and construction site erosion control
ordinances and commitment of a county revolving loan fund to support better stormwater management.
Audience Targeting
Targeting the audience is critical to effective education programs. Focusing on desired behavior change
requires the educator to focus on a specific collection of people that will do that behavior change. In
Wisconsin, a joint Information and Education plan identifies three types of audiences for their efforts, 1)
those that must act (elected officials, homeowners, business owners, developers), 2) those that must support
change (conservation groups, civic organizations, media and concerned citizens) and 3) those who are future
supporters and actors (youth, teachers) (Dane County, 2003). In Ohio, their NEMO program targets
decision makers and recognizes in particular that local officials and decision makers have high turnover
rates and a process must be in place to educate new decision makers as change of leadership occurs. A cadre
of professional staff have been trained to provide continuous support as this audience turns over
periodically.
Partnering Education with Technical Expertise
It is critical to engage the technical expertise of consultants and engineers when developing and
implementing stormwater education programs. For many aspects of stormwater management, the devil is in
the details, and the stormwater professional is the most appropriate person to help address technical
questions and provide analysis of options. During the county Stormwater Technical Advisory Committee
process in WI, the technical engineers regularly paired up with the Extension educator to present detailed
concepts and alternatives to their audience. The best role of the educator is to work with the technical
experts to communicate the technical messages to a variety of audiences in understandable ways (Koles and
Neiswender 2002).
Incorporating Stormwater into Natural Resources planning processes
Stormwater management fits logically into other natural resources and land use planning efforts. Often the
same measures taken to protect natural resources and manage sprawl (such as conservation design, and
reducing impervious surfaces) serve the dual purpose to protect stormwater infiltration areas like wetlands
and vegetated areas, foster on-site treatment and infiltration and reduce runoff via traditional curb and gutter
designs. Multi-agency coordination will strengthen the ability of planners to integrate various natural
resources and land-use planning elements together.
The Ohio NEMO program highlights the interconnections between stormwater and natural resources
management planning and works with local government officials to build their capacity to integrate these
programs.
Public Participation
Public participation is one of the 6 minimum measures of a stormwater plan and when done correctly, can
build the support needed to fund and implement changes that will affect nearly everyone in the community.
In Dane County, WI a public participation plan was developed prior to the development of the stormwater
ordinance. The public was engaged to help design the ordinance by providing the parameters and
guidelines. A team of specialists then developed the technical specifications to meet these criteria. The use
of public participation prior to ordinance development enabled the county to minimize potential conflict
resulting from ordinance changes.
269
-------
Coordination of multi-jurisdictional and multi-agency efforts
Efficiencies can be gained by coordinating educational efforts and messages and pooling educational
dollars. Several examples exist. In Minnesota, the Metro area of Minneapolis-St. Paul recognizes that
county tax dollars and university resources are most effectively used when there is coordination among the
host of organizations that have an educational role or need. In Wisconsin, the 19 municipalities in Dane
County all pooled their local resources to fund a joint stormwater educator position that would serve all of
the communities. Additionally, Ohio's Stormwater Task Force, comprised of several local and state
agencies, consultants and environmental groups, guides implementation of Phase n in Ohio and coordinates
educational activities across agencies.
Evaluation
To know that scarce education dollars are spent well and desired behaviors are changed, it is important to
evaluate educational programs. Evaluation measures a variety of outcome data against the program's intent
(Bennett and Rockwell, 1995). Evaluation should occur for short, medium and long-term desired outcomes
to ensure the educational program is on track.
The Metro Educational program in Minnesota and the Dane County Joint Education Plan in Wisconsin are
excellent examples of educational initiatives that have built in an evaluation plan at the beginning of the
effort. In Dane County a scientifically designed pre-assessment survey will be delivered to 500 residents in
the communities to assess perceptions, behavior and willingness to change behavior. After five years, a
post-assessment survey will be administered to evaluate the effectiveness of the stormwater program.
Conclusion
The success of these education approaches does not mean the stormwater learning needs will subside. On
the contrary, enhanced regulatory measures, continued growth, and related environmental factors are
effectively increasing the demand for quality outreach education. The expectation that individual and
collective behavior changes will improve stormwater quantity control and quality necessitates continuous,
multi-tiered, education strategies.
The authors encourage stormwater professionals and educators to use outcomes-based educational
principles when developing their education strategies. Additionally, professional facilitation and process
skills are critical to development of educational plans and public participation initiatives required by the
new stormwater rules. The University Extension System has expertise in these areas and in many places is
working with or taking the lead on stormwater educational programming and collaboration.
Such programs are critical to achieving desired results and behavior changes that will have a positive impact
on stormwater quality and quantity. The authors challenge states and communities to consider stormwater
educational programming a valid and serious part of their overall stormwater management plan and design
strategies that are targeted to local situations.
For more information
For more information on the programs described above contact the author at
catherine.neiswender@ces.uwex.edu. The Ohio NEMO program is found on the web at
http://nemo.osu/edu A listing of University Extension Water Quality contacts is available at
http://www.usawaterqualitv.org/contacts/WOCDirectory.pdf
270
-------
References
Bennett, C., & Rockwell, K. (1995). Targeting outcomes of programs (TOP): An integrated approach to
planning and evaluation. From: http://deal.unl.edu/TOP/
Dane County, WI, 2003. Draft Joint Stormwater Permit Group Information and Education Plan, A
stormwater Information and Education Plan for 19 Dane County Municipalities, in draft.
Earle, R., 2000. The Art of Cause Marketing. Lincolnwood, IL: NTC Business Books.
Federal Register, 1999. Appendix 6 & 7, Volume 64, No. 235, pages 68812-68837.
Habecker, M., 2002. Dane County University of Wisconsin-Extension. Personal Communication.
Hill, R.P., 1996. Marketing and Consumer Research in the Public Interest Thousand Oaks, CA: Sage.
Koles, M. and C. Neiswender, 2002. University of Wisconsin-Extension Steps up to the Plate for
Stormwater Education, P103-PUB, StormCon Conference, August 2002, San Marcos, Florida.
Lawrence, T., 2002. University of Ohio Extension. Personal Communication.
Rice, R., and C. Atkin, 2001. Public Communication Campaigns, 3rd. Ed. Thousand Oaks, CA: Sage.
Rogers, E.M., 1995. The Diffusion of Innovations, 4th Ed. New York: Free Press.
Scarborough, V., Killough, S., Johnson, D., and J. Farrington (Eds.)., 1997. Farmer-led Extension.
Intermediate Technology Publications, Ltd.: London.
Seevers, B., Graham, D., Gamon, J. and N. Conklin, 1997. Education through Cooperative Extension.
Delmar: Albany, NY.
Shepard, R.L., 1999. Taking Aim in Water Quality Education: Designing Audience-Specific Programs. In
Proceedings of the Sixth National Watershed Conference, Austin TX: Water Environment Federation.
Struss, R., 2002. University of Minnesota Extension. Personal Communication.
Van den Ban, A.W. and H.S. Hawkins, 1996. Agricultural Extension (2nd ed.). Blackwell Science Ltd.:
Cambridge, Massachusetts.
Wade, S., 2002. University of Wisconsin-Extension. Personal Communication.
271
-------
A CONSERVATION PLAN FOR THREE WATERSHEDS WITHIN THE
MILWAUKEE METROPOLITAN SEWERAGE DISTRICT (MMSD)
By Mark J. O'Leary, Applied Ecological Service, Inc.
Doug J. Eppich, Applied Ecological Services, Inc.
Peg Kohring, The Conservation Fund
Steve Apfelbaum, Applied Ecological Services, Inc.
Abstract
Previous watercourse studies completed for the Menomonee River, Oak Creek, and Root River have
indicated that demographic and community development trends over the next 20 years will exacerbate
flooding problems within these watersheds. These studies have provided recommendations for traditional,
engineered strategies to combat flooding: and they have acknowledged the importance of maintaining
existing open space to prevent future flooding. As a result, the Milwaukee Metropolitan Sewerage District
(MMSD) retained a team led by The Conservation Fund to develop a Conservation Plan for the acquisition
and protection of important open space at risk of development. The objectives of the plan were as follows:
1) Identify undeveloped private properties potentially at risk for development that could provide future
flood-reduction benefits; 2) Assess opportunities for MMSD to partner with public, private, or non-profit
entities that would assist with the acquisition, management, and maintenance of identified properties; 3)
Assess mechanisms and strategies to leverage MMSD funding for this effort; 4) Provide recommendations
for the acquisition of parcels (or easements on these parcels) at risk for development; and 5) Consider how
the ecological restoration of identified parcels could reduce future flooding. The Project Team used GIS-
based remote sensing techniques (aerial photography, soils maps, wetland maps, etc.) and field visits to
identify more than 28,000 acres of undeveloped land containing hydric soils that provide future flood
reduction benefits. A subset of 199 sites that were 25 acres or larger in size (a total of 17,146 acres) was
identified for further investigation. Thirty-four sites totaling 2,417 acres (representing 4,835 potential acre-
feet of storage) were eliminated during field visits because they had been developed. Other sites were
eliminated or ranked as low priority for acquisition if they contained a high number of parcels, were aligned
in an impractical configuration, or were known to contain environmental hazards. Forty-two sites were
identified as high priorities for acquisition. These were ranked based on several factors including: 1) surface
area; 2) potential storage capacity of the site relative to runoff produced by the sub-watershed tributary to
the site; 3) Potential storage to reduce flooding along the main stem of the watercourse; and 4) importance
of the site in reducing future flood risks. This study provides the scientific and practical rationale for
protecting these parcels from development in perpetuity, and for using public, private and non-profit entities
to manage these properties to maximize flood control benefits. Furthermore, this study identifies funding
mechanisms and strategies to leverage monies earmarked for land acquisition.
Introduction and Background
Watershed Changes
"While much attention of late has focused on the construction of engineering works as a means of
meeting water deficiencies . .. comparatively little consideration has been given to the regulatory
influence of the soil and rocks of the watersheds, or of the part played by herbaceous range plants in
maintaining the efficiency of these natural reservoirs." (Pearse and Wooley, 1937).
272
-------
Flooding is a natural process in which a stream or river spills over its banks and into the adjacent floodplain.
Flooding usually occurs because the volume of water running off of the contributing tributary area is greater
than the capacity of the receiving waterway, and the rate of water running off of the landscape is too great
for the receiving waterway to convey within its channel. Flooding also occurs when obstructions within the
channel or floodplain create bottlenecks that elevate water levels upstream.
Flooding has many positive effects in a healthy watershed including dissipating the energy of water and
thereby minimizing in-channel erosion; depositing nutrient-rich silt and sediment into the receiving
floodplain; temporarily storing water in the floodplain and then slowly releasing it into the primary channel
as water levels drop; and providing a plethora of habitat benefits, especially for wildlife that depend on
floodplain habitat during important times of their life cycle such as breeding and migration. Flooding can
result in devastating damage to property, water quality, wildlife habitat, and channel stability when the
ability of the floodplain to slow down and store water is impaired.
The frequency and degree of impact of floods is based on a number of watershed factors including
precipitation, topography, soil type, vegetation type and cover, and in developed watersheds, the type and
extent of land use.
Precipitation drives the storm water runoff of the watershed. Precipitation, while varying with event, is
relatively constant over time.
Topography influences the rate and volume of water running off of the landscape. All things being equal,
steeper landscapes convey more water at a higher rate than flatter landscapes. Flatter landscapes, or
landscapes with depressed areas, provide more opportunities for water to infiltrate, evaporate, and slowly
release into the waterway.
Soil type affects the infiltration of water into the ground. Highly pervious soils such as sand infiltrate water
more quickly into the ground than tight soils such as clay. Hydric soils, or soils created under anaerobic
conditions, often occur in depressed areas of the landscape.
Vegetation cover and type can dramatically affect the rate and volume of runoff. Living vegetation and
organic debris (duff) retard runoff. Roots provide channels for water to infiltrate into the ground and build
organic matter that has a higher water holding capacity than mineral soil. Vegetation type has a dramatic
influence as well. In general, native vegetation such as prairie plants have a much greater ability to capture
and infiltrate runoff than introduced species such as turf grass (Weaver and Clements, 1938; Weaver, 1954).
Changing land uses have the most dramatic effect on the frequency and impact of flooding. But before
listing the most important reasons, it is useful to consider how the historic Midwest landscape functioned to
manage storm water runoff before it was plowed, plumbed and peopled.
Today's Midwest landscape was shaped and formed over the last 10,000 years following the last glacial
period. The major land forms - plains, hills, valleys, wetlands, rivers and lakes - are artifacts of the glaciers
carving during encroachment, depositing debris during glacial retreats, and creating drainage ways for
melting ice to the Gulf of Mexico.
273
-------
Plants colonized the raw earth left by the retreating glaciers and evolved and adapted to climatic and
edaphic conditions that persist today. By the time the first Europeans established a firm foothold 150 years
ago, the ecosystems of the tall grass prairies, savannas, woodlands and wetlands were firmly established.
From a storm water management perspective, it is important to note that the capacity and morphology of
today's streams and rivers were formed (some might say "sized") when the contributing watershed was
vegetated in native prairie, savanna, woodland and wetland. Impervious surfaces only existed in localized
areas where bedrock was exposed. All other areas were vegetated or inundated. Storm water runoff was
minimal due to the great water holding capacity and natural infiltration of native vegetation and localized
natural depressions. In the prairie lands, many of the major rivers of today were little more than large
vegetated swales.
The character of our historic watersheds and receiving waterways began to change shortly after the arrival
of Europeans. In 1859, Henry F. French records the effects of agricultural practices on stream flows in his
Farm Drainage monograph:
"The effect of drainage upon streams and rivers, has, perhaps, little to interest merely practical men,
in this country, at present; but the time will soon arrive, when mill-owners and land-owners will be
compelled to investigate the subject... If now, this surplus of water, this part which cannot be
evaporated, and must therefore, sooner or later, enter the stream or pond, be, by artificial channels,
carried directly to its destination, without the delay of filtration through swamps and clay-banks; the
effect of immediate agricultural drains furnish those artificial channels. The flat and mossy swamp,
which before retained the water until the Midsummer drought, and then slowly parted with it, by
evaporation or gradual filtration, now, by thorough-drainage, in two or three days at most, sends all
its surplus water onward to the natural stream. The stagnant clay-beds, which formerly, by slow
degrees, allowed the water to filter through them to the wayside ditch, and then to the river, now, by
drainage, contribute their proportion, in a few hours, to swell the stream. Thus, evaporation is
lessened, and the amount of water which enters the natural channels largely increased; and, what is
of more importance, the water which flows from the land is sent at once, after its fall from the
heavens, into the streams. This produces upon the mill-streams a two-fold effect; first, to raise
sudden freshets to overflow the dams, and sweep away the mills; and, secondly, to dry up their
supply in dry seasons, and to diminish their waterpower."
Engineering News printed in 1892 a story with a similar message, titled "The Drainage of the Kankakee
Marsh," and excerpted as follows:
"But when the whole swamp is drained and under cultivation the rainfall will drain off from it as
rapidly as from any other tract of cultivated land of similar slope and character of soil. The swamp
will no longer be a great shallow storage reservoir to hold the floods which pour down from other
parts of the watershed. It is certain, then, that when the drainage enterprise is carried out, a
considerable increase in the flood volume of the Kankakee will result. The exact amount of the
increase it will be the duty of the engineers of Chicago drainage canal accurately to determine, for in
future years, when the compensation for flood damages in the Illinois valley arises the increased
flow from the Kankakee must be considered as well as that from the Chicago River".
274
-------
These early investigators write of draining the land and changing the plant communities from native prairie,
savanna, woodland and wetland, to agricultural land. It wasn't long before we started removing the
vegetation all together and began constructing impervious roofs, roads and parking lots.
The sequence of events beginning with a healthy undeveloped watershed with minimal to no flooding to an
urbanized watershed with severe flooding are summarized as follows (Coffman, 2002):
• In a healthy, undeveloped landscape, water falling on the ground is intercepted by vegetation,
retained in depressed areas such as wetlands, and is evaporated and infiltrated. Essentially, water
falling on the land stays on the land, or is slowly released into receiving streams.
• Urbanization results in compressed soils, an increase in impervious surfaces, and improved
conveyance systems such as streams straightened to ditches, agricultural drain tiles, and storm
sewers. Rather than remaining on the land as in a natural setting, water is piped off of the land as
quickly as it falls on to the ground.
• Streams and rivers, "sized" over the millennia to receive water from the native landscape, respond to
increased runoff by becoming wider and deeper. Flooding occurs as the effects of urbanization
outpace the ability of the waterways to receive and convey water; water quality drops as the channel
erodes, and water is conveyed through pipes rather than through native vegetation that filters water;
wildlife habitat is lost.
It wasn't long before the historic prairie streams - moving marshes with a current, really - were well
beyond their capacity to convey the volume and rate of water racing off of the urbanizing landscape. And
flooding began in earnest.
The MMSD Model
Studies completed for the Menomonee River, Oak Creek, and Root River watersheds in southeast
Wisconsin indicate that demographic and community development trends over the next 20 years will
exacerbate flood problems. These studies provide recommendations for conventional, engineered strategies
to combat flooding, as well as acknowledging the importance of maintaining existing open space to prevent
future flooding (SWRPC, 1990; COM, 2000, a,b,c).
Conventional engineered strategies include constructing massive storm water detention facilities where
storm water runoff is temporarily stored and released downstream at a controlled rate, or improved
conveyance to move water more quickly from one point in the watershed to another point downstream.
While detention and improved conveyance has been proven to reduce flooding within a localized region, in
many cases, these strategies have failed to adequately protect downstream communities from flooding,
degraded water quality and wildlife habitat, and eroding waterways for a number of reasons:
• New developments are still mass graded and sewered to drain water from the site as quickly as
possible. Conveyance is maximized while infiltration and evaporation are minimized.
• Proactive communities require detention ponds designed to release water from new developments at
the same rate water was released before the site was developed. However, release rates for detention
ponds are usually calculated based on the land cover type immediately prior to development rather
than the historic vegetation cover that likely had a much slower release rate. As a result, release rates
are often over estimated.
275
-------
• Detention facilities do not account for the increased volume of runoff from developed areas due to
the reality that much less water infiltrates into the ground than under historic conditions (Ferguson,
2002).
• Most storm water regulations address individual development projects but do not take into account
the cumulative affect of multiple detention facilities constructed along the same waterway.
• Some communities continue to allow development of naturally depressed storage areas such as
wetlands and floodplains. Even if existing regulations do protect these depressed storage areas,
regulations can change. Isolated wetlands, for example, are no longer protected from filling under
Section 404 of the U.S. Clean Water Act.
• Runoff characteristics of a watershed are very complex and storm water runoff models often
underestimate the actual rate and volume of runoff (Apfelbaum, 2001).
The construction of detention facilities over the last 30 years has provided tremendous flood protection
benefits and will continue to do so in the future. However, the persistence of flooding in areas where
detention facilities and other conventional storm water management strategies are in place, and the failure of
conventional techniques to adequately address water quality and habitat goals, makes the objective observer
question whether there aren't alternatives to at least supplement conventional strategies.
MMSD took the judicious approach of adopting a conventional storm water management plan per the
recommendations of Watercourse Reports prepared by Camp Dresser McKee. But in addition, they
launched an aggressive land acquisition program targeting land at threat to development that provided
important, natural storm water management functions.
MMSD retained a team led by The Conservation Fund to develop a Conservation Plan with the following
key components: 1) Identify undeveloped private properties potentially at risk for development that could
provide future flood-reduction benefits; 2) Assess opportunities for MMSD to partner with public, private,
or non-profit entities that would assist with the acquisition, management, and maintenance of identified
properties; 3) Assess mechanisms and strategies and leverage MMSD funding for this effort; 4) Provide
recommendations for the acquisition of specific parcels (or easements on those parcels) at risk for
development; and 5) Consider how the ecological restoration of identified parcels could reduce future
flooding.
The Conservation Plan was completed during 2001 and provides a technical basis and justification for
identifying undeveloped properties to purchase that have the greatest potential to protect against future
flooding. The plan also describes a land acquisition strategy, partnership opportunities, additional funding
sources, and how the plan can be expanded to target additional objectives such as water quality and wildlife
habitat with the implementation of an ecological restoration strategy. MMSD allocated $15 million dollars
over five years to develop the Conservation Plan and purchase property.
Project Area
The project area consisted of the watersheds of the Menomonee River, Root River and Oak Creek that are
within the MMSD Planning Area (Figure 1). The MMSD planning area is in southeast Wisconsin and
includes portions of Washington, Waukesha, Milwaukee, and Ozaukee counties. The Menomonee River
drains an approximately 135 square mile area including at least portions of the cities of Brookfield,
Milwaukee and Germantown. The Root River drains an approximately 197 square mile area including at
least portions of the cities of Franklin and New Berlin. Oak Creek drains an approximately 27 square mile
area including the city of Oak Creek, Milwaukee, and South Milwaukee.
276
-------
Figure No. 1: The study area consists of the Menomonee River, Root River and Oak Creek watersheds.
Study Area
Map
minsD
Tin
(MMIMMm
Fund
Legend
fflvw
C I
277
-------
Methods and Results
Base GIS Information
An extensive Geographical Information Systems (GIS) database was developed using Arc View ™ to
assemble, store, manipulate and display geographically referenced information. Digital data was obtained
from Southeast Wisconsin Regional Planning Commission (SEWRPC), MMSD, participating counties,
townships and municipalities, and the World Wide Web. Data layers developed included watershed
boundaries, sub-watershed boundaries, digital elevation models, aerial topography, 2' topography (where
available), planned and existing environmental corridors, governmental boundaries, parcel boundaries and
other layers.
Digital ortho-rectified aerial photography (1995 were the most current images available during the study
period), hydric soils, floodplain, private/public land, and land use/land cover data were obtained from
SEWRPC. Watershed boundaries and characteristics were obtained from Wisconsin DNR, Geographic
Services Section (April 1997). USGS 7.5" Digital Elevation Model (DEM) data were used to create an
elevation model.
Hydrologic Impact Site Analysis
The primary objective of the Hydrologic Impact Site Analysis was to identify undeveloped, privately held
parcels and evaluate their potential ability to store runoff and reduce flood risks.
An undeveloped site can reduce flooding in two ways. One, reduce the rate and volume of water running off
of the site; and two, reduce the rate and volume of water running off of lands tributary to the site. Several
criteria were used to evaluate and rank potential sites for restoration for floodwater runoff reduction
including: area; the potential floodwater storage capacity of the site relative to runoff tributary to the site;
the effectiveness of a site to store water; and the importance of a site to reducing flooding downstream along
the mainstem.
Site Selection - We began our initial investigations for potential sites by intersecting privately held,
undeveloped lands with hydric soils and floodplain. More than 28,000 acres of land were identified in the
initial query. Sites less than 25 contiguous acres were dropped leaving a subset of 199 (Figure 2) sites
totaling 17,146 acres. The smaller sites were dropped to create a more manageable data set to work with,
and because smaller sites would likely have less potential to affect floodwater runoff.
278
-------
Figure No. 2: Each floodplain was mapped to assist in the hydrologic analysis.
Floodplain
Locations
rarnso
Tho
Coniorvotion
Fond
Root Rtvef Watershed |
Legend
Rjww
Etu4yAn»n« mnnhm Bouidary
K* ~i>
I
*iMd JL
.— "T
QrtC«l
Hoc*
FiBur.;2
279
-------
Each of the 199 sites was field-verified with mapped data. Sites already developed or in the process of being
developed were removed. Thirty-four sites totaling 2,417 acres were eliminated during field visits because
they had been developed between 1995 and 2001.
Capacity Relative to Runoff - Each of the 199 sites were evaluated and ranked as to their potential to
efficiently handle runoff from their tributary watershed during a 100-year, 24 hour duration, storm event.
We assumed that the land cover of the tributary watershed was a typical, residential urban development
(Cn=75). This resulted in approximately 3.5" of runoff during a 100-year event (duration 24 hours, Huff 3rd
quartile precipitation distribution, precipitation 6.24") for the watershed.
We also assumed that 2 feet of storage was available within the open space site, so a site with a watershed
seven times the size of the site (7:1 watershed to site ratio) would most efficiently handle 3.5" of runoff
(watershed area x 3.5 inches/12/foot = storage area x 2 feet). Table 1 describes the ranking system created to
develop the Watershed/Site Area Ratio Score.
Table No. 1: Watershed/Site Area Ratio. A weight of 0 is assigned to sites with negligible on-site storage capacity for
runoff relative to the size of the contributing watershed. A weight of 10 is assigned to sites with optimum on-site
storage capacity for runoff relative to the size of the contributing watershed. Note each weight is assigned to a range
of ratios.
Watershed area: Weight
site area ratio
0:1 to 2:1
2:1 to 4:1
4:1 to 6:1
6:1 to 8:1
8:1 to 10:1
10:1 to 12:1
12:1 to 14:1
14:1 to 16:1
16:1 to 18:1
18:1 to 20:1
>20:1
3
6
8
10
9
8
7
4
2
1
0
Storage Effectiveness - The storage effectiveness of each site was calculated as a function of the area of the
site, and the ratio between the area of the site and the area of the contributing watershed. Larger sites that
efficiently store water are ranked higher than smaller sites that do not efficiently store water. The storage
effectiveness score was used to identify the 42 highest priority sites (7,065 acres) for protection (Figure 3).
280
-------
Figure No. 3: Soil analysis contributed to site assessment and prioritization.
Hydric Soils
Of 25 Acraa Hi Greater)
ThQ
Conservation
Oak Creek Watershed |
Legend
Slufy AnA «nd A i/Hfhl
Mjn Nnfid -A,
u^j i — Hr
281
-------
Site Importance to Flood Risk Reduction - Each of the 42 high priority sites were assessed as to their
importance for reducing flooding risks along the main stems of the primary channels of their watersheds.
The importance of the site was based on the proximity of the site to areas along the main stem projected to
have flood increases between the 1995 design year and 2020.
Flood projections were taken from Hydrologic Simulation Program-Fortran (HSPF) models prepared Camp
Dresser and McKee (2000 a,b,c). Sites were assigned a high priority location rank if they were located in
sub-watersheds that discharged into reaches of the main stem projected to have significant increases in the
100-year design flood substantially greater than projected increases on the main stem immediately upstream
of the site.
Sites were assigned a medium priority location rank if they were located in sub-watersheds that discharged
into reaches of the main stem projected to have increases in the 100-year design flood that were similar to
projected increases on the main stem immediately upstream of the site.
Sites were assigned a low priority location rank if they were located in sub-watersheds that discharged into
reaches of the main stem that were not projected to have increases in the 100-year design flood.
Final Ranking of Each Site - Each of the 42 high priority sites were ranked in order of 1-42 using weighted
variables described above. The rank of each site is described within each of the three watersheds as well as
within the entire project area. Table 2 indicates the final rank of each of the 42 high priority sites.
282
-------
Table No. 2: Final ranking of high priority sites by watershed as well as within the entire study area.
Restoration
Site
8
2
7
27
28
35
5
15
13
52
21
40
3
12
37
30
58
51
32
1
9
65
17
64
66
6
19
103
114
108
144
137
174
156
139
146
142
145
140
143
163
186
Watershed
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Menomonee River
Oak Creek
Oak Creek
Oak Creek
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Root River
Site
Area
(Acres)
250.2
667.2
265.3
105.4
104.3
71.7
312.7
188.5
208.9
51.4
145.5
64.3
354.7
226.1
68.7
95
47.4
55.2
84.4
673.7
230.6
43.1
155.8
44.1
42.6
292.8
152.6
138.9
65.3
73.8
135.3
420.3
44.8
88.3
239.7
119.4
188.6
120
195.3
148.9
54.9
29.9
Storage
Effectiveness
Score
85
80
76
63
62
60
87
81
74
42
69
36
58
55
51
45
27
43
21
80
75
15
71
23
22
19
26
68
43
55
76
59
31
44
38
25
54
50
36
9
16
14
Location
Rank
H
H
H
H
H
H
M
M
M
H
M
H
M
M
M
M
H
M
H
L
L
H
L
M
M
M
L
H
H
M
M
M
H
M
M
M
-
-
-
M
-
-
Watershed
Rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
1
2
3
1
2
3
4
5
6
7
8
9
10
11
12
Study
Area
Rank
1
2
3
5
6
7
8
9
11
13
14
15
17
18
21
22
23
25
27
28
29
30
32
33
34
35
40
4
12
19
10
16
20
24
26
31
36
37
38
39
41
42
283
-------
Parcel Prioritization
Each parcel within each of 42 high priority sites was evaluated and prioritized for acquisition based on the
potential storm water runoff storage each parcels would provide. The parcel evaluation methodology
consisted of a two-step process:
• Identification of parcels, boundaries and ownership within each of the high priority sites;
• Evaluation of the storage potential of each of the individual parcels.
Parcel Identification - Parcel boundaries and ownership was defined according to available land parcel
ownership records.
Parcel Storage Evaluation - The storage potential for each parcel within each of the 42 high priority sites
were determined as follows:
1. A site digital elevation model (SDEM) using Arc View ™ software was developed for each site.
2. The minimum elevation value (site runoff evaluation) along the perimeter of the site was extracted
from the SDEM.
3. A reservoir surface model was generated based on the minimum elevation value along the perimeter
of the site.
4. Ownership parcel boundaries were defined and put into the SDEM.
5. The potential volume of each parcel was calculated by using the SDEM elevation grid as the product
of the difference between the grid elevation and the minimum elevation along the site perimeter for
each SDEM grid and the area of the grid cell. Iterations were calculated based on existing
conditions, and the construction of 2-foot, 4-foot, and 6-foot berms.
6. Parcels were ranked and prioritized based on their potential storage at various berm heights.
While the parcel storage evaluation method provided an effective way to compare the potential storage
capacity of one parcel to another, the topographic drawings available to us were at too coarse of a scale to
permit an accurate representation of actual storage per parcel.
Site Action Plan - A site action plan was developed for each of the high priority sites. The site action plan
included an aerial base map indicating site limits and parcel boundaries within the site. Parcels were color
coated to indicate parcels with the most potential for storing water. Parcels were linked to a Microsoft 2000
ACCESS database that provided additional information useful to land negotiators, including ownership,
size, potential storage, and other information.
Partnership Opportunities and Potential Funding Mechanisms
Concurrently with the preparation of the Base GIS Information and Hydrologic Impact Site Analysis, staff
from The Conservation Fund investigated opportunities for partnering with land trusts, local units of
government and private landowners to own, hold easements, or manage Conservation Plan Sites. Staff from
Heart Lake Conservation Associates investigated methods to leverage the $15 million MMSD had allocated
to this effort to obtain additional monies through grants or gifts.
284
-------
Partnership Opportunities - Partnership opportunities with local units of government were evaluated by
identifying the overlap between each of the 13 local government's park and open space plans with
Conservation Plan sites. Eleven local units of government were surveyed. Eight of the 11 governments were
interested in working with MMSD to manage Conservation Plan sites long term.
Partnership opportunities with non-profit land trusts were evaluated by developing a list of land trusts
operating in the project area, and by determining whether the land trust met the minimum requirements for a
profile The Conservation Fund developed. Sixteen organizations were identified and 10 were interviewed to
determine interest and whether or not the organization met the profile. Two organizations expressed interest
and have the capability to own and manage 23 of the 42 Conservation Plan sites.
The Conservation Fund also explored potential partnership opportunities with the private sector including
private landowners, residential developers and commercial developers. Private landowners would be more
inclined to explore easement arrangements such as the Wetland Reserve Program, Crop Reserve
Enhancement Program and the Wisconsin Stewardship program. Commercial and residential developers
would more likely be interested in incentive for conservation developments.
Potential Funding Mechanisms - Heart Lake Conservation Associates identified and researched 30 grants
that MMSD might pursue to purchase and/or manage Conservation Plan sites and interviewed 18 agencies
and organizations. Public and private entities exhibited a high level of interest in supporting a Conservation
Plan they viewed as an innovative and exciting approach to deal with multiple objectives (flooding, water
retention, wildlife habitat, water quality, open space protection, etc.). Heart Lake estimated that MMSD had
the potential to double its $15 million investment through leveraging.
Heart Lake identified two broad categories of funding that might be leveraged. The first, existing grant
programs, is available to grant applicants that meet the criteria of the grant program. The second, that Heart
Lake termed "money to be found," has even greater potential for leveraging funding than grants. "Money to
be found" refers to MMSD developing successful partnerships and relationships with organizations that can
provide funds. It is not uncommon for agency staff to direct discretionary funds to a project because the
project is attractive, a priority for the agency, or will help an organization achieve its goals.
One nearly universal rule when soliciting funds from outside sources is that funding agencies tend to look
more favorably on projects that meet multiple objectives. A project that provides flooding, water quality,
wildlife habitat and recreational benefits and opportunities would be looked on more favorably than a
project with just flood reduction benefits.
Discussion
A Case for Protection
State and federal statues and regulations govern much of the activities that are permitted in floodplains,
floodways, wetlands and shore land zones. However, most of these resources are not given outright
protection by these statues or regulations, but are merely regulate as required by the statutes.
For example, floodplains and wetlands are frequently impacted by agricultural operations and development.
These impacts often result in filling, and reduced size and capacity to function. Many of these impacts are
285
-------
permissible by state and federal regulations with a permit. Whether or not these permits compensate for lost
resources is subject to debate.
Studies of wetland mitigation areas across the country have suggested that most wetland mitigation projects
designed to compensate for wetland fills fail to meet design standards. Isolated wetlands, which have been
regulated by the Corps of Engineers for more than 15 years, have lost their protection since February 2001
due to changing regulations.
Protection through acquisition or easement offers the very best way to ensure that areas currently used for
floodwater storage will be allowed to function in this way in the future. Where protection has not been
granted, the range of impacts and alterations to these important areas have contributed greatly to the current
flooding problems now experienced in our communities.
Flood Benefits of Protected Sites
An undeveloped open site provides two opportunities for floodwater runoff reduction. 1) Reduce the rate
and volume of runoff from the site itself; and 2) Reduce the rate and volume of runoff from the site through
on site management of floodwater runoff from a watershed tributary to the site.
Volume reduction is accomplished through retention (surface water is prevented from leaving the site).
Rate reduction is accomplished both by retention and by detention (surface water is temporarily stored on
the site and then slowly discharged at a controlled rate).
The type of land cover and vegetation on the landscape has a substantial effect on the amount of surface
water running off of the land. A typical urban development will result in surface runoff of approximately 3.5
inches from a 100-year recurrence interval design storm (duration 24 hours, Huff 3rd quartile precipitation
distribution, precipitation of 6.24"). An undeveloped fallow field with deep-rooted vegetation (i.e. prairie
plants) decreases surface runoff of a fallow field from 2.9 inches to 1.1 inches, providing retention of 1.8" of
floodwater runoff.
The construction of low berms provides an additional (and greater) volume of floodwater storage. Perimeter
berms can reduce floodwater runoff to zero inches. The installation of additional berms at strategic locations
throughout the site can retain storm water runoff to a depth of two feet that in turn provides two feet of
retention on a site. Such a strategy has the potential to reduce runoff to zero inches for an off-site tributary
area up to 6.5 times larger than the site itself.
Cost Effectiveness of Preservation
It is difficult to accurately measure the cost effectiveness of preserving and restoring open space to the
extent that flood benefits are realized. While the Conservation Plan provides a technically defensible
method for identifying and prioritizing land to protect, budget and data limitations prevented us from
precisely quantifying how much runoff each site or parcel could store.
The budget for preparing the Conservation Plan was less than $200,000. In the absence of funds to prepare a
1' or 2' topographic survey, we were forced to use U.S.G.S. 7.5" topographical data to quantify the potential
storage in sites and parcels. Storage numbers cited in the plan are most useful for comparisons between
sites and parcels rather than as a precise representation of actual storage provided.
286
-------
However, common sense and the use of reasonable assumptions indicate that preserving open space can be
very economical when compared to the costs of flood damages, conventional flood damage studies, the
costs of implementing conventional flood damage strategies, and costs associated with the loss of water
quality, habitat, and other open space opportunities when conventional strategies are exclusively used.
For example. MMSD has a goal of purchasing 5,000 to 7,000 acres of land over the next 5 years using the
$15 million budgeted for the project. If we assume that each acre of land would provide an average of two
acre-feet of storage (Eppich et al. 1998), the acquisition of 7,000 acres of land could provide approximately
14,000 acre feet of storage (7,000 acres x two feet of storage per acre = 14,000 acre feet of storage). That
translates into $1,071 per acre-foot of storage for land costs.
Cost per acre-foot of storage would increase once you add construction costs associated with restoring a site
to maximize its capacity to store floodwater. Costs for restoration can range from $1,000 to $5,000 per acre
which raises total cost per acre-foot of storage to $2,071 to $6,071 per acre-foot of storage.
It is useful to consider how these costs compare with traditional storm water detention facilities. The
Village of Arlington Heights, Illinois provides one such comparison. The Village allows some developers to
purchase storm water storage from a regional storm water detention facility in lieu of providing storm water
detention on site at a cost $l/cubic foot of storage, or $43,560 per acre-foot of storage.
Costs associated with a Phase n Corps of Engineers flood damage reduction project on the Des Plaines
River in Illinois provide another useful comparison. The maximum flood of record in 1986 caused $35
million in damage. The cost of just the study to determine what can be done is $9.8 million.
Logic suggests that costs associated with flood damages, preparing engineering studies to deal with flood
damages in conventional means, and constructing conventional flood damage reduction projects are far
greater than costs associated with protecting open space important in storing floodwaters.
Restoration ecologists and storm water management experts will argue without cease as to the virtues and
pitfalls of their respective approaches. If approached objectively, and with humility, such arguments are
healthy. Ecologists must have the numbers to back up assertions for alternative approaches; engineers must
recognize that models can turn into black boxes with simplistic answers to complex questions. However, no
alternatives to conventional practices will exist without the land on which to work.
Water Quality Benefits
Water quality benefits associated with storing storm water runoff in the natural landscape when compared
with no storm water management, or even conventional storm water management strategies where water is
piped to detention ponds, are substantial.
Coffman (2002) prepared a table summarizing research completed by the Center for Watershed Protection
that cites 16 papers published between 1979 and 1994 examining the relationship between urbanization and
stream water quality. These papers indicate significant reductions in the diversity of aquatic fauna once
total impervious cover in the contributing watershed approaches 10%.
287
-------
Liptan and Thomas (2002) cite a Portland Bureau of Environmental Service experiment in which a swale
planted in turf grass is compared with an identically configured swale planted in native prairie grasses and
forbs. The investigators found that runoff attenuation in the native swale was 41% compared with the turf
grass swale that was 27%. 68% of the total suspended solids (TSS) in the runoff were retained within the
native swale compared with 59% in the turf grass swale. It is important to note that if sewers were used for
conveyance rather than swales, attenuation of runoff and TSS would not be significant.
The Storm water Treatment Train™ concept uses constructed landscape features of upland prairies, swales
vegetated in native plants, wetlands and lakes to retain and treat runoff. Apfelbaum et al (1995) used HSPF
modeling to predict the effectiveness of this system in treating runoff from the Prairie Crossing conservation
development in Grayslake, Illinois, with the following results: Surface runoff would be reduced by 65%;
TSS would be reduced by 98%; total nitrogen would be reduced by 85%; and total phosphorus would be
reduced by 95%.
Lessons Learned and Additional Research
• This paper provides an original approach for quantifying the potential efficiency of open space to
provide storage for storm water runoff. While the topographic information at our disposal was too
coarse to provide a precise quantification of potential storage, the technique used permitted us to
make objective comparisons between sites and parcels. Higher resolution topographic data would
have allowed us to make precise quantification of potential storage using the techniques we
developed.
• Costs associated with flood damages, preparing studies to reduce flood damages, and implementing
conventional storm water management strategies to combat flooding, are enormous. This study
justifies allocating more resources toward studying alternative strategies that rely on preservation
and restoration as a cost effective means to combat flooding, as well as address other objectives such
as water quality, habitat, and open space benefits.
• The investigators were restricted to considering only privately held open space. We recommend
expanding the study to include publicly held open space for additional passive floodwater storage
opportunities.
• The ranking system did not include restoration measures on each site that could maximize the
potential for each site to store floodwater. We recommend expanding the study to consider how
restoration could maximize the potential for each site.
• This study concentrated on floodwater benefits of open space. We recommend additional work to
demonstrate how preserved open space will provide multiple benefits including water quality,
habitat, and other open space benefits.
• The investigators learned that it is absolutely essential to be sensitive and humble when proposing
alternative methods for combating flooding. Communities may wait years for flooding relief that
may or not be consistent with alternative strategies described in this paper. The investigators
acknowledge the value conventional storm water strategies have had in the past and will continue to
have today and into the future.
Conclusion and Summary
• This Conservation Plan identified 199 sites total 17,146 acres for further investigation. Thirty-four
sites totaling 2,417 acres were eliminated during field visits because they were already developed.
288
-------
Forty-two sites totaling 7,065 acres were identified as high priority sites. Remaining sites were
identified as low to medium priority for acquisition due to limited flooding benefits, an impractical
configuration for acquisition, or an excessive number of parcels.
• Interviews with potential partners (local governments, land trusts, others) indicate that 61% of the
high priority sites have entities that are "definitely" interested with MMSD.
• Thirty-four high priority sites containing up to 4,835 acre-feet of potential storage have been lost or
altered since 1995.
• Approximately $15 million is earmarked for the implementation of the Conservation Plan. While
variable land costs prohibit an accurate estimate of the amount of land that might be purchased with
available funds, this study indicates that costs associated with preserving and restoring important
open space is less than the cost of constructing traditional detention facilities to deal with existing or
future flood problems.
• This study provides an original approach for quantifying the potential efficiency of open space to
provide storage for storm water runoff. While the topographic information at our disposal was too
coarse to provide a precise quantification of potential storage, the technique used permitted us to
make objective comparisons between sites and parcels. Higher resolution topographic data would
have allowed us to make precise quantification of potential storage using the techniques we
developed.
• Conceptual cost estimates indicate that securing undeveloped sites and maximizing their natural
flood storage potential is cost effective compared with conventional flood control alternatives.
A cknowledgements
Portions of this paper first appeared in the MMSD's Conservation Plan (2001). Steve McCarthy was the
MMSD's project manager. Peg Kohring of The Conservation Fund led the consulting team. Mark O'Leary
of Applied Ecological Services, Inc. was the project manager for the technical team and writing the
Conservation Plan. Doug Eppich of Applied Ecological Services, Inc. assisted with the development of the
ranking system. Neal Thomas of Research Data, Inc., and Jason Carlson of Applied Ecological Services,
Inc. completed the GIS analysis. Steve Apfelbaum of Applied Ecological Services, Inc. provided leadership
and vision. Peter McKeever of Heart Lake Conservation Associates contributed to the funding portion of the
study. Megan Kelly, David Johannesen and Pam Sullivan provided additional assistance on the
Conservation Plan. Karen Sands and Kevin Shafer provided important reviews of the Conservation Plan. Iva
McRoberts edited and formatted this paper. The Project Team gratefully acknowledges their support and
assistance.
Bibliography
Apfelbaum, S.I., Eppich J.D., Price T.H, and M. Sands, 1995. The Prairie Crossing Project: attaining water
quality and stormwater management goals in a conservation development. Pages 33-38 in Using Ecological
Restoration to Meet Clean Water Act Goals, Chicago, Illinois.
Apfelbaum, S.I., 2001. A brief history and analysis of floodwater management engineering: experiments
with undefined certainty. From the Milwaukee Metropolitan Sewerage District's Conservation Plan. 2001.
Camp Dresser and McKee, 2000a. Watercourse Study for the Oak Creek Watershed. Prepared for the
Milwaukee Metropolitan Sewerage District.
289
-------
Camp Dresser and McKee, 2000b. Watercourse Study for the Menomonee River Watershed. Prepared for
the Milwaukee Metropolitan Sewerage District.
Camp Dresser and McKee, 2000c. Watercourse Study for the Root River Watershed. Prepared for the
Milwaukee Metropolitan Sewerage District.
Coffman, L.S., 2002. Low-impact development: an alternative stormwater management technology. Pages
97-124 in Handbook of Water Sensitive Planning and Design. Edited by Robert L. France. Lewis
Publishers. Boca Raton.
Engineering News, 1892. The drainage of the Kankakee Marsh. Drainage Journal, 14:323-324.
Eppich, J. D., Apfelbaum S.I., and L. Lewis, 1998. Working paper # 2: small wetlands use for stormwater
runoff management in the Red River of the North basin. From a Compendium of Technical and Scientific
Advisory Committee Working Papers. In press.
Ferguson, B.K., 2002. Stormwater management and stormwater restoration. Handbook of Water Sensitive
Planning and Design. Edited by Robert L. France. Lewis Publishers. Boca Raton. Pages 11-28.
French, Henry F., 1859. Farm Drainage. A.O. Moore & Co., New York, NY.
Liptan, T., and R. K. Murase, 2002. Water gardens as stormwater infrastructure (Portland, Oregon).
Handbook of Water Sensitive Planning and Design. Edited by Robert L. France. Lewis Publishers. Boca
Raton. Pages 125-153.
Pearse, C.K., and Samuel B. Wooley, 1937. The Influence of Range Plant Cover on the Rate of Absorption
of Surface Water by Soils.
Southeastern Wisconsin Regional Planning Commission, 1990. A stormwater drainage and flood control
system for the Milwaukee Metropolitan Sewerage District. Waukesha, WI. Community Assistance Planning
Report No. 152.
The Conservation Fund; Applied Ecological Services, Inc.; Resource Data, Inc.; Heart Lake Conservation
Associates; Velasco & Associates, K. Singh & Associates. 2001. Conservation plan. Prepared for the
Milwaukee Metropolitan Sewerage District.
Weaver, I.E., 1954. North American Prairie. Johnsen Publishing Company. Lincoln, Nebraska.
Weaver, I.E., and F.E. Clements, 1938. Plant Ecology. McGraw-Hill Book Company, Inc. New York.
290
-------
CRITICAL COMPONENTS FOR SUCCESSFUL PLANNING, DESIGN,
CONSTRUCTION AND MAINTENANCE OF STORMWATER BEST
MANAGEMENT PRACTICES
Richard A. Claytor, Jr., P.E.
Director of Watershed Services
Horsley & Witten, Inc.
Sandwich, Massachusetts
Abstract
This paper presents a common nomenclature for structural stormwater best management practices (BMPs)
and reviews the several critical elements that must be addressed to ensure that BMPs meet watershed
protection goals. A set of key planning, design and implementation elements is reviewed. The paper
documents some of the many possible pitfalls that planners, designers, and local officials are faced with
during the BMP implementation process. Several real world examples of successful and failed BMP
implementation are cited as illustrations. The old adage, "the devil is in the details," is illustrated to alert
stormwater management practitioners to critical components throughout the BMP implementation process.
Introduction
This paper presents a series of suggestions to help implement successful stormwater management best
management practices (BMPs). A nomenclature is introduced to understand the context of how planning,
design, and construction decisions vary depending on which stormwater practice is being discussed. Next, a
series of BMP performance factors are presented to help the reader understand the complex nature that
governs BMP effectiveness. Finally, several planning, engineering, construction and maintenance
considerations are reviewed that identify specific measures to help engineers, plan reviewers, and regulators
implement successful BMPs.
Background
Stormwater BMPs are commonly grouped into one of two broad categories, as so-called "structural"
management measures or as "non-structural" measures. For purposes of discussion, structural measures are
those that consist of a physical device or practice that is installed to capture and treat stormwater runoff for
a prescribed precipitation amount, frequently referred to as either the "water quality volume" or "first flush"
volume. Structural BMPs include a wide variety of practices and devices, from large-scale retention ponds
and constructed wetlands, to small-scale underground treatment systems, and manufactured devices. Non-
structural practices are generally defined as the operational and/or behavior-related practices that attempt to
minimize the contact of pollutants with stormwater runoff.
Over the years, there has been a great deal of confusion and uncertainty regarding BMP nomenclature. For
example, one person may use the term "wet pond" to describe a retention pond. Another may use the term
"retention pond" to describe an infiltration basin because runoff is "retained" within the pond until it is
infiltrated into the ground. Both are technically correct, since a wet pond "retains" runoff in a permanent
pool and an infiltration basin "retains" runoff within the underlying soils of a basin. This confusion arises
291
-------
because stormwater practitioners do not have a consistent BMP nomenclature whereby everyone knows
what everyone else is talking about. To help provide a consistent basis for comparison and discussion of
BMPs, many organizations, state agencies and others are developing naming conventions for the most
common stormwater treatment practices. Table 1 lists some of the various widely accepted structural
practices and provides a brief description of each. As illustrated in Table 1, the so-called structural
practices can be grouped into one of six major categories as ponds, wetlands, infiltration practices, filters,
open channels, and other practices. While Table 1 certainly cannot be offered as the "standard" for BMP
nomenclature, it recently has been adopted in a series of statewide programs in Vermont, New York,
Maryland, and Georgia. Figure 1 illustrates four of the more widely applied of these structural BMPs.
Another area of particular interest and concern to stormwater managers is the question of how effective
BMPs actually will be in meeting watershed protection goals, such as helping to achieve total maximum
daily load (TMDL) targets or implementation as part of EPA's Phase n Stormwater Program. This raises
the question, what watershed management objectives are BMPs being designed to solve? In general,
stormwater management measures are called upon to meet one or more of four major watershed planning
objectives, including:
Promoting groundwater recharge
Reducing pollutant loading to receiving waters
Mnimizing or eliminating accelerated stream channel erosion
Mnimizing or eliminating flooding
The management objective along with any site constraints will dictate which practice, or suite of practices,
is employed for implementation. For example, the typical dry detention pond or underground vault does
little to reduce pollutant loading, but can be reasonably effective in meeting channel protection and flood
control goals (Winer, 2002). Infiltration practices certainly promote groundwater recharge, but rarely are
capable of meeting flood control objectives. This paper will concentrate on those components that go into
the successful planning, design and implementation of BMPs to reduce pollutant export to receiving waters.
All of the structural stormwater management measures have some capability to remove pollutants, but their
effectiveness varies widely depending on the type of practice, design characteristics, site characteristics,
target pollutant constituents, and construction and maintenance factors. Watershed managers are
increasingly aware that there are limitations and uncertainty to structural BMP effectiveness. Consequently,
there is frequently a need to also employ a suite of "non-structural" practices to help meet watershed
protection goals. While the uncertainty of the effectiveness of non-structural practices is probably an order
of magnitude higher than that of structural BMPs, many practitioners recognize the need to do both.
While there are certainly several options available to watershed managers, the reality is that many practices,
both structural and non-structural, may simply be infeasible or impractical in certain situations.
Furthermore, there are other considerations, such as cost, unintended environmental consequences,
neighborhood acceptance, or maintenance burden that will affect the ultimate selection and implementation
of any given stormwater management strategy. The remainder of this paper will focus on those factors
affecting structural BMP performance and longevity. This is not to underestimate the role of non-structural
BMPs in the stormwater manager's toolbox, but simply to acknowledge that data in this arena is currently
under-represented in the literature.
292
-------
Table 1: Naming Convention of Common Structural Stormwater Management Practices for Water Quality
Management and Treatment (Adapted from CWP, 2002)
BMP
Group
Ponds
Wetlands
Infiltration
Filters
Open
Channels
Other
Practices
Practice Name
Dry Detention Pond
Dry Extended
Detention Pond
Wet Pond
Wet Extended
Detention Pond
Multiple Pond System
Shallow Marsh
Extended Detention
Wetland
Pond/ Wetland System
Gravel Wetland
Infiltration Trench
Infiltration Basin
Surface Sand Filter
Underground Sand
Filter
Perimeter Sand Filter
Organic Filter
Bioretention
Dry Swale
Wet Swale
Grass Channel
Hydrodynamic Devices
and Swirl
Concentrators
Oil and Grit Separator
Filter Strips
Practice Description
Dry ponds or vaults are generally designed to temporarily detain runoff from a
set of defined storm frequencies to provide peak flow attenuation for flood
control purposes.
Ponds that treat a prescribed water quality volume through extended
detention, a design option that holds runoff over a fixed detention time.
Ponds that provide storage for a water quality volume in a permanent pool.
Ponds that treat a water quality volume by detaining runoff above the
permanent pool for a specified minimum detention time.
A group of inter-connected ponds that collectively treat a water quality
volume.
Constructed wetlands that provide water quality treatment primarily in a wet
shallow marsh.
Wetland systems that treat a portion of a water quality volume by detaining
storm flows above the marsh surface.
Wetland systems that treat a portion of a water quality volume in a permanent
pool of a wet pond that precedes the shallow marsh wetland.
Wetland systems composed of wetland plant mats grown in a gravel matrix.
Infiltration practices that store a water quality volume in the void spaces of a
gravel trench or within a chamber or vault before being infiltrated into
underlying soils.
Infiltration practices that store a water quality volume in a surface depression,
before being infiltrated into underlying soils.
Filtering practices that treat stormwater by settling out larger particles in a
sediment chamber, and then filtering stormwater through a sand matrix.
Filtering practices that treat stormwater as it flows through an underground
sediment chamber and then into a sand-matrix filtering chamber.
Filters that incorporate a shallow sediment chamber and a sand filter bed as
parallel vaults.
Filtering practices that use an organic medium such as compost in the filter, or
incorporate organic material in addition to sand (e.g., peat/sand mixture).
Practices that incorporate shallow depressions with vegetation that treat
stormwater as it flows through a soil matrix.
Open vegetated channels or depressions explicitly designed to detain and
promote the filtration of stormwater runoff into a prescribed underlying soil
media.
Open vegetated channels or depressions with wetland vegetation designed to
retain water or intercept groundwater for water quality treatment.
Open vegetated channels or depressions designed to convey and detain a
water quality volume at a very slow maximum velocity with a minimum
residence time.
Hydrodynamic solids separation devices characterized by an internal
structure that creates a swirling vortex.
Flow separation devices designed to remove pollutants from stormwater
runoff through gravitational settling and trapping.
Vegetated areas with prescribed dimensions and slopes, designed to treat
sheet flow runoff from adjacent surfaces and remove pollutants through
filtration and infiltration (a.k.a., grass filter strips, filter strips, and forested
buffers).
293
-------
IDSCAPING AROUND POOL
PLAN VIEW
Wet Pond Schematic
(TO DETENTION FACILITY)
Infiltration System
Schematic
PLAN VIEW
OVERFLOW BERM.
6 INCH DIAMETER PERFORATED PVC RUNOFF FILTERS THROUGH GRASS
f- OBSERVATION WELL BUFFER STRIP (20' MINIMUM); GRASS
WITH SCREW TOP LID CHANNEL; OR SEDIMENTATION VAULT
•-.
: * « "^
•Of ft
-^
•»J~~*~~*~ PEA GRAVEL OR SAND FILTER LAYER
^— PROTECTIVE LAYER OF FILTER FABRIC
FILLED WITH 2 TO 5 INCH DIAMETER
ASHED STONE
(BANK RUN GRAVEL PREFERRED)
-— ;AND FILTERS" DEEP
>-^"^ OR FABRIC EQUIVALENT)
R NOFF EXF LTRATES THROUGH
ND T RBED SUBSOILS WITH A
MNM M RATE OF 0.5 INCHES PER HOUR
SECTION
PARKING LOT SHEET FLOW
i I I I I
j—yND£RDRiaiN{;:aUEC"O^6VSfS:M
I ,-FH.TFS BK-D
- STONE DIAPHRAGM
PLAN VIEW
UNDERDRAIN COLLECTION SYSTEM
PLAN VIEW
PROFILE
25-4'PLANTING SOIL
6" PERFORATED
PIPE IN 8" GRAVEL
JACKET
TYPICAL SECTION
Bioretention System
Schematic
Surface Sand
filter
Schematic
TYPICAL, SECTION
Figure 1: Illustration of four common structural stormwater BMPs (source, CWP, 2002) (the figure
illustrates the plan and profile schematic view of four BMPs: the wet pond, infiltration trench,
bioretention system and surface sand filter)
294
-------
Pollutant Removal Effectiveness
What are the characteristics or criteria that govern BMP pollutant removal effectiveness and how can one be
reasonably certain that BMPs will meet watershed management objectives? These are key questions that
watershed managers need to address in order to reliably predict benefits of stormwater implementation.
From the author's experience there are at least six separate variables that govern BMP pollutant removal
effectiveness. These include:
1. The estimated pollutant removal capability of the practices themselves, based on prior monitoring
2. The contributing drainage area that is physically directed to one or more BMPs
3. The fraction of the annual rainfall that is effectively captured by practices
4. The criteria that are employed for the design and implementation of new BMPs
5. The construction inspection and enforcement capabilities of watershed managers and/or agencies to
ensure that the design criteria are applied and implemented
6. The maintenance performance of BMPs over the long term
While several of these variables are self explanatory, it is worth a brief explanation to describe them in
greater detail. The estimated pollutant removal capability of specific BMPs is simply the pollutant removal
efficiency that has been calculated from monitoring data of actual field studies of BMP performance.
Generally, quoted removal efficiencies are based on the median removal values from a dataset of
performance monitoring studies. There are several factors that will govern the pollutant removal of a given
practice, including inflow concentration, internal geometry, storage volume, and several site characteristic
parameters such as soil type/sediment particle size, catchment size, watershed land use, and percent
impervious. Two of the most extensive datasets available are the National Pollutant Removal Database for
Stormwater Treatment Practices, 2nd Edition (Winer, 2000), and the US EPA/ASCE National Stormwater
BMP Database (www.bmpdatabase.org).
Unfortunately, watershed load reduction is not necessarily a direct function of the BMP removal efficiency
because often a portion of a watershed cannot be captured by stormwater BMPs. Watershed managers must
account for watershed areas and loads that do not drain directly to structural BMPs.
The next important factor is the fraction of the annual rainfall and resulting runoff that cannot be effectively
treated by structural BMPs. The pollutant removal rates for most BMPs represented in pollutant removal
databases are specific to a certain prescribed runoff volume. If BMP sizing criteria in a given watershed is
either higher or lower than the norm, watershed managers may need to adjust removal estimates
accordingly. Furthermore, the flow path, depth, area, and topographic complexity within a BMP site can
influence performance. For example, it has been surmised that pond and wetland geometry is an equally
important parameter to design volume in defining pollutant removal performance (Schueler, 1992 and
Strecker, et al., 1992). Designs that do not consider internal geometry criteria or ignore "short-circuiting"
possibilities are likely to be less effective.
The final two factors that govern BMP effectiveness relate to the quality of construction and the
maintenance performed over time. Many structural BMPs have unique and often subtle design features that
facilitate pollutant removal. For example, shallow marsh wetlands must have shallow water depths and
complex topographical features to maximize pollutant removal. Filtering practices must be constructed
within very tight elevation tolerances to ensure proper inflow and distribution across the surface area of the
295
-------
practice. Even small variances in the construction of these facilities can result in significant impacts to
pollutant removal performance.
Finally, long-term maintenance must be performed to achieve the stated pollutant removal estimates
established from prior monitoring studies. While there is not a great deal of research documenting BMP
effectiveness over time, at least one study of a constructed wetland in Minnesota found a significant
reduction in pollutant removal ten years after initial construction, primarily as a result of a lack of
maintenance (Oberts, 1997). Furthermore, the vast majority of facilities being evaluated in BMP
performance studies are less than three years old (Winer, 2000). The net result should be that watershed
managers and those developing watershed loading assessments should be prepared to discount pollutant
removal effectiveness in relationship to anticipated maintenance.
Planning for BMP Implementation
It all starts with planning. Remember the six P's? Poor Planning Produces Piss Poor Performance! Well,
it could not be any more appropriate than for stormwater BMP implementation. Stormwater practitioners
must understand the broad watershed management objectives, site-specific physical limitations, and a host
of other issues to select and locate the most effective BMP system. The selection of appropriate stormwater
practices involves a combination of the process of elimination and the process of addition. Typically, no
single practice will meet all of the stormwater management objectives at a given site. Instead, a series of
practices are generally required. Certain practices can be eliminated from consideration, based on one
limiting factor, but several practices may ultimately "survive" the elimination process. The most
appropriate practices are those that are technically feasible, achieve the benefits for watershed protection,
can be most easily maintained, and meet budget constraints of the owner.
The basic considerations for arriving at the most appropriate practice or suite of practices are governed by a
variety of factors, including:
Land use
Which practices are best suited for the proposed land use at the site in question? Conversely, some
practices are ill suited for certain land uses. For example, infiltration practices should not be utilized where
runoff is expected to contain high levels of dissolved constituents, such as metals or the gasoline additive,
MTBE.
Physical feasibility factors
Are there certain physical constraints at a project site that restrict or preclude the use of particular
practices? This involves an assessment of existing onsite structures, soils, drainage area, depth to water
table, slope or head constraints at a particular site. For example, stormwater wet ponds generally require a
drainage area approaching 25 acres unless groundwater interception is likely. They can also consume
significant land area.
Watershed factors
What watershed protection goals are needed within the watershed that the site drains to? This set of factors
involves screening out those practices that might be in conflict with overall watershed protection strategies.
For example, practices that contribute to thermal loading should be restricted in cold-water fisheries.
296
-------
Stormwater management control capability
What is the capability of a particular stormwater practice or suite of practices to meet the multiple
objectives of water quality control, channel erosion mitigation and/or flood control? Certain practices have
limited capabilities to manage a wide range of storm frequencies. For example, the filtering practices are
generally limited to water quality treatment and seldom can be utilized to meet large storm management
objectives.
Pollutant removal capability
How do each of the stormwater management options compare in terms of pollutant removal? Some
practices have a better pollutant removal potential than others or have a better capability to remove certain
pollutants. For example, stormwater wetlands provide excellent total suspended solids (TSS) removal but
only modest total nitrogen (TN) removal.
Environmental and maintenance considerations
Do the practices have important environmental drawbacks or a maintenance burden that might influence
the selection process? Some practices can have secondary environmental impacts that would preclude their
use in certain situations. Likewise, some practices require frequent maintenance and operation that is
beyond the capabilities of the owner. For example, infiltration practices are generally considered to have
the highest maintenance burden because of a high failure history.
Key Planning Considerations
Choosing the right BMP
While designers and reviewers alike may be familiar with the list of selection criteria cited above, many still
select BMPs primarily based on a single factor, cost. This is particularly true in the private sector, where
cost seems to be the overriding selection criteria. This includes the cost to design as well as the capital costs
of construction. Design firms submit competitive bids to clients and tend to select BMPs that are easy and
quick to design. The easiest designs are those that involve the implementation of proprietary products,
where vendors provide sizing computations and ready-drawn cad files. As a result, many sites end up with
"stormwater in a can" as the proposed BMP, yet in general, these practices provide no groundwater
recharge, little or no channel protection or flood control benefits, and often do little to remove pollutants of
concern. One example is from Lake George, New York, where a propriety product was installed to help
mitigate fecal coliform delivery to a downstream swimming beach. Unfortunately, this product had no
documented capabilities to remove bacteria and as it turned out, actually exported bacteria to the beach
(West, et al, 2001). Apparently, the right conditions existed in the system for bacteria reproduction.
In this climate of intense competition and modest profit margins, developers are increasingly unwilling to
weigh other factors beyond cost in the BMP selection process unless forced by regulatory agencies.
Another preferred practice has historically been the standard dry detention pond. In some jurisdictions,
however, the dry pond no longer meets required water quality performance criteria. For example,
Massachusetts requires an 80% total suspended solids (TSS) removal rate as part of the statewide
stormwater policy. The dry pond is not rated to remove this percentage and therefore developers frequently
turn to the wet pond as a substitute. The problem is that wet ponds are being proposed in several
applications where they likely will not function. In one example in Mattapoisett, Massachusetts, the
engineer and developer of a five-acre condominium project are implementing a 5,000 square foot, four-foot
deep wet pond with a drainage area of 4.3 acres, where the groundwater elevation is below the pond bottom
297
-------
for most of the year (Rizzo Associates, 2002). At best, one should expect to see eutrophic conditions at this
pond and frequent complaints from homeowners living nearby.
Site Surveys and Physical Investigations
A comprehensive site survey and physical investigations are perhaps the two most important BMP planning
considerations. At a minimum, a soils test and a simple site visit should be performed at all sites. Aside
from flat terrain, site soils and groundwater elevation are the most common limiting factors inhibiting
successful BMP implementation. Only a few of the filtering systems and the proprietary products can be
implemented in most soil conditions. Other practices such as ponds and wetlands must have soils suitable
for embankment construction and water retention. All infiltration practices must have soils with appropriate
percolation rates and separation between groundwater. Even open channels rely on either porous soils for
infiltration, or impermeable soils for retention. Poor underlying soils are perhaps the greatest single factor
leading to infiltration system failure. For example, approximately 55% of infiltration trenches installed in
one Maryland county had failed within five years of construction, most as a result of poor underlying soils
(Galli, 1992). In Massachusetts and several other states, at least a two-foot separation distance is required
between the seasonal high groundwater elevation and the bottom of any infiltration facility (MADEP/CZM,
1997). Failure to document water table elevations can lead to potential groundwater contamination and
inadequate treatment where groundwater mounds-up into the bottom of infiltration facilities.
The site visit can reveal limitations that may not appear in topographic surveys or geographic information
system (GIS) mapping. For example, specimen trees can be identified, located and avoided in subsequent
design plans, underground and surface utilities can be documented, subtle drainage patterns that might have
a significant impact on the design can be identified, or design constraints from adjacent property owners
might be revealed.
Development of the Stormwater Management Concept Plan
Before developing full-scale engineering construction drawings, designers should prepare a conceptual
design that clearly defines the location, type, and approximate size of the practice. At this stage,
preliminary hydrologic computations should be performed to arrive at the basic configuration of a facility.
Potential permitting issues can be identified and hopefully addressed. Typically, a preliminary cost estimate
is developed to give the owner some sense of the ultimate capital costs of implementation. Figure 2
illustrates the level of detail typically found at the conceptual stage. The primary purpose of the conceptual
plan is to present the design intent in sufficient detail so owners, reviewers, and regulatory staff can
understand the project plans and provide input prior to the development of more expensive engineering
construction drawings and specifications.
298
-------
Figure 2: Illustration of a typical stormwater management concept plan (Sourial and Claytor, 2002)(the
figure shows the level of detail typical of a stormwater management concept plan)
299
-------
BMP Design
Assuming an appropriate BMP has been identified and selected in the planning stage, the next opportunity
for success or failure is at the design stage. Generally, this stage is where most engineers do all right.
Engineers typically have a good education and training background to develop a set of sound construction
plans and specifications. However, there are a couple of key considerations that consistently seem to be the
vulnerable points in the design process.
Hydrologic and Hydraulic Computations
The development of hydrologic and hydraulic computations is the first point in the design process of a
stormwater management system, and the most crucial to get right, since all other design depends on the
answers. While the examination of hydrologic methods is beyond the scope of this paper, the following
considerations are worth noting:
Get the rainfall amount right. Many designers rely on the venerable National Weather Service
Technical Paper 40 (TP-40), which dates to the early 1960's, to obtain precipitation values for selected
storms (NRCS, 1986). While TP-40 is widely referenced in regulatory documents, more recent
research is probably more accurate. For example, the Northeast Regional Climate Center at Cornell
University has published recent data that is significantly different than those values represented in TP-
40 (Wilks and Cember, 1993).
Estimate a realistic time of concentration. The time of concentration is the single most sensitive
hydrologic variable that hydrologists rely upon to estimate peak flow rates. The use of an excessively
long overland flow condition can artificially distort the travel time and reduce peak discharge rates.
Examine land use assumptions to ensure that values are based on current and projected future
conditions.
Examine hydrologic soil group assumptions to make sure they are representative of actual watershed
conditions. In one example in the Catskill Mountains of New York, engineers used hydrologic group
"C" soils in an attempt to mimic a shallow-shale based soil profile that had large initial infiltration
potential and equally large interflow rates, but no relationship to the hydrologic conditions
representative of the "C" soil group.
Utilize appropriate assumptions when performing hydraulic modeling. Many errors occur in
describing the storage and outlet conditions of facilities that are very different from what ultimately
makes it to the design plan. Examples include: applying large infiltration rates where soil data show
modest or poor infiltration, over estimating the storage capacity of a pond, describing an outlet as a
single orifice where multiple releases are proposed, getting the invert elevations wrong, or simply
ignoring a contributing area in the hydrologic routing to a facility.
Soils and Structural Design
Almost all stormwater designs involve some requirement for soils information and in some cases,
reasonably complex geotechnical calculations for soil compaction, seepage diaphragm design or rapid
drawdown analyses, for example. Yet few BMP designs incorporate these measures. As a consequence,
poor soils analyses ranks as perhaps the most common factor leading to BMP practice failure. Designers
and reviewers must involve a reliable soils evaluator or geotechnical engineer in the design process and
incorporate their recommendations in the design. Again, according to Galli, (1992), soil limitations ranked
among the highest factor contributing to infiltration system failure. Design of infiltration BMPs must
include adequate subsurface investigations and reporting.
300
-------
Structural design is another key component for many BMPs. Typical examples include: adequate
foundation design for pond outlet control structures or underground vaults, retaining wall design for weir
walls or large outlet facilities, and concrete slab design for load bearing structures. Many hydrologic and
hydraulic engineers are unfamiliar with this component of design to the level of expertise required for some
applications. For example, one of the more notable stormwater facilities designed by this author was the
Wheaton Branch Retrofit facility constructed in Maryland in the early 1990s (Claytor, 1998). The Wheaton
Branch facility design required the modification of a nearly 30-year old riser that wasn't adequately
evaluated for structural integrity. As a consequence, the newly constructed facility developed failure cracks
that had to be remediated shortly after the facility was finished, at great expense and embarrassment to all
parties involved, especially, this author. So the point is, one must recognize that sometimes stormwater
design involves detailed structural calculations that involve an experienced structural engineer, do not be
bashful in seeking their expertise.
Seeking Adequate Storage Volume
The storage volume design element involves simply making sure a facility is large enough to accommodate
the appropriate design criteria. However, one cannot imagine the difficulty that this criteria imposes on
BMP designers. For one thing, a site is often simply not big enough to accommodate the required storage,
so designers tend to make the "hole in the ground" deeper to accommodate the criteria. Ponds can end up
excessively deep and frequently with steep side slopes. Another common problem arises when designing
shallow marsh wetlands. Designers are trying to meet the duel objectives of obtaining a minimum water
quality volume, while maintaining a shallow marsh system. Invariably, one or the other design objective
looses. Two examples illustrate this point. The first was one of the pilot stormwater retrofit projects
implemented in Montgomery County, Maryland in the late 1980's. In this facility, the planners and
engineers were trying to meet a minimum water quality volume within a limited area constraint. The result
was a 2-foot deep permanent pool that was intended to be a shallow marsh and instead resulted in a shallow
pond (see Figure 3a). Likewise, for a project completed on Staten Island as part of the "South Richmond
Bluebell Restoration" effort, a shallow marsh stormwater facility was planted with Pickerelweed
(Pontederia cor data) in 18 inches of water. Unfortunately, Pickerelweed does not typically survive in
depths over about 12" (Thunhorst, et al., 1993) and, again, another shallow open water pond was created
(see Figure 3b).
Figure 3a: 2-foot deep pond in Montgomery County
Maryland (illustrating open water where a
shallow marsh should be present)
Figure 3b: 18-inch deep pond in Staten Island,
New York (illustrating open water where
Pickerelweed should be growing)
301
-------
BMP Construction and Maintenance
This last area of successful BMP implementation involves the often-grueling process of getting designs
constructed properly, and ensuring that practices are maintained over the long term. Construction of BMPs
can be a very rewarding process. The satisfaction of seeing a set of design plans mature to a real world
facility is very fulfilling. Unfortunately, the construction process is often where the "successful
implementation" part of the process breaks down. There seem to be a number of commonalities, as
discussed below.
BMP Construction
There are a number of elements that contribute to a successfully constructed facility. Based on the author's
experience, it is hard to say whether one element is more crucial than another. However, it is certainly true
than any one flawed component can lead to a failed system. The following considerations are worth
particular attention:
Design drawings, details and specifications need to be clear, concise, unambiguous and correct.
While there are certainly many places where construction problems can occur, it all starts with the
engineering drawings. Engineers must take extra caution to produce plans that are error-free. Details
should be easy to interpret and free of vague information. Designers need to consider the "twelve-year
old rule." If one's twelve year old child will not understand it, then one is asking for interpretation
problems by the contractor. Interpretation problems often lead to contract change orders and usually
increase construction costs.
The design engineer should be involved in the construction process, if possible. Where it is not
possible, or preferable to retain the original designer, then an equally qualified engineer, who has
design experience with the specific BMPs being constructed, should be involved in the project.
"Involved with the project" means that the engineer supervises construction inspections, reviews shop
drawings, participates in construction progress meetings, and coordinates directly with the contractor
on critical construction issues.
The contractor should have prior experience building the specific BMPs being proposed. Most
construction contracts go to the low bidder. In fact, most municipal laws require that contracts go the
"lowest qualified bidder." The key word is "qualified." Bidding documents should contain specific
requirements for contractors to submit prior work experience that are used as part of a "qualified
bidder" assessment process. Many construction problems can be attributed to the fact that a contractor
has never seen anything like an "underground sand filter" before, for example. Conversely, a qualified
contractor can solve many unforeseen problems, often before they become problems.
Do not start construction in November when working in a cold climate. Many stormwater practices
involve earth moving operations, dewatering, and or stream diversions. Winter construction
complicates almost everything. A good example was the University Boulevard Retrofit project in
Maryland that started in the late fall of 1992 and finished about a year later. The original construction
duration was estimated to be 120 days with a anticipated start date in May. But the county
procurement process took over six months from the contract award to the "notice to proceed." While
the project resulted in a very successful BMP, the construction process was brutal. The contractor
could not meet compaction specifications due to excessive soil moisture, construction equipment was
routinely mired in muck, concrete curing required tenting, and stabilization of disturbed areas was next
to impossible. Not to mention the joy of attending weekly progress meetings in freezing weather (see
Figure 4).
302
-------
Figure 4: University Boulevard Retrofit project in Maryland - during and after construction
(illustrating the complexity of winter construction on the left and the successfully completed
project on the right)
"Work in the dry." Most BMPs are constructed at the bottom of a drainage system of one kind or
another, and projects are not usually completed before at least a few precipitation events. Designers
and contracts need to work together to divert storm flows around construction stages to prevent costly
delays and/or downstream sediment transport.
Make sure a professional land surveyor stakes out the project. Many projects end up being constructed
with just a small variance from the original design drawings. In most cases, this is all right, but in
some it means the difference between a successful project and failure. Shallow marsh wetlands require
the maintenance of extremely tight tolerances to foster the different depth zones required for a
complex wetland plant community. Filter strips function properly only when sheet flow is maintained.
The slightest imperfection in a level spreader will result in concentrated flow. Sand Filters, which also
rely on the distribution of flow across a level filter bed, need to be built to within very tight tolerances.
Provide construction inspections to ensure facilities are built in accordance with approved design
plans. This involves a commitment from the approving regulatory agency to develop inspection
standards, train personnel on how to perform inspections, and provide enforcement mechanisms for
those facilities that are not constructed in accordance with approved plans.
BMP Maintenance
The key to successful BMP implementation is to provide needed maintenance in a manner that ensures that
facilities will remain effective over the long term. A successful maintenance program should include at
least the following three components:
Inspection of facilities to identify and document material deficiencies
Technical resources on how to correct facility deficiencies
Enforcement provisions on how to deal with owners/operators who are unwilling or unable to correct
material deficiencies
In practice, the key to a successful maintenance program is to develop an adequate funding source to
perform inspections, correct facility deficiencies, and provide technical capabilities to owners/operators.
Adequate funding is perhaps the greatest single hurdle for small municipalities that seek to implement
successful stormwater management programs. The few communities that have succeeded have developed
303
-------
either aggressive fee structures funded by new development, stormwater utilities that collect fees from
existing residents and businesses that contribute to stormwater runoff impacts, or stormwater tax systems.
While a review of stormwater funding is beyond the scope of this paper, it is generally agreed that the
stormwater utility option appears to provide the most reliable source of funding for long-term maintenance
implementation
Conclusion
To summarize, successful implementation of stormwater management BMPs requires careful attention to
detail at several stages across the planning, design, construction and maintenance process. As
municipalities move into the implementation of EPA's Phase n Stormwater Rule, practitioners should be
aware of the several critical elements to successful BMP implementation. From the author's experience,
successful programs include a number of key ingredients, such as:
A comprehensive BMP design criteria that specifies such elements as practice selection, sizing
requirements, geometry, landscaping, and maintenance provisions
A training program for engineers and reviewers on the application of the design criteria
A well-defined permitting process that includes adequate protections to ensure that facilities are
constructed in accordance with approved plans (e.g., review fees, design checklists, surety,
enforcement provisions)
An adequately staffed and trained inspection force to ensure facilities are constructed in accordance
with approved plans
A long-term inspection and maintenance program to ensure facility function over time, and
A funding source to ensure that above provisions are capable of being implemented
While stormwater BMPs are conceptually relatively easy to understand, they are too often used as a blunt
instrument in a watershed manager's toolbox. They are a relatively simple technology that is being applied
to help solve a very complex interaction between natural systems and human activities. The unfortunate
message is that it may only take one lapse in judgment or lack of training on the part of any one of a variety
of individuals, organizations, or institutions to implement a measure that may be partially or wholly
ineffective at meeting the challenge of watershed protection. The hopeful message is that, from that
author's experience, with thoughtful attention and diligent effort from those involved in the process,
stormwater BMPs can be implemented successfully in a variety of applications to help meet a variety of
watershed management objectives.
References:
Center for Watershed Protection (CWP). 2002. The Vermont Stormwater Management Manual - Final
Draft. Final Report to the Vermont Agency of Natural Resources. Center for Watershed Protection,
Ellicott City, Maryland.
Claytor, R.A. 1998. An Eight Step Approach to Implementing Stormwater Retrofitting, in proceedings
from National Conference on Retrofit Opportunities for Water Resource Protection in Urban Environments.
February 9-12, 1998. Chicago, Illinois.
Galli, J. 1992. Analysis of Urban BMP Performance and Longevity in Prince George's County, Maryland.
Metropolitan Washington Council of Governments. Washington, DC.
304
-------
Massachusetts Department of Environmental Protection and Massachusetts Office of Coastal Zone
Management (MADEP/CZM). 1997. Stormwater Management: Volume One: Stormwater Policy
Handbook. Commonwealth of Massachusetts, Executive Office of Environmental Affairs, Boston,
Massachusetts.
Natural Resources Conservation Services (NRCS), formerly the Soil Conservation Service. 1986. Urban
Hydrology for Small Watersheds, Technical Release Number 55. US Department of Agriculture.
Washington, DC.
Wilks, D. S. and R. P. Cember. 1993. Atlas of Precipitation Extremes for the Northeastern United States
and Southeastern Canada. Northeast Regional Climate Center. Cornell University, Ithica, New York.
Oberts, G. 1997. Lake McCarrom Wetland Treatment Systems - Phase II Study Report. Metropolitan
Council Environmental Services. St Paul, Minnesota.
Rizzo Associates. 2002. Drainage Report: The Village at Mattapoisett. Mattapoisett, Massachusetts.
Schueler, T. 1992. Design of Stormwater Wetland Systems - Guidelines for Creating Diverse and Effective
Stormwater Wetland Systems in the Mid-Atlantic Region. Metropolitan Washington Council of
Governments. Washington, DC. 131pp.
Strecker, E. 1992. Pollutant Removal Performance of Natural and Created Wetlands for Stormwater
Runoff. Final Report to US EPA, Woodward Clyde Consultants, Inc. Portland, OR. 112pp.
Sourial, J., and R. Claytor. 2002. The Urban Japanese Stormwater Garden, finalist in the MIT/EPA
Stormwater Competition. Cambridge, Massachusetts/
Thunhorst, G.A., D. R. Biggs, and B. E. Slattery. 1993. Wetland Planting Guide for the Northeastern
United States. Environmental Concern, Inc. St. Michaels, Maryland.
Winer, R. 2000. National Pollutant Removal Performance Database for Stormwater Treatment Practices,
2nd Edition. Final Report to US EPA, Office of Science and Technology. Center for Watershed Protection.
Ellicott City, Maryland. 29pp.
West, T. A., J. W. Sutherland, J. A. Bloomfiled, and D. W. Lake, Jr. 2001. A Study of the Effectiveness of
a Vortechs Stormwater Treatement System for the Removal of Total Suspended Solids and Other Pollutants
in the Marine Village Watershed, Village of Lake George, New York. New York State Department of
Environmental Conservation. Albany, NY.
www.bmpdatabase.org. 2002. National Stormwater Best Management Practice (BMP) Database, website.
Geosyntec Consultants, Wright Water Engineers, and the Denver Urban Drainage and Flood Control
District. Sponsored by USEPA, Washington, DC. and the American Society of Civil Engineers, Reston,
Virginia.
305
-------
EVALUATION OF NPDES PHASE 1 MUNICIPAL STORMWATER
MONITORING DATA
Robert Pitt, Alex Maestre, and Renee Morquecho
Dept. of Civil and Environmental Engineering
University of Alabama
Tuscaloosa, AL 35487
Ted Brown, Chris Swann, Karen Cappiella, and Tom Schueler
Center for Watershed Protection
Ellicott City, Maryland 21043
Abstract
The University of Alabama and the Center for Watershed Protection were awarded an EPA Office of Water
104(b)3 grant in 2001 to collect and evaluate storm water data from a representative number of NPDES
(National Pollutant Discharge Elimination System) MS4 (municipal separate storm sewer system)
municipal stormwater permit holders. The data are being collected and reviewed to both describe the
characteristics of this data and to provide guidance to permit writers for future sampling needs.
There have been serious concerns about the reliability and utility of Phase 1 stormwater NPDES monitoring
data, mainly due to the wide variety of experimental designs, sampling procedures, and analytical
techniques used. On the other hand, the cumulative value of the monitoring data collected over nearly a ten
year period from more than 200 municipalities throughout the country has a great potential in characterizing
the quality of stormwater runoff and comparing it against historical benchmarks. This project is creating a
national database of Phase 1 stormwater monitoring data, providing a scientific analysis of the data, and
providing recommendations for improving the quality and management value of future NPDES monitoring
efforts.
Each data set is receiving a quality assurance/quality control review, based on reasonableness of data,
extreme values, relationships among parameters, sampling methods, and a review of the analytical methods.
The statistical analyses is being conducted at several levels. Probability plots are used to identify range,
randomness and normality. Clustering and principal component analyses are also being utilized to
characterize significant factors affecting the data patterns. The master data set is also being evaluated to
develop descriptive statistics, such as measures of central tendency and standard errors. We are testing for
regional and climatic differences, the influences of land use, and the effects of storm size and season, among
other factors.
This paper describes our data collected to date and presents some preliminary data summaries. We have
been collecting much data to date, and encourage any other communities with wet weather outfall data
collected as part of their NPDES permit program to contact us so we can include as much data as possible in
our final effort.
306
-------
Project Description and Background
The importance of this project is based on the scarcity of nationally summarized and accessible data from
the existing NPDES stormwater permit program. There have been some local and regional data summaries,
but little has been done with nationwide data. A notable exception is the CDM national stormwater database
(Smullen and Cave 2002) that combined historical NURP (Nationwide Urban Runoff Program) (EPA
1983), available urban USGS, and selected NPDES data. Their main effort has been to describe the
probability distributions of this data (and corresponding EMCs, the event mean concentrations). They
concluded that concentrations for different land uses were not significantly different, so all their data was
pooled.
Other regional databases also exist, mostly using local NPDES data. These include the Los Angeles area
database, the Santa Clara and Alameda County (CA) databases, the Oregon Association of Clean Water
Agencies Database, and the Dallas area stormwater database. These regional data are (or will be) included
in this comprehensive NPDES national database. However, we will not be including the USGS or historical
NURP data in this NPDES database due to lack of consistent descriptive information for the older drainage
areas. Much of the NURP data are available in electronic form at the University of Alabama student
American Water Resources Association web page at: http://www.eng.ua.edu/~awra/download.htm The
results from these other databases will be compared to our results during our final analyses to indicate any
important differences.
This new NPDES database is unique in that detailed descriptions of the test areas and sampling conditions
are also being collected, including aerial photographs and topographic maps for many locations which we
are collecting from public domain Internet sources. The land use information used is as supplied by the
communities submitting the data, although aerial photographs and maps are also used to clarify any
questions. Most of the sites have homogeneous land uses, although many are mixed. These characteristics
are all fully noted in the database.
This project is collecting stormwater runoff data from existing NPDES permit applications and permit
monitoring reports; we are conducting QA/QC (quality assurance/quality control) evaluations of these data;
and statistical analyses and summaries of these data. The final information will be published on the Internet
(such as on an EPA OW-OWM, Office of Water and Office of Wastewater Management, site and on the
Center for Watershed Protection's SMRC, Stormwater Manager's Resources Center, site at:
http ://www. stormwatercenter.net/). Some of the information is currently located at Pitt's teaching and
research web site at: http ://www.eng.ua.edu/~rpitt/.
The phase 1 NPDES communities included areas with:
• A stormwater discharge from a MS4 serving a population of 250,000 or more (large system), or
• A stormwater discharge from a MS4 serving a population of 100,000 or more, but less than 250,000
(medium system)
More than 200 municipalities, plus numerous additional special districts and governmental agencies were
included in this program. Part 2 of the NPDES discharge permit application specified that sampling was
needed and that the following was to be included in the application:
• Proposed monitoring program for representative data collection during the term of the permit.
• Quantitative data from 5 to 10 representative locations,
307
-------
• Estimates of the annual pollutant load and event mean concentration (EMC) of system discharges,
• Proposed schedule to provide estimates of seasonal pollutant loads and the EMC for certain
detected constituents during the term of the permit.
The permit applications were due in 1992 and 1993. For Part 2 of the application, municipalities were to
submit grab (for certain pollutants) and flow-weighted sampling data from selected sites (5 to 10 outfalls)
for 3 representative storm events at least 1 month apart. In addition, the municipalities must have also
developed programs for future sampling activities that specified sampling locations, frequency, pollutants to
be analyzed, and sampling equipment.
Numerous constituents were to be analyzed, including typical conventional pollutants (TSS, TDS, COD,
BODs, oil and grease, fecal coliforms, fecal strep., pH, Cl, TKN, NOs, TP, and PO/t), plus many heavy
metals (including total forms of arsenic, chromium, copper, lead, mercury, and zinc, plus others), and
numerous listed organic toxicants (including PAHs, pesticides, and PCBs). Many communities also
analyzed samples for filtered forms of the heavy metals. Our database includes information for about 125
different stormwater quality constituents, although the current database is mostly populated with data from
44 of the commonly analyzed pollutants (as summarized later in Table 3). Therefore, there has been a
substantial amount of data collected during the past 8 or 9 years from throughout the country, although most
of these data are not readily available, nor have detailed statistical analyses been conducted and presented.
Data Collection and Analysis Efforts to Date
As of mid-December 2002, 3,757 events from 66 agencies and municipalities from 17 states have been
collected and entered into our database. These locations are listed in Table 1. Table 2 lists 27 states where
municipalities have been contacted and we plan to target for our next phase of data collection. Figure 1
shows the locations of these municipalities on a national map. We anticipate excellent national coverage,
although we may have few municipalities from the northern west-central states of Montana, Wyoming,
North and South Dakota (where cities are generally small, and few were included in the Phase 1 NPDES
program).
Some of the municipalities that we have contacted (and some where we actually received data) have
information that could not be used for various reasons. One of the most common reasons for not being able
to use the data was that the samples had been collected from receiving waters (such as Washington state,
Nashville, and Chattanooga). We are using data only from well-described stormwater outfall locations.
These can be open channel outfalls in completely developed areas, but are more commonly conventional
outfall pipes. The other major problem is that the sampling locations and/or the drainage areas were not
described. We are using data with some missing information for now, with the intention of obtaining the
needed information later. However, there will likely still be some minor data gaps that we will not be able to
fill. In addition, the list of constituents being monitored has varied for different locations. Most areas
evaluated the common stormwater constituents, but few have included organic toxicants. The most serious
gap is the frequent lack of runoff volume data, although all sites have included rain data. Finally, if we
collect all the data we have asked for, our current project resources will not permit us to fully utilize them,
as it requires a great deal of time to enter and review this information.
The assembled data has been entered into a database which contains site descriptions (state, municipality,
land use components, and EPA rain zone), sampling information (date, season, rain depth, runoff depth,
sampling method, sample type, etc.), and constituent measurements (concentrations, grouped in categories).
308
-------
In addition, more detailed site, sampling, and analysis information has been collected for each sampling site
and included as supplemental information. We are using the reported land use information supplied by the
communities, and are verifying some with aerial photographs and maps. In many cases, the sampled
watersheds have multiple land uses and those designations are included in the database (we list the
percentages of the drainage as residential, commercial, industrial, freeway, institutional, and open space).
Our final data analyses will consider these mixed sites also, although the following preliminary results are
only for the homogeneous land use sites.
Preliminary Summary of Phase 1 Stormwater Data
We plan to acquire additional stormwater data before our final data analysis, and to complete many of the
missing records. The following data and analysis descriptions should therefore be considered preliminary
and will change with these additional data and analyses. However, we are presenting only our most basic
and robust analyses here for consideration. Our final report and data presentations will obviously be much
more comprehensive.
Table 1. Municipalities
ALABAMA
GEORGIA
Clayton County
Cobb County
De Kalb County
Fulton County
Gwinnett County
Atlanta
whose Data has been Entered into Database
IDAHO MINNESOTA
TEXAS
Jefferson County
Mobile
ARIZONA
Maricopa County
Tucson
CALIFORNIA
Alameda
Caltrans
COLORADO
Denver
Colorado Springs
Ada County Highway
District
KANSAS
Topeka
Wichita
KENTUCKY
Jefferson County
Louisville
Lexington
MASSACHUSETTS
Boston
Minneapolis
NORTH CAROLINA
Charlotte
Fayetteville
Greensboro
OREGON
Clackamas County
Eugene
Gresham
Portland
Salem
ODOT
Arlington
Dallas
Dallas County
Fort Worth
Garland
Harris County
Houston
Irving
Mesquite
Piano
Tarrant County
VIRGINIA
Arlington County
MARYLAND
_Anne Arundel County
Baltimore County
Baltimore City
Carroll County
Charles County
Harford County
Howard County
Montgomery County
Prince Georges County
State Highway
PENNSYLVANIA
Philadelphia
TENNESSEE
Knoxville
Memphis
Chesapeake County
_ Chesterfield County
Fairfax County
Hampton County
_Henrico County
Newport News County
Norfolk County
Portsmouth County
Virginia Beach County
309
-------
Table 2. Communities Targeted for Next Phase of Data Collection
ALABAMA ILLINOIS NEBRASKA
Madison
Huntsville - Madison
Montgomery
ALASKA
Anchorage
ARIZONA
Pima County
Mesa
Phoenix
Tempe
CALIFORNIA
Various Communities
COLORADO
Aurora
Lakewood
Littleton
DELAWARE
Wilmington
New Castle County
Rockford
INDIANA
Indianapolis
KANSAS
Kansas City
LOUISIANA
New Orleans
Shreveport
MASSACHUSETTS
Worcester
MICHIGAN
Ann Arbor
Flint
Grand Rapids
Sterling Heights
Warren
MISSISSIPPI
Jackson
Lincoln
Omaha
NEVADA
Las Vegas
Reno
Clark County
NEW MEXICO
Albuquerque
NEW YORK
Various Communities
NORTH CAROLINA
Durham
Raleigh
Winston-Salem
OHIO
Akron
Columbus
Dayton
Toledo
PENNSYLVANIA
Allentown
SOUTH CAROLINA
Greenville County
Richland County
Columbia
TEXAS
Abilene
Amarillo
Austin
Beaumont
Corpus Christ!
El Paso
Laredo
Pasadena
San Antonio
Waco
UTAH
Salt Lake County
Salt Lake City
WISCONSIN
FLORIDA
Various Communities
HAWAII
Honolulu County
Milwaukee
MISSOURI
OKLAHOMA
Independence
Kansas City
Springfield
Oklahoma City
Tulsa
Table 3 is a summary of the Phase 1 data we have collected and entered into our database as of mid
December 2002. The data are separated into six major land use categories: residential, mixed residential
(but mostly residential), commercial, industrial, institutional, and freeways. Our open space and other mixed
land use data are not included on these tables due to lack of space in this paper. This table also summarizes
all data combined. The total number of events included in the database is 3,757, with most in the residential
category. Many of the monitoring locations are characterized by mixed land uses. With the exception of the
mixed residential area, only the main land use categories are shown separately on this table. For most
common constituents, we have detectable values for almost all monitored events. However, filtered heavy
metal observations, and especially organic analyses, have many fewer detected values. This table shows the
percentage of analyzed samples that had detected values. The median and coefficient of variation (COV)
values are only for those data having detectable concentrations. If we included the non-detected results in
these calculations, extreme biases would invalidate many of the COV calculations. Our final analyses will
110
-------
further examine issues associated with different detection limits, multiple laboratories, and varying
analytical methods on the reported results and statistical analyses. See Burton and Pitt (2002), and the many
included references in that book, for further discussions on these important issues.
v>
Figure 1. Data has been obtained and entered in our database for the communities shown in black. The other
communities are targeted for our next data collection phase (plus Delaware, Alaska, Wisconsin, Southern
California, Florida, and Hawaiian communities).
Ill
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database
Land Use (Number of Events)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Cond. Hardness
Area % Precip. (uS/cm (mg/L
(acres) Imperv. Depth (in) @25°C) CaCO3)
3562
94
45
7.79
937
94
57.3
4.91
582
97
104
2.46
442
90
32
4.83
448
93
37.9
1.70
18
100
36
0.00
182
85
0.99
0.72
2036
100
50
0.44
558
100
37
0.44
239
100
40
0.28
211
100
80
0.11
255
100
71.8
0.32
18
100
45
0.00
154
100
80
0.13
3063
100
0.47
0.97
831
100
0.455
0.99
421
100
0.56
0.75
399
99
0.39
1.05
395
100
0.47
1.00
17
100
0.18
0.91
182
100
0.54
1.05
887
78
121
1.75
164
65
96
1.51
137
77
116
1.15
73
90
118.5
0.98
129
84
136
1.31
0
n/a
n/a
n/a
86
100
99
1.01
1115
81
39
1.45
223
76
31
0.98
146
75
43.4
0.90
120
94
36
1.04
114
79
37.3
1.09
0
n/a
n/a
n/a
128
99
34
1.85
PH
1690
86
7.4
0.11
247
74
7.13
0.12
341
88
7.3
0.10
152
91
7.1
0.13
205
86
7.2
0.11
0
n/a
n/a
n/a
111
100
7.1
0.11
112
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Fecal Fecal
Coliform Strep.
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
IDS
(mg/L)
3062
97
78
4.13
802
97
69
2.17
470
98
85
5.68
378
98
74
1.92
380
97
84
4.11
18
100
52.5
0.67
97
99
77.5
0.80
TSS
(mg/L)
3525
98
63
6.05
923
98
50
6.25
570
99
74.8
7.89
446
98
48
4.85
434
98
90
4.74
18
94
17
0.83
134
99
99
2.53
BOD5
(mg/L)
3135
94
8.3
4.45
867
96
9.05
3.34
557
92
7.16
1.37
410
94
12
1.12
377
94
9
6.34
18
89
8.5
0.70
26
85
8
1.26
COD (mpn/
(mg/L) 100 mL)
2796
96
52
4.79
746
97
55.5
3.49
444
98
40
1.47
353
96
60
1.01
339
96
61
2.17
18
89
50
0.91
67
99
100
1.06
1764
89
5000
4.64
382
87
7750
5.06
342
93
11000
3.21
215
87
3000
3.93
272
86
2400
6.11
0
n/a
n/a
n/a
49
100
1700
1.95
(mpn/
100 mL)
1142
91
16000
3.85
267
90
24000
1.89
160
94
25000
2.21
152
90
9200
2.84
176
92
13050
6.89
0
n/a
n/a
n/a
25
100
17000
1.21
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Nitrogen,
Total Phos.,
N02+NO3 Ammonia Kjeldahl filtered
(mg/L) (mg/L) (mg/L) (mg/L)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
3127
96
0.6
1.99
863
97
0.58
1.93
542
96
0.56
1.01
415
96
0.62
1.07
398
94
0.75
0.96
18
100
0.6
0.64
25
96
0.28
1.23
1874
75
0.44
3.45
564
87
0.31
2.14
255
57
0.36
2.96
285
85
0.57
2.52
243
91
0.52
3.60
18
89
0.31
0.53
79
87
1.07
1.73
3304
95
1.32
3.64
879
96
1.42
3.87
562
94
1.2
1.85
426
95
1.6
4.86
411
95
1.4
2.53
18
100
1.35
0.50
125
97
2
1.37
2470
89
0.12
2.44
656
90
0.16
0.98
399
90
0.11
3.70
295
85
0.1
3.25
301
90
0.1
1.25
18
83
0.14
0.53
22
95
0.197
2.13
Phos., Oil and
total Grease
(mg/L) (mg/L)
3307
96
0.27
8.74
885
96
0.31
8.13
554
95
0.27
7.98
425
96
0.23
7.36
403
97
0.27
6.79
18
94
0.17
1.04
128
99
0.25
1.76
1830
71
4
4.50
473
66
3.3
7.79
254
74
4
2.53
260
77
5
3.13
287
74
4
3.28
0
n/a
n/a
n/a
60
72
8
0.62
114
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Sb, As, As, Be,
total total filtered total
(H9/L) (ng/L) (ng/L) (ng/L)
Cd, Cd, Cr,
total filtered total
(lig/L) (ng/L) (ng/L)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
755
9
3
2.56
214
2
40
1.11
74
4
1
1.59
91
3
69
0.79
123
18
4.8
1.37
0
n/a
n/a
n/a
14
50
3
0.25
1425
49
3.3
2.42
366
37
3
2.42
170
65
4
3.78
165
38
2.5
0.79
219
58
5
0.94
0
n/a
n/a
n/a
61
56
2.4
0.70
209
27
1.5
1.00
32
6
1.48
0.50
18
28
2
0.84
21
10
1.5
0.47
23
13
1
0.43
0
n/a
n/a
n/a
72
50
1.43
1.15
842
10
0.31
2.74
239
11
0.4
2.92
76
16
0.3
2.86
112
6
0.5
1.99
164
12
0.345
2.55
0
n/a
n/a
n/a
12
17
0.3
0.47
2481
49
1
4.42
599
38
0.5
5.20
398
51
0.9
3.53
303
54
0.86
5.02
329
60
1.9
3.77
18
17
0.5
0.69
95
72
1
0.90
389
31
0.5
1.69
85
6
0.7
0.55
30
40
0.3
0.64
48
25
0.33
2.26
42
55
0.6
1.10
0
n/a
n/a
n/a
114
26
0.68
1.03
1561
63
7
1.47
383
50
4.55
1.31
172
72
8
1.62
201
66
6
1.38
215
72
15
1.13
15
0
n/a
n/a
76
99
8.3
0.71
115
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Cr, Cu, Cu, CN,
filtered total filtered total
(H9/L) (ng/L) (ng/L) (ng/L)
Pb, Pb, Hg,
total filtered total
(lig/L) (ng/L) (ng/L)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
260
61
2.08
0.74
33
27
1.28
0.59
21
52
2
0.80
27
41
2
0.59
36
56
3
0.73
0
n/a
n/a
n/a
101
78
2.3
0.70
2770
86
16
2.24
719
84
11.1
1.60
421
85
18.7
1.31
360
96
15
1.55
372
91
21.8
2.01
17
41
17
0.59
97
99
34.7
0.95
413
83
8
1.68
91
64
7
1.92
30
73
5.75
2.33
49
80
8
1.50
42
90
8
0.67
0
n/a
n/a
n/a
130
99
10.9
1.50
1012
8
5
2.62
325
7
5
1.93
82
6
0.01
2.20
144
15
0.013
1.69
177
10
5.92
1.60
0
n/a
n/a
n/a
3
0
n/a
n/a
2902
80
15.9
1.89
704
75
12
1.95
501
78
19
1.34
345
95
17
1.70
372
83
23.7
1.90
0
n/a
n/a
n/a
100
100
27.5
1.44
446
50
3
2.01
109
34
3
1.84
30
47
3
0.68
59
54
5
1.61
51
53
5
1.58
0
n/a
n/a
n/a
126
50
1.8
1.65
1014
11
0.2
1.17
252
10
0.2
1.14
100
19
0.3
0.85
133
11
0.2
0.79
178
11
0.1
1.89
0
n/a
n/a
n/a
34
6
0.19
0.80
116
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Ni, Ni, Se, Ag,
total filtered total total
(|ig/L) (ng/L) (ng/L) (ng/L)
Zn, Zn,
total filtered
(lig/L) (ng/L)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
1602
40
9
2.08
381
33
6
1.19
179
28
10
0.84
203
58
7
1.82
225
53
20
0.87
15
0
n/a
n/a
79
87
9.2
0.92
246
64
4
1.47
25
44
2
0.51
25
72
5.5
0.87
23
48
3
0.84
36
58
5
1.43
0
n/a
n/a
n/a
95
67
4
1.38
912
9
2
1.48
246
7
2
0.54
80
9
4
0.89
118
7
2.5
0.82
175
10
2
0.98
0
n/a
n/a
n/a
16
6
2
n/a
1149
14
3
4.63
297
17
5
4.33
92
10
2800
2.02
148
20
5
3.02
216
23
1
4.28
0
n/a
n/a
n/a
21
19
0.35
0.87
3053
95
112
4.59
728
96
73
4.33
505
92
97
1.06
366
100
150
1.26
387
98
220
2.28
18
100
305
0.81
93
97
200
1.01
383
96
51
3.91
90
90
32
0.85
28
100
48
0.88
49
100
59
1.37
42
95
111.5
3.62
0
n/a
n/a
n/a
105
99
51
1.86
117
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Bis(2-
ethylhexyl) Di-n-butyl
Methylene- phthalate phthalate Fluoranthene
chloride (\iglL) (ng/L) (jig/L) (\iglL)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
251
36
11.2
0.77
104
33
11.3
0.93
23
43
9.05
0.51
42
21
9.2
0.40
33
33
9.7
0.40
0
n/a
n/a
n/a
0
n/a
n/a
n/a
250
30
9.5
1.13
143
20
4.5
1.68
26
15
5.1
0.38
72
44
10.1
1.07
49
43
10
0.81
0
n/a
n/a
n/a
0
n/a
n/a
n/a
93
16
0.8
1.03
22
18
10
0.64
8
13
14
n/a
20
25
0.7
1.39
12
25
0.7
0.09
0
n/a
n/a
n/a
0
n/a
n/a
n/a
259
19
6
1.31
145
3
3
1.21
26
0
n/a
n/a
75
35
5.9
4.38
51
25
3.8
0.97
0
n/a
n/a
n/a
0
n/a
n/a
n/a
118
-------
Table 3. Summary of Available Stormwater Data Included in NPDES Database (cont.)
Phenanthrene Pyrene Diazinon
(|ig/L) (ng/L) (jig/L) 2, 4-D (ug/L)
All Data Combined (3757)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Residential (983)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Mixed Residential (584)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Commercial (464)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Industrial (471)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Institutional (18)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
Freeways (185)
Number of observations
% of samples above detection
Median of detected values
Coefficient of variation
233
13
3.95
1.00
136
3
1.7
0.70
23
0
n/a
n/a
70
31
4.05
4.50
47
17
9
0.72
0
n/a
n/a
n/a
0
n/a
n/a
n/a
249
14
5.2
1.24
140
4
2.2
0.30
26
0
n/a
n/a
75
35
5
4.57
47
21
7.2
0.73
0
n/a
n/a
n/a
0
n/a
n/a
n/a
79
22
0.06
1.90
11
36
30
0.40
1
0
n/a
n/a
19
42
0.045
0.49
9
33
0.72
1.40
0
n/a
n/a
n/a
1
100
0.05
n/a
101
35
3
0.86
11
64
8
0.72
2
50
5
n/a
13
69
3
0.94
3
100
2
1.14
0
n/a
n/a
n/a
1
0
n/a
n/a
119
-------
Data Analyses
Statistical analyses are being conducted at several levels. First, probability plots are used to identify range,
randomness, and normality. Figure 3 (end of paper) is an example of log-normal probability plots for some
of the constituents and for all data pooled. Probability plots shown as straight lines indicate that the
concentrations can be represented by log-normal distributions. This is important as it indicates that data
transformations, or the use of nonparametric statistical analyses, will be needed. Other plots with obvious
discontinuities (such as for bacteria, phosphorus, lead, and zinc) imply that multiple data populations may
be included. Our future analyses will identify the significance of these different data categories (such as
land use, region, and season).
Clustering and principal component analyses (PCA) are also being utilized to characterize expected factors
influencing sample variability. Figure 4 is an example dendogram from a cluster analysis of all of the
preliminary data combined. This plot indicates very close relationships between rain depth and the nutrients
(total phosphorus, dissolved phosphorus, nitrite plus nitrate, ammonia, and Total Kjeldahl Nitrogen). Some
of the heavy metals (cadmium, nickel, and chromium) are closely related to each other, but copper, lead and
zinc are much more independent. BODs, COD, dissolved solids, and suspended solids are poorly related to
other pollutants for the pooled data. Pearson correlation analyses did show relatively strong relationships
between suspended solids and the total forms of most of the heavy metals, substantiating the observation
that most of the stormwater metals are not in filtered forms.
The master data set will also be evaluated to develop descriptive statistics, such as measures of central
tendency and standard errors. The runoff data will then be evaluated to determine which factors have a
strong influence on event mean concentrations, including sampling methods. We will test for regional and
climatic differences, the influence of land use, and the effect of storm size, among other factors. Figure 5
includes example scatter plots of COD vs. BODs and filtered copper vs. total copper, illustrating these
suspected close relationships. Also shown on this figure are scatter plots of suspended solids and
phosphorus concentrations for different rain depths. Little variation of these concentrations with rain depth
are seen when all of the data are combined, implying little likelihood of important "first-flush" effects at
stormwater outfall locations. Specific comparisons of concentrations from first-flush samples with
concurrent composite samples will be a more direct test and will be conducted later.
Figures 6 and 7 are example grouped box and whisker plots of suspended solids, total Kjeldahl nitrogen,
fecal coliforms, and copper, grouped for different major land uses and for different seasons. The TKN and
copper observations are lowest for open space areas, while the freeway locations had the highest values.
Suspended solids and fecal coliform variations are not as obvious, although it is likely that the freeway
bacteria values are significantly lower than those found in residential areas. The seasonal variations are not
as obvious, except that the bacteria values appear to be lowest during the winter season (a similar
conclusion was obtained during the NURP, EPA 1983, data evaluations). Preliminary statistical ANOVA
analyses for all land use categories (using SYSTAT) found significant differences for land use categories
for all pollutants. Our final analyses will further investigate this important finding and will also examine
possible confounding factors.
A major goal of these analyses will be to provide guidance to stormwater managers and regulators.
Especially important will be the use of this data as an updated benchmark for comparison with locally
collected data. In addition, this data may be useful for preliminary calculations when using the "simple
method" for predicting mass discharges for unmonitored areas. This data can also be used as guidance when
320
-------
designing local stormwater monitoring programs (Burton and Pitt, 2002), especially when determining the
needed sampling effort based on expected variations.
We will also be examining trends of concentrations with time. A classical example would be for lead, which
is expected to decrease over time with the current use of unleaded gasoline. Older stormwater samples from
the 1970s typically have had lead concentrations of about 100 ng/L, or higher, while most current data
indicate concentrations in the range of 1 to 10 [ig/L. Figure 8 is a plot of lead concentrations for residential
areas only, for the time period from 1991 to 2002. This preliminary plot shows likely decreasing lead
concentrations with time for all residential sites combined. However, more work is needed to investigate
interacting factors and other relationships of potential interest in order to reduce the variability inherent in
this (and the other preliminary) plots.
Our final analyses will expand on these preliminary examples and will also investigate other stormwater
data and sampling issues. As an example, we will compare "first flush" samples with composite samples for
a number of locations and conditions (the above data only represent composite samples) and will also
compare data collected manually vs. automatically.
As we are still collecting information for the database, we encourage all local and state agencies who have
Phase 1 municipal stormwater data but have not previously sent it to us, to please contact us so we can
arrange to have your data included in our final analyses.
References
Burton, G.A. Jr., andR. Pitt, 2002. Stormwater Effects Handbook: A Tool Box for Watershed Managers,
Scientists, and Engineers. CRC Press, Inc., Boca Raton, FL. 911 pgs.
Smullen, J.T. and K. A. Cave, 2002. "National stormwater runoff pollution database." In: Wet-Weather Flow
in the Urban Watershed, edited by R. Field and D. Sullivan. Lewis Publishers. Boca Raton, pgs. 67-78.
U.S. Environmental Protection Agency, Dec. 1983. Results of the Nationwide Urban Runoff Program.
Water Planning Division, PB 84-185552, Washington, D.C.
Acknowledgements
Many people and institutions need to be thanked for their help on this research project. Project support and
assistance from Bryan Rittenhouse, the US EPA project officer for the Office of Water, is gratefully
acknowledged. The many municipalities who worked with us to submit data and information were
obviously crucial and the project could not be conducted without their help. Finally, a number of graduate
students at the University of Alabama (especially Veera Rao Karri, Sanju Jacob, and Sumandeep Shergill)
and employees of the Center for Watershed Protection are also thanked for their careful work.
321
-------
s:
•f,
J •
r
• n
r
IIJ -IIIJ
Figure 3. Log-normal probability plots of selected stormwater quality data.
322
-------
Cluster Tree
ACRE
BODS
IMPERV
I
0
100 200 300
Distances
400 500
Figure 4. Cluster analysis (dendogram) showing relationships between stormwater pollutants.
323
-------
a 10
Pf'«*pitSliOi1 DUfJlft (irsi
0*p*h (if I
Figure 5. Example scatter plots of stormwater data.
324
-------
UJ
"a i»—
* 1
I I
,-f
s X c/' *<
,„
I
JC*"
1 1 1 . , i
' '
i i • ;
*
*
I
u
*
i !.'
I—
7 I
f f~'
^ y
c-F
Figure 6. Example stormwater data sorted by land use (no mixed land use data included in plots).
325
-------
•»-« .,
*
0 '«-5 - * *
; • S .
I ; ,
i '•*.* . „ *
i . •
a. £
i * * « •
(A I
1*11
§ | 1 | 1
/' / <* /
E *
. I * i
-' i *
§ E f I I I
Mi:
H ,e-l i 1 ' ' 1 I
! ! • I
B [ .
S If'4 -
» i
a. i
10 a)
*
>-' i an ,
£ *
i *
o" * 1
! I
1 1 *
•z i
i
0 1 - i *
(030 , *
1
i- 1 i
0 •
o •
ID ,
II
, 1
*
S
i
i I
1 :
i •
*
«
i 1
.
1.
I
Figure 7. Example residential area stormwater pollutant concentrations sorted by season.
326
-------
10000
1000
100
•en
"i 10
fl>
ra
-H-I
o
[_ 1
0.1
I.Q1
• *• •
. **jj$ ff1^ 0pT^rT^rSf
m «* * * *^«*%
• _.** .••••••
.-"*
Figure 8. Residential lead concentrations with time.
327
-------
Funding Phase II Storm Water Programs
Andrew J. Reese
AMEC Earth & Environmental, Inc.
Nashville, TN
615-333-0630
andrew.reesefo) amec.com
Abstract
Most Phase n cities are now in the midst of looking at how to fund their stormwater Phase n programs. The
cost of Phase n is widely variable but expected to be in the range of $3.75 to $6.00 per citizen per year
when the program is fully formed. Not all of those costs are new line items in a local budget. This paper
explores an approach for funding that combines a variety of methods or sources available to most local
governments - many of them not requiring new funds at all but using human resources instead. A hierarchy
of methods is established and a cost effectiveness method of program development defined.
Introduction
NPDES Phase n programs are in the final stages of planning. Assuming you have the authority and
organizational issues worked out (a BIG assumption), at about this point in the process Municipal Separate
Storm Sewer system owners and operators are asking the difficult question: "so how do we pay for the six
minimum controls?" Perhaps a better question is, "how can I best define a program that I can pay for?"
Under Phase I many communities defined a program, often in a vacuum, and then attempted to find ways to
fund it. Under Phase n the majority of the efforts under the six minimum controls required are highly
integrated with current stormwater program efforts. Thus, it makes sense to formulate a stormwater
program by working from both ends toward the middle - funding or resource sources and program
requirements.
Phase II Costs
There have been several attempts to
estimate the probable costs of the NPDES
Phase n stormwater program. EPA's
overall annual estimate for all permittees
is nearly one billion dollars. Most
individual MS4 estimates are expressed in
terms of cost per person per year, though
the actual costs do not always lend
themselves very well to this yard stick.
EPA itself, based on very scattered data
and surveys, established their cost estimate
as $1,525 per permittee + $3.50 per person
which, plotted, looks like Figure 1.
Reese, et al, (2000) provide cost estimates
for model stormwater Phase n programs
for a small town and a city of 50,000 in
EPA Annual Cost Estimates
Final Rule
$400
0
0)
"CD $250 -
to
O
15
-------
population. However, as they point out, there are great variations in the potential costs of any stormwater
Phase II program due to such things as:
• Character of the MS4
• Climate and geology
• Preferences of the permit writer and specific requirements of the state
• Maturity of current stormwater program
• Character of stream quality and need for improvement
• Ability to share costs with others
Based on that analysis and subsequent work by the Denver Urban Drainage and Flood Control District
(personal communication) a range of cost (on a per person per year basis for a fully developed Phase n
program) was established between about $1.50 and $8.00 in today's dollars for a very minimal and fairly
well developed stormwater program for a city of 50,000 (Reese, et al, 2000). This range is not very helpful
in actually estimating Phase n program costs other than to point out and illustrate the great variability and
flexibility in the program.
Another way to arrive at the potential cost is to recognize that most MS4s that have already implemented a
fairly advanced stormwater quality program spend about 15 to 25 percent of their total stormwater dollars
on stormwater quality aspects - a subset of which is Phase n compliance. Figure 2 shows typical
stormwater program costs for a range of stormwater program maturities on a per developed acre per year
basis. This is based on the author's firm's experience in over 100 cities and counties.
Dollars per Acre per Year
$200
$150
$100
$50
$0
Figure 2. Average Annual Per Developed Acre Stormwater Program Costs
Assuming typical numbers of about three persons per acre (2000 per square mile), and that stormwater
quality compliance aspects make up roughly 15 percent of the program then for a moderate program the cost
of the stormwater quality program is in the range of $3.75 to $6 per person per year.
329
-------
However it needs to be stated that not all of these costs are monetary, and not all of them are new costs.
These numbers simply reflect a level of effort necessary to implement the permit, not a budgetary line item
in some City's comprehensive annual financial report. As we will see below, that effort can be realized in
many ways, not all of them fully budgetary.
The MEP Standard and Cost
The NPDES regulatory compliance program for stormwater is based on the dual standard of "prohibition"
and "maximum extent practicable (MEP)." Prohibition means keeping non-stormwater from the stormwater
system. MEP means addressing and mitigating all the ways pollutants get into the system including dirty
stormwater, and doing so to one's maximum ability.
MEP consists of the mix of Best Management Practices (BMPs) and measurable goals that will attain
reduction of pollution to attain water quality standards. This is described in 40 CFR 68754, Dec. 8th, 1999,
as follows (italics mine):
The pollutant reductions that represent MEP may be different for each small MS4, given the
unique local hydrologic and geologic concerns that may exist and the differing possible
pollutant control strategies. Therefore, each permittee will determine appropriate BMPs to
satisfy each of the six minimum control measures through an evaluative process. EPA
envisions application of the MEP standard as an iterative process. MEP should continually
adapt to current conditions and BMP effectiveness and should strive to attain water quality
standards. Successive iterations of the mix of BMPs and measurable goals will be driven by
the objective of assuring maintenance of water quality standards. If, after implementing the
six minimum control measures there is still water quality impairment associated with
discharges from the MS4, after successive permit terms the permittee will need to expand or
better tailor its BMPs within the scope of the six minimum control measures for each
subsequent permit. EPA envisions that this process may take two to three permit terms.
MEP really depends on the consideration of several things as illustrated in Figure 3:
• Do I have, or can I obtain, the legal authority to carry out the program I am describing?
• Is my technical approach sound in that it is a "proven" approach, structural or non-structural that
addresses pollutants of concern in an effective manner?
• Are my defined procedures, policies, staff resources and equipment appropriate for the level and
type of program described?
• Do I have, or can I obtain, dedicated and sufficient funding to support the program I am describing?
Currently there are no specific numeric criteria for stormwater discharges (unless established under a TMDL
or court induced program), and there will not be until 2013. MEP is considered a flexible, narrative,
technology-based standard. If you do what you say you are going to do you are, by definition, in
compliance - regardless of the actual water quality. Monitoring may be required in the second round for a
percentage of MS4's to prove that water bodies are attaining water quality standards. If not.. .the
requirements will be tightened. Remember that the congressionally mandated goal is to meet water quality
standards (as they are currently defined or may change as newer wet weather approaches are developed),
330
-------
Maximum
Extent
Practicable
Public Education
Public Involvement
Illicit Connection
CuHslfUUilMi
ru\iii
Pollution Prevention
s
I/
s
s
s
s
I/
s
V
Figure 3. Definition of Maximum Extent Practicable
and EPA plans to negotiate a change in the definition of MEP for you on the basis of existing or collected
monitoring information in each successive permit period.
Language throughout the preamble to the permit language and in the congressional record describing MEP
definitions also contains the term "cost effective" when it describes BMP programs. This term "cost
effective" has not been defined either but can serve as a critical basis when selecting among BMP options,
the level of the stormwater quality program, and funding needs.
The fact that cost should and can be considered when developing an MEP program is incontrovertible - to
what extent, that is a source of controversy and must be balanced with other considerations. Consider:
• President Clinton's Clean Water Initiative (USEPA, 1994) addressed a number of issues associated
with NPDES requirements for storm water discharges and proposed establishing a phased
compliance with a water quality standards approach for discharges from municipal separate storm
sewer systems with priority on controlling discharges from municipal growth and development areas
and clarifying that the maximum extent practicable standard should be applied in a site-specific,
flexible manner, taking into account cost considerations as well as water quality effects.
• EPA has stated (see footnote 1) that MS4s need the flexibility to optimize reductions in storm water
pollutants on location-by-location basis. EPA envisions that this evaluative process will consider
such factors as conditions of receiving waters, specific local concerns, and other aspects included in
comprehensive watershed plan. Other factors may include MS4 size, climate, implementation
schedules, current ability to finance the program, beneficial uses of receiving water, hydrology,
geology, and capacity to perform operation and maintenance.
331
-------
• In California the State Water Quality Board provided the following explanation of MEP1: "There
must be a serious attempt to comply, and practical solutions may not be lightly rejected. If, from the
list of BMPs, a permittee chooses only a few of the least expensive methods, it is likely that MEP
has not been met. On the other hand, if a permittee employs all applicable BMPs except those where
it can show that they are not technically feasible in the locality, or whose cost would exceed any
benefit to be derived, it would have met the standard. MEP requires permittees to choose effective
BMPs, and to reject applicable BMPs only where other effective BMPs will serve the same purpose,
the BMPs would not be technically feasible, or the cost would be prohibitive. Thus while cost is a
factor, the Regional Water Board is not required to perform a cost-benefit analysis."
Funding Sources
The objective of a local stormwater manager in setting up his or her Phase n program is to find a program
that attempts to meet the long-term objective of the Clean Water Act while being affordable - knowing
there is both an ability to consider cost (and funding) in developing the program and a mandate to not let
cost rule the final outcome.
Much has been written about the program side of the equation - focusing first on the worst problems and on
those problems that are important to the local community and then filling in the rest of the six minimum
controls. Lets focus on the funding_ side of the equation.
There are many ways to help resource the NPDES program that cost little - but it will take some
imagination. As local communities look at the potential program needs they have a variety of ways to
resource the program. These ways fall naturally into a hierarchy of ease of resource acquisition or use. A
local community should systematically look to the following resource sources prior to looking to the general
fund and the other usual culprits. In this discussion I will assume that there is currently little or no actual
stormwater quality work being done in the community.
1. Modify local programs The first step in the resourcing analysis is to look at the current local program and
see what is being done that looks and smells like Stormwater Phase n. Based on looking at several
stormwater programs we have found that, perhaps, 25 percent of a typical Phase II program is already being
done to some extent by current staff, or similar things are being done. With suitable adjustment and refocus
some responsibilities can be covered by current staff as part of, or a redefinition of, their current duties. In
some cases it will take little effort to redefine or describe current practices. Table 1 contains a set of
potential areas to look for each of the six minimum controls.
2. Share costs with neighbors or region/state-wide Much of what can be done can be done more cheaply
sharing the cost. After determining what you can already do in-house, or offer to others, the next step is to
see what others can offer to you. Phase I saw large numbers of group permits issued causing regional
approaches to spring up. There are various types of relationships that can be formed for sharing. In one set
of cities each agreed share costs for a minimal program and go independently for a more advanced
program. Costs can be shared for all activities that each community has to do in a similar fashion. This
includes a whole host of things for each of the minimum controls including things like models, joints and
bulks:
1 California State Water Quality Board Order WQ 2000-11, page 19.
332
-------
• "Models" - model brochures, ordinances, bill staffers, checklists, instruction manuals, white papers,
curriculum, etc.
• "Joints" -joint design criteria, videos, billboards, procedure manuals, brochures, web sites,
advertising, etc.
• "Bulks" - Bulk orders for printing, stencils, placards, other PR materials, manual printing, etc.
Table 1. Some Potential Existing Stormwater Program Modification Areas
1. Public Education
• Inserts in other bills
• Speakers bureau
• PAO staff person
• Brochure printing and
distribution capability and
channels
• Public access TV
• Web site
• Watershed signage
• Library
2. Public Involvement
• Citizen advisory group or
panel
• Festivals
• Scout troops
• Internships
• Non-profit groups
• Clubs
• Web site
• Storm drain labeling
programs
• Stream walks
3. Illicit Connections
• GIS coverage
• SARA Title III program
• Pretreatment program
• Land use mapping
• System inventory
• Mayor's complaint hotline
• Water and wastewater
monitoring program
• Camera and smoke testing
capability in water and
wastewater
• Household hazardous waste
collection day
• Recycling programs
• Field personnel
• Used oil programs
• Web site
4. Construction BMPs
• Current ordinance and
development process
• Site inspections
• Other building inspectors
(e.g. electrical, plumbing)
• Mayor's complaint line
• Web site
• Bonding program
• Plan review chicklists
5. Post Construction BMPs
• Current zoning, stormwater
and subdivision ordinances
• Current design criteria
manual
• Open space and related
ordinances
• Current overlay districts
• Master plans
• Floodplain program
6. Municipal Housekeeping
• Street, storm drain and other
maintenance programs
• Current employee training
programs
• Current materials handling
programs
• Current flood control
specifications and in -place
structures
• Recycling program
• Adopt a highway programs
• Neighborhood and non-profit
groups
• Street sweeping program
• Waste disposal program
3. Get free information on the web The Internet has hundreds of sites giving examples of BMPs, manuals,
ordinances, documents, guidance, pamphlets, etc. Literally almost every written document that might be
necessary has been developed somewhere and is available free of charge. The experience of other Phase I
cities is especially helpful for Phase U cities. Fort Worth (http://ci.fort-worth.tx.us/dem/sitemap.htm)
especially has a helpful web site with multiple links to other sites. The Center for Watershed Protection
333
-------
(http://www.cwp.org/) offers a multitude of helpful documents and links and their stormwater center
(http://www.stormwatercenter.net/) has hundreds of references and assistance tools. Other useful sites
include http://www.mtas.utk.edu/bmptoolkit.htm, http://www.dfwstormwater.com, which have links sorted
by each of the six minimum controls. EPA's website (best found from a search as it changes quite often)
offers significant Phase n guidance as well as information on many related programs.
4. Partner with non-profits There are hundreds of non-profit organizations created to accomplish various
environmentally related functions. Often these groups will adopt a watershed, provide workers, perform
monitoring, do public education and involvement campaigns (they are a public involvement campaign), and
find sources of money not available to local governments (501(c)(3) grants to non profits). Some local
communities actually assist them in finding and applying for grants. They also are less willing to file a
lawsuit against a local government when they are partners with it. Areas to investigate beyond the obvious
watershed type grants include Greenspace, parks, quality of life, sustainable development, education, etc.
Sites include: http://www.adopt-a-watershed.org/, http://www.cwn.org, http://www.iwla.org
http://ctic.purdue.edu, http://www.nrdc.org/nrdc/, http://www.tnc.org, http://www.waterkeeper.org,
http://www.rivernetwork.org/ (provides a complete listing of other organizations as well as a funding source
catalog).
5. Federal, regional and state consulting programs Various Federal programs provide consulting either
gratis or cost share.
• For example, TVA supplies Stream Teams to any local community willing to pursue a watershed
protection program (http://www.tva.gOv/river/landandshore/landuse_contacts.htm).
• The National Park Service provides a Rivers, Trails and Conservation Assistance Program that
provides meeting facilitators and planning assistance for river corridor development
(http://www.ncrc.nps.gov/programs/rtca/index.html).
• Several Phase n communities received significant assistance from the Corps of Engineers in
their Phase n permit application and parts of their implementation.
• The USGS cooperative program will provide monitoring and data analysis
(http ://water.usgs.gov/coop/).
• In many cases a regional flood control authority, planning agency, or a state league of counties or
municipalities is more than willing to step in and serve as an integrator programs.
• Pseudo state/university programs often provide consulting free or at greatly reduced rates or can
use other Federal grant monies to provide consulting or product services. For example, in
several states a university, through a 319 grant, developed a statewide BMP manual to serve all
communities in the state. The Ohio Department of Natural Resources "Rainwater and Land
Development Manual" is an excellent BMP source in Ohio.
• Sometimes state programs can serve to partially fulfill one, or more, of the minimum controls.
For example in several states an erosion control or channel protection and permitting program
operated by the state is being relied on for part of the construction minimum control.
6. Federal, State and regional grants States and federal agencies administer or provide grant monies for
local governments to pursue environmental projects:
• State administered programs such as Section 319 (recent congressional action extending the
ability to use 319 money for Phase n for one year, after that some agencies allow "horse
334
-------
trading"), 604(b), 104(b)(3), HUD block grants (http://www.hud.gov/progdesc/cdbgent.cfm),
Coastal Zone (http://www.epa.gov/owow/watershed/wacademy/fund/coastzone.html), Well head
protection, FEMA (http://www.fema.gov/regions/iv/2000/r4_06.shtm), etc. provide funds for
various programs.
• Much of this information can be gleaned from Federal web sites including
http://www.epa.gov/efmpage/fundings.htm (the environmental finance program),
http://www.epa.gov/OWOW/watershed/wacademy/fund.html (watershed Academy funding
site), and EPA regional sites.
• The TEA water quality mitigation retrofit demonstration projects also can be used along with
other TEA-21 mandatory set asides (http://www.fhwa.dot.gov/tea21/).
• Several states have grants set aside for environmental education projects through schools.
• Greenspace programs abound at both the Federal, state and private grant areas and could be
explored as part of a Low Impact Development or Smart Growth approach .
While some of these programs are not, per se, to be used for compliance activities many Phase I cities and
regulators have been cagey about how to bend rules and waive requirements in order to secure funding for
key projects and programs.
7. Special fees for service Another source of funding is to charge special fees for added services including
inspection fees for BMPs, additional construction program related fees, plans review fees, etc. These fees
can be scaled to cover part of or a whole program area. Some communities have instituted a simple
"environmental" surcharge on a water bill as a special assessment. There are really four basic ways local
governments get money: taxes, service charges, exactions and assessments. Each of these basic ways have
rules that vary somewhat state to state, so it is important to know what you are getting into. I recently
visited a city that had 108 different fees and charges based on specific services offered - not sure if that was
a good thing !
8. Private resources Having your corporate name associated with a clean environment is still considered a
good thing. This leads naturally to looking to private resources to fund public environmental projects. This
can take the form of corporate grants, corporate involvement in adopt a stream programs, and other visible
volunteer-based activities:
• Several communities have benefited from industry providing bags, gloves, vests, hats, key chains,
pens, trinkets, coffee cups, new cars... well ok not new cars.
• Others sponsor stream clean ups, partner in restoration projects, construct greenways, etc.
• Another innovative approach is to allow them to put their logos on such things as storm drain
plaques or banners. A firm called adopt-a-storm-drain specializes in this approach... perhaps among
others (http://www.adoptastormdrain.com/).
9. Stormwater Utility The surest and best way to fund stormwater, if you don't have lots of gambling loot
that is, is through a user fee system based on demand on the stormwater infrastructure. If it looks like water
and wastewater it should be funded like those other two public utilities. There is lots of information about
how to set a stormwater utility up, some of it has even been developed by persons have set up a large
number of them. Here are a few good sources: http://www.florida-stormwater.org/manual.html,
http://www.forester.net/sw_001 l_utility.html, http://stormwaterfinance.urbancenter.iupui.edu/. With the
demands of Phase n coming there might just be sufficient planetary alignment to attempt it for even the
335
-------
most reluctant Public Works director. I would not blame EPA for the utility, but it certainly can be the
straw that breaks the camels back, amidst the other pressing stormwater program needs. It IS an unfunded
Federal mandate after all. Again a word of caution. Do it right. Your opportunity cost of failure due to
cutting corners on public education and consensus building is five to seven years of stormwater revenue -
maybe millions. The cost to do it right versus cutting corners is less than two months revenue. Do the
math.
10. Partner with local organizations/agencies_Mmy local/county organizations may be already
implementing programs that fall right in line with the Phase n requirements. For example, educational
school programs, teacher monitoring workshops, watershed festivals, storm drain labeling and stream
walk/community clean-up events, and watershed signage programs are often taken on by county Soil &
Water Conservation Districts (SWCDs). Additionally, construction site plan reviews, inspections, and
enforcement procedures are carried out by SWCD offices. Other organizations such as a Public Works
Departments or Engineers may have the storm sewer systems and detention areas within the county mapped
out. The Health Department may have a map of the septic system locations, thereby making it easier to
determine where illicit discharges may be located.
Defining a Program that Can Be Paid For
Environmental Cost Effectiveness is a term that has evolved over the years principally through the Federal
government's attempt to quantify habitat or ecological benefits of potential projects (COE, 1994).
Traditional benefit-cost analysis is, of course, not possible because costs and benefits are expressed in
different units. Costs are expressed in terms of: dollars, volunteer man hours, level of effort ("hassle
factor"), resources consumed, etc. Benefits are expressed in a wide variety of metrics in stormwater
management including such "measurable goals" as: contact hours, pounds of pollutant removed, stream
miles removed from the 303(d) list, increase in some biotic integrity or bio-assessment measure, bank-miles
restored, "habitat units" restored or protected, delivered information pieces, constructed BMPs, specific
actions taken, etc. Recreational activities such as fishing, boating, biking, etc. can have an associated dollar
value.
Because it is difficult to evaluate cost effectiveness in absolute terms, most cost effectiveness analyses seek
to determine effective programs relative to other potential options. The goal is not to lead to perfect
environmental or economic solutions, but to elevate the decision process above the often emotional cost
oblivious arguments. Steps in a typical cost effectiveness analysis modified to fit a Phase n program might
include (see figure 4):
336
-------
GOALS & OBJ.
PROG MINS
BMP UNIVERSE
COST EFFECTIVENESS
INCREMENTAL COST ANAL.
CONFIGURE PROGRAM
Figure 4. Cost Effectiveness Analysis for Phase II
1. Establish Value. Define the goals and objectives of the overall program focusing on solving
apparent water quality problems or protecting key assets or resources, while keeping in mind the
need to have a program under each of the six minimums. Identify key streams or other water bodies,
ecological systems, habitat areas, and key pollutants of concern. Discuss MS4 values and the
environmental characteristic of the community. Seek to define, in some way, what the community
wants to achieve - besides compliance at minimum cost. Then insure that you have defined a
complete set of goals for all of the minimum controls - even those where you would not normally
chose to focus. Your eventual cost effectiveness consideration will be a bit different for those goals
and objectives that are "essential" and those that are more "fillers" to round out the program.
2. Define the Universe of Possible Solutions. Brainstorm and screen individual and combinations of
BMP programs (both structural and non-structural) including cost or resource estimates, potential
type and availability of funding sources, fit with local program, ability to impact the goals and
objectives, level of expected impact and benefit, mutual exclusivity. Focus first on the "real" goals
and objectives and secondly on meeting each of the six minimum controls. The end product is a set
of feasible BMP or combinations.
3. Perform Basic Cost Effective Analysis. Seek to eliminate inefficient and ineffective (economically
irrational) solutions. Often a certain level of environmental benefit, or program level can be
obtained in several different ways.
• Efficiency is determined by selecting the BMP programs that can produce a given level of
environmental benefit or output at the lowest resource expenditure combination. This analysis
would be most appropriate for this minimum control areas that are not seen as key to the overall
thrust of the local program.
337
-------
• Effectiveness is determined by determining the highest level of environmental benefit or output
at the lowest cost. This analysis would be most appropriate for those areas of the program
identified in step one that are key to the overall surface water health of the community - the
"compelling case".
For example there are several potentially viable options for stream clean up: (1) hiring students
during the summer, (2) using non-profit watershed groups, (3) hiring full-time staff, (4) working
through scouting agencies, (5) working through neighborhood groups, (6) using local businesses
in a way similar to adopt-a-highway. Student hires for stream trash removal may be more cost
effective than full-time staff. However, with a higher initial cost and effort, it might be possible
to set up self funded and largely self managed "adopt-a-stream" groups as 501(c)(3) non profit
groups who will be self sustaining, increase public involvement and education, and provide other
ancillary benefits. This option may then be seen as the most cost effective of the options when
considering the long term program and the character of the community.
Perform Incremental Cost Analysis The Attempt is to optimize cost effective solutions. The goal is
to answer the question: "is the increment in environmental benefit worth the increment in cost?" For
each cost effective BMP a range of effort and cost may be defined and, if possible a range of
environmental outputs in response to that effort input range. That is, if we increase the level of
effort for a particular BMP program will the range of environmental benefit also increase - and
how?
For example, there will be diminishing returns in public education programs as saturation is reached.
Each incremental brochure, billboard, or other means will not yield as high a return - though
sometimes only intuition and experience will often define those points, or that curve.
Or using the example from step three, it might be found that student summer hires are the most cost
effective way to achieve stream clean up. This step then looks at this options and seeks to find ways
to maximize the effectiveness of that particular solution. It might be that providing a certain level of
resources, finding private grant money, forming a student organization, etc. will provide maximized
returns for this option.
Configure the Program. Blend the various BMPs into a cohesive program, seeking synergy and
practicality. Insure the program is at a level that is both acceptable to the permit writer and doable
within the legal, social, financial, political, technical and physical constraints within the community.
Lay out a program and funding strategy, leaving "outs" if anticipated funding sources do not emerge.
Develop processes to manage the program and attain measurable goals.
338
-------
References
Reese, A.J., E. Treadway and D. Noel, 2000. Estimating Costs for the Phase n Stormwater Management
Program, Water Environment and Technology, Water Environment Federation, April 2000, pp. 33-39.
US Army Corps of Engineers, 1994. Cost Effectiveness Analysis for Environmental Planning: Nine Easy
Steps, IWRRpt. 94-PS-2, Inst. For Water Resources, 7701 Telegraph Rd., Alexandria, VA, 22315.
U.S. Environmental Protection Agency, Office of Water. 1994. President Clinton's Clean Water Initiative.
Washington, D.C. EPA 800-R- 94-001.
339
-------
PROTECTING WATER RESOURCES WITH HIGHER DENSITY DEVELOPMENTS
Lynn Richards, Geoffrey Anderson, Mary Kay Santore
US Environmental Protection Agency
Washington, DC
ABSTRACT
How and where development occurs can affect water quality. The purpose of this paper is to examine the
water quality impacts from low and high-density development at the site level and watershed level.
Considerable evidence in the literature demonstrates that dispersed, low-density development can
exacerbate non-point source pollutant loadings by consuming absorbent open space and increasing
impervious surface area relative to compact development. Some case studies have demonstrated that higher
density development can minimize impacts on regional water quality by consuming less land and
minimizing impervious surface cover. This paper discusses the relationship between water quality and
growth patterns; uses modeling results to compare pollutant loadings from different types of residential
development; and discusses measures to mitigate potential increased pollutant concentrations, which may
result from higher density development
INTRODUCTION
In the face of droughts, oil spills, beach closures, and overall declining water quality, communities are
increasingly concerned about managing their watersheds to maintain hydrologic integrity and water quality.
The nation's aquatic resources are among its most valuable assets. Although environmental protection
programs in the United States have improved water quality during the past 25 years by focusing on point
sources, many challenges remain. EPA estimates that of the causes of pollution in the states' impaired
waters, only 10 percent is presently attributable to point source pollution, such as industrial discharges. The
rest is ascribed to non-point source pollution or some combination of point and non-point source pollution,
which can include increased sedimentation from land development, stormwater runoff, and on-site sewage
systems.
The National Water Quality Inventory: 1998 Report to Congress identified urban runoff as one of the
leading sources of water quality impairment in surface waters.1 Of the 11 pollution source categories listed
in the report, emissions from urban runoff and storm sewers was ranked as the sixth leading source of
impairment in rivers, fourth in lakes, and second in estuaries. In addition, recent water quality data find that
more than a third of assessed rivers and streams (291,000 of 840,000 miles) do not meet water quality
standards. For these impaired surface waters, urban and agricultural runoff are the primary sources of
pollution.2
Of special concern are the problems associated with non-point source storm water runoff in our urban
streams, lakes, estuaries, aquifers, and other water bodies caused by runoff that is inadequately controlled or
treated. These problems include changes in flow, increased rates of sedimentation, higher water
temperature, lower dissolved oxygen, degradation of aquatic habitat structure, loss offish and other aquatic
1 U.S. Environmental Protection Agency (USEPA). 2000a. National Water Quality Inventory: 1998 Report to Congress.
www.epa.gov/305b/98report. Last updated October 5, 2000.
2 U. S. Environmental Protection Agency, Office of Water. June 2000b. "Water Quality Conditions in the United States: A
Profile from the 1998 National Water Quality Inventory Report to Congress" Washington, DC.
EPA841-F-00-006 (also available atwww.epa.gov/OWOW/305b/).
340
-------
populations, and decreased water quality due to increased levels of nutrients, metals, hydrocarbons, bacteria,
and other constituents.
Recent research has revealed a strong relationship between impervious cover and water quality. These
studies have demonstrated that at 10 percent imperviousness, a watershed will become impaired.3 In
addition, water quality suffers not only from the increase in impervious surface, but also from the associated
activities: construction, increased travel to and from the development, extension of infrastructure, and
chemical maintenance of the areas in and surrounding the development. Oil from motor vehicles, lawn
fertilizers, and other common solvents, combined with the increased flow of runoff, contribute substantially
to water pollution. These findings suggest that as imperviousness increases, so do associated activities,
thereby delivering an increased impact on water quality. In an effort to protect water resources,
communities may apply the 10 percent impervious cover threshold from the watershed level to the site level.
The purpose of this downscaling is to reduce development densities and therefore reduce overall impervious
surfaces at the site level. While intended to address overall impervious within the watershed, when the 10
percent figure is applied to the individual site level within the watershed, it suggests that only lower
densities can protect water quality.
This study suggests that the opposite may in fact be true- attempts to ensure low densities at the site level
can often lead, not to better, but to worse overall water quality. Other recent studies have demonstrated that
dispersed, low-density development can exacerbate non-point source pollutant loadings through increased
consumption of pervious open space and greater amounts of transportation-related impervious
infrastructure, such as roads, driveways, and parking lots. On the other hand, a compact development
approach accommodates more activity while consuming less space. In turn, this reduces overall
imperviousness and helps to maintain watershed functions.
The purpose of this paper is to examine the water quality impacts from low and high-density development at
the site level and then to extrapolate these findings to the watershed level. This paper discusses the
relationship between water quality and growth patterns; uses modeling results to compare pollutant loadings
as a function of residential density; and summarizes existing research on the subject. We conclude that
accommodating new growth in a compact, higher density fashion (in undeveloped areas or developed areas)
will likely be more protective of water quality than lower density development.
DEVELOPMENT'S IMPACT ON WATERSHED FUNCTIONS
One of the most noticeable trends in recent history has been the dramatic expansion in the geographic size
of metropolitan areas. Virtually every urban area in the United States has expanded substantially in land
area in recent decades. Between 1954 and 1997, urban land area has almost quadrupled, from 18.6 million
acres to about 74 million acres in the contiguous 48 states.4 Moreover, from 1992-1997, the national rate of
development more than doubled. During this five-year period, more land was developed (nearly 16 million
acres) than during 1982-1992 (about 13 million acres).5 The newly developed land has typically come from
forest land, pasture and range land, and crop land. A 1994 study by the American Farmland Trust showed
3 See, for example, Montgomery County Department of Environmental Protection, 2000; Center for Watershed Protection, 1998;
Schueler, 1994; Arnold and Gibbons, 1996.
U.S. Department of Agricultural, Economic Research Service, Natural Resources and Environmental Division. Agricultural
Resources and Environmental Indicators (AREI) Updates, No. 3. "Major Land Use Changes in the Contiguous 48 States." June
1997.
5 Ibid.
341
-------
that urban development already has consumed nearly a third of the country's most highly productive
farming regions.6
Direct environmental impacts of current development patterns include habitat loss and fragmentation and
degradation of water resources. Building on undeveloped land consumes and fragments habitat and thus
displaces or eliminates wildlife communities. The construction of impervious surfaces such as roads and
rooftops leads to the degradation of water quality by increasing runoff volume, altering regular stream flow
and watershed hydrology, reducing groundwater recharge, and increase stream sedimentation.
Watersheds and their streams and rivers provide critical ecological and economic services. Ecologically,
small watersheds and streams sustain larger ecosystems. In addition, the stream corridor, with its rich flood
plains, wetlands, and forests, is home to unique plant and animal species. Streams support diverse aquatic
communities and perform the vital ecological roles of processing the carbon, sediments, and nutrients upon
which downstream ecosystems depend. Economically, small watersheds are the ultimate source of our
drinking water; watershed and riparian buffer zone soils act as filters for water that might ultimately be
consumed. Slow-order streams and their associated flood plains serve as temporary storage for floodwaters,
and thereby act as natural flood control. The services provided by small watersheds are maximized when
their land area is maintained in a natural condition.
The extent of beneficial watershed services begins to diminish when the natural condition of land is altered
through development. Construction exposes sediments and construction materials to precipitation, which
then washes material into storm drains or directly into nearby bodies of water. After construction,
development usually replaces native meadows, forested areas, and other natural landscape features with
compacted and fertilized lawns, pavement, and rooftops. These largely impervious surfaces generate
substantial quantities of surface runoff. In addition, engineers traditionally design drainage systems to move
rainwater as quickly as possible by directly it over the ground towards curbs, gutters, streets, and sewers.
These conventional drainage systems prevent water from flowing into the ground and filtering through soil
before being released into surface and ground waters. To compound problems, traditional construction
practices seek to "connect" all of the impervious surfaces in a development to direct water to a minimal
number of drainage outlets. For a typical retail protect, the storm water system connects water from all
rooftops, several parking lots and the interior road network. Even when landscaped islands are built into the
project, the grading typically directs water away from the landscaping, thus losing any opportunity to
"disconnect" the imperviousness for infiltration. This connected system instead creates more surface
runoff—and this results in increased flooding, erosion, and pollution. Consequently, an urban watershed
produces a greater volume of stormwater runoff, which in turn degrades the physical, chemical, and
biological quality of streams.7
Some communities are taking steps to preserve undeveloped parcels or regional swaths of open space, in
order to preserve watershed functions, among other environmental, economic, and social goals. Preserving
open space can reduce total watershed impervious surfaces. Indeed, since 1998, nearly $20 billion has been
approved for open space preservation in local and state referenda. Since all land has differing ecological
value, some communities are beginning to develop open space conservation programs that target the most
American Farmland Trust. 1994. Farming on the Edge: A New Look at the Importance and Vulnerability of Agricultural Near
American Cities.
1 Woodworth, James et al. 2002. Out of the Gutter: Reducing Polluted Runoff in the District of Columbia. NRDC: Washington,
DC.
342
-------
critical areas for preservation.8 However, strategic and targeted open space preservation planning is in its
nascent stages and the overall impact of these measures tends to be somewhat limited from an ecological
protection standpoint. While open space preservation is certainly part of the solution for development-
related water quality problems, it is critical to address overall densities in the watershed in order to minimize
total land consumption.
Low DENSITY DEVELOPMENT-BAD FOR WATER QUALITY? CRITIQUING CONVENTIONAL WISDOM
Knowing that development has the ability to impair the natural functions performed by watersheds, state and
local governments are asking, "If we are going to grow, how do you minimize development's impacts on
water quality? Are some patterns of development less harmful than other development patterns? Are there
critical thresholds of which to be aware? How much development can a watershed absorb without
significant harm occurring?"
There are some answers to these questions. Studies have demonstrated that watershed's suffer impairment
at a 10 percent impervious cover. Over 25 percent, the watershed is considered severely impaired.9
Conventional thinking has translated these findings into the notion that low-density development will result
in better water quality. The reasoning behind these policies is: a 1-acre site will typically have one or two
residential units with a roadway passing by the property, the driveway, a home with an average footprint of
2,265 ft2. 10 The remainder of the site is lawn. The impervious cover is approximately 35 percent.11 The
lawn, however, while still pervious cover, contributes to stormwater runoff because of its disturbed nature,
e.g., the soils have been compacted due to scraping and the traversing of construction equipment. The effects
of this compaction can remain for years, and be increased due to mowing. Therefore, sites with fewer houses
minimize impervious cover and maximize lawn cover or other types of variably pervious surface. Given
indications that watershed impairment begins at 10 percent impervious cover, it is thought that a low-density
development scenario may be one approach to the improvement of water quality. However, in a higher-
density scenario, which will typically have eight to ten residential units per acre, the parcel is likely to be built
out with upwards of 85 percent impervious cover.12 The majority of this impervious cover is due to the
footprints of the housing units. Lawn space is generally minimized. This scenario seems less protective of
water quality because it has more impervious cover due to housing footprints.
Because impervious surface area appears to vary with specific land use, a common approach to local land
use regulation in support of water quality is to specify maximum development densities. The reasoning here
is that if each site minimizes water quality impact through density alone, e.g, the number of residential
unites per acre, then overall parcel-level impervious cover is regulated, with the putative benefits apparent at
the watershed or regional scale.13 While this seems to make sense, there are some significant flaws in this
thinking.
Trust for Public Land and the National Association of Counties. 2002. Volume 1: Local Greenpr Ming for Growth: Using Land
Conservation to Guide Growth and Preserve the Character of Our Communities.
9 There are different levels of impairment. In general, when the term is used in EPA publications, it usually means that a water
body is not meeting its designated water quality standard. However, the term can also imply a decline or absence of biological
integrity, e.g., the water body can no longer sustain critical indicator species, such as trout or salmon. Further, there is a wide
breadth of levels of impairment, e.g., endangered trout versus spontaneous combustion.
10 National Association of Home Builders. 2001. Housing Facts, Figures, and Trends: 2001. NAHB: Washington, DC. The
average house built in 2001 includes 3 or more bedrooms, 2.5 baths, and a 2-car garage.
11 Soil Conservation Service, 1986. Technical Release No. 55 (TR-55). Urban Hydrology for Small Watersheds.
12 Ibid.
13 See, for example, the code for Durham, NC: www.ci.durham.nc.us/departments/planning/zoneord/Section5/556.html
343
-------
1) Density and imperviousness are not equivalent. Depending on the actual design of the development,
two houses may actually create as much imperviousness as four houses, for example. The impervious
area on site associated with given number of residential dwelling units can vary widely due to road
infrastructure, housing design (single story or multi-story), or length and width of driveways. For
example, a multi-story apartment of 10 units on one acre can have less impervious surface than 6 single-
family homes on the same acre. Even at the level of a single house, impervious area can vary widely, and
therefore assumptions about the impervious area per dwelling unit are questionable. For example, in some
dispersed low-density communities, such as Fairfax County, Virginia, some homeowners are paving their
front lawns to create more parking space for the large number of cars each household owns.14 This
phenomenon has also been noted in some San Francisco, California neighborhoods with large households
and high vehicle ownership rates.15
2) Much of the "pervious " surface left on low-density development acts like impervious surface for water
quality purposes. All else being equal, undisturbed land is better for water quality than disturbed land,
including lawns and other maintained areas. However, disturbed and impervious areas vary widely in
the amount, speed, and type of runoff per square foot. At one time, lawns were thought to provide
"open space" for infiltration of water. However, development can involve wholesale grading of the site,
removal of topsoil, severe erosion during construction, compaction by heavy equipment and filling of
depressions. Research now shows that the run-off from highly compacted urban lawns is almost as high
as paved surfaces.16 Therefore, a one or two acre lawn does not offer the same watershed services that a
one or two acre undisturbed forest does. The idea that minimizing impervious surfaces by limiting
housing structures and maximizing larger lawns does not address the loss of ecological services that the
area provided before development.
3) Low-density developments mean more off-site impervious infrastructure. Development in the watershed
is not simply the sum of the sites within it. Rather, total impervious area in a watershed is the sum of site
developments plus all the infrastructure supporting those sites, such as roads, parking lots, ditches, and
other impervious surface infrastructure. Furthermore, recent research has demonstrated that impervious
surfaces attributed to streets, driveways, and parking lots can represent upwards of 75 percent of total site
imperviousness, and this is on sites with two residential units per acre.17 That number decreases to 56
percent on sites with 8 residential units per acre. This indicates that as density decreases, off-site
transportation-related impervious infrastructure often increases. In a density-limiting policy environment,
densities are generally calculated absent this infrastructure, and low-density development requires
substantially higher amounts of this infrastructure per capita and per acre than do the more dense
developments, which are paradoxically prohibited by some types of zoning regulation.
4) The scale of the finding that 10 percent impervious cover impairs watersheds is for the watershed level.
Often, this finding is applied at the site level, and, as discussed in the previous point, does not take into
account the transportation-associated infrastructure. Applying this finding at the site level is flawed
since the research behind this finding was conducted at the watershed level, not the site level.
Extrapolating from the site to the watershed would be incorrect because other factors come into play at
14 Rein, Lisa and David Cho, "In Defense of the Front Lawn: Fairfax Attacks Crowding With Ban on Oversize Driveways,"
Washington Post, June 4, 2002, p. Al.
15 Brown, Patricia Leigh, "The Chroming of the Front Yard," New York Times, June 13, 2002, p Fl.
16 Schueler, T. 2000. The Compaction of Urban Soil. Techniques for Watershed Protection. Center for Watershed Protection,
Ellicott City, MD.
17 Cappiella, K. and Brown, K. 2001. Impervious Cover and Land Use in the Chesapeake Bay Watershed. Ellicott City. MD.
344
-------
the watershed level. However, what the 10 percent finding does suggest is that it is better to cluster
development or to increase the density of existing communities.
5) Growth is coming to the region, limiting density on a given site doesn 't eliminate that growth. Density
limits are responses to—and attempts to manage—growth. Yet they do not in fact manage growth; they
only manage some growth—the growth on the density-limited area. The rest of the growth that was
going to come to the region still comes, but goes elsewhere. Is that elsewhere better or worse for
regional water quality than accommodating the growth at the density-limited site? Rarely if ever are
density limits part of a watershed plan that answers that question. If growth is coming to a region, it will
come regardless of density limits in a particular place. There is a lively debate in economic
development circles about whether certain types of development are especially attractive to residents
and/or businesses, and will therefore draw additional growth. But no one argues that pursuing a
particular kind of growth will slow or stop growth in a region.18 (This issue is discussed in more detail
in on page 11). At most, covering a large part of a region with density limits will drive growth to other
parts of the region. If the excluded growth's destination is upstream from the density-limited area, then
the area with the density limits will still be affected by the growth, and, depending on local conditions,
may actually be made worse off from a water quality perspective than if the growth had been
accommodated and well-managed in the area.
TESTING THE ALTERNATIVE: CAN COMPACT DEVELOPMENT IMPACT REGIONAL WATER QUALITY?
The debate over how best to protect water quality, and how to continue to enjoy the ecological and
economic services of watersheds, begins with the expanding United States population. The Census Bureau
projects that U.S. population will grow by 50 million people between 2000 and 2020.19 Where and how
these people will be accommodated is fundamental to all water quality protection strategies.
What is the alternative to the density-limiting approach? Compact development can accommodate more
people on less land, leaving more undisturbed land, i.e., greenfields, available to serve critical ecological
functions as previously described.20 The fundamental debate, then, is over which scenario is better for
regional, or watershed, water quality—lower density or higher density ("compact") development. The two
arguments can be summarized as follows:
1. Low-density development is better for watershed water quality because it limits impervious cover at the
site level.
Or
18 There are, of course, minor exceptions to this dynamic. An area that is desirable will probably experience an increase in
housing prices and would consequently experience a very modest displacement of development to other parts of the region. For
example, housing prices in some neighborhoods in Manhattan, New York, San Francisco, California, or Washington, DC have
increased significantly because of the urban form and high densities. It is likely that the higher housing prices have fostered
development in areas further from these central locations.
19 "Annual Projections of the Total Resident Population as of July 1: Middle, Lowest, Highest, and Zero International Migration
Series, 1999 to 2100." Population Projections Program, Population Division, U.S. Census Bureau, Washington, D.C. 20233.
Internet Release January 13, 2000, revised February 14, 2000 at www.census.gov/population/www/proiections/natsum-T 1 .html.
20 In addition, higher densities make public transit profitable, increase walkability, and generally increase other livability factors
that are absent in dispersed, low-density sites. For more information on these positive externalities associated with compact
development, see EPA document 231 -R-01 -002 "Built and Natural Environment: A Technical Review of the Interactions between
Land Use, Transportation, and Environmental Quality."
345
-------
2. High-density development is better for watershed water quality because overall it disturbs less land to
accommodate the similar numbers of people and therefore leaves more land available to serve critical
ecological functions.
Although the previous section gave numerous reasons to doubt that the density-limiting approach was
protective of watershed quality, a complete evaluation needs to test the density-limiting approach against
one that encourages compact development. We test the competing approaches by comparing higher- and
lower-density developments by using hypothetical site plans that represent typical low-density and compact
development patterns.21
Assumptions
In order to construct scenarios and conduct the modeling in a way that produces policy-relevant results, certain
assumptions drive the analysis. Because the relevant question concerns selection of an approach that produces
less runoff and pollutant loadings, the analysis examines the comparative differences in the impacts of low
density and compact development patterns. The analysis is driven by two major assumptions:
1. Metropolitan regions will continue to grow. This assumption is consistent with US Census projections
that the US population will grow by roughly 50 million people by 2020.22 Given this projected
population growth, communities across the country are or will be grappling with how to accommodate
expected population increases to their regions.
2. Shifting growth represents a shift in growth, not additional growth within the region. Individual states
and regions grow at different rates depending on a variety of factors including macroeconomic trends
(e.g., the technology boom in the 1980s spurring development in the Silicon Valley region in
California); historical growth rates; and demographic shifts. These factors are not significantly impacted
by the prevailing distribution of density of development. The question for a state or a region is, "If we
are going to receive X number of new jobs and X number of new residents, what is the effect of
accommodating those jobs and residents in a higher density pattern of development versus a low density
pattern of development?"
AN ILLUSTRATIVE EXAMPLE
To determine which development pattern is more protective of water quality, we have developed two
scenarios in order to examine water quality impacts from a high density and lower density developments.
These scenarios take place within a fictional watershed and are simplified in order to isolate and examine
the impacts of density on water quality. Issues such as slope, ground water hydrology, commercial,
industrial, and agricultural land uses are important to watershed health, but are not considered in these
scenarios.
Two communities in this watershed are each growing by the same amount. The region's council of
governments has forecasted that over the next 20 years, the metro area will grow by 270,000 persons. As
the region looks to accommodate this new growth, they are also looking for ways to protect water quality
and the overall health of the watershed.
21 For more information and other tests, please see EPA's draft document, Minimizing the Impacts of Development on Water
Quality, 2003.
22 U.S. Census Bureau.
346
-------
Two communities in this region have different average densities. Community A is dominated by lower
density development, and has an average residential density of three residential units per acre.23'24
Community B, the higher density area, has a density average of approximately nine residential units per
acre. Each residential unit in both communities generates a certain volume of stormwater runoff and a
proportional amount of pollution. For both communities, we assumed that development would have the
following features:
• The entire acre is disturbed land; e.g., no forest or meadow cover would be preserved.
• Each residential unit in both communities has a footprint of 2,200 square feet.
• The same percentage of transportation-associated infrastructure, such as roads, parking lots,
driveways, and sidewalks is allocated to each community acre.
• No best management practices, structural or otherwise, are implemented.
In general, impervious surfaces, such as housing footprint, driveways, and roads will have higher amounts
of runoff and associated pollutants. Lawns, while pervious, still contribute to runoff due to their compacted
and disturbed nature. Based on these assumptions, the overall percent imperviousness for Community A is
approximately 30 percent for an average density of 3 residential units per acre and the overall percent
imperviousness for Community B is 70 percent for an average density of 9 residential units per acre.25
While these assumptions are based on an illustrative example and not on actual site plans, the size of
housing units is based on national trends from the National Association of Home Builders.26
The percentage of infrastructure that is attributable to each acre is based on the curve number methodology
from the Natural Resources Conservation Service (NRCS); and the overall site imperviousness is based on
NRCS studies of urban hydrology.27
The model used to generate the results described below is Smart Growth Water Assessment Tool for
Estimating Runoff (SG WATER)28—a peer reviewed sketch model that was developed specifically to
compare water quantity and quality differences among different development patterns. SG WATER'S
methodology is based on the NRCS curve numbers,29 event mean concentrations, and daily rainfall data.30
23 Densities at three or nine residential units per acre are conservative and used here for illustrative purposes only. Many
communities now are zoning for one unit per two acres at the low-density end of the spectrum. Low density residential zoning
exists in places as diverse as Franklin County, OH that require no less than 2 acres per unit
http://www.co.franklin.oh.us/development/franklinco/LDR.html#304.041) to Cobb County, Georgia outside of fast growing
Atlanta that requires between 1 and 2 units per acre in its low density residential districts
(http://www.cobbcountv.org/communitv/planbzacommission, htm). By comparison, some communities are beginning to allow
higher densities upward to 20 or high units per acre. For example, Sonoma County, California's high density residential district
permits between twelve (12) and twenty (20) units per acre (http://www.sonoma-countv.org/prmd/Zoning/article 24.htm) and the
City of Raleigh, NC allows up to 40 units per acre in planned development districts.
(http://www.raleigh-nc.org/planning/DPRC/BROCHURES%20PDF/HIGH DENSITY.PDF)
24 For this example and throughout this paper, residential units instead of commercial units are compared. Most communities do
not zone for density limits for commercial and retail properties.
25 Soil Conservation Service, 1986.
26 National Association of Home Builders. 2001.
27 The NRSC estimate for average imperviousness for 8 units per acre is 65 percent. They do not have an estimate for 9 units per
acre. Given our calculations and NRSC estimates of average site imperviousness, we are extrapolating average impervious for 9
unit per acre to be 70 percent.
28 Technical Approach for SG WATER: Smart Growth Water Assessment Tool for Estimating Runoff, 2002.
29 Soil Conservation Service. Technical Release No. 55 (TR-55). Urban Hydrology for Small Watersheds.
30 Daily time-step rainfall data for the three year period (1997-1999) was used.
347
-------
It does not take into account wastewater or drinking infrastructure, slope, or other hydrological interactions
that the more complex water modeling tools use.
Please note that SG WATER uses a general and simple methodology based on curve numbers. One
limitation of curve numbers is that they tend to under predict stormwater runoff for smaller storms. This
under prediction can be significant since the majority of storms any given area experiences in any year are
small storms. In addition, the curve numbers tend to over-estimate runoff for large storms. However, curve
numbers will more accurately predict runoff in areas with more impervious cover because the runoff for
impervious cover is similar using the curve number approach and the small storm hydrology approach.31
For the analysis here, the runoff from the low-density site will be under predicted to a larger degree than the
runoff from the higher density site because the higher density site has more impervious cover. Simply put,
the difference in the numbers presented here are conservative—it is likely that the comparative difference in
runoff between the two sites will be much greater if more extensive modeling was used.
RESULTS
In the lower-density Community A, the total average annual volume of runoff from the one-acre site, with
three housing units, is 21,400 ft3- and the total average annual volume of runoff from Community B, with 9
housing units is 42,900 ft3. These totals represent the amount of water measured at one hypothetical outfall.
Community B, with more housing units, has a greater amount of impervious surface cover and thus
generates a larger volume of runoff at the site level.
Exhibit 1: Total Average Annual Stormwater Runoff Per Acre for Both Communities. (These totals represent
the amount of water measured at a hypothetical outfall.)
Community A
Community B
Density
3 residential units per acre
9 residential units per acre
Imperviousness
30 percent
70 percent
Average Annual
Runoff32 per acre
21,400ft3
42,900 ft3
Now, looking at how much runoff each individual housing unit produces, we see that in Community A,
each house yields 7,133 ft3 of average annual runoff, whereas in the more dense Community B, each unit
produces 4,767 ft3 average annual runoff. Therefore, when examined at the housing unit-level, each house
in Community B produces approximately 33 percent less runoff for each house in Community A. This is
because houses in Community B have smaller yards and less site-infrastructure on a per unit basis.
Therefore, on a per unit basis, each home in the higher-density communities contributes less stormwater
runoff. Exhibit 2 demonstrates.
Most existing stormwater models incorrectly predict flows associated with small rains in urban areas. Most existing urban
runoff models originated from drainage and flooding evaluation procedures that emphasized very large rains (several inches in
depth). These large storms only contribute very small portions of the annual average discharges. Moderate storms, occurring
several times a year, are responsible for the majority of the pollutant discharges. The effects caused by these frequent discharges
are mostly chronic in nature, such as contaminated sediment and frequent high flow rates, and the interevent periods are not long
enough to allow the receiving water conditions to recover.
32 Calculated by SG WATER using Atlanta, Georgia daily time step rainfall data and assuming hydrologic soil type C.
348
-------
Exhibit 2: Total Average Annual Stormwater Runoff Per Housing Unit for Both Communities. (These totals
represent the amount of water measured at a hypothetical outfall.)
Community A
Community B
Density
3 residential units per acre
9 residential units per acre
Imperviousness
30 percent
70 percent
Average Annual
Runoff per Acre
21,400ft3
42,900 ft3
Average Annual
Runoff per Unit
7,133ft3
4,767 ft3
In sum, our model showed that when density is tripled, total stormwater runoff doubles at the per acre level,
but is decreased by one-third at the housing unit level. In other words:
• density triples; and
• imperviousness doubles; and
• total average annual doubles; and
• runoff per housing unit falls by 33 percent.
Exhibit 3 illustrates the relative differences between Community A and Community B. At the one-acre
level, the lower total average annual runoff produced by Community A's low-density development would
be better for water quality than the Community B's high-density development. On the other hand, at the
individual housing unit level, the high-density development of Community B produces less stormwater
runoff on a per-dwelling-unit basis.
Exhibit 3: Average Annual Stormwater Runoff in Community A and Community B. (These totals represent the
amount of water measured at a hypothetical outfall.)
Community A: One acre, three houses
21,400 ft3 of runoff
(better) per acre or
7,133 ft3 of runoff
per unit
Community B: One acre, nine houses
42,900 ft3 of runoff
per acre or
4,767 ft3 of runoff
per unit (better)
On a strict site-level basis, the density limiting approach is more environmentally protective. Recalling
from the previous section the conclusion that the watershed is the correct level of analysis, rather than the
site-level, we turn next to examining the implications for the watershed, by extrapolating these site-level
results.
The assumptions establish that Communities A and B will grow at the same rate. Thus our initial model run,
placing only three units in Community A, did not test the situation actually faced by Community A.
Community A will also needs to accommodate the same nine dwelling units, so the correct scenarios must
compare nine new dwelling units in Community A to nine new dwelling units in Community B.
Where is Community A put the six additional houses that Community B accommodated? Assuming the
same development densities, Community A will need to develop two additional acres, or three acres total, to
accommodate the same number of housing units that Community B accommodated on one acre. In this
scenario, total average annual runoff from nine houses in Community A is 64,200 ft3 (21,400 ft3 x 3 acres),
349
-------
which is 50 percent more runoff as the same nine houses produce in Community B (the same 42,900 ft3 total
average annual runoff). Exhibit 4 illustrates.
Exhibit 4. Each community accommodates nine houses. (Average annual runoff—assuming one hypothetical outfall.)
Community A: Nine houses, three acres
'64,200 ft3 of runoff
total or
7,133 ft3 per unit
Community B: Nine houses, one acre
42,900 ft3 of runoff
total OR
4,767 ft3 per unit
Better by either
measure
From this example, we can see that with higher densities, the per unit runoff rates are dramatically less
(approximately 33 percent) than their low-density counterparts. If we only look at runoff from the 1-acre
site level (not looking at the per unit rates or the rates for accommodating the same number of houses given
permitted densities) we see that lower densities can create less impervious cover and produce less runoff.
But if we treat the watershed as a whole—expecting that the region will be accommodating a given amount
of new growth, regardless of whether that growth is low or high density— the lower density developments
will necessarily require developing further into the watershed. In turn, each low-density unit, requiring
more space for driveways, roadways, and compacted lawns, will create more runoff and watershed
degradation. If these impacts are extrapolated to the watershed level, Community A will develop land at a
rate three times faster than Community B. Exhibit 5 is intended to illustrate the potential regional build out
of these two different community scenarios. These illustrations33 give us a pictorial view of how
Community A and B might end up developing at a watershed scale. Clearly development in Community B
disturbs less land, thereby preserving more critical ecological functions than the low-density development
patterns in Community A. Yet, both communities are accommodating the same number of people.
; Provided by the New Jersey Office of Planning; http://www.state.nj.us/osp/plan2/p2full/colorsOO.htm.
350
-------
Exhibit 5: Comparison of Water shed Build Out for Communities A andB
Community A Community B
FINDINGS FROM THE ILLUSTRATIVE EXAMPLE
Using average densities to project stormwater runoff for two communities, we were able to demonstrate that
a higher density scenario generates less stormwater runoff on a per housing unit basis. Specifically, this
example illustrates:
• For a given site, less compact development can create less impervious cover, less runoff, and may
better protect water quality;
• With more compact development, runoff rates per residential unit fall dramatically, to approximately
1/3 of their less compact counterparts;
• For the same amount of development, the more compact development will produce less runoff than
the less compact development pattern; and
• For a given amount of growth, then lower density developments must force development further into
the watershed.
Taken together, these findings lead to the conclusion that, all else being equal, including amount of growth,
at the watershed level, higher densities are more environmentally protective. These results were also tested
for comparative development sites at the square mile area and 10-acre area in addition to the one-acre
analysis. At all levels, the ratios remain the same: when density is tripled, total stormwater runoff doubles
at the per acre level, yet the housing level stormwater runoff is decreased by one-third.
351
-------
ISSUES TO CONSIDER IN THIS ANALYSIS
1. Is growth really fixed?
A basic assumption for our modeling is that the amount of growth coming to either Community A or B is
fixed—and the question to be examined is how can certain strategies influence the density and pattern of
that growth. When developing and examining the consequences of regional growth trends, regional
forecasters ask, "how much growth is expected to come to this region in a given period of time?" In
standard regional population modeling practice, wage or amenity (a firm-location criterion based on
pleasant locational attributes—such as climate or culture—rather than on transport or production cost34)
differentials with other areas of the country seem to account for most of the ingress or egress to a
metropolitan area.35 Growth is also a function of birth and death rates in a region. Regional growth models
do not typically employ density drivers of regional jobs or population. That is, growth is apparently not a
function of regional development patterns. Development density is independent of regional growth, there is
no reason to believe that low-density zoning limits the number of people moving to a region, and many
reasons to believe that such zoning does not limit the number of people moving to a region, but rather
simply pushes them further out.
Estimates of future growth are rarely precise and despite this imprecision, regions have used this fixed
amount of growth to test the effects of adopting different growth planning strategies. This is possible if we
accept the premise that development patterns do not significantly change the amount of regional growth. A
wide variety of regions have used this approach. One of the best-known studies and planning processes is
Portland, Oregon's "Vision 2040." Portland understood that the region would grow substantially by 2040;
the question was not if, but where and how. In response, it developed a base case and three alternative
growth concepts that all absorbed the same amount of growth; approximately 720,000 additional residents
and 350,000 additional jobs in the region.36 These four alternative futures are schematically illustrated in
Exhibit 6. Although they all absorb the same amount of people and jobs, they vary substantially in
infrastructure requirements, open space preservation, and impact on both the urban and natural environment.
Each option was analyzed for effects on:
• land consumption
• travel times and distances
• open spaces and air quality
• various urban landscapes."
34 Mills, Edwin, B. Hamilton. 1994. Urban Economics: Fifth Edition. Harper Collins College Publishers.
35 The most widely-used such model—the REMI® Policy Insight™ model—uses an amenity variable. However, even this is
implemented as an additional change in the wage rate. See www.remi.com/Overview/Evaluation/Structure/structure.html. All
other regional population models in a survey by ICF use only economic and demographic drivers. The in-house model used by
San Diego Association of Governments is an advanced example of the type used by COGs around the country.
www.sandag.cog. ca.us/resources/demographics_and_other_data/demographics/forecasts/index.asp.
36 http://www.metro-region.org/growth/tfplan/2040.html
37 Metro, "The Nature of 2040: The region's 50-year plan for managing growth," 2000. See http://www.metro-
re gion.org/growth/tf/2040history.pdf
352
-------
Exhibit 6: Same amount of growth, different locations and densities: Portland's Vision 2040 alternatives
analysis.
Base case - Continuing pattern
Concept A
Groxving out
Concept B
Cm wing up
Concept C
PJoiqhboring cities
Greatest expansion of UGB; continuation of
development patterns occurring between 1985
and 1990.
354.000 acres in UGB
(121,000 acres added to UGB)
ihTiilicjnl BfarrKfi ol ttw
LJi~.B new rjowh a. Li-ban
«JgC' fcwfcp1. meetly n ttv
Itarmcf rnuilno;
2B4.GO3 irai n UGB
P1JQ DO aaz added
HUGE)
HD UOflfiup'-rGKri 'TCfWtti Modvztt tnpanvzn of tha
-KcormiDrtatgd mrauqh UGB;gmwthfoanDclln
dwcJ[ipfTiErt of minting land ccfibrs. ccrndcrs ^nd
wthn-H-Burbcngrowtti rKigrtoinngciiiK
ZS7.MD acres in UGfi
IZ2JOM acnn addbd to
boundary.
234JODO. 1CRB n UQB
The Minneapolis-St. Paul region took the same approach in its Blueprint 2030, developing alternative
growth scenarios that all absorbed the same amount of growth—in this case, 280,000 households—and then
forecasting the impacts associated with each scenario.38 As in Portland's study, the growth scenarios varied
substantially in where in the region they located the 280,000 new households, and how dense those
households were developed in those locations. Total growth, however, was held constant across scenarios.
This approach has been used at the statewide level as well. New Jersey, in their State Plan, explicitly
addressed the question whether population and jobs would change under the PLAN versus business-as-usual
TREND, and found that, "It is anticipated that the TREND and PLAN scenarios will have essentially the
same population and household growth at the state and regional levels, but significantly different growth by
type of community and State Plan planning area. It is also anticipated that under the PLAN regimen there
will be more growth in communities with more densely developed planning areas and in communities with
urban, regional, and/or town centers, and that there will be less growth in these areas under the TREND
regimen." So, both PLAN and TREND scenarios analyzed "Accommodating a growth of 462,000
households and 802,500 jobs over the period 2000 to 2020 [requiring] approximately 486,500 housing units
and 422.5 million square feet of nonresident!al space."39
Although these three studies are excellent examples, they are by no means only examples of this approach to
regional and statewide growth planning. Other examples include:
40
• Puget Sound Regional Council's Vision 2020 (where and how to absorb 1.4 million people),
• San Francisco Bay Area's Smart Growth Strategy,41 developed by the Association of Bay Area
Governments (where and how to absorb 1 million new residents and 1 million new jobs),
Envision Utah (where and how to absorb 600,000 new residents by 2020),
42
Metropolitan Council, Blueprint 2030: "[EJach alternative future illustrates a distinct way in which the Twin Cities can
accommodate the Region's next 280,000 households (approximately 580,000 people) and 360,000 jobs.
http://www.metrocouncil.org/planning/blueprint2030/overview.htm
39 http://www.state.ni.us/osp/plan2/ias/sp3economic.pdf and http://www.state.ni.us/osp/plan2/ias/ia2000en.htm.
40 http://www.psrc.org/projects/vision/2020overview.htm
41 Association of Bay Area Governments, "Smart Growth Strategy: Shaping the Future o f the Nine-County Bay Area,"
Alternatives Report, April 2002. See http://www.abag.ca.gov/planning/smartgrowth/AltsReport/SmartGrowthStrategv.pdf
353
-------
While these studies have forecast the environmental impacts of a fixed amount of growth absorbed in
various locations and in various densities, they have not, in most cases, looked explicitly at water impacts.
The population and growth assumptions outline in this paper, then,
• Follows the standard model of growth impacts analysis by examining the impact on the
environment of a fixed amount of growth, absorbed in different locations and in different densities;
• Seeks to contribute to the standard approach by demonstrating that it is both possible and important
to add water to the list of impacts that is examined in this type of alternatives analysis for regional
and statewide growth.
In sum, the approach in this study is both consistent with the current state of the practice, and builds on it.
Finally, as we establish the assumptions for this analysis, it is important to note: we do not argue that the
projected 270,000-person growth increment is necessarily the correct number and that the growth is fixed
and known. It may be 240,000 persons, or it may be 340,000 persons. There is uncertainty in these
projections, as in all growth forecasting. However, we also know that some amount of growth is coming,
and that whatever the amount it will not vary as a result of lower or higher density development. That is the
sense in which it is fixed for the purposes of this policy analysis.
2. What happens if high-density development occurs and the remaining green space is developed as well?
Higher density development performs better at the watershed level because some green space is "saved" by
concentrating development regionally—see Exhibit 5 for an illustration of this dynamic. In other words,
accommodating more people in closer proximity can relieve development pressures at the edge. However,
critics argue that the undeveloped lands will be developed anyway, thereby further degrading water quality
by allowing higher densities and by developing on all the absorbent open space. However, there are two
issues with this critique:
(1) Growth is fixed. As discussed in the previous section. More growth will not arbitrarily come to
a region simply because there is space to expand.
(2) Comparisons between built out densities must keep the number of housing units accommodated the same. For
example, if critics argue that the high-density approach will bring more development to the remaining open spaces,
that same amount of development must be added to the comparison watershed that has developed at lower densities.
We have already explored the first issue. For the sake of exploring the second issue, we'll examine two
comparative watersheds in three stages: (1) each watershed accommodates the same number of housing units
but at different densities; (2) as the critics argue, the more dense watershed is fully built out, while no growth
is added to the comparison watershed; and (3) the comparison watershed accommodates the growth of the
more dense watershed, which means that each watershed accommodates the same number of housing units.
We're assuming that the watershed in question is 10,000 acres.
The first step in this process is to examine each watershed accommodating the same number of housing
units- but at different densities. Initial growth projections suggested that at 3 housing units per acre the
watershed would be fully built out. However, at the higher density level of 9 housing units per acre, only
one third of the watershed would be built out. The runoff associated with each of these scenarios is shown
in Exhibit 7.
42 See http://www.calthorpe.com/Project%20Sheets/Envision%20Utah.pdf
354
-------
As Exhibit 7 shows, if development occurs at a lower density, e.g., 3 housing units per acre, the entire
watershed will be built out. Any additional development that occurs in this community will have to go into
another watershed, since this watershed is built out. This total buildout will generate 214 million ft3 average
annual stormwater runoff, assuming one hypothetical outfall. This is approximately one-third more
stormwater runoff that the watershed that is developed at the higher density. In this situation, developing at
the lower density seems worse for watershed water quality.
Exhibit 7: Hypothetical 10,000-acre Watershed Developed at Different Densities
Scenario 1: The 10,000-acre watershed is fully
built out at 3 housing units per acre. 30,000
housing units are accommodated. This translates
to:
10,000 acres x 3 housing units x 7,133 ft3 of runoff
214 million ft3 average annual stormwater runoff
30,000 housing units accommodated
Scenario 2: The 10,000-acre watershed is only
partially built out because development is occurring at
higher densities—9 housing units per acre. 30,000
housing units are still accommodated. This translates
to:
1/3 (10,000 acres) x 9 housing units x 4,767 ft3 of
runoff
141.57 million ft3 average annual stormwater runoff
30,000 housing units accommodated
But what happens if the remaining 2/3 of the watershed in Scenario 2 is built out, as was initially suggested?
Exhibit 8 examines those numbers considering the worst case situation—that the remaining land in the
watershed is developed at the higher density of 9 housing units per acre.
355
-------
Exhibit 8: Hypothetical 10,000-acre Watershed Developed at Different Densities
Scenario 1: The 10,000-acre watershed is fully built
out at 3 housing units per acre. 30,000 housing
units are accommodated. This translates to:
10,000 acres x 3 housing units x 7,133 ft3 of runoff
214 million ft3 average annual stormwater runoff
30,000 housing units accommodated
Scenario 2: The 10,000-acre watershed is fully built
out at 9 housing units per acre. 90,000 housing units
are accommodated. This translates to:
10,000 acres x 9 housing units x 4,767 ft3 of runoff
429 million ft3 average annual stormwater runoff
90,000 housing units accommodated
Now, both watersheds are fully built out and the watershed developed at the higher density, e.g., developed at
9 housing units per acre, is generating approximately double the total stormwater runoff. This would be worse
for watershed water quality if both scenarios accommodated the same amount of growth. However, note that
the watershed with the higher density is accommodating 60,000 more units of housing, or three times the
number of housing units. And, as was discussed in the previous section, growth is fixed. A region will not
accommodate unlimited growth. In essence we are projecting what would happen if the regional growth is
three times higher than initially projected. So, where are those additional housing units accommodated in the
watershed that was developed at the lower-density? They were built in nearby or adjacent watersheds. So, to
continue with the analysis, if regional forecasts were wrong and 90,000 housing units were needed and not
30,000 housing units, then the watershed developed at the lower density level, e.g., 3 housing units, will need
to expand into two additional watersheds to accommodate the same growth! Exhibit 9 illustrates this situation.
356
-------
Exhibit 9: Hypothetical 10,000 acre Watershed Developed at Different Densities
Scenario 1: The 10,000-acre watershed is fully built
out at 3 housing units per acre. But an additional
60,000 housing units must be accommodated. This
will require total build out of another 2 entire
watersheds. This translates to:
10,000 acres x 10,000 acres x 10,000 acres x 3
housing units x 7,133 ft3 of runoff
642 million ft3 average annual stormwater runoff
90,000 housing units accommodated
Scenario 2: The 10,000-acre watershed is fully built
out at 9 housing units per acre. 90,000 housing units
are accommodated. This translates to:
10,000 acres x 9 housing units x 4,767 ft3 of runoff
429 million ft3 average annual stormwater runoff
90,000 housing units accommodated
As Exhibit 9 demonstrates, accommodating an additional 60,000 housing units requires disturbing and
developing another 2 watersheds. Total average annual stormwater runoff from accommodating 90,000
housing units at 3 housing units per acre generates 642 million ft3 average annual runoff. While the
watershed developed at the higher density, e.g., 9 housing units per acre, has still just disturbed one
watershed and is generating approximately one third less stormwater runoff—or 429 million ft3 average
annual runoff.
3. Urban water infrastructure is failing — how can it accommodate more users?
It is better to preserve public investments by investing where the public has already invested. It is a poor
strategy economically and environmentally to divert development away from any area because
infrastructure is failing. For example, in a report by the Office of Technology Assessment, one official of a
large western city reported that it costs the city $10,000 more to provide infrastructure services to a house
on the suburban fringe than one in the urban core.43 Myron Orfield, a member of Minnesota's House of
Representatives, calculated that by 1992, the central cities of Minnesota were paying over $6 million
annually to subsidize growth in edge areas. This was especially troubling to areas like Minneapolis, which
had 22 percent existing sewer service that in 1990 remained undeveloped. Rather than directing growth to
this area, between 1987 and 1991, the region provided new capacity to 28 square miles of land at the cost of
$50 million per year.44 The capacity went primarily to serve expansion into the development affluent
43
US Congress, Office of Technology Assessment, The Technological Reshaping of Metropolitan America. Washington, DC: US
Government Printing Office, 1995. OTA-ETI-643.
44
Orfield, Myron. 1997. Metropolitics: A Regional Agenda for Community and Stability. Washington, DC: Brookings
Institution Press.
357
-------
southwest suburbs. This kind of infrastructure spending subsidizes and encourages development at the
fringe.
The implications of building new infrastructure instead of maintaining existing infrastructure is that it is
apparently more important to provide new infrastructure than to maintain good service in existing
communities. This signal leads to an unwillingness to invest on the part of private owners. Thus, a catch-22
situation begins—an area is degraded and no one, including the local government, wants to invest in it,
which causes further degradation of the area. The result of this type of disinvestment causes the movement
of people and businesses out of the community to newer developed areas. This movement can lead to
sprawl even in the absence of significant population growth. This has been evidenced in numerous cities
such as Buffalo, New York, Cleveland, Ohio, and Pittsburgh, Pennsylvania. All these cities experiences
population loses at the same time as their land consumption and urbanized area grew. The result of this
outward growth, with or without population growth, is a significant increase in watershed or regional
impervious cover, which will further degrade regional water quality.
4. Shouldn 't increasing densities be accompanied by open space offsets?
This question essentially asks that if the benefits of density are derived from undeveloped open space,
shouldn't there be a requirement that this land be preserved? Earlier, we discussed the issue that the
resulting open space will be developed in addition to the higher density development is in essence a fear that
the region will receive more growth than anticipated. The implication here is that without some active
preservation, the open space will be developed anyway. As discussed, this growth would have come to the
region in either the low density or the higher density scenario and we asked which density development
pattern would accommodate this new growth with the least impact to water quality.
There is a fixed amount of growth coming to any given region. Once that growth is accommodated,
developers (in the private or public sector) will not continue to develop land independent of the demand for
that development. For example, if the market anticipates that 1,000 new households will be coming to the
region, it will supply 1,000 new units of housing. Once those units are supplied the market will not then add
another 1,000 units. This is unaffected by the density at which the 1,000 are supplied. Thus, the open space
remains undeveloped simply by virtue of the fact that the higher density development alleviates the need for
the development of additional land. If, on the other hand forecasts are incorrect and an additional 500 units
of housing comes to the region, then we are left with the original question, "What is the best way to
accommodate this growth?"
The second problem with linking open space preservation requirements to higher density development is
that it can create unintended consequences that may harm water quality. For example, there are two ways to
link open space preservation to higher density development: require the developer to provide the open space
offsets of some type, or use public tools such as downzoning or open space purchases to achieve
preservation while increasing densities elsewhere. Either approach adds a barrier for the developer who
wants to build higher density development. The "high-density" developer would be faced with the
additional time and cost of complying with these rules while the "low-density" developer would have no
such barrier or cost. In essence, either strategy puts an extra burden on the development product that is, by
itself, more protective of water quality. Water quality professionals are not in the business of making
developments easier for developers to build. However, by tilting the playing field towards "low density"
developers, it is likely that more low density projects will be built, thereby further consuming absorbent
open space, increasing transportation-related impervious cover, and overall, increasing the footprint of a
358
-------
region. As a result, in an attempt to guarantee water quality benefits by linking open space preservation to
higher density development we actually hasten water quality declines. One could argue that it is more
important is the role of open-space offsets in low-density zoning environments. Given that low-density
development drives subsequent development further into undisturbed land, it would appear more important
to attach offset requirements to low-density zoning than to the land-conserving approach of compact
development.
The question underlying this issue is how can communities determine where to develop and where to
preserve? In all development scenarios, ensuring adequate open space for water quality, flooding
mitigation, sports and recreation, habitat, and biodiversity is a critical part of the planning process.
Hydrologically speaking, it is generally accepted that more open space is needed, and specifically removal
of development from flood plains. Not all land has equal ecological value and it is critical for local
governments to determine where the critical ecological systems exist within their region and to take steps to
preserve these areas. Once this process of determining how to minimize new development and maximize
retention and reclamation of open space, a community will perhaps have in place a significant network of
green infrastructure.45
In addition, open space preservation specialists argue that for an open space plan to be effective, preserved
parcels must be large enough to serve a critical environmental function and, if possible, connected. By
requiring any development to have an open space offset, a community has the potential of creating a hodge
podge of spaces that may or may not have significant environmental value. In addition, open space offsets,
in the worse case scenario, cause leapfrog development. What some communities have done to address the
issue of preserving open space in the face of mounting development pressures is to require all new
developments, high and low densities, to pay a fee into a general fund. The local government then uses
these funds to acquire or purchase the lands they have identified as having high environmental, economic, or
social value.
5. Do infill sites (such as brownfields and grey fields) represent a particular opportunity?
This paper has demonstrated that compact development produces less stormwater runoff on a per-unit basis
than does low-density dispersed development. Communities can enjoy a further reduction in runoff if they
take advantage of underutilized properties, such as infill, brownfield, or greyfield46 sites. For example, an
abandoned shopping center (a greyfield property) is often almost completely impervious cover, and is
already producing high volumes of runoff. If this property is redeveloped, the net runoff increase will likely
be zero since the property was already predominately impervious cover. In many cases, redevelopment of
these properties will break up or remove some portion of the impervious cover, converting it to pervious
cover and allowing for some stormwater infiltration. In this case, redevelopment of these properties can
produce a net improvement in regional water quality by decreasing total average annual . Exhibit 11
illustrates this opportunity.
45 For more information on the environmental and ecological benefits of preserving open space, please see Trust for Public
Land's " Economic Benefits of Preserving Open Space;" and "Local Greenprinting for Growth: Using Land Conservation to
Guide Growth and Preserve the Character of Our Communities."
46 Greyfield sites generally refer to abandoned or underutilized shopping malls, strip malls, or other areas that have significant
paved surface and little or no contamination (in order to distinguish it from brownfield sites). For more information on greyfield
sites and the potential for redevelopment, please see Urban Land Institute's publication, "Turning Greyfields into Goldfields."
359
-------
Exhibit 10: Redevelopment of a Grey field Property
Before Redevelopment
After Redevelopment
Utilization of brownfield and grey field sites can reduce regional land consumption and ensure
accommodation of projected growth thus decreasing its environmental impact. A recent George
Washington University study found that for every brownfield acre that is redeveloped, 4.5 acres of open
space are preserved.47 In addition to redeveloping brownfield sites, regions can identify underutilized
proprieties or land, such as infill or greyfield sites, and target those areas for redevelopment. For example, a
recent analysis completed by King County, Washington demonstrated that property that is vacant and
eligible for redevelopment in the county's growth areas can accommodate 263,000 new housing units—
enough for 500,000 people.48 Redeveloping this property represents an opportunity to accommodate new
growth without degrading water quality. As discussed, much of the abandoned properties in areas are
already close to 100 percent impervious cover. By taking advantage of these properties, a community
experiences the benefits of growth without the costs of water quality degradation. Finally, in addition to
water quality benefits, if these properties are developed at higher densities, a local government can ensure
that more people are accommodated in areas with existing infrastructure, housing choices, and
transportation choices.
6. What about localized hot spots ?
One of the largest benefits about developing at higher densities are the other community opportunities that
become more viable because of more people living in closer proximity to each other. For example, bus
transit becomes viable at 7 units an acre, while light rail and subway become viable at 15-20 units an acre.49
Mixed use, such as first floor retail, becomes viable only at higher densities. And, community walkability
and livability increase dramatically as densities increase.50 Increasing densities on a regional scale is more
Deason, Jonathan, et al. "Public Policies and Private Decisions Affecting the Redevelopment of Brownfields: An Analysis of
Critical Factors, Relative Weights and Area Differentials." Prepared for US EPA Office of Solid Waste and Emergency Response.
The George Washington University, Washington, DC. September, 2001. Available at
www.gwu.edu/~eem/Brownfields/project report/report.htm.
48 Pryne, Eric. "20 Years' Worth of County Land?" Seattle Times, Monday, May 20, 2002.
49 Ewing, Reid. "Pedestrian and Transit-Friendly Design: A Primer for Smart Growth. ICMA: Washington, DC. 1999.
50 For more information on the other benefits of density, please see, ICMAs publication, "Getting to Smart Growth: 100 Policies
for Implementation;" www.smartgrowth.org: and www.smartgrowthamerica.org.
360
-------
protective of water quality, overall, but has the potential to create localized hot spots that affect proximate
water bodies. EPA estimates that over 70 percent of urban water bodies are impaired. If a local community
increases densities in their development patterns, while better for overall regional watershed health, there is
a real potential to increase pollutant loadings in water bodies new or adjacent the new development. Of
course, even with low-density development, creating hotspots is also a real potential, but because of the
slightly higher runoff and pollution levels of the higher-density development patterns, as demonstrated,
localized hot spots are a greater concern.
This paper suggests that the answer to this question is to protect pristine watersheds and overall watershed
health through compact development and mitigate hot spots. There are two approaches for mitigating
hotspots:
(1) Address increased pollutant loads at the site- and development-level, reducing the amount of runoff and
associated pollutants entering the system through structural or non-structural best management practices,
such as riparian buffer zones or conservation easements, or low-impact development; and
(2) Reduce the overall levels of "background" pollution, thereby allowing the streams and water bodies to
absorb more pollution from localized hotspots while still maintaining water quality standards.
EPA and other organizations, such as the Center for Watershed Protection, have written extensively about
numerous best management practices and low-impact development techniques that reduce site- or
development-specific stormwater runoff and associated pollutants.51 For example, low-impact development
is increasingly recognized as one mechanism to reduce effective impervious cover and to allow natural
features to serve their ecological functions. Some LID techniques include:
• Rain gardens and bioretention;
• Rooftop gardens or simple roof storage;
• Tree preservation and planting;
• Vegetated swales, buffers, and strips;
• Roof leader disconnection;
• Rain barrels and cisterns;
• Impervious surface reduction and disconnection;
• Soil amendments;
• Permeable pavers; and
• Pollution prevention and good housekeeping.52
The Center for Watershed Protection recently released a document that details 11 techniques for reducing
water quality impacts from development. While this document, "Redevelopment Roundtable Consensus
Document,"53 is geared for urban infill redevelopment opportunities, many of the practices described, such
as, "Design sites to maximize transportation choices in order to reduce pollution and air and water quality,"
can also be applied to high-density greenfield developments.
51 See, for example, www.bmpdatabase.org and www.stormwatercenter..net.
52 Woodworth, "Out of the Gutter."
53 For more information on this document, please see http://www.cwp.org/pubs_download.htm.
361
-------
Unlike reducing site-specific impacts that only require innovation and desire on the part of the developer,
reducing background levels of pollution generally require some type of local government involvement. For
example, stormwater management utilities provide an opportunity for the local government to address the
most pressing stormwater problems. Residents, commercial, and industrial users of wastewater treatment
plants pay into this fund, giving the localities the funds and flexibility to address the area's most severe
problems. Other regional examples include:
• Variable sewer hookup fees, such as in Sacramento, California, which recently changed its hookup
fees to vary by location and type of development. This results in developers having to pay almost
twice as much to hook up sewer lines in fringe or edge areas as in urban areas.
• Maine charges "compensation fees" to residents and commercial entities for not meeting statewide
phosphorus reduction requirements. These fees enable the state to address the increasing
phosphorous problem at the source—either in locations with hot spots or at the waste water
treatment facility.
• North Carolina has established density averaging of non-contiguous parcels, and density trading
with buffer zones. The goal of this program is to encourage density in clusters, that is, encourage
density without a net increase in watershed development density.
These and other regional policies are described in an EPA document, "Protecting Water Resources with
Smart Growth: 100 Policies.'"54 This report describes both site-specific and regional policies that local
communities have put in place to address localized hotspot and associated water quality issues.
To demonstrate the importance of these principles in reducing site- and development-related hot spots, the
University of Oregon conducted a study entitled: "Measuring Stormwater Impacts of Different
Neighborhood Development Patterns."55 The study site near Corvallis, Oregon, was created to compare
stormwater management strategies in three common neighborhood development patterns.56 For example,
BMPs, such as disconnecting residential roofs and paving from the stormwater system, introducing swales
and water detention ponds into the sewer system, and strategically locating open space had significant
impacts on peak water runoff and infiltration. The study concludes that:
"Some of the most effective opportunities for reducing stormwater runoff and decreasing peak flow
are at the site scale and depend on strategic integration with other site planning and design decisions.
"Reduced street networks of narrower streets and planting strips significantly reduce the amount of
pavement and as a result, runoff, in urban areas. Best management practices such as swales,
constructed wetlands and ponds integrated with urban streets and open space networks are also
important to collect, clean, store and slow the flow of runoff. However, these facilities and their
physical relationships must be planned early to be well orchestrated and effective."57
54 This document will be ready for distribution by June 1, 2003.
55 Study description and results on neighborhood.uoregon.edu/projects/research/owrri/owm.html.
56 The University of Oregon used the PCSWMM model developed by Computational Hydraulics, in Guelph, Ontario, Canada.
57 http://neighborhood.uoregon.edu/proiects/research/owrri/owrri conclusion.html
362
-------
7. Won't increasing densities create sacrifice zones?
In the mid-1990's advocates for the environment, affordable housing, farmland preservation, transportation
reform, and community reinvestment started calling on communities to develop in new ways. Since then,
citizens across the nation are demanding it - in polls, in the market, and at the ballot box. Americans want
fewer hours in traffic and more opportunities to enjoy green space; housing that is both affordable and close
to jobs and activities; healthy cities, towns and suburbs; air and water of the highest quality; and a landscape
our children can be proud to inherit. Increasing densities and determining where we should develop and
where we should preserve offers the best chance of attaining those goals. Not only will our communities
thrive economically and socially, but also environmentally. Increasing densities provides a mechanism for
communities to accommodate growth, enjoy economic development and jobs in the most environmentally
protective way possible.
WHAT HAS OTHER RESEARCH FOUND?
Current research suggests that compact development and/or redevelopment in existing areas will impact
water quality less than scattered, low-density development. Several site-specific studies have been
conducted across the country to predict the runoff and pollutant loading responses to changing land use.
This section highlights five case studies that approach the research question with varying levels of
complexity. Jordan Cove in Connecticut; Belle Hall in South Carolina; a statewide analysis of New Jersey;
Chicago in Illinois; and an analysis done by the Chesapeake Bay Foundation each analyze the differences in
runoff and associated water pollution from different types of development.
Researchers at Jordan Cove58 development in Waterford, Connecticut are finding that, when compared to
high-density design development, the large lot development, or low-density design, produces 95 more
runoff during construction. Using monitoring data from two study sites and a control site, these paired sites
will evaluate "Traditional" suburban development, "BMP" development, and the control subdivision. Early
results from storm events during construction indicate that construction of the large lot neighborhood is
causing significant impacts on runoff quality and quantity, including observed increase in mean weekly flow
volume (99 percent), runoff frequency (from 16 to 95 percent), and mean weekly peak discharge (79
percent).
The Belle Hall study, completed by the South Carolina Coastal Conservation League (1995), examined the
water quality impacts of two development alternatives for a 5 83-acre site in Mount Pleasant, South
Carolina. In the "Sprawl Scenario," the property was analyzed as if developed along a conventional
suburban pattern. The "Town Scenario," was analyzed if using the development incorporated traditional
neighborhood patterns instead. In each scenario, the overall density and intensity (the number of residential
unit, square feet of commercial and retail space, and so forth), was held constant, although the building
types and sizes vary. The results found that "Sprawl Scenario" consumed 8 times more open space,
Cote, M.P., Clausen, I, Morton, B., Stacey, P., Zaremba. S. 2000. Jordan Cove Urban Watershed National Monitoring Project.
Presented at the National Conference on Tools for Urban Water and Resource Management Protection, Chicago, IL. See also
Engdahl, J. 1999. Impacts of Residential Construction on Water Quality and Quantity in Connecticut. University of Connecticut,
Storrs, CT.
h2osparc.wq.ncsu.edu/96rept319/CT-96.html
www.epa.gov/owow/estuaries/coastlines/summer98/jordancove.html
www. canr.uconn. edu/iordancove/
363
-------
generated 43 percent more runoff, 4 times more sediment, almost 4 times more nitrogen, and 3 times more
phosphorous as compared to the "Town Scenario" development.59
New Jersey's State Plan calls for increasing densities in the state by directing development to existing
communities and existing infrastructure ("Plan"). Researchers at Rutgers University analyzed the water
quality impacts from "Trend" versus "Plan" development. The study found that compact development
("Plan" development) would generate significantly less water pollution than low-density development
("Trend" development) for all categories of pollutants.60 The reductions ranged from over 40 percent for
phosphorus and nitrogen to 10 percent for lead. The smaller impervious areas would produce 30 percent less
runoff, and concentrating this development in areas served by sewers would reduce its impact on the
environment by another 10 percent.61 These conclusions supported a similar statewide study completed in
1992 that concluded that compact development would result in 30 percent less runoff and 40 percent less
water pollution than would a sprawl scenario.62
Researchers at Purdue University examined two possible project sites in the Chicago, Illinois area.63 The
first site was in the urban core and currently consists of a mix of residential, industrial, and commercial
properties. The second site was on the urban fringe. The results found that placing a hypothetical low-
density development at the Chicago fringe area would produce 10 times more runoff than a higher-density
mixed-use development located in the urban core.
Finally, a study published by the Chesapeake Bay Foundation in 1996 comparing conventional and
clustered suburban development on a rural Virginia tract found that clustering would convert 75 less land,
create 42 percent less impervious surface, and produce 41 percent less stormwater runoff.64
CONCLUSIONS
As metropolitan areas continue to grow in population, the area of the region's built environment will
continue to expand. How and where this development occurs will have a profound impact on water quality.
EPA believes that increasing densities of all developments can minimize water quality impacts from
development. Nationwide, state and local governments are considering the environmental implications of
development patterns. A growing body of research clearly documents that the creation of impervious cover
causes a predictable and profound decline in critical elements of aquatic ecosystems.65 Conventional low-
density development and its attendant infrastructure consume previously undeveloped land and create
stretches of impervious cover throughout a region. In turn, these land alterations are not only likely to
degrade the quality of the individual watershed, but are also likely to degrade a larger number of
watersheds.
59 South Carolina Coastal Conservation League, EPA, NOAA, SC Department of Health and Environment; Town of Mount
Pleasant. 1995. The Belle Hall Study: Sprawl vs. Traditional Town: Environmental Implications. Dover, Kohl, and Partners,
South Miami, FL.
60 Ibid.
61 University of Rutgers. 2000. The Costs and Benefits of Alternative Growth Patterns: The Impact Assessment of the New
Jersey State Plan. Center for Urban Policy and Research.
62
63
Pollard, Trip. "Greening the American Dream." Planning Magazine: American Planning Association, October 2001.
Harbor, J., Engel, B., et al. "A Comparison of the Long-Term Hydrological Impacts of Urban Renewal versus Urban Sprawl."
Purdue University: West Lafayette, IN. 2000.
64 Pollard.
65 See Arnold, Chester L. Jr., C. James Gibbons. See also EPA, Urbanization and Streams: Studies of Hydrological Impacts.
Washington, DC: EPA Office of Water. 1997. EPA # 841-R-97-009.
364
-------
Concentrating development in urban areas maintains the functions of smaller watersheds because at the
regional or watershed scale, impervious cover is minimized and undisturbed open space is maximized.
Further, development decisions can often affect transportation-related imperviousness across multiple
watersheds. While a low-density scenario often subjects numerous watersheds to possible degradation, a
compact scenario can limit the number of watersheds affected by development.66 This review of the effects
of different development densities on water quality suggests three conclusions:
1. Compact development is better good for water quality than less compact development. It minimizes the
consumption of land needed to support critical watershed functions, which in turn minimizes the
creation of impervious surfaces that lead to increased runoff, and associated pollutants. And intensifies
activity in a smaller area - e.g., less motor traffic outside of cities.
2. There is no reason to expect that lower density development reduces total or even necessarily
(depending on site design and building type) site-level runoff,67 or are protective of watershed water
quality. Rather, this paper and the literature suggest that, all else being equal, accommodating new
growth through higher densities will likely be more protective than lower density development.
3. The denser development should be given preference, because its lower per-unit runoff minimizes the
impact of a given increment of growth, and leaves more room for additional growth.
4. Regions can enjoy a substantial bonus from re-using existing brownfields, greyfields, and other sites that
are already impervious. Building on these saves land elsewhere, can often accommodate higher
densities, and can reduce flows from the developed parcel.
In sum, compact development is an environmental protection strategy, and should be included in any set of
such strategies that are reviewed as part of a search for ways to protection water quality, whether at the
local, state, or national level.
ACKNOWLEDGEMENTS
The authors would like to thank Will Schroeer from ICF Consulting for his research assistance. This paper
is currently being reviewed by EPA's Office of Water and outside peer reviewers. Their comments will be
incorporated into the final draft of this paper.
66 See review of several cases in US EPA, Development, Community and Environment Division, Our Built and Natural
Environments: A Technical Review of the Interactions between Land Use, Transportation, and Environmental Quality, EPA #123-
R-01-002, 2001. pp. 41-43.
67 Keeping the number of housing units similar to a higher density development.
365
-------
REFERENCES
American Farmland Trust. 1994. Farming on the Edge: A New Look at the Importance and Vulnerability of
Agricultural Near American Cities.
Arnold, C. L. and C. J. Gibbons. 1996. Impervious Surface Coverage: The Emergence of a Key
Environmental Indicator. Journal of the American Planning Association 62(2) 243-258.
Association of Bay Area Governments. 2002. Smart Growth Strategy: Shaping the Future of the Nine-
County Bay Area, Alternatives Report. See
http://www.abag.ca.gov/planning/smartgrowth/AltsReport/SmartGrowthStrategy.pdf
Brown, Patricia Leigh, "The Chroming of the Front Yard," New York Times, June 13, 2002, p Fl.
Cappiella, K. and Brown, K. 2001. Impervious Cover and Land Use in the Chesapeake Bay Watershed.
Ellicott City. MD.
Center for Watershed Protection (CWP). 1998. Rapid Watershed Planning Handbook—A Comprehensive
Guide for Managing Urban Watersheds. Ellicott City, MD.
Chesapeake Bay Foundation. 2001. State of the Bay Report. Annapolis, MD.
Cote, M.P, Clausen, J., Morton, B., Stacey, P., Zaremba. S. 2000. Jordan Cove Urban Watershed National
Monitoring Project. Presented at the National Conference on Tools for Urban Water and Resource
Management Protection, Chicago, IL.
Deason, Jonathan, et al. 2001. Public Policies and Private Decisions Affecting the Redevelopment of
Brownfields: An Analysis of Critical Factors, Relative Weights and Area Differentials. Prepared for US
EPA Office of Solid Waste and Emergency Response. The George Washington University, Washington,
D.C.
Engdahl, J. 1999. Impacts of Residential Construction on Water Quality and Quantity in Connecticut.
University of Connecticut, Storrs, CT
Ewing, Reid. 1999. Pedestrian and Transit-Friendly Design: A Primer for Smart Growth. ICMA:
Washington, D.C.
Krugman, Paul. 1991. Geography and Trade. MIT Press, Boston, MA.
Krugman, Paul . 1995. Development, Geography, and Economic Theory. MIT Press, Boston, MA.
Harbor, J., Engel, B., et al. 2000. A Comparison of the Long-Term Hydrological Impacts of Urban Renewal
versus Urban Sprawl. Purdue University: West Lafayette, IN.
McGrath ,Daniel T 2001. 2025 Urban Land Area Forecasts for the US Top 20 Coastal Metropolitan
Regions. Great Cities Institute, unpublished, available at
http://iisgcp.org/research/cbe/2025forecast.htm.
366
-------
Metropolitan Council, Blueprint 2030. http://www.metrocouncil.org/planning/blueprint2030/overview.htm
Mills, Edwin, B. Hamilton. 1994. Urban Economics: Fifth Edition. Harper Collins College Publishers.
Montgomery County Department of Environmental Protection. 2000. Stream Conditions Cumulative
Impact Models for the Potomac Subregion. MCDEP, MD.
National Association of Home Builders. 2001. Housing Facts, Figures, and Trends: 2001. NAHB:
Washington, DC.
New Jersey Office of Planning. 2002. http://www.state.nj.us/osp/plan2/p2full/colorsOO.htm.
Orfield, Myron. 1997. Metropolitics: A Regional Agenda for Community and Stability. Brookings
Institution Press, Washington, D.C.
Pollard, Trip. 2001. Greening the American Dream. Planning Magazine: American Planning Association.
Portland Metro, 2000. The Nature of 2040: The region's 50-year plan for managing growth." See
http://www.metro-region.org/growth/tf/2040history.pdf
Pryne, Eric. "20 Years' Worth of County Land?" Seattle Times, Monday, May 20, 2002.
Rein, Lisa and David Cho, "In Defense of the Front Lawn: Fairfax Attacks Crowding With Ban on Oversize
Driveways," Washington Post., June 4, 2002, p. Al.
Schueler, T 1994. The Importance of Imperviousness. Watershed Protection Techniques 1(3): 100-111.
Schueler, T. 2000. The Compaction of Urban Soil. Techniques for Watershed Protection. Center for
Watershed Protection, Ellicott City, MD.
Soil Conservation Service. 1986. Technical Release No. 55 (TR-55). Urban Hydrology for Small
Watersheds. Washington, DC.
South Carolina Coastal Conservation League, EPA, NOAA, SC Department of Health and Environment;
Town of Mount Pleasant. 1995. The Belle Hall Study: Sprawl vs. Traditional Town: Environmental
Implications. Dover, Kohl, and Partners, South Miami, FL.
Trust for Public Land and the National Association of Counties. 2002. Volume 1: Local Greenprintingfor
Growth: Using Land Conservation to Guide Growth and Preserve the Character of Our Communities.
University of Rutgers. 2000. The Costs and Benefits of Alternative Growth Patterns: The Impact
Assessment of the New Jersey State Plan. Center for Urban Policy and Research.
U.S. Census. 2000. Annual Projections of the Total Resident Population as of July 1: Middle, Lowest,
Highest, and Zero International Migration Series, 1999 to 2100. Washington, D.C.
367
-------
U.S. Congress, Office of Technology Assessment. 1995. The Technological Re shaping of Metropolitan
America. Washington, D.C., 1 OTA-ETI-643.
U.S. Department of Agricultural, Economic Research Service, Natural Resources and Environmental
Division. June 1997. Agricultural Resources and Environmental Indicators (AREI) Updates, No. 3.
"Major Land Use Changes in the Contiguous 48 States."
U.S. Environmental Protection Agency. 1997. Urbanization and Streams: Studies of Hydrological Impacts.
Washington, DC: EPA Office of Water. EPA # 841-R-97-009.
U.S. Environmental Protection Agency (USEPA). 2000a. National Water Quality Inventory: 1998 Report to
Congress, www.epa.gov/305b/98report. Last updated October 5, 2000.
U. S. Environmental Protection Agency, Office of Water. June 2000b. Water Quality Conditions in the
United States: A Profile from the 1998 National Water Quality Inventory Report to Congress.
Washington, DC. EPA841-F-00-006 (also available at www.epa.gov/OWOW/305b/).
Woodworm, James et al. 2002. Out of the Gutter: Reducing Polluted Runoff in the District of Columbia.
NRDC: Washington, DC.
368
-------
PREDICTING THE IMPACT OF URBAN DEVELOPMENT ON STREAM
TEMPERATURE USING A THERMAL URBAN RUNOFF MODEL
(TURM)
A. Roa-Espinosa1, T.B. Wilson2, J.M. Norman2, and Kenneth Johnson3
!Dane County Land Conservation Department, Fen Oak Resource Center, 1 Fen Oak Court,
Madison, WI 53718. Biological Systems Engineering, University of Wisconsin, Madison, 460
Henry Mall, Madison, WI 53706
2Department of Soil Science, University of Wisconsin-Madison, 1925 Linden Drive, Madison,
WI 53706
3 Wisconsin Department of Natural Resources
Abstract
In this paper, we present a Thermal Urban Runoff Model (TURM) developed by the Dane
County Land Conservation Department and the University of Wisconsin-Madison to predict the
effect of urban development on runoff thermal regime. The model can predict the temperature
increase of runoff from impervious surface by calculating the heat transfer between runoff and
the heated impervious surfaces that commonly exist in urban areas. The model mainly assumes a
complete mixing of runoff water to predict the heat transfer and the thermal gradient within the
impervious media in contact with the runoff flow. Runoff temperature measurements indicate
that the hot paved surfaces receiving rainfall initially produce energy released by evaporation,
but high temperature runoff is quickly generated by the gradual increase in rainfall intensity.
TURM can also predict the temperature reduction after the runoff passes through rock-filled
channels; open vegetated swales, infiltrating surfaces; conduits and rock-filled chambers that can
be used to cool the first flush of heated storm water runoff. Data collected during summertime
storms indicate that determination of the air and rainfall temperatures is critical in predicting the
runoff temperature.
TURM was used to evaluate the heating of runoff water during summertime and its impact
characteristics at two urban subdivisions in Dane County, Wisconsin with different proportions
of imperviousness. The percentage of imperviousness and rainfall depth defined the changes in
runoff rate, volume and the timing of runoff. The model predictions for the temperature increase
in runoff agree very well with site-specific measurements.
This study is an attempt to fill the knowledge gap that currently exists in determining the thermal
impact of urban runoff on coldwater systems. The justification of this research effort is to
provide a useful tool to urban planners, fishery managers, biologists, and the engineering
community in Wisconsin to better manage impact of large urban development. TURM is still at
the early stages of development and additional work is required to make the model applicable to
wide array of practical situations.
Keywords: urban imperviousness, runoff: cold water stream, heat transfer, and thermal impact.
369
-------
1. Introduction
The increase in temperature of stream waters has historically received little attention.
However, recent studies have suggested that the expanding urbanization has a strong thermal
impact on small streams, and as a result, water temperature is now being considered as a part of
the permitting process for urban development throughout Wisconsin. Stream water temperature
is a limiting factor for cold-water fisheries and is the "narrowest door" in the water system, as all
biological activity depends on temperature. Over time, the cumulative impact of hundreds of
individual development sites will slowly increase water temperature, affecting the habitat for
every stream biota.
Temperature is a characteristic of water quality and is very important in chemical and
biochemical processes, particularly those involving biochemical activity. Higher stream
temperatures result in lower dissolved oxygen (DO) concentrations and may cause biological
oxygen demand (BOD) to increase. Temperature increases in streams can also result in changes
in the behavior offish and macro invertebrates (aquatic insects). Stenothermal fish are very
sensitive to temperature changes, with a physiological optimum temperature of <20 ° C, while
temperatures above 26 ° C are considered lethal. Brown Trout (Salmo trutta\ for example, have
an optimum temperature range of 7 to 17 ° C and become stressed at temperatures above 19 ° C.
Macro invertebrates, such as Stoneflies (Plecoptera sp) and Caddis flies (Trichptera sp), have a
maximum temperature of 17 ° C and are important not only because they are the primary food
source for trout, but because they are indicators of the overall health of the ecosystem. As a
result, cold-water streams are apparently the most ecologically sound at temperatures between 7
and 17 ° C (Lyons and Wang, 1996, Simonson, 1996)
Urban runoff heating is recognized as the biggest threat to cold-water streams. The
permanent warming of streams is often due to the increase in imperviousness and the heating of
runoff water in contact with warm surfaces. The runoff is heated as it passes over the impervious
surfaces with large heat storage due to solar radiation. In Dane County, Wisconsin, measured
runoff temperatures from urban impervious areas have been as high as 29 °C. Excessive heated
runoff can substantially and permanently harm runoff receiving cold-water streams. Widely
elevated water temperatures can impair the health of aquatic organisms and are responsible for
habitat degradation in the headwaters of cold-water streams in urban areas. It is also warmed by
the displacement of stored runoff heated by summer conditions that are in line with the storm
water conveyance systems, such as wet detention basins.
Increased area under impervious surface in urban areas is a major source of thermal
heating in cold climates and can threaten the health of cold-water ecosystems. Impervious areas
absorb energy from the sun, which causes them to become warmer. As water runs over these
areas, it absorbs some of that heat energy and is warmed, causing thermal pollution in lakes,
rivers, and streams. Impervious areas also compound the problem by reducing infiltration, which
in turn increases the volume of runoff that is created, leading to higher permanent stream
temperatures in the summer months. By mitigating runoff and water temperature impacts, the
stream community will benefit not only from temperature reduction, but also from a decline in
the amount of sediment, nutrients, and pollution that reaches receiving waters.
370
-------
The issues of urban runoff thermal impacts require the use of detailed models of the
urban surface-water-atmosphere system. Modeling the heat transfer from warm surfaces to
runoff water provides a means of assessing the contributions of various factors to the overall rise
in water temperature. Some of these factors that may significantly affect the water temperature
are solar radiation, air temperature, relative humidity, wind speed, the temperature and amount of
rainfall or runoff, and the temperature and amount of ground water entering the river or stream.
The objective of this analysis is to develop a reliable urban rainfall-runoff model that
includes a thermal component for impervious areas. The justification of the Thermal Urban
Runoff Model (TURM) is to enable communities with cold-water streams to better manage
development and minimize thermal impacts to streams. The focus of this paper is therefore on
three specific objectives: (1) to provide evidence of the thermal impact of urban imperviousness,
(2) to validate the performance of TURM, and (3) to evaluate the effectiveness of using a rock
crib as a temperature moderating device.
2. Methods
2.1. Field measurements
Data were collected at several sites from May 28th to September 30th, 2000 (Figure 1).
Stream discharge data was collected in the Token Creek subwatershed at 6 locations, temperature
data at 11 locations, and rainfall data at 6 locations. The University of Wisconsin Geology
Department collected stream flow and temperature data at two locations. A weather station
located at Shonas Heights recorded the following measurements: wind speed with cup
anemometer, solar radiation with a silicon cell pyranometer, rain and air temperature with
thermocouple wires, relative humidity with a humidity probe, and rainfall with a tipping bucket
rain gauge. The flow and rainfall data was collected every 5 minutes and the temperature every
15 minutes. The data was summarized for seven rainfall events in four-hour intervals. Interflow
and groundwater discharge, as a base flow, is an important source of cool water for streams. The
average base flow temperatures measured in the study area ranged from approximately 9 to 10 °
C and remained nearly constant during the entire summer season. The exact study area is
described in Table 1 below.
Token Creek Subwatershed illustrated in Figure 1 and 2 was selected to study the impact
that imperviousness has on stream temperatures. This subwatershed extends west from Sun
Prairie to Cherokee Marsh, and north to the Dane County line, and contains naturally occurring
springs as well as urban, agricultural, and naturally vegetated areas, encompassing an area of
22.2 square miles (14,212 acres). The storm water in this subwatershed is discharged to streams
or stream segments that are classified as either existing or proposed cold water communities by
the Wisconsin Department of Natural Resources and thus are more susceptible to thermal
impacts than other streams. Token Creek is a major contributor of fresh water to Lake Mendota,
with a base flow of about 22.21 cubic feet per second (cfs) during July 2000 (data collected by
the USGS); contributing 93% of all stream flow the lake receives (A Water Resources Study,
U.W, 1997).
371
-------
Table 1. Site descriptions and measurements.
Site
Culver Springs
St Albert Pond
Shonas Upstream Pond
Shonas East
Subdivision (Figure 3)
Shonas West
Subdivision (Figure 3)
Rock Crib (Figure 3)
Token Branch
Stonehaven
Highway C
Highway 51
Description
Naturally spring- fed area
Urban area with 20% imperviousness
Urban area with 20% imperviousness
Urban development
Urban area with 35% imperviousness
Rock chamber of 255 m3
Urban drainage area
Natural grass area
Confluence of streams
Confluence of streams
Measurement
Base flow
Runoff temperature
Runoff temperature
Runoff temperature
Runoff temperature
Base flow/Runoff
temperature
Base flow/Runoff
temperature
Runoff temperature
Base flow/runoff
temperature
Base flow/runoff
temperature
The thermal impact on Token Creek was measured over 7 rainfall events that occurred
between June 1st and August 5th 2000. Table 2 and Figure 4 present the summary of runoff
temperatures for the study area in 4-hour increments (0, 4, 8, 12, 16, 20). Table 2 and Figure 4
show that the runoff temperature from this area is consistently above the threshold for many
cold-water species. Other impervious areas, such as Shonas Upstream Pond, Shonas East, and
Shonas West, also had temperatures that were consistently above the threshold, but showed a
period between the hours of 4 and 8 when the temperatures were lower and suitable for cold-
water species. The total thermal impact on the Token Creek sub watershed is clear when the
temperatures observed at Culver Springs (-10° C) are compared with those observed at Highway
C. Highway C has temperatures that are approximately 7 to 8° C higher than those at Culver
Springs (Table 2). The cooling effect that Culver Springs has on Token Creek can be seen at
Highway 51, where the temperatures measured remained within the optimum temperature for
trout
372
-------
Figure 1 Data Collection Locations in the Token Creek Subwatershed
Explanation
A Rainfall Gage
[I] Flow Gage
0 Temperature Gage
£ UW Temperature Gage
S UWWsather Station
Road
Drainage Network
Stream or Ditcti
1 5 Mile;
Data 5QLTC9S.
Raintell. llort. and tarn perature gage inforrnabon tmrn U:".c-s (4-200 G)
Subwatershed boundary and drainage network, dsnved from a
10m DEM by Dane Co LCD O'MS) SireBm^drtcti inlomralion Irom
i:24.ooo-S':aie WDNR data
373
-------
Figure 2. Land Use in the Token Creek Subwatersned
Explanation
Cropland
Pasture
Woodland
Wetland
Water
Building
Low Density Residential
Medium Density Residential
High Density Residents i
Commercial
Institutional
Road
O.i
Miles
Data Sources
Land use data tram — MRCS wellands. digitalorthophuto 11995),
Regional Planning CCTTImission land use data (19901, Dane Co
LCD term tract and iteid data [1998), WDNR i:24,cjoo-scaie
hydrography data, Cit^orSun Praine land use data f2QOO)
and lieldvenficabco (2000).
374
-------
Figure 3, Detail of the TURM Model Testing Site
Collection sire for Flo
Temperature Calibration
. "''
Drainage Area
(Shonas East Subdr^ision)
Pw1lbSl>lin>tr
0 300
900 Feet
lniiOin
-------
2.2. TURM model in brief
To estimate the thermal impacts of the study, TURM was used for urban sewer sheds.
This model accounts for the fact that storm water not only absorbs heat from impervious surfaces,
but that it also cools these surfaces, reducing the ability of the impervious surface to heat runoff
from additional rainfall. Other model considerations include: the amount and temperature of
impervious surfaces, the ambient air temperature, the gain or loss of heat from the passage of
water through swales, detention basins, and streams, the gain or loss of heat due to tree canopy,
the heat loss due to evaporation, heat loss due to heat exchange in rock cribs, and the time and
duration of storm events. In addition, the model accounts for the time difference between the
runoff from impervious surfaces (TCimperv) and from vegetated areas (TCveget). However, TURM
does not account for the inherited variability of rainfall due to changes in intensity and the type
of storm, as the model assumes that the rainfall is uniform over the entire duration of the event.
The specific theoretical developments of TURM are listed as follows:
1) The convective transfer coefficient from Raney and Mihara (1974) was inappropriate for
use in TURM, and under-estimated the heat lost to the air. The equation from Ryan and
Harleman (1973) seems more appropriate.
2) Equations were developed to estimate the temperature of pavement on a clear day, before
rain falls. This simple model formulation for estimating the difference between the
surface pavement temperature and air temperature produces reasonable results when
compared with field measurements.
3) The inclusion of air and rainfall temperature as inputs into the model indicates that wet
bulb temperature during the rainfall period is a reasonable approximation of raindrop
temperatures.
4) A routine was developed in the model to account for the cooling effect of dry and water
fill rock crib.
3. Measurements
The thermal analysis data from TURM indicates that storm-water runoff from Token Creek
subwatershed's impervious areas can increase the temperature of the stream. Furthermore,
depending on the time of day that the rainfall event occurred, the impact of runoff on the
receiving waters can cause the stream temperature to reach lethal levels (Figure 4).
The model results are based solely on the data obtained between June 1st and August 15th,
although the model is capable of producing year-round results. These months were chosen
because, historically, they produce the largest amount of rainfall and the highest temperatures of
the year. Runoff volumes for urban areas were calculated using the rainfall data collected by the
University of Wisconsin's Soil Science Weather Station, the Wisconsin Department of Natural
Resources, and the United States Geological Survey. The runoff was measured as a continuous
stream flow in cubic feet-per-second at six gauging stations, (Figure 1), while the flow rate at the
outfall of each sewershed was determined by the ratio of rainfall depth to the individual land use
(curve numbers).
376
-------
Table 2. Runoff Tern
Site
Time
St Albert Pond
Shonas Upstream
Pond
Shonas East
Subdivision
Shonas West
Subdivision
Token Branch Up
After Rock Crib
Stonehaven
Grassed Area
Highway C
Culver Springs
Highway 51
Rainfall (inches)
perature Summary (June 1st -August 15th, 2000)
Time (In 4 Hour Intervals)
0
19.51
17.84
18.43
18.25
16.60
17.45
15.06
17.45
10.02
17.12
0.88
4
18.34
16.66
17.82
17.39
15.80
16.09
14.80
16.76
9.89
16.44
2.40
8
18.11
17.08
17.79
17.16
15.24
15.38
15.22
16.26
9.89
15.49
0.22
12
18.95
18.75
18.12
17.82
15.53
15.52
16.30
16.73
9.93
15.12
1.94
16
20.91
20.70
18.49
18.67
16.16
16.50
16.62
17.62
10.07
16.15
2.49
20
21.58
20.81
19.48
19.40
17.85
17.30
17.91
19.18
10.18
17.57
2.67
The data presented in tables 2 through 5 and displayed in Figures 4 through 10 represents
a summary of the field data collected by the USGS and WDNR Fisheries Department during
June, July, and August 2000.
Figure 4. Average Summer Runoff Temperature
(June 1st to August 15th, 2001)
26.00 '
24.00 '
22.00 '
°_ 20.00 "
= 18.00-
E
a> 16.00 '
o> 14.00 '
12.00
10.00
8.00 "I
» — .
/
4
0
f
A
A
4
=P=
\ J
\ /
v
¥
/
3£
_/-*—
/
8 12
Time (hours)
^
s*r~
JR"
16
—
. — •
^
20
o
-2.5
"w"
" 2 Jj
o
_c
- 1.5 -
1
1 K
- 0.5
' 0
>— St Albert Pond
* Shonas East
Subdivison
— *— HyC
Culver Springs
Hy51
* Rainfall (inches)
377
-------
3.1. Measurements of June 4, 2000
The 0.37 inches of rain fell in the early morning hours of June 4th. During this event, a
temperature difference of approximately 3° C was observed between the non-urban (Stonehaven,
Token Branch Up) and urban areas at the onset of the rain. The urban areas measured, on
average, 6° C higher than Culver Springs, or twice the temperature increase measured at the non-
urban areas. Because this rainfall occurred in the early morning hours, the impervious areas had
not yet been warmed by the sun, resulting in a lower than average high water temperature for
urban areas. The high temperatures for the urban and non-urban areas for this event were very
similar, separated by -2° C.
The rain began at hour 4 and ended at hour 16 with 0.1 inches of rain. The runoff
temperatures for the entire study area declined throughout the storm event and were compared to
the air temperature, rainfall temperature, and the solar radiation data (measured at the weather
station). The resulting temperatures were the lowest recorded during the entire study period
(Table 3 and Figure 5). This data relates that thermal impact is closely related to the weather
conditions and the time of day when the rainfall event occurred.
The thermal impact of the runoff from the Shonas urban area on the crib was minimal, measuring
only 2.5° C. This is significant because the temperature of the runoff after the crib represents the
temperature of the ground water during and after the rainfall event (Figure 5). Due to the low
intensity and long duration of this storm event, the runoff from the Shonas urban areas did not
overwhelm the heat exchange capacity of the rock or the cooling effect of the water in the rock
crib. Thus, the temperature of the runoff after crib was within 1° C of the temperature of the
stream at Flighway 51 (table 3).
Table 3. Temperature and rainfall data summary for June 04, 2000
Site
Shonas St. Albert
Shonas Upstream
Pond
Shonas East
Subdivision
Shonas West
Subdivision
Token Branch Up
After Rock Crib
Stonehaven
Ffighway C
Culver Springs
Ffighway 51
Rainfall (Inches)
Time (Hour)
Date
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
04-Jun
0
16.63
16.53
15.64
16.03
16.08
17.69
13.79
10.01
16.94
0.37
4
15.00
15.36
15.36
16.02
14.77
15.00
13.43
9.83
15.77
0.01
8
14.84
14.22
14.82
14.68
13.78
13.72
13.31
9.73
14.43
0.1
12
14.80
14.02
14.39
14.79
13.18
13.08
13.09
9.73
13.71
0.09
16
14.47
14.34
14.22
15.37
13.01
12.72
12.89
9.76
13.38
0
20
15.48
15.04
13.85
15.83
12.98
12.26
13.30
9.73
13.00
0
378
-------
Figure 5. Temperature and Rainfall Summary for 6-04-00
20.00
o 18.00 <
£ 16.00-
3
^
flj 14.00
0
12.00
o -i n nn -
|_ I U.UU
V
\
"^
N,— __
^H
7
^^
•- — i
^^
—- — «
k
X
^^
: q
u.^
- 0.35 ~
-0.3 ®
> 0.25 |
.2 _
rO.15 £
- 0.1 -5
-------
Table 4. Temperature and Rainfall Data Summary for July 10, 2000
Site
Shonas St. Albert
Shonas Upstream
Pond
Shonas East
Subdivision
Shonas West
Subdivision
Token Branch Up
After Rock Crib
Stonehaven
Highway C
Culver Springs
Highway 51
Rainfall (Inches)
Time (Hour)
Date
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
10-Jul
0
23.98
21.87
20.76
20.86
20.08
19.28
16.75
20.14
19.25
10.22
0
4
22.59
20.04
20.55
19.53
18.96
17.39
16.05
19.07
18.86
10.02
0.75
8
23.36
21.84
21.55
21.96
18.48
16.79
19.18
18.66
17.16
10.11
0.02
12
23.72
22.35
23.14
22.39
19.78
17.42
22.09
20.06
15.79
10.13
0
16
27.15
24.99
24.19
21.42
20.29
20.56
22.22
20.48
18.08
10.48
0
20
26.03
24.43
24.13
21.83
21.93
21.41
21.28
21.88
20.33
10.51
0
Figure 6. Temperature and Rainfall Summary for 7-10-00
0.80
0
8
12
16
Time (Hours)
20
"St Albert Pond
Shonas East
Subdivison
Shonas West Sub
Division
After Rock Crib
'HyC
Culver Springs
Hy51
"Rainfall (inches)
0.00
380
-------
3.3. Comparison of model results with measurements at St. Albert Shonas
Figure 7 presents the average temperature of runoff that drained into St. Albert's dry
detention pond. The measured temperature averaged 19.54° C, 0.42° C greater than the modeled
temperature, 19.12° C. The model, represented in Figure 7, over-predicts the initial runoff
temperature by -2° C because the model assumes that runoff is produced immediately after the
rainfall event starts. However, some rainfall evaporates when it meets the pavement, while some
is stored in the micro-depressions present in pavement and other impervious surfaces. Because
of this assumption, this immediate runoff has the highest modeled temperature for the event. At
mid-day (between hours 12 and 16), pavement temperatures in urban areas are often
considerably higher than air temperatures, and TURM requires an initial temperature before an
estimate of runoff temperature can be made. To correct this problem, maximum air temperature,
minimum air temperature, and mean wind speed at midday hours were used to solve for the
temperature difference. The approach for this simple set of equations is to solve for the
temperature difference between the black top surface and the air at hour 17, when the conduction
flux into the pavement is zero, and also at hour 5 when surface heat conduction equals zero.
Figure 7. Measured and predicted runoff temperatures at St Albert Shonas
25
24
^ 23
O
0 00 -
^ 22
UJ
(£ 04 -
D 21
< 20
UJ
1 19-
UJ ._ .
i- 18
17
16
M*
* <
15 1
0
**U4AU4
>***
> * .
» * '
AA*44it*t
Tpmrv
***
4AAAA4
irature M
Station
* * * '
Temperature Modeled
a
t Station
#8
easured <
#8
\
>".
f
/
it
. A A A <
' * J.
* * '
mmmmmmmmto
L
2345678
Time (hours)
381
-------
Shonas Pond The temperatures predicted for the St. Albert Shonas areas were assumed to be the
initial temperature of the runoff when it was delivered to the pond. The average runoff
temperature over the duration of the rainfall and the duration of the effective runoff was
measured at 20.27° C, while the model predicted a temperature of 20.76° C, a difference of 0.5°
C. However, the actual runoff temperature was influenced by the storage water in the pond,
which was assumed to be equal to the air temperature prior to the rain, which may account for
some of the difference in the results.
Shonas East. The runoff from Shonas East is delivered to the rock crib by a rock-lined
channel. The rock-lined channel's time of concentration is related to the voids of the channel
bottom, the interception by the rock, and the roughness coefficient of the rocks that impede the
free flow of the runoff. These characteristics reduce the inherited variability of the rainfall,
stabilizing the time of concentration and allowing the model to more accurately predict the
temperature of the runoff. The rock temperature was assumed to be the same as the air
temperature when the rain began to fall, while the initial temperature of the runoff was assumed
to be the same temperature of the pond. The temperatures changed according to the relationship
developed in Figure 8. The heat exchange in the rock channel was modeled to occur initially at
air temperature (before the rain fell) and the heated runoff from the impervious areas and was
calculated utilizing the model for dry rock basins. The predicted runoff temperature (19.25° C),
proved to be very close (within 0.27° C) to the field data (19.52° C).
Figure 8. Measured and predicted runoff temperatures at Shonas East
or n
ZO.U
"IA r\ -
z4.U
OQ n -
zo.U
*""• oo n -
o £.£.\j
o
*""" o-i n _
UJ £. \ .1)
tt
^ 20.0 -
DC -IQ n -
tu iy-u
Q.
S 180-
LJj IO'U
h-
4 ~7 n -
\ 1 .1)
•ic n -
Ib.U
•i K n -
hUUi
.l.l.i.1.1.
Mea
il»*i^
IMIs>
IVIO
sured Temperatu
at Station #7
u>
X
A
deled Te
Stati
«LULL
mperatu
on #7
re
^\
*^MT
/
re at / ~
^
^\
/
/
^k ^
1 1 1
lO.U n
0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
Time (Hours)
382
-------
Shonas West. The runoff was delivered to the rock crib directly by a 24-inch concrete pipe. The
temperature of the runoff from Shonas West (12 acres and 40% impervious) was modeled as a
direct heat transfer from the impervious areas to the runoff. Figure 9 presents the average runoff
temperatures from the impervious areas that drain to the outfall at Shonas West. The average,
calculated over 12 hours, was measured at 21.25° C and modeled at 19.82° C, for a difference of
1.4° C over the entire runoff event. The model, represented in Figure 9, shows that the model
over-predicted the initial runoff temperature by 1.5 ° C. As previously discussed, the model
assumes immediate runoff and does not allow for the delays that occur in the field. TURM also
does not account for the inherited variability of the rainfall due to changes in intensity and the
type of storm; rather, it assumes a uniform rainfall over the entire duration. This relationship is
shown in Figure 9.
Figure 9. Measured and predicted runoff temperatures at Shonas West
O
LU
OL
OL
LU
Q.
s
UJ
25
24
23
22
21
20
19
18
17
16
15
14
Temperature Measured
at Station #6
Temperature Modeled at
Station #6
4 5
Time (hours)
In a study conducted by Steve Greb of the WDNR in 1996, measurements were made of
the runoff flow and temperature, and pavement temperature from a parking lot in the City of
Madison, Wisconsin. The weather conditions during this study showed 0.43 mm of rain, wind
speed of 3.3 m/s, air temperature of 26° C, relative humidity of 92%, as well as the temperature
of the pavement and rooftops (40° C and 50° C). The rainfall lasted 39 minutes. The results
from the study proved to be very encouraging. The measured runoff outflow temperature
averaged 29.3° C. Using the TURM, the predicted the temperature of the runoff was 29.4° C,
indicating a difference between measured and predicted temperatures of only 1° C.
383
-------
3.4. Runoff Thermal Regime Best Management Practice (BMP)
The thermal impact of the impervious areas on stream temperature for the St. Albert
events was moderated by rock crib and by the base flow from Culver Springs, resulting in little
change in stream temperature. Overall, after the rainfall event began, the temperatures predicted
by TURM remained within 1° C of the actual measured temperatures for this event (Figures 7 to
10).
The rock crib was monitored for flow and temperature at the two inlets that drain into the
crib, as well as 50 feet below the crib. The crib was built with the assumption that water that
runs off impervious areas could be cooled by passing it through an underground rock chamber.
From initial calculations, if the crib is empty, the conduction of heat from the rock limits heat
transfer to the water, rather than the convective transfer coefficient of the moving water. As a
result, the problem is one of transient heat conduction from spheres. If the space between the
rocks is filled with water, the heat exchange in the crib is one of mixing, displacement, and the
convective transfer coefficient of the moving water. Unfortunately, no analytical solution to this
transient heat conduction is available, so a numerical solution was used.
The 255 m3 rock crib received runoff from Shonas East (140 acres) and Shonas West (12
Acres). The runoff flowed through paving blocks (25% porous) on the surface to an opening
filled with pea gravel, where the runoff was filtered into the rock crib, which is filled with
ground water and stone. The temperature of the rock in the crib was assumed to be the same as
the ground water temperature (15° C). The runoff (initially 30° C, enters and filters into the
ground at the moment of the rain, with the temperature of the runoff (measured at each time step)
changing according to the relationship developed in Figure 10. As previously discussed, the
temperature of rain and the air are closely related, and the runoff temperature depends on the
heat exchange with the stone in the riprap channel. The model has two heat exchange processes:
the initial exchange between the heated runoff and the stone, followed by the heat exchange
caused by the mixing of ground water with the runoff. The effectiveness of the rock crib
depends on the ratio of the volume of runoff and the volume of the rock crib, as well as the
volume of water stored in the crib prior to the rainfall event. In this case, the crib was filled to
capacity (9,000 ft3), resulting in an effective thermal treatment for the 140 acres of urban
development it drained. The field data shows that the rock crib mitigates the thermal impact
caused by impervious areas until the initial volume of the crib has been completely replaced by
the runoff. After the volume has been replaced, the rock crib no longer provides a thermal
reduction for stormwater.
384
-------
Figure 10. Measured and Predicted Runoff Temperatures at the Rock Crib
O
o
LU
£
LU
Q.
^
LLI
30.00
28.00
26.00
24.00
22.00
20.00
16.00
14.00
measured runoff temp no BMP
measured runoff temp with BMP
modeled runoff temp with BMP
Measured Temperature
at Point #3
\l
Runoff Temperature from
Roofs and Streets
Modeled Temperature at
Point #3
2345
Time (hours)
8
Table 5 shows a summary of the measured temperatures (recorded every 15 minutes) and
the modeled temperatures (predicted every 5 minutes). The thermal impact was modeled for a
14-hour period because runoff and heat exchange continue after the rain has stopped. When
runoff is directly discharged into an outfall from an impervious surface, the model does not
account for the variability in rainfall intensity changes. However, when the runoff is delivered to
a BMP, such as a pond, grassed channel, or rock crib, the variability of the attenuation factor is
reduced because BMPs temporarily store runoff before releasing it at a preset rate to a receiving
streams.
385
-------
Table 5. Measured vs. Modeled Runoff Temperatures at Selected Sites
(15-Minute Intervals)
Time of
Rainfall Event
12:00
12:15
12:30
12:45
13:00
13:15
13:30
13:45
14:00
14:15
14:30
14:45
15:00
15:15
15:30
15:45
16:00
16:15
16:30
16:45
17:00
17:15
17:30
17:45
18:00
18:15
18:30
18:45
19:00
19:15
19:30
19:45
20:00
Average
Standard
Deviation
Measured
Temperature
Albert Shonas
(Station #8)
(°C)
18.32
18.16
18.32
18.32
18.48
20.12
20.28
19.79
19.95
19.95
19.79
19.79
19.63
20.12
19.63
19.63
19.79
19.95
20.12
20.28
20.28
20.12
20.12
20.12
20.12
20.12
20.12
19.95
19.79
19.95
19.79
19.95
20.12
19.54
Calculated
Temperature
at St. Albert
Shonas
(Station #8)
(°C)
19.61
19.54
19.48
19.42
19.37
19.32
19.27
19.23
19.19
19.17
19.16
19.14
19.13
19.12
19.10
19.09
19.08
19.07
19.06
19.05
19.05
19.04
19.03
19.02
19.02
19.01
19.00
19.00
18.99
18.99
18.98
18.98
18.97
19.12
0.90
Measured
Temperature
at Shonas
East
(Station #7)
(°C)
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.02
18.18
18.51
18.99
19.16
19.81
20.29
20.46
20.62
20.62
20.62
20.62
20.62
20.62
20.62
20.46
20.46
20.46
20.46
20.29
20.29
19.41
Calculated
Temperature
at Shonas
East
(Station #7)
(°C)
18.0C
18.0C
18.0C
18.0C
18.0C
18.0C
i8.oe
18.2E
18.42
19.0-
19.45
19.76
20.41
20.4S
20.45
20.47
20.45
20.4-
20.42
20.41
20.3S
20.35
20.3e
20.35
20.3:
20.32
20.3C
20.2S
20.27
20.3:
20.26
20.2E
20.2E
19.70
0.73
Measured
Temperature
at Shonas
West
(Station #6)
(°C)
19.55
19.38
19.71
21.84
22.01
22.01
22.01
21.84
21.34
21.84
22.01
21.84
21.67
21.34
20.36
20.03
20.52
20.68
21.01
20.68
20.52
20.68
21.01
21.34
21.67
21.84
21.84
21.84
21.51
21.34
21.51
21.67
21.67
21.25
Calculated
Temperature
at Shonas
West
(Station #6)
(°C)
21.05
20.88
20.68
20.57
20.42
20.34
20.25
20.14
20.03
19.96
19.86
19.8C
19.75
19.72
19.71
19.69
19.66
19.64
19.61
19.6C
19.58
19.57
19.55
19.54
19.52
19.52
19.5C
19.4S
19.48
19.52
19.47
19.46
19.46
19.82
1.90
Measured
Temperature
at Rock Crib
Shonas
(Station #3)
(°C)
15.35
15.51
15.51
15.35
15.35
15.35
15.51
15.35
15.35
15.35
15.35
15.35
18.71
18.71
18.87
19.03
19.36
19.84
20.17
20.17
20.33
20.33
20.33
20.33
20.33
20.17
20.17
20.01
20.01
20.01
19.84
19.84
19.68
18.40
Calculated
Temperature
at Rock Crib
Shonas
(Station #3)
(°C)
15.5C
15.5C
15.5C
15.5C
15.5C
15.5C
15.5C
15.6C
15.7C
15.8C
15.90
17.33
18.59
19.67
19.66
19.66
19.65
19.64
19.63
19.63
19.62
19.61
19.6C
19.6C
19.59
19.59
19.58
19.57
19.57
19.56
19.56
19.55
19.55
18.33
.57
386
-------
Figure 11. Daily Measured Temperature in Token Creek
(June 2000) Letha, to
6.00
Trout become
Stressed
Optimum Temperature
"Culvers Spring
'ST. Albert
Shonas.
"Rock Crib
HwayC
"Hwy51
6/1 6/4 6/7 6/10 6/13 6/16 6/19 6/22 6/25 6/28
Date
7/1
Figure 12. Daily Measured Temperature in Token Creek
(July 2000)
30.00- x ' '
o
o
I
+J
5
a
E
a
28.00
26.
Trout become Stressed
6.00
7/1 7/4 7/7 7/10 7/13 7/16 7/19 7/22 7/25 7/28 7/31
Date
'St Albert
Shonas
Rock Crib
Hyway C
Culver
Springs
'Hyway 51
387
-------
4. Conclusions
In this study, we have clearly illustrated that the existing urban development in the Token
Creek sub watershed causes an increase in runoff temperatures. Further, the increases recorded at
Highway 51 suggest that these increases cause a permanent rise in stream temperature.
The daily temperatures of Token Creek are presented in Figures 11 and 12 for the months
of June and July. During June, the temperature of Token Creek did not increase above the lethal
limit (26° C, but did rise above 19° C at several points, including St. Albert Pond and St. Albert
Shonas, as well as at the major points of confluence in Token Creek Subwatershed (Ffighway C
and Ffighway 51). The runoff temperatures at the outfall of St. Albert Pond were directly
impacted by the heat exchange between the black top and the rainfall. The temperatures
recorded during June continued to rise after each rainfall, reaching levels above the threshold for
trout (19° C), while temperatures recorded during July often reached the lethal threshold (26° C).
The cumulative temperature increase caused by new developments will have a profoundly
negative impact on the sensitive cold-water community in Token Creek if provisions to reduce
the thermal impact are not implemented. In order to predict the conduction of heat to the runoff,
TURM takes into account two factors:
1) Time of concentration. In the present model, the runoff from impervious areas is
delivered instantaneously to the conveyance system. However, due to micro-
depressions and evaporation, peak runoff flow does not begin instantaneously.
2) A correction in the convective transfer coefficient, resulting in less heat being lost to
the air. The result is that the pavement heats rainwater up more and during longer
periods, which is the case for runoff in impervious surfaces in urban settings.
TURM was validated successfully, predicting temperatures within 1° C of the actual
temperatures recorded. The standard deviation was less than one, and significant at the 1% level
for all sites when the field data was compared having the same mean (|j,i=|j,2). When used for
estimating the difference between pavement and air temperatures, TURM produces reasonable
results compared to the field data collected by USGS (pavement temperatures during this
research were from 10 to 20° C above air temperature at midday). The results indicate that rain
and air temperature are very closely related, a unique finding as little, if any, data has been
published previously on the subject. Due to this correlation, it is possible to have an analytical
solution based on the atmospheric variables that were incorporated into TURM.
Thermal impact analysis accounts for the impact that impervious areas have on stream
temperatures. These impervious areas are generally associated with urban development and are a
major source of thermal pollution in cold climates, not only because they remove water's ability
to infiltrate into the soil, increasing the quantity of runoff, but because they store heat. As
rainfall passes over impervious areas such as rooftops, roadways, and parking lots, it absorbs a
portion of the energy stored in the surface. Cumulatively, the rise in runoff temperature causes
an increase in the temperature of the stream, degrading the habitat and the diversity of the
stream. To reduce the thermal impacts on streams, effective Storm Water Best Management
Practices should be used. Some examples of BMPs include: rain gardens, rock catchment basins,
swales, deep tilling, constructed wetlands, reforestation, and buffer strips
388
-------
The model for the rock crib indicates that cooling can be obtained from rock storage for a
limited time and that the cooling depends on the size of the crib. Because TURM is based on an
analytical solution, we can extend the model to other temperature reduction devices, such as
detention ponds, dry ponds, deep rock trenches, drain tiles filled with pea gravel, grassed swales,
and green areas. These practices and many others can be utilized to reduce runoff temperatures;
however, any device selected should be integrated as part of a storm water management plan.
As a result of this study, it is clear that municipalities and developers alike should
implement a system of BMPs to reduce the impact that impervious areas have on lakes, rivers,
streams, and wetlands. Rock cribs, which are a relatively new thermal reduction device
developed by the Dane County Land Conservation Department, proved to be very successful at
reducing runoff temperatures during this study. In addition, they are a practical, attractive option
for new developments and can be used to augment existing thermal reduction systems. The use
of BMPs is critical to reduce the thermal impacts caused by urban areas and imperviousness and
to ensure the future diversity and health of aquatic ecosystems.
Acknowledgements: This work was undertaken with support from the Wisconsin Department of
Natural Resources (WDNR), Kevin Cornnors and the Dane County Land Conservation
Department, and the University of Wisconsin-Madison.
5. References
Aida, M. 1982a. Urban albedo as a function of the urban structure-A model experiment.
Boundary Layer Meteorol. 23, 405-413.
Campbell, G.S. and J.M. Norman. 1998. Introduction to environmental biophysics. Springer
VerlagPubl. Co., NY.
Galli and Robert Dubose, 1990. Water temperature and Fresh water Stream Biota. An Overview.
Dept. of Environmental Programs. Metro. Wash. Council of Governments.
David, Marshall, 2001. DNR Biologist. Personal communication.
Lyons, J. L. Wang and T .D. Simonson. 1996. Development and validation of and index of
biotic integrity for cold water streams in Wisconsin. North America Journal of Fisheries
management 16:241-356.
Norman, J.M. 1967. An application of harmonic analysis to an in-site determination of soil
thermal diffusivity. M.S. Thesis, University of Mnn., St. Paul, MN
Norman, J.M and A. Roa-Espinosa. 2001. Modeling the effect of the natural environment and
urban runoff on stream temperature.
Raney, F.C. and Y. Mhara. 1974. Water and soil temperature. In Irrigation of Agricultural
Lands, eds. R.M. Hagen, H.R. Haise and T.W. Edminster. No. 11 in Agronomy Series,
Amer. Soc. Agron. Publ., Madison, WI. pp.1024 - 1036
Ryan, P.J. and D.R.F. Harleman. 1973. An analytical and experimental study of transient
cooling pond behavior. M.I.T. Dept. of Civil engr., Ralph M. Parsons Lab, Tech. Rept.
No. 161, Cambridge, MA.
389
-------
RAIN BARRELS - TRUTH OR CONSEQUENCES
Karen Sands, AICP and Thomas Chapman, P.E.
Milwaukee Metropolitan Sewerage District
Milwaukee, Wisconsin
Abstract
Rain barrels are a centuries-old technique used by many cultures to collect rainwater from rooftops for later
use or consumption. Recently, rain barrels have become popular in parts of the United States and Canada for
a variety of uses, particularly among "green" proponents. Their uses may include garden and lawn watering
(particularly during drought conditions), and even possible combined sewer overflow volume (CSO)
reduction. In addition to the logistical and cost issues surrounding the use of rain barrels, proponents boast
the right to a "free" resource on the grounds of environmental ethics. Although there are many potential
benefits, there are a number of factors that sponsoring agencies must consider before embarking on a rain
barrel program. The Milwaukee Metropolitan Sewerage District (MMSD) will begin advocating for and
implementing such a rain barrel program. While considering whether to pursue rain barrels, staff worked to
quantify the potential benefits to CSO reduction. No CSO volume reductions were demonstrated, but
treatment cost reductions were realized. Therefore, a program to subsidize, distribute, and educate people
about how to use rain barrels will be crafted and launched by the MMSD.
This paper describes rain barrels and how they work. It also explores how well rain barrels perform against
some of the benefit assumptions, including water quality issues not generally discussed in this context. This
paper also compiles a list of assumptions and suggestions for their use that encompasses barrel size and
shape, key barrel features, climate considerations, algae and mosquito control, and home foundation
protection, among other issues. It will explore these issues from a neutral position, with the end result being
a recommendation for or against the use of rain barrels in the MMSD service area—a recommendation that
may be applied by other sewerage agencies looking to reduce treatment costs while also reaping other
environmental benefits.
Introduction
Rain barrels are on-site rainwater collection systems. Rainwater can be collected as a valuable resource for
lawn and garden watering, as well as possibly retained to reduce CSO volume and storm water management
costs. Implementing a rain barrel program first requires an evaluation of the potential to meet desired
results. The Milwaukee Metropolitan Sewerage District (MMSD) has studied the effectiveness and benefits
of such a program, and is in the early stages of crafting and eventually implementing it. Aspects of that
evaluation and factors to be considered during implementation are described below.
Program Function
There are a number of factors to consider before implementing a rainbarrel distribution program. These
include setting goals for the program, educating the user public about how to operate and care for
rainbarrels, and being realistic about the benefits. While a potential program for the Milwaukee region may
produce only modest results, there are side-benefits to be gained, such as educating people and getting them
involved in possibly reducing the volume of CSOs. The effectiveness of any potential program could be
enhanced through promoting an integrated management plan featuring compatible stormwater management
concepts, including things like downspout disconnections, green roofs, raingardens, and grassy swales. In
390
-------
fact, this is the direction to be taken by the MMSD. While some rainbarrel users also use the rainwater for
drinking, the MMSD will not likely recommend this use (see below).
Uses
Recently, rain barrels have become popular in parts of the United States and Canada for a variety of uses,
particularly among "green" proponents. Their uses may include garden and lawn watering (particularly
during drought conditions) or even CSO reduction. In addition to the practical and cost issues surrounding
their usage, rain barrel proponents boast the right to a "free" natural resource on the grounds of
environmental ethics.
Although there are many potential benefits, there are a number of factors sewerage agencies must consider
before embarking on a rain barrel program. These include water quality issues, climate considerations, algae
and mosquito control, physical site suitability, homeowner ability and willingness to operate effectively, and
home foundation protection.
Water Quality. Rainwater collected in a barrel can provide a relatively clean, safe, and reliable source of
water as long as the collection system is properly built and maintained. Rainwater that is to be used outside
to water lawns or gardens is typically not a water quality concern. The roof construction materials should
not be treated cedar shakes or materials containing asbestos. The gutter system should not have lead solder
or lead-based paint, and bird droppings should be cleaned from gutters and the roof as needed. Depending
on the location, an awareness of the dry deposition of pollutants from the air may also be warranted.
Overall, rain barrel water quality is not a major concern unless the water is intended to be consumed.
Filtration and disinfection would be necessary for consumption-based water use, and is beyond the scope of
this analysis.
Climate. Climate considerations apply particularly where temperatures regularly reach freezing during
winter months. Where this occurs, rainbarrels should be disconnected during winter months to ensure that
water in rainbarrels doesn't freeze and damage barrels and/or allow water to back up into downspouts or
overflow into building foundations. When rainbarrels are disconnected for winter months, they should be
stored upside down so they may fully drain and remain relatively clean. During this time, downspouts
should be reattached so that winter precipitation doesn't damage foundations. In the Milwaukee, Wisconsin
region, CSOs occur an average of 2.5 times per year. These have occurred overwhelmingly during non-
winter months and, when they do occur in winter, are typically due to mechanical malfunction. Therefore,
disconnecting rainbarrels in the winter will not likely reduce the effectiveness of rainbarrels as a CSO
volume reduction approach.
Algae. Algae are microscopic, photosynthetic plants. When exposed to sunlight, chlorophyll in algae
converts carbon dioxide (CO2) and water into glucose and oxygen (©2). Generally, algal growth in water is
influenced primarily by the amount of nutrients (phosphorus, nitrogen, carbon, etc.) in water, and
secondarily by the availability of light incident on the water. However, water temperature, water flow,
available substrate, and pH also influence the growth of algae.
The primary factor controlling algal growth--nutrient content in water--generally comes from leaves, lawn
clippings, fertilizer, pet waste, and non-contact cooling water that enter the water cycle after water is
discharged from a rainbarrel. It would follow, therefore, that the nutrient content of rainbarrel water is not
likely to be high, and may not be a large determinant in rainbarrel algae growth. Intuitively, there are
exceptions to this: (1) rainbarrels that collect runoff from a green roof or rooftop garden and (2) rain gutters
391
-------
that are filled with leaves, allowing rainwater to filter through. Raingardens remain rare in the Midwestern
U.S., and the problem of leaves in the gutter is easily avoided by periodic cleaning to reduce this primary
influence.
The secondary factor controlling algal growth is light. Light incident upon standing water in a rainbarrel is a
function of rainbarrel design. Rainbarrels with open or screened tops or that allow light to penetrate will
provide more light inside the barrels. Therefore, open and/or light colored rainbarrels would be more likely
to contribute to algal growth. On the other hand, rainbarrels with openings limited to the size of the
downspout or gutter tube would allow less light to reach water stored inside. Therefore, partially closed
and/or darker rainbarrels would be less likely to contribute to algal growth.
Other factors listed above include water temperature, flow, and pH. Water temperature may be relatively
high when rainbarrels are placed in full sun, thus increasing the risk of algal growth. Placing rainbarrels in
shade can reduce this risk. Flow is virtually nonexistent, thus further increasing the overall risk of algal
growth. Overall pH can be affected by roofing materials, and higher pH levels contribute to algal growth.
Rainwater typically has a slightly lower pH and, therefore, higher pH is not likely an issue. Further study of
this is suggested.
There are a number of factors, such as low nutrients, that tend to minimize algae growth. Other factors, such
as incident light and water temperature, can be managed to further minimize (but not eliminate) the potential
for algae growth. While algae is typically considered undesirable, small amounts of algae that may grow in
a rainbarrel may actually help to fertilize gardens and lawns. Given that some causal factors are not
favorable and that others may be minimized, algae growth in rainbarrels can be kept in check by selection of
barrel characteristics that limit algal growth and proper barrel placement.
Mosquitoes. West Nile virus is increasingly becoming a concern in the Midwest, as an increasing number
of illnesses and deaths are blamed on the virus. Mosquitoes tend to breed in wet areas, and the Culex
mosquito that carries and transmits West Nile virus is found where there is decaying organic matter and wet
conditions. Recommendations to reduce populations of Culex mosquitoes include source reduction of
mosquito breeding sites and avoidance of biting mosquitoes. Recommendations for reducing breeding sites
include eliminating or emptying artificial water collection containers described as "prime breeding spots for
the mosquito species implicated in the transmission of West Nile Virus." (See:
http://www.cfe.cornell.edu/erap/). This potential connection between standing water breeding sites and
rainbarrels may have implications for rain barrel use. Mosquitoes can breed in as little as 10 days. In
rainbarrels that allow mosquitoes to enter, therefore, rainbarrels should be emptied in less than 10 days.
Another potential solution is to screen the rainwater inlet so mosquitoes don't enter in the first place. In
either case, user education is key to reducing the potential for Culex mosquito breeding sites.
Physical Site Suitability. Homeowners--rather than professionals—typically install rainbarrels, so it is very
important that any distribution program make homeowners aware of the risk to their home foundations.
Because water pooling near a foundation can eventually work its way into a home's basement, it's important
to make sure the collection system keeps water away from the foundation. This includes properly
channeling water from the inlet to the rainbarrel, provisions for rainbarrel overflow during larger storms,
and drip-free spouts and hose connections. This also involves instructions on how to reattach downspout
connections prior to winter months. With proper care, foundation and basement damage can be avoided.
392
-------
There are some situations where rainbarrels may not be appropriate. These include high-density urban
settings where there may not be a significant use for the collected water. Moreover, homes that are close
together may not have an adequate area to contain rainbarrel overflow. Such homes in Milwaukee are more
likely to be located within the combined sewer service area and, therefore, should be carefully evaluated,
particularly when disconnecting direct downspouts to the combined sewer. Finally, where homes are located
on smaller lots, there may be less opportunity for garden watering simply due to space constraints.
Homeowner Willingness and Ability. Proper care includes a willingness on the part of the homeowner to
periodically check to see that connections and fittings are in proper working order, empty the barrel after a
rainstorm (in advance of new rainstorms), remove the barrel and store it for winter, and reconnect the
downspout. Some homeowners may see this work as bothersome, and still others may not be physically
capable of performing the work. To have or not to have a rainbarrel is an individual decision. Incentive and
assistance programs could be developed to encourage rain barrel use and proper maintenance.
CSO Volume Reductions
The MMSD has responsibility for sewage conveyance and treatment as well as for flood management.
MMSD's sewerage system includes a regional collection/conveyance system and two wastewater treatment
plants. In the late 1970s and early 1980s, MMSD undertook a Water Pollution Abatement Program
(WPAP), which included over $2 billion in improvements to the conveyance system, treatment plants, and
an inline storage system known as the "deep tunnel." Together, projects from the WPAP virtually
eliminated separate sanitary overflows (SSOs) and reduced combined sewer overflows to an average of 2.5
times per year.
While the SSO and CSO goals of the WPAP were attained, the media and the public expect MMSD to
further reduce CSO volumes. With this in mind, MMSD conducted an evaluation of a program that would
utilize rain barrels in the combined sewer system area to reduce the volume of stormwater runoff. The study
assumed 40,000 single-family homes in the combined sewer service area. Each home was estimated to have
1,200 square feet of roof area that emptied into two 90-gallon rain barrels, each collecting rainwater from
600 square feet of roof. Homeowners were assumed to empty the rain barrels after each storm event and the
water would be released to infiltrate into the ground and not into the combined sewer system. An analysis
of the precipitation record from 1940 to 1997 showed the following results:
Numb er of events: 78.2
Mean Volume: 0.40 inch
Median Volume: 0.19 inch
Mean Duration: 15.1 hours
Median Duration: 9 hours
The distribution of the storm events show half of all events are 0.19 inch or less, but we found that these
events account for only 8.5 percent of the total rainfall volume. A 90-gallon rain barrel can hold 0.24 inch
of rainfall from a 600 square foot roof. The annual capture amount from the 40,000 residences using two
90-gallon barrels was calculated to be 243 million gallons. With proper disposal, this volume represents
water flow that would not need to be treated at the treatment plants. Most storm events that are 0.24 inch or
less do not typically result in a CSO event. In fact, these relatively small storms with low rainfall volume
are easily conveyed to the treatment plants. Even in a large storm the rain barrel volume collected in the
beginning of the storm would not reduce the volume of a CSO, which happens much later in the storm. The
study showed that an extensive rain barrel program would not have an impact on CSOs but that such a
393
-------
program could reduce costs at the treatment plants. Further research is needed to determine if rain barrels
used in conjunction with other on-lot treatments (rain gardens, storm water trees, boulevard swales, etc.)
could be integrated to decrease runoff volumes enough to reduce the volume of a CSOs. While none of
these other on-lot treatment programs may make a significant impact as a stand-alone solution, in
combination there would likely be a greater benefit.
Recommendations
An extensive rainbarrel distribution and use program may not provide reduction in CSO volumes, but would
save treatment costs at the plants. There are a number of considerations that program sponsors must take
into consideration before sponsoring a distribution program. These include:
0 A realistic understanding of the goals to be met
0 A public education program that includes the benefits, costs, and considerations of rainbarrels
0 The likely need to provide technical assistance to homeowners
Likewise, homeowners must take into consideration a number of factors before deciding whether to become
rainbarrel owners. These include:
0 An understanding of how to operate rainbarrels, including the need to drain them within a reasonable
period after a rainstorm
0 A physical ability and personal commitment to operating rainbarrels as recommended
Conclusion
There are a number of factors to consider before implementing a rainbarrel distribution program. These
include setting goals for the program, educating the user public about how to operate and care for
rainbarrels, and being realistic about the benefits. While a potential program for the Milwaukee region will
produce only modest benefits, there are additional benefits to be gained by getting people involved in
reducing treatment costs and by educating them in the process. And, the effectiveness of any potential
program could be enhanced through promoting an integrated management plan that also promotes
compatible stormwater management concepts, including things like green roofs, raingardens, storm water
trees and grassy swales.
394
-------
THE STRANGER AMONGST US:
URBAN RUNOFF, THE FORGOTTEN LOCAL WATER RESOURCE
Neal Shapiro
City of Santa Monica
Santa Monica, CA.
Abstract
Urban runoff is an ignored and misunderstood local water resource. As a pollution problem, runoff is the
single greatest source of water pollution in Southern California, specifically in the Santa Monica Bay and as
an ecological problem causes degradation of water quality and impairment of beneficial uses, threatening
the long-term health of marine ecosystems and local economies. As a water resource, capturing stormwater
for groundwater recharge can add a significant regional water supply, lowering the region's dependence
upon imported water, which causes ecological degradation and water supply disruption to distant
watersheds. The City of Santa Monica adopted a strategy to solve both problems: harvest stormwater, treat
it and infiltrate back into the ground, and keep a pollution source out of the Bay. The City's comprehensive
watershed-urban runoff management approach includes: (1) an ordinance to require the harvesting of
stormwater runoff from new development; (2) a philosophy of treating all dry weather and some wet
weather urban runoff leaving the City; (3) a first-of-its-kind innovative recycling facility for dry weather
runoff.
This runoff management approach allows for the development of a toolbox of innovative structural
solutions, best management practices (BMPs), which can be tailored for each site's specific land use
characteristics. A critical component of this successful toolbox is the unique management style: a shift
from the traditional stormwater management approach of plumbing land, paving it over to move the
maximum amount of runoff to receiving waters, to a low-impact site design approach of allowing the land to
work within nature's hydrologic cycle, maximizing permeability and runoff infiltration into the ground.
The City ordinance requires low-impact BMP designs in new developments. These design techniques
harvest precipitation and infiltrate it back into the ground, keeping urban runoff and its pollutants out of
receiving waters. Not only are water quality objectives improved and beneficial uses restored as runoff is
treated while passing across, through and into landscapes, but aquifers are recharged for future extraction.
The Santa Monica Urban Runoff Recycling Facility turns a perceived "waste" product into a natural
resource, a commodity, for reuse in landscape irrigation and indoor plumbing, and eliminates dry weather
runoff into the Bay. Secondary project goals include public outreach through urban runoff educational
exhibits at the facility, and strong artistic and architectural elements into a highly functional design and
community asset.
Introduction
Studies cite contaminated urban runoff as the greatest single source of water pollution in the country. This
non-point source urban runoff pollution problem in Southern California, specifically in the Santa Monica
Bay, is one such major ecological problem, threatening the long-term health of marine ecosystems and local
economies. The City of Santa Monica took a three-prong integrated management unique approach to this
problem:
395
-------
> Ground-breaking municipal ordinance to require the harvesting of stormwater runoff from new
development;
> City goal of treating with Best Management Practices (BMPs) urban runoff from new City
development and all urban runoff from its storm drain system before runoff leaves City boundaries;
and
> Construction of a year-round dry weather runoff facility to treat and reuse in place of imported
potable water urban runoff, the country's first dry weather urban runoff recycling facility.
The City redirected its approach to managing urban runoff from the traditional approach of moving runoff
as fast as possible from the City and into the Bay, to a watershed approach in which the land is viewed as
part of the hydrologic cycle and can absorb runoff for treatment and storage, keeping runoff out of the Bay.
Instead of disrupting the water cycle, the City objective is to work with nature. Figure 1 demonstrates this
approach, making a building and its surrounding hardscapes appear invisible to precipitation and runoff
through the placement of BMPs and site planning so that rain runoff goes back into the ground to the
maximum extent possible, instead of running off hardscapes into the street and water ways.
Figure 1. Making a building seem like it is not there in terms of precipitation and stormwater runoff to the land. On
right, existing building and its hardscapes collect rain and runoff, and direct them onto the street and into the Bay, the
Traditional Approach. On left, strategically-placed BMPs within the landscape receive runoff from the building for
infiltration, keeping runoff out of the street and giving the appearance to the land that the building is not there, the
Low-Impact Approach.
Studies (May, 1997; Schueler, 1995; Schueler, 1994) have shown that as impermeable surfaces increase,
replacing permeable surfaces, water quality decreases and impacts on aquatic flora and fauna increase, even
with as small as 5-10% increase of impermeable over permeable.
Many studies have documented the health risks and dangers to beach-users and aquatic habitats and life
from urban runoff. The Southern California Coastal Water Research Project, a leading marine research
group in Southern California, reported that storm water and urban runoff are the leading source of water
pollution in the Los Angeles area (Cone, 2000); storm water pollution has increased 200-700 percent during
the last 20 years. Stormwater has become a lethal cocktail of pollutants that now constitutes the single
greatest source of water pollutants, contributing 50-60 percent of the pollutant load. According to the US
EPA, urban stormwater is the largest source of water quality damage in estuaries, the second largest for
wetlands degradation, third largest impairment of lakes and fourth largest source of river damage (Mehta,
396
-------
2002; Sheppard, 2000; Coastal Alliance, 2000; Los Angeles County, 2000; American Oceans Campaign,
2000). An epidemiological study (Haile, 1996) showed that people who recreate near flowing storm drains
are much more likely to contract intestinal, ear, and nose illnesses. In light of numerous studies mentioned
above and with the passage of stricter regulations for urban runoff discharges, the City leadership believes
that all dry weather and some initial wet weather runoff leaving the City should receive some treatment to
remove pollutants of concern before entering the local receiving water body, the Santa Monica Bay. To
achieve this goal, the City has installed BMPs in many of its storm drain outlets and in catch basins within
the storm drainage system. The City has every expectation to have BMPs in all storm drain outlets in the
near future.
The purpose of this paper is to describe the City's urban runoff management program and some examples of
BMPs that have been implemented to reduce problems associated with urban runoff, namely water quality
and quantity issues. The City's program integrates the resources of many departments to comply with urban
runoff regulations and the City's Sustainable City Program. Instead of disconnecting staff, the program
seeks to connect personnel and goals to achieve success. The program is a hands-on, proactive and
watershed approach in which solutions seek to mimic nature, not disrupt it. Ultimately, the program seeks
to convert a perceived waste into a valuable resource and at the same time keep pollutants out of the Bay.
Santa Monica
Santa Monica is about 20.5 kilometers2 (8.1 miles2) in size with a residential population about 90,000. The
daytime population increases by more than double. The City is surrounded by the Pacific Ocean (Santa
Monica Bay) on the west, Santa Monica Mountains to the north, and cities of Los Angeles and Venice to the
east and south. Attractive beaches and the Santa Monica Pier, pleasant year-round climate and proximity to
attractions in Southern California make Santa Monica a popular destination. The City is completely built
out.
The City's urban runoff management program is strongly supported by a City Council and management
concerned about environmental stewardship and responsibility. To this end, the City enacted a Sustainable
City Program (Santa Monica, 1994) to promote sustainable practices, including the reduction of pollution
found in urban runoff. The Council has a history of political activism for environmental protection, which
is critical to a City that depends upon a healthy Bay to support a healthy economy.
Due to recent media reports about the dangers of urban runoff and impacts to beach-goers and aquatic life,
the City responded quickly and implemented many changes in how the City does business on a daily basis.
The rest of this paper describes the many programs to improve urban runoff quality and reduce runoff
quantity.
Source Control & Prevention
The best solution to pollution found in urban runoff is to prevent pollutants from coming in contact with
urban runoff, whether dry weather runoff or storm runoff. The pollutants of concern are familiar to us:
petroleum products from vehicular use, heavy metals from vehicle brakes, organic chemicals and fertilizers
(nutrients) from lawn care use, overwatering of landscapes, broken irrigation systems, sediments from
exposed land, detergents from cleaning hardscapes, and pathogens from pets, wild animals and transients.
Education
The City has printed materials that are distributed to residents and businesses, explaining the problems
associated with urban runoff and suggested solutions. People can obtain these materials from City offices,
397
-------
at community events, through the mail, at City-sponsored presentations, or from the City's web site. The
City also collaborates with other municipalities and regional groups to disseminate educational materials
through newsprint and radio.
Signage
The City maintains signage on all City catch basins, warning people not to dump materials into basins, and
providing a phone number to call in incidents of dumping. Unfortunately, a mix of materials, some
hazardous, still finds its way into catch basins and storm drains. Over the years, the City has used painted
stencils, ceramic tiles and thermoplastic stencils to alert people about dumping materials into the City's
storm drain system.
The City also maintains signage on the Pier, warning visitors not to dump materials over the side and into
the Bay, nor to feed the birds. Dumping materials over the side, such as food and fish guts, attracts birds,
and birds defecate into surrounding waters, adding pathogens.
Some City parks and pet walk parks contain dispensers with bags to clean up after pets for pet-owners who
forget to bring bags with them. A City ordinance requires that anyone walking a pet outdoors must have a
visible means of cleaning up after the animal.
The Santa Monica Urban Runoff Recycling Facility (discussed below) has numerous educational signs to
explain what urban runoff is, its causes, and solutions. The City has additional plans for signage at new
installations of BMPs so that people can learn more about runoff and how to prevent pollution.
Good House-Keeping Measures
The City's Urban Runoff Pollution Mitigation (Santa Monica, 2000) ordinance requires people in existing
buildings or at existing properties without new or redevelopment to take steps to prevent pollutants from
coming into contact with urban runoff. For example, people should clean up any spilled household
hazardous materials immediately. Lawn care chemicals should be used as per instructions and not overused,
nor applied before rain. Sprinkler systems should be properly maintained; any leaks should be repaired
immediately. Containers of chemicals and trash receptacles should not be left outside uncovered.
Construction BMPs
The Mitigation ordinance also requires construction sites to be well maintained. Responsible parties at a
construction site must take steps to prevent pollutants from coming into contact with urban runoff, and to
prevent erosion and the escape of polluted runoff and sediment from a site. As with Good House-Keeping
BMPs, containers of chemicals must not be left open and exposed to the elements. Trash containers must be
covered. A sediment rack must be at the entry/exit to minimize tracking sediments offsite. Mounds of dirt
must be covered to prevent wind and water erosion offsite. These are some examples of BMPs to prevent
pollutants from entering storm drains.
398
-------
Figure 2.
ocean.
Concrete washouts are collected for disposal instead of released to the street, storm drain system and
Onsite Treatment
The Urban Runoff Pollution Mitigation ordinance requires new developments that exceed a specific
threshold to incorporate best management practices, BMPs, such as infiltration trenches, french drains,
permeable paving, biofilters and other low-impact structures into the post-construction design of a project.
The design should be linked to how urban runoff will be managed onsite instead of dumping the problem
into the public right-of-way. The express purpose of these low-impact development techniques is for
harvesting precipitation, infiltrating it back into the ground and keeping urban runoff and its low-level
pollutants out of receiving waters. Not only is water quality improved as the runoff passes through soil, but
aquifers are recharged for future extraction.
Private & Public Development: Infiltration Trenches, Biofilters, Permeable Paving
A menu of BMPs is available to choose the best ones to incorporate into the design of a new building.
These are post-construction BMPs to harvest, infiltrate and treat runoff. As shown in Figure 1, the goal is to
design a low-impact development that minimizes the hardscapes, maximizes permeable surfaces and returns
as much water as technically possible into the ground. The most common BMP for single-family
developments is the infiltration trench, a sub-surface retention basin filled with large gravel, stackable
plastic pallets or long concave-shaped plastic cylinders to store a certain amount of runoff for infiltration.
Surface infiltration depression basins in yards also serve to retain runoff for infiltration. Biofilters and
swales are other BMPs suitable for site-specific situations. Porous concrete and permeable paving products,
modular and rolled, replace asphalt and concrete for parking lots, driveways and alleys.
Effectiveness
To date, over 600 new developments, including single-family, multi-family, commercial and City, have
implemented this requirement of post-construction BMPs, keeping over 4,540,000 liters (1,200,000 gallons)
of runoff out of the Bay per 0.25 centimeters (0.10 inch) or greater storm. To put this in perspective, this
amount of water, if harvested and used directly represents about 9% of daily water use. Moreover, the City
contains about 22,500 parcels. About 2.5% of properties in the City have had to comply with the ordinance
and install BMPs since 1995. The City recognizes that each project is site-specific and in some cases BMPs
will not be possible onsite. The ordinance allows for variances.
399
-------
Figures 3, 4. On left, cylindrical infiltration system some 20 feet deep under a subterranean garage for a multi-family
building receives roof runoff during a storm and infiltrates into the ground; on right, common box-shaped, sub-surface
infiltration trench at a single-family development collects roof and other hardscape runoff for infiltration.
Figures 5, 6. Use of plastic in-fill instead of gravel allows greater storage volume, 94% versus 40%.
surface infiltration trench filled with RainStore', on right, trench filled with StormCell.
On left, sub-
Figure 7, 8. On left, biofilter/swale system in parking lot of a school receives all runoff. For almost all storms, all
runoff remains onsite for infiltration. On right, permeable pavers in a parking lot of a business allows runoff to infiltrate
instead of run off into the street.
400
-------
Figures 9, 10. On left and right, before and after photos of Grassy Pavers permeable pavers at a multi-family building
in the parking stalls. Left photo shows pavers exposed before infill. Right photo shows pavers filled with colored rock.
Figure 11. Porous concrete V-swale in a City alley to harvest runoff and reduce flooding of adjacent properties.
Public Surface Systems
As mentioned earlier, City leadership believes that all dry weather and some initial wet weather runoff
leaving the City should receive some treatment to remove pollutants of concern before entering the local
receiving water body, the Santa Monica Bay. The City continues to install BMPs in its storm drain system.
The City has every expectation to have BMPs in all storm drain outlets in the near future.
Catch Basin Inserts
The newest generation of basin and storm drain BMPs, inserts and screens, avoid many of the pitfalls of the
earlier efforts—pieces of wood over the openings of catch basins. Water can pass into the catch basin, trash
can be removed, and high flows still bypass into the basin, avoiding flooding. Many insert types are on the
market. Some filter only trash and debris; some filter both trash and soluble chemicals via a special filtering
medium. The City uses both types of strategies. The City places inserts for trash and debris in areas of high
pedestrian traffic, such as the downtown Promenade area. Inserts that filter hydrocarbons, in addition to
trash, are placed along streets with automotive businesses.
401
-------
Figures 12,13. A catch basin insert, DrainPac, captures trash and debris, preventing these materials from entering
the receiving waters.
Catch Basin Screens
With inserts, City staff must clean them out on a regular basis to maintain the removal efficiency of the
BMP, a time-consuming and costly requirement, especially in confined spaces. With screens attached to the
curbface, trash and debris are kept out of the runoff, water can pass into the basin or drain, and street
sweepers or City staff can remove easily these materials. However, if not properly installed, vehicles can
brush against screens and damage both screens and vehicles. And in some installations, flooding might be
an issue if the screens are covered with trash or in a flood-prone location.
Figure 14. Catch basin screen operating during storm. Water can flow through the openings while keeping trash out.
The City has found inserts and screens to be effective when the best device is chosen for a site, installed
properly and maintained regularly. Many other types of BMPs that fit into catch basins and storm drains
exist. More information about these BMPs, and those used by the City, is available from the author.
Public In-Line Systems
The City installed a number of these BMP devices as off-line centralized treatment systems. The advantage
of centralized BMPs is that all the collection of pollutants and maintenance occurs in one location, instead
of City crews driving to hundreds of locations to clean BMPs. Time and money spent for maintenance are
reduced. To date, the City has found these devices very effective in removing trash, debris, oil and grease,
and solubles attached to sediments. City staff is gathering data on amounts of solid pollutants removed from
catch basins, storm drains and in-line BMPs, as well as characterizing pollutant types. These devices also
allow the City to pinpoint sources of some pollutants depending upon BMP locations.
Many other types of in-line BMPs that fit into storm drain systems exist. More information about these
BMPs, and those reviewed by the City, is available from the author.
402
-------
Figures 15, 16. Left, muffler and concrete pieces captured in a CDS unit during a rain storm. Right, trash, mostly
plastics, removed by the same CDS unit (Continuous Deflective Separation). This CDS unit receives runoff from the
City's highly congested downtown area, rich in pedestrians, visitors, trendy shops and restaurants, and the weekly
Farmer's Market.
Santa Monica Urban Runoff Recycling Facility (SMURRF)
The SMURRF is a first-of-its-kind facility that harvests on an annual basis dry weather urban runoff (93%
of the City's total runoff) from the City's two main storm drain lines, treats the runoff through five systems,
and reuses the new water resource for landscape irrigation and indoor toilet flushing. Santa Monica has
become a leader in its efforts to safeguard and enhance the natural environmental and the community's
health through innovative programs and policies.
What is truly revolutionary about the SMURRF is that not only does it represent an innovative 'wastewater'
(not really wastewater) treatment facility, but it also represents a critical shift in philosophy and
management of a natural resource. The traditional perspective is to dispose of a waste product "out of site,
out of mind." In the case of urban runoff, the City has chosen a watershed perspective, transforming a waste
product--urban runoff—into a valuable local natural resource.
Figure 17. The Santa Monica Urban Runoff Recycling Facility.
This project is an outstanding example of how the City effectively integrated art, engineering, and education
to develop a project that is embraced by the public. This project safeguards and enhances water resources,
prevents harm to the natural environment and human health, and enhances the community and local
economy for the sake of current and future generations. The SMURRF is also an example of how cities
work together to solve a shared problem. In this case, Santa Monica and Los Angeles are partners in this
project. Some 1.1 million liters (300,000 gallons, almost 1 acre-foot) per day of dry weather runoff are
being diverted from the ocean, treated to a high level and reused, or treated and returned to the ocean,
removing a pollution source, especially pathogens.
403
-------
SMURRF Project Goals
The primary objective of the SMURRF, which began operations in February 2001, is to dramatically reduce,
if not eliminate, dry weather urban runoff pollution into Santa Monica Bay. To date, this goal is being met.
Secondary project goals include raising public awareness about problems and solutions of urban runoff
pollution through educational exhibits at the facility and combining strong artistic and architectural elements
into a highly functional design. These goals have also been met through regular tours for interested visitors,
from around the world: tourists, engineers, government officials, students and residents.
In addition, and no less important than any other secondary goal, the development of an additional water
source for use throughout the City is critical. If the City has to treat urban runoff anyway to meet stricter
regulations, why dump the treated effluent into the Bay? Is there not an advantage to reusing the treated
local water resource and reduce imported water supplies? Every acre-foot of water recycled through the
facility equates to one less acre-foot of potable water that must be imported from Northern California and
the Colorado River. In doing so, the SMURRF benefits the entire region as well as Santa Monica.
Water Quality Challenges of Dry Weather Flow
Dry weather runoff captured by the SMURRF originates in a 153 kilometers2 (4,200-acre) drainage area in
the cities of Santa Monica and Los Angeles. Sources of dry weather runoff arise from the inefficient use of
potable water by people: over-irrigation, broken irrigation systems, washing of paved surfaces and business
equipment, car washing on hard surfaces, pool draining, leaking water pipes and hydrants, and illegal
dumping. The average daily flow is estimated to be 1.1 million liters (300,000 gallons) per day, which
represents slightly more than two percent of Santa Monica's overall water demand of 49 million liters (13
million gallons) per day. The facility has a capacity of 1.9 million liters (500,000 gallons) per day.
A variety of pernicious contaminants are found in urban runoff. The presence and concentration of these
contaminants appear to vary significantly over time. Contaminants found in the dry weather runoff treated
by the SMURRF include:
• Suspended and Dissolved Solids
• Oil and Grease
• Trash and other debris
• Pathogens
• Heavy metals (lead, copper, zinc, and chromium)
Initial laboratory tests of influent and effluent SMURRF water samples confirm significant reductions of
these pollutants when found at elevated levels in influent.
Demand Challenges for Recycled Water
The two most likely uses for recycled urban runoff are landscape irrigation and toilet flushing in dual-
plumbed buildings. To date, recycled water is being used for irrigation at the City's cemetery and two
parks, and along a section of the Santa Monica Freeway within City boundaries. Additional users for indoor
flushing will come online over the next few months at a major commercial development and next few years
at the City's new Public Safety Facility next to City Hall and an international consulting firm.
404
-------
Treatment Challenges of Urban Runoff
The five-stage treatment train at the SMURRF consists of bar screens, flow equalization, air floatation,
microfiltration, and UV disinfection. Because the SMURRF is a new system, combining proven
technologies to treat a new water resource presents challenges. Pre-treatment is critical to remove solids
and sediments that can foul secondary and tertiary treatment systems. Daily maintenance is required. Oil
and grease need to be monitored to avoid high concentrations (from spills) from entering the facility and
exceeding the system's parameters. The microfiltration system requires special monitoring to ensure proper
operation and long-term durability and reliability. A major challenge is the control of algae, which is very
common in urban runoff. Initial designs required the inj ection of a background level of chlorine within the
distribution line. However, the City has found that algae grows almost everywhere within the facility,
especially in the finished reservoir. Weekly cleaning is required to prevent the buildup of algae. The City is
considering adding chlorine earlier in the treatment train to reduce algal growth.
Figure 18. Diagram of SMURRF treatment train.
Challenges of Public Education and Artistic Allure
Placing a treatment facility near a prominent tourist site, the Santa Monica Pier and attractive sandy
beaches, presented many challenges. The City took extraordinary steps to include educational, artist and
architectural features, bringing drama to signage, landscape, and architecture; presenting educational
material in a fresh fashion; and providing talking points for visitors. These features are key elements of a
public information campaign that stresses the future importance of stormwater and wastewater recycling as
a local water resource.
Because the SMURRF is open continuously, other types of challenges occur, the types of social challenges
presented by youth and those without shelter. The City has had to balance the openness and unmonitored
design of the SMURRF against the need for operational continuity and system security. During the first
year of operation, City staff visit the facility daily for maintenance and damage control, in addition to its
maintenance of the City's other water distribution systems: potable, waste, storm and recycled.
The daily activity of SMURRF reduces pollution into the Santa Monica Bay and provides a sustainable
alternative water supply for the City of Santa Monica, with the displacement of up to four percent of potable
water demand. The supply is sustainable in the sense that society is wasting hundreds of thousands of
gallons a day through inefficient uses of water.
The collaborative design approach between the artist, architect, engineer, and public works department has
transformed a potentially unsightly treatment facility into an important community asset. The more than 2
million visitors who come to the Santa Monica Beach and Pier each year will have an opportunity to learn
about the benefits of pollution prevention and watershed protection.
405
-------
The SMURRF is a reflection of the shift in how society manages all water resources. No longer is the
traditional approach of removing any and all water resources from our midst acceptable. In a time of
unstable, unreliable potable water supplies, water management needs to shift from the old traditional
approach of over tapping existing potable supplies and think outside the box - use to the maximum extent
practicable all existing and local water supplies, with an emphasis on water efficiency and conservation —
water efficient appliances and landscapes, elimination of leaks, and reuse of "waste" water supplies.
Funding Resources
The City has been fortunate to have a stormwater utility fee, an annual fee incorporated into the annual
property tax bill. This annual revenue source is approximately $1.2 million. However, with the additional
requirements on municipalities from regulations, such as the new NPDES permit and TMDLs, to reduce
urban runoff pollution and improve water quality of receiving water bodies, this revenue source is
inadequate. This fee can no longer support the anticipated future operating and capital expenses of the
City's urban runoff management program.
The City has received many federal, state and county grants, local rebates and state loan funds to implement
many BMPs. A proactive staff and supportive management have allowed the City to seek out and obtain
these grants. Grants cover most if not all of the construction cost of these systems. The City provides a
certain level fiscal resources for planning, design, community outreach and education, and water quality
monitoring. The City also works with neighboring cities to share expenses where appropriate.
Urban Runoff Management Plan
The City recently began a major effort to codify into an urban runoff management plan its dispersed runoff
management program, bringing together the activities of the City's many divisions involved in urban runoff
management. To date, the City has a variety of activities to curb runoff pollution and meet the requirements
of its Phase I National Pollutant Discharge Elimination System permit, through the County of Los Angeles.
Almost all City divisions participate in runoff management, from legal to planning to engineering to open
spaces to enforcement. Since 1990, the City has operated its program without a formalized document, a
repository of all requirements, whether regulatory and City policy, a document that anyone can review,
share and update—a living, dynamic document. Without such a document, City finds it difficult to present a
unified and centralized approach. When other government agencies contact the City for a copy of our plan,
we do not have one document to present. Though the City has many clear objectives and policies, and a
Sustainable City Program, for urban runoff, the City has been lacking in a written plan.
Beginning in November 2002, the City will work with a consultant to begin a year-long process to develop
this document, incorporating the latest hydrologic and hydraulic data about the City's storm drain system,
GIS information and maps, regulatory requirements, and low-impact design solutions. The unique aspect of
this plan is its low-impact approach, seeking watershed solutions upstream for any storm drain system
deficiencies, soft and permeable BMPs instead of traditional hardscapes solutions. Wherever possible, to
upgrade the storm drain system, low-impact design BMPs are preferred and requested, or the installation of
treat and release systems to give a minimum of treatment to meet new standards. The plan's approach is to
treat runoff as the valuable local resource it truly is, and not as a waste product to be easily discarded.
Conclusion
The City's Urban Runoff Management Program has two goals: treat runoff to the highest possible
standards, given economic and regulatory realities, and release; and treat runoff and reuse it as a valuable
406
-------
resource. These goals have three implementation and guiding strategies within the management plan: treat
all dry weather and initial wet weather runoffbefore leaving the City's boundaries; harvest wet weather
flows for groundwater recharge; and harvest, treat and reuse dry weather runoff for landscape and in-door
plumbing purposes. These goals and strategies make up the new Urban Runoff Management Plan. What
makes this plan unique is the toolbox of human, technical and fiscal resources that the City employs to reach
these goals and strategies: numerous divisions working together to meet regulatory requirements; a
supportive City Council and management with a Sustainable City Program with guiding principles; City
employees who are trained and believe in the goals and strategies; a stormwater user fee; grants; and tested
and effective technologies.
SMURRF is the centerpiece of the City's integrated urban runoff management program, being the linchpin
of the City's commitment to protecting the Bay's water quality, wildlife and beachgoers, and an important
best management practice for the Santa Monica Sustainable City Program. Not only can urban runoff be
treated and released back into the environment, the SMURRF demonstrates the feasibility of taking a local
polluted resource, urban runoff, and turning it into a valuable natural resource for reuse, helping to displace
the need for more expensive and energy-intensive imported water. This BMP and those BMPs installed by
new development to harvest stormwater for infiltration establish a precedence for exhausting efforts to first
reuse local water resources of various qualities before turning to distant water resources, the removal of
which may cause significant ecological damage and water supply disruption to distant aquatic habitats and
cities. These BMPs also keep potential pollutants of concern out of surface waters, improving water quality
for beneficial uses and protection of wildlife and human visitors to the ocean.
References
American Oceans Campaign, 2000. About Stormwater and Polluted Runoff, www.americanoceans.org.
Coastal Alliance, 2000. Pointless Pollution: Preventing Polluted Runoff and Protecting America's Coasts,
Washington, DC.
Cone, Maria, 2000. Study Finds Widespread Runoff Peril on the Coast, Los Angeles Times, November 29.
Los Angeles County, 2000. Reducing the Health Risks of Swimming at Los Angeles County Beaches,
1999-2000 Grand Jury Final Report, 192-246.
Haile, R.W., Alamillo, J., Barrett, K., Cressey, R., Dermond, J., Ervin, C., Glasser, A., Harawa, N, Harmon,
P., Harper, J., McGee, C., Millikan, R.C., Nides, M., and J.S. Witte, 1996. An Epidemiological Study of
Possible Adverse Health Effects of Swimming in Santa Monica Bay, Final Report, Santa Monica Bay
Restoration Project, Los Angeles.
May, C.W., Horner, R.R., Karr, J.R, Mar, B.W., and E.B. Welch, 1997. Effects of Urbanization on Small
Streams in the Puget Sound Lowland Ecoregion, Watershed Protection Techniques, 2:483-494, Center for
Watershed Protection, Maryland.
Mehta, S., 2002. State OKs Tough New Regulations to Reduce Runoff, Los Angeles Times, January 19.
Santa Monica, City of, 2000. Urban Runoff Mitigation Ordinance, Chapter 7.10 of the Santa Monica
Municipal Code, #1992, November 28.
407
-------
Santa Monica, City of, 1994. Sustainable City Program adopted by City Council, September 24.
Schueler, T. and R. Claytor, 1995. Impervious Cover as an Urban Stream Indicator and a Watershed
Management Tool, Effects of Water Development and Management on Aquatic Ecosystems, ASCE, New
York.
Schueler, T., 1994. The Importance of Imperviousness, Watershed Protection Techniques, 1(3), Center for
Watershed Protection, Maryland.
Sheppard, H., 2000. Testing the Waters: Health of region's shores threatened by urban runoff, Daily News,
Los Angeles, May 17.
408
-------
A PROCESS FOR DETERMINING APPROPRIATE
IMPACT INDICATORS FOR WATERSHED PROJECTS
Robin Shepard
Wisconsin Water Quality Coordinator
University of Wisconsin, Madison, WI;
Catherine Neiswender
Great Lakes Regional Water Quality Liaison; and
on behalf of Cooperative Extension's
Great Lakes Regional Water Quality Leadership Team*
Abstract
Watershed project evaluation, especially in urban-focused efforts, typically focuses on water quality
improvements, habitat expansion or improvement, and a variety of other positive changes in the physical
and biochemical realms. However, watershed projects are ultimately about influencing human behaviors and
changing how people interact with the natural resources in the watershed. By including both physical and
social indicators of change, a more holistic approach to watershed project evaluation can emerge. A Logic
Model for Program Performance was used in group discussions by State Nonpoint Source Pollution (Section
319 Project) Coordinators from the Great Lakes Region to identify a set of common impact indicators for
assessing Section 319 projects. These multi-state discussions confirmed the lack of focus on the behavioral
and socio-economic components of water quality efforts. Results of these and ongoing discussions will
establish a set of impacts that can be used both to develop state and regional reporting procedures and to
create a training program for Section 319 project staff.
Introduction
Increased pressures from politicians and agency personnel through program reviews and audits, as well as
the federal enactment of the Government Performance and Results Act (GPRA) in 1993, are examples of
the ever-expanding focus on program results and impacts. As the demand for accountability in natural
resources programming increases, so too will the need for thoughtful, well-planned program evaluations
(Davenport, 2002).
Evaluation is a critical dimension of any watershed project. It is most often used in summative or conclusive
ways to identify what was accomplished by a project after a specified period of time. But, evaluation can
also be a formative element in program planning and implementation, to ensure that projects within those
programs are meeting short- and long-term goals. Building evaluation skills and developing the confidence
to use those skills is critical for watershed-based staff if they are to answer questions about the effectiveness
and efficiency of their programs. While it may not be necessary for educators to become evaluation experts,
they do need a fundamental understanding of methods and ethical standards if they are to make evaluation
part of overall program design.
Evaluation is the systematic collection of information about the activities, characteristics, and outcomes of
programs, personnel, and products, in order to reduce uncertainties, improve effectiveness, and make
decisions with regard to what those programs or products are doing and affecting (Patton, 1982). While
evaluation includes a look at program impacts, it is different from impact reporting, which focuses on
409
-------
specific program results that may only be important to program stakeholders (Patton, 1997; Bickman, 1985;
and Cronbach, 1982) Evaluation measures a variety of outcome data against the program's intent (Bennett
and Rockwell, 1995).
Approach
To improve how evaluation is used in watershed projects, six land grant universities in the Great Lakes
region (i.e., Illinois, Indiana, Michigan, Minnesota, Ohio and Wisconsin) are working with state and
regional coordinators from nonpoint source pollution projects (Section 319). This multi-state effort, which
includes participation by the U.S. Environmental Protection Agency Region V office, has been initiated to
identify consistent and reliable impact indicators and evaluation processes. A series of small group
discussions and interactive training sessions on evaluation is currently being offered to state-level 319
coordinators. Those meetings and interactions will encourage cross-state problem solving and lead to the
development of common success indicators for watershed projects.
Discussion
Typically, evaluation is not addressed until late in, or even at the end of, a project. This reactive evaluation
is often merely a hunt for positive impacts, and has limited value in either describing the success of a
program or in planning future efforts. A more planned, formative evaluation that is integrated into the
project from the very beginning can track changes over time.
Formative evaluation (Scriven, 1967) examines issues such as audience needs, current knowledge gaps,
prevalent behaviors, and information preferences. Because they are assessed prior to a project's start, these
issues can be used to influence the design and implementation of the outreach efforts (King & Rollins, 1999;
Lanyon, 1994; Mattocks & Steele, 1994). One barrier associated with formative evaluation approaches is
deciding what to measure.
Water quality projects are by nature directed at protecting or improving physical water quality. Biophysical
changes to the water are normally the measure of success (Davenport, 2002). While the ultimate goal of
water quality projects may be to protect or enhance water quality, there are other impacts to assess, such as
increased knowledge, improved skills or the adoption of improved management practices (Rogers, 1995).
Research has shown certain management practices to be beneficial to water quality and farm profits, and the
promotion of these practices by project staff is at the heart of most water quality outreach efforts. Therefore,
both long-term indicators (i.e., physical changes to water quality) and more immediate impacts (i.e., changes
in farm management and behavior) were assessed in this study to determine the level and type of evaluation
support needed by and from state water quality coordinators.
In prior internal assessments of evaluation processes (Shepard, 2002) used by water quality program staff,
only three (10 percent) of the states actually conducted a formative assessment strategy for their project.
This involved documenting pre-project needs and audience characteristics specifically for USDA Water
Quality program efforts pertaining to the Cooperative State Research Education and Extension Service
(CSREES) Water Quality Initiative of the 1990s. When individual project coordinators were asked what
information they intended to use to determine program impact, they mentioned a range of indicators, from
biophysical environmental (e.g., sediment loading, biotic indexes, etc.) to behavioral (e.g., awareness,
knowledge or adoption of practices). When a range of potential indicators was assessed for intended use, it
was shown that many states intend to rely on such indicators without any true baseline from which change
can be adequately assessed (Figure 1).
410
-------
ftoiic indenes
UttfoS usa paUo«ns
—]
J 30",
1
0)
2
•5
§
tm pigmentation
use
waif •" vf" prr>t i.-— «
I <)• I 'MS,
participation
(sat
0
H Hitve ti»Mline dala
[J Surafnam/e evaijahcn only
20
40
60
Percent
Figure 1. Evaluation Measures Used by CSREES Water Quality Coordinators.
Presentation Focus
This presentation will summarize results from the Section 319 Project Coordinators' group discussions
about evaluation and the proposed training program (suggested in the Approach Section above). Results will
offer ideas from state and regional project staff as to: 1) the purposes for evaluation, 2) suggested processes
and methods, and 3) recommendations for strengthening watershed evaluations. As watershed-based efforts
come under more scrutiny, watershed program administrators and funders need to know how to evaluate the
success of these efforts. Results from this project are planned to be implemented in 319-funded and other
watershed projects by 2004.
An Overview of Results
In fall 2002, an interactive process began with a small group discussion of State Nonpoint Source Pollution
(Section 319 Project) Coordinators from Illinois, Michigan, Ohio, Indiana and Minnesota. That meeting on
October 23-24 was subsequently followed with a series of email discussions among the state coordinator in
order to share ideas about what can and should be the basis of project-level reporting and evaluation.
As a starting point for the exchange of ideas on reporting, the October meeting focused on using the Logic
Model for Program Performance as a framework to identify the potential range of program and project
impacts. Over the next several months, the ideas generated by that meeting will continue to be discussed and
further refined with the intent of developing set of primary program and project-level impacts that can be
tracked over time and reported through the existing regional network of Section 319 projects. Again, this
paper is a progress report on the development of common indicators for Section 319 projects, and is meant
411
-------
to foster broader discussion through its presentation. The information and data presented here are
preliminary and will continue to be refined as a training program is developed in 2003.
To guide the discussion pertaining to what is currently, and what can be, evaluated, the Logic Model for
Program Evaluation was used (Figure 2). The Logic Model has been used in a number of disciplines to help
identify three levels of programmatic impact referred to as: (1) input, (2) outputs and (3) outcomes.
PLANNING
INPUTS
Programmatic
Investments
OUTPUTS
Activities
Participation
OUTCOMES
Short
Medium
Long
EVALUATION
Figure 2. The Logic Model for Program Evaluation (Taylor-Powell, 1998).
Inputs are a category of program investment that includes staff time and dollars invested to conduct the
program or project. Outputs refer to those actions that are immediately caused or supported by the initial
inputs. Outputs include watershed activities and events. Outputs also can include the initial participation in
such activities, like the number of farmers attending a demonstration or field day. Outcomes are those
impacts that result from the activities and events of the project. Outcomes are commonly divided into short-,
medium- and long-term impacts. Short-term outcomes could include changes in knowledge or the
acquisition of specific skills introduced at a demonstration or field day. Medium-range outcomes would
include the application of skills or behaviors such as the adoption of improved management practices that
were demonstrated by the project. And long-term indicators are most often considered to be actual changes
to the environment, such as biophysical improvements in water quality. The Logic Model has relevance to
both program planning and program evaluation. If programs/projects begin by identifying the outcomes they
are hoping to achieve (top arrow), they will plan the program/project from right to left. As the
program/project is implemented, it actually unfolds from left to right (bottom arrow).
In discussions with states in USEPA Region V (during the October 23-24 meeting), the Logic Model was
used to help identify the three categories of inputs as they pertain to the Section 319/watershed projects
(Figure 3). States and EPA Regional Staff readily identified inputs and outputs, but short- and medium-
range outcomes were more problematic.
412
-------
PLANNING
INPUTS
base funds
amount of funds
to sub-state
recipients
number of state
employees
OUTPUTS
Activities:
TMDL
identification
Participation:
brnp related
activities
OUTCOMES
Short:
bmp
adoption
rates
Medium :
stream
bank/
shoreline
restoration
(miles)
Long
NPS
pollutant
reductions
load
reductions
EVALUATION
Figure 3. The Logic Model as Built by USEPA Region V Staff (adapted from Taylor-Powell, 1998).
Results from this process have focused much attention on the lack of behavioral and socio-economic
indicators in the short- and medium-outcome categories. This finding has not been totally unexpected, given
the biophysical orientation of technically trained watershed staff and the emphasis placed on biological and
chemical changes to water quality parameters. Few would disagree that water quality programs are
primarily about changing or protecting water quality - the natural resource itself. However, concern over the
extent of biophysical change that is possible, and the time it takes for those biophysical indicators to change,
may be well beyond the political life of a watershed or water quality project. This means our staff and
programmatic resources are often focused on five-to-ten year windows of time, while the biophysical
indicators may take many more years to show change. Therefore, if biophysical changes in water resources
do indeed take much longer than the life of a particular program, then social indicators of change (i.e., short-
and medium-range indicators like practice adoption) may be more useful and obtainable as measures of
success in the lifespan of the watershed project. Social indicators, in this context, are not considered
exclusive, but rather are valuable complements to long-term biophysical outcomes. Watershed projects are
about changing the way resources are managed and cared for. After all, human behavior and interactions
with the resource may in fact be the true focus of many environmental protection programs, and social
science indicators should be given more attention and not merely written off as "soft" or too difficult to
measure adequately.
Future Implications
During winter 2002-03, email and conference calls will be used to further complete the Logic Model(s) for
each of the Region V states. The goal of this process is to (1) better define a set of impact indicators that can
be built in to state and regional reporting procedures; and (2) identify a training and professional
development program for Section 319 projects that will help build local/watershed capacity that will support
413
-------
and conduct program evaluation. At this time it is premature to identify the exact curriculum and format for
this training and professional development, however, those concepts are expected to be developed by
February 2003.
References
Bennett, C. and K. Rockwell, 1995. Targeting Outcomes of Programs (TOP), an Integrated Approach to
Planning and Evaluation. (A program planning guide prepared for USDA employees.) Washington, D.C.:
Cooperative State Research, Education and Extension Service.
Bickman, L., 1985. Improving Established Statewide Programs: A Component Theory of Evaluation.
Evaluation Review, 9(2), 189-208.
Cronbach, L.J., 1982. Designing Evaluation of Educational and Social Programs. San Francisco: Jossey-
Bass.
Davenport, T.E., 2002. The Watershed Project Management Guide. Lewis Publishers: New York.
King, R and T. Rollins, 1999. An Evaluation of Agricultural Innovation: Justification for Participatory
Assistance. Journal of Extension, 37(4).
Lanyon, L.E., 1994. Participatory Assistance: An Alternative to Transfer of Technology for Promoting
Change on Farms. American Journal of Alternative Agriculture, 9(3), 136-142.
Mattocks, D. and R. Steel e, 1994. NGO-Government Paradigms in Agricultural Development: A
Relationship of Competition or Collaboration? Journal of International Agriculture and Extension
Education, 1(1), 54-61.
Patton, M., 1997. Utilization-focused Evaluation. Thousand Oaks, California: Sage, p.20 and p.200.
Patton, M., 1982. Practical Evaluation. Sage: Newbury Park, California.
Rogers, E.M., 1995. The Diffusion of Innovations. (4th ed.). Free Press: New York, NY.
Scriven, M., 1967. The Methodology of Evaluation. In Tyler, R.W., Gagne, R.M. and M. Scriven (Eds.),
Perspectives of Curriculum Evaluation (pp. 39-83). Rand McNally: Chicago, Illinois.
Shepard, R., 2002. Evaluating Extension-based Water Resource Outreach Programs: Are We Meeting the
Challenge? Journal of Extension, 40(1). Accessible at [http://www.joe.org/joe/2002february/a3.html].
Taylor-Powell, E., 1998. The Logic Model: A Program Performance Framework. Cooperative State
Research Education and Extension Service and University of Wisconsin-Extension: Madison, Wisconsin.
Accessible at [http://www.uwex.edu/ces/pdande/].
414
-------
* The Great Lakes Regional Water Quality Leadership Project is a collaborative effort among CSREES
land grant universities. Team members include: Jim Anderson (Minnesota), Jon Bartholic (Michigan),
Joe Bonnel (Ohio), Jane Frankenberger (Indiana), Mike Hirschi (Illinois), Ruth Kline-Robach (Michigan),
Lois Wolfson (Michigan) and Robin Shepard (Wisconsin). The Great Lakes Regional Water Quality
Liaison is Catherine Neiswender (Wisconsin).
415
-------
DUAL FUNCTION GROWTH MEDIUM AND STRUCTURAL SOIL FOR USE AS
POROUS PAVEMENT
John J. Sloan and Mary Ann Hegemann
Texas Agricultural Experiment Station
Dallas, Texas
Abstract
Porous grass-covered parking surfaces can reduce the quantity of urban storm water runoff and filter out
potentially harmful chemicals. The objective of this study was to develop porous structural soils that
promoted and sustained healthy turfgrass growth and also reduced the effects of contaminated pavement
runoff. The basic medium for all soils was a 50:50 mixture of expanded shale and sand. The expanded shale
component consisted of: 1) a large diameter particle (3 to 6 mm), 2) a small diameter particle (1 to 3 mm),
or 3) a 50:50 mixture of the two. The basic blends were mixed with 0, 10, and 20% peat moss (v/v) and 0,
10, and 20% zeolites (v/v) and placed in 15-cm pots in a greenhouse. Bermudagrass plugs were planted in
each pot. Grass growth was evaluated to determine which mixtures promoted establishment of vigorous turf.
When added alone to the sand/expanded shale medium, peat moss increased bermudagrass growth and also
improved plant response to added fertilizer, but the effect diminished in the absence of regular fertilization.
Zeolites had no significant effect on plant growth in the absence of peat moss. Growing mediums that
contained both 10-20% peat moss and 10-20% zeolites consistently produced more bermudagrass biomass
than the unamended sand/expanded shale mixture. Changing the ratio of small to large diameter expanded
shale in the basic medium did not affect bermudagrass yield. Very low amounts of Cd, Cu, Pb, and Zn were
recovered in leachate after the addition of 10 mg metal per pot, suggesting that most heavy metals (>99%)
were retained in the growing medium.
Introduction
Urban areas are increasingly covered by impermeable parking surfaces that contribute to greater quantities
and intensities of storm water runoff with elevated concentrations of particulates, heavy metals, and organic
chemicals (Barrett et al, 1998; Harrison and Wilson, 1985; Morrison et al, 1984; Stotz, 1987). For
example, the Elm Fork Branch of the Trinity River, which passes through the Dallas metroplex, was
included in the Texas 1998 Clean Water Act list of impaired water bodies due to elevated concentrations of
dissolved lead. Paved and rooftop surfaces also contribute to an increasing trend in nighttime surface
temperatures (Gaffen and Ross, 1999). Data from the Urban Heat Island Pilot Project, a joint
USEPA/NASA venture, showed that surface temperatures of paved surfaces and rooftops was much higher
than the air temperature (111°F vs. 85°F), whereas vegetated areas had lower surface temperatures (83°F)
(Johnson, 1999; Lo et al., 1997).
Urban water quality could be improved by increasing the amount of vegetated surfaces within the urban
limits. Use of strategically positioned grass-covered permeable surfaces for intermittent parking would
decrease the amount of impermeable surfaces in the urban environment and potentially decrease the quantity
and pollutant load of runoff water. When runoff water from impermeable pavement passes over a
permeable surface, the concentration of pollutants is reduced (Legret et al., 1996; Pratt, 1989; Stotz and
Krauth, 1994). Use of grass-covered permeable surfaces for intermittent parking would decrease the
amount of impermeable surfaces in the urban environment and decrease the quantity and pollutant load of
416
-------
runoff water (Barrett et al., 1998). In addition to improving runoff water quality, vegetated surface will help
reduce urban heat buildup (Johnson, 1999; Lo et al., 1997).
The objective of our study was to evaluate combinations of expanded shale, sand, peat moss and zeolites as
growing mediums for turfgrass. We also evaluated the ability of each mixture to remove heavy metals and
phosphorus from contaminated runoff water.
Materials and Methods
Porous Pavement Mixes
Table 1 shows the ingredients in each of the nine porous pavement mixes evaluated in this study. The major
component of each mix (60 to 100% by volume) was a base blend that contained 50% greens-grade sand
plus 50% of a equal portions of small (1 to 3 mm) and large (3 to 6 mm) diameter expanded shale (Fig. 1A).
The greens-grade sand met the specifications for construction of a U.S. Golf Association putting green,
containing mostly medium to course grained sand (0.25 to 1.0 mm). Expanded shale is a light-weight
porous aggregate made by heating crushed shale to >1200 C. For seven of the nine porous pavement mixes,
small and large diameter expanded shale were mixed in ratios of 1:1. The eighth and ninth porous pavement
mixes were included in the study to determine the effect of expanded shale particle size on the porous
pavement mixes. The eighth base blend was a 50:50 mixture of sand plus small diameter (1 to 3 mm)
expanded shale, whereas the ninth was a 50:50 mixture of sand and large diameter (3 to 6 mm) expanded
shale.
Table 1. List of ingredients in the base blend of each porous pavement mixture plus the content of peat
moss and zeolites added to each base blend.
Mix No.
1
2
3
4
5
6
7
8
9
Base Blend
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50/50 Small/Large diameter shale + 50% Sand
50% 100LS + 50%Sand
50% 100SS + 50% Sand
Peat Moss
0
10
20
0
0
10
20
10
10
Zeolites
0
0
0
10
20
10
20
10
10
Sphagnum peat moss is partially decomposed sphagnum moss harvested from peat bogs found mostly in
Canada. Peat moss was added to the porous pavement mixes to improve water holding capacity and to
provide a source or organic matter for promoting biological activity. Sphagnum peat moss was added to
some of the base blends at rates of 0, 10, and 20% by volume.
Natural zeolites are aluminosilicate minerals with a unique interconnecting crystal lattice structure that gives
them a large internal surface area and a very high cation exchange capacity. We added zeolites from New
Mexico to the porous pavement blends for two reasons. First, we thought they would help retain fertilizer
nutrients in the porous pavement mixtures so that the mixes would need less frequent fertilization. Second,
417
-------
we thought zeolites would absorb heavy metals from contaminated urban runoff water as it percolated
through the porous pavement. Zeolites were added to the base blends at rates of 0, 10, and 20% by volume.
The various porous pavement mixes were chosen so that we could test the effect of peat moss and zeolites
alone or in combination with each other. Mixes 2 and 3 showed the effects of peat moss, mixes 4 and 5 the
effects of zeolites, and mixes 6 and 7 the effects of the combined ingredients. A comparison of mixes 8 and
9 showed the effect of expanded shale particle size.
•t
^B
Figure 1. (A) Base blend for porous pavement mixes consisting of 50% sand plus 50% of equal
portions of small (1-3 mm) and large (3-6 mm) expanded shale; and (B) bermuda grass growing
on the base blend in 5-in pots.
Physical Properties
A portion of each porous pavement mix was used to determine bulk density and approximate water holding
capacity. Each porous pavement mix was placed in a 1 L polyvinyl chloride leaching column of known
volume and saturated with water for a period of 24 hours. Then excess water was drained from the porous
pavement mix for 24 hours prior to measuring the wet weight. The mixes were then dried at 105°C for 48
hours before measuring the dry weight. Water holding capacity (equivalent to soil field capacity) was
calculated on a volumetric basis using a value of 1 g/mL for water. Bulk density was equal to the oven-dry
weight divided by the volume of porous pavement mix.
Grass growth
The porous pavement blends were placed in greenhouse pots measuring approximately 12.5 cm height by 10
cm depth. Bermudagrass sprigs (Cynodon dactylon [L.]) were collected from bermudagrass plots on native
soil. Sprigs were washed to remove soil prior to planting 3-4 sprigs in each porous pavement pots. During
the first 2 weeks after planting the sprigs, each pot received three applications of soluble 20-20-20 fertilizer
for a total N, P, and K rate of 0.48, 0.21, and 0.40 g/pot, respectively. After grass was established (Fig. IB),
each pot was periodically fertilized (approximately every 200 to 260 days) with a slow-release form of 18-6-
12 fertilizer at a rate of 1.08, 0.16, and 0.60 g/pot of N, P, and K, respectively. Pots were maintained in a
greenhouse environment most of the time, but were periodically moved outdoors when mealy bugs
(Pseudococcus Spp.) became a problem. Growth rates varied depending on the time of year and time after
418
-------
fertilizer application. Grass tissue was clipped to a 3.8 cm height whenever necessary. Clippings were oven
dried at 65°C and weighed to determine biomass production.
Heavy metal and phosphorus leaching
After grass was well established on each pot, 10 mL of an aqueous solution containing 250 jig each of Cd,
Cu, Pb, and Zn was added to the top of each pot. Pots were then leached with 250 mL of deionized water at
1, 3, 7, and 14 days after metal addition. The leachate volume was measured, filtered through a medium
grade filter paper, and analyzed for Cd, Cu, Pb, and Zn content by atomic absorption spectroscopy. In most
cases, the concentration of these heavy metals was below detection limits of the instrument. Therefore,
another 750 jig of the same heavy metals was added to the top of each pot and the pots were leached with
375 mL deionized water 5 days after metal addition Heavy metal concentrations were still very low, so two
months later we added 10 mg each of Cd, Cu, Pb and Zn to each pot and leached them with 375 mL
deionized water at 1 and 4 days after metal addition. In all cases, leachate volume was measured and filtered
prior to subsequent analyses. To determine the effect of the porous pavement mixes on absorption of heavy
metals, we calculated the cumulative amount of heavy metals leached from each pot following the three
additions of heavy metals. The cumulative amount was calculated by summing the product of leachate
volume and heavy metal concentration for all the leaching events. However, since we did not collect
leachate every time we added water to the porous pavement mixes, the cumulative values should be
interpreted as qualitative measurements rather than the total flux of heavy metals leached from the porous
pavement mixes.
Another purpose for collecting leachate was to determine the fate of fertilizer P added to the porous
pavement mixes. An aliquot of the same leachate sample analyzed for heavy metals was also analyzed for
dissolved inorganic P content. Inorganic P was determined using the colorimetric method of Olsen and
Sommers (1982). Inorganic P measurements should also be interpreted as a qualitative indicator of the
ability of the porous pavement mixes to absorb P. Leachate P data was interpreted by considering the
number of days the leachate was collected after the last application of fertilizer.
Results and Discussion
Physical properties
For each physical property, the nine porous pavement treatment means are presented in a single bar graph.
However, the data will discussed in terms of how each specific variable (peat moss content, zeolites content,
or expanded shale particle size) affected the physical property of interest. Most treatment means were
compared to the simplest porous pavement mix that contained only the base blend without peat moss or
zeolites. In the following bar graphs (Figs. 1, 2, and 3), the control treatment is the bar furthest to the left
(Mix No. 1) with the other mixes located by increasing mix number (Table 1) to the right. For statistical
purposes, the means for all nine treatments were compared simultaneously using Duncan's multiple range
test. In general, significant differences among treatment means were easily discerned for all physical
properties due to a low degree of variability in the data.
Bulk density
Peat moss was the ingredient that had the greatest effect on soil bulk density (Fig. 2). When added at a 10%
rate (v/v), peat moss was no different that the base blend (Mix No. 2 vs. 1), but a 20% addition of peat moss
419
-------
(Mx No. 3) significantly decreased bulk density. Peat moss is an organic material with a lower bulk density
than mineral materials such as soil and expanded shale (Sloan, et al. 2002) Therefore, replacement of the
expanded shale/sand base blend by peat moss in the porous pavement mix caused the bulk density to
decrease. On the other hand, zeolites have a higher bulk density than expanded shale, which comprised
50% of the base blend, so addition of zeolites to the porous pavement mixes caused the bulk density to
increase (Mixes 4 and 5 vs. 1). Addition of 10% peat moss and 10% zeolites did not significantly change
bulk density of the porous pavement mix (Mx No. 6 vs. 1), probably because the addition of one negated
the effect of the other. Therefore, it was somewhat unexpected to see that the porous pavement mix with the
lowest bulk density was the blend that contained 20% peat moss and 20% zeolites (Mx No. 7 vs. 1). With a
20% addition of each of these ingredients, the base blend comprised only 60% of the porous pavement mix.
In reality, we did not measure the final volume of the porous pavement mix after we blended the
ingredients. The ingredients probably combined in such a way that there was a looser arrangement of
individual particles in the final mix, especially the heavier expanded shale and sand particles. Expanded
shale particle size had no effect on bulk density of the porous pavement mix (Mx No. 8 vs. 9). The bulk
densities for all porous pavement mixes ranged from 1.0 to 1.4 g/cm3, which suggested there would be no
impediment to root growth.
1.6
•. Bulk Density
u
O)
£1-2
1 1.0
X
_i
D
CO 0.8
I I I I
I I I I
I I I I
I
I II
I I I
U.D
Mix No.-:
at Moss—:
Zeolites-:
se Blend-:
1
0
• 0
50/50
2
10
0
50/50
3
20
0
50/50
4
0
10
50/50
5
0
20
50/50
6
10
10
50/50
7
20
20
50/50
8
10
10
100SS
9
10
10
100LS
POROUS PAVEMENT MIX
Figure 2. Effect of peat moss and zeolites content or expanded shale particle size on the bulk
density of porous pavement blends.
Water Holding Capacity
Water holding capacity of the porous pavement mixes is equivalent to soil field capacity because it is the
amount of water retained in the mix after all excess water has drained gravimetrically. Both the 10% and
20% additions of peat moss increased the water holding capacity of the base blend (Fig. 3) (Mixes 2 and 3
vs. 1). The 20% addition of zeolites (v/v) also increased water holding capacity of the base blend (Mx No.
5 vs. 1), but not the 10% addition (Mx No. 4 vs. 1). However, the increase in water holding capacity due to
zeolites was not as great as the increase due to peat moss (Mix No. 5 vs. 3). Nus and Brauen (1991) found
420
-------
that sand amended with 10 to 20% peat moss retained more moisture than sand amended with equal
amounts of natural zeolites. Mix No. 7, which contained 20% peat moss and 20% zeolites, exhibited the
highest water holding capacity relative to all other porous pavement blends. This is consistent with the low
bulk density for the same mix (Fig. 2). Apparently a combination of 20% peat moss and 20% zeolites has a
greater potential to retain water than a 20% addition of either ingredient alone. Once again, it is probably
related to the physical arrangement of peat moss and zeolites with the expanded shale/sand base blend.
Expanded shale particle size had a small but statistically significant effect on water holding capacity of the
porous pavement mix. The porous pavement mix that used only small diameter (1 to 3 mm) expanded shale
in the base blend had a higher water holding capacity than the mix that used only large diameter (3 to 6 mm)
expanded shale (Mix No. 8 vs. 9). This is consistent with the effect of particle size on water holding
capacity of natural soils.
50
> 40
30
8 20
10
Container Capacity Water Content
ll i i i i
ii
ii
i i i i i i i i
u
Mix No.-->
Peat Moss->
Zeolites-->
Base Blend~>
1
0
0
50/50
2
10
0
50/50
3
20
0
50/50
4
0
10
50/50
5
0
20
50/50
6
10
10
50/50
7
20
20
50/50
8
10
10
100SS
9
10
10
100LS
POROUS PAVEMENT MIX
Figure 3. Effect of peat moss and zeolites content or expanded shale particle size on the water
holding capacity of porous pavement blends.
Grass Growth
Bermudagrass clippings were collected twenty two times during a 26-month period. After 15 clippings, we
noticed the pots were infected with mealy bugs (Pseudococcus Spp.), so we clipped the bermudagrass to the
crown level. Mealy bugs continued to be a problem, so after the nineteenth harvest, porous pavement pots
were moved outside the greenhouse. Bermudagrass was clipped at intervals ranging from 14 to 65 days,
depending on the rate of growth. Since we were interested in the long term ability of the porous pavement
blends to sustain plant growth, we calculated the cumulative clipping weights per pot (Table 2). Peat moss
was the only ingredient that significantly increased bermudagrass clipping weights compared to the
unamended base blend. Zeolites had no effect on bermudagrass growth when added to the base blend alone
or with peat moss.
Fertility was the main factor controlling the rate of bermudagrass growth. Bermudagrass required clipping
at 2 to 3 week intervals during the first two months after fertilization, but less frequently after that. Pots
were fertilized only five times during 26 months. Therefore, the structural soil was probably depleted of
nutrients prior to the each fertilization. The length of time between fertilizations ranged from 200 to 260
421
-------
days. After fertilization, the structural soil blends responded differently to the added nutrients. Those
porous pavement blends that contained peat moss or peat moss plus zeolites responded to fertilizer more
favorably than those blends that did not in terms of bermudagrass clipping weights (data not shown). In
general, there was no difference in bermudagrass clipping weights among the various porous pavement
blends beyond 80 to 90 days after the last fertilizer application.
Table 2. Treatments 1 through 7 show the effect of peat moss and zeolite on cumulative
bermudagrass clipping rates when mixed with a 50:50 blend of small (1-3 mm) and large (3-6 mm)
diameter expanded shale at rates of 10 and 20% (based on volume). Treatments 8 and 9 show
the effect of expanded shale diameter on cumulative bermudagrass clipping weights.
TrtNo
Base BlendT
Peat
Moss
Content
Zeolite
Content
Cumulative
Clipping
Weights
SD*
2
3
4
5
6
7
8
9
50/50
50/50
50/50
50/50
50/50
50/50
50/50
100SSh
100LSh
0
10
20
0
0
0
0
Linear effect of peat moss
10
20
Linear effect of zeolites
10 10
20 20
Linear effect of combined peat moss and zeolites
10 10
10 10
Linear effect of expanded shale diameter
(g/pot)
91.7
98.7
107.9
100.9
95.1
NS
106.2
116.9
110.3
106.3
NS
17.0
9.8
9.0
18.0
13.5
9.1
20.6
6.2
12.0
NS, ** Not significant and significant at the 0.05 level of probability, respectively.
t Base blends were mixed in a 50:50 ratio with sand before mixing with the other ingredients.
t Standard deviation of the treatment mean.
Leachate Chemistry
Leachate was not collected continuously throughout the study, but rather at specific times in relation to the
addition of heavy metals to the pots. We generally collected leachate for several days after heavy metals
were added. The leachate was analyzed for heavy metals (Cd, Pb, and Zn) and inorganic phosphorus.
Additional leachate was periodically collected to assess the effect of fertilization on inorganic P
concentration and other nutrients (data not shown). Since we did not collect all leachate from the porous
pavement mixes, we cannot calculate a mass balance for the heavy metals and nutrients added to the pots.
However, the leachate data is a good indicator of the effect of the porous pavement ingredients on the
leaching loss of potential environmental pollutants.
422
-------
Heavy metals
Our hypothesis was that the addition of zeolites to the porous pavement mixes would increase their ability to
remove heavy metals from contaminated runoff water. The results shown in Figure 4 for Cd, Pb, and Zn are
somewhat inconclusive. In most cases, the concentrations of heavy metals in the leachate waters were very
close to the analytical detection limits. This introduced a high degree of variability in the data and made it
more difficult to discern significant differences among porous pavement mixes. In the case of Cd, neither
zeolites nor peat moss affected the amount of Cd in leachate relative to the unamended base blend (Mix
Nos. 2 to 7 vs. Mix No. 1). For some reason, the two porous pavement mixes that contained only small or
large expanded shale in the base blend (Mix Nos. 8 and 9) resulted in significantly higher Cd concentrations
in the leachate water. The reason for this is unclear, but it could be related to the physical arrangement of
particles in the porous pavement blend. The results for Zn were very similar to those for Cd. Essentially,
peat moss and zeolites did not affect the amount of Zn in leachate water, either when applied alone or
together. Only expanded shale particle size affected the amount of Zn. Both small and large diameter
expanded shale increased leachate Zn when they were the only form of expanded shale in the base blend.
Mix No.-> 1
8
9
Peat Moss-> 0
Zeolites-> 0
10 20 0 0 10 20 10 10
0 0 10 20 10 20 10 10
SS LS
Figure 4. Concentrations of Cd, Pb, and Zn in leachate from the porous pavement mixes after
applying 10 mg of each heavy m etal to the top of each pot.
Lead was the only heavy metal that appeared to be affected by the presence of zeolites in the porous
pavement mix (Fig. 4). Leachate from porous pavement mixes that contained 10% and 20% zeolites
without peat moss (Mix Nos. 4 and 5) contained significantly lower concentrations of Pb than the
unamended base blend (Mix No. 1). Leachate from the porous pavement mix that contained both 10% peat
moss and 10% zeolites (Mix No. 6) also had lower levels of Pb than the unamended base blend, but not the
mix that contained 20% of both ingredients (Mix No. 7).
423
-------
In general, the effect of zeolites on heavy metal removal from leachate water is still unclear based on the
results of this study. Zeolites have been used to successfully remove heavy metals from wastewater
(Ibrahim et al., 2002), so it is logical to expect them to remove heavy metals from contaminated runoff
water. However, there is a high degree of variability in the properties of natural zeolites (Mumpton, 1999),
so some zeolites sources may be better than others. In our study, the failure to see definite effects due to the
inclusion of zeolites in the porous pavement mixes was probably due to a combination of two factors. First,
the amount of zeolites added to the porous pavement blends may have been insignificant compared to the
overall porous pavement matrix, and second, the amount of heavy metals added to the top of each column
was very low.
Phosphorus
Grass growing on porous pavement would require periodic fertilization in order to maintain healthy growth.
Fertilizer nutrients, especially phosphorus, can be environmental contaminants when present in runoff or
drainage water at high concentrations. For that reason, we looked at phosphorus concentrations in leachate
water, particularly in relation to when the fertilizer was applied. Table 2 shows concentrations of P in the
leachate from each porous pavement blend at times ranging from 5 to 254 days after fertilization. From 5 to
97 days after fertilization, there was a significant difference among porous pavement mixes in the levels of
P in leachate water. Peat moss was the ingredient that had the greatest effect on P leaching. Leachate P
concentrations increased with the amount of peat moss in the porous pavement mix. Zeolite content and
expanded shale particle size had little effect on the amount of P leached from the porous pavement mix.
Time after fertilization also had a significant effect on the amount of P leached. The amount of P leached
decreased with time and by 162 days after fertilizer application, there was no significant difference among
the porous pavement blends. In general, inorganic P concentrations were relatively low in the porous
pavement leachate, suggesting that most of the fertilizer P remained in the porous pavement matrix or was
removed by grass. Sloan et al. (2000) found that expanded shale has a relatively high capacity to adsorb
fertilizer P.
Table 3. Effect of porous pavement ingredients and days after last fertilization application on the P
concentration in leachate water.
Leachate P Concentration
Mix No.
Base
Blend
Peat
Moss
Zeo
5
Days
18
after last fertilizer application
40
97
162
254
(%) (%) (mg/L)
1
2
3
4
5
6
7
8
9
50/50
50/50
50/50
50/50
50/50
50/50
50/50
100LS
100SS
LSD1"
p-level*
0
10
20
0
0
10
20
10
10
0
0
0
10
20
10
20
10
10
0.192
0.244
0.435
0.274
0.414
0.333
1.203
1.299
0.856
0.274
***
0.162
0.115
0.316
0.090
0.225
0.149
0.491
0.468
0.191
0.134
***
0.181
0.141
0.377
0.163
0.295
0.337
0.481
0.403
0.285
0.188
**
0.447
0.129
0.242
0.198
0.416
0.166
0.658
0.353
0.061
0.263
***
0.243
0.278
0.425
0.385
0.427
0.306
0.434
0.419
0.220
0.182
Ns
0.473
0.483
0.563
0.513
0.494
0.613
0.467
0.652
0.719
0.184
ns
t Least significant difference between treatment means.
t Level of significance.
ns, **, *** Not significant or significant at the 0.01 and 0.001 level of probability, respectively.
424
-------
Conclusions
Our study evaluated the ability of 9 porous pavement mixtures to maintain healthy grass growth and to
remove potential contaminants from urban runoff water. Sphagnum peat moss provided the greatest
benefits to plant growth but had little effect on the ability of the porous pavement blends to remove
contaminants from polluted runoff. Zeolites provided little benefit to plant growth, but showed some
potential to remove heavy metals from runoff water. Further testing is needed with higher concentrations of
heavy metals. The expanded shale particle sizes tested in this study had no effect on grass growth and there
was not effect of particle size on the amount of heavy metals leached. Field scale testing of the porous
pavement mixes is needed in order to evaluate their performance under actual environmental conditions and
to begin to develop best management practices for turfgrass growing on porous pavement surfaces. Since
the porous pavement blends are proposed as temporary parking surfaces, engineering tests are needed to
determine load-bearing strengths as it relates to the handling of vehicular weights.
References
Barrett, M.E., L.B. Irish Jr., J.F. Malina Jr., and R.J. Charbeneau. 1998. Characterization of highway
runoff in Austin, Texas, area. J. Environ. Eng. 124:131-137.
Gaffen, D.J. and R.J. Ross. 1999. Climatology and trends of U.S. surface humidity and temperature. J.
Climate. 12:811-823.
Harrison, R. and S. Wilson. 1985. The chemical composition of highway drainage waters I. Major ions and
selected trace metals. Sci. Total Environ. 43:63-77.
Ibrahim, K.M., E. Nasser, T. Deen, and H. Khoury. 2002. Use of natural chabazite-phillipsite tuff in
wastewater treatment from electroplating factories in Jordan. Environ. Geol. 41(5):547-551.
Johnson, K.V. 1999. Sweaty cities learn natural ways to chill out. USA Today. Monday, June 7, 1999.
Legret, M, V. Colandini, C. le Marc, R.S. Hamilton, and R.M. Harrison. 1996. Effects of porous pavement
with reservoir structure on the quality of runoff water and soil. Special issue: Highway and Urban
Pollution. Proceedings of the Fifth International Symposium, Copenhagen, Denmark. Sci. Total Environ.
189-190:335-340.
Lo, C.P., D.A. Quattrochi, and J.C. Luvall. 1997. Application of high-resolution thermal infrared remote
sensing and GIS to assess the urban heat island effect. J. Remote Sensing. 18:287-304.
Morrison, G., D. Revitt, J. Ellis, G. Svensson, and P. Balmer. 1984. Variation of dissolved and suspended
solid heavy metals through an urban hydrograph. Environ. Technol. Letters. 7:313-318.
Mumpton, F.A. 1999. La Roca Magica: uses of natural zeolites in agriculture and industry. Proc. Natl.
Acad. Sci. 96(7):3463-3470.
Nus, J.L. and S.E. Brauen. 1991. Clinoptilolitic zeolites as an amendment for establishment of creeping
bentgrass on sandy media. HortSci. 26:117-119.
425
-------
Olsen, S.R. and L.E. Sommers. 1982. Phosphorus. In Page, A.L. et al. (eds.) Methods of Soil Analysis,
Part 2 - Chemical and Microbiological Properties. 2nd ed. Agronomy Mono. No. 9. p. 413-414, 421-422.
Pratt, C.J., J.D.G. Mantle, and P. A. Schofield. 1989. Urban stormwater reduction and quality improvement
through the use of permeable pavements. Water Sci. Technol. 21:769-778.
Sloan, J. J., W. A. Mackay, and S.W. George. 2000. Growing mediums for porous pavement and rooftop
gardens. In W. Burghardt and C. Dornauf (eds.) Proc. 1st Intl. Conf. Soils of Urban, Industrial, and Traffic
Mining Areas. Universitat-GH, Essen, Essen, Germany.
Sloan, J.J., S.W. George, W.A. Mackay, P. Colbaugh, and S. Feagley. 2002. The suitability of expanded
shale as an amendment for clay soils. HortTech. 12:646-651.
Stotz, G. 1987. Investigations of the properties of the surface water run-off from federal highways in the
FRG Sci. Total Environ. 59:329-337.
Stotz, G. and K. Krauth. 1994. The pollution of effluents from pervious pavements of an experimental
highway section: first results. Sci. Total Environ. 146-147:465-470.
426
-------
THE WASH PROJECT - THINKING OUTSIDE THE CULVERT
Betty Solek,
City of Boulder
Water Quality & Environmental Services
Boulder, Colorado
David Hollingsworth
City of Longmont
Public Works
Longmont, Colorado
The Watershed Approach to Stream Health (WASH) project grew out of a 1999 local storm water round
table. Water quality professionals representing various communities in the Boulder Creek and St. Vrain
Creek watersheds attended these meetings. The group agreed to develop a way to identify storm water
management and data gaps and create consistent storm water quality management approaches throughout
Boulder County, which includes much of the Boulder Creek watershed. This effort was initially funded by
a grant from the U.S. Environmental Protection Agency (EPA).
The primary goal of the WASH project is to implement a regional storm water management program not
only to comply with Phase n regulations but to address broader water quality issues at a watershed scale.
The WASH partners recognize the advantages of creation of cost-effective solutions to storm water
problems through collaboration on compliance with the Phase n Storm Water Regulations. Countywide
collaboration supports and implements the spirit of the watershed approach envisioned in the Federal Phase
n Storm Water Regulations. The project has already enjoyed side benefits of increased communication and
cooperation, and has created a collaborative process for discussing water quality issues.
There are a number of ways in which WASH uses novel approaches to addressing storm water issues. The
WASH Project's implementation strategy provides one example of innovation. WASH Implementation
strategy evolved out of the need to allow flexibility within local jurisdictional boundaries. For instance,
jurisdictional issues relating to local land-use control were considered when developing programs. The
program structure outlined three approaches to collaboration:
1. Shared program elements: common themes and common implementation procedures. An example
would be the development of common ordinance language.
2. Individual programs elements: exclusively the responsibility of individual entities to implement. An
example would be individual community enforcement of an adopted ordinance that contains the
common ordinance language.
3. Shared Program: shared by all entities. An example would be the implementation of one education
program servicing all participating communities.
427
-------
A LOCAL WATERSHED APPROACH
Much has been written about "watershed approaches" to water quality protection. Often these efforts focus
on scientific assessments and technical solutions to problems and issues. However, the relationships
between entities to achieve a watershed approach are just as important and provide the foundation to tackle
technical challenges. In particular, cooperative approaches to compliance with the Phase n Storm Water
Quality Regulations involve internal agreements within an organization, agreements between entities within
a watershed and between the group of entities and the state agency.
The Watershed Approach to Stream Health (WASH) Project has been operating informally since August
1999. It has made significant strides in building cooperative relationships among municipal, county, and
regional water quality professionals in the Boulder Creek and St. Vrain Creek watersheds. The WASH
project developed its own unique solution in order to share Phase n programs. The process of developing
these programs and the benefits and challenges of program development are described. The WASH process
provides an example of a compliance strategy that builds on existing innovative local programs and
agreements to create a program that fits local conditions. This paper describes the process of developing the
WASH collaborative approach. It also provides a summary of lessons learned from this process that WASH
participants hope will be helpful to other efforts.
THE WASH PROCESS
During a storm water round table in April of 1999 and a subsequent focus group in August 1999, it was
discovered that there are many gaps in Boulder County storm water data. Initially, the WASH Project
provided a forum for Boulder County water quality professionals to identify these data gaps, create
workable solutions for filling these gaps, and begin to implement a countywide system of sharing and using
storm water quality data to improve water quality in Boulder County.
EPA funding was provided for the WASH project in fiscal year 1999 under the 104(b)(3) grant program.
Boulder County used these funds to conduct a workshop on watershed approaches to water quality issues.
The grant was also used to facilitate initial exploratory meetings of potential county partners. In the initial
WASH meetings, a working agreement and work plan were created between the WASH partners.
The grant also funded a workshop on watershed management presented by the Center for Watershed
Protection. This workshop provided information about storm water quality problems and created greater
understanding of the issues associated with watershed management strategies. Facilitated meetings of
county entities to explore the potential benefits of a watershed approach to storm water permitting followed
this workshop. County entities have a history of cooperative, intergovernmental approaches to land
management but cooperation on water issues has sometimes been lacking and at times contentious. Thus,
facilitation of this discussion was key to identifying common ground and starting the process of developing
cooperative programs.
During the first WASH Project meeting, participants came together to discuss the potential benefits of
working together and resources that each municipality brings to the table. Participants agreed that in
working together, communities would benefit from sharing data, resources, programs, and ideas.
Participants were also interested in presenting the public, elected officials and developers with a unified
storm water quality message from all Boulder County municipalities.
428
-------
Participants saw Boulder County municipalities collaborating by:
> Sharing monitoring and data
> Sharing development standards
> Creating a model for accomplishing standards in the basin
> Being unified in defining incentives
> Cooperating between agencies
> Being a model community with respect to water quality
The participants agreed to initiate monthly meetings to explore the opportunities for cooperation. Over the
course of a few months, the participants agreed that next steps should include:
1. Create a Memorandum of Understanding/Agreement (MOU)
2. Begin sharing data
3. Educate the public
4. Find additional resources
5. Comply with new regulations
Participants also noted the importance of clearly defining problems, solutions, and common ground at each
step in the process, as well as the importance of continuing to build relationships with each other.
Building Common Goals & Objectives
Participants were asked to break into small groups and answer the following questions together:
What would it look like if we were successful?
How would things be different than the current situation?
What are the possibilities for what a plan like this could create/accomplish?
What are our "key leverage areas?"
What specific issues can we focus on to move us forward toward our new vision?
As a result of discussion in the small groups, individuals were asked to jot down thoughts or phrases
regarding their needs, desires, values, and goals with respect to the county's storm water quality. Shared
values and goals identified by individuals included:
> enhance and improve water quality
> get councils and boards to believe in enhanced water quality
> educate self and community on Phase n regulations
> use one anther as resources; collaborate
Individuals then formed small groups to find any overlap among their shared values and goals and small
groups formulated language to describe overlaps. Key words which described the overlaps included:
> stream health
> cost effective
> water quality programs
> improve and protect water quality
429
-------
• storm water management
• watershed approach
• clearinghouse for education efforts
From these key words, the WASH Project partners developed the following goal and objectives:
Goal: Develop a cost-effective watershed approach to enhance and improve water quality through storm
water management to protect public and environmental health.
Objectives:
> Develop common storm water education programs to raise public awareness and increase public
participation in water quality protection.
> Coordinate training and inspection programs for erosion control.
> Coordinate implementation of Best Management Practices (BMPs) to mitigate impacts of storm water
runoff.
> Share and coordinate resources to monitor storm water quality throughout the Boulder County
watersheds.
> Develop common Phase n programs to ensure cost-effective compliance strategies for WASH
communities.
> Provide a forum for coordination of storm water quality concerns and related watershed issues.
The facilitator also led the group in thinking through an agreement regarding the operating ground rules for
the project. Following is the working agreement for the project developed by Boulder County participants.
WASH Project Partners Are:
> Dedicated to the stated goal and objectives of the project.
> Active participants, attending meetings and voicing opinions equally.
> Willing to share resources and data.
> Clear about their agency' s needs and interest in participating in the proj ect.
> Completing the bulk of WASH Project work in subgroups.
WASH Project Partners Will:
> Be prompt to meetings and participate to the highest level of their ability.
> Maintain focus, prioritize all actions, and encourage involvement of all.
> Understand that not all communities have the resources to attend every meeting.
> Complete assigned tasks that are agreed upon in the group.
> Stay informed about discussions and decisions that take place at WASH meetings in their absence.
WASH Project Partners Are:
> Participating in good faith and working towards the identified common goal and objectives.
> Committed to the protection of water quality within the Boulder Creek and St. Vrain River watersheds.
430
-------
> Committed to sharing information and resources with other WASH partners.
> Committed to developing strategies and solution that benefit the general public and represent the shared
goal and objectives of the WASH Project.
WASH Project Decisions:
> Will be discussed in an organized manner and the process will be open to all.
> Will be made by consensus, an approach to find an inclusive solution that everyone can support.
WASH Project Partners:
> Understand compromise may be necessary to reach WASH common goals.
> Show a commitment to mediate disagreements.
The size of the communities involved in WASH varies considerably and the working agreement
acknowledged the variable resource pool available due to size differences. It allowed small Boulder County
communities to remain involved without committing scarce personnel resources. This was valuable since
initially it was not clear that the smaller communities would be designated by the state for compliance with
the Phase n permit requirements. When the state finally designated these communities, the smaller
communities were linked to the WASH project and the groundwork had already been laid to include them in
the project as Partners.
Memorandum of Understanding: The Power of Non-Binding Agreements
Early in the process, WASH participants recognized that the six minimum control measures (MCM's) were
especially suitable for sharing resources between communities. Thus, exploration of the possibility seemed
realistic and appropriate. A MOU was created to document the willingness of the entities involved in the
WASH project to explore a watershed approach to compliance with the Phase n permit requirements. The
MOU was intended to explain to community decision makers the importance of protecting county streams
through a watershed approach. The agreement also pointed to the connection between watershed protection
and the opportunity that the Phase n regulations represented. The agreement was a non-binding agreement.
However, it created a vehicle for senior management to endorse commitment of staff resources to this
approach. The MOU provided formal support for the WASH goals and objectives and the working
agreement developed during early WASH work sessions. It also laid the groundwork for development of a
formal intergovernmental agreement.
The MOU signature process presented further opportunities to educate decision makers. Senior
management of Boulder County entities were informed about Phase n Storm Water Regulations and the
benefits of a cooperative, cost effective approach to compliance. Ultimately, a year later, the MOU was
signed by the majority of the original WASH participants. Actual signature of the document provided
experience in the logistics, which will be useful when a formal agreement is signed.
Subgroups: The Real Workhorse of the Process
The WASH participants agreed that meeting once per month for a half-day meeting was a realistic time
commitment; however, it quickly became apparent that in order for work products, such as the MOU, to be
completed, more frequent meetings of smaller subgroups were needed. During the first year, the subgroups
focused on the following tasks:
431
-------
> MOU: develop MOU and obtain signatures
> Data: inform WASH Project partners of available storm water resources that can be shared
throughout Boulder County
> Education: create widespread awareness of water quality issues including implementation of web
page, brochures, media products, school materials and presentations
> Additional Resources: explore available and applicable funding and resources possibilities in order
to secure additional resources for the WASH project
> Regulations: inform and educate the WASH project partners about Phase n storm water regulations.
This group also investigated the options for a cooperative permit arrangement under the state of
Colorado's permit system.
Initially, these work groups focused primarily on gathering information and educating the WASH
participants about many issues.
During the second year, one of the most important decisions made by the group was to coordinate
compliance under the Phase n Storm Water Regulations. Implementation of the following six "Minimum
Control Measures" is required under the Phase II Stormwater Quality Regulations :
1. Public education and outreach on stormwater impacts
2. Public involvement and participation
3. Illicit connections and discharge detection and elimination
4. Construction site stormwater runoff control
5. Post construction stormwater management in development and redevelopment
6. Pollution prevention and good housekeeping in municipal operations
As a result of this decision, the participants re-organized into three workgroups, each workgroup taking on
the task of developing two of the above six MCM's called for in the Phase II Storm Water Regulations.
These three workgroups each tackled two of the six MCM's as follows:
> Pollution Prevention and Good Housekeeping
> Construction and Post Construction
> Education and Public Involvement
The WASH participants recognized the need for an organized effort to track the progress of the workgroups
and prepare an overall schedule for the WASH project in order to coordinate submittal of a joint application
for a Storm Water permit. A WASH Project Steering Committee was formed which included
representatives from three of the largest jurisdictions in the Boulder Creek/St. Vrain Creek watersheds.
These include Boulder County, the city of Boulder and the city of Longmont. The Steering Committee was
charged with planning and oversight of the overall WASH Project. Additionally, the Steering Committee
developed a schedule for WASH Project activities leading up to storm water permit submittal in March
2003.
The workgroups allowed an interested group of participants to focus on a key aspect of the process. The
flexibility of the workgroup tasks allowed the project to progress by making the most of available
432
-------
personnel. The WASH Programs that resulted from workgroup efforts ultimately became the foundation of
the WASH Plan.
Technical Panels: Educating Ourselves
WASH Project participants organized and attended a series of panel discussions. The WASH Steering
Committee invited technical experts to speak on these panels at the WASH general meetings. The panel
discussions served to educate all of the WASH participants on the complex issues of storm water quality.
These panel presentations began in November 2000 and continued through April 2001.
In May, after the completion of the panel presentations, the WASH Project partners considered all
information, which had been gained as a result of the panels. The WASH Project partners answered the
question: What specific storm water problems will the WASH Project address?
After much group discussion, those present agreed that urbanization is the underlying cause of increased
and undesirable storm water runoff issues. While halting urbanization is neither desirable nor practical,
urbanization can be accomplished in ways that minimize runoff concerns. In urbanized areas, storm water
quality and quantity has been impacted and is different than in non-urbanized areas. The group agreed that
by addressing four distinct, yet interrelated areas, the WASH Project could lessen the impact of storm water
runoff. The four focus areas are:
1. Sediment
2. Nutrients
3. Spills
4. Erosion
The WASH participants agreed that the WASH project would develop programs to mitigate the impacts of
urbanization on the quantity and quality of storm water runoff. This includes the development of programs
that address sediment and nutrient loading, illicit discharges (spills), and erosion. The WASH participants
agreed that programs would focus on prevention rather than treatment and be easy to implement,
enforceable, and cost effective. The WASH Project focus was integrated into draft program proposals under
development in each of the workgroups.
Management Transition
Initially, the Boulder County WASH Project consisted of a group of county and city staff, representatives of
non-governmental organizations, university researchers and the regional flood management agency. This
diverse group of representatives might have encountered difficulty in coordinating decisions and steps
needed to make the WASH proj ect a reality. The EPA grant provided the funding to hire a county
facilitator. This facilitator provided a focus for group activities and was a tremendous organizational
resource as the group worked through common goals and agreements. Facilitation was also key to
developing relationships between WASH participants as the group developed an identity and focus.
The WASH participants recognized that the management system developed over the history of the project
was working well. This management system was reflected in the Intergovernmental Agreement (IGA).
Part of this management system included establishment of a WASH Project Coordinator to track budgets,
program development and permit compliance. This position reflected the importance of the role of the
433
-------
facilitator in the evolution of the WASH partnership. The Steering Committee also became a formal part of
the WASH management system and was incorporated into the IGA.
Over the first two years, the facilitator essentially served as the WASH Project Coordinator. While not
initially recognized by the group, the skills and background of a facilitator are substantially different from a
project manager. This became apparent as the relationship between the WASH partners became formalized
and the skills and focus needed for management of the group changed. The need for different management
skills and resulted in a shift of project personnel.
Other shifts in organizational needs also came to light. The informal contribution of staff resources began
to shift towards commitment of financial resources for additional WASH staff and consulting resources.
This transition time involved some uncomfortable discussions and changes in personnel. In retrospect, this
transition from informal to formal organization is predictable and is likely to continue as the group
continues to progress towards a formal permit arrangement.
GOING BEYOND THE MINIMUM: HOW?
Building on Previous Successes
The history of storm water quality management in Boulder County provides an important foundation for the
development of the WASH Program. A number of innovative and progressive programs were developed
before implementation of the storm water regulations. These programs were already applied regionally
through the county and local school district. The existence of these programs quickly was recognized as a
resource for development of WASH programs for compliance with the storm water regulations.
In 1989, it appeared that the Storm Water regulations were to be finalized. In anticipation of those
regulations, the city of Boulder established a Storm Water Quality Program; however, the Storm Water
Quality regulations were not actually finalized until 10 years later. The experience and expertise developed
during this interim period were an important foundation for the WASH project.
The city Storm Water Quality program developed and implemented an award winning watershed education
program, WatershED. WatershED was developed in cooperation with the Boulder Valley School district
and a local watershed organization. The teacher training in the curriculum includes:
> Information on the local watershed
> Classroom and water quality monitoring activities
Community action programs were also developed:
> Storm Drain stenciling
> Raise and release of native species
> Adopt A Stream
The Watershed Outreach program gives adults and kids proactive means to protect, conserve and improve
community water. This program was incorporated into the WASH Education program with plans to expand
application to another school district located in the Boulder Creek watershed.
434
-------
Additionally, the city of Boulder and Boulder County have cooperated to develop the Partners for a Clean
Environment (PACE) Program. The PACE program offers a voluntary certification of good business
practices for environmental protection. To become PACE-certified, businesses must meet industry-specific
criteria that reduce hazardous materials and pollution from their routine operations. The certification
involves:
> Inspection of business activities for their impacts on the environment.
> Documentation of current business practices which are protective of the environment
> Recommendations to improve practices
> Certification of implementation of protective practices
> Placard announcing PACE certification
Over the years, this certification program has been extended to public entities in addition to businesses.
The WASH partners are building on this existing program. Storm water quality protection will be added to
the PACE programs. The certification will be extended to all WASH partner municipal and county
operations for the WASH Pollution Prevention and Good Housekeeping programs.
Cooperation as Innovation
In the latter part of 2001, the WASH Project subcommittees completed the proposed programs for each of
the six minimum controls measures required by the Phase n Storm Water Regulations. These proposed
programs were summarized in tables that outlined the following program components:
> Required Minimum Control Measure
> Program Goals
> Regulatory Compliance
> Community Standards
> Local and National Existing Resources
> Best Management Practices (BMP) Selection
> Implementation Strategy
> Coordination and Responsible Agencies
> Estimated Costs and Funding Options
> Measurable Goals
> Implementation Schedule
Within each of these programs, shared elements, shared programs and individual programs were identified.
This approach was developed in recognition of the extent of shared programs that was possible. The WASH
implementation strategy evolved out of the need to allow flexibility within the structure of local
jurisdictional boundaries. For instance, jurisdictional issues relating to local land-use control were
considered when developing programs. The program structure outlined three approaches to collaboration as
follows:
1. Shared program elements: common themes and common implementation procedures. An example
would be the development of common ordinance language.
2. Individual programs elements: exclusively the responsibility of individual entities to implement. An
example would be individual community enforcement of an adopted ordinance that contains the
common ordinance language.
435
-------
3 . Shared Program: shared by all entities. An example would be the implementation of one education
program servicing all participating communities.
The following graphic shows the relationship between these program elements:
/Approach to ^tream
Program Structure
(W/Vf3J~~1)
INDIVIDUAL
PROGAMS
COMMON
ELEMENTS
,—^ S>K%X
SHARED %\VV
PROGAMS
Public Education
-School Based
- Community Based
Goodhouskeeping
- PACE Certification
436
-------
Sharing programs was recognized as particularly challenging for the Construction and Post Construction
programs. The group recognized the challenge of coordinating these two programs in particular due to the
variable approaches in regulations and community philosophies. It would not be politically feasible or
practical for an entity to relinquish jurisdiction for inspections and approval of development plans.
The work group researched types of ordinances, guidance and enforcement resources currently in place in
each community. This provided background on similarities and differences between communities in
existing programs. The group identified the following differences:
> Status and patterns of community land development
> Varying levels of funding and resources
> Approaches to storm water quality management
Common elements identified included:
> Guidance manuals
> Challenges in inspection and enforcement
> Management approaches to open space and stream buffers
This analysis allowed the group to realistically identify potential areas of co-operation and sharing between
WASH Project partners.
Elements that could be shared included common ordinance language and minimum inspection and
enforcement procedures. It was agreed that sharing these elements would create a consistent regulatory
environment for businesses in Boulder County. The added benefit of enacting consistent regulations across
the county could be expected to protect the health of the Boulder Creek watershed. The common elements
of the regulations still allow a regional approach to erosion control and stream protection.
The group agreed that adding a certification of erosion control training to the Construction Program would
be an appropriate way to ensure consistency of application of erosion control standards throughout the
county. This certification is not a required element of the storm water regulations but was recognized as a
cost effective approach to supplement inspection resources available to WASH entities.
The proposed WASH program structure is an innovative, local response that allows maximum sharing of
resources for those programs that are readily shared but retains the ability of local jurisdictions to implement
their regulations and standards. This flexibility was important for WASH participants, allowing for regional
cooperation and maintaining local autonomy. WASH participants recognized that cooperative programs
and a regional approach was, in itself, going beyond regulatory requirements.
BENEFITS
The WASH participants recognize and have reaped the benefits of a regional, watershed approach during
the three years of program development. The watershed approach employed by the WASH participants as a
compliance strategy has generated grant income to support and advance the project. After March 2003,
development of programs will no longer be eligible for grant funding because the programs will be
considered regulatory requirements. However, since a regional approach is not a regulatory requirement,
the WASH participants are hopeful the project will continue to attract grant funding.
437
-------
More importantly, stream protection benefits are anticipated from the regional, watershed approach. The
application of common regulatory requirements will allow for consistent standards to be applied to business,
public and construction activities throughout the county. This reduces the potential for one entity to apply
lower standards in one portion of a watershed, perhaps undoing the benefits achieved by another entity
applying protective standards in another portion of the watershed.
The complete sharing of the WASH Education program is anticipated to provide similar benefits. It is
hoped that the power of a consistent message and look from the WASH program will capture the public's
attention. This is particularly important given the nature of non-point pollution sources that are literally in
everyone's "backyard."
WASH participants have already reaped the benefits of sharing personnel, experience and expertise during
development of the WASH programs. The collaborative nature of the process has multiplied the resources
available to each entity for development of a permit application. A comparison of the resources available
within each entity versus the combined resource base of all county entities quickly shows the power of
combining resources.
During the development of the WASH budget for the proposed programs, the WASH consultant's research
indicated that a cost savings of 25 percent to 30 percent for program costs could be expected from a
collaborative approach. This was confirmed by an analysis which indicated a 25 percent cost savings could
be expected by a selected WASH entity.
Further benefits are anticipated from the expansion of innovative existing programs that have already
achieved substantial recognition. These programs have been tested and gained the benefit of experience.
The programs are now well positioned for expansion.
LESSONS LEARNED
WASH participants have learned a lot of lessons over the course of the project's evolution.
Be Flexible-Adjust Directions
The evolution of a program can lead in many directions and there are many ways to achieve the same result.
Be flexible in order to take advantage of innovative ideas and directions that produce a program that is
appropriate for local needs.
Goals are Key-Be Firm
Achievement of collaboration and a common approach may seem unrealistic in the face of individual
regulatory systems. Detailed examination of the components of various options can yield unexpected
opportunities. Commitment to agreed upon goals and objectives facilitates is key to progress through these
challenges.
Money-Rubber Meets the Road!
The level of scrutiny of proposals increases when it is time to make financial commitments. Factor in the
necessary time and energy to address this additional scrutiny. Additional time will often be required when it
seems that development of the program components is final. The commitment of each jurisdiction to the
process will be tested as the budget is finalized.
438
-------
It Takes Time
The process of collaboration takes time. It is common to experience a long period for development of a
program within one jurisdiction. That time period should be at least doubled for development of a regional
program.
Patience — Don't Force Square Peg into Round Hole
The time required to develop these collaborative approaches dictates the need for patience during the
process. Don't frustrate your efforts further by being rigid. There are many options and it is important to
choose those options that work well for your particular group of organizations and individuals.
439
-------
NECESSITY AND OPPORTUNITY: URBAN STORMWATER MANAGEMENT IN
ROCKVTLLE, MARYLAND
Lise Soukup, P.E.
Department of Public Works
Rockville, Maryland
The City of Rockville, with 50,000 residents and substantial areas of commercial and office development, is
located in the Maryland suburbs of Washington, D.C. Rockville began a building boom in the 1940s that
continues today (City of Rockville Planning Commission, 2002). The Mayor and City Council encourage
residents to take ownership in their local government, and the City prides itself on being responsive to their
needs as much as possible. The City government is committed to "enhancing the quality of life in Rockville
by providing premium services in response to the needs of everyone who visits, works, and lives in our
city", according to the City's mission statement.
Much of Rockville was built prior to stormwater management (SWM) requirements. Many existing
stormwater management systems are ineffective or undersized by today's standards. The resulting riparian
tree loss, stream erosion, siltation and struggling aquatic species in the City's streams indicate that
stormwater management is an ongoing process that continually needs fine-tuning.
Rockville's Department of Public Works (DPW) has 25 years of experience with comprehensive watershed
management, beginning with the first SWM ordinance in the State of Maryland. Current City law and
regulations, which mirror the State's requirements, provide for stringent water quality and quantity control
for new development or redevelopment. They also support a strong public stormwater retrofit and stream
restoration program. DPW is challenged with creating practical and effective watershed management plans
for existing development in a city that is 87% built out. DPW also must demonstrate to residents that the
proposed solutions are achievable, effective, safe, attractive, compatible with many other neighborhood
needs, and above all, necessary.
Rockville's Watershed Management Plans
The purpose of the watershed management plans are to make the City's stream corridors environmentally
stable and enjoyable for residents, and to mitigate Rockville's nonpoint source effects on downstream
conditions in the Potomac River and the Chesapeake Bay. These plans recommend projects for subsequent
Capital Improvement Program (CIP) implementation that will make a substantial difference to local stream
conditions. To work in Rockville, these need to be politically as well as technically viable. The City's
watershed management strategy has evolved into a flexible, opportunistic approach that matches available
funding, developers, and complementary projects to needed watershed improvements. The plans also
involve stakeholders to an unprecedented degree.
Over the last six years, DPW completed watershed management plan studies for the City's three
watersheds, each more detailed and comprehensive than the last (Figure 1). Each had stream inventories of
aquatic conditions and an opportunities assessment to identify possible SWM improvements and stream
restoration sites, and each resulted in projects now being implemented through the City's CIP. The
complexity and controversy of the public process varied greatly, however. Residents often had different
opinions about stream problems, solutions and acceptable trade-offs, most notably in the last plan.
440
-------
Watts Branch
Rock Creek
Figure 1. City of Rockville Watersheds
The Watts Branch Watershed Study (Center for Watershed Protection and City of Rockville, 2001) was a
lightening rod for controversy. The area had 4,000 acres of residential, office and highway uses, and two
major mixed-use developments pending in the headwaters as the study commenced. Vocal residents were
protective of their parks and distrustful of the City's environmental judgement in previous projects. To
many, the stream problems lay with the newcomers building upstream, not with their own 30-year old
developments. Still, they wanted solutions to the acknowledged erosion and water quality problems through
Watts Branch Stream Valley Park, the City's largest natural area. Table 1 presents data on the City as a
whole and on the Watts Branch.
Size1
CITY
OF ROCKVILLE
Population2
Land
Use3
Residential
Mixed Use
Office
Industrial
Retail
FACTS
13.3 mi2
47,388
73%
12%
7%
4%
4%
Drainage Area
WATTS
1
BRANCH FACTS
5
Watershed Imperviousness4
Watts Branch
Watts Branch
Streams4
Streams in
parkland4
18.7
7
.9 mi2
28%
miles
miles
1: City of Rockville CIS
2: Census 2000
3: City of Rockville Planning Commission, 2002
4: Center for Watershed Protection and City of Rockville, 2001
441
-------
Essential Public Process
In previous watershed studies, the City began by studying technical issues. Resident involvement came
towards the end of the process when there were recommendations to react to. For Watts Branch, the City
needed much earlier involvement and better communication.
Before DPW developed the watershed study's scope, it held a public meeting to solicit the residents'
watershed and neighborhood concerns. Afterward, staff invited attendees and other stakeholders to join
staff for regular meetings to review the study and deal with community concerns about balancing tree loss,
appearance, safety and recreation needs against watershed improvements. The City also asked civic
associations and developers to send representatives. The resulting Watts Branch Partnership was comprised
of residents from across the watershed, City staff from the Recreation and Parks Department, the Planning
Department and DPW, and eventually the consultants. The City Manager's Office had recently established
the new Project Implementation Coordinator position to manage the public process for all City projects.
This person served as a facilitator at Partnership meetings, and focused on keeping discussions within the
ground rules and staying on the agenda. Table 2 lists the stakeholders invited to join; business and
development interests did not participate, but residents and institutional agencies were very involved.
Table 2. Watts Branch Watershed Stakeholders
Non-agency Stakeholders Agency Stakeholders
Homeowners Association(s) Rockville Recreation and Parks Departments
Civic Associations Rockville Public Works Department
Watts Branch Partnership Rockville City Forester
Developers (e.g., King and Thomas Farms) Rockville Environmental Specialist
Watershed Property Owners State and Federal Regulatory Agencies
Business Interests (industrial, commercial Gas, Oil and Utility Companies
business owners) Montgomery County Public Schools
Montgomery College Rockville Mayor & Council
Lakewood Country Club
Center for Watershed Protection and City of Rockville, 2001
The Partnership's first task was to review the scope of the watershed study. Staff incorporated most
suggestions, then had a Partnership resident participate in the consultant selection. The Center for
Watershed Protection was selected because of their innovative watershed management approach and
experience with local governments. The Center teamed with a local engineering firm and an environmental
resource assessment firm to augment their staff (primarily in surveying, stream inventory, and some concept
designs).
The Partnership met monthly or more often for two and a half years. City staff set agendas for the meetings
and the study schedule, and evaluated and summarized technical information and study results for the
Partnership. The Partnership's resident members acted as liaisons between their civic associations and the
City to convey opinions and explain projects, attended lectures to learn about current SWM and stream
protection practices, and reviewed drafts of the study report. Partnership members visited existing City
SWM facilities and stream restoration sites to see marshes, bio-engineering and gabions that had been in
operation for several years.
They used their new knowledge to evaluate the consultant's analysis and plans. Project details mattered
greatly to these members, even seemingly small things. DPW incorporated their advice and comments
442
-------
wherever feasible, and explained the staffs reasons when we disagreed. This process helped assure the
residents that their involvement was productive. It resulted in better integration of important issues for both
the residents and the City in the final results, rather than each side losing essential features or issues. It also
offered a sense of fairness that is absolutely necessary to belief in good government - even if the residents
did not always get what they wanted, they agreed that the study was fair and reasonable.
The City needed residents to support the management plan. It was not only politically difficult to get a
controversial set of recommendations adopted, but also complaints of inaccuracies, unresponsiveness and
unfairness would cloud unrelated City projects. To demonstrate the City's commitment to working with the
residents, staff tried a new process. In 1997, the City had begun training all employees on a new process
called Citizen Participation by Objectives (Bleiker, 1995). This process demands that the City convey to all
potentially affected interests, or stakeholders, that:
1. There is a serious problem or an important opportunity that must be addressed;
2. The City is the right entity to address it, and that it would be irresponsible for us to ignore it;
3. Our approach is reasonable, sensible and responsible; and
4. We are listening and we care about the costs, the negative effects or the hardships that our actions will
cause people.
The Citizen Participation by Objectives approach was time consuming but worthwhile. DPW did not
abdicate its responsibility to manage the watershed study or give in on controversial projects. However,
staff tried to look at the decision-making process from the residents' point of view as well as from the
City's. Sometimes, the staff would argue for a worthy project where the benefits were particularly helpful
and the negatives could be overcome or minimized to suit most of the affected people. The Partnership
generally saw the same thing and helped design improvements to overcome neighborhood concerns. They
advocated the projects and the goals of the Watts Branch study in discussions with their civic associations.
This was difficult for some people since they were sometimes viewed as 'selling out', or were caught
between displeased neighbors and the City. Nevertheless, the Partnership maintained representation from
thirteen out of twenty-one neighborhoods within the watershed. Neighborhoods containing stream valleys
or with potential SWM projects tended to participate more. Meetings typically had ten to sixteen residents
in attendance.
The Partnership did not vote on decisions. It was explained at the beginning of the study that this would be
an effort to uncover opinions and concerns, and to look at all reasonable alternatives within the confines of
the study assumptions. The Partnership would seek consensus where possible, but dissent was also
acceptable. Staff emphasized that the Mayor and Council were the final arbiters of the management plan
recommendations, and that the study would try to fairly present both pros and cons of proposed projects. At
most key decision points, after discussion had elicited all viewpoints, the large majority of resident
members agreed on their recommendations. Those who held opposing positions seemed satisfied that their
concerns would be recorded in the final study to be further evaluated when the individual project moved
into final design.
In addition to educating the Partnership members, the City also shared the study with the larger Watts
Branch community. The Center for Watershed Protection hosted a charette, which was sponsored by the
Partnership, for the public early in the study to present findings of existing conditions. Charette participants
tried watershed management activities such as creating an educational campaign and designing SWM for
several sites. Staff held a month-long Open House to present project concepts and information about the
443
-------
study, which were also posted on the City's website for the remainder of the study. Notification postcards
were mailed to all homeowners near proposed projects so they would be aware of the study
recommendations. Partnership members paired with DPW staff at their own civic association meetings
where proposals for local projects were explained. The presence of a neighborhood member who had
worked with the City on the study recommendations proved invaluable. With the Partnership in attendance,
the Mayor and Council adopted the Watts Branch Watershed Management Plan in 2001.
Watershed Study Methods
The Watts Branch Study uses the Rapid Watershed Planning Handbook (Center for Watershed Protection,
1998) methods to predict future watershed conditions based on impervious cover, set realistic and
measurable goals, and assess whether improvements are working. This generates recommendations based
on defensible science and measurement. It emphasizes local commitment by requiring community
involvement and an implementation plan adequate to carry out the recommendations. Figure 2 illustrates
milestones in the study.
S ~\ S~
July, 1998- Begin
Public Discussions
and Assemble
Staff Team
V J V
October, 1999-
Public Charette for
Stormwater
Management Options
J
1998
February, 1999-
Consultant
Begins Study
1999
November, 1999-
~> Evaluation of
Initial SWM and
Stream Project
Inventories
J V
Summer,
House o
Storr
Manage
Stream
V
2000
\
/
2000- Open | [ August, 2001-
F Proposed Mayor & Council
nwater Adoption
ment and
J
2001
January, 2001-
Final Report
Presented to
Mayor & Council
Figure 2. Watts Branch Watershed Study Timeline (Center for Watershed Protection and City of Rockville, 2001)
Phase I of the study consisted of the initial data gathering and analysis, leading to a list of needs and
opportunities. Rockville was fortunate to have recent GIS-based topographic, property and utility
information for the entire city, and 2' contour topography and tree surveys for almost all parks. The
consultants did an RSAT (Rapid Stream Assessment Technique) survey of stream habitat and physical
conditions at 400-foot intervals to assess the general health and level of erosion in Watts Branch and its
tributaries (Galli, 1996). Potential and existing SWM facilities around the watershed were screened by
drainage area and capacity, effectiveness and feasibility of modernization. They were field-checked to
evaluate natural resource constraints and expansion concerns.
Data from a Rapid Geomorphic Assessment evaluated physical parameters related to channel widening,
downcutting and accretion (Center for Watershed Protection and MacRae, 1999). Based on this data and
historic cross-sections from the 1950s- 1960s, a new technique developed by the Center for Watershed
Protection was used to predict the ultimate size of the channel at various points. It correlated the pre-
urbanization and current channel cross-sectional area to imperviousness changes in the sub-watersheds, then
444
-------
predicts final stream sections for the built-out imperviousness after factoring in the stream's response time
(Caraco, 2000). This was considered to be more accurate than short-term monitoring with bank pins.
After the staff and Partnership members evaluated and prioritized the Phase I results, a list emerged of the
most promising SWM opportunities and the most significant reaches of stream erosion. Phase n produced a
30% engineering concept design for each of these projects. The SWM concepts provided basic hydrologic
and sizing computations, a conceptual grading plan that included maintenance access and limits of
disturbance, and a count of significant trees (>12" DBH) that would be removed by the proposed project.
Stream concept plans showed proposed restoration techniques, including rock vanes, step pools, coir fiber
logs, bank laybacks and planting, and imbricated rip-rap or gabions. Stream plans also showed the limits of
disturbance for access paths, stockpiles, and construction to give the Partnership a better sense of whether
the stabilization justified the disturbance and tree loss. On several projects, the consultants were asked for
alternate SWM concepts to explore Partnership requests that would reduce tree loss or relocate the footprint.
During Phase n, the City met with representatives from Maryland Department of the Environment and the
Army Corps of Engineers to consider wetland and waterway permitting issues. Their comments resulted in
abandonment of one SWM concept and revisions to several others to better protect existing wetlands and
maintain streams through the proposed ponds. The regulatory agencies were very supportive of the
management plan's intent to mitigate a developed watershed, and helped identify permitting constraints and
acceptable alternatives during the concept process. This is expected to facilitate the later project design
stage when Section 401-404 permits will be sought.
Phase in focused on watershed-wide issues. Several Partnership meetings were devoted to discussing
members' views on environmental education, watershed outreach and effective ways to change behavior in
residents and businesses. The Center for Watershed Protection developed a schematic education/outreach
approach based on research into other successful programs (Schueler, 2000a, 2000b). The Center also
produced a map of wetland enhancement and forestation opportunity sites that staff will integrate either
with specific stream restoration/SWM CIP projects or through developer obligations under the City's Forest
Conservation and SWM ordinances. These and other non-structural watershed rehabilitation strategies will
be implemented across Rockville in the next few years through the City's upcoming National Pollutant
Discharge Elimination System - Phase II (NPDES-II) permit requirements.
Study Assumptions
City staff specified numerous study assumptions that shaped the solutions. The City Department of Parks
provided parameters such as no net loss of active playing fields or other recreation features due to SWM or
stream projects. The City Forester and Environmental Specialist specified access paths and helped
characterize forest and wetland resources to avoid extensive impacts. For cost-effectiveness, DPW chose 25
acres as a desired minimum drainage area for retrofit consideration, although a few opportunities for small
facilities were also evaluated. This limitation automatically reduced feasible SWM choices to various forms
of ponds and marshes. Bioretention, surface sand filters and underground pipe storage become impractical
with drainage areas larger than a few acres, although the City regularly uses these methods for smaller sites.
With erosion and riparian tree loss topping the list of community concerns, water quantity control became
the most important SWM parameter to address on a comprehensive scale. Therefore, it was decided in
consultation with the Center that the first priority would be to achieve 100% of the Channel Protection
Volume (i.e., 1-year, 24-hour extended detention control) in a facility. This has been designated by the
445
-------
State of Maryland as the most critical SWM control for preventing downstream erosion (Maryland
Department of the Environment and CWP, 2000). Water quality treatment was also included to the
maximum extent feasible. About half of the recommended SWM sites could accommodate 100% of the
water quality volume for 0.5" of runoff over the watershed area, which was consistent with the City's water
quality standards in 2000 and deemed reasonable for a retrofit situation (Center for Watershed Protection
and City of Rockville, 2001). One inch of water quality treatment was not practical due to storage
limitations.
Stream erosion problems were found in almost all tributaries and throughout the mainstem. To help
prioritize these, DPW applied an existing City policy that limits use of City funds to improvements on City
lands. From the City's perspective, these funds should be spent on repairs to the City's first responsibility,
its own parks. For stream reaches owned by private homeowners' associations or residents, this assumption
has caused problems. Even if erosion was significant on these reaches, the City's ranking system
discounted the site, resulting in stream restoration recommendations only for publicly owned streams. The
City is now debating whether this policy can be modified without incurring large and unplanned financial
burdens.
The public process also operated under assumptions. First, staff believed that the Citizen Participation by
Objectives methods would be effective in fostering cooperation and open exchange of ideas with residents,
so that compromise would be achievable. This assumption was generally met, and resulted in high
satisfaction with the study process from both Partnership and non-Partnership residents. Second, staff
assumed that the civic and homeowners' associations were the main conduits to convey information
between residents and the City. This tended to work well in active associations, but was ineffective at
informing communities where neighborhood meetings were informal and infrequent. This gap was partially
filled with the City's publicity and notification process through local mailings, papers, and City Cable TV
shows.
Study Findings and Recommendations
At the end of Phase I, 54 SWM opportunities were considered in Partnership meetings from both the City's
perspective (such as pollutant removal efficiency, capacity to control the drainage area, cost, access and
maintenance burden) and from the community perspective (including appearance, safety concerns, impacts
to trees and to recreation). Since these perspectives often worked at cross-purposes, staff chose a two-
variable system to compare SWM projects. Each project received two scores that were plotted on an x-y
coordinate system to graph the relative values of environmental management vs. community impacts.
Scores reflected that a project could be neutral or negative in a category, as well as positive. Projects that
scored well in both categories were agreed to be worthy of further investigation at the Phase n concept
stage. A few projects that were highly rated in one category and had few negative effects in the other
category also went to concept stage. This method simplified the comparisons while helping the Partnership
visualize distinctions. In all, 18 SWM projects moved forward for Phase n concepts.
Similarly, 62 RSAT sample points, covering 4.7 stream miles, were culled through a ranking system based
on severity and extent of erosion, land ownership and forest impacts. 2.7 miles of stream were selected as
high priority restoration areas.
446
-------
Major Roads
Streams
Priority SWM Facilities
Stream Restoration Areas
Figure 3. Adopted Watershed Projects in Watts Branch (Center for Watershed Protection and City of Rockville,
2001)
447
-------
Some controversial projects moved forward to concept design. The Partnership members agreed that more
information was needed before deciding whether these were viable or not. This aided the concept
evaluation process at the end of Phase II, since the Partnership could then assess questionable projects with
better information.
The City and the Partnership had to balance the impacts of projects against the threat of doing nothing. The
Rapid Geomorphic Assessment projections showed that, as a whole, Watts Branch stream cross-sectional
area may expand to two to four times its existing size over the next 40-50 years as it adjusts to a new state
of equilibrium with the watershed's built-out impervious condition (Center for Watershed Protection and
City of Rockville, 2001). This would lead to continued extensive undermining and toppling of large trees
along most of the stream valley, add more sediment to the stream system, and degraded the biological
activity of the fish and macroinvertebrate populations throughout Watts Branch. Given that the community
was clamoring for the City to do something about sediment-laden streams and undermined trees at the start
of the study, it became clear to the Partnership that the null alternative would not serve the goals. This
made it easier for the Partnership to defend the inevitable tree loss, construction impacts and SWM facility
changes they needed to endorse, and helped the members move onto seeking realistic ways to minimize
these impacts rather than declare them unacceptable.
The projects adopted in the management plan are shown in Figure 3. The Watts Branch Management Plan
established fourteen SWM retrofit projects covering 925 acres of untreated or under-treated development
(roughly 25% of the total watershed), of which eleven would be public facilities. The plan provides four
new SWM facilities and ten modernizations to existing SWM ponds, as well as nine separate stream
restoration projects. Combined with new SWM systems for 700 additional acres of mixed-use development
in the Watts Branch headwaters, this represents effective management of a substantial portion of a built-out
watershed. Over 50% of the watershed will be treated by modern SWM controls of 1-year, 24-hour
extended detention and quality treatment of at least 0.5" runoff.
Problem Projects
Not all projects evaluated in Phase n survived in the final recommendations. In following the Citizen
Participation by Objectives method, staff dropped environmentally valuable projects that might create more
neighborhood problems than they would solve, such as on a potential pond site that would clear a 200 foot
wooded buffer between houses and an interstate highway. Technically, the facility would work; the noise
and visual impacts to the houses facing the site were estimated by City staff to be insurmountable and could
not be adequately mitigated without the State Highway Administration's commitment to a noise wall. The
City maintained credibility by showing that the watershed goals were based not only on environmental
benefits but community benefits as well.
Knowing neighborhood history helped the staff and consultant avoid unnecessary impacts. For example,
the study recommended a new wet pond at a park site that was just receiving a new playground through the
efforts of the Parks Department and a local Girl Scout troop. The proposed pond would necessitate
relocation of the playground. DPW decided to schedule the pond project later in the CIP to coincide with
the expected lifespan of the playground. This would give the community ten years to enj oy their
playground and agree on a satisfactory new location in the same park for the next set of play equipment.
As expected, the most controversial projects were proposals for new ponds in active parks. The College
Gardens Park pond produced a long stalemate between the staff and a neighborhood civic association. This
448
-------
project called for the expansion of a small farm pond to almost three times its current size in a heavily used
urban park. This project was popular with the staff and the Partnership because it was fairly neutral in
community impacts while providing exceptional water quality and quantity benefits for an 89-acre mixed-
use watershed. Although the expansion would remove grassed playing area, all other recreation features
and trails could be retained or relocated.
Several residents, including the civic association president, were polarized against any changes to this park,
and demanded more 'innovative' alternatives be investigated, including underground SWM proprietary
measures and moving the pond downstream into a wooded stream valley. Community opposition
materialized with the first presentation to the civic association and took fourteen months and eight formal
meetings with association representatives before the Watts Branch management plan was finally adopted.
Some of the difficulty came from issues of control as people who were not involved in the Partnership tried
to negotiate separate oversight of the study.
To counter this, the City followed the original methods of Citizen Participation by Objectives, reiterating
the history of the public process. The City also pointed out that several association members had, in fact,
been on the Partnership since the beginning, including the association's president at the start of the Watts
Branch study. DPW also obtained a lengthy alternatives analysis from our consultant in the final months of
the study that investigated the association's requests and demonstrated that there were high costs for
proprietary treatment and wetland/stream impacts for the in-line alternative that proved unacceptable to the
state and federal regulatory authorities.
The project was conditionally recommended in the management plan after an extensive section on benefits
and concerns describing the civic association's issues. At the request of the Mayor and Council, a further
alternatives analysis will be completed before selecting a final design. Since traditional SWM approaches
have already been investigated, staff will use this required evaluation to look at feasibility and
implementation of concepts that were previously outside of the Watts Branch watershed study assumptions.
The alternatives analysis will compare expected benefits and disadvantages from a watershed
education/behavior modification program for residents, businesses and institutions in this community, a
small-scale SWM retrofit program focusing on the high-impervious non-residential uses (about 30% of the
watershed), the management plan's recommended central SWM facility, and stream restoration/storm drain
outfall stabilization. DPW hopes this will help clarify the pros and the cons of each choice to find a solution
that has both reasonable environmental benefits and acceptable public understanding and support. Staff
expects the civic association to be an active participant in this follow-up analysis, much as the Partnership
was for the Watts Branch study. This investigation will also assist DPW in testing approaches for the
NPDES-n requirements.
Post-Study Evaluation of the Public Process
The Partnership's two and a half year review period left enough time for watershed education and gathering
feedback from the participating neighborhoods. Residents were welcome at any time to start attending
meetings, and several active Partnership members joined at the Phase n concept stage. Staff had more
difficulty explaining the study's background, scientific basis and findings to non-Partnership residents &
civic associations in the space of only a few meetings. Most associations and residents were able to
appreciate the validity of the recommendations and agreed to support their local projects. One
neighborhood did not participate at all in public meetings or the Partnership, then protested the proposed
449
-------
project during the last few months before adoption. The City will need to work closely with these residents
when the design stage begins, since they have no previous commitment through the watershed study.
The Partnership members delivered a statement at the final plan's introduction to the Mayor and Council
regarding their support of the management plan process and recommendations. Not all controversy could
be avoided. The Mayor and Council heard opposing views during the eight months between introduction
and adoption of the management plan, but still believed that staff had been fair and objective in making the
recommendations. The fact that only two of the recommended projects drew any negative comments
showed that there was general satisfaction among the stakeholders. Many residents commented that the
projects showed an awareness of collateral neighborhood issues and preserved features important to them.
A year after the Watts Branch Watershed Management Plan was adopted, the Partnership members received
a survey from the City asking for their opinions on the effectiveness of the study process, their satisfaction
with the study's methods and recommendations, and their viewpoint on whether their involvement made a
difference. The responders were extremely pleased with the staffs cooperative efforts and the public
process, citing it as much improved over previous City projects and an example of how government should
work. They recommended that this process be used for other controversial projects. Although some
members felt that solution options were too limited, they agreed that the City had made a valid effort to
explore alternate ideas and the final recommendations were compatible with their neighborhood needs.
They also liked that SWM and stream concepts had been revised to incorporate most of their project-
specific comments.
The public process led to compromise on both parts, a willingness to explore alternatives, and
acknowledgement that not every problem could be solved. Once the members could tie watershed goals to
community goals, or at least balance conflicts between them, many watershed projects became palatable. In
general, residents are much less fearful of the short-term impacts and long-term effects on their quality of
life. The study built credibility and support within the neighborhoods that will be essential as DPW
continues to work with the residents during design and construction.
Implementation - From Paper to Ponds
A watershed management plan will succeed only if it is implemented. In the past decade, DPW has built at
least ten stormwater management retrofit and five stream restoration projects from its watershed studies.
Watts Branch Plan projects on City parkland are proceeding through design and construction in the City's
CIP over a 10-year period. Non-City projects are also advancing through other mechanisms, such as a low-
cost retrofit of a State Highway Administration dry pond in an Interstate-270 interchange that is being
designed and constructed through the Recreation and Parks Department to fulfill its SWM obligation for a
new bike trail. Through private development, dozens of other SWM and stream projects are built and then
turned over to the City to maintain. Although Rockville has had its share of planned SWM projects that
were never built due to changing wetland standards, land constraints or public outcry, the City's long-term
implementation rate is impressive.
Watershed plans are dynamic documents. They guide CIP planning, but DPW also forwards the watershed
goals through cooperative planning with developers and teaming projects that need more immediate
attention. The City's watershed management strategy continues to include a bigger toolbox of private/non-
parks opportunities. Given Rockville's built-out condition, equivalent SWM alternatives such as stream
restoration or stabilization, retrofit of an existing but outdated SWM facility, or control of a different piece
450
-------
of imperviousness on the site (parking lot instead of rooftop) may offer more environmental benefit than a
traditional onsite SWM system. Regular performance monitoring and stream surveys are still needed to
identify the solutions that work and the needs that remain. DPW expects to revisit each watershed
management plan every ten years to evaluate its progress.
The public process continues through the final design and construction phases for individual projects.
Projects in parks or near residences are heavily publicized. Several meetings are held at various points to
get feedback on design details and neighborhood concerns. DPW, the Project Implementation Coordinator
and other staff make sure residents have access to information. Good groundwork at the management plan
level helps to prepare communities for upcoming changes.
The City's dedicated SWM Fund makes the watershed management program self-supporting (Table 3).
Money is primarily collected from monetary contributions collected in lieu of on-site SWM from projects
too small to support their own facilities and, to a lesser extent, from developers' SWM and sediment control
permit fees. The fund supports the operating budget expenditures for maintenance on City-owned SWM
facilities and for DPW staff who review or inspect SWM and sediment control in both private development
and the City's CIP. The fund also covers design and construction of public SWM facilities and stream
restoration, watershed studies, policy planning, and some additional programs that will be needed for the
City's upcoming NPDES-II permit.
The estimated design and construction cost for all of the Watts Branch Management Plan projects is a total
of $2.8 million. Based on a 2000 fiscal analysis, the fund should manage expected costs for the foreseeable
future, including full funding of projects from all three watershed management plans. However, as
development slows with the City's near build-out, a SWM utility fee for residential and business owners
may become necessary. DPW also solicits and receives limited State grant funding for design and
construction of SWM and stream restoration projects.
Table 3. City Stormwater Management Fund
Stormwater Management Fund
Unreserved Fund Balance (FY2002) $5.2 million
Monies Earned (FY97-2000)* $963,000/year
Operating Expenses (FY97-2000)* $290,000/year
Capital Expenses (FY1997-2000)* $550,000/year
City of Rockville Department of Finance, 2002
*Note: Average taken over 4 years for better picture of
income and expenditures over time.
Conclusion
Rockville's watershed management plans have benefited from a dedicated funding source, a compact and
flexible city government, a strong development community, a spirit of teamwork among City staff, and
resident interest in streams and parklands that is reflected by the Mayor and City Council. Problems and
priorities change, so these plans only capture a snapshot in time of watershed conditions. Therefore, DPW
will continue to advance effective and innovative watershed stream protection with a variety of strategies.
In watershed management, everything is an opportunity.
451
-------
References
Bleiker, H. and A.M., 1995. Citizen Participation Handbook for Public Officials and Other Professionals
Serving the Public, 9th ed., Institute for Participatory Management and Planning.
Caraco, D.S., 2000. The Dynamics of Urban Stream Channel Enlargement, Watershed Protection
Techniques, 3(3)-729-734.
City of Rockville Department of Finance, 2002. FY2003 Adopted Budget and FY2003-2008 Adopted
Capital Improvements Program.
City of Rockville Planning Commission, 2002. Comprehensive Master Plan for the City of Rockville,
Planning Commission Draft.
Center for Watershed Protection and City of Rockville, 2001. Watts Branch Watershed Study and
Management Plan Final Report.
Center for Watershed Protection, 1998. Rapid Watershed Planning Handbook.
Center for Watershed Protection and C.R. MacRae, 1999. State of Vermont Watershed Hydrology
Protection and Flood Mitigation Project.
Galli, L, 1996. Rapid Stream Assessment Technique (RSAT) Field Methods, Final Technical Manual,
Metropolitan Washington Council of Governments.
Maryland Department of the Environment and Center for Watershed Protection, 2000. 2000 Maryland
Stormwater Design Manual, Volumes I & n.
Schueler, T., 2000a. On Watershed Education, Watershed Protection Techniques, 3(3)-680-686.
Schueler, T., 2000b. Understanding Watershed Behavior, Watershed Protection Techniques, 3(3)-671-679.
452
-------
RE-INVENTING URBAN HYDROLOGY IN BRITISH COLUMBIA:
RUNOFF VOLUME MANAGEMENT FOR WATERSHED PROTECTION
Kim A Stephens, PEng, KSA Consultants Ltd1, West Vancouver, British Columbia
Ted van der Gulik, PEng, BC Ministry of Agriculture, Food and Fisheries2, Abbotsford , British Columbia
Laura Maclean, Environment Canada3, North Vancouver, British Columbia
Ed von Euw, PEng, Greater Vancouver Regional District4, Burnaby, British Columbia
ABSTRACT
There is a logical link between changes in hydrology and impacts on watershed health, whether those
impacts are in the form of flooding or aquatic habitat degradation. The link is the volume of surface runoff
that is created by human activities as the result of alteration of the natural landscape (i.e., through removal
of soils, vegetation and trees). When trees, vegetation and soils are replaced by roads and buildings, less
rainfall infiltrates into the ground or is taken up by vegetation, which results in more rainfall becoming
surface runoff. The key to protecting urban watershed health is to maintain the water balance as close to the
natural condition as is achievable and feasible by preserving and/or restoring soils, vegetation and trees.
But accomplishing this requires major changes in the way we approach urban drainage and in the way we
develop land. Drainage engineers have traditionally thought of reconciling pre- and post-development
runoff in terms of flow rates, not volumes. At the site level, however, we need to focus on how much
rainfall volume has fallen, how to capture the excess, and what to do with it. The Province of British
Columbia in the Pacific Northwest is leading the way in North America in developing and implementing
innovative criteria and methodologies for reducing excess runoff volumes at the source, where rain falls.
Science-based performance objectives and targets have been established to mimic the hydrology of a natural
forest. Performance targets are being implemented through demonstration projects, notably at two large-
scale 'sustainable communities':
a UniverCity - A high-density urban community that is being developed by Simon Fraser University to
house 10,000 people at the top of Burnaby Mountain in the heart of the Greater Vancouver urban region
a Headwaters - A medium-density residential community that is being developed to house 14,000 people
in the East Clayton area of the City of Surrey, a suburban municipality in the Greater Vancouver region
that is the Province's second largest city (with a population 300,000).
Through an Inter-Governmental Partnership, a decision support tool called the Water Balance Model for
British Columbia is being enhanced to help local governments integrate land use planning with volume-
based analysis of stormwater management strategies. The WBM is used to evaluate the potential for
developing or redeveloping communities that function hydrologically like naturally forested or vegetated
systems. The tool creates an understanding of how, and how well, stormwater source control strategies for
runoff reduction would be expected to achieve watershed protection and/or restoration objectives.
453
-------
What Can be Done at the Site Level to Protect Watershed Health
The Logical Link
There is a logical link between changes in hydrology and impacts on watershed health, whether those
impacts are in the form of flooding or aquatic habitat degradation. The link is the volume of surface runoff
that is created by human activities as the result of alteration of the natural landscape. The key to protecting
urban watershed health is to maintain the water balance as close to the natural condition as is achievable and
feasible by preserving and restoring soils, vegetation and trees. Accomplishing this requires major changes
in the way we approach urban drainage and in the way we develop land. In the future, there will be more
runoff volume to manage in the urban regions of British Columbia due to the combination of:
a Population Growth - resulting in more land development plus re-development and densification of
existing urbanized areas
a Climate/Weather Change - likely resulting in both increased seasonal rainfall and more frequent
'cloudbursts'
The financial and staff resources of local government are limited. Therefore, those resources must be
invested wisely to maximize the return-on-effort. Common sense says that the best return will be at the site
level where local government exerts the most influence, and can therefore make a cumulative difference at
the watershed scale. The term 'source control' is used in this context to describe the suite of strategies
available to capture and retain rainfall volume at the development site.
Water Balance Model for British Columbia
The practice of low impact development often involves efforts to reduce the impacts of stormwater runoff
using various types of source controls designed to minimize runoff volumes. The effectiveness of these
source controls varies with their design, with precipitation patterns, and with soil type, among other factors.
The overall performance of these source controls is obviously of great interest to developers, homeowners
and local governments alike.
In June 2002, the British Columbia Ministry of Water, Land and Air Protection published the document
Stormwater Planning: A Guidebook for British Columbia The Guidebook lays out targets for reducing
runoff volume to achieve watershed protection objectives. The Greater Vancouver Regional District
(GVRD) recently completed a study to evaluate the effectiveness of a suite of such stormwater source
controls with these targets in mind. The results of the GVRD study are incorporated in the Guidebook.
In order to answer questions about the effectiveness of source controls, the GVRD's consultant developed
and applied a water balance model, an interactive tool that can simulate the performance of impervious
controls, absorbent landscaping, infiltration facilities, green roofs and rainwater harvesting under various
development scenarios. After exploring the capabilities of the model, a group of municipal, regional,
provincial and federal government representatives saw the potential to use it to integrate volume-based
analysis of stormwater management strategies into land use planning throughout British Columbia.
An Inter-Governmental Partnership was struck in the summer of 2002 to secure access to the model and
develop a more user-friendly version, to be called the Water Balance Model for British Columbia. The
Inter-Governmental Partnership is chaired by the BC Ministry of Agriculture, Food and Fisheries, and co-
chaired by Environment Canada. The GVRD is the host organization, providing logistical support as
required. A number of municipalities are currently engaged in the project, and others who share an interest
454
-------
are invited to join as the project evolves. The end result will be a user-friendly model that can be used to
inform and evaluate land use planning decisions for their ability to meet stormwater management
objectives, both at the scale of the individual development site and the watershed.
In Phase 1, to be completed by June 2003, the existing water balance model is being converted to a new
operating platform complete with graphical user interface (GUI) that will allow for more efficient data
storage procedures, faster performance, increased portability, more flexible output options, and easier
technical enhancement as the state-of-the-science evolves.
Members of the Inter-Governmental Partnership are participating actively in enhancement of the model and
graphic user interface, and will be the first recipients of the resulting Water Balance Model for British
Columbia (hereinafter referred to as 'the WBM'). Subsequent project phases may involve field testing and
calibration of key model assumptions, and linking the model to regional GIS and precipitation databases.
Project Vision for WBM Application
A "project vision" is the image or understanding of what the project will accomplish, and what will be
different at the end of the project. The British Columbia Guidebook demonstrates how to establish science-
based performance objectives to mimic the hydrology of a natural forest. This outcome can be achieved
through a combination of rainfall capture and runoff control techniques. The WBM is an extension of the
Guidebook, and is intended to be a 'decision support /scenario modeling tool' that will help local
governments and landowners make better land development decisions.
The over-arching project goal in enhancing the WBM is to facilitate changes in land development practices
so that in future sites and subdivisions will be designed to function hydrologically like a natural forest that
has 10% impervious area. To accomplish this goal, the GUI (graphical user interface) for the WBM must be
easy to understand and simple to use.
The enhanced WBM will be an Access-based, web-accessible platform. There are two audiences for the
model output: engineers and planners who want detailed data; and elected councils and the public who want
only the big picture. Account access privileges will be tiered as follows:
a Public access will be to the completed product and with limited model flexibility.
a Project partners will have access to developmental models, including opportunities to download model
databases.
a Scientific authority will have access to manipulate algorithms, manage and update user profiles.
A distinguishing feature of the WBM is the level of detail that it enables with respect to site design. This
provides a significant capability to test 'what if scenarios related to zoning bylaw changes.
Reducing the Volume of Runoff
Drainage engineers have traditionally thought in terms of flow rates rather than volumes. In fact, at the site
level, we need to focus on how much rainfall volume has fallen, how to capture the excess, and what to do
with it. British Columbia is leading the way in North America in developing and implementing innovative
criteria and methodologies for reducing excess runoff volumes at the source, where rain falls.
455
-------
What the Science is Telling Us
A science-based understanding of how land development impacts watershed hydrology and the functions of
aquatic ecosystems provides a solid basis for making decisions to guide early action where it is most
needed.
The science is explicitly telling us that major biophysical changes occur once the impervious percentage of
a watershed reaches about 10%. Beyond this threshold, a change in the water balance may trigger be
expected to trigger watercourse erosion, which in turn would degrade or eliminate aquatic habitat. This
implies that, where urban land use densities approach this threshold level, the focus should be on what
needs to be done at the site level to effectively mimic a watershed with less than 10% impervious area and
reduce runoff volumes to similar levels. As documented in the British Columbia Guidebook, the science
also indicates that capturing rainfall at the source for the frequent, lower intensity events will in large part
help maintain or restore the natural Water Balance.
Research on the Effects of Urbanization on Fish
Aquatic habitats that influence the abundance of salmon and trout are the outcome of physical, chemical and
biological processes acting across various scales of time and space. The environmental conditions that
result from these processes provide the habitat requirements for a variety of species and life history stages
offish and other stream organisms.
Decline of Wild Salmon
Whether in pristine or heavily urbanized watersheds, the basic requirements for survival of salmon and trout
are the same. These basic requirements include: cool, flowing water free of pollutants and high in dissolved
oxygen; gravel substrates low in fine sediment for reproduction; unimpeded access to and from spawning
and rearing areas; adequate refuge and cover; and sufficient invertebrate organisms (insects) for food.
Over the past century, salmon have disappeared from over 40% of their historical range, and many of the
remaining populations are severely depressed (Nehlsen et al. 1991). There is no one reason for this decline.
The cumulative effects of land use practices, including timber harvesting, agriculture and urbanization have
all contributed to significant declines in salmon abundance in British Columbia (Hartman et al. 2000).
Puget Sound Findings
In the Puget Sound region of Washington State, a series of Ksearch projects have been underway for over
10 years to identify the factors that degrade urban streams and negatively influence aquatic productivity and
fish survival. The streams and sites under examination represent a range of development intensities from
nearly undisturbed watershed conditions to watersheds that are almost completely developed in residential
and commercial land uses (Horner 1998).
For each watershed, detailed continuous simulation hydrologic models were prepared and calibrated to
rainfall and runoff data. Physical stream habitat conditions, water quality, sediment composition, sediment
contamination, and fish and benthic organism abundance and diversity were measured and documented for
each site.
The studies found that stream channel instability is a result of the urbanization of watershed hydrology. The
alteration of a natural stream's hydrograph is a leading cause of change in instream habitat conditions. The
physical and biological measures generally changed most rapidly during the initial phase of watershed
456
-------
development, as total impervious area changed from 5% to 10%. With more intensive urban development
in the watershed, habitat degradation and loss of biological productivity continues, but at a slower rate
(Horner 1998).
The role of large woody debris in streams was recognized as a key factor in creating complex channel
conditions and habitat diversity for fish. Both the prevalence and quality of large woody debris declined
with increasing urbanization. In addition, development pressure has had a negative impact on streamside
(riparian) forests and wetlands, which are critical to natural stream functioning.
The impacts of poor water quality and concentrations of metals in sediments did not show significant impact
to aquatic biological communities until urbanization increased above approximately 50% total impervious
area.
Instream habitat conditions had a significant influence on aquatic biota. Streambed quality, including fine
sediment content and channel stability, affected the benthic macro invertebrate community (as measured by
the multi-metric Benthic Index of Biological Integrity (B-ffil) developed by Karr (1991)). Negative
impacts to fish and fish habitat from sedimentation related to urban development have been documented
(Reid et al. 1999). The composition of the salmonid community was also influenced by a variety of
instream physical and chemical attributes.
Summary of Puget Sound Findings
Alterations in the biological community of urban streams are a function of many variables representing
conditions that are a result of both immediate and remote environmental conditions in a watershed. The
research findings clearly demonstrate that the most important impacts of urbanization that degrade the
health of streams, in order of importance, are:
a Changes in hydrology
a Changes in riparian corridor
a Changes in physical habitat within the stream, and
a Water quality
British Columbia Findings
Within the Georgia Basin of British Columbia, population pressures have caused urban sprawl, resulting in
habitat loss (B.C. MELP 2000). Freshwater fish population declines in this region are a partial result of
rapidly expanding urban development (Slaney 1996).
The aquatic ecosystems most directly affected by urbanization are the small streams and wetlands in the
lowlands of the Georgia Basin and lower Fraser River Valley. These ecosystems are critical spawning and
rearing habitat for several species of native salmonids (both resident and anadromous). In the Lower Fraser
Valley, 71% of streams are considered threatened or endangered, and a further 15% have been lost
altogether as a result of urban growth (B.C. MELP 2000).
457
-------
A Science-Based Understanding
The widespread changes in thinking about stormwater impacts that began in the mid to late 1990s reflect
new insights in two areas:
a Hydrology, and
a Aquatic ecology
These new insights are the result of improved understanding of the causes-and-effects of changes in
hydrology brought about by urban development, and the consequences for aquatic ecology. As we gain new
knowledge and understanding of what to do differently, a central issue for watershed protection becomes:
a What is the proper balance of science and policy that will ensure effective implementation and
results?
King County in Washington State addressed this question in 1999 as part of the Tri-County response to the
listing of chinook salmon as an endangered species in Puget Sound. A significant finding was that scientists
and managers think and operate differently. This led to the following recommendations:
a An interface is needed to translate the complex products of science into achievable goals and
implementable solutions for practical resource management. This interface is what we now call
a science-based understanding.
a A reality for local government is that management decisions need to be made in the face of
significant scientific uncertainties about how exactly ecosystems function, and the likely
effectiveness of different recovery approaches.
a The best path forward is a dynamic, adaptive management approach that will allow local
governments to monitor the effectiveness of their regulatory and management strategies and
make adjustments as their understanding grows.
a In a co-evolving system of humans and nature, surprises are the rule, not the exception; hence,
resilience and flexibility will need to be built into the management system.
Through a science-based understanding of the relationship between hydrology and aquatic ecology, the
British Columbia Guidebook has derived a comprehensive set of water balance, hydrology/water quality
and biophysical objectives that provide an over-arching framework for watershed protection.
Eliminate the Source of Problems
Understanding the cause-and-effect relationship between hydrology and biology has provided the basis for a
paradigm-shift in stormwater management in British Columbia - from a traditional approach that only deals
with consequences, to one that also eliminates the sources of problems.
Dealing with consequences is the traditional end-of-pipe engineering approach that is reactive in solving
problems after the fact. Eliminating the causes of problems involves an integrated approach to source-
control that is proactive in preventing problems from occurring.
In addition to being a partner in both the Guidebook and WBM initiatives, the GVRD has also developed
Integrated Stormwater Management Planning - Terms of Reference Template6 as part of its regulatory
commitment to the Province. The Template supports and encourages the use of the water balance
methodology for both greenfield and retrofit watersheds, particularly to assess the effectiveness of
stormwater source controls.
458
-------
Regulatory Overview
In British Columbia, the Local Government Act has vested the responsibility for drainage with
municipalities. With the statutory authority for drainage, local governments can be held liable for
downstream impacts that result from changes to upstream drainage patterns - both volume and rate. The
Act also enables local governments to be proactive in implementing stormwater management solutions that
are more comprehensive than past practice. Furthermore, a stormwater component is a requirement for
approved Liquid Waste Management Plans (LWMPs). Guidelines for developing an LWMP were first
published in 1992. LWMPs are created by local governments under a public process in co-operation with
the Province.
An Official Community Plan Provides the Foundation for a Stormwater Management Plan
There is a clear link between the land use planning required of local governments in the Local Government
Act and the LWMP process. In most cases where an Official Community Plan (OCP) is in place, the local
government planning statement (bylaw) will form the basis for an LWMP. The purposes of an LWMP are
to minimize the adverse environmental impacts of the OCP and ensure that development is consistent with
Provincial objectives.
OCPs tend to be led by planners, with input from engineers on infrastructure sections. LWMPs tend to be
led by engineers, with little cr no input from planners. Both processes involve approval by a Local Council
or a Regional Board. In some cases, an LWMP process may be a trigger that focuses attention on
stormwater management. In other cases, public concern related to flooding or habitat loss may be the
trigger. An OCP public process may communicate public interest in raising local environmental and habitat
protection standards. Whatever the motivation, at the end of the process an OCP should include goals and
objectives for stormwater management. These goals and objectives, or a variant of them, might first reside
in an LWMP, and then be adapted to the OCP in the next review process. Or they may originate in the OCP
process, and then be detailed through an LWMP. Either approach is entirely acceptable.
Integrated Stormwater Management Planning
In British Columbia, the term Integrated Stormwater Management Plan (ISMP) has gained widespread
acceptance by local governments and the environmental agencies to describe a comprehensive approach to
stormwater planning. The purpose of an ISMP is to provide a clear picture of how to be proactive in
applying land use planning tools to protect property and aquatic habitat, while at the same time
accommodating land development and population growth.
Stormwater Planning: A Guidebook for British Columbia
Stormwater management in British Columbia is a key component of protecting quality of life, property and
aquatic ecosystems. The science and practice of stormwater management is constantly evolving, in British
Columbia and around the world. Within British Columbia, the range of stormwater management activity
varies from completely unplanned in many rural areas, to state-of-the-art in some metropolitan centres. The
purpose of Stormwater Planning: A Guidebook for British Columbia is to provide a framework for
effective stormwater management that is usable in all areas of the province.
459
-------
The Guidebook presents a methodology for moving from planning to action that focuses the limited
financial and staff resources of governments, non-government organizations and the development
community on implementing early action where it is most needed. The Guidebook is organized in three
parts: Part A defines the problem, Part B provides solutions and Part C defines the process. The Guidebook
provides a comprehensive understanding of the issues and a framework for implementing an integrated
approach to stormwater management. Case study experience underpins the approaches and strategies that
are presented in the Guidebook.
Guidebook Overview
Part A - Why Integrated Stormwater Management?
Part A identifies problems associated with traditional stormwater management and provides the rationale for
a change from traditional to integrated stormwater management. Some guiding principles of integrated
stormwater management are introduced. Part A also builds a science-based understanding of how natural
watersheds function and how this function is affected by land use change.
Part B - Integrated Stormwater Management Solutions
Part B outlines the scope and policy framework for integrated stormwater management, and presents a
three-step, cost-effective methodology for developing stormwater solutions.
Step #1 - Identify At-Risk Drainage Catchments: A methodology is presented for identifying at-risk
drainage catchments to focus priority action. The methodology relies on a roundtable process that brings
together people with knowledge about future land use change, high-value ecological resources and chronic
flooding problems. The key is effective integration of planning, engineering and ecological perspectives.
Step #2 - Set Preliminary Performance Targets: A methodology is presented for:
a Developing watershed performance targets based on site-specific rainfall data, supplemented by
streamflow data (if available) and on-site soils investigations
a Translating these performance targets into design guidelines that can be applied at the site level to
mitigate the impacts of land development
This portion of the Guidebook also documents British Columbia case studies of stormwater policies and
science-based performance targets applied to both greenfield and urban retrofit scenarios.
Step #3 - Select Appropriate Stormwater Management Site Design Solutions: Guidance is provided for
selecting appropriate site design solutions to meet performance targets source control and runoff
conveyance. Case study examples are provided of:
a Design and performance of stormwater source controls for various land uses
a Watershed scale modelling of the effectiveness of site design solutions
460
-------
Part C - Moving from Planning to Action
Part C describes a process that will lead to better stormwater management solutions. The role and design of
action plans are introduced to bring a clear focus to what needs to be done, with what priority, by whom,
with related budgets. Tips are provided on processes that produce timely and high-quality decisions. Part C
also provides guidance for organizing an administrative system and financing strategy for stormwater
management. A final section on building consensus and implementing change describes how to develop a
shared vision and overcome barriers to change.
Two acronyms, ADAPT and CURE, provide a useful summary of the principles and elements of
integrated stormwater management, as described below.
ADAPT- The Guiding Principles of Integrated Stormwater Management
The acronym ADAPT summarizes five guiding principles for integrated stormwater management. The
Guidebook is based upon these five principles.
gree that stormwater is a resource
Design for the complete spectrum of rainfall events
ct on a priority basis in at-risk drainage catchments
Ian at four scales - regional, watershed, neighbourhood & site
solutions and reduce costs by adaptive management.
Guiding Principle 1 - Agree that Stormwater is a Resource
Stormwater is no longer seen as just a drainage or flood management issue but also a resource with
both benefits and deleterious effects on:
a fish and other aquatic species
a groundwater recharge (for both stream summer flow and for potable water)
a water supply (e.g., for livestock or irrigation)
a aesthetic and recreational uses
461
-------
Guiding Principle 2 - Design for the Complete Spectrum of Rainfall Events
Integrated stormwater solutions require site design practices that provide:
a Rainfall Capture for Small Storms (runoff volume reduction and water quality control) - Capture
the low intensity, frequently occurring rainfall events at the source (building lots and streets) for
infiltration and/or re-use.
a Runoff Control for Large Storms (runoff rate reduction) - Store the runoff from the infrequent large
storms (e.g., a mean annual rainfall), and release it a rate that approximates the natural forested
condition.
a Flood Risk Management for the Extreme Storms (peak flow conveyance) - Ensure that the drainage
system can safely convey extreme storms (e.g., a 100-year rainfall).
The Integrated Strategy for Runoff Volume Management
Guiding Principle 2 forms the foundation of integrated stormwater solutions that mimic the most effective
stormwater management system of all - a naturally vegetated watershed. The 'integrated strategy' for
managing the complete spectrum of rainfall events is built around an understanding of the Natural Water
Balance. The strategy has three components - retain the small frequent events, detain the large events, and
convey the extreme events - as illustrated below.
Total Rainfall Volume
75%
Small Storms
20%
Large Storms
5%
Extreme Storms
Evaporation-
Transpiration
Reuse
Rainfall Capture |
Infiltrate or Reuse
Small Storms at the
Source to Reduce
I
Total Runoff Volume!
1
Runoff Control
Provide Storage to
Control the Rate
of Runoff from
Large Storms
Flood Risk
Management
Ensure that the
Stormwater System
can Safely Convey
Extreme Storms
fntei
^Storage
Release
Deep
Groundwater
Runoff
Integrated Strategy for Managing the
Complete Spectrum of Rainfall Events
Typical
Rainfall
Volume
Distribution
Stormwater
Management
Strategy for
Impervious
Areas
Hydrologic
Pathway
462
-------
The WBM enables modelling of all three components of the integrated strategy. It can be used to evaluate
how well alternative strategies (including combinations of storm water source control and off-site detention)
can reduce the runoff from development areas, and how this translates into benefits at the watershed level.
Source control options include bioretention, infiltration facilities, rainwater capture and re-use, and green
roofs. The WBM can also be used to evaluate the impacts of population growth and climate change
scenarios.
The Target Condition for a Healthy Watershed
The target condition for any watershed is defined by the Water Balance, water quality and streamflow
characteristics of that watershed with less than 10% impervious area. The target relates to existing
conditions for relatively undeveloped watersheds (i.e., new development scenarios) and historical
conditions for developed watersheds (i.e., retrofit scenarios). In order to achieve the target condition, the
total annual runoff volume must be limited tolO% (or less) of total annual rainfall volume. This means that
90% of annual rainfall must be returned to natural hydrologic pathways (e.g., infiltration and evapo-
transpiration) or harvested for re-use. Capturing the frequent small rainfall events at the source will, in large
part, maintain or restore the natural Water Balance and achieve the above targets. The Guidebook explains
how to achieve the above water balance targets at the site scale, and how to apply the Water Balance Model
to assess the feasibility of reducing runoff volume at the watershed scale over time in conjunction with land
redevelopment.
Comparison with Conventional Stormwater Management
Conventional 'flows-and-pipes' stormwater management is limited because it focuses only on the fast
conveyance of the extreme storms and often creates substantial erosion and downstream flooding in
receiving streams. Similarly, a detention-based approach is only a partial solution because it allows the
small storms that comprise the bulk of total rainfall volume to continue to create erosion and impacts on
downstream aquatic ecosystems. Neither of these approaches fully prevents the degradation of aquatic
resources or flooding risks to property and public safety. In contrast, the Guidebook approach is to eliminate
the root cause of ecological and property impacts by designing for the complete spectrum of rainfall events.
Solutions described in the Guidebook include conventional, detention, infiltration and re-use approaches for
rainfall capture, runoff control and flood risk management.
Guiding Principle 3 -Act on a Priority Basis in At-Risk Drainage Catchments
Focus priority action should be focused in at-risk drainage basins where there is both high pressure for land
use change and a driver for action. The latter can be either:
a a high-value ecological resource that is threatened
a an unacceptable drainage problem
The stormwater management policies and techniques implemented in at-risk catchments become
demonstration projects.
463
-------
Guiding Principle 4 - Plan at Four Scales - Regional, Watershed, Neighbourhood and Site
Integrated stormwater management must be addressed through long term planning at each of the regional,
watershed, neighbourhood and site scales.
a At the Regional and Watershed Levels - Establish stormwater management objectives and priorities
a At the Neighbourhood Level - Integrate stormwater management objectives into community and
neighbourhood planning processes
a At the Site Level - Implement site design practices that reduce the volume and rate of surface runoff
and improve water quality
Guiding Principle 5 - Test Solutions and Reduce Costs by Adaptive Management
Performance targets and stormwater management practices should be optimized over time based on:
a monitoring the performance of demonstration projects
a strategic data collection and modeling
As success in meeting performance targets is evaluated, the stormwater management program can be
adjusted as required.
CURE- The Elements of an Action Plan
The acronym CURE focuses attention on the four key types of actions that must all work together to
implement integrated stormwater management solutions:
a CAPITAL INVESTMENT - Short-term capital investment will be needed to implement early action in
at-risk drainage basins. Improvements to existing drainage system are often the most significant capital
investments required. A financing plan should provide an ongoing source of funds for watershed
improvements.
a UNDERSTANDING SCIENCE - Improved understanding of a watershed, the nature of its problems,
and the effectiveness of technical solutions is key to an adaptive approach. Stormwater management
practices can be optimized over time through the monitoring of demonstration projects, combined with
selective data collection and modeling.
a REGULATORY CHANGE - Changes in land use and development regulations are needed to achieve
stormwater performance targets. Changes to land use planning and site design practices are needed to
eliminate the root cause of stormwater related problems. These changes must be driven by regulation.
a EDUCATION AND CONSULTATION - Changes to land use planning and site design practices can
only be implemented by building support among city staff, the general public and the development
community through education and consultation.
464
-------
Translating a Vision into Action
It is important to establish a long term shared vision at the start of any watershed planning initiative. A
vision that is shared by all stakeholders provides direction for a long-term process of change. The vision
becomes a destination, and an action plan provides a map for getting there. Actions plans must be long term,
corresponding to the time frame of the vision. Action plans must also evolve over time. Ongoing
monitoring and assessment of progress towards a long term vision will improve understanding of the policy,
science and site design components of integrated stormwater management. This improved understanding
will:
a Lead to the evolution of better land development and stormwater management practices
a Enable action plans to be adjusted accordingly
An adaptive management approach to changing stormwater management practices is founded on learning
from experience and adjusting for constant improvement.
Building Blocks
The Guidebook elaborates on three fundamental objectives that become building blocks for a long-term
process of change:
a Achievable and Affordable Goals - Apply a science-based approach to create a shared vision for
improving the health of individual watersheds over time
a Participatory Decision Process - Build stakeholder consensus and support for implementing change,
and agree on expectations and performance targets
a Political Commitment - Take action to integrate stormwater management with land use planning
The Water Balance Model: A Tool for Stormwater Source Control
Modelling in a Watershed Context
For the past thirty years, there has been a fixation on peak flow control through the use of detention ponds
for all flood events from the 2-year through 100-year floods, and the conveyance of major flood events
caused by urban developments of all kinds. The recently developed software focus has been on the user
interfaces, but not on the hydrology engine; and certainly not on improvements in the science of infiltration.
Traditional applications of hydrology models reflect "peak flow thinking" at a watershed or macro scale.
But the models may not be appropriate for simulating what happens at the site scale, nor for assessing the
effects of storm runoff volume changes caused by urban development.
The missing link in urban hydrology has been a tool that quantifies the benefits, in terms of reducing
stormwater runoff volume at the site level, of installing source controls under a variety of circumstances.
The water balance modeling approach was developed to demonstrate how to meet performance targets for
water balance management at the site, neighbourhood, drainage catchment, and watershed scales. The
WBM assists local governments to integrate land use planning with volume-based analysis of stormwater
management strategies.
465
-------
The volume-based approach that is being implemented in British Columbia picks up the baton that Dr. Ray
Linsley started more than a generation ago. As a professor of Civil Engineering at Stanford University, and
later as a consulting engineer, Linsley pioneered the development of continuous hydrologic simulation as
the foundation for water balance management. He has received world-wide recognition for his vision and
his contributions to the field of hydrology and continuous hydrologic simulation modelling:
• In the 1960s, Linsley championed the paradigm-shift from empirical relationships to computer
simulation of hydrologic processes. He had little or no use for "simple hydrology" and the many
simple equations that were used to represent the hydrologic cycle.
• Linsley fought a difficult war to replace the established procedures that had been used for many
years, and that continue to be used in most urban hydrologic analyses throughout North America and
in other locations around the world. He believed that continuous simulation was the only hydrology
that should be used for most design and analysis applications.
• Linsley's pioneering efforts resulted in development of the well-known HSPF Model. This continues
to be the hydrologic simulation tool of choice in many parts of North America, notably Washington
State where its use is mandated by the Department of Ecology, even though it is a complex model
with great data input needs.
Somewhat ironically, the "hydrology engine" for HSPF and other contemporary models (such as SWMM)
is based on 1930s and 1940s science. As reported by Linsley in a 1976 article:
• In 1933 - Horton first proposed the concept of infiltration, which is at the heart of continuous
simulation.
• In 1934 - Zoch first suggested the use of routing to develop the runoff hydrograph.
• In 1942 - Linsley and Ackerman introduced the idea of continuous soil moisture accounting.
The power of the WBM is in the engine that instantly, interactively, and transparently models hydrologic
processes at the site level, including the processes that govern the movement of water through soil and
vegetation. This engine incorporates algorithms that simulate how runoff is generated at the site level and
generates a continuous simulation of the runoff from a development site, neighbourhood, drainage
catchment, or watershed. The WBM simulates five source control categories:
a Impervious Controls
a Absorbent Landscaping
a Infiltration Facilities
a Green Roofs
a Rainwater Re-Use
The WBM provides local governments with the means to integrate land use planning with stormwater
management. It is a decision support and scenario modelling tool that is used to:
a Visualize the 'how to' details of source control implementation
a Model scenarios at the site, neighbourhood and watershed scales
a Make decisions through a scientifically defensible, interactive and transparent process.
466
-------
The WBM has a wide range of application possibilities, including:
a Design of volume-based stormwater controls
a Site performance assessment
a Evaluating opportunities for urban retrofits
a Volume-based watershed trading for urban stormwater management
a Watershed management optimization
a Analysis of changes in rainfall patterns
a Public education and outreach
The WBM has enabled evaluation of the hydrologic performance of stormwater source controls (e.g.,
bioretention, infiltration facilities, rainwater capture and re-use, green roofs) and stormwater detention. It
provides a continuous simulation of the runoff, given these inputs:
a Continuous rainfall data (any time increment)
a Evapotranspiration data
a Extent and distribution of land use types
a Site design parameters for each land use type
a Soil and groundwater information
a Information on stormwater controls
a Seasonal change in rainfall patterns due to climate change
The sensitivity of source control performance to any of these model inputs can be tested by comparing
modelled scenarios. The output hydrograph generated by the WBM can become an input to a wide range of
hydraulic routing models. WBM hydrographs represent a major improvement over conventional hydrologic
simulation. In the Greater Vancouver Region, the WBM has been used to assess the potential for urban
watershed restoration over a 50-year timeframe. The WBM has made it possible to:
a Identify affordability and feasibility thresholds
a Develop evaluation criteria for cost-benefit analysis
a Generate watershed-specific performance relationships
The following figures illustrate the types of relationships that have been developed using the WBM, and
that are presented in the Guidebook:
Achievable Level of Runoff Volume Reduction
Using Infiltration Facilities
North Surrey Rainfall (wet year, 1999)
_i
If
"- = 60-
t 'ro
o >-
K 2
o £ 40
Q)
3
o 20_
0-
2
1
I
I
I
/
y
cy • "- ,j '""""""""""" ^|^^^^^^^
Hydraulic Conductivity
of Local Soils
Very Low (1
mm/h)
Low(2.5mm/h)
Medium to High
(greater than 13
mm/h)
No Source Control
0 30 40 50 60 " " 70 ' ' 80^100
Percent Impervious Lot Coverage
Where soils have medium or
better hydraulic conductivity,
runoff volume could be
reduced to about 10% of total
rainfall for all but the highest
coverage land uses.
Significant levels of runoff
volume reduction can also be
achieved in soils with poor
hydraulic conductivity.
467
-------
Infiltration Facility Performance (Runoff Rate Reduction)
25
E 20
o f ^
it o
£ I £ 15
- S 10
50% lot coverage (e.g. single family)
- — \ !
\:
N^
Affordability
Threshold
1
I
I
Feasibility
Threshold
5 10 15 20
% of Lot Used for Infiltration
25
Hydraulic
Conductivity of
Local Soils
Very Low (1
mm/h)
Low(2.5
mm/h)
-Medium to
High (greater
than 13 mm/h)
No Source
Control
Reductions in runoff rates
using infiltration facilities
depend on the hydraulic
conductivity of local soils
and the amount of area
provided for infiltration.
Affordability thresholds
govern infiltration facility
sizes for lower surface
coverage land uses, and
feasibility thresholds govern
for higher coverage land
uses.
Summary
Recent stormwater initiatives in British Columbia include:
a Publication of Stormwater Planning : A Guidebook for British Columbia
a Publication of Integrated Stormwater Management Planning - Terms of Reference Template
a Development of the Water Balance Model for British Columbia
a Evaluation of Stormwater Source Control Effectiveness at the site, neighborhood and watershed scales
To protect property, aquatic habitat and water quality, British Columbia has:
a Recognized the logical link between surface runoff volume and impacts on watershed health
a Embraced the integration of land use planning with stormwater management
a Established performance objectives for designing communities that function hydrologically like
naturally forested systems
The paradigm-shift from an approach that only deals with consequences, to one that also eliminates the
causes, has resulted in a re-invention of urban hydrology:
a There was a need for a tool that realistically simulates how runoff is actually generated at the site level
a The WBM is a stormwater planning and site design tool that evolved in two stages:
• Initially through the Burnaby Mountain Project - to achieve watershed protection objectives
• Subsequently through the GVRD Project - to evaluate the effectiveness of a range of source control
options (e.g., absorbent landscaping, infiltration facilities, rainwater re-use, green roofs) under a
range of operating conditions (i.e., land use, soil and rainfall)
468
-------
Conclusion
The Water Balance Model for British Columbia provides an effective decision support tool for local
governments to integrate land use planning with stormwater management, and to evaluate the potential for
developing or re-developing communities that function hydrologically like naturally forested or vegetated
systems. The tool creates an understanding of how, and how well, stormwater source control strategies for
runoff reduction would be expected to achieve watershed protection and/or restoration objectives.
Bibliography
The following is a partial list of publications on the aquatic impacts from urban development and
stormwater in British Columbia and Washington State:
• BC Ministry of Environment, Lands and Parks and Environment Canada. 2000. Water Quality Trends in
Selected British Columbia Waterbodies. http://wlapwww.gov.bc.ca/wat/wq/trendsWQS/index.html
• BC Ministry of Environment, Lands and Parks. 2000. Environmental Trends in British Columbia, 2000.
http://wlapwww.gov.bc. ca/soerpt/
• Hartman, G.F., Groot, C., and Northcote, T.G. 2000. The Ball is Not in Our Court. In Sustainable Fisheries
Management: Pacific Salmon. Edited by E. Eric Knudsen, C.R. Stewart, D.D. MacDonald, J.E. Williams and
D.W. Rieser. Lewis Publishers, Boca Raton, FL. pp. 31-49.
• Homer, Richard, and May, C. 1998. Watershed Urbanization and the Decline of Salmon in Puget Sound
Streams. In Salmon in the City May 20-21, 1998, Mount Vernon Washington, Abstracts. Edited by
Anonymous, pp. 19-40.
• Karr, J.R 1991. Biological Integrity: A long neglected aspect of water resources management. Ecological
Applications 1(1): 66-84 in Homer, Richard, and May, C. 1998. Watershed Urbanization and the Decline of
Salmon in Puget Sound Streams. In Salmon in the City May 20-21, 1998, Mount Vernon Washington,
Abstracts. Edited by Anonymous, pp. 19-40
• Nehlsen, W. Williams, J.E. and J.A. Lichatowich. 1991. Pacific salmon at a crossroads: stocks at risk from
California, Oregon and Washington. Fisheries. Vol 16, No.2 American Fisheries Society. Bethesda, MD. Pp
4-21.
• Reid G, and Michalski, T. 1999. Status of Fish Habitat in East Coast Vancouver Island Watersheds. In At
Risk. Proceedings of a Conference on the Biology and Management of Species at Risk. Edited by L. Darling.
BC Ministry of Environment, Lands and Parks, Victoria, BC. pp. 355-367.
• Slaney, T.L, Hyatt, K.D., Northcote, T.G, and Fielden, RJ. 1996. Status of Anadromous Salmon and Trout
in British Columbia and Yukon. American Fisheries Society 21: 20-35.
1 IGP Liaison & Project Coordinator
2 Chair, British Columbia Inter-Governmental Partnership for Development of a Water Balance Model
3 Co-chair, Inter-Agency Steering Committee for Stormwater Planning Guidebook; and
Co-chair, British Collumbia Inter-Governmental Partnership for Development of a Water Balance Model
4 Facilitator, Greater Vancouver Regional District Stormwater Inter-Agency Liaison Group
5 The Stormwater Planning Guidebook is available on the Ministry's website at :
http://walpwww.gov.bc.ca/epdpa/mpp/stormwater/stormwater.html
6 Integrated Stormwater Management PlanningTerms of Reference Template, prepared by KWL for the GVRD, 2002.
http://www.gvrd.bc.ca/services/sewers/drains/Reports
469
-------
Development of the San Diego Creek Natural Treatment System
Eric Strecker, GeoSyntec Consultants, Portland, OR, estrecker@geosyntec.com
Peter Mangarella, GeoSyntec Consultants, Walnut Creek, CA, pmangarella@geosyntec.com
Norris Brandt, Irvine Ranch Water District, Irvine, CA, brandt@irwd.com,
Todd Hesse, GeoSyntec Consultants, Portland, OR, thesse@geosyntec.com
Rachata Muneepeerakul, GeoSyntec Consultants, Walnut Creek, CA, rmuneepeerakul@geosyntec.com
Klaus Rathfelder, GeoSyntec Consultants, Portland, OR, krathfelder@geosyntec.com
Marc Leisenring, GeoSyntec Consultants, Portland, OR, mleisenring@geosyntec.com
ABSTRACT
A Natural Treatment System (NTS) Master Plan that includes a watershed-wide network of constructed
wetlands was evaluated for treatment effectiveness of dry weather base flows and runoff from smaller more
frequent storms in a 120 square mile (311 km2), urban watershed. The goal of the 'regional retrofit' wetland
network is to serve as an integral component in watershed-wide BMPs for compliance with pollutant
loading limits (TMDLs) requiring discharge limits of sediments, nutrients, pathogen indicators, pesticides,
toxic organics, heavy metals, and selenium. The NTS Plan was assessed with 'planning-level' water quality
models that account for the integrated effects of the planned 44 NTS facilities. The NTS Plan is estimated to
achieve total nitrogen (TN) TMDL for base flows and reduce in- stream TN concentrations below current
standards at most locations. Total phosphorous TMDL targets would be met in all but the wettest years.
The fecal coliform TMDL would be met during the dry season, but not all wet season base flow conditions,
and not under storm conditions. The NTS Plan is not designed to meet the sediment TMDL, but would
capture, on average, about 1,900 tons/yr (1,724,000 kg/yr) of sediment from urban areas. The wetlands are
estimated to remove 11% of the total copper and lead, and 18% of the total zinc in storm runoff. The NTS
Plan provides a cost-effective alternative to routing dry-weather flows to the sanitary treatment system.
Introduction
San Diego Creek and Newport Bay in Orange County, California have been identified as having impaired
surface water quality under California State and U.S. Environmental Protection Agency (USEPA)
regulations. The creek and the bay receive runoff from storm events and from agricultural and urban
activities in the San Diego Creek Watershed, in addition to natural flows. Federal regulations for impaired
water bodies require the establishment of and compliance with discharge limits for the pollutants that are
determined to be causing the impairments. These limits are called total maximum daily loads (TMDLs),
and are linked to discharge permits established under the National Pollutant Discharge Elimination System
(NPDES).
Orange County and NPDES co-permittees, including the local municipalities, are seeking comprehensive
solutions for meeting the TMDL requirements. As a component of this effort, the Irvine Ranch Water
District (IRWD) has developed a Natural Treatment System (NTS) Plan. The NTS Plan addresses runoff
water quality from a watershed-wide perspective, utilizing a network of constructed wetlands. The NTS
Plan would build on IRWD's successful use of constructed wetlands by expanding their use throughout a
highly urbanized and nearly fully developed watershed. The NTS Plan, therefore, is viewed as an urban
470
-------
retrofit using constructed wetlands as an integral component for compliance with TMDL requirements. The
advantage of the NTS system to IRWD, the primary provider of sanitary and potable water services for the
watershed, is avoiding the increasingly costly trend in Southern California of routing low flows to sanitary
treatment systems.
This paper describes the NTS Plan, the evaluation approach, and the evaluation results of the Plan's
effectiveness for contributing to TMDL compliance. An example of the NTS retrofit concept is provided at
the end of the paper.
Project Area
Setting. The San Diego Creek Watershed is located in Orange County, California (Figure 1) and covers
approximately 120 square miles (311 km2). The watershed is drained by Peters Canyon Wash and San
Diego Creek, and by a number of smaller channels and drainages. San Diego Creek flows into Upper
Newport Bay, which contains the 752-acre (3.04 km2) Upper Newport Bay Ecological Reserve, one of the
largest remaining coastal estuaries in Southern California. The San Diego Creek Watershed drains almost
80% of the 154 square miles (398.9 km2) that are tributary to Upper Newport Bay.
The western and central portions of the watershed are a relatively flat alluvial plain, bordered by the
Santiago Hills to the northeast and the San Joaquin Hills to the south. The alluvial plain rises gently from
sea level at Upper Newport Bay to about 400 ft (122 m) above mean sea level (msl) at the El Toro Marine
Base. The peak elevation in the Santiago and San Joaquin Hills is 1,775 ft (541 m) and 1,160 ft (355 m)
above msl, respectively.
The climate is characterized by warm dry summers, and cool intermittently wet winters. The main wet
season is from November to April during which widespread general winter storms may last for several days.
The average annual rainfall is about 13 inches per year, with 90% occurring in the wet season. Average
base flows in San Diego Creek are less than 16 cfs (0.45 cms) during dry weather. The estimated peak 100-
year flood discharge is 42,500 cfs (1,203 cms) in San Diego Creek at Newport Bay.
Table 1: Estimated existing and fully developed land uses acreages in the San Diego Creek Watershed.
Land Use
Agriculture
Urban1
Open2
Existing
(acres)
11,510
40,210
24,690
Estimated when fully
developed (acres)
1080
52,160
23,170
% Change of watershed from
existing to fully developed
-13.7
+15.6
-2.0
Urban is the sum of commercial/light industrial, industrial, mixed use, all residential, roads, and transportation corridors.
2 Open is the sum of open space-preserve, open space-other, parks, golf courses, and water land use categories.
Land Use. The San Diego Creek Watershed experienced rapid growth and development after World War II.
Land-use estimates show that most of the developable lands in the watershed are currently developed (Table
1), with about 15 percent remaining. Much of remaining development would come from continued
conversion of agricultural land and from land-use conversion of recently decommissioned military bases.
471
-------
San Diego
Creek
Watershed
Boundary
Newport
Ds^n r*l\
•raCo
no -
M ^3V
San ••*'./> 4932
o - iaPt GS^J
O 42 Q -."^OP
•so^^B
Type 1- Offline Water Quality
Type 2 - Inline Water Quality
Type 3 - Water Quality Wetlands within Existing
Proposed Detention Basin
Sites excluded from selected NTS Plan
Figure 1: Aerial photograph of the San Diego Creek Watershed showing the locations of NTS Facilities and the types
of wetland facilities.
Water Quality Issues and Regulatory Requirements.
Coinciding with rapid growth and development over the past 50 years, water quality in San Diego Creek and
Newport Bay has been affected by:
• Excessive sediment loads and sedimentation in Upper Newport Bay, impacting beneficial uses of the
bay and wildlife habitat;
• Excessive nutrient concentrations, primarily nitrate from fertilizers, which contribute to the
formation of algae blooms in Newport Bay;
• Elevated fecal coliform concentrations in the Newport Bay, especially in storm runoff, which impact
shellfish harvesting and recreational uses;
• Elevated concentrations of toxics in portions of Newport Bay, primarily the pesticides Diazinon and
Chlorpyrifos, which contribute to acute and chronic toxicity;
472
-------
• Elevated concentrations of heavy metals in portions of Newport Bay, primarily copper, which "may
be causing, or contributing to, toxicity to aquatic life" (RWQCB, 2000); and
• Elevated concentrations of selenium in San Diego Creek from natural origins, with the major source
thought to originate from groundwater discharge to San Diego Creek in areas of a historic ephemeral
lake in Peters Canyon Wash
Water quality has been affected by both low-flows resulting from irrigation return flows, car washing, and
groundwater recharge to streams, as well as stormwater discharges. Dry weather flows have increased with
urbanization of open space and remained about the same, as compared to agricultural activities. The normal
generalization that urbanization dries up base flows is typically not true in southern California because
irrigation levels significantly exceed natural rainfall. These low flows have caused leaching of pollutants
from soils, as well as transport of dissolved nutrients from planted areas.
As a result of these water quality problems, Newport Bay has been designated as an impaired water body by
the State of California. In response, TMDLs have been established or drafted for the impairing pollutants
(Table 2) (USEPA, 1998a,b; 2002). To address TMDL requirements, Orange County and local
municipalities have implemented an array of Best Management Practices (BMPs) for load reduction,
regional monitoring activities for the assessment of BMP effectiveness, and public education and
coordination efforts. These activities are generally directed towards source control and do not fully address
regional treatment needs for compliance with the TMDL requirements.
Table 2: A listing of the constituents included in the San Diego Creek TMDLs, general information about each, and
the TMDL loading limits for watershed land uses.
Constituent
Sediment
Nutrients
(TN and TP)
Pathogens
Selenium
(draft)
Heavy metals
Chlorpyrifos &
diazinon
Organochlorine
compounds
General Information
Load is strongly correlated with rainfall.
Annual average load estimate: 250,000
tons; 1998 load was 620,000 tons.
Declining trends in 1990's
1 986 TN load = 1,448,000 Ibs
1 998 TN load = 632,000 Ibs
Fecal coliform bacteria used as an
indicator. Goal is to achieve contact
recreation standards by 2014.
Natural sources from groundwater
discharge and surface runoff
1998/99 estimate: 3,248 Ibs/year
Loads highly variable with rainfall:
Total load (Ibs) 1998 1999
Copper 15,087 1,643
Lead 10,385 449
Zinc 63,021 3,784
Widely used pesticides that are currently
being phased out for non-commercial use.
Both exceed the chronic concentration
criteria in base flow and storm flow
conditions.
Legacy compounds that tend to
bioaccumulate and have considerable
persistence in soils, sediments, and biota.
Sources are unknown.
TMDL
62,500 tons/year to Newport Bay,
62,500 tons/year to the rest of the watershed, based on a
10-year running average.
Annual total load targets:
298,225 Ibs Total Nitrogen/year by 2012
62,080 Ibs Total Phosphorus/year by 2007
5 samples/30-days with a geometric mean concentration
of 200 organisms /lOOmL, and no more than 10% of the
samples to exceed 400 organisms/ lOOmL
Annual total load targets = 891 .4 Ibs. Loads are
partitioned into four flow tiers .
Concentration based TMDLs expressed at four flow tiers.
Concentrations are based on the California Toxics Rule
objectives using average hardness values of the associated
flow tier
SD Creek acute and chronic concentration targets,
respectively, by 2005:
Diazinon - 80 & 50 ng/L
Chlorpyrifos - 20 & 14 ng/L
Annual load limits to Newport Bay (g/yr):
Chlordane = 346.2; Dieldrm = 287.7; DDT = 475. 9;
PCBs = 310.3; Toxaphene = 9.8
473
-------
Natural Treatment System Plan
Plan Development
Various treatment-type control options were evaluated in developing the NTS strategy, including: (1) on-site
controls for new development; (2) complete or partial diversion of dry weather base flows and portions of
wet weather discharges to the sanitary sewer system; and (3) a regional treatment approach
Given the urbanized nature of the watershed, a strategy that focuses on on-site controls for new
development (or re-development) could not, by itself, meet regulatory requirements in a timely manner,
since that strategy would not address pollutants associated with existing urbanization in the San Diego
Creek Watershed, nor disperse sources such as groundwater discharges. Diversion of streamflow to the
sanitary sewer was determined to be mostly infeasible, given the stringent total dissolved solids
requirements for water recycling (an important IRWD water conservation tool), the cost for providing
storage and treatment for the large volumes of water, and the need to maintain in-stream flows for riparian
habitat and wildlife.
The NTS approach, based on a regional network of constructed wetlands, was determined to be the best
strategy for addressing regional water quality treatment needs because: (1) constructed wetlands are an
effective and cost-competitive approach for water quality treatment, based on the experience and success of
the existing IRWD constructed wetlands in the San Joaquin Marsh (a low-flow treatment marsh already
operated by IRWD near Upper Newport Bay), as well as other wetlands both regionally and nationally; (2)
constructed wetlands address pollutant sources from existing and future development, as well as disperse
sources; and (3) constructed wetlands can enhance habitat and natural resources in the watershed.
Constructed Wetlands
The facilities envisioned in the NTS Plan are constructed wetlands to improve the water quality of dry
weather base flows and the runoff from smaller storms. Constructed wetlands are engineered systems
designed to improve water quality by taking advantage of processes occurring in natural wetlands, but in a
more planned and controlled system. Constructed wetlands have evolved and gained acceptance during the
past 25 years as a practical and cost-effective means for advanced treatment of municipal wastewater and
for treatment of urban runoff (Kadlec and Knight, 1996; Strecker, 1996).
A local example is the IRWD constructed wetlands at the San Joaquin Marsh near the mouth of the San
Diego Creek Watershed. The IRWD constructed wetlands consists of five treatment cells with 45 acres of
open water and 11 acres of marshland vegetation. Water is pumped from San Diego Creek into the
wetlands at an average rate of about 7 cfs and has a retention time of about two weeks. Monitoring data
indicate that about 200 Ibs (91 kg) of nitrate are removed per day during dry weather, reducing the total load
to Upper Newport Bay by about 30%. The strategy of the NTS Plan is to expand the success of the IRWD
wetlands throughout the San Diego Creek Watershed.
Facility Designs
Each of the over 40 NTS facilities will be tailored to local conditions and constraints; however, most of the
NTS facilities share common design features (see Figure 2). Throughout most of the year the water quality
wetlands will primarily treat low flows because rainfall events are infrequent in Orange County (10-15
events per year over 0.1 inch (0.25 cm)). During non-storm conditions, water levels in the typical wetlands
will be in two general regimes:
474
-------
Shallow water
with emergent
Cattails
Open
water
areas-
Shallow water
with emergent
Bulrush
Inflow
Shallow water
depth = 1-2 ft
Outlet
structure
& piping
Open water pool,
depth = 4-6 ft.
Outflow
Small storrrwater
quality pool,
depth
-3-4 ft.
Main
pollutant
removed
Inlet Pool
Sediment,
pathogens
Cattail Stand
Open Water
Pool
Nitrate, Pathogens
phosphorus, metals, Organic
pathogens transformation
Secondary
pollutant
removed
Metals,
phosphorus,
organics
Organics
Sediment
Primary
Removal
Mechanisms
Sedimentation,
UV radiation
Nitrification, plant
uptake, filtration,
sedimentation
UV radiation,
volatilizatrion
Wildlife
Mosquito fish,
birds
Birds, insect larvae
e.g. dragonflies,
mosquito fish
Ducks,
mosquito fish
Bulrush Stand
Organics,
pesticides
Nitrate,
phosphorus, metals
Nitrification, plant
uptake,
sedimentation
Birds, insect larvae
e.g. dragonflies,
mosquito fish
Outlet Pool
Pathogens
Organics
UV radiation
Ducks,
mosquito fish
Main
pollutant
removed
Secondary
pollutant
removed
Primary
Removal
Mechanisms
Wildlife
Extended
Detention
Sediment
Oil and grease,
phosphorus,
metals, organics
Sedimentation
None
Figure 2: Generic Design and Removal Mechanisms of NTS Facilities, showing a plan view and providing information
on intended pollutant removals in each sub-area of the wetland.
Open water regions typically 4-6 ft (1.2-1.8 m) deep are intended to help distribute the flow uniformly
through the wetland vegetation and to trap course sediments. These areas are most effective at removing
sediments and pollutants associated with sediments such as phosphorus, metals, and some organic
compounds. Open water areas also facilitate destruction of pathogens by exposing them to sunlight.
Shallow water regions 1-2 ft (0.3-0.6 m) in depth are intended to support the growth of emergent
wetland vegetation, primarily cattails and bulrushes. These areas are most effective at removing
nutrients, and to a lesser extent metals, pathogens, and toxic compounds.
The time required to obtain effective pollutant removal during low flows is estimated to be typically 7-14
days, depending on site conditions and temperature (Kadlec and Knight, 1996). Most NTS sites are
designed for a 10-day retention time during low flow conditions.
Sediments and pollutants that tend to attach to sediments are primarily transported by higher flows from
storm events. Many of the NTS facilities are designed to detain and treat stormwater runoff by means of
reduced flow outlets that drain the stormwater over a period of about 36 hours. The depth of the stormwater
475
-------
quality pool is typically 3-4 ft (0.9-1.2 m) above the normal low flow water level (Figure 2), thus inundating
the wetland vegetation. Wetland vegetation would not be destroyed by inundation for short detention
periods.
Removal of pollutants from storm runoff will primarily occur by settling processes. Therefore the primary
pollutants removed from storm runoff are sediments and pollutants associated with sediments such as
phosphorus, metals, and some organic compounds. There will be little or no removal of dissolved nutrients
(e.g., nitrate) during detention of storm runoff.
Habitat enhancement is an important aspect of the NTS Plan. The selection and planting of riparian
vegetation between the wetlands and the surrounding habitat affects the habitat characteristics of the
wetlands. Where feasible, native riparian vegetation will be selected to enhance habitat for endangered
avian species.
San Diego Creek has consistently high levels of selenium, which originate from natural sources. A major
source of selenium is groundwater discharge to the San Diego Creek in a historical ephemeral lake and
marsh region. Selenium was historically immobilized and trapped in the marsh due to the presence of
reduced anoxic conditions. Drainage of the swamp in the early 1900's for agriculture allowed oxygenated
groundwater to flow through the marsh, creating soluble and mobile forms of selenium that are now being
flushed to the creek
Elevated selenium levels must be reduced in accordance with the draft TMDL for selenium. To address the
TMDL, the NTS Plan includes one facility for selenium removal (Site 67) located in the historical
ephemeral marsh region. The selenium treatment concept is to mimic the selenium sequestrating processes
that occurred in the historical marsh in a subsurface flow treatment wetland. Stream water would be
diverted through organic rich native soils under anoxic conditions, creating reduced forms of selenium that
are immobilized by sorption to the soil particles.
Facility Selection
Potential NTS sites were selected using a simple screening process. Staff at IRWD developed an initial list
of potential sites based on their knowledge of the watershed and information contained in their databases.
Following field visits, the initially selected sites were assessed by preliminary technical analyses and
institutional and community acceptance assessments. This process was followed by successive rounds in
which some sites were removed from further consideration, due to technical constraints or other
considerations, and replaced with new sites. In total, more than 60 sites were considered for the NTS Plan,
of which 44 were retained for detailed assessment. The location of all NTS sites is shown in the aerial
photograph in Figure 1.
The NTS facilities are categorized by their location in reference to stream channels and whether they are
being added to a flood retarding basin: Type I off-line facilities are adjacent to existing channels and
require diversion structures for influent and effluent to the facility; Type II in-line facilities are wetlands that
are established within existing stream channels; and Type III facilities are established within existing or
planned retarding basins, and make use of the local storm drains.
476
-------
Evaluation of the NTS Plan
The NTS Plan was evaluated using planning-level water quality models that primarily rely on local
hydrologic and water quality data, and data collected on the performance of local and national wetlands.
The purpose of the water quality models was to provide planning-level assessments of the NTS Plan
alternatives, and to evaluate the NTS contribution to TMDL compliance. The modeling strategy used to
evaluate the NTS Plan is summarized in the following steps:
1. Forecast future land uses: The NTS Plan was evaluated under the assumptions of complete
development in the watershed ("build-out" conditions) and full implementation of the NTS facilities.
The intent was to obtain a measure of the total effectiveness of the NTS Plan under ultimate
watershed conditions. Build-out land use conditions were estimated from zoning maps and local
agency land-use plans.
2. Forecast hydrology and pollutant loads under build-out conditions: Estimates of flow conditions and
pollutant loads were forecasted for future land use conditions using available monitoring information
and statistical correlations between current and projected land uses. In cases where there was
insufficient monitoring data, land-use based pollutant load estimates were developed from regional
monitoring information.
3. Estimate load reductions in the NTS facilities: Water quality models were developed to estimate
pollutant loads and load reductions occurring in individual NTS facilities and as a network of NTS
facilities. The water quality models take into account the interrelationships of individual facilities
that occur when pollutant removals in up-stream facilities affect pollutant loads at down-stream
facilities. Separate models were developed for low flow and storm flow conditions and different
pollutants were modeled for different flow regimes, depending on the pollutant characteristics and
TMDL requirements.
Low Flow Conditions: Load reduction estimates for low flow conditions were modeled as a first
order kinetics process using coefficients derived from data collected at local constructed
wetlands. Seasonal rate coefficients were used to account for temperature differences. Flow and
load estimates included evaporation losses, and pollutant contributions from groundwater
discharge to stream channels. Pertinent assumptions are summarized in Table 3..
Storm Conditions: The treatment effectiveness of runoff from storm events was assessed on an
average annual basis. A 21-year period of recorded rainfall was used to estimate: the annual
runoff quantities. Pollutant concentrations were estimated with the event mean concentration
(EMC) values from available local and regional monitoring information. Load reduction was
estimated with data from the USEPA's Nationwide BMP database (ASCE, 2001; Strecker et. al.,
2001). Pertinent assumptions are summarized in Table 4.
477
-------
Table 3: Approach and Assumptions used in the Low Flow Model.
Parameter / Process
Load reduction
Steady state
Atmospheric sources
Stream flow
Evapotranspiration
Infiltration
Background concentration
First-order rate constant
Residence time
Open water ratio
Period of operation
Influent concentration
Assumption / Approach
Evaluated with a first-order kinetics model with background concentration.
Seasonal average steady state conditions were assumed.
Water and pollutants from atmospheric sources were assumed negligible compared
with influents flows and loads.
Estimated with seasonal based empirical relationships that account for projected
land-use and groundwater contributions. Equations were developed by regression
analysis using available stream flow data and geographical information.
Estimated with available monthly average reference evapotranspiration.
Assumed negligible based on planned use of liners in areas with poor soil
conditions.
1 mg/L for total nitrogen; 50 MPN/100 mL for fecal coliform bacteria
TN removal: 0.55 and 0.25/day for the dry and wet seasons, respectively .
Fecal coliform: 75 m/year (area based)
7- 14 days
Open water areas constitute 20% of the wetlands, except near airports where no
open water areas were included.
165 days in the dry season; 150 days in the wet season
Average seasonal concentrations estimated from available monitoring information
Table 4: Approach and Assumptions used in the Storm Flow Model.
Parameter /Process
Annual model
Sediment sources
Annual rainfall depth
Runoff volume
Stormwater pollutant
concentrations
Capture efficiency
Background concentration
BMP performance
Assumption / Approach
Uses annual rainfall depths to estimate annual runoff volume and pollutant loads.
Post-construction sediment sources from urban and open space areas. Does not
address in -stream sediment sources.
Determined from monthly rainfall records. Rainfall was reduced by a correction
factor to account for events that produce no appreciable runoff.
Estimated as a function of land-use with the rationale formula where the runoff
coefficient is expressed as a linear function of percent imperviousness.
Estimated with land-use based Event Mean Concentration (EMC) values from
available local and regional stormwater monitoring data.
Estimated by routing stormwater runoff volumes obtained from hourly rainfall data
through the NTS facilities. Different routing rules were used depending on the
facility type.
1 mg/L for total nitrogen; 50 MPN/100 mL for fecal coliform bacteria
Data available from the USEPA's Nationwide BMP data was assumed to be
representative of the treatment performance in the NTS facilities.
Estimated Nitrogen Removal
Nitrogen removal was modeled only for low flow conditions, consistent with the TMDL requirements. The
modeling results indicate that the NTS facilities would remove about 227,500 Ibs (103,200 kg) of total
nitrogen (TN) annually, and that both dry and wet season TMDLs would be met (Table 5). In general, wet-
season TMDLs are more difficult to achieve because loads are higher in the wet season and removal rates
are smaller due to lower temperatures and resulting biochemical activity.
The modeling results reveal that a large proportion of the TN removal occurs at the larger sites located in
the downstream reaches of the watershed. Smaller sites distributed in the upstream reaches remove less TN
on a percentage basis, but contribute to the improvement of 'local' in-stream water quality. Model
478
-------
predictions indicate the NTS Plan would significantly reduce in-stream TN concentrations (Figure 3),
meeting water quality objectives at nearly all locations.
Table 5: Summary of Estimated TN Loads to Newport Bay that show that TMDL loading limits are predicted to be met
by implementation of the NTS Plan.
Load to Newport Bay
Without Plan (Ibs/season)
Load Removed by NTS (Ibs/season)
With Plan (Ibs/season)
TMDL (Ibs/season)
Dry Season Low Flow
200,000
119,500
80,500
153,861 (2007)
Wet Season Low Flow
237,500
108,000
129,500
144,364 (2012)
In-Stream TN Concentration - San Diego Creek
20 I
18-
HT 16 '
"3> 14 -
v 10 -
o
o 8"
0 6-
z
1- 4 •
2 -
B Average
Reach 1, Dry Season
measured TN concentration (1991-2000)
~| •Estimated without Plan I
n Estimated with Plan
1
38^=
0 c 0
Q- 2 0)
™ £ 5
o
Q
n
II
1
13mg/LTIN
InilL
i i 1 f S i t 1 1 I
B- to ^ ro^WO (jg^i
^ 5 ^ -1-1 B- mE ^
CO t r^ ^ o CO
2 2 s m
In-Stream TN Concentration - San Diego Creek Reach
14 -I
2"
D) 12
^10-
TO
"£ 6 •
0)
O
C 4 .
O
O
2 2-
2, Dry Season
n Average measured TN concentration (1991-2000)
I II
• Estimated without Plan
n Estimated with Plan
t5mg/LTIN
II
n« «i = 0^5,0^-5
°^|2°I""<
-------
The NTS Plan was not intended to treat in-stream sources of phosphorus; therefore it was assumed that bank
stabilization measures and other BMPs would effectively control in-stream sources at build-out.
Table 6: Summary of Sediment Sources, TMDL Allocations, and Modeling Approach
Sediment Source
In-stream erosion & scouring from
In-Line sediment basins
Dedicated open space
Agricultural
Urban (commercial, residential,
transportation, and industrial)
Construction activities
TMDL Allocation (tons/year)
None
28,000 discharged to Newport Bay
28,000 retained in sediment basins
1 9,000 discharged to Newport Bay
1 9,000 retained in sediment basins
2,500 discharged to Newport Bay
2,500 retained in sediment basins
13,000 discharged to Newport Bay
13,000 retained in sediment basins
Modeled in NTS Evaluation
No
Yes
Yes
Yes
No
The storm flow model is based on rainfall/runoff relationships for the annual precipitation record from
1978-1998, as well as the average annual rainfall for this 21-year period. Model results estimate that NTS
facilities remove about 1,600 tons/yr (1,451,000 kg/yr) of sediment during average rainfall conditions, or
about 25 percent of the mean annual sediment load attributed to urban and open space land sources under
build-out conditions. The NTS facilities would remove an estimated 7,300 Ibs (3,311 kg) of TP per average
year (Figure 4), or about 11% of the annual TP load from urban and open space sources. The 2012 TMDL
target for TP (62,000 Ibs/yr or 28,120 kg/yr) would be met in all but the wettest rainfall years. The two
years where the TMDL was not met were the two highest rainfall years in the 21-year record, with 1998 also
being a record rainfall El Nino year.
Annual Sediment Load to Newport Bay from Land
Sources at Build-out
14,000
12,000
— 10,000
to
c
O 8,000
6,000
4,000
2,000
•Without Plan
Q Primary NTS Plan
water year
Annual TP Load to Newport Bay from Land Sources at
Build-out
90000
80000
70000
60000
50000
40000
30000
20000
10000
• Without Plan
D Primary NTS Plan
water year
Figure 4: Estimated Sediment and TP Loads to Newport Bay from Storm Runoff.
Estimated Coliform Removal
The TMDL for pathogen indicators (fecal coliform bacteria) is valid throughout the year under all flow
regimes. Therefore, fecal coliform removal was modeled for both low flow and storm flow conditions.
Low flow conditions were modeled as a time series for comparison with monitoring data from a one-year
monitoring period beginning in April 1999. Modeling results (Figure 5) indicate that during dry weather
base flow conditions, fecal coliform concentrations would be reduced below the 30-day geometric mean
standard of 200 MPN/lOOmL. The maximum 400 MPN/lOOmL standard would be met in most, but not all,
of the dry season low flows. The standards are not met during the wet season base flow conditions.
480
-------
The removal of pathogen indicators from storm runoff was modeled as equivalent fecal coliform loads.
Modeling results suggest the NTS facilities will reduce fecal coliform concentrations by about 20 percent,
but that concentrations entering Newport Bay will remain well above the TMDL targets during storm flow
conditions. The inability to meet TMDL targets in the wet season runoff is attributed to the overwhelming
pathogen loads generated during storm events.
Fecal coliform concentration - dry
weather low flow conditions
100000
*~ Measured at Campus Drive
10000 i -°- Primary NTS Plan
8 -i
4/5/99 6/5/99 8/5/99 10/5/99 12/5/99 2/5/00
30 day geometric mean -
dry weather low flow conditions
10000
El
o s.
£ =
HI U
§§
Si
1000
100
10
1
-Measured at Campus Drive
-Primary NTS Plar
4/5/99 6/5/99 8/5/99 10/5/99 12/5/99 2/5/00
Figure 5: Measured and Estimated Fecal Coliform Concentrations
Estimated Metals Removal
Monitoring data indicate that the majority of metal loads in San Diego Creek are sorbed metals associated
with sediment loads from winter storm events. Therefore, assessment of metal load reduction was carried
out for total metal loads under storm flow conditions. Removal of total metals in NTS facilities was
evaluated for copper, lead, and zinc. Translators were used (Table 7) to estimate the dissolved metals
fraction of the estimated total metal loads for comparison with the draft TMDL.
Table 7: Fraction of Dissolved Metals in Total Metal Concentration Measurements
Metal
Copper
Lead
Zinc
Estimated Fraction Dissolved -
storm flow (1)
41.4%
17.5%
37.3 %
Estimated Fraction
Dissolved- low flow (2)
82.8 %
37.9 %
61.8%
(1) Based on average concentrations in storm monitoring data.
(2) Based on average concentrations in base flow (dry weather) monitoring data.
Average annual loads to Newport Bay from urban and open land sources for total copper, lead, and zinc are
estimated at about 2,700, 1,100, and 21,000 pounds, respectively. The NTS Plan is estimated to remove
about nine percent of the total copper and lead loads, and about 13 percent of the total zinc load attributable
to urban and open land sources. The estimated annual total metal loads were converted to average annual
dissolved metal concentrations to allow comparison with the TMDL objectives. Results indicate (Table 8)
that the TMDL objective at the large and medium flow regimes is achieved on 'average' at build-out for
both with and without NTS Plan conditions. The results suggest that TMDL compliance is most easily
achieved for lead and zinc and is more difficult to achieve for copper. These 'average' results to do not
indicate the frequency at which occasional exceedances could occur.
481
-------
Table 8: Estimated Average Annual Dissolved Metal Concentration in Storm Flows
Metal
Copper
Lead
Zinc
Average annual total metal
load in Ibs at build -out ^
Without
Plan
2970
1240
23800
With
Plan
2680
1130
20400
Initial
Phase
2790
1170
21600
Average annual dissolved
metal concn in storm flow at
build-out (ug/L) (2)
Without
Plan
12.1
2.1
87.4
With
Plan
10.9
1.9
74.9
Initial
Phase
11.4
2.0
79.3
TMDL for medium
flow regime
(182-814 cfs)
Acute
(ug/L)
30.2
162
243
Chronic
(ug/L)
18.7
6.3
244
TMDL for large
flow regime
(>814cfs)
Acute
(ug/L)
25.5
208
135
Selenium Removal
The design of the selenium treatment wetland at Site 67 was partially based on a successful treatment
facility operating near the San Francisco Bay, which has similar site characteristics (Hansen et al., 1998).
This facility was able to achieve selenium reduction below the water quality standard of 5 ppb. The
proposed selenium treatment wetland at Site 67 is located in the historical marsh region, which is thought to
be a significant source area in the watershed. This facility is estimated to remove between 235-500 Ibs
(107-227 kg) per year, or about 20 to 50 percent of the low flow selenium loads to Newport Bay. While the
facility will significantly contribute to the reduction of low flow selenium loads, it may not, by itself, allow
for attainment of the proposed TMDL targets. This is because other tributaries also contribute selenium
loads to Newport Bay.
As selenium removal is relatively less well-under stood, and in particular, is much less well-understood as an
anoxic treatment system, the project has conducted column tests of different materials including chopped
cattails, coconut shells, and green waste, as potential carbon-providing media for the anoxic treatment
design. The next testing that is currently underway is at the mesocosm scale. The media that was chosen for
further testing was the chopped cattails. Two side-by-side mesocosm facilities have been built to provide
longer-term testing. The latest results of this testing will be presented at the conference and will also be
available on the project web site when complete. Initial results are showing that selenium is being reduced
to below laboratory detection limits.
Toxics Removal
The effectiveness of the NTS Plan for removing pesticides and organic compounds was not quantified
because there is insufficient information about the sources of these compounds and about their treatment
effectiveness in constructed wetlands. A literature review suggests the pesticides diazinon and chlorpyrifos
have characteristics amenable for effective treatment in constructed wetlands; namely they are relatively
insoluble, they are moderately to strongly sorbing, and they exhibit low to moderate persistence in soils.
Limited data from the existing water quality treatment wetlands at the San Joaquin Marsh indicate that a
high level of diazinon removal is occurring in the marsh.
Elements of NTS Plan
Maintenance
Regular and unscheduled maintenance activities will be required for all NTS facilities. Safe Harbor and
access agreements will be processed to ensure that maintenance requirements can be carried out.
Maintenance activities will include: trash and debris removal, pump servicing, vegetation removal and
planting, sediment removal, installation and removal of seasonal weirs, vector control activities, and
482
-------
emergency repairs. Minimization measures will be undertaken to limit impacts to wildlife and habitat from
maintenance activities.
Monitoring
Monitoring is a key component of the NTS Plan. There are three aspects to the monitoring program: routine
monitoring, site performance monitoring, and TMDL compliance monitoring. Routine monitoring activities
include site inspections, sediment accumulation monitoring, vegetation monitoring, monitoring of pollutant
accumulation and distribution, and vector pest monitoring. Detailed performance monitoring will be
conducted for a few selected NTS facilities to evaluate their treatment effectiveness and operating
constraints. Experience gained from these assessments will be used to improve designs and operation
practices of the NTS facilities. Regional monitoring will be conducted to assess the performance of the
entire NTS network, in combination with other BMPs, for meeting the TMDL and other goals.
Vector Control
Wetlands can provide breeding habitat for numerous pests and vectors, most notably Mosquitoes. A
comprehensive Vector Control Plan was developed, which includes the use of Mosquito Fish and the
application of a natural microbial pesticide (Bacillus thuringiensis israeliemus, Bti) for the control of
mosquitoes. With the increasing attention being paid to West Nile Virus, the control of Mosquito's will be
increasingly important. The Vector Control Plan was developed with the local vector control agency.
Implementation of the plan will be carried out by the same agency to ensure its success. With the West Nile
virus concerns, the Vector Control Plan is receiving additional attention, as it should.
Program Modification
The NTS Plan is intended to be flexible. The NTS Plan would be formally evaluated on a regular basis to
ensure that it is working as intended and to evaluate changes to the program that can improve the overall
performance. Sites could be added or deleted in response to new opportunities, needs, or constraints. Site
designs and operation practices could be changed as monitoring experience is gained.
Example Designs
The first example of an urban retrofit for establishment of constructed wetlands is the El-Modena/Irvine
Retarding Basin. This 9.5-acre (2.84 hectare) retarding basin is located within a fully developed residential
and highly urban setting. The basin was designed to retard peak flood flows in the adjacent El-
Modena/Irvine Channel, which drains approximately 1.6 mi (4.14 km2) of residential areas in the upper
reaches of the Peters Canyon Watershed.
The basin was originally designed with a water park in the floor of the basin, below the flood allocation
pool, which is considered dead storage. The water park was to include a live stream and a waterfall, but was
never implemented. The dead storage area is seen as the bare earth region in the photos shown in Figure 6.
Notice the mounded area in Photo 2, which was to have been an island in the center of the water park. The
basin is dry throughout most of the year, as winter storms of the magnitude that would cause any flow into
the basin occur very infrequently. A portion of the flood flows that are infrequently diverted into the basin
are retained in the dead storage area below the flood allocation pool. This water either infiltrates or
evaporates.
483
-------
Figure 6: El Modena/lrvine Retarding Basin. View in Photo 1 is from the upper end, near the diversion location. View
in Photo 2 is from the lower end, near the discharge location.
Gravity diversion
from El Modena /
Irvine Channel
Marsh area with emergent
plants; Inundated 1-2 ft
Open water pools;
Inundated 6ft
N
i
0 ' 200 ft.
Ill I
Approximate boundary
of storm water quality
storage pool
Island planted
with riparian
vegetation
Pumped discharge to El
Modena / Irvine Channel
Section A-A'
Cattails &
Bulrushes
Water depth- 1-2 ft
during low flow
conditions
Section B-B'
Open water pool
depth ~ 6 ft during
low flow conditions
Not To Scale
Small storm water
quality pool,
depth -2-4 ft.
Figure 7: Conceptual Design of a Constructed Wetlands Retrofit in the El Modena/lrvine Retarding Basin.
484
-------
The retrofit concept is to establish constructed wetlands within the dead storage area in the bottom of the El-
Modena/Irvine Retarding Basin. The wetlands would treat nuisance (low) flows, and runoff from smaller
storms, as well as the first-flush flows from larger storms. A geotextile clay liner would eliminate
infiltration losses from the wetland. Figure 7 shows a conceptual design of the proposed facility. The
wetlands consist of 0.66 acres (0.27 hectares) of shallow water marsh with emergent cattails and bulrushes,
0.17 acres (0.07 hectares) of open water areas 4-6 ft (1.2-1.8 m) deep, and 0.5 acres (0.2 hectares) of re-
vegetation area for native riparian habitat. The estimated average low flows during the dry and wet seasons
are 0.07 cfs (2 L/s) and 0.12 cfs (340 L/s), respectively. The average residence time during low flow
conditions is about 10 days. The stormwater quality treatment pool is on top of the low flow water level.
The stormwater treatment capacity is about 2.7 acre-ft (3,330 m3) (average depth of 2 ft or 0.6 m), with a
detention time between 48 and 96 hours (draw-down time).
A second example site includes an "in-line" facility. This is one that will only treat low flows. These
facilities will be located with the drainage system and will provide treatment of low flow discharges.
During storm events they would not be expected to provide any treatment. One of these sites is the
Woodbridge In-line facility. Figure 8 shows several photographs of the existing channel. The channel in
much of the reach is an earthen channel with limited habitat value. However, the placement of wetlands
within such a system is expected to improve habitat while also improving water quality. In California, the
use of "in-stream" treatment facilities has been controversial, with at least one Regional Water Quality
Control Board not allowing the use of "regional" treatment systems such as these. It is the author's opinion
that not allowing regional treatment or not allowing treatment within a highly degraded stream such as this
one is not a wise ecological approach.
Photo 1 & 2 -San Diego Crk, looking downstream at grade control structure between East Yale Loop and Creek Rd.
Photo 3 - San Diego Crk, looking upstream from
grade control structure toward E. Yale Loop overpass.
Figure 8. Woodbridge Site Photographs
-:.. -Si-
Photo 4 - San Diego Crk looking downstream from
grade control structure at energy dissipaters.
485
-------
Figure 9 shows an aerial photograph and conceptual layout of the facilities. Figure 10 shows a concept
sketch of one of the facilities. The weirs may also require more maintenance than off-line facilities,
including removing materials over the course of the year to maintain pooled water above the weirs. In a
very space-constrained watershed, however, where dry-weather water quality is an issue, these types of
facilities can provide significant benefits.
•Jto£^?»S»Rpck Weirs
Figure 9. Aerial Photograph and Conceptual Layout of Wood bridge Facility showing the planned series of shallow
linear wetlands within San Diego Creek.
1000ft.
f %^
Gravel and
rock d^m
Section A-A' T^
Figure 10. Conceptual Drawing of In-line Facility, showing a plan view along with cross-sections of the planned
gravel and rock dams (2 to 3 feet in height).
486
-------
Some professionals argue against "in-line" or "in-stream" treatment. However, man-made earthen or rip-
wrapped channels with engineered drop structures are not natural streams. Because of the degraded status
of these highly maintained flood control channels, the NTS Plan would improve both habitat and water
quality. In a highly urbanized watershed such as the San Diego Creek Watershed, in-line treatment such as
this may be one of the few options for improvement in water quality over the shorter-term.
Discussion
The estimated cost to provide low-flow treatment of urban runoff in a sanitary treatment plant is greater than
$60 million in construction costs, with annual operation and maintenance costs of about $5 million. The
NTS System is expected to cost about $12.2 million for first-phase construction of the 13 NTS sites, and
$1.1 million annually for ongoing operations, maintenance, and monitoring. This does not include the cost
of projects funded by local developers or costs of second-phase regional project sites. A comparison of the
capital cost per unit pollutant removed, indicates that the treatment plant is about three times more costly for
TN removal from low flows, and about twice as costly for removal of copper from storm runoff.
The San Diego Creek Natural Treatment Systems Plan has been designed to result in a cost-effective
solution that meets many goals. The effectiveness of the NTS Plan will ultimately be determined through
the long-term coordinated efforts, spanning the planning, implementation, and program evaluation stages.
Observations and conclusions from the development and initial evaluation of the NTS Plan are:
• Retrofit options are necessary to meet water quality goals in watersheds that are highly developed. It
is possible to develop cost-effective regional retrofit solutions on a large watershed basis that would
result in significant water quality improvements;
• Existing flood control basins and conveyance facilities can be cost-effectively retrofitted;
• The NTS Plan has resulted from a cooperative problem-solving focus by municipalities,
development interests, water and sewer providers, and environmental groups. This effort has not
focused on just meeting single-purpose requirements, and therefore has resulted in a more robust
plan. Consequently, the NTS approach can achieve multiple benefits, including habitat and aesthetic
values;
• The NTS Plan was developed in a relatively short 15-month time frame, demonstrating that planning
efforts can be accelerated when there are motivated interests; and
• Cost-recovery from other sources of funds is possible when urban runoff treatment requirements
include treating dry weather flows.
Acknowledgments
The authors gratefully acknowledge the support of the Irvine Ranch Water District, The Irvine Company,
County of Orange, Cities in the San Diego Creek Watershed, and local environmental groups. Efforts from
all members of the NTS project team have led to the successful development of the NTS Plan. In particular,
we acknowledge the contributions of John Tettemer, Sat Tamaribuchi, Dr. Alex Home, Dick Diamond, and
Bon Terra Consulting.
487
-------
References
ASCE, 2001. EPA/ASCE National BMP Database, www.btnpdatabase.org
Hansen, D., PJ. Duda, A. Zayed, and N. Terry, 1998. Selenium removal by constructed wetlands: Role of
biological volatilization, Environ. Sci & Technol., 32, 591-597.
Kadlec, R.H. and R.L. Knight 1996. Treatment Wetlands - Theory and Implementation, CRC Lewis
Publishers.
RWQCB, 2000. Final Problem Statement for the Total Maximum Daily Load for Toxic Substances in
Newport Bay and San Diego Creek, California Regional Water Quality Control Board, Santa Ana Region,
December 15, 2000.
Strecker, E.W., Spring 1996. Use of wetlands for stormwater pollution control, Infrastructure^ 48-66.
Strecker, E.W., M.M. Quigley, B.R. Urbonas, I.E. Jones, and J.K. Clary, 2001. Determining urban storm
water BMP effectiveness, ASCE J. of Water Resources Planning and Management., 125(3), 144-149.
USEPA (Region 9), 1998a. Total Maximum Daily Loads for Sediment and Monitoring and Implementation
Recommendations; San Diego Creek and Newport Bay, California
USEPA (Region 9), 1998b. Total Maximum Daily Loads for Nutrients; San Diego Creek and Newport Bay,
California
USEPA, 2002. Total Maximum Daily Loads for Toxic Pollutants, San Diego Creek and Newport Bay,
California. Public Review Draft April 12, 2002, http://www.epa.gov/region09/water/tmdl/
488
-------
"SHERLOCKS OF STORMWATER"
EFFECTIVE INVESTIGATION TECHNIQUES FOR ILLICIT
CONNECTION AND DISCHARGE DETECTION
Dean C. Tuomari/Susan Thompson
Wayne County Department of Environment
Watershed Management Division
3600 Commerce Court
Wayne, MI 48184
ABSTRACT
"Come, Watson, Come! The Game is afoot!'..." (Doyle, 1930) Wayne County has operated an Illicit
Connection and Discharge Elimination Program for over 15 years. Its staff has gained valuable
investigative expertise by experimenting with many different methods, committing lots of trial and error,
and having a little bit of luck. Investigating for illicit discharges in the field is very similar to Holmes and
Watson solving a case - it requires a mix of science, detection, deduction, and persistence.
This paper presents investigation techniques used effectively to identify illicit connections and discharges.
These techniques are: Identifying priority areas (i.e. "hot spots"), outfall survey, facility dye testing,
televising sewer systems, intensive water sampling, smoke testing, and other creative means. Each
technique, its advantages and disadvantages, and the best application for each method are described in
detail.
In 1999, the Illicit Connection Discharge Elimination Training Program was created and implemented by
the Wayne County Department of Environment, Watershed Management Division (WCDOE-WMD).
The program was developed to provide training for local and regional governments responsible for locating
and eliminating illicit discharges to surface waters. Wayne County determined that such a program is an
effective means of transferring technology to others. The key goals of the training program are: Sharing our
expertise with other local units of government involved in stormwater management and collaborating efforts
to reduce improper discharges to surface water.
The Wayne County Training Program is consistent with the Illicit Discharge Elimination Plan (IDEP)
requirements of the Michigan Voluntary Storm Water Permit (MIG6100000) and the EPA Phase n
Stormwater Permit Regulations. The training program consists of five modules and two specialty training
sessions. The modules are: Overview, Basic Investigations, Advanced Investigations, Construction Related
Illicit Discharges, Combined Basic/Advanced Investigations and two specialty training sessions. The
specialty training sessions are titled "Recognizing and Reporting Illicit Discharges" and "Illicit Discharge
Investigation Exercise." Nearly 800 people, representing various local units of government, attended the
training sessions through September 23, 2002. As a result of these training efforts, 82 illicit discharges were
eliminated, preventing an estimated 3.5 million gallons/year of polluted water from entering Michigan
surface waters. Wayne County will explain its experiences and those of other agencies with selected
investigative methods. A case study based on an actual investigation exemplifying how some of the
techniques are used in the field is presented.
The "Sherlocks of Stormwater" will assist others needing to prepare and implement an Illicit Discharge
Elimination Plan.
489
-------
Introduction
"Come, Watson, Come! The Game's afoot!'...." (Doyle, 1930) Wayne County has operated an Illicit
Connection and Discharge Elimination Program for over 15 years. Its staff gained valuable investigative
expertise by experimenting with many different methods, committing lots of trial and error, and having a
little bit of luck. Investigating for illicit discharges in the field is very similar to Holmes and Watson
solving a case - it needs a mix of persistence, science, detection, and deduction. A brief overview of the
Rouge River National Wet Weather Demonstration Project, geography of the Rouge River Watershed and
illicit connection and illicit discharge definitions are provided. Wayne County's Illicit Discharge
Elimination Program, training program, and the reasons why they are necessary is introduced.
The Illicit Discharge Elimination Plan (IDEP) Training curriculum, formulation and content are outlined.
The primary focus of this paper is introducing the variety of techniques used to identify illicit connections
and discharges used by Wayne County and other local agencies. Based on Wayne County's experience in
IDEP investigations, each technique is described and the advantages and disadvantages to each method
listed. A case study illustrates how the different techniques are used in field investigations. In conclusion,
Wayne County presents the successes achieved by the implementation of its IDEP Plan and IDEP Training
Program.
Rouge River Project Overview
The Rouge River is located in the southeast region of lower Michigan. It encompasses an area of about 467
square miles and is highly urbanized. Approximately 1.5 million residents of 48 municipalities live and
work in the watershed.
The Rouge River is tributary of the Detroit River, a part of the southeast Michigan area identified as an
"area of concern", by the International Joint Commission (UC). In response to this bleak assessment and
demands of local residents for improved water quality, the State of Michigan created a series of Remedial
Action Plans (RAP) to address specific sources of pollution of the state surface waters.
The Rouge River Remedial Action Plan (RAP) is an ambitious 20-year plan to clean up and restore the river
to a fishable and swimmable state. The RAP focused on sources of pollution such as Combined Sewer
Overflows (CSOs), Industrial Pollutant Discharges, and Non-Point Source Pollution. The RAP contains a
recommendation that "programs to eliminate improper connections to storm drains should be
implemented..." (SEMCOG, 1988). In 1987, Wayne County developed and implemented a program for
reducing pollutant loadings to the Rouge River. This program detects and eliminates illicit discharges
and/or improper/illegal connections to Wayne County storm sewers and surface waters. An illicit
connection is defined as a pipe intended for a sanitary sewer that is directly connected to, or indirectly
drains to a storm sewer system or surface water body. An illicit discharge is the indirect migration of
pollutants by storm water to a surface water body. Examples of illicit discharges are: failing on-site sewage
disposal systems, spilling or dumping of materials, and illicit connections.
In November 1999, the U.S. Environmental Protection Agency (USEPA) promulgated Phase U of the
National Pollutant Discharge Elimination System (NPDES) storm water regulations, which affects virtually
all communities in southeast Michigan, including Wayne County. Wayne County, through its Rouge River
National Wet Weather Demonstration Project (Rouge Project) assisted the Michigan Department of
Environmental Quality (MDEQ) in the development of a new watershed-based General Permit for
490
-------
municipal storm water discharges (General Permit). The MDEQ General Permit was approved by EPA as
an option available to local communities and other public agencies to comply with the requirements of the
Phase n federal NDPES storm water regulations.
One of the requirements of the federal Phase n NPDES storm water regulations and the MDEQ General
Permit is to develop, implement, and enforce a program to eliminate improper connections to the storm
sewer system and other improper discharges to surface waters. During 1999, over 45 communities and
agencies in the Rouge River watershed, including Wayne County, have received coverage under the MDEQ
storm water General Permit and have initiated the illicit discharge elimination program (IDEP) requirements
of the permit. Wayne County recognized this as an opportunity to share our considerable expertise in illicit
discharge investigations with others.
The Wayne County Illicit Connection/Discharge Elimination Plan (IDEP) Training Program was created
and implemented in 1999-2000. The training program was developed to provide training for county and
local community staff responsible for locating and eliminating illicit discharges to surface waters, as
required under the federal National Pollutant Discharge Elimination System (NPDES) regulations for
municipal storm water discharges. The training program consists of five modules and two specialty training
sessions. The modules are: Overview, Basic Investigations, Advanced Investigations, Construction Related
Illicit Discharges, and Combined Basic/Advanced Investigations. The Specialty Training Sessions are
entitled "Recognizing and Reporting Illicit Discharges" and "Illicit Discharges Investigation Exercise".
This paper provides a basic overview of the Advanced Investigations IDEP Training Module. It introduces
the techniques used to effectively identify illicit connections.
Finding the Problem Area: "In Quest Of A Solution"
Where to begin an investigation for illicit discharges?
Wayne County sewer sleuths begin compiling information on the targeted area. Information and data can be
gathered from many different sources; outfall surveys, referrals from other departments, known areas of
concern, review of existing water quality data, and complaint response. All this data is reviewed and
compared with existing data to determine if potential problems (i.e., "hot spots") exist. The goal of
identifying "hot spots" is to isolate the area where the problem exists and then locate the pollutant source.
Specifying the problem allows the investigator to select the type of parameters for field measurement. For
example, if sewage is the suspect problem, sampling for bacteria is useful in verification. Once the problem
is identified, additional sampling is performed, upstream and downstream of the "hot spot." Data from the
sampling events is compared and utilized to determine the area where the values are the highest. Once the
suspect pollutant is identified, sampling may be repeated as necessary to narrow down the geographical area
to a manageable size.
Outfall survey is also used as a screening tool to define investigation areas. It involves field observations of
the stream channel and conditions at outfall locations. If suspicious discharges or signs of past discharges
are seen, physical and chemical parameters are selected to identify the type of discharge. If observations at
an outfall triggers an investigation, tracking the suspect source moves upstream from the outfall along the
storm sewer system. Storm sewer manholes are opened and visual and physical observations for signs of
suspicious discharges are made. This process continues upstream and along sewer laterals until signs of a
491
-------
discharge are found. Once this area is isolated, the investigator can choose from the variety of techniques to
help track the source.
IDEP Tracking Techniques: "The Science of Deduction"
Like Sherlock Holmes, modern day IDEP detectives utilize their skills and knowledge to solve their cases.
However, there are techniques available to make detection of illicit discharges and connections more than
just deduction. There are four techniques commonly used by Wayne County and other local governmental
agencies in southeast Michigan when searching for sources of illicit discharges. Each of the techniques has
its advantages and disadvantages and there's no one "right" way. In some cases, a combination of methods
may be used in quest of a solution.
Investigative Methods: "The Sign of Four"
Intensive Sampling
Intensive sampling is defined as one of the two following situations: 1) many samples collected at many
locations, and 2) many samples collected at the same location over a specific period of time. This method is
effective if intermittent flows in a storm sewer or when a source is active in hours where routine sampling is
ineffective. For example, if a suspected source is a residence where persons are not home during the day,
peak flows typically occur during early morning or early evening.
Taking many samples at many locations is useful when isolating the area of a suspected illicit discharge,
especially when the survey area is large. The sample data can help narrow down an area where a problem
may exist, by comparing sampling data from different locations along the storm sewer line or stream
channel. Degree of concentration, or presence and/or absence of a pollutant demonstrated in the data, can
lead an investigator to an area of the potential source.
Intensive sampling techniques are good for isolating source areas for investigation, completing field data
gaps present between sampling events, off-hour sampling events (because staffing is unnecessary for
automatic sample collection), and in residential areas where intermittent flows are common.
There are also several disadvantages to the method. For example, it does not pinpoint the pollutant source
exactly and data variances may exist which makes it difficult to establish trends. Also, limited holding
times for certain parameters make it difficult to time sampling sessions and collecting many samples may be
expensive and require laboratory analysis and holding times. Finally, placing flow meters or automatic
samplers at a site may involve confined space entry, which requires additional training and equipment.
Dye Testing
Dye testing is an investigative technique that involves placing tracing dyes in a sewer system to determine
path of the flow. This method is effective for determining if illicit connections exist in a facility, of if there
are interconnections between sewer systems.
Wayne County extensively uses dye testing for illicit connection detection. When performing a dye test,
field staff walk through the facility to determine where the plumbing fixtures are and observe interior and
492
-------
exterior housekeeping practices. A dye testing plan is prepared and tracing dye is placed into plumbing
fixtures. The dye is flushed through the system with running water. A person is stationed at the sanitary
manhole down stream of the tested facility and alerts the team member inside the building when the dye is
observed in the sanitary sewer. Alternate dye colors are used so multiple fixtures can be tested
simultaneously. If dye is not observed in the sanitary sewer, the dye test is repeated until it is confirmed in
the sanitary sewer or in a storm sewer, or surface water body. If the dye from a fixture inside the building is
discovered in a location other than a sanitary sewer, it is an illicit connection.
Advantages of using dye testing for illicit connection detection are that dye testing is inexpensive, relatively
easy to do, points to a specific source, and does not require confined space entry.
Disadvantages to dye testing are that it may be difficult to see the dye in high-flow or turbid conditions, it is
time consuming in low flows, and entering a facility is necessary in order to conduct the test.
Televising
A remote camera with a video recorder is another means to search for illicit connections and discharges.
The self propelled camera is placed into the sewer line and the operator can view live footage of the sewer
line, so the condition of the sewer line and evidence of illicit taps can be seen. Televising is an effective
technique because it views active taps, provides a record of observations, and is the only way to observe
pipes between manholes. It can, however, be expensive, ineffective in determining if inactive taps convey
illicit discharges, time-consuming to interpret results, and in practical in water-filled or obstructed sewers.
Other Techniques
There are other methods various agencies use to search for illicit connections. Some municipalities use
smoke testing of storm sewers. A non-toxic smoke is introduced into a storm sewer and an illicit tap is
suspected if the smoke is observed in a sewer vent from a building. Some communities have sent stout -
hearted workers to survey storm sewers big enough to walk into in order to do "up-close" illicit discharge
surveys. Searching for illicit discharges is part art and part science. Imaginative ways are created to do
investigations often because no specific equipment exists, or because of cost-effectiveness. Ingenuity leads
to effective methods like placing a rope with oil-absorbent pads tied at measured intervals into a storm
sewer manhole located upstream of a facility where leaky underground oil storage tank was suspected. The
rope was pulled out of the sewer, and measured off to where oil was present. The distance was walked off
on the surface and the investigators ended up in front of the suspect facility. When presented with the
evidence, the owner admitted to the problem and repaired the leaky tank.
A Case Study: "Sherlock Holmes Gives a Demonstration"
"A Study in Scarlet" - Restoring Rouge River Recreational Opportunities
A goal of Wayne County is to return canoeing to selected portions of the Rouge River. To support that goal,
water quality must meet the State of Michigan bathing beach standards for Escherichia coliform (E. coli).
This standard is the daily geometric mean of three samples must be less than 300 cfu/lOOml of water. The
geometric mean of five sample events collected over 30 days must be less than 300 cfu/lOOmls of water.
493
-------
Review of available data and additional sampling found two impoundments, and a mile of river between
them, safe for canoeing during dry weather. Canoeing on the upstream impoundment became a reality and
received an overwhelmingly positive public response. Based on the success upstream, a second goal to
extend canoeing downstream of the second impoundment was set. Sampling of this stream section found
that E. coli. levels were very high below the dam of the second impoundment. A nine-foot wide storm
sewer outfall discharged at this location, making it a prime target for investigations. Wayne County sewer
sleuths were assigned to the case and set out to unravel the matter.
The sewer drainage area is approximately 157 acres, contains over 5.5 miles of enclosed storm sewer and
over 350 manholes. The storm sewer has one main line with many connecting branches. The land use in
this area is primarily residential.
Samples were collected along the main line at a variety of locations. A branch coming into the main line
near the outfall had higher E. coli levels than the others. Confirming sampling found very high levels of E.
coli., indicating a significant bacteria source upstream at this suspected branch. After a using a very
extensive sample regime designed to narrow the search area, efforts switched to sampling the main line at a
variety of locations to isolate the branch that contained the source. One Vi mile long branch line had
significant levels of E. coli. Storm sewers from an adjacent branches with similar land use had extremely
low levels ofE. coli..
Sampling up the line revealed increasing levels ofE. coli. Results at one manhole were over
160,000 cfu/lOOml. On the basis of this result, it was suspected that this short section of sewer line
contained a problem. This suspicion, and supporting data, was shared with the City representatives. During
this meeting, the County learned that a storm sewer separation project occurred in this area several years
ago. A sewer line or a tap may be misconnected, discharging sewage into the storm sewer and causing the
bacteria problem. The City and the County agreed on a plan of action to find the source of the E. coli. The
City agreed to televise the storm sewers. At one location on the sewer line, the camera dipped below the
water. Otherwise, no taps were found. The City agreed to dye test the homes along the sewer line for illicit
connections. The County drafted a letter and provided the City educational materials for mailing to the
homeowners. Dye testing did not uncover any illicit connections.
Subsequent sampling of the sewer line found very low levels ofE. coli. Repeated sampling up to two years
later found extremely low E. coli. levels. The problem seemed to disappear. What happened? The theory
is that someone in the neighborhood owned a recreational vehicle and discharged its holding tank into the
storm sewer or performed some other inappropriate action. With all of the activity in the neighborhood, the
County and City staff interacting with the residents, and the public education mailing, the culprit realized
their actions caused a problem and stopped. This is an investigation where a blend of intensive sampling,
sewer televising, dye testing and public education techniques were used in attempting to resolve the
problem.
Success of Wayne County's IDEP and Training Program
From October 1987, when the Illicit Discharge Elimination Plan was implemented, through December 2001,
Wayne County inspected 4,887 commercial, retail and industrial facilities for illicit connections. During
these inspections, field staff discovered 1,243 illicit connections at 326 facilities. Finding and eliminating
these illicit connections prevents and estimated 18 million gallons/year of polluted water and 4,600
494
-------
pounds/year of biological oxygen demand (BOD) and 16,000 pounds/year of total suspended solids (TSS)
from entering Wayne County surface waters.
Nearly 800 people, representing various agencies and communities throughout Michigan, and two
neighboring states, attended training sessions conducted by the County through September 30, 2002.
Ninety-eight percent of training session participants surveyed encourage persons with similar
responsibilities to attend the training course(s). One participant commented "This is the best training
session I have had in 20 years." The information these individuals gained from attending the training
session helped them in creating their own IDEPs. Successful programs include those implemented by
neighboring counties. Eighty-two illicit discharges were identified by IDEP investigations performed in the
Counties of Oakland, Washtenaw, and Wayne. The pollutant load into Michigan's surface waters from
these discharges is estimated to be 3.5 million gal/year of polluted water, 7,200 Ibs/yr BOD, and 25,000
Ibs/yr TSS.
Conclusion: "Light In The Darkness"
There is no "cookbook" or standard operating procedure for investigating illicit discharges and connections.
A combination of using the techniques presented here and ingenuity, plus a little luck, will go far in the
"Quest of a Solution."
"Eliminate all other factors, and the one which remains must be the truth." - Sherlock Holmes (Doyle,
1930).
Happy Hunting!
495
-------
ACKNOWLEDGEMENTS
We want to express our thanks to all those who helped in the development of the IDEP Training Program,
including Glenn Brown, P.E., John O'Meara, P.E., and Kelly A. Cave, P.E.. A special thanks to those who
took the time to share their expertise with IDEP training session attendees: James W. Ridgway, P.E., John
O'Meara, P.E., Annette Demaria, Earl Friese, Tom McNulty, P. E., Glenn Brown, P.E., Bonnie Pawloske,
Thomas Williams, Patrick Cullen, Noel Mullett, Gary White, and Steve Lichota.
Production support provided by Chris O'Meara, Judy Holt, and Latisha Martin. Appreciation to those took
the time out of their busy schedule to participate in the "Ad-Hoc" committees. In addition, recognition is
extended to the Wayne County staff who supported the effort by helping with a multitude of tasks.
Special Acknowledgement goes to Sir Arthur Conan Doyle, whose fictional detective inspired investigators
to use "the Science of Deduction" in their IDEP Activities.
This paper represents a summary of select elements from the ongoing efforts of many individuals and
organizations who are involved in the restoration of the Rogue River. The Rouge River National Wet
Weather Demonstration Project is funded, in part, by the United States Environmental Protection Agency
(EPA) Grant #XP995743-01, 02, 03, 04, 05, and 06 and #C995743-01. The views expressed by individual
authors are their own and do not necessarily reflect those of EPA. Mention of trade names, products, or
services does not convey, and should not be interpreted as conveying, official EPA approval, endorsement,
or recommendation.
References:
Conan Doyle, Sir Arthur "The Complete Casebook of Sherlock Holmes", Double Day, 1930.
Note: All sections with titles in quotations are either titles of Sir Arthur Conan Doyle's short stores or
novels.
Remedial Action Plan for the Rouge River Basin, (SEMCOG), 1988)
496
-------
Low Impact Development Strategies for Rural Communities
Neil Weinstein, Executive Director
Low Impact Development Center
Beltsville, Maryland
John Tippett, Executive Director
Friends of the Rappahannick, Inc.
Fredericksburg, Virginia
Abstract
The Friends of the Rappahannock and the Low Impact Development Center, Incorporated (both non-profit
organizations) are developing guidance and strategies for rural communities in Virginia to incorporate LID
into their local resource protection and regulatory programs. This project was funded by the National Fish
and Wildlife Foundation, under a grant from the Chesapeake Bay Program. The Town of Warsaw, Virginia
is the municipal partner in the grant. The first part of this effort includes evaluating state and local codes to
determine what, if any, necessary legislative, code, or local regulations need to be modified to include LID.
Identifying areas in the Town and land uses that are appropriate for LID technologies follow this effort. The
next step will be to develop materials for developers and plan reviewers to help guide them through the
development process when the use of LID is appropriate. The final step will be to design and implement a
small demonstration project that showcases LID features, such as rain gardens, soil amendments, permeable
pavers, and infiltration devices. This paper will document this effort and identify key issues that other
communities should consider when contemplating the use of LID.
Background
The Town of Warsaw, Virginia is a rural locality in Virginia's Northern Neck, located between the
Rappahannock and Potomac Rivers. Figure One shows the vicinity of the town in the watershed. The Town
and County have historically had strong economic ties to the surrounding rivers, although this has declined
in recent years due in particular to the decline of oyster harvests. The Town does not have a strong
economic base, and recently lost a major employer, a Levi's plant. The Town recently annexed a portion of
its "parent" County for the purposes of economic development. This former agricultural land is highly
suitable for development, and is situated along the area's major 4-lane highway. The nature of future
development in Warsaw is currently unclear, although current trends tend toward assisted-living and
retirement communities, along with supporting services. Town officials expressed an interest in Low
Impact Development strategies after seeing presentations at various local government and watershed
management conferences. They were concerned about the stormwater infrastructure costs associated with
new development in the annexed land, as well as with the aesthetic and environmental impacts of
conventional pond treatment of stormwater runoff. The Town currently has only one stormwater
management pond, that was recently put in as part of a new shopping center. There have been numerous
complaints by the property manager and adjacent property owners about the maintenance and aesthetics of
the facility. The town is also concerned about the inspection, ongoing maintenance, and potential
rehabilitation costs of conventional end of pipe pond systems. This has caused the town to revaluate its
existing stormwater program. Figure 2 is a map of the annexed areas and drainage master plan that shows
existing drainage problems and projected stormwater pond locations. The Town views LID strategies as a
means of reducing costs while also increasing community aesthetics and environmental protection. The
Low Impact Development Center, the Friends of tie Rappahannock, and the Town of Warsaw teamed up on
497
-------
a joint grant proposal to the EPA Chesapeake Bay Program (through the national Fish and Wildlife
Foundation) to develop a model approach for incorporating LID in rural communities.
Chesapeake
Bay
Figure 1: Rappahanock Watershed
Warsaw: Stormwater Infrastructure, and Possible Regional Stormwater Pond Sites
/
Figure 2: Potential Pond Locations
498
-------
Evaluation of Local Codes
The project initiated with an evaluation of local (Town and County) codes and ordinances to determine
compatibility with LID. Most local governments, especially rural ones, reference the State BMP design
manual and Stormwater handbooks for guidance (VADCR 2001). A review of the local and state guidance
indicated that the codes allowed the use of many (though not all) types of LID stormwater management
practices. However, there were no mechanisms in the language to promote LID designs in lieu of
conventional approaches. Additionally, the conventional approach was designed around detention/retention
of the 2-year storm, while the LID approach is designed around the replication of pre-development
hydrology, which focuses on infiltration of the increase in "initial abstraction" on a site, and maintaining
pre-development Time of Concentration.
While practices such as bioretention were permissible in the state guidance, there were other practices
without design guidelines or standards by which to calculate pollutant removal or water volume detention.
Most notable was the LID practice of "amended soils." Another deficiency in the stormwater guidance was
a table used to determine appropriate BMPs for a site. The guidance recommended using bioretention only
on projects with low levels of impervious cover. Another weakness was a specific recommendation against
the use of infiltration practices under parking lots.
Project leaders met with Commonwealth of Virginia officials to discuss these barriers. Most were agreed
upon for revision in subsequent volumes of the stormwater guidance. On the issue of the conventional
versus LID approach to stormwater management design, it was generally agreed that the LID approach
meets or exceeds the Commonwealth water quality and quantity requirements, as long as the designs also
meet the Commonwealth's provision for having an "adequate receiving channel" (Minimum Standard 19
VADCR).
Assessing Local Government Needs
The Town Manager's interest in LID stemmed form a desire to reduce infrastructure and maintenance costs,
to increase community aesthetics, and to reduce impacts to the local aquatic resource. Figure Four is a map
of the potential number of conventional ponds that could be constructed at the ultimate buildout of the
community under conventional stormwater management scenario. Based on the towns maintenance and
construction experience with the recently conventional management pond it recognized that the pond
strategy would potentially be unsustainable and would be impact other funded programs. Consequently, the
project was designed around developing a plan to institute LID as the standard development approach
Town-wide, and possibly to be expanded to the county in which the Town resides.
Project staff conducted meetings with Town and County officials to determine their needs in regard to
instituting an LID development program. The issue that emerged in the forefront was the lack of criteria
that local government plan reviewers had for assessing an LID site design. There were significant concerns,
based on prior experience, that "token LID" plans would be submitted (i.e., plans that included some LID
practices, but did not achieve the quantitative LID goals) and that staff would not have the means by which
to evaluate the merits of the plans. Additionally, there concern on the parts of local officials that the
development community was unfamiliar with the LID approach to site design and stormwater management,
and that it would be difficult to have quality LID plans submitted.
This project has far reaching implications for many rural Virginia communities. It demonstrates how local
governments can work with regulatory agencies to develop and implement a stormwater program that meets
499
-------
both regulatory requirements and community environmental and fiscal programs. For a relatively small
cost, communities can develop their own programs, obtain resources to train review and maintenance
personnel to deal with more complex storm water design and construction issues, and gain acceptance by
political, business, and citizens within the community for innovative programs.
Developing an Action Plan
Based on the evaluation of codes and local government needs, the following action items were developed:
1. Develop policy language for instituting LID as the standard practice for project site design and
stormwater management
2. Create easy-to-use LID review guidelines for local plan review staff
3. Create a reference document for developers to use in designing LID plans
4. Create an LID educational brochure targeted to citizens
5. Develop a list of specific recommendations for changes to Commonwealth stormwater design
guidelines to better support LID at the local level
Demonstration Project
A demonstration project to model the LID design approach is planned for a Virginia Department of
Transportation (VDOT) Commuter parking lot. A rain garden and pervious pavers are planned for the
demonstration. The project is currently pending funding from VDOT.
Project Products
The policy language developed for the Town establishes the LID approach as the standard methodology
within the jurisdiction for stormwater management methodology for new developments. The language
includes references to the LID National Manual for design guidelines, and to other guidance products
created under this project. The language is currently under review by the Town and County officials for
inclusion in the local stormwater management ordinance.
The guidelines for developers and plan reviewers underwent an iterative process of revision between the
project leaders, state stormwater management officials, and town staff. The resulting guidelines are
designed to lead a developer with little familiarity of LID through the process of creation a viable LID site
design. These guidelines are outlined through a series of checklists, flow charts, and references to guidance
documents and technical information that can be incorporated into the development process. Figure 3 is a
design process flow chart that was developed as part of the guidelines. The goal is the development of a site
to mimic pre-development levels of infiltration, runoff, and Time of Concentration. The guidelines include
the development of pre-, post- and "LID" curve numbers, and recommended means of accounting for
volume storage achieved by practices such as bioretention and amended soils. Also included is a flowchart
depicting the LID design process. An option for a hybrid approach (using conventional practices to make
up for excess volume not managed by LID practices) is built in to the guidelines, but is discouraged.
500
-------
Virymiu Stmmvater Maru.iy«;ri*jiit (VA SWW) ti*$Utti0n$
Figure 3: LID Flowchart
Project Follow Through
The projects have just been delivered to the local government and are currently being evaluated for formal
adoption in the ordinance. The rate of development in the Town is currently very low. The first project to
be reviewed under the new LID approach is expected to be completed within the next several months. The
products of this project are being made available to other local governments to help guide their adoption of
LID strategies. Additionally, a multimedia CD is currently being developed which chronicles the Warsaw
project and includes the project deliverables. Project products are available on the web at
http ://for. communitypoint.org.
Background and References
Friends of the Rappahannock Website, http//:for.communitypoint.org
Low Impact Development Center. 2002, Low Impact Development Stormwater Resource Website.
http://www.lid- stormwater.net
Prince George's County Maryland, 1999, Low Impact Development Design Strategies, Prince George's
County, Maryland
Virginia Department of Conservation and Recreation, 2001. Virginia Stormwater Management Handbook,
Richmond Virginia, http://www.dcr.state.va.us/sw/stormwat.htm
501
-------
AN ASSESSMENT AND PROPOSED CLASSIFICATION OF CURRENT
CONSTUCTION AND POST CONSTRUCTION STRUCTURAL BEST
MANAGEMENT PRACTICES (BMPs)
David A. Woelkers
Hydro Compliance Management, Inc.
Ann Arbor, Michigan
Marc S. Theisen, M.S., CPESC
SI Geosolutions
Chattanooga, Tennessee
Jerald S. Fifield, Ph.D., CPESC
Hydrodynamics, Inc.
Parker, Colorado
Abstract
The more stringent NPDES Phase II Storm Water regulations of the US Environmental Protection Agency (EPA)
Clean Water Act are set to take effect in March of 2003. This legislation will require a growing number of
municipalities, construction and industrial sites to develop, implement, and enforce storm water management
programs to reduce the discharge of pollutants to the "maximum extent practicable" to protect water quality.
Compliance with these enhanced EPA policies will lead to an inevitable increase in the development and use of
sediment control measures and other storm water treatment Best Management Practices (BMPs).
During the past few years a growing number of sediment control and storm water treatment devices have entered the
market. Unlike products or techniques designed only to limit or control erosion, these devices are intended to help
filter, capture and contain sediment transport (the by-product of erosion) and other pollutants that are generated and
transported during and after construction related activities. As with many emerging technologies, confusion may
develop as appropriate applications for specific products or families of products are not yet clearly developed and/or
sufficiently defined. This may result in end-users lacking clear direction on the proper selection and/or use of these
devices for specific applications.
This paper will propose a comprehensive and logical system to organize into classifications the growing range of
BMPs and techniques for specific prescribed functions or applications while integrating these applications into the
pre-construction, construction and post-construction phases of land disturbing, site development activities. This
classification system is intended to assist planners, contractors, designers, and regulatory agencies so that they may
have a better understanding of BMP selection based on application needs for protecting the environment from the
negative impacts of construction and post-construction storm water runoff. It is hoped that these proposed
classifications combined with increased field experience will evolve into practical and cost-effective methods of BMP
selection for an increasingly diverse array of storm water treatment measures and applications.
-------
Background
With the application deadline for National Pollutant Discharge Elimination System (NPDES) Phase II Storm Water
Permit coverage rapidly approaching storm water professionals, contractors, and end-users will need a systematic and
logical method for establishing techniques, management tools and classifications of Best Management Practices
(BMPs) to be integrated into the construction phase for storm water management. The new requirements of Phase II
lower the threshold for permit coverage for construction activities from 5 acres to 1 acre. In addition, regulations
affecting municipalities and public entities with Municipal Separate Storm Sewer Systems (MS4s), within urbanized
areas may also result in additional local construction requirements.
Regulators have two primary concerns that will underlie storm water requirements in the site plan approval processes.
These are the control of water quantity and quality both during and after the construction phase. Water quantity
outputs from sites will generally be limited to pre-development levels. Water quality issues will focus on the
reduction of contaminants from the runoff prior to its discharge from the site. Sediment has been recognized by EPA
and others as the most prevalent constituent of concern for US receiving waters. (Northcutt 1992 and Theisen 1991).
It will be the focus of most of the BMPs discussed in this paper. Other problematic constituents include nutrients,
metals, hydrocarbons and other organic compounds, bacteria, and others, and each site must be analyzed to determine
specific application needs. Understanding what types of structural BMPs are available and how they interact with one
another will help provide guidance in selecting the right mix for a specific site.
A major consideration to be determined is how maintenance will be assured and performed over the long run. Thus,
planners need to think of BMP selection as a revolving process of Installation, Inspection, Maintenance and
Enforcement (I2ME). While this paper focuses on the selection aspects, decision maker need to consider the latter
three components to insure quality-based selections of appropriate BMPs. Many techniques and technologies may
involve lower upfront costs, but maintenance costs over time must be factored into the equation.
In order to ensure that the maximum benefit is achieved planners will need to evaluate various BMPs in the pre-
construction, construction, and post-construction phases to ensure their plans are approved in a timely and cost-
effective manner.
Phases of Construction
Pre-Construction
The pre-construction phase will require a careful analysis of the specific site. The first step will be to gain a clear
understanding of what storm water controls are required by state regulations, local ordinances and site plan approval
processes. Nearly all will require controls during the construction phase to control sediment and to limit runoff from
the site in order to ensure minimum impacts on downstream receiving waters. The primary construction concern will
be sediment control and a wide range of both temporary and permanent BMPs will be needed. Each application must
be examined to determine site specific needs for laying out the sequence of selecting both temporary and permanent
BMP's. This sequence is commonly referred to as the "treatment train" and a clear understanding of all available
options is critical for a successful site plan.
According to EPA's Preliminary Data Summary of Urban Storm Water Best Management Practices an urban storm
water BMP is a "technique, measure or structural control that is used for a given set of conditions to manage the
quantity and improve the quality of storm water runoff in the most cost-effective manner." Many people only have a
vague understanding of the range of BMPs available, and with ongoing research, new BMPs are constantly emerging.
In fact, the term 'Best Management Practices' would be more accurately phrased as better Management Practices'
because what is 'best' varies with each situation
-------
In devising an effective organization of BMPs to assist planner and end users in the selection process several factors
must be considered. First, the proposed land use of a project must be determined. These possible uses include
industrial, commercial, residential, and streets and highways. For each of these various uses the specific site
application needs must be determined. Consideration should be given to whether the project is new or re-
development. A detail review of receiving water concerns along with an analysis of the potential pollutants of
concern that might be generated on the site and that could have a negative impact also needs to be completed prior to
BMP selection.
Once a review of the land use and receiving water concerns is completed then a review of the appropriate BMP
options can be evaluated. The wide range of BMP options can be organized into several classifications by
determining what the BMP can accomplish. Many are designed to control erosion and contain sediment transport.
This is particularly important in the active construction phase where site stabilization has not yet occurred. Other
BMPs deal with controlling the quantity of run-off that will occur as a result of both construction activities and post-
construction changes in flow that will occur as a result of increased imperviousness on the completed site. Again, this
will be a factor of the intended land use. Finally, many BMPs are utilized for treatment of run-off to reduce pollutants
that are generated during the construction and post-construction phases.
Many quality and quantity issues can be resolved through efficient site designs that incorporate practices that prevent
the transport of water and pollutants from increasing as a result of development. These preventive measures can
greatly reduce the need for reactive designs and technologies that are needed to contain water and remove pollutants
of concern. It is, however, beyond the scope of this paper to analyze Better Site Designs. Instead the focus will be on
the organization of structural BMPs and related Storm Water Treatment Devices (SWTDs). SWTDs are structural or
non-structural BMPs that positively impact Storm Water quality before, during or after construction or construction
related, land-disturbing activities. SWTDs may be temporary or permanent depending upon their desired application
or function.
Structural BMPs can be divided into three primary types. These include Vegetative Techniques and Open Space
Designs, Designed Structures, and Manufactured Technologies. The following chart lays out a proposed organization
of BMPs based on type and function.
Classification of Structural BMPs
Vegetative Techniques and Open Space Designs
• Constructed Wetlands
• Bio-retention Systems
• Swales
• Filter Strips
• Rain Gardens
• Green Roofs
Structural Designs
• Porous Pavement
• Below Surface Chamber Systems
• Infiltration Basins/Trenches
• Drywells
• Detention Basins
• Oversized Pipes
• Retention Ponds (Wet Ponds)
• Design-Sand Filters
-------
Manufactured or "Proprietary" Devices
• Hydrodynamic Separator Systems
• Filtration Systems
• In-Line Filtration Systems
• Catch Basin Inserts - Long Term/Short Term
• Exterior Treatments
• Storm Water Underground Storage Tanks
• Fabricated Underground Piping Systems
A broad overview of various BMP types is provided below in the post-construction phase section to help clarify the
assessment and selection process for meeting construction and post-construction requirements.
Active Construction
Sediment-Containment Systems
The role of sediment control systems is to create conditions for sedimentation, allowing for the settlement of soil
particles that are held in suspension. When soil-particle transport mechanisms flow at slow rates, particles may settle
out of suspension. How deposition occurs may depend upon several parameters.
Sediment-control systems are generally hydraulic controls that function by modifying the storm-runoff hydrograph
and slowing water velocities. This allows for the deposition of suspended particles by gravity. Some of the more
common names for these structures are sediment basins, sediment ponds and sediment traps. When designed
correctly, sediment-containment systems should provide containment storage volume sufficient to handle incoming
waters, create uniform flow zones within the containment storage volume for deposition of suspended particles and
discharge water at a controlled rate.
When all runoff waters are captured, efficiency of the containment system is near 100%. However, the feasibility of
retaining all runoff waters from a construction site is usually impossible since large containment areas and volumes
are required. In addition, evaporation and infiltration might not be sufficient to drain the system before the next storm
event occurs, which may cause flooding problems. Finally, retained waters may hamper maintenance of the system
since removal of captured sediments becomes more complicated with the presence of water.
Due to the above concerns, rather than attempting to retain all runoff waters, a containment system should provide
sufficient volume for capturing suspended particles while allowing discharge to occur. This provides the advantage of
detaining incoming runoff to control the discharge of suspended particles while not requiring large areas to store
runoff waters. Flooding problems from sequential storm events are reduced since contained waters will usually be
drained from the system between events. Finally, frequent maintenance is facilitated because the sediments do not
remain saturated with water.
If detention of runoff from construction sites is to be effective in removing suspended particles, contained waters must
remain long enough for deposition of suspended particles within the system. Since outflow from the system will
occur, 100% reduction of all incoming suspended particles will not be possible. However, high efficiencies can occur
for sediment-containment systems developed for design-sized particles. (Fifield, 1995 and 1996.)
Sediment-containment systems may be characterized using the following assumptions. Goldman (1986) defined a
structure that treats runoff from 2.0 ha (5.0 ac) or less as a "sediment trap." When the contributing area to the
structure exceeds 2.0 ha, then a "sediment basin" is used. Both structures are "sediment-containment systems" that
function on the principles discussed previously.
EPA has suggested that the design of any sediment-containment system be based upon capturing the volume of runoff
resulting from a 2-year, 24-hour storm event (US EPA 1992 and 1998). The problem with considering only the
-------
volume from a contributing area is that it does not take into account the size of the particles generated by upstream
eroding soils. Table 1 provides suggested definitions for sediment-containment systems.
Table -1 - Defining Sediment-Containment Systems Using Particle Diameters (Fifield, 2001)
Sediment-Containment System Type
Type- 1 Sediment-Containment System
Type- 2 Sediment-Containment System
Type- 3 Sediment-Containment System
Design Particle Size
Design- Size Particle <_0. 045mm
0.045 mm < Design-Size Particle <_0.14 mm
Design-Size Particle > 0. 14 mm
Type- 1 Sediment-Containment Systems
A Type-1 sediment-containment system will require development of a structure to capture the maximum possible
number of medium silt and smaller suspended particles. Since particles of this size have low settling velocities, large
storage volumes, long flow-path lengths, and controlled discharges are required. Type-1 systems are designed to have
the highest possible net efficiency and are best represented by the traditional sediment basin and trap.
Type-2 Sediment-Containment Systems
The Type-2 sediment-containment system will capture suspended particles having higher settling velocities than
particles requiring Type-1 structures. Consequently, smaller storage volumes and shorter flow-path lengths can be
used. As with a Type-1 structure, these sediment control systems will also have controlled discharges. While their net
effectiveness for the entrapment of all suspended solids may be low, Type-2 systems will still have a high apparent
effectiveness.
Type-3 Sediment-Containment Systems
The least effective methods to control suspended particles in runoff waters are represented by Type-3 sediment-
containment systems. These are not necessarily design structures, but are often temporary BMPs found on
construction sites. Examples include straw or hay bales and silt-fence barriers, inlet control structures, and drainage
ditch check structures.
Whenever significant runoff occurs, all Type-3 systems have very low net and apparent effectiveness to control
suspended particles. However, when runoff quantity is low, the Type-3 sediment control systems can be effective in
reducing suspended particles as long as they are continuously maintained.
The Effectiveness and use of Sediment-Containment Systems
Documentation on the effectiveness of containment systems for trapping suspended solids is limited, and there are
conflicting opinions on their actual effectiveness. However, if properly designed, constructed, inspected, and
maintained, containment systems are effective in trapping some sediment.
This discussion will focus on selected, man-made non-structural Type-3 sediment-containment systems that act as
barriers or filters. Since their effectiveness is minimal for large runoff events, they do not require the detailed designs
needed for Type-1 and Type-2 containment systems. These devices must be carefully installed and in conjunction
with Type-1 and Type-2 systems to minimize downstream problems since their usefulness is generally limited to low
-------
volume flows from smaller storm events. As such, these systems are typically only used and installed during the pre-
and active-construction phases of a project.
A barrier is any structure that obstructs or prevents the passage of water. If runoff cannot pass through a barrier, then
water will either be contained or flow over the structure. Consequently, small sediment barriers may function as a
Type-3 system or as a method to reduce flow velocity. Commonly used man-made barrier devices include silt fences,
continuous geotextile-wrapped berms, turbidity barriers, and geosynthetic silt dikes. An example of a silt fence is
shown in Figure 1.
Appropriate places to use sediment control barriers include:
• Along sections of a site perimeter
• Below disturbed areas subject to sheet and rill erosion
• Below the toe of exposed and erodible slopes
• Along the toe of stream and channel banks
• Low flow swales and ditches
• Around area drains or inlets located in a sump
• Turbidity barriers are used in low flow streams, tidal areas or lakes
Inappropriate places to use sediment control barriers include:
• Parallel to a contour when installed on a hillside
• In channels where concentrated flows occur, unless properly reinforced
• Upstream or downstream of culverts where concentrated flows occur
• In front of or around inlets where concentrated flows occur and sump conditions do not exist
• In continuously flowing streams or ephemeral channels
Other Type-3 devices designed to provide filtration include geotextile catch basin inserts, geosynthetic drainage and
curb inlet filters, geotextile tubes, and geotextile filter bags. These materials allow water to flow through them while
filtering or capturing sediment. Selection of the correct geotextile or fiber consistency will reduce the possibility of
blinding or clogging of the device with excessive sediment. An example of a Type I geotextile catch basin insert is
shown in Figure 2.
Appropriate places to use geosynthetic filters would be in front of or around gutters and drain inlets where sump
conditions exist and areas of de-watering of detention/retention ponds or dredging of construction and/or industrial
spoils.
Inappropriate places to use geosynthetic filters would include in front of or around inlets where concentrated flows
occur and sump conditions do not exist in channels where concentrated flows occur or in continuously flowing
streams or ephemeral channels.
Man-made geosynthetic Type-3 barriers and filters have numerous advantages over traditional sediment control
practices derived from natural materials. They are normally easier to transport, install and maintain versus straw and
hay bales or soil and rock structures. Manufacturing and fabrication consistencies enable performance of geosynthetic
devices to be more predictable and generally superior to natural materials. In many cases these devices may be
washed and reused which makes their usage highly cost effective versus using traditional practices or nothing at all.
Thus the acceptance and usage of geosynthetic sediment- and erosion-control devices has increased dramatically over
the past few years (Theisen, 1991, Theisen and Hunt, 2001).
-------
Figure 1 - Example of Silt Fence Containing Sediment - Geotex® by SI Geosolutions
Figure 2 - Example of Type 1 Geotextile Catch Basin Insert - Siltsack by ACF Environmental
-------
Post-Construction
Structural BMP's are techniques that can be used to address flow quantity control and pollutant removal in wet
weather runoff. These BMPs can include site-specific engineered designs as well as proprietary systems. The
challenge with any attempt to organize or classify BMPs by type or function is that many fit into multiple categories.
However, in the interest of clarity structural BMPs can be grouped into several subcategories by function that includes
the following.
• Infiltration systems
• Detention systems
• Retention systems
• Vegetated systems
• Filtration systems
• Hydrodynamic separation systems
Infiltration Systems
Infiltration systems are designed primarily to reduce the quantity of storm water runoff from a particular site.
Increasing urbanization and percentage of impervious surfaces has resulted in substantial increases of surface runoff,
causing serious degradation of urban streams and the corresponding negative impacts on aquatic health BMPs for
Phase II The use of infiltration techniques can reduce the amount of surface flow and direct the water back into the
ground. Advantages of infiltration techniques include the recharging of groundwater supplies and the removal of
certain pollutants such as sediments. Care must be exercised, however, in determining whether infiltration is best for a
specific application, especially when groundwater is the source of drinking water in the area. Infiltration can result in
groundwater contamination since soils that allow good infiltration also allow rapid migration of certain pollutants. In
these situations, infiltration should not be used without effective pretreatment. Conversely, poorly permeable soils can
prevent an infiltration system from functioning.
Infiltration techniques can be divided into several different classifications depending on site needs. Regardless of the
classification a careful understanding of the soil type is necessary since certain soils, such as clays, are poor
infiltration types. If the soil type is appropriate for infiltration then the next step in the evaluation is determining
which method is most appropriate. A site with minimal land space would be a likely candidate for porous pavement,
and sub-surface chamber systems that can store water below impervious surfaces and allow for slow infiltration after
the end of a wet-weather event. Conversely, sites with sufficient space should utilize infiltration basins, vegetative
practices, constructed wetlands and open space designs.
Detention Systems
These BMPs are designed to temporarily hold storm water runoff for gradual release into receiving waters. Detention
systems are used primarily to reduce peak discharges to prevent flooding, stream bank erosion, and channel
alterations. Straight up Detention systems are generally not very effective for removing pollutants unless combined
with other BMPs. Many detention systems incorporate characteristics normally utilized with retention ponds, such as
permanent pools, to prevent subsequent scouring. Examples of detention systems include detention basins,
underground tanks, oversized pipes, and fabricated underground high-density polyethylene piping systems such as
Storm Compressor™.
Retention Systems
Retention systems are intended to capture and hold runoff from entering receiving waters. Because retention systems
are designed for permanent containment of storm water, they can also be a good infiltration and or filtration BMP
-------
with the right conditions, thus providing both water-quantity and water-quality control. Retention systems can be in a
variety of forms such as green roofs, but most retention systems are in the form of ponds or basins, (also commonly
referred to as wet or detention basins) and when certain types of aquatic vegetation or aerators are added, the systems
can actually provide further water treatment (see figure 3 below). As with all BMPs, regular maintenance is essential
to maintain a healthy retention pond. Clay siltation can result in a substantial loss of infiltration, resulting in a sharp
increase in overflow from the basin during wet-weather events. Without maintenance, retention ponds will eventually
fill in and become ineffective. In addition, certain pollutants can become concentrated in the area, potentially
requiring remediation.
Most storm water collection ponds are in fact combinations of retention and detention applications. While these
ponds are designed to hold most flows they are usually equipped with some sort of overflow system to prevent
flooding over their banks. These overflow systems are either reset in the middle or end of the ponds or a spillway of
rip-rap, other coarse materials or vegetated turf reinforcement mats. When the runoff into the pond is from an
impervious area with high vehicle traffic, post-treatment devices in the riser can provide initial management of
floating oils and other toxins prior to discharge into the receiving waters.
Figure 3 - Example of Wet Pond (courtesy of Hydro Compliance Management, Inc.)
Vegetative Systems
Constructed Wetland Systems
Constructed wetlands are a very effective BMP for both pollutant removal and runoff storage (see figure 4 below).
When properly designed, they incorporate the processes of sediment removal, microbial decomposition, and aquatic
plant uptake. Sites for constructed wetlands must be carefully selected to ensure that sufficient waters are available in
dry weather to sustain the wetlands. Areas with shallow groundwater levels are ideal. Heavy sediment loads can
quickly degrade a constructed wetland. Pretreatment of sediment flows must be considered if this is the case.
-------
Generally, natural wetlands should be preserved and not used as a BMP because changing hydrology can significantly
degrade a natural wetland.
Other wetland BMPs include wetland basins and channels. These BMPs do not necessarily require open waters and
can instead be in the form of wetland meadows that have surface water only for short periods of time after
precipitation events.
Figure 4 - Example of Constructed Wetland (courtesy of Ingham County, Michigan, Drainage Commission)
Bio-retention and other Vegetated Systems
Bio-retention and vegetated systems, such as buffers and swales, are variations of infiltration and filtration systems.
The media in these systems are actually natural vegetation and soil beds that allow ponding and gradual infiltration.
The vegetation and underlying soils can filter a variety of pollutants from runoff. In addition, these systems can be
used to reduce the quantity of flow. This category of BMP includes large bio-retention systems, swales, rain gardens,
grass filter strips, and even green roofs. The use of these "natural" systems in site development can significantly cut
down on surface runoff and reduce the need for other more costly structural BMPs. An example is shown in Figure 5
below.
-------
Figure 5 - Example of Bio-retention Swale (courtesy of Hydro Compliance Management, Inc.)
Filtration Systems
Filtration systems are BMPs that use media to remove participates from runoff. They are typically used when
circumstances limit the use of other types of BMPs, such as where space is limited-particularly in a highly urbanized
setting-or when it is necessary to capture particular industrial or commercial pollutants such as hydrocarbons or
metals. In these circumstances, other BMPs might be cost-prohibitive or not as effective. Filtration devices can also
work well as pretreatment systems for other types of BMPs. For example, infiltration systems that move water
directly to ground aquifers might require pre-treatment for certain contaminants to maintain effective well-head
protection of drinking-water supplies.
Filtration systems can be either designed into a site plan, such as sand filter systems, or be manufactured technologies
such as catch-basin inserts or in-pipe systems. An example of a filtration device is shown in Figure 6. Many different
filtration media are available, such as sand, peat, absorbents, and activated carbon. The choice depends on the
particular application.
When considering filtration systems, planners need to consider flow rates. As a result of the volume of water being
moved in a wet-weather flow, filters generally need to focus on treating at least the first quarter inch of runoff and
allow bypass for high-flow events. Filters should incorporate pre-settling sediment chambers to remove sediments
that can clog the filters and reduce flow rates and effectiveness. An effective filtration system should be able to
demonstrate removal efficiencies for specific contaminants. Again, as with all BMPs, regular maintenance is
essential.
Proprietary filtration devices are catch-basin inserts or in-pipe designs that remove various pollutants. Effective
designs should use non-leaching media, incorporate pre-filtration sediment removal chambers or other measures to
reduce plugging, and be accessible for regular maintenance. In addition, filtration devices need to be designed with
overflow bypasses to prevent flooding caused by high flow rates or plugging of the filters. A properly designed
-------
filtration system can be a useful device for urban hot-spot applications where a particular pollutant is being targeted.
It also can be cost-effective where land use does not allow other economical BMP options. This is particularly true
with existing sites in urban settings. Proprietary systems can be effective pre-treatment or post-treatment devices for
infiltration systems and other BMPs.
1/16-
STAINLESS FRAMING
SEDIMENT
CHAMBER
BOTTOM
DRAIN
TM
Figure 6 - Example of Catch-Basin Filtration System — Hydro-Kleen Storm Water Filtration System
Hydrodynamic Separator Systems
These systems remove sediment, debris, and surface oils and grease through various hydrodynamic designs. Effective
separator systems trap and separate pollutants to prevent them from being reintroduced into runoff, which can result
from "scouring" or other actions prompted by the powerful energies created from heavy volumes of storm water
runoff. Effective systems have protective zones for pollutant storage to prevent re-suspension or washout of
contaminants and stabilize the flow regime to minimize turbulence. Systems with stabilized rotary flow regimes tend
to have smaller footprints than conventional gravity separators. An example is shown in Figure 7.
-------
Access Ports
Support Frame
Dip Plate -
Tangential Inlet Pipe
Benching Skirt
Floatables Lid
Outlet Pipe
Center Shaft and Cone
Concrete Manhole
Sediment Storage Facility
Figure 7 - Example of Hydrodynamic Separator System - Downstream Defender® - Hydro International, Inc.
Functions of Storm Water Treatment Devices
SWTDs may be "proactive" or "reactive" in their approach or application. Examples of proactive SWTDs include
erosion control practices, green roofs, vegetative filter strips, or rain barrels. Reactive techniques might employ
sediment control practices, in-line treatment devices, sedimentation ponds, and detention/retention systems.
Basic functions of SWTDs may be grouped into five major categories. These are Sediment Containment,
Filtration, Separation, Infiltration, and Underground Detention. Again, it is beyond the scope of this paper to describe
and classify all the BMPs that may be used to fulfill these functions. Various manufactured SWTDs may be grouped
by primary function as shown below.
Basic Functions of Storm Water Treatment Devices
• Sediment Containment
• Filtration
• Separation
• Infiltration
• Underground Detention
It is beyond the scope of this paper to describe and classify any and all BMPs or SWTDs that may be used to fulfill
these functions. This paper, however, does describe various man-made SWTDs may be grouped by primary function
as shown below.
-------
Sediment Containment
• Silt Fences (SF)
• Continuous Berms (CB)
• Wattles (W)
• Drain Inlet Barriers (DIB)
• Channel Silt Dikes (CSD)
• Turbidity Barriers (TB)
• Geotextile Filter Bags (GFB)
• Geotextile Tubes (GTT)
Filtration
• Catch Basin Inserts (CBI)
• Type 1 -Geotextile Filtration Systems (GFS)
• Type II - Multi-Chamber Permanent Structures (MPS)
• Curb Inlet Filters (GIF)
• Type 1 - Exterior - Geotextile Filtration Systems (GFS)
• Type II - Interior - Multi-Chamber Interior Filtration Systems (MIF)
Separation
• Hydrodynamic Separation Devices (HSD)
Infiltration
• Infiltration Chamber Systems (ICS)
Detention
• Underground Piping Systems (UPS)
Once the function required of a SWTD has been determined, it is then time to consider when and where it should be
employed. These two considerations are as important as the selection of the correct SWTD to be used. Failure to
properly install a SWTD in the correct location or sequence of a land-disturbing activity may result in failure or
compromised performance.
Once the application or function and appropriate construction phase of the required storm water treatments have been
determined, these parameters may be coupled to facilitate selection of the most appropriate SWTD. Table 2 presents a
matrix that combines function with construction phases for identifying potential SWTDs for selection consideration.
Table 2 - Function and Typical Construction Phase(s) for Application of Manufactured Storm Water Treatment
Devices
Function
Sediment Containment
Filtration
Separation
Infiltration
Detention
Construction Phase
Pre-Construction
SF, CB, TB
Active Construction
SF, CB, CBI, DIB, CIF,
CSD, TB, GFB, GTT
CBI, CIF, GFB, GTT
Post-Construction
CBI, CIF, HSD
GFB, GTT, HSD
HSD
ICS
UPS
-------
Finally, where to use a SWTD must be considered. Again, it is beyond the scope of this paper to present specific site
locations for the vast potential variances of SWTD applications. Good discussions for placement of several of these
materials during active construction may be found in publications by Fifield as well as in EPA publications. Table 3
below presents a matrix coupling site location with the various construction phases. Combining Tables 2 and 3 may
help end users to make informed decisions when considering SWTDs for various functions, construction phases and
site locations.
Table 3 - Site Location and Typical Phase(s) of Construction for Application of Manufactured Storm Water Treatment
Devices
Site Location
Perimeter
Catch Basin Inlet,
Curb Inlet
Channel
Slopes
Waterway
Sediment Basin/Trap
Below Impervious
Surfaces
Construction Phase
Pre-Construction
SF, CB
SF, CB, W
TB
Active Construction
SF, CB
CBI - Type 1 & II, DIB, CIF,
HSD
CSD
SF, CB, W
GTT
GFB, GTT
Post-Construction
CBI - Type II, CIF,
HSD
GTT
ICS
UPS
Conclusion
In order to insure that regulators, planners, engineers and contractors have a clear picture of what techniques and
measures can be utilized in the various construction phases for proper BMP management, a solid understanding of the
options is essential. By classifying the various sediment controls and post-construction BMPs into proper
applications, storm water professionals are far more likely to develop efficient yet cost-effective storm water plans for
specific projects. The result will be cleaner water and a more satisfied general public. A thorough understanding of
the Installation, Inspection, Maintenance, and Enforcement requirements will also result in a more comprehensive and
realistic cost analysis of the project.
Literature Cited
Goldman, Steven J., Jackson, K, and Bursztynsky T.A. 1986. Erosion and Sediment Control Handbook. McGraw-Hill
Book Company, New York, NY.
Fifield, Jerald S. 2001. Designing for Effective Sediment and Erosion Control on Construction Sites. Forester Press.
Fifield, Jerald S. 2002. Field Manual on Sediment and Erosion Control - Best Management Practices for Contractors
and Inspectors. Forester Press.
Northcutt, Ben. 1992. The Erosion Control Industry - A Look at It's Past, Present and Future. Proceedings: High
Altitude Revegetation Workshop No. 10, Ft. Collins, CO USA, 1992.
Theisen, Marc S., 1991. "The Role of Geosynthetics in Erosion and Sediment Control: An Overview", Proceedings of
the 5th GWSeminar, Philadelphia, PA, pp. 188-203.
-------
Theisen, Marc S. and Hunt, Daniel L., 2001. "An Overview of Current Rolled Erosion Control Product (RECP)
Design Practices - Science with a Bit of Art", Proceedings of the 15th GRI Conference, Houston, TX, pp. 130-154.
US Environmental Protection Agency, 1992. "Final NPDES General Permits for Storm Water Discharge from
Construction Activities." Federal Register, Part II, Notice. US Government Printing Office, Washington, D.C.
September 9.
US Environmental Protection Agency, 1998. "Reissuance of NPDES General Permits for Storm Water Discharges
from Construction Activities." Federal Register, Part II, Notice. US Government Printing Office, Washington, D.C.
February 17.
US Environmental Protection Agency, 1999. Storm Water Technology Fact Sheet: "Hydrodynamic Separators",
EPA-832-F-99-017.
US Environmental Protection Agency, 2000. " Storm Water Phase IIFinal Rule, An Overview ", EPA 833-F-00-001.
US Environmental Protection Agency, 2000. " Storm Water Phase II Final Rule, Small MS4 Storm Water Program
Overview", EPA 833-F-00-002.
-------
OVERCOMING CHALLENGES IN ESTABLISHING A REGIONAL PUBLIC
EDUCATION AND OUTREACH PARTNERSHIP
Michael A. Worlton, PE and J. Ryan Christensen
RBF Consulting
Phoenix, Arizona
ABSTRACT
Over the past twelve years, many Phase I Municipal Separate Storm Sewer System (MS4) operators have
established programs for public outreach. Often these programs have focused on specific municipalities
using varying approaches. With the implementation of the National Pollutant Discharge Elimination
System (NPDES) Phase II rules by USEPA, smaller municipalities are faced with the challenge of creating
effective public outreach programs. Although Phase II rules provide more comprehensive guidance, Phase
n municipalities typically have fewer resources at their disposal.
Phase II rules emphasize the importance of forming partnerships for public outreach and education. These
partnerships can provide the benefits of pooled resources, reduced costs, and a more consistent and effective
outreach program. While there are clear benefits of forming regional public outreach and education
partnerships, many challenges must be overcome to establish an effective and equitable program.
This paper discusses key issue areas that were addressed in the successful establishment of a regional public
outreach partnership involving ten municipalities in the metropolitan Phoenix, Arizona area. These include
issues related to membership, local perceptions, funding, the decision-making process, and leadership.
Introduction
The USEPA's National Pollutant Discharge Elimination System (NPDES) Stormwater Permit Program was
introduced to reduce the number of impaired surface water bodies within the United States. When one
considers the requirements of the Phase I and Phase n programs, one may simply envision BMPs being put
into place to minimize polluted Stormwater runoff flowing into our nation's treasured streams and lakes.
These water bodies not only serve as a valuable natural resource, but also may enhance quality of life.
When applied to the desert southwest, this vision of the NPDES program is not so easy to grasp. First, very
little rainfall is received in desert areas. Secondly, in the desert, the term river is more commonly associated
with a dry riverbed than a flowing body of water. These realities play a significant role in influencing
public opinion about Stormwater pollution.
With these realities in mind, regulated MS4s throughout the Phoenix Metropolitan area recently came
together to form a regional public outreach organization. This paper describes how Phase I and Phase II
municipalities worked together to change the way Stormwater quality concerns are perceived in an area
where some view these concerns on the same level as UFO sightings. This paper discusses the methods
used, challenges encountered, and lessons learned in forming a Stormwater public outreach group in the
Phoenix Metropolitan area.
518
-------
History
NPDES permit requirements implemented in 1990 brought new connotations to the word "stormwater" in
Arizona, as larger municipalities were faced with the challenge of regulating stormwater quality. When
speaking of stormwater, in an arid climate that only receives an average rainfall of about six-inches per year
(http://ag.arizona.edu/oals/watershed/highlands/climate.html), pollution is not the first issue that comes to
mind. Nonetheless, Arizona's Phase I municipalities worked diligently to successfully implement effective
stormwater programs. Representatives from these municipalities often shared ideas and information, but
their respective NPDES permit applications and programs varied from municipality to municipality. For
example, permitted municipalities each developed unique programs to address public outreach, inspections,
enforcement, representative rainstorms, and other program requirements.
This individual approach to NPDES issues in Arizona would change in 1997, when Pima County and the
cities of Tempe, Tucson, Mesa, and Phoenix petitioned against numeric limitations on water quality
standards (Case Name: Defenders of Wildlife V. Browner; Case Number: 98-71080; Date Filed: 09/15/99).
The submission of the petition, and its subsequent defense in a lawsuit brought by Defenders of Wildlife,
helped these municipalities form strong working relationships and unify their visions. In the late 1990's, the
cities of Glendale and Scottsdale were also issued NPDES permits, and began to interact with
representatives from other Phase I communities in Arizona. By this time, Phase I communities had
organized themselves to form a fairly cohesive unit, with a unified voice.
In early 2000, the State of Arizona's Department of Environmental Quality (ADEQ) began working toward
NPDES Permitting program approval
(http://www.adeq.state.az.us/environ/water/permits/azpdes.htmWquest). This event sparked the interest of
other municipalities, many of which would be designated as regulated MS4s under the Phase n NPDES
program. Many of the larger municipalities throughout the state worked together as stakeholders in
ADEQ's quest for NPDES. This process resulted in building a working relationship among the Phase I
communities and several of the larger Phase n communities.
These events, which helped form the foundation for the NPDES Program in Arizona, played an important
role in bringing municipalities together as partners. Relationships were developed, ideas were shared, and
assistance was offered. This atmosphere provided a good foundation for the creation of a regional public
outreach group. Several communities realized the benefits of working together on a regional level, and they
began exploring the idea of a regional public outreach program.
Forming a Regional Public Outreach Group
The first recommendation in the EPA's Phase n Rule for developing public education and outreach
programs is to form state or regional partnerships (EPA 2000). The EPA Fact Sheet on Public Education
and Outreach (Fact Sheet 2.3, Public Education and Outreach Minimum Control Measure, January 2000,
EPA ) suggests that regional programs are more cost-effective since they utilize shared resources and
existing education and outreach materials. As will be discussed later, there are additional benefits to
regional public outreach groups. These benefits stem from the collective creativity and the variety of
experience and interests shared by the group.
The concept of regulated communities in Arizona forming partnerships was not new, but, when public
works planners from the City of Scottsdale met with the City of Phoenix's chief water quality inspector to
talk about public education and outreach, a new enthusiasm was generated. This enthusiasm was translated
519
-------
into action, including the coordination of a meeting among several regional MS4s to discuss developing a
cooperative public outreach and education effort.
Identifying Membership
The first challenge faced in organizing a regional public outreach group was identifying membership.
Before the first meeting could be convened, a list of potential group members had to be created. It made the
most sense to select municipalities affected by the regulation and located within a common geographical
region and influenced by the same television and radio stations. It was also important to consider the
communities that intermingle within the region. For example, a person who lives in Mesa may work in
Phoenix, and shop in Scottsdale. Someone from Peoria may work in Glendale and watch Cactus League
baseball games within the City of Surprise. Maricopa County was generally identified as the region of
focus for the public outreach group. The original list of potential members included all known Phase I and
Phase n municipalities in the selected region, Maricopa County Flood Control District, Maricopa
Association of Governments (MAG), ADEQ, and various municipalities that were potential Phase n
candidates. Key contacts for each municipal stormwater program were invited to attend. Once the potential
members were identified, it was important that everyone had the opportunity to participate. Meeting
announcements were distributed via email, and RSVPs were requested. When a municipality did not
respond, a follow up call was made.
Maintaining Focus
The idea of this first meeting was to identify the level of interest for participation in the group, provide
background information about the Public Education and Outreach requirements of the NPDES stormwater
program, and discuss the viability of implementation. The inaugural meeting was held in June of 2001.
Twelve municipalities participated. This meeting marked the commencement of a public outreach
organization for the Phoenix Metropolitan area, now known as STormwater Outreach for Regional
Municipalities (STORM). Enthusiasm at the meeting was very encouraging, and many municipalities
showed an interest in participating.
Although there was consensus support at the meeting for forming a regional organization, there was no
decision regarding where to go from there or immediate follow-up. Consequently, Phase I municipalities
forged ahead with their permit reapplications independently, while Phase U communities attended NPDES -
related seminars, began to budget, and contracted with consultants to prepare for completing their individual
permit applications. Other priorities and lack of follow through from this initial meeting caused a loss of
focus. This loss of focus was the second challenge to the establishment of the regional public outreach
group. It would be almost another year before a second meeting was held.
520
-------
Establishing Leadership
One key to moving forward with the formation of STORM was establishing leadership. At first, several
people seemed willing to fill the leadership role. As time passed, however, it became apparent that someone
would need to assert him or herself as the leader. This person needed to take the initiative and assume the
role of coordinating with the selected municipalities and planning meetings. While the majority of the
representatives from the municipalities were willing to participate in the organization, they did not have
extra time needed to perform leadership duties such as setting a meeting time, arranging for meeting space,
inviting members to attend, and establishing an agenda. A consultant who has represented several Phase I
and Phase n communities in the region assumed this role. This leader ensured that the organization was
established, interest did not wane, and that the group would move forward.
In May of 2002, a second meeting was held to reinitiate the regional public outreach effort. The goal of this
meeting was to reconvene the group and establish a plan for the future. Some of the participants had
changed, so this meeting brought new faces and new questions. The meeting was very well attended and
the results were encouraging. During this meeting, it became clear that this was the first exposure to
NPDES program requirements for some municipalities in attendance. It was necessary, therefore, to
provide background information about the requirements of the Phase n Program and the associated
responsibilities of the affected municipalities. The meeting also served as forum to identify common goals
and outline advantages to the group members. It quickly became apparent that some of the municipalities
desired to have a high level of participation, while others wanted to become involved only after the group
had been established.
Both the Phase I and Phase n communities shared a desire to make this regional public outreach effort a
success. The Phase I municipalities saw an immediate need to begin a regional partnership so they could
integrate it into their existing programs, and the Phase n municipalities wanted to capitalize on the
experience and resources of the Phase I municipalities. Many were interested in the group's success
because there was a feeling that this group could truly have a positive impact on their community, and that
those who participated in organizing this group would be part of something great. Another perceived
benefit of the group was that it could reduce the public outreach burden on the individual municipalities.
The Phase I municipalities with years of public outreach experience played a significant role in guiding the
group.
Making Decisions
As subsequent meetings were held, more issues began to surface. One of the first issues to be tackled was
determining the process by which the group would make decisions. Buy-in from the group as a whole was
important, but there was always some disagreement among members about what the best decision might be.
Consequently, the organization established a policy of majority rule and general consensus. This meant that
decisions were narrowed down to the point where a vote could be taken, followed by a poll to ensure that all
members could live with the results. This process was tested in the selection of a name and mission
statement for the group. The group decided to adopt "STormwater Outreach for Regional Municipalities,"
or STORM, as the name. The mission statement agreed to was "STORMpromotes regional stormwater
public education through outreach. "
521
-------
Another major issue addressed was financing the organization. This was a very difficult issue because it
involved city budgets, intergovernmental cooperation, financial management, and finding an equitable way
to distribute the projected costs of the program. Discussion among regional Phase n municipalities revealed
common concerns about acquiring the resources to pay for the new program. Their budgets for the entire
NPDES stormwater program ranged from $10,000 to $500,000. Most of the Phase I communities had
already established budgets for public education and outreach, but there was concern about how much could
be allocated to the group.
The City of Phoenix had already made a significant investment in outreach and educational materials. They
freely shared all of the information and materials they had developed with other group members. These
materials included a storm drain marker design, BMP pamphlets, and a comic book series detailing the
adventures of "Storm Drain Dan," a stormwater quality superhero. Phoenix also volunteered to send
electronic copies of their printed materials so that other municipalities could customize them by changing
the logos and contact information. While these materials came at no cost, another goal of STORM was to
enable member municipalities to capitalize on the buying power of the group, and to share the costs of
developing television and radio spots.
Because most budgets for the 2002-2003 Fiscal Year had already been established at the time the group got
started, STORM members had some time before the next budget cycle to consider the benefits of
participating on the group and determine their levels of commitment. Before the group could publish any
materials, they needed to identify funding mechanisms. The following funding ideas were considered.
> Base membership fees on distinct population categories; similar to what is done by the National
Association of Flood and Stormwater Management Agencies.
> Assess membership fees on a per capita basis, (i.e., $0.05 per person within the municipality).
> Establish in-kind contributions in lieu of membership fees.
> Assess a flat membership fee for all members of the group.
> Pay as you go. Develop public education and outreach materials that municipalities can buy
individually.
> Provide no funding. Use the group to share resources and ideas.
After much discussion, the group decided that the most equitable funding method was a fee-based approach,
set according to each municipality's population. Table 1 lists the first-year fees for the members of
STORM. These fees are subject to change based on the programs the group chooses to implement in the
future.
522
-------
Table 1. STORM Population Based Fee Structure
Population
0 - 25,000
25,001 - 50,000
50,001 -100,000
100,001 -250,000
Greater than 250,000
Fee
$1,000
$1,500
$2,000
$2,500
$5,000
Another issue involved dealing with perceptions by some of the local governments that stormwater
pollution prevention is insignificant and a low priority. These perceptions were shared by the public and
even some potential members of STORM. When City Managers and Councils do not consider stormwater
runoff a high priority, it is unlikely that sufficient funding will be dedicated to stormwater quality programs.
An independent effort was initiated by Maricopa Association of Governments (MAG), an established
regional planning organization, to educate and offer assistance to city managers. Another approach to
educating decision-makers was for group members to work individually with their municipality's
management. This presents an additional opportunity for the group to make an impact. The group
discussed these issues and provided recommendations that would assist members in approaching decision-
makers.
The group also addressed the issue of public perception. These perceptions will govern the types of
outreach activities that each of the municipalities conducts. Group members stressed that the stormwater
pollution prevention message had to be tailored to meet the needs of the area. While many areas of the
country can use storm drain markers with slogans such "No dumping... Drains to River," a more
appropriate slogan for the Phoenix area would be "No dumping... drains to dry river bed." Therefore, more
creative solutions must be presented, such as "Only Rain in the Drain" or "Storm Drains... No Dumping."
The general feeling was that the message had to strongly target pollution prevention and have stormwater
under tones. A regional group speaking to the public with a common voice and a consistent message has a
much better chance of educating the public than inconsistent messages from independent sources.
The municipalities also expressed concern about how the group would be controlled. Members have to be
committed to STORM either financially or through in-kind service in order to accomplish the organization's
mission. Decisions will be made as a group, but someone has to be responsible for following through. The
Flood Control District of Maricopa County (FCDMC) expressed a willingness to be the fiscal agent for the
group, and will handle the funding through letter agreements with the member municipalities.
In order to address these concerns, a subcommittee of STORM has researched several models for the
management of the funds and coordination of contracts. These models include several existing programs
administered by the City of Phoenix, FCDMC, and various non-profit organizations. Based on these models
a structure was established for the administration of STORM.
In establishing the organizational model for STORM, the members agreed that a board would be elected
which would be responsible for organizing and facilitating the meetings, developing the meeting agendas,
maintaining meeting minutes, and managing the group's money. The STORM board now consists of four
members, including a chairman, a vice chairman, a secretary, and a fiscal agent. The board members are
elected annually, and can serve for an unlimited number of terms.
523
-------
The contractual agreement among the organizations was also addressed. Two different methods for the
administering the contracts were put forth, a formal intergovernmental agreement, and an informal
agreement. The members of STORM decided that the informal agreement would be easier and more
effective to administer, since it would bypass the need for City Council approval. The informal agreement
will be administered through the fiscal agent (FCDMC) who will submit a letter each year to the group's
members assessing the fees due.
Lessons Learned
Many challenges have been faced in the establishment of STORM, and many lie ahead. In the process of
overcoming these challenges various lessons were learned that might assist others in developing a regional
education and public outreach program.
Understanding Needs
Since the needs of each municipality dictate the direction of the regional education and public outreach
group, it is important that these needs be identified. It was interesting to observe that the goal of some of
the municipalities was to utilize the efforts of STORM to totally fulfill the public outreach requirements of
their permits. Other municipalities only desired a minimal amount of participation, seeing the organization
as merely a purchasing entity that would allow them additional buying power. Respecting and
understanding these and other group needs lead to a balanced approach in establishing the objectives of the
group. Understanding the needs of the group members also helped the group to remain focused on the
issues that are most important.
Taking the Initiative and Sustaining the Effort
A lesson learned from the year-long lag in between the first and second meeting of STORM, was that
finding someone to take the initiative in assuming leadership of the group was critical to establishing the
organization. Additionally, if the effort is not sustained over time, little will be gained. Leaders and
members of the group must be committed to the effort. Success in sustaining the organizational effort for
STORM was realized through the following processes.
> Prepare and organize meeting details and agendas. It is important to meet in a central location and
have an agenda that catches the attention of potential members.
> Identify and Invite potential members. A key to getting such a strong showing of Phase n
municipalities, was getting the larger Phase I municipalities involved.
> Follow up on invitations personally. When a municipality neglected to RSVP, a personal follow up
call was made to extend the invitation.
> Make assignments. When the members participate there is a sense of ownership and greater buy-in.
> Sustain the effort for future meetings. This was done by setting a date and time for the next meeting
before adjourning.
524
-------
The process of creating a successful regional public education and outreach organization does not happen
over night. It takes careful planning, consistent effort, discipline, and cooperation to build the foundation of
an organization that will have a lasting impact. STORM has found success in applying these principles.
Realizing Results
This stormwater public outreach organization that began as a dream is now thriving. Great momentum has
been growing, and though the trail has been rough and the path ahead is long, the results are truly amazing
to see. Some of the group's key accomplishments along the way are listed below:
• Existing Resources from Phase I MS4s have been shared with Phase n MS4s
• An Organizational Model, Strategic Plan, and Funding Mechanism have been formally adopted
• A Fiscal Agent has been assigned
• A Governing Board has been elected
• A grant application for $250,000 in funding has been submitted to EPA on behalf of STORM
• A STORM website is being created
• STORM has been recognized as the cover story in the November/December issue of Stormwater
Magazine
• Municipalities are budgeting for participation in the group by July 2003
• A new bond has been forged among participating municipalities
Each step toward these accomplishments was small. But steadily these steps moved STORM down the path
to monumental accomplishments. The future of STORM is looking brighter all the time, and the leadership
of STORM on regional stormwater education has been significant.
References
ADEQ's Quest for NPDES Program Approval, 2002. Retrieved August 20, 2002 from Arizona Department
of Environmental Quality Web site: http://www.adeq.state.az.us/environ/water/permits/federal.htmWquest
Climate, 2002. Retrieved August 20, 2002 from University of Arizona Web site:
http://ag.arizona.edu/oals/watershed/highlands/climate.html
Environmental Protection Agency, 2000. Fact Sheet 2.3 Public Education and Outreach Minimum Control
Measure, Storm Water Phase II Final Rule Fact Sheet Series.
Michael A. Worlton, P.E. and J. Ryan Christensen work in the water resources department of RBF
Consulting's Phoenix office. They have assisted in the development of stormwater management programs
for multiple NPDES Phase I and Phase U municipalities throughout Maricopa County, Arizona. Mr.
Worlton led the establishment of STORM and is currently serving as facilitator of the organization.
525
-------
The Watershed Partners:
An Education Collaboration That Works.
Tracy Fredin,
Director, Hamline University's Center for Global Environmental Education.
Founding Steering Committee Member, Watershed Partners
Executive Summary
The Watershed Partners is a coalition of over 50 non-profit and public organizations in the Twin
Cities metropolitan area. Its mission is to promote public understanding that inspires people to
act to protect water quality in watersheds. Formed in 1995, it is directed by a Steering
Committee appointed from its member organizations that operates on a consensus basis. The
activities of the Watershed Partners are coordinated through Hamline University's Center for
Global Environmental Education.
The Watershed Partners' hallmark work over the past seven years has been the development of a
traveling educational exhibit that has been viewed by over 750,000 people. Additional projects
have extended the reach of the Partners' conservation messages to a total estimated audience of
2.5 million people—an audience that continues to grow. Most recently, the Partners have begun
work with a collaboration of over 100 Minnesota cites under the direction of League of
Minnesota Cities and the Minnesota Pollution Control Agency to develop a guide plan for
implementing the NPDES Phase n in Minnesota. We are also developing an integrated
education initiative that will address the six minimum control education measures with
multimedia, printed materials, exhibits, community outreach and education, and K-12 education
projects.
Introduction
The Watershed Partners has grown from a small group of educators into a collaborative of over
50 nonprofit organizations, universities, businesses, and government agencies (local, regional
and national). This consensus-based coalition continues to grow and create new projects. We
have six main programming areas that take differing approaches to educating the public about
watersheds and non-point source pollution. Over the past seven years, our educational messages
have reached over 2.5 million people and we have administered over $1.5 million in grant
funds—all without existing as an official entity!
The Watershed Partners' education initiatives help citizens make informed, environmentally
conscious decisions and take responsible actions. These efforts target a key underlying issue that
makes non-point source pollution (NPS) such a challenge: few people are aware of the impacts of their
own daily activities on their watershed. There is a tendency to think it is the "other person" who is
526
-------
responsible for NFS. Watershed Partners' projects emphasize that we are all potential polluters and
that there are some basic things that can be done to minimize the pollution of our rivers. Toward this
end, we are guided by three primary goals:
1. Educate the public about what a watershed is and how a watershed functions
2. Provide the public with an understanding of their personal connection to their watershed
and their impact on NFS
3. Provide motivation to act in support of a healthy watershed
To achieve our goals, the Watershed Partners have created six integrated programs that have a
cumulative effect greater than the sum of their parts. Our program growth has resulted from a
combination of strategic planning and entrepreneurship. These six programs include: the
Watershed Exhibit, the Metro Media Campaign, the Volunteer Stream Monitoring Partnership,
Project NEMO, The Watershed Education Network, and the League of Minnesota Cities NPDES
Phase n Education Initiative.
Watershed Exhibit
The Watershed Exhibit is a suite of museum-quality, hands-on interactive educational modules
that may be used independently or together that has engaged more than 750,000 people over the
past seven years. Four exhibit elements, each of which occupies a collapsible 8-foot table against
a colorful fabric banner backdrop, focus on the following topics:
• What is a Watershed?
• What is your Watershed Address?
• Your Street Flows to the River
• Clean Water Starts with You
In addition, an interactive multimedia kiosk program (in English, Spanish, and Hmong) with six
modules that reinforce the messages conveyed through the four table displays can be
incorporated into the exhibit or used separately. The exhibit can be set up under a specially
designed tent. A van has been dedicated to transporting the exhibit.
The goal of this interactive exhibit is to educate participants so that they will leave it knowing
what a watershed is, understanding their personal connections to their watershed, and being
motivated to take stewardship actions. For example, the interactive kiosk effectively delivers the
message to participants that polluted runoff from their homes and yards flows, untreated, directly
to the Mississippi River.
The development and implementation of the exhibit was made possible by significant funding
from the Metropolitan Council and by other Watershed Partners.
527
-------
This past year, the Watershed Exhibit accomplished the following:
• Served over 160,000 people
• Increased capacity through Watershed Ambassadors, Watershed van and multimedia kiosk
system has proved successful
• Created new urban sprawl education module for the multimedia kiosk
• Developed new models to use local students as educators and distribute kiosks to public
buildings as a service learning project
• Conducted national marketing survey to explore market for NFS education tools
• Expanded impact at the Minnesota State Fair by integrating exhibit into MN Pollution
Control Agency booth
"Think Clean Water" Campaign
In 1999, Metro Watershed Partners implemented the "Think Clean Water" communications
campaign that used broadcast media to communicate two main messages: 1) Keep grass
clippings, leaves and fertilizer out of the street, and 2) Use low phosphorus lawn fertilizer.
Evaluation at the conclusion of the campaign showed that 57% of individuals surveyed heard or
read information about using low phosphorus fertilizer and 49% heard or read information about
keeping grass clippings and leaves out of the street. Over 2 million media exposures were
created by this campaign. The campaign was supported by a $200,000 grant from the
Metropolitan Council and untold in-kind labor of Metro Watershed Partner members.
Additionally, this past year the following has been accomplished:
• Minnesota has passed a no phosphorus bill at the state legislature
• Watershed Partners have created an educational brochure that addresses the no phosphorous
fertilizer regulation
• $90,000 has been secured to initiate another media campaign
Volunteer Stream Monitoring Partnership (VSMP)
In the past five years, interest in volunteer monitoring has exploded in the Twin Cities area, with
39 volunteer monitoring groups identified in a recent inventory having begun their activities after
1994. In November 1999, the Watershed Partners coordinated the development of a strategic
plan for coordination of volunteer stream monitoring in the seven-county Twin Cities
metropolitan area. Representatives from 15 organizations participated in the development of the
strategic plan.
528
-------
Volunteer stream monitoring programs are based on three experience-tested principles:
1. Volunteers can collect reliable, meaningful data that can be used by decision makers in
watershed planning and management
2. Volunteer monitoring programs promote watershed stewardship by engaging volunteers in
understanding and managing natural resources
3. Successful volunteer monitoring is a blend of education and science, supported by local
units of government, educators, agencies, industry, and non-profits
The goals of the Volunteer Stream Monitoring Partnership are as follows:
• To facilitate the collection and management of quality volunteer stream monitoring data
• To effectively involve local, regional, and state agencies, including encouraging them to use
volunteer generated data
• To engage volunteers
In the process of achieving these goals, we anticipate that the public will become more aware of
river issues and more inclined to protect water resources. We also expect that water quality-
monitoring resources will expand as the quality and amount of data available for decision-
making improves at the state, community, and individual level. Finally, we anticipate that a
centralized data management system with a watershed perspective will be developed and made
accessible to agencies and volunteers.
Quality assurance and quality control of the collection of data is an important component of this
partnership. All data collection is based on standard EPA protocol in coordination with the
Minnesota Pollution Control Agency and the Metropolitan Regional Council. Data is collected
by volunteer students and their teachers. The data is then checked by the county coordinators
and double-checked by the VSPM coordinator. At that time the data is entered into a database
coordinated by the Metropolitan Council. This information has been used by various counties in
their year end reports and in developing their water plans.
In the past year the VSMP has:
• Convened first official year of operation focusing on ensuring quality data, strengthening
partner collaboration and expanding outreach, securing funds and increasing capacity
• Worked with nearly 1,900 citizen volunteers
• Hosted a River Summit for over 230 students and professionals
• Provided 13 trainings for local partners
529
-------
• Monitored 57 different stream sites in the five-county metro area
Watershed Education Network for Teachers
The Metro Watershed Education Network uses communication, coordination and collaboration
to educate youth about NFS; integrate watershed education into school systems; and leverage the
energy created by the award-winning Watershed traveling exhibit. The Watershed Education
Network Project has successfully increased the capacity of teachers and schools to integrate
water quality education into their students' learning activities.
The Watershed Education Network has:
• Developed a Watershed Education Network Web site with educational resources and calls
for stewardship action
• Integrated water issues into the Sharing Environmental Education Knowledge (SEEK) Web-
based environmental education clearing house
• Created a Watershed Listserve for over 180 educators
• Created a Watershed Hotline for individuals
• Recruited and trained teachers as Watershed Ambassadors through hands-on summer
institutes, the watershed training sessions, and online graduate course work
• Created an award-winning graduate course for 20 St. Paul educators that infused watershed
education into the St. Paul Chamber of Commerce's Teacher in the Work Place program
• Created the framework for the Watershed Partners to act as watershed content experts
• Infused watershed education into the St. Paul, Minneapolis, and suburban schools systems,
and the Grand Excursion 2004 Special Event
Project NEMO (Nonpoint Education for Municipal Officials)
Project NEMO is an educational program for land-use decision-makers that addresses the
relationship between land use and natural resources protection, with a focus on water resources.
NEMO was created in 1991 at the University of Connecticut and, due to the success of the
program, has grown to become a national network of projects in 15 states. "Linking Land Use to
Water Quality and Linking Town Hall to Technology" is the NEMO motto.
Land use in the United States is largely decided locally by elected and appointed officials serving
on county and municipal boards and commissions who have not been chosen for their knowledge
of natural resource protection and often have little or no professional staff to support them. As a
result, local land-use decision makers have been largely left out of the nonpoint source pollution
reduction equation. A primary goal of Project NEMO nationwide is to provide education for
these officials and to inspire them to take action.
530
-------
For the Minnesota NEMO program, the first goal of the project is to develop and provide locally
adapted NFS educational materials for a targeted audience of local land-use officials and thereby
enable them to plan for growth while addressing water quality through wise land-use decisions.
The second goal is to incorporate this educational message into actual changes in policies,
practices, and plans at a local level. The third goal is to bring together and develop relationships
between regional and state agencies, water management organizations, conservation districts and
other associations interested using land-management decisions to protect water quality and
natural resources. The fourth goal is to establish a coordinator position that will be responsible
for implementation and expansion of the NEMO program in the Twin Cities Metropolitan area
and the achievement of goals 1-3.
Project NEMO has successfully:
• Presented Project NEMO programs to over 40 cities
• Received additional funds from the Met Council
• Been identified by the Minnesota Environmental Quality Advisory as a major initiative for
next year
• Received EPA 319 funds
• Impacted over 200 city council members and decision makers
League of Minnesota Cities NPDES Phase n Guide Plan and Educational Initiative
Hamline University and Watershed Partners have recently worked with the over 100 cites,
coordinated through the League of Minnesota Cites and the Minnesota Pollution Control
Agency, to assist in developing the NPDES Phase n Guide Plan. This plan provides cites a
template to develop their Phase n plans in Minnesota. It is an attempt to provide guidance and
standardization for the cities in order for them to best meet their needs. The Watershed Partners
has focused on the educational component's six minimum control measures, while the
engineering firms of Boonestro and Associates and AMEC have provided the primary structure
of the document and other technical information.
In the past year we have:
• Assisted in writing the education component of the Phase n Guide Plan, with an emphasis on
educational components of the six minimum control measures
• Provided training sessions for over 80 cities
• Made plans to develop an integrated campaign with media, printed materials, exhibits,
community outreach and education and K-12 education projects
531
-------
Results
Over the past seven years, over 2.5 million people have been impacted in the Twin Cities Metro
Area by Watershed Partners projects. Since the Watershed Partners programs differ, not all of
the impacts have been the same, nor can they all be measured by the same gauge. While we are
still in the formative evaluation stage of these initiatives, evaluations from two independent
sources have begun to document impacts.
An independent evaluation of the Watershed Exhibit by the Wilder Research Center has
indicated:
• 75% of the visitors to the Watershed are be able to accurately describe what a
watershed is
• 60% of the visitors to the Watershed exhibit gain ideas for keeping their watersheds
healthy and reducing their NFS pollution contributions
An independent evaluation of the Watershed Partners' Media Campaign indicates that:
• 57% of individuals surveyed heard or read information about using low-phosphorus
fertilizer
• 49% heard or read information about keeping grass clippings and leaves out of the
street
One of the most interesting components of this project is comparing the level of NFS awareness
in Minnesota to the nation at large. Using the National Environmental Education Training
Foundation's National Report Cards on Environmental Knowledge, Attitudes and Behaviors as a
guide, Hamline University conducted a survey of the citizens of the State of Minnesota and
compared it to the national standard. In most categories, Minnesotans scores were similar to
national averages. Regarding knowledge of non-point source pollution, however, Minnesotans
scored over 100% higher than the national average. That is to say, only 24% of the nation
understood NFS, while 52% of Minnesotans understood this concept.
While this cannot be directly correlated to the work of the Watershed Partners, we believe our
efforts have played a small role in this outcome.
Conclusion
The Watershed Partners is a collaboration that is effective in many different ways. By engaging
educators and organizations in the Twin Cities that have a stake in educating the public about
watersheds and non-point source pollution, we have been able to build on and greatly magnify
the impacts of our partner organizations.
532
-------
Three critical components for success can be identified:
1. Information Sharing-Monthly meetings provide a forum for the Partners to share information,
network, and gain new knowledge. Sometimes, a lot of business gets done during the
informal time before and after the meetings. There is much less "reinventing the wheel" in
local areas.
2. Pooled Resources-By working together in a coordinated fashion, the Partners are able to
create products and services that would be difficult to create individually. This system
provides incentive for collaboration. For example, the Watershed Exhibit is a resource that
can be used by any of the partners when they need it, and they do not all need to own a
$100,000 exhibit.
3. Coordinated Efforts-By coordinating efforts, organizations can more effectively focus on
their particular niches and put forward an integrated effort to educate the public about how to
protect their watersheds. Synergies can be built around programming and fund-raising
opportunities.
Through collaboration and consensus, the Watershed Partners has been able to serve over 2.5
million people in the Twin Cities Metro Area in rich and diverse ways. Our hope is to be able to
leverage this partnership to assist others in the local, regional and national setting to better
educate the public and get them to take action about important watershed issues.
533
-------
Metro Watershed Partners Milestones
1992-3
An idea by naturalist Karen Kobey of Hennepin Parks stimulates a network of agencies, non-
profits and educators interested in the future of the Mississippi River to collaborate on a
conference, "The Ever Changing Mississippi" held in Feb, 1993.
1994-5
"Summer of the River," coordinated by Shelley Shreffler of Macalester College, and an
informal partner network provides exhibits at outdoor events under banner entitled
"Watershed" (term coined by Ron Erickson of National Park Service). Displays include
historical and water quality topics.
1995
A fall conference entitled "Awakening the Watershed," sponsored by Summer of the River
and the Mississippi National River & Recreation Area (MNRRA) is held in Red Wing, and
provides a springboard for educational partnership development.
In December, a group is convened by MNRRA to further develop a watershed education
partnership and an interactive watershed exhibit. (Exhibit Goal agreed: "The visitor to the
Watershed will leave with the knowledge of what a watershed is, an understanding of their
personal connections to their watershed, and the motivation to act in support of a healthy
watershed.")
First regular meeting attendees: Anoka County Parks, Army Corps of Engineers, Center for
Global Environmental Education (CGEE), Friends of the Mississippi River, Greening the
Great River Park, Metropolitan Council Environmental Services, Minnesota Valley National
Wildlife Refuge, MNRRA, Science Museum of Minnesota.
1996
Monthly meetings established, convened and facilitated by MNRRA.
Group name and Mission Statement established: "The Water Shed Partners is an informal
association of organizations committed to addressing shared goals pertaining to watershed
education through educational projects, networking and sharing resources." Focus is on
educating to prevent runoff pollution.
Prototype exhibits developed and utilized at outdoor, summer events.
First $100,000 grant received from Metropolitan Council to create Watershed interactive,
mobile exhibits. CGEE coordinates grant process for WSP.
Internal processes and structures established, including consensus decision making and
Steering Committee.
First Steering Committee members include: Cliff Aichinger, Marie Asgian, Tracy Fredin,
Pauline Langsdorf, and Lyndon Torstenson.
Exhibit design development begins facilitated by Science Museum of Minnesota (SMM), and
involving a committee of the Watershed Partners.
534
-------
Magnets and set of brochures created by NEC, CBE, MNRRA & WSP.
1997
New "Watershed" exhibits built by Science Museum of Minnesota are completed. Four
tables include: 1) What is a Watershed? 2) What is your watershed address? 3) Your street
flows to the river, and 4) Clean water begins with you. Custom modified tent is purchased to
house exhibits in events.
New mission established: "The Watershed Partners promote a public understanding that
inspires people to act to protect water quality in their watershed."
150,000 people interact with Watershed exhibits in 1997, including over 45,000 at the
Minnesota State Fair.
Number of partners grows from 12 to 32.
Committees include steering, exhibits, education, evaluation, public outreach.
1998
Public media campaign ("Water Quality Action Campaign") undertaken in collaboration with
Board of Water & Soil Resources reaches 2.5 million households.
Computer interactive developed.
Staffing support for exhibits and partnership established through CGEE.
Exhibit evaluation conducted by Wilder Foundation reveals notable learning occurring in
response to exhibit interactions: 92% correctly define watershed.
CGEE establishes Watershed Partners website.
WSP receive Partnership Minnesota award.
Metro Watershed Education Network initiated at CGEE with $35,000 grant for the
Metropolitan Council
1999
"Water Education Resource" book of ready-to-use educational materials created.
WSP awarded top honors by the Minnesota Environmental Initiative.
WSP receive MN GREAT award (Minnesota Government Reaching Environmental
Achievements Together.)
Water Quality Monitoring initiative undertaken.
Partners number over 40; monthly meeting attendance regularly over 25.
Sponsorship of national conference considered.
Watershed Stewards Curriculum established and modeled at Farnsworth Elementary School.
2000
Volunteer Stream Monitoring Project receives $500,000 grant over 3 yrs from Met Council.
LCMR funding proposals developed and presented to Legislative committee.
"Project NEMO" (Nonpoint Education for Municipal Officials) launched.
2001
McKnight awards $150,000 two-year grant to CGEE for WSP programs.
535
-------
Over $1.5 million in grants have been successfully administered by the WSPs.
First annual River Summit held for stream monitoring volunteers.
Project NEMO gets $93,000 grant from Metropolitan Council.
Van purchased and outfitted for exhibits and events use, thanks to 50/50 matching challenge
grant from Ramsey Washington Metro Watershed District.
Name "Metro Watershed Partners" (MWSP) adopted and officially registered after
"Watershed Partners" name is found to be already registered by a real estate company.
MWSP officially becomes project of CGEE, providing liability and other benefits.
2002
State phosphorus legislation passes, thanks to MN Dept of Agriculture, and several WSPs.
Over half a million people have interacted with the Watershed exhibits since 1997; the
exhibits have been displayed at national conferences and have been a national model.
Nearly 100,000 people interact with Watershed exhibits at State Fair alone.
Project NEMO receives $125,000 "319" grant from MN Pollution Control Agency, and
$50,000 from Metropolitan Council.
$50,000 grant for "Think Clean Water" media campaign from Metropolitan Council
$40,000 grant for "Think clean Water" media campaign from Office of Environmental
Assistance.
536
-------
EDUCATING THE LAS VEGAS COMMUNITY ABOUT STORM WATER
POLLUTION
Betty Hollister, APR
Clark County Regional Flood Control District
Las Vegas, Nevada
Abstract
The Clark County Regional Flood Control District, located in Las Vegas, Nevada, is the umbrella agency
that administers the region's National Pollutant Discharge Elimination System (NPDES) permit. While the
majority of the District's outreach efforts have been focused on flood safety education, the District has
moved forward with increased public outreach about urban runoff and storm water pollution in the last two
years. The Las Vegas Valley drains to the Las Vegas Wash, which drains to Lake Mead, the area's primary
source of drinking water. With more than 6,000 new residents moving to the community each month, the
education process about flood safety and storm water quality are continuous. New and innovative measures
are needed to provide multiple impressions and reminders to the community about the impact their behavior
can have on the environment.
Background
The current population of the Las Vegas Valley is 1.5 million, with only 24% of those residents being born
in Nevada. An average of 6,000 new residents move to the Valley each month, making Las Vegas one of
the fastest growing cities in the nation. Almost one-half of the area's residents have lived in Las Vegas less
than 10 years, and one-third of those have lived in the Valley less than five years (Las Vegas Perspective,
2002). To put this growth into perspective, in 1950 the city's population was 47,000, and every 10 years
since, the population has doubled. The area is experiencing all of the challenges associated with other major
metropolitan areas. In addition, the arid desert climate and drought conditions facing many of the western
states make water quality and water availability major concerns for the area.
537
-------
Watershed Basins in the
Las Vegas Valley
Ltftd
:
Figure 1 Las Vegas Valley Watersheds
The average annual rainfall for Las Vegas is approximately 4 inches. However, in 2002, the area received
less than 1.5 inches of rain (Historical Rainfall Data, 2002). The geography of the Valley slopes from the
west to east with seven major washes passing through the urban area (Figure 1). All of these washes
converge on the east side of town at the Las Vegas Wash, which drains to Lake Mead, the area's primary
source of drinking water. Five percent of the flow through the Las Vegas Wash into Lake Mead is from
storm water; 5% is from over-irrigation, surface and groundwater; the remaining 90% is highly treated
wastewater.
In accordance with the Federal Water Pollution Control Act, the Clark County Regional Flood Control
District, as lead agency, was granted a National Pollutant Discharge Elimination System (NPDES) permit in
December 1990. The Nevada Division of Environmental Protection issued the permit to six co-permittees
representing the various city, county and state agencies owning and operating municipal separate storm
sewer systems in the Las Vegas Valley.
538
-------
The Storm Water Quality Management Committee was formed with the Regional Flood Control District as
the umbrella organization funding the majority of storm water activities, like dry and wet weather testing of
water entering Lake Mead through the Las Vegas Wash. Public outreach activities until recently were
limited primarily to environmental "fairs" like Earth Day events. With the addition of a Public Information
Manager in 2000, the Regional Flood Control District placed more emphasis on public outreach and
education about storm water quality.
Research
An initial brainstorming session was held with members of the Storm Water Quality Management
Committee to determine the focus each of these organizations hoped to take with the outreach efforts. From
this two-hour session, it became clear that this group of 20 people had differing opinions about the content
of the information campaign and the target audiences to be reached.
Mall Intercept Survey
An informal survey was taken at the area's three largest shopping malls to determine residents' awareness
level of the problem. While this was a non-scientific survey, it was hoped the results would point the
communication efforts in a certain direction.
After surveying 150 residents, the results showed that approximately 50% of the respondents were not
aware that floodwater and urban runoff flowed through the storm drain network untreated to Lake Mead.
Discussions also showed that residents were unaware of proper disposal of various pollutants, especially
how to drain their swimming pool. Most of those surveyed were aware that Lake Mead was the Valley's
primary source of drinking water.
By partnering with the Southern Nevada Water Authority the following year, the District was able to
include two questions on their next telephone survey of residents at no charge. These questions were similar
to the mall intercept questions and the results were also similar. While 71% of respondents knew Lake
Mead was the area's primary source for drinking water, 32% said they believed urban runoff was treated
before entering the Las Vegas Wash and Lake Mead. Thirteen percent did not know. From this survey, it
was determined that the first step in public outreach should be education about untreated runoff and
stormwater.
539
-------
Stormwater Quality
Management Committee
Figure 2 Storm Water Logo
Website/Logo Creation
Other environmental websites were researched and evaluated. Information was compiled and edited using
information from several sources. A member agency staff person agreed to construct a storm water website
as a volunteer service. The Regional Flood Control District paid registration costs associated with the site.
The logo that was used prior to the website creation did not clearly communicate that water flowing through
storm drains was untreated. The committee, with permission, modified a logo from a California community
so that it better represented the Las Vegas environment. (Figure 2).
Lesson Learned: Research, evaluate, coordinate and borrow ideas (with permission). Feel free to borrow
any of our ideas at http://www.lvstormwater.com/. We were also fortunate to have a volunteer Webmaster
who is highly capable and dedicated to the effectiveness and accuracy of our site.
Public Service Announcements
The primary objective of our mall intercept survey was to first educate the community that urban runoff and
storm water are not treated before entering Lake Mead. The concept of a toy boat floating through gutter
water, falling into a storm drain and being "found" in Lake Mead was used. The 30-second spot was
produced at no charge by Clark County's Communication Team who operate the county's government
access station. They were enthusiastic about producing a commercial that allowed a large amount of
creativity and clever camerawork. An award-winning public service announcement (PSA) resulted that has
received five first-place local and national awards.
Lesson Learned: While the Toy Boat spot was clever and award winning, it had no news or event hook
for the television stations. Each of our local network affiliate stations are bombarded with about 35 new
public service announcements each month, many of which are tied to an event or are co-sponsored by the
TV station. Consequently, the Toy Boat PSA saw very little airtime. The District recognized that the next
PSA must have some "hook" for it to be used by the media.
540
-------
The District also explored producing 10 or 15 second spots but learned that there are only so many "natural"
breaks of these shorter time slots with the network affiliates - that would have limited even more the
possibility of airtime. In addition, production costs would have been essentially the same.
Two other public service announcements were produced following the Toy Boat educational spot. Each
PSA was produced at a negotiated rate of $2,300. These spots focused on behavior changes that could help
improve the quality of urban runoff and storm water. One pointed out proper fertilization and irrigation of
landscaping and was distributed in April to coincide with the Las Vegas Valley Water District's water
conservation campaign. This PSA was aired by all three network affiliates in both April and May, 2002.
The "hook" was two-fold: 1) The fertilizer/irrigation PSA was distributed in the Spring during a time when
people begin working in their yards, and 2) The Las Vegas Valley Water District's water conservation
campaign (paid advertising) was running heavily during this time.
Only one television station in Las Vegas provides documentation of PSA airtime, KWU-Fox 5. With a
program called PR-Trak, the District was able to document -just from this one station - that the fertilizer
spot aired 70 times in a two-month period and was viewed by several hundred thousand people. This
program uses actual Neilson ratings for individual markets. This program is also helpful in summarizing
media coverage, both quantitative and qualitative, and provides accountability for the communications
effort.
The third PSA focused on proper disposal of pet waste and was distributed in June. Knowing that the news
or event "hook" was missing, television advertising departments were contacted about placement of the spot
as a commercial. A "bonus" schedule was agreed to that gave free and extra placement of the spot in July
for paid time in June. A competitive advertising rate request (Request for Avails) was conducted to ensure
the best available television schedule, ratings and prices. Each of the three stations received $3,000 from
the Regional Flood Control District.
All three PSAs can be viewed from the www.lvstormwater.org website.
Homeowners' Associations
A one-page camera-ready article was produced and mailed to a database of 300 Homeowners' Associations.
A cover letter from the Storm Water Quality Management Committee explained the importance of
educating the community about how they could help protect Lake Mead, which is our primary source for
drinking water. The same article was also sent to the neighboring cities and county for inclusion in
newsletters they mail to residents.
Lessons Learned: The one-page camera-ready article was apparently not widely used by the
Homeowners' Associations. The District received seven phone calls thanking it for the information, but did
not put in place a method to secure a copy of the next newsletter from each Association. While the text and
layout were standard and "ready-to-use," it appears that personal phone calls to the major associations may
have worked to build better response than just a blind mailing. The article was, however, widely used in the
newsletters produced by the cities and county. The District also revised the text of the article to focus on
business best management practices and sent the mailing to related businesses. In response to this mailing,
11 businesses called to discuss concerns they had regarding their policies and to ensure that they were in
compliance. The District plans to work with Homeowners' Associations again in the spring of 2003. This
will coincide with new and more expensive watering rates that go into effect along with stricter water
conservation guidelines and citations.
541
-------
DON'T
^POLLUTE!
DRAINS TO LAKE MEAD!
Figure 3 Storm Drain Plaques
Storm Drain Markers
Various types of markers were evaluated based on the hot, desert climate. A plastic version was chosen that
used a special adhesive. A $65,000 grant from the state and the local conservation district funded the
purchase of 12,000 plaques (along with other collateral material) to be distributed to the five city and county
entities (Figure 3).
Lesson Learned: The funding did not include the installation of the plaques. The job for installation fell
on the Public Works/Maintenance Departments to "fit in" as they had time. After a year, only a few
hundred plaques had been installed in violation of the terms of the grant. Meetings were held to determine
alternative ways of installation. Because of the toxic nature of the glue and liability issues (some students
had been killed while picking up trash on a roadside), the only alternative was to contract the job out or seek
direction from top management. A combination of the two was used with the Regional Flood Control
District assisting with contract labor costs. All the storm drain makers were placed by December 2002.
The Flood Channel Television Program
The District produces six 30-minute television programs each year under consultant contract for $15,000 per
episode. Two programs in the last l!/2 years were devoted to storm water quality - education and behavior
change. Several awards were received for the "Protecting the Environment" episode. These programs air
on our two local government access television stations and receive about 40 airings each month. The County
Government Access Station (C-4) airs its programming on the Internet via the county's website,
www.co.clark.nv.us. The Flood Channel television program can be viewed from the county's website.
Lessons Learned: Segments of the program educated the community about environmentally friendly
businesses and the actions they were taking to conserve and protect the environment. Other segments
showed what actions residents could take to improve water quality. The interviews with businesses were
difficult to obtain because they were reluctant to go on camera - perhaps they were not doing all they could
do or were afraid of repercussions from regulators. These companies included pool cleaners, carpet
cleaners, mobile dog groomers, automotive service and car washes. With the second environmental
episode, the District made the initial phone calls using public relations contacts and other relationships built
over the years.
542
-------
Figure 4 Bus Stop Shelter Ad
Bus Stop Shelter Ads
The City of North Las Vegas received a grant for public outreach about storm water quality. It chose to
focus on proper disposal of pet waste as a reinforcement of the public service announcements. The city
produced 25 bus stop shelter posters (Figure 4) that were in place from September through December 2002
(four months). The size of the posters was 4 feet by 6 feet. Total cost of artwork, production and placement
was $8,000. A similar version of this message was also distributed to North Las Vegas residents via utility
bills one month prior to the bus stop shelter posters being put in place.
Lessons Learned: The artwork for this effort was incredibly eye-catching. An out of focus woman held
a bag of pet waste (in focus) with the words "Do Your Doody" written on the bag. The sub-heading was
"Protect the Environment" (our tag line for all the PSAs) and the words "Pick Up After Your Pet." While
the District did not receive any feedback from residents, those associated with the campaign were really
grabbed by the artwork. One change would be to downplay the woman's fingernails (they were emphasized
543
-------
in such a way that they distracted from the pet waste bag). A copy of the artwork is also on the
www.lvstormwater.com website.
Brochures
The District is currently finalizing a best management practices brochure for those wash water and urban
runoff related businesses seeking new licenses. This two-color brochure was created with simple graphics
and examples of various low to high impact activities. Funding for this effort was shared with the Southern
Nevada Water Authority. A limited number of brochures are being printed (5,000) and an evaluation of its
success will determine if more should be produced and if any revisions are needed. A copy of the brochure
is available on the www.lvstormwater.com website.
Lessons Learned: Because this was a committee effort, with many agencies and government entities
involved, the process of producing this brochure took about six months. The committee met once a month.
Additional reviews were required for every suggested change. Moreover, one entity would suggest a
change that was not in agreement with the regulations of another community, which would necessitate
further changes. Because of such difficulties, the committee decided to print and distribute only a limited
number of copies.
Community Events and Collateral Material
The Regional Flood Control District takes part in spring and summer environmental fairs, as well as events
geared toward pet owners. The District has produced several collateral materials for distribution: pet food
lids, pooper scoopers, sponges, stickers and coloring books to name a few. It also uses an enviroscape
model (a landscape topographical model) that shows how various pollutants are carried by rainwater into a
lake. These are all helpful in getting the message out about storm water quality.
Lessons Learned: It is best to participate in smaller, organized events that provide the crowd a schedule
of when demonstrations (like with the enviroscape) will be held. This allows for coordinated presentations
with better audience participation. While brochures are a standard in these events, we believe that focusing
our participation helps to more effectively get the message out.
School Outreach
A four-page school curriculum was produced after a year of research to determine how much information
was needed and in what format. The curriculum focuses mainly on flood safety, but storm water and the
pollutants it contains are also included. Personal school presentations last year reached 45 schools and
approximately 8,000 elementary students. As requested by teachers, the material was also mailed to
schools, reaching 15,000 additional elementary students. A six-minute video and student activity book were
also included.
Lessons Learned: The research was crucial to ensuring that the curriculum met both local and state
education requirements. The curriculum met both science standards and health and safety requirements.
Four pages of teacher information, a student test and teacher evaluation helped ensure teachers' use of the
material and gave the District immediate feedback. From the feedback, the District saw a need to produce
the video in Spanish for schools with higher populations of non-English speaking students. The District
also included an interactive version of the activity book on its website at www.ccrfcd.org.
544
-------
Conclusion
Because of the tremendous growth the Las Vegas Valley has been experiencing, the District's two major
audiences are new residents and construction companies. The Nevada Division of Environmental Protection
is stepping up enforcement of construction best management practices in Las Vegas and the District will be
assisting them with their education efforts. The District is also exploring the next focus of its Public Service
Announcements. One possible topic is boating on Lake Mead, because of new regulations restricting
certain types of watercraft on the lake.
Currently, three sanitation districts discharge highly treated wastewater into the Las Vegas Wash that flows
into Lake Mead. The sanitation districts contribute approximately 90% of the annual flow in the Las Vegas
Wash. The sanitation districts are now considering systematically eliminating their Las Vegas Wash
discharges by piping their flows farther into the lake or to the Colorado River. If and when this occurs, the
capacity for dilution of urban runoff and storm water pollutants in the Las Vegas Wash will be decreased,
thus resulting in greater concentrations of these pollutants as they reach Lake Mead. While the various
agencies involved with the Wash are building grade control structures and wetlands in the Lower Las Vegas
Wash to help improve water quality, the District continues to evaluate how to most effectively educate the
public on behavioral changes that have positive impacts on the environment.
References
Historical Rainfall Data, 2002, www.ccrfcd.org compiled from the National Weather Service official gage
at McCarran International Airport.
Metropolitan Research Association, 2002, Las Vegas Perspective.
545
-------
ILLICIT AND INDUSTRIAL STORM WATER CONTROLS:
A MUNICIPAL PERSPECTIVE
Michael J. Pronold, Industrial Stormwater Manager
Ali Dirks, Illicit Discharge & Spill Protection Environmental Technician
Bureau of Environmental Services
City of Portland
Portland, Oregon
Abstract
As part of the Environmental Protection Agency (EPA) Phase 1 storm water requirements, the City of
Portland, Oregon (City) was responsible for developing a program to monitor and control pollutants in
storm water runoff from industrial facilities to the municipal separate storm sewer system (MS4). In
addition, certain classes of industries are required to obtain National Pollutant Discharge Elimination
System (NPDES) Industrial Storm Water permits. The EPA, or a State Agency that has been delegated by
EPA, administers these permits. Addressing storm water runoff from industries under these separate
programs can result in redundant efforts and a less than efficient program. EPA and/or State agencies
may not have the resources to adequately administrate and enforce the permitting program while still
leaving the municipality liable for the discharges from the MS4.
The City chose to meet the requirement in their municipal storm water permit to control industrial storm
water sources of pollution by developing a Memorandum of Agreement (MOA) with the Oregon
Department of Environmental Quality (DEQ), (which is the delegated authority) to administrate the
permit program. The MO A outlines the responsibilities of the City and DEQ for the implementation of
the program, including notification of permit requirements, inspections, compliance, and enforcement
issues.
To implement the provision of the Illicit Discharge Elimination Program, the City identified and
prioritized 109 major outfalls in the MS4. Maps were developed that outlined the drainage basin and over
3,000 industrial and commercial facilities were researched using building and plumbing records to
identify illicit connections. Outfalls are inspected monthly during dry weather and flows sampled to
detect the presence of illicit discharges. The City has also developed a citizen complaint program to
facilitate the reporting of spills and illicit discharges.
Industrial Storm Water Program
Storm water discharges have been increasingly identified as a significant source of water pollution in
numerous nationwide studies on water quality. To address this problem, the Clean Water Act
Amendments of 1987 required EPA to publish regulations to control storm water discharges under
NPDES. EPA published storm water regulations (55 FR 47990) on November 16, 1990 which require
certain dischargers of storm water to waters of the United States to apply for NPDES permits. The
regulations include NPDES permit application requirements for storm water discharges associated with
industrial activity. EPA has defined this phrase in terms of 11 categories of industrial activity. The DEQ
has been delegated by EPA to administrate the program and started issuing Industrial Stormwater permits
in 1991.
546
-------
As a Phase 1 city, Portland was required to develop a program to monitor and control pollutants in storm
water runoff from industrial facilities in accordance with 40 CFR 122.26(d)(2)(iv)(C). This creates the
potential for redundant efforts and an inefficient program. The City is ultimately responsible for
discharges from their MS4. To meet the requirement in their municipal storm water permit and to provide
the oversight necessary to protect itself from liability, the City developed the legal authority and entered
into an MOA in 1994 with the authorized NPDES state authority (DEQ), to administrate the permits for
those discharges to the MS4. The City also inspects and notifies industries that may be required to obtain
a permit. The program is administered by a dedicated work group in the City because of the large
industrial base and number of NPDES Industrial Storm Water permits (approximately 250) within the
City.
Program Elements
Legal Authority
Code was developed in March 1994 to allow the City to have legal authority over storm water discharges
to the MS4. Key elements of the code included the requirement for permit holders to submit their Storm
Water Pollution Control Plan (SWPCP) and monitoring results to the City, the authority for the Director
to adopt administrative rules, inspection authority, and enforcement capability. It was important that the
City reviewed the NPDES Industrial Storm Water permit when code was developed to ensure that any
City identified inadequacies of the state issued permit were addressed. One example would be the
requirement to submit SWPCP and monitoring results to the City as this was not included in the permit.
Another provision that was critical was the ability of the City to implement measures to address facilities
that may not be required to obtain a permit. Currently, federal regulations base the requirement for
obtaining a permit based on Standard Industrial Classification (SIC) Code and exposure. City experience
has shown this to be cumbersome as certain facilities that have activities similar to those facilities that are
required to obtain a permit fall under an unregulated SIC Code. There are provisions in the federal
regulations to request that the permitting authority issue a permit but this could require that the City
undertake sampling and additional work to prove this. This reduces the efficiency of the program in terms
of resources and uniformity. This matter was partly addressed by including provisions in the code that
allows the City to develop its own permit. However, because of concerns about confusion for the
regulated community, plus the current workload of inspecting facilities that may need a permit under the
SIC Code criteria, the City has not pursued this effort to date. Other measures, including the requirement
for secondary containment and the development of Accidental Spill Prevention Plans, are included in code
and used to address non-permitted sites.
Enforcement capabilities, including fines, have been developed for violations of the City's code.
Provisions of the code include general discharge prohibitions, reporting requirements, right of entry,
inspections, and sampling by City staff, and measures to prevent the entry of wastes to the MS4.
Enforcement capability by the City is especially important for "low level" violations, such as late reports.
The DEQ is reluctant to enforce on those "low level" violations, other than with notification letters,
because the minimum fine is $1,000. Where the City does not have enforcement capability, the City must
seek voluntary compliance and refer those violations to DEQ when they are unable to obtain compliance.
Failure to apply for a permit and/or develop a SWPCP in a timely manner are referred to DEQ for formal
enforcement. This has worked to date, but requires coordination between the City and DEQ. To make this
effective, the City worked with DEQ to identify which violations merited referral to DEQ's formal
enforcement process.
547
-------
Memorandum of Agreement
The City entered into a MOA with the DEQ in March 1994, which was revised in 1999. The MOA
delineates the responsibilities for the implementation of the program between the two agencies. Language
is broad enough to not constrict how the City implements the program. There were two key provisions in
the 1999 update of the MOA. One was the submittal of the permit application materials to the City. The
City reviews the applications for completeness and then forwards them to the DEQ. This allows the City
to track the industries' compliance with applying for a permit once the City has notified them. Previously,
the application was submitted directly to the DEQ which proved cumbersome for the City to track
compliance with submittal deadlines. In addition, if the application was incomplete, it was returned by
the DEQ to the applicant with no clear submittal deadline. Another benefit of submittal to the City is the
facilitation of obtaining the Land Use Compatibility Statement (LUCS), which is issued by the City
Planning Department. This allows the applicant to submit all the materials at once as opposed to
obtaining a LUCS separately. The second provision was the authority granted the City to administrate the
permits for those facilities within the City limits but that had storm water discharges through private
outfalls. Prior to this, these facilities were rarely inspected nor was there the level of oversight as with the
other permittees. To account for the added workload, the MOA included provisions for revenue sharing
of permit fees. With approximately 250 permits citywide, this provided adequate funding for one
additional City staff person.
Table 1. Oregon DEQ and City of Portland Select Responsibilities and Funding Allocations Under the MOA
for City Administration of the NPDES Industrial Storm Water Permit
MOA Element
Permit Application and Review
Permit Issuance
Permit termination
Site Inspections
Storm Water Monitoring
Review of Self Monitoring Data
SWPCP
Enforcement
Staffing
Application Fee ($670)
Annual Fee ($275)
Oregon DEQ
Review for applicability
DEQ responsibility, notify City
DEQ responsibility, consult City
Upon request, at discretion
Upon referral
1 FTE Northwest Region of Oregon
50%
25%
City of Portland
Track application submittal, review for
completeness, forward to DEQ.
Notification of non-compliance and
referral to DEQ for enforcement.
Confer with action
Annual at a minimum
Annual, weather permitting
Review for compliance, notification of
non-compliance, and referral to DEQ for
enforcement.
Track submittal, review for completeness,
notification of non-compliance and referral
to DEQ for enforcement.
Enforce City Code, seek voluntary
compliance where City doesn't have
authority and refer to DEQ whhen unable
to achieve voluntary compliance.
Approximately 3.0 FTE
50%
75%
Permitted Industries
When the City took over the administration of the permits in 1994, 66 facilities were permitted and less
than half of them had developed the required Storm Water Pollution Control Plan (SWPCP). Since that
time, the City has identified, through inspections, facilities that are required to obtain a permit. At the
548
-------
time of this report, approximately 250 facilities were permitted. Therefore, the rate of compliance for
obtaining a permit has increased dramatically.
Inspections are performed after a review of the SWPCP and other pertinent information in the industry's
file. The City utilizes a checklist that includes all the required elements of the SWPCP. This provides a
very succinct evaluation to provide to industry. Inspections are usually scheduled in advance with the
facility operator but can be performed without notice. Inspection forms are filled out during the
inspection and any readily noticeable issues addressed during a post inspection meeting. Inspectors
provide technical assistance and information in the form of recommendations, including best management
practices (BMPs), using flyers that the City has developed. Each flyer addresses a specific BMP, such as
storage of waste materials, sandblasting, employee education, and catch basin maintenance. This allows
the City to target specific activities on site and reduces printing costs. Facilities are also evaluated for the
presence of illicit discharges. Approximately 15% of the industries had illicit discharges, primarily
washwater, identified during the initial inspection. All inspections are followed up with correspondence
outlining the findings of the inspection and expectations of the industry. Any item where the industry is
not in compliance with the permit is highlighted with a deadline to meet compliance before escalating
enforcement is pursued. It is the goal of the program to perform annual inspections, at a minimum, of all
permitted facilities.
Table No. 2 Number of Industrial Storm Water Permits Administered by the City of Portland, Oregon
1 Fiscal Year 1 94/95
| No. Permitted Facilities || 66
95/96
70
96/97
100
97/98
110
98/99
125
99/00
200
00/01
245
01/02 I
259 |
Storm water sampling of permitted facilities is performed by collecting grab samples at the sample
point(s) identified in the facility's SWPCP. Analyses are performed by the City lab and include the
parameters listed in the permit. This includes pH, total suspended solids, copper, lead, zinc, and oil and
grease. The City may also test for additional parameters that are not included in the general storm water
permit. The City's sampling does not relieve the facility from their storm water sampling responsibilities.
The results are relayed to the industry and used as a basis to assess the effectiveness of the SWPCP. The
City strives to obtain at least one sample annually, weather permitting.
For the City's situation, placing the responsibilities within a dedicated work unit has worked very well.
The work section is able to develop expertise in the area while having access to existing information from
other City programs, including the City's Pretreatment Program for discharges to the sanitary sewer.
Approximately 25% of the facilities that have storm water permits also have industrial pretreatment
permits issued by the City. There are currently five staff members that administrate the program for the
City, but approximately one-half of their time is spent conducting other activities for the City including
addressing non-storm water discharges and source investigation work for programs addressing
contaminated sediment.
Other municipalities have adopted this approach while others have incorporated the responsibility into the
pretreatment program or other existing programs including fire and safety inspections. The municipality
needs to consider several items when determining who will be responsible for implementing a program
like this, including the number and type of industries, level of oversight, and oversight of industries by
existing programs within the municipality (e.g., pretreatment, hazardous materials, etc.).
549
-------
Non-Permitted Industries
Industries are also inspected if they are identified as potentially needing a permit. There are approximately
3,000 facilities within the City that have the SIC Code listed in the federal regulations. To perform a
general survey of all facilities would have generated much more work than resources allowed. Each site
would have to be evaluated prior as the City is a mixture of combined sewers, sumps, and separated storm
sewers. Staff spends a considerable amount of time determining where the storm water from the facility
discharges to. A municipality may be able to perform a survey if the industrial base is smaller. The City
chose to prioritize the search in a systematic manner. Federal guidance states that a system-wide
approach to establishing priorities for inspections should be developed.
Initially, the City identified facilities to inspect by searching storm water outfall basins. The basins were
prioritized using criteria such as size of outfall, land use (industrialized), water quality concerns of the
receiving water, and reported pollution complaints. The basins were delineated for drainage, the
industrial facilities identified using our database, and facilities selected by SIC Code. It became readily
apparent from these inspections that for the City, certain classes of industries pose more of a pollution risk
than others. Auto wreckers, recycling facilities, and certain manufacturing facilities were identified as an
inspection priority. Certain light manufacturing, including leather products, electronic equipment,
printing, and warehousing and storage facilities posed a much lower risk as their activities are typically
indoors. Therefore, the City has adopted an approach that includes sweeps of industries based on SIC
Code. Inspections are also performed in response to referrals, field observances, complaints, and an
industrial survey performed in support of the pretreatment program. The City has identified approximately
150 additional facilities since 1994 that were required to obtain storm water permits.
In addition, investigation efforts by the City identified the Wholesale Distribution of Construction
Equipment (5082) and Heavy Construction Equipment Rental (7353) as significant sources of pollutants.
The City identified 20 of these facilities as impacting the MS4 and petitioned the DEQ to issue NPDES
General Storm Water permits. These classes are not included in the federal regulations but any municipal
program should evaluate these facilities and consider including them in their programs. The number of
inspections varies each year depending on the number of permitted industries, staff vacancies, and
requests for source investigation work. Generally, each staff member is able to inspect about the same
number of non-permitted facilities as permitted facilities. However, as the number of permitted facilities
increase, the efforts in this area will decrease.
Table No. 3 Inspection Priorities for the City of Portland, Oregon
Higher Priority SIC Codes || Lower Priority SIC Codes
5015
5093
33-
347-
7353
20-
40-41-42-
5082
Motor Vehicle Parts, Used
Scrap and Waste Materials
Primary Metals Industry
Coating, Engraving and Allied Services
Heavy Construction Equipment Rental
and Leasing
Food and Kindred Products
Transportation
Construction and Mining Machinery and
Equipment
23-
25-
27-
31-
38-
Apparel and Other Finished Products
Furniture and Fixtures
Printing, Publishing and Allied Industries
Leather and Leather Products
Measuring, Analyzing, and Controlling
Instruments; Photographic, Medical...
The City has developed several "partnerships" to expand the inspection program. Informational flyers
and a poster were developed for Multnomah County Sanitarians to use when they inspect restaurants. A
550
-------
simple storm water checklist was developed for City commercial recycling staff to use when inspecting
retail establishments. In both of these cases, it is important to note that the facilities targeted would not
ordinarily be inspected for storm water issues (unless a complaint was received), and that any follow-up
issues are then addressed by storm water staff.
Phase IINPDES Storm Water Program
The Phase n regulations did not expand on the category of industries for inclusion in the permitting
program. However, there were two significant changes that impact industry. Previously, operators of
certain facilities within category eleven (xi), commonly referred to as "light industry," were exempted
from the definition of "storm water discharge associated with industrial activity," and the subsequent
requirement to obtain an NPDES permit, provided their industrial materials or activities were not
"exposed" to storm water (EPA 2000). A light industry operator was expected to make an independent
determination of whether there was "exposure" of industrial materials and activities to storm water and, if
not, simply not submit a permit application.
As revised in the Phase n Final Rule, the conditional no exposure exclusion applies to ALL industrial
categories listed in the 1990 storm water regulations, except for construction activities (category (x)). In
addition, an operator seeking to qualify for the revised conditional no exposure exclusion, including light
industry, must submit written certification that the facility meets the definition of "no exposure" to the
NPDES permitting authority once every five years. A No-Exposure Certification (NEC) form which
contains guidance on determining whether a condition of no-exposure exists was developed by EPA
(2000) for use in those states where they are the permitting authority. The DEQ has adopted a similar
form for use in Oregon, which is a delegated state. It serves as the necessary certification provided they
are able to answer all of the questions in the negative. Regulated industrial operators need to either apply
for a permit or submit a NEC form in order to be in compliance with the NPDES storm water regulations.
The City is in the process of re-inspecting facilities that previously were not required to obtain a permit
because a condition of no exposure existed. Based on inspection results, approximately 20% of the
facilities that previously were not required to obtain a permit had exposure of industrial materials and
activities to storm water. These sites were then required to apply for a permit or remove the exposure.
The City and DEQ have agreed that any submitted no exposure certification would have to be verified
with an inspection by the City. The City is also evaluating whether certain facilities and/or sites will need
to be inspected prior to the five-year re-certification period.
The Phase n program for municipalities do not include a specific requirement for an industrial storm
water control program. However, since municipalities are ultimately responsible for discharges to their
MS4, if they have significant industries present, they should consider programs such as described here.
Illicit Discharge Elimination Program (IDEP)
The IDEP program was developed as part of the City's response to 40 CFR 122.26(d)(2)(iv)(B), which
requires the municipality to describe a program, including a schedule, to detect and remove (or require the
discharger to the municipal separate storm sewer to obtain a separate NPDES permit for) illicit discharges
and improper disposal into the storm sewer. The specific elements addressed in the City's IDEP include
conducting on-going field screening activities during the life of the permit, investigating the storm sewer
system when the results of the field screening or other appropriate information indicate a probable
551
-------
presence of illicit discharge, procedures to contain and respond to spills, and procedures to promote and
facilitate public reporting of the presence of illicit discharges.
Program Elements
Outfall Prioritization
A plan was developed to rank outfalls on the potential for the presence of pollutants found to commonly
contaminate receiving waters. Criteria included land use, pipe size, historical problems, pollution
complaints and information from outfall monitoring data (both analytical and visual). The prioritization
process made it possible for the City to utilize staff and resources in an effective manner by focusing on
the outfalls that have the highest potential for pollutant problems. From a total of over 300 storm water
outfalls, the City used the criteria to identify 109 on an Outfall Priority List. The list allows the City to
develop a reasonable schedule for Dry Weather Outfall Monitoring. After the creation of the outfall
priority list, maps of each outfall's drainage basin were created. Maps were made using existing sewer
maps, public work as-builts and field inspection records. The largest outfall basin is 475 acres while the
smallest is 15 acres. There are approximately 30,000 acres within the MS4 area.
Connection Verification
The Connection Verification Program is a methodical search and documentation of current City building
and plumbing records on connections to the MS4. The research was conducted to evaluate all connections
to the MS4 from individual property. It took two years, using staff part-time and a summer intern, to
evaluate all the properties located in the drainage area of the priority outfalls. Information collected was
reviewed looking for questionable connections to the storm sewer system (example - wash racks, trench
drains, or loading dock drains going to the storm sewer). If questions arose from a review of the records, a
site inspection was performed or referral made to the agency responsible for building inspections. The City
identified 15 businesses (out of approximately 3,000) with questionable connections. The process was very
time consuming for the results achieved. If a Phase n municipality is considering this work, they need to
understand that most illicit connections are mistakes made during the construction phase and reviewing
records does not identify these. A benefit of the creation of these records is that it provides information
when trying to identify the source of illicit discharges identified at the storm water outfall, and to industrial
storm water or similar inspection programs. In addition, once the task is completed, building plan review is
in place to address any new development.
Dry Weather Outfall Monitoring
This program has been developed to collect and analyze samples from storm water outfalls using portable
field test kits for pollutants that the EPA determined commonly contaminate storm water. This is an effort
to obtain defendable evidence of illicit connections and discharges. Monitoring and analysis are conducted
on "dry days" (>24 hours with no measurable rainfall) due to the fact that increased flows caused by
transient rainfall related storm water runoff dilute pollutant concentrations and make analytical detection and
pollutant tracing difficult. The outfall sampling schedule for any given dry day is established by the Outfall
Priority List. Outfalls are inspected/sampled at least once a month during the dry weather season (June
through September). Outfalls that have tested positive for pollutant(s) are tested more frequently during
the month.
552
-------
The analyses for commonly found storm water pollutants are performed using field meters and test strips.
Emphasis during dry weather monitoring is on looking for indicators of pollutants, instead of a long list of
individual pollutants. The City currently samples for the following pollutants: pH, temperature,
conductivity, copper, iron, residual chlorine and E Coli. All samples are analyzed in the field except for E
Coli. This has been scaled back from a much longer list than the City originally analyzed for. This was
necessary because of the excessive time required to analyze for the pollutants on the original list.
Additional pollutants may be sampled for, depending on the observed or suspected pollutants in the flow.
When pollutants are detected at concentrations that indicate the presence of illicit connections or
discharges, procedures to identify the source of the pollutant are implemented. Of the 109 storm water
outfalls monitored, approximately 40% have flow present. Many times the flow is from groundwater
infiltration or stream and ditch diversions. Of the outfalls that have flow, analyses indicate pollutants high
enough to warrant an upstream investigation approximately 25% of the time.
Pollutant Discharge Investigation
This program was developed to investigate problems identified through the Connection Verification and Dry
Weather Outfall Monitoring. If an outfall tests positive for a pollutant, an upstream investigation is
conducted to track and identify the source of the pollutant. Investigations consist of going upstream of the
outfall and checking manholes for similar flow and/or visual inspection of streets, driveways and parking
lots looking for runoff. Once the discharge is identified, the next step is to determine the severity of the
discharge and proceed accordingly.
The City has identified and corrected six illicit connections and twenty illicit discharges. Illicit connections
include wastewater from a photo processing lab, two improperly connected bathrooms, and a zamboni pit
connected to a storm sewer. Illicit discharges include discharge from a produce company, a broken City
sanitary sewer line infiltrating into the storm sewer, a commercial building with a failing septic system
leaking to the storm sewer and a steel manufacturing facility with a broken potable water line leaking into
City storm sewer. Even though outfalls are inspected monthly, illicit discharges have proven hard to
identify. This is most likely due to the number of outfalls and the intermittent nature of the discharges.
The City currently has one staff person that utilizes approximately 50% of their time performing the tasks
identified with the IDEP. The program has resulted in reduced illicit discharges overall.
Spill Protection and Citizen Response
The City has also developed a citizen complaint program to facilitate the reporting of spills and illicit
discharges. A dedicated phone number is staffed 24 hours a day. After hour reporting is recorded on
voice mail and the duty officer is paged to retrieve the information. This allows the duty officer to screen
the calls and respond accordingly. The duty officer carries a limited amount of spill materials, but works
directly with the appropriate agencies, including the City's Fire and Maintenance Bureau's, Coast Guard,
and a City contractor to provide containment and clean-up. On average, the City receives approximately
1,500 calls to the complaint line per year. Of these, nearly 50% are registered as water pollution
complaints. The remaining calls are referred to the appropriate agency and include noise and nuisance
complaints, air quality concerns, etc. Approximately 300 (20%) of the calls come after normal working
hours with 25% of these requiring an on scene response either immediately or the next day. The City
staffs the position with one full time employee for regular business hours, and utilizes staff on a rotating
basis from the Industrial Storm Water, Industrial Pretreatment, and IDEP for after-hours response.
553
-------
Conclusions
The development of an industrial stormwater program is not one of the six BMPs that Phase n permit
holders will be required to be developed. This may be due, in part, to the assumption that all industrial
permits would be in place because of Phase I requirements. However, our efforts have shown that only
25-30% of the industries requiring permits had applied prior to the administration of the program by the
City.
A municipality may become co-applicants with Phase 1 permit holders. If this occurs, the applicant will
become subject to an industrial control program but may be able to utilize the existing program of the
permit holder. If a municipality does not develop a program, it is recommended that they work with the
permitting authority to identify who has a permit and the status of their compliance. The municipality
should also evaluate the industrial base in the MS4 and provide this information to the permitting
authority if they identify a facility that may be subject to the program. It may be prudent to incorporate
these activities into the illicit discharge elimination program, which is a requirement of the Phase n
permit. Whatever the municipality chooses, they need to understand that they are ultimately responsible
for discharges from their MS4.
Work to date in the implementation of the IDEP has shown that researching building and plumbing
records of facilities was a very time intensive use of resources with very little benefit in identifying illicit
connections. Most illicit connections are the result of in-field errors in connections during construction.
Time would be better spent conducting dry weather monitoring to identify illicit discharges, although
identifying them can be difficult due to the intermittent nature of the discharge. In addition, some illicit
discharges may be low in volume and never reach the storm water outfall. These pollutants would then be
discharged with the next storm event. Based on this, it may be necessary to move the inspection program
upstream in the collection system. However, this would dramatically increase the points of inspection.
An alternative would be to monitor storm water quality at the outfall and identify where there are water
quality concerns. Upstream inspections of facilities could then be used to identify illicit discharges. The
City's Industrial Storm Water program has identified illicit discharges in approximately 15% of their
facility inspections.
References
EPA 2000. Guidance Manual for Conditional Exclusion from Storm Water Permitting Based On "No
Exposure" of Industrial Activities to Storm Water, U.S. EPA, Office of Water, EPA 833-B-00-001.
554
-------
A Reassessment of the Expanded EPA/ASCE National BMP Database
Eric W. Strecker, P.E. *, Marcus M. Quigley, P.E. **, and Ben Urbonas, P.E. ***
* Principal, GeoSyntec Consultants, 838 SW First Avenue, Suite 430, Portland, OR 97204; PH 503-222-
9518: estreckerfSlgeosvntec.com
** Project Engineer, GeoSyntec Consultants, 629 Massachusetts Avenue, Boxborough, MA 01719; PH 978-
263-9588; mquigelv@geosvntec.com
*** Chief, Master Planning and South Platte River Programs, Urban Drainage and Flood Control District, 2480
West 26th Avenue, Suite 156-B, Denver, CO 80211, PH 303-455-6277; burbonas@udfcd.org
ABSTRACT
The USEPA/ASCE National BMP Database has grown significantly since the first evaluation
of BMP performance data in the database was completed in 2000. The project team is
currently performing a re-evaluation of the data contained in the database to assess the
overall performance of BMPs as well as compare BMP design attributes to performance.
Although this analysis has not been fully completed, several initial results are presented in
this paper.
The evaluations include the assessment of various BMP types as categorized in the database
with regards to their ability to reduce runoff volumes as well as improve effluent quality.
Certain BMP types may reduce the volume of runoff through evapotranspiration and/or
infiltration, as opposed to BMPs that are more "sealed," such as wet ponds, wetlands, and
vaults. Runoff reductions directly reduce pollutant loading as does improved effluent
quality. On average, dry detention basins were found to reduce runoff volumes by an
average of 30% (comparison of inflow to outflow), while biofilters reduced volumes by
almost 40%. As expected, wet ponds, wetlands, and hydrodynamic devices, and retention
ponds show little or no runoff volume reductions. BMP types vary with regards to effluent
quality that is achieved. BMPs such as wet ponds and wetlands appear to achieve lower
concentrations in effluent quality than other BMPs such as detention ponds (dry) and
hydrodynamic devices. These differences vary with pollutant type. With more data
available, analyses of BMP design versus performance show statistically valid results. For
example, a relationship (ratio) between the treatment volume of retention ponds (with wet
pools) versus the average size storm event volume monitored has been established, showing
that those with a ratio of 1 or greater have been observed to achieve significantly belter
effluent quality.
This paper also briefly overviews the Urban Stormwater BMP Performance Monitoring
("Manual") (Strecker, et. al., 2002) that was 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.
555
-------
INTRODUCTION
The USEPA (Environmental Protection Agency)/ASCE (American Society of Civil
Engineers) National Stormwater BMP (Best Management Practice) Database has been under
development since 1994, under a USEPA grant project with the Urban Water Resources
Research Council (UWRRC) of ASCE (Urbonas, 1994). The project has included the
development of recommended protocols for BMP performance (Urbonas, 1994 and Strecker
1994), a compilation of existing BMP information and loading of suitable data into a
specially designed database (www.bmpdatabase.org), and an initial assessment of the results
of the analyses of the database (Strecker et. al., 2001). In addition a detailed guidance
document on BMP monitoring has been developed, entitled "Urban Stormwater BMP
Performance Monitoring: A Guidance Manual for Meeting the National Stormwater BMP
Database Requirements" (available for download at: www.bmpdatabase.org).
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 has been 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 varying
reporting protocols make wide-scale assessments difficult, if not impossible. (Strecker et al.
2001; Urbonas 1994) 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 wide ranges of
BMP "efficiency" (typically percentage removal) that has been reported in literature to date.
Municipal separate storm sewer system owners and operators, industries, and transportation
agencies 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 designs; and, within a specific
category of BMPs, to what degree design and environmental static and state variables
directly affect BMP performance. The protocols developed under this project and the Urban
Stormwater BMP Performance Monitoring 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.
One of the major findings of the EPA/ASCE BMP Database efforts to date has been that
BMP pollutant removal performance for most pollutants is believed best assessed by the
following: (Strecker et. al., 2001):
556
-------
• How much stormwater runoff is prevented? (Hydrological Source Control)
• How much of the runoff that occurs is treated by the BMP or not?
• Of the runoff treated, what is the effluent quality?
For some pollutants, the amount of material captured could also be important, as well as how
the BMP mitigates temperature and/or flow changes. Percent removal of pollutants is a
highly problematic method for assessing performance and has resulted in some significant
errors in BMP performance reporting (Strecker, et. al., 2001).
Urban Stormwater BMP Performance Monitoring: A Guidance Manual for Meeting the
National Stormwater BMP Database Requirements (available for download at:
www.bmpdatabase.org) is intended to improve the state of the practice by providing
recommended methods for meeting the EPA/ASCE BMP Database protocols and standards
(Urbonas 1994) 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. Furthermore, it provides insight into and guidance for strategies, approaches, and
techniques that are appropriate and useful for monitoring BMPs. The overall focus of the
document is on the collection, reporting, and analysis of water quantity and quality
measurements for quantitative BMP performance studies. 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 performance under some
circumstances.
RE-EVALUATION OF THE NATIONAL BMP DATABASE
The project team is completing a detailed assessment of the expanded database. Table 1
presents an overview of the BMPs currently in the database, including the number of data
records for each BMP type. New BMP information is being provided to the database team at
about a rate of 15 to 20 studies per year. These are studies that meet the protocols established
for BMP monitoring and reporting. The 170 studies now in the database compares with the
total of just over 60 BMP studies in the database during the initial evaluation.
Each study has again been analyzed in a consistent manner as described in Strecker, et. al.
2001) and on the project web site. The data being produced includes lognormal distribution
based summary statistics, comparisons of influent and effluent water quality through
parametric and non-parametric hypothesis tests, and a large number of other summary
statistics. In this evaluation, the project team has been investigating the effects of BMPs on
hydrology and effluent quality. The project team is currently working on evaluation of the
design attributes versus BMP performance, which will be highlighted in more detailed at in
the presentation.
557
-------
Table 1. Number of BMPs and Data Records (events or event mean concentrations) in
the National BMP Database as of 11/01/02
BMP Type
Detention Basins
Grass Filter Strips
Media Filter
Porous Pavement
Retention Pond
Percolation Trench and
Dry Well
Wetland Channel and
Swale
Wetland Basin
Hydrodynamic Devices
Total
# of BMPs
in Category
with Design
Information
24
32
30
5
33
1
14
15
16
170
Precipitation
Records for BMP
type
129
227
187
5
378
3
53
221
169
1372
Flow Records for
BMP Type
229
385
327
5
817
3
113
681
309
2,869
Water Quality
Records for BMP
Type
4209
6,251
6,144
55
14,293
21
1,241
7,320
6,186
45,720
Hydrology Evaluation
One of the goals of the data base was to provide better information on the effects of BMPs on
hydrology and whether some BMPs may have some benefits over others in terms of reducing
volume of runoff (Hydrological Source Control-HSC). For example, one would expect that a
wet pond might not significantly decrease the volume of runoff, but a biofilter might, given
the contact with more frequently drier soils and resulting evapotranspiration and/or
infiltration. Accurately measuring flow during storm conditions is very difficult (EPA,
2002). In a field test of over 20 different flow measurement technologies and approaches,
FHWA (2001) found that flow measurements can be upwards of 50% or more off of the
expected true flow. Therefore assessments of the database will likely show some variability
in flow changes due to measurement errors.
Figure 1 presents plots of inflow versus outflow for Biofilters (Swales and filter strips),
Detention Basins (dry ponds), Retention Ponds (wet ponds) and Wetland Basins. Biofilters
showed an average of 20% less volume of runoff on a storm-by-storm basis and were
consistently lower for almost all storm events. The other BMPs showed a large scatter, but
generally showed an increase in runoff volumes. While showing an increase on a storm-by-
storm basis, dry ponds tended to have many more storms that were lower in outflow.
558
-------
Table 2, presents the results of removing the smaller more insignificant storms from the
analyses (storms resulting in flows less than 0.2 watershed inches removed). The term
"watershed inches" refers to an area-normalized volume (the total volume divided by the
total watershed area). From these analyses, it is apparent that detention basins (dry ponds)
and biofilters (vegetated swales, overland flow, etc.) appear to contribute significantly to
volume reductions, even though they were likely not specifically designed to do so. One
needs to note that although in our protocols we ask for the total storm volume of the influent
and effluent over the entire event, it is possible that some studies may have cut-off effluent
sampling before the BMP returned to pre-storm conditions. Based upon the recommended
criteria above for assessing BMP performance, it appears that there is a basis for factoring in
volume and resulting pollutant load reductions into BMP performance. This has significant
implications for Total Maximum Daily Loads (TMDLs) implementation planning and other
stormwater management planning. It is also expected that as BMPs that are specifically
designed to reduce runoff volumes (e.g., lower impact development, etc.) are tested and
information added into the database, that these results will improve.
Water Quality Performance
The analysis of water quality performance data of the BMPs that we are being conducted by
the authors performing is comprised of three levels: 1) a comprehensive evaluation of
effluent versus influent water quality; 2), comparisons of effluent quality amongst BMP
types; and 3) comparisons of performance versus design attributes for BMP types and
individual BMPs. Even with the increase in data in the database since the last evaluation, the
total number of BMPs in any one category is still small as compared to the number of design
parameters that can be potentially investigated. The approach that the team has taken is to
develop groupings of BMPs by Design Factors. That is, our approach has been to develop
categories of design parameters that are expected to affect performance, group BMPs into
those that meet all or most of the factors (e.g., length to width ratios; volume of facility as
compared to average storm inflow, etc.) and then explore if a difference in performance can
be established and potentially explained by these assessments of these grouped design
factors.
Figure 2 presents plots shows a box plot of the fractions of reported Total Suspended Solids
(TSS) concentrations removed and the box plots of effluent quality of BMP types. As has
been found previously (Strecker et. al, 2001), the effluent quality is much less variable than
fraction removed. It appears that percent removal is more or less just a function of inflow
concentration Recent analysis of the expanded database shows that effluent quality can be
assumed to be different among different BMP types. It appears that Retention Ponds (wet
ponds) and Wetlands can achieve lower concentrations of TSS than other BMPs, while
hydrodynamic devices were the lowest performers (higher effluent concentrations) on
average. Similar results have been found for other constituents with some variations. One
should note (discussed below) that there are serious questions regarding the validity of TSS
as an accurate measure of suspended solids. However, the problems with TSS methods are
likely not large with effluent quality as most of the potentially missed larger fractions would
likely have been removed if the BMP is "working" at all.
559
-------
Figure 3 shows the result for comparing Total Phosphorus and Total Copper concentrations
for the same BMP categories. Wetlands and wet ponds are more consistent performers,
while the other BMPs vary with regards to effluent quality results. The lowest effluent
quality achieved for Phosphorus is on the order of 50 to 60 ug/1. This contrasts with some
water quality efforts where the ultimate phosphorus goal has been selected to be in the range
of 10 to 20 ug/1 and then showing achievement of such goals by misapplication of percent
removal approaches.
Biofilters(N=16)
(Swale and Filter Strips)
'%.0 0.3 0.6 0.9 1.2 1.5
Inflow (watershed inches)
Retention Ponds (N=20)
(Wet Ponds)
o
c
T3
-—
CO
o
0
Average Ratio
(Out/In) = 1.12 ,
0.
10 0.3 0.6 0.9 1.2 1.5
Inflow (watershed inches)
Detention Basins (N=11)
(Dry Ponds)
Average Ratio
(Out/In) = 1.12
•°0.0 0.3 0.6 0.9 1.2 1.5
Inflow (watershed inches)
Wetland Basins (N=10)
Average Ratio
(Out/In) = 1.34
0.0 0.3 0.6 0.9 1.2 1.5
Inflow (watershed inches)
Figure 1. Comparison of Individual Storm Inflow and Outflow Volumes for Indicated
BMPs (N= number of BMPs included; n= number of storm events)
560
-------
As mentioned above, we are exploring individual BMP designs (sizing, etc.) relative to
performance. Some initial results of the expanded database have been encouraging. For
example, the previous effort during the initial work was not able to statistically find a
potential relationship between performance of retention ponds and wetlands and their
treatment volume relative to measured storm events. Figure 4 shows a scatter plot of
Retention Ponds (with a permanent pool) effluent quality versus the ratio of the treatment
volume to mean monitored storm event volume, and a box plot of Retention Pond mean
effluent quality for sites with ratio less than one and greater than one ratio of the treatment
volume to mean monitored storm event volume. The plots clearly demonstrate that at those
sites where the treatment volume was greater than the average size storm event monitored,
the effluent quality was significantly lower. In addition, the variability of effluent quality for
the larger retention ponds was lower. These results are expected, but it is one of the first
times that they have been demonstrated statistically.
Table 2. Ratio of Mean Monitored Storm Event Outflow to Inflow for Storms Greater
than 0.2 watershed inches.
BMP Type
Detention Basins
Biofilters
Media Filters
Hydrodynamic
Devices
Wetland Basins
Retention Ponds
Wetland Channels
Mean Monitored Outflow/Mean Monitored
Inflow for Events Where Inflow is Greater Than
or Equal to 0.2 Watershed Inches
0.70
0.62
1.00
1.00
0.95
0.93
1.00
Some of the other assessments that are being preformed are the potential reductions in
toxicity of heavy metals by BMPs. More recent BMP studies have been collecting data on
water hardness and therefore there is the ability to assess potential toxicity issues via
comparisons of effluent quality with EPA acute and chronic criteria values (as benchmarks as
the criteria apply in receiving waters). One trend that we have noticed in the data is that for
many BMPs, hardness levels are increased in effluent versus the influent and therefore this
could contribute along with concentration reductions to reduce toxicity (as defined by EPA's
Acute Criteria for Aquatic Life). We will also be looking at the effects of BMPs on load
561
-------
reductions considering both hydrological source control performance as well as effluent
quality.
Fraction of TSS Removed
o o o o o -
o KJ j^ b> bo c
ff
-S -j
M
J.....L
1 1
Biosw
Detention
Basins
ales Me
Filt
1
Hydro -
Dynamic
Devices
dia Rete
=rs Bas
—
1 i
Wetlands
ition
ns
100.00
~O)
•§ 10.00
c/)
c/)
H 1.00
0.10
:
: *
Detention
Basins
Hyd
Bioswales DYna
Devic
1
f :
Z
:
1
Retention
Basins
ro Wet
mic
es
ands
BMP Type
BMP Type
Figure 2. Box plots of the fractions of Total Suspended Solids (TSS) removed and of
effluent quality of selected BMP types
10.0
to
o
.c
Q.
W
O
Q_
1.0
0.10
£ 0.010
0.001
A r
I I I
I
100.0
BioswaleOetentionMedia Hydro RetentiorWetlands
Basins Filters Dynamic Basins
Devices
O)
CD^ 10.0
Q.
O
O
1.0
0.1
BMP Type
\ \
I I I I
Bioswales Detention Media Hydro- Retention Wetland;
Basins Filters Dynamic Basins
Devices
BMP Type
Figure 3. Box plots of effluent quality of selected BMP types for Total Phosphorus and
Total Copper.
562
-------
50
G"
^? 40
E
s" — "'
w
^ 30
c
0
2 20
-i— •
CD
O
§ 10
0
"E
o
-
o
F.O
'o
- —
o
-DO
o
o 0
o
~° ° 0
0
i i n i i i i i i i
^ 0123456789 10
Permanent Pool Volume (watershed meters)
Mean Monitored Effluent Volume (watershed meters)
50
~
> 40
£,
CO
CO
^ 30
o
IE
1 20
0
o
O
| 10
^
LU
n
k
K
)
/
/
L
.
i
J\
T
\
\
_J
-
.
J
\
\
-
-
•
-,
/
(
* / \ ~
u
<
1 >
1
Permanent Pool Volume (watershed meters)
Mean Monitored Effluent Volume (watershed meters)
Figure 4. Scatter plot of 1) Retention Pond (with permanent wet pool) TSS effluent
quality versus the ratio of the permanent pool volume to mean monitored
effluent volume and 2) Box plots of the TSS effluent quality of sites grouped
by a ratio of less than or greater than 1 for the ratio of the permanent pool
volume to mean monitored effluent volume. (Note: watershed meters are
calculated by dividing the volume by the total watershed area)
AN OVERVIEW OF THE URBAN STORMWATER BMP PERFORMANCE
MONITORING
The Manual contains two main sections following the introduction:
Overview of BMP monitoring. A detailed discussion is provided on the context of BMP
monitoring, difficulties in assessing BMP performance, and understanding the 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
BMP performance 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. A recommended approach focusing on effluent quality and the amount of
runoff treated (and not) is specified.
Developing and Implementing a Monitoring Program. This section provides specifics on
how to develop a monitoring program, including selecting monitoring methods and
equipment, installing and using equipment, implementing sampling approaches and
563
-------
techniques, and reporting information consistent with the National Stormwater Best
Management Practices Database.
Supporting Materials. In addition, four appendices that focus on statistical methods for
improving BMP monitoring studies and data reporting have been included in the guidance
document. The first appendix describes detailed methods for estimating potential errors in
field measurements. The second provides detailed information about the estimating the
number of samples expected to be necessary 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 (e.g., inlet and outlet water
quality) 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.
Understanding Variability and Sources of Error in BMP Performance Monitoring
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 statistically valid assessments of water quality. The Manual
provides direct and applicable guidance on approaches to integrating quantitative evaluations
of potential sample results variability to improve attainment of study goals via the collection
of adequate data. As the National Stormwater Best Management Practices Database is
founded on the quantitative assessment of water quality performance of BMPs, the Manual
focuses on providing practitioners with firm statistical footing for study design and
implementation within that context. Specifically the manual focuses on the four factors that
influence the probability of identifying a significant temporal and/or spatial changes in water
quality, including:
1) Overall variability in BMP influent and effluent water quality data.
2) Minimum detectable change in water quality (difference in the mean and variability of
concentrations).
3) Number of influent and effluent samples collected.
4) Desired confidence level from which to draw conclusions.
The manual recommends that statistical analyses should be conducted to estimate how many
events need to be monitored to achieve a specified level of confidence in a desired
conclusion (i.e., power analysis). Performing a power analysis requires that the magnitude of
acceptable error in effluent quality and/or detectable change in pollutant concentration, the
confidence level, the estimated variability of future samples collected and the statistical power
or probability of detecting a difference are defined or can be estimated. A complete set of
nomographs provided by Pitt (2001) were included in the Manual.
In addition to drawing attention to the need to better integrate improved understanding of the
inherent variability found in water quality data, the authors would like to emphasize the
564
-------
importance of collecting accurate flow data. Flow measurement data is often one of the most
often overlooked sources of error and variability in BMP monitoring studies. In nearly all
studies involving assessments of water quality, flow is used as a primary factor underlying all
collected data. Not only are flow measurements used directly to calculate loads and event mean
concentrations (depending on approach take), flows are often used to pace samplers for
collection of flow-weighted samples. They are also used in an attempt to understand watershed
hydrology and effects of BMPs on flow reduction and/or attenuation. Very few studies look
quantitatively at the likely errors introduced into BMP performance studies as a result of flow
measurement errors. 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, improperly installed or selected equipment, or poor maintenance.
However, even under ideal conditions, errors in flow measurement can be significant.
Quantitative analyses should be conducted to determine the likely errors associated with lower
flow rates that in many climates result in the majority of total runoff volumes. Flow equipment
should be designed to accurately quantify flows that may be orders of magnitude above and
below the mean flow rate. This is particularly the case for very small watersheds (less than an
acre) which have extremely peaky flows and are receiving increased monitoring attention with
the growing installation of "in watershed" controls. Many flumes and depth measurement
approaches which work for large watersheds do not function well when the flow rates rapidly
vary by more than three orders of magnitude with extremely low flows occurring during light
rainfall periods. It is recommended that primary devices be used where possible and their
selection be made carefully with full knowledge of the magnitude of likely errors associated
with the selection. For example h cases in wmch there is a need for measurement of extreme
flow ranges and a free overflow (no backwater conditions exist down stream) is available, the
H, HS, or HL flumes should be considered. The range of flows that can be measured
relatively 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. H flumes are also not prone to issues associated with sediment build-up and are
relatively unaffected by upstream turbulence.
Weirs are generally recognized as more accurate than flumes (Grant and Dawson 1997). A
properly installed weir can typically achieve accuracies within 2 to 5% of the actual 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, sediment accumulation within
storms, 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 ±5 to 10% or more (Grant and Dawson 1997). Circumstances
present in many stormwater monitoring locations can result in errors well in excess of 100%.
565
-------
There is a potential that certain BMPs could be more difficult to monitor accurately, as well
as the outflow of some BMPs (those with significant storage) may be less peaky and
therefore easier to measure. These both could affect the Qout/Qin (Table 2) results.
Other Sources of Error
A number of other sources of error are important to obtaining and reporting monitoring
program data effectively. These errors should 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 (or transferred from an
electronic data deliverable-EDD), typically with separate records for each monitoring station
and each storm event. Inconsistencies 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. Newly emerging tools for
field data collection and observation such as personal digital assistant (PDA) deployed
databases, which close the "paper gap" in collecting field data hold promise for decreasing
some of the sources of these types of errors.
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 the sources of error described above, 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. As discussed
previously in this paper, 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.
Data Analysis Methods
The ASCE/EPA project team reviewed available methodologies for data analysis as part of the
publication of the first comprehensive analysis of data stored in the National Stormwater Best
566
-------
Management Practices Database (available on the project website at www.bmpdatabase.org)
and continues to look at more recent methods that have been proposed which are being used to
re-evaluate the much more complete data set now available in the Database. In the manual, the
authors recommend an effluent focused approach to efficiency evaluations labeled the Effluent
Probability Method.
The Effluent Probability Method quantifies BMP efficiency in two steps. The first of these
steps is to determine if the BMP is providing treatment (that the influent and effluent mean
EMCs are statistically different from one another). The second step then focuses in on an
examination of either a cumulative distribution function of influent and effluent quality or a
standard parallel probability plot (essentially the same information in two different formats).
It is recommended that before any 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 and 80% power).
The Effluent Probability Method is straightforward and directly provides a clear picture of
one of the ultimate measures of BMP effectiveness, effluent water quality. Curves of this
type may be the single most instructive piece of information that can result from a BMP
evaluation study. Although an exact format has yet to be agreed upon, the authors of this paper
strongly recommend that the stormwater industry accept this approach as a standard "rating
curve" for BMP evaluation studies. An example in the recommended format is shown in Figure
5, alternately the y axis can include "percent less than" instead of the expected value of the
standardized normal distribution. It is critical that the BMP study also report on how much of
the runoff is actually treated versus bypassed as well as infiltrated or evapotranspirated as
appropriate for some BMPs. This is the hydraulic performance of the BMP and effects
evaluation of the effectiveness of various BMP sizes.
The Urban Water Resources Research Council and the Co-Principal Investigators for the
ASCE/EPA National Stormwater Best Management Practices Database at the time of the
writing of the paper are in the process of recommending a final format or standard "cut sheet"
that will be recommended for inclusion in any BMP monitoring study to clearly and succinctly
provide vital information to practitioners on the performance of a particular BMP. This
standard "cut sheet" will be posted on the project website (www.bmpbatabase.org) both in
generic format with guidelines for use and will be created for each BMP study that is included
in the National Database.
Selecting Parameters
Stormwater runoff may contain a variety of substances that can adversely affect the
beneficial uses of receiving water bodies. The Manual recommends that the following
factors are important to examine when selecting parameters to be included in a BMP
monitoring program:
567
-------
Permit requirements (if any). Monitoring to comply with a permit may specify the
parameters that must be measured in stormwater discharges. However, monitoring for
additional parameters may help attain overall program objectives.
Q
CO
Figure 5.
- Log Mean (Log Arith. Median)
, I ,/VV , I , , , , I , , , , I , , , ,
0.01
1 10
EMC (mg/L)
100
Inflow
Outflow
Normal Distribution
- — 95% CL
1000
Example Normal Probability Plot Recommended for Inclusion in All BMP
Monitoring Studies as Part of the Effluent Probability Method.
• Land uses in the catchment area. Land use is a major factor affecting stormwater quality.
Developing a list of the pollutants commonly associated with various land uses is helpful
for deciding what to look for when monitoring.
• Existing monitoring data (if any) for the catchment area. Previous monitoring data can
be helpful in refining the parameter list and developing estimates of the potential
variability of the BMP influent data. However, if there is uncertainty about the
monitoring methods and/or analytical data quality, or if the existing data pertain to
baseflow conditions or only one or two storms, caution should be used in ruling out
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 Astern 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.),
568
-------
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 below. Refer to Strecker (1994) and Urbonas (1994)
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 of
parameters to monitor for a given BMP evaluation, it is important to consider the upstream land
uses and activities.
The base list represents a basic set 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).
Dissolved versus 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, along with total
metals. If selenium or mercury is of concern, total concentrations should be measured to
enable comparison with criteria based on bioaccumulation by organisms.
569
-------
Table 3: Typical urban stormwater runoff constituents and recommended detection limits
Parameter Units Target Detection Limit
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
pH
mg/L
mg/L
mg/L
mg/L
MPN/lOOml
MPN/lOOml
MPN/lOOml
mg/L
mg/L
mg/L
mg/L
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)
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.
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
and organic matter 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 and
organic matter that acts as adsorption sites for pollutants to attach to. It is difficult to
570
-------
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.
Hardness values are typically higher in hardened conveyance systems that in receiving waters
or earthen channels.
If loads to the receiving waters are of concern (e.g., discharge to a lake known to be a water
quality limited water body) than analyzing for total recoverable metals is particularly
recommended. Finally, total recoverable metals data together with dissolved metals data can
be used to assess potential metals sediment issues.
Measurements of Sediment Concentration
A variety of methods have been employed in stormwater quality studies for quantifying
sediment concentrations. The most frequently cited parameter is "TSS" or total suspended
solids. The "TSS" label is used, however, to lefer 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.
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 can 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 very 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 h 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).
The authors recommend that both TSS (for comparison to existing data sets) and SSC be
measured for BMP monitoring studies. The difference between TSS and SSC in samples
from BMPs that are even mildly performing should be minimal (e.g., if the BMP is
functioning at all then the sands and larger particles should be removed. Therefore, assessing
effluent data from past BMP performance studies, rather than percent removal eliminates, is
likely to be a much more valid approach.
571
-------
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 even within a given technique and include (USGS 2001):
Dry sieve.
Wet sieve.
Visual accumulation tube (VA).
Bottom withdrawal tube.
Pipet.
Mcroscopy.
Coulter counter.
Sedigraph (x-ray sedimentation).
Brinkman particle size analyzer.
Laser diffraction spectroscopy.
Light-based image analysis
At this time the authors recommend selecting and using a consistent and appropriate method
from the above (i.e., no single method has been established as the standard).
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. For example the depth at which the sample was collected can
significantly impact results. Again, the impacts would be much greater on influent data
rather than effluent data due to the fact the BMP should be removing the larger particles.
CONCLUSIONS
An evolving tool is available to practitioners who are assessing the performance of BMPs via
the National Stormwater Best Management Practices Database Project. Practitioners can
perform their own evaluations by downloading information from the web site.
Results of the analyses of the now expanded database have reinforced the initial finding that
BMPs are best described by how much they reduce runoff volumes, how much of the runoff
that occurs is treated (and not) by the BMP, and of the runoff treated what effluent quality
(concentrations and potential toxicity) is achieved. These basic BMP performance
descriptions can then be utilized to assess effects on total loadings, frequency of potential
exceedances of water quality criteria or other targets, and other desired water quality
performance measures. The results show that the effluent quality of various BMP types can
572
-------
be statistically characterized as being different from one another. Additionally, some design
parameters may be statistically significant with regards to performance.
A new guidance tool is available to practitioners who are conducting BMP monitoring
studies and wish to comply with the standards established as part of the National Stormwater
Best Management Practices Database Project. The Manual contains a comprehensive and
practical discussion on all elements of water quality, flow, and precipitation monitoring and
discusses them within the specific framework of the National Database.
ACKNOWLEDGEMENTS
The project team would like to thank Jesse Pritts, P.E. and Eric Strassler of the
Environmental Protection Agency for their support for and participation in the ASCE/EPA
National Stormwater Best Management Practices Database Project. The authors would also
like to thank the following members of ASCE's Urban Water Resources Research Council
for their thorough reviews and contributions to the Database Project: 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), and Tom Langan (Wright Water Engineers). The initial database evaluation
findings discussed in this paper are those found by the authors only and not the listed
individuals above.
REFERENCES
American Society for Testing and Materials. (2000). Standard Test Method for Determining
Sediment Concentration in Water Samples. ASTM Designation D 3977-97, 395-400.
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.
EPA (2002). Urban Stormwater BMP Performance Monitoring: A Guidance Manual for
Meeting the National Stormwater BMP Database Requirements. EPA 821-C-02-005. U.S.
Environmental Protection Agency, Office of Water, Washington, D.C. April.
FFIWA. (2001). Guidance Manual for Monitoring Highway Runoff Water Quality. FFIWA-
EP-01-022. Federal Ffighway Administration, U.S. Department of Transportation. Federal
Ffighway Administration. Washington, D.C.
Grant, D.M. and B.D. Dawson. (1997). ISCO Open Channel Flow Measurement Handbook.
ISCO Environmental Division. Lincoln, NE.
573
-------
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.
National Urban Runoff Program. (1983). Final Report. U.S. Environmental Protection
Agency Water Planning Division, Washington, D.C.
Pitt, R. (2001). Personal Correspondence. Peer review meeting, August.
Spitzer, DW (ed.) (1996). Flow Measurement Practical Guides for Measurement and Control.
Instrument Society of America. NC.
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.
Strecker, EW., J.M. Kersnar, E.D. Driscoll, and R.R. Horner. (1992). The use of wetlands
for controlling storm water pollution. The Terrene Institute, 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.
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.
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.
574
-------
EPA's Management Measures Guidance
to Control Nonpoint Source Pollution
from Urban Areas
It's Time to Develop and Implement Your Storm Water
Management Program...Are You Ready?
Rod Frederick and Robert Goo
U.S. Environmental Protection Agency
Nonpoint Source Control Branch (4503F)
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Martina Keefe
Tetra Tech, Inc.
10306 Eaton Place, Suite 340
Fairfax, VA 22030
Poster Session for Urban Storm Water:
Enhancing Programs at the Local Level
February 17-20, 2003
The Westin Michigan Avenue
Chicago, Illinois
-------
EPA's Management Measures Guidance to Control
Nonpoint Source Pollution from Urban Areas
It's Time to Develop and Implement Your Storm Water
Management Program...Are You Ready?
Introduction
Urban runoff/storm sewers were listed among the top
three sources of water quality impairment in rivers,
lakes, and estuaries, according to the National Water
Quality Inventory: 1998 Report to Congress (USEPA,
2000). This indicates that urban areas have been a
substantial contributor to the decline of water
resources in the U.S. As population continues to grow
and urban areas expand (see Figure 1), the quality of
water bodies near urban centers will continue to be
threatened unless actions are taken to reduce the
impact of everyday human activities on water
resources.
This is not just an issue of pollutant loading, although
urban areas can be a significant source of several
pollutants, especially nutrients, sediments, heavy
metals, and toxic chemicals. Also of concern are the
increase in the volume of runoff and the change in
runoff timing that results when land in a
predominantly pervious condition (i.e., forested or
meadow) is converted to impervious surfaces—
buildings, streets, sidewalks, parking lots, or other
infrastructure.
. Census 2000
B Urban area
Urban cluster
| | State boundary
Figure 1. Urban areas and urban clusters according to the 2000
U.S. Census (USCB, 2002)
The complicating factor in mitigating urban storm
water is that the sources of pollution are diffuse and
are therefore difficult to locate and manage. For
example, nutrient pollution in urban areas can come
from a variety of sources that include failing septic
systems, improper connections to the storm drain
system, overfertilization of lawns, and poorly
managed pet waste. Each source can require a
different strategy for elimination, which can seem
overwhelming to small programs faced with pollution
problems.
Because managing urban storm water is not a simple
task, EPA has developed guidance to help watershed
managers put together a comprehensive and effective
program to address a myriad of urban sources. The
most recent guidance, called National Management
Measures to Control Nonpoint Source Pollution from
Urban Areas—Draft, is an update of Chapter 4 of the
1993 Guidance Specifying Management Measures for
Sources of Nonpoint Pollution in Coastal Waters.
The 1993 document was designed to aid coastal states
in developing nonpoint source control programs to
meet the requirements of the Coastal Zone Act
Reauthorization Amendments of
1990. The 2002 guidance document
is intended to provide technical
assistance to state and local program
managers and other practitioners on
the best available, most economically
achievable means of managing urban
storm water. It describes how to
develop a "comprehensive runoff
management program" that deals
with all phases of development—
from predevelopment watershed
planning and site design, through the
construction phase of development,
to the operation and maintenance of
structural controls. It also provides
information for other situations such
as retrofitting existing development,
implementing nonstructural controls,
-------
and reevaluating the storm water management
program (see Figure 2).
How Does This Relate to NPDES Phase II
Storm Water?
The publication of the guidance is timely because
thousands of small municipalities (with a
population between 1,000 and 10,000 with a
population density of at least 1,000 people per
square mile) and other entities (e.g., private
institutions, Department of Defense facilities) that
own and operate separate storm sewer systems will
need to apply for a permit to discharge municipal
storm water under Phase II of the National
Pollutant Discharge Elimination System (NPDES)
Storm Water Program. NPDES permit coverage
must be obtained by March 10, 2003. To meet the
Program
Evaluation
Operation &
Maintenance
Existing
Development
Pollution
Prevention
Construction
Sites
Program
Framework
& Objectives
Comprehensive
Runoff
Management
Framework
Bridges &
Highways
Watershed
Assessment
Watershed
Protection
Site
Development
New
Development
Onsite Wastewater
Treatment Systems
Figure 2. Comprehensive Runoff Management Program
Box 1. NPDES Storm Water Phase II
Program Requirements
Regulated municipalities must develop and
implement a Storm Water Management
Program (SWMP) that will reduce pollutants in
storm water to the Maximum Extent Practicable
(MEP). The SWMP must include BMPs for
each of the 6 Minimum Control Measures,
which are:
1. Public Education and Outreach on Storm
Water Impacts
2. Public Involvement/Participation
3. Illicit Discharge Detection and
Elimination
5.
4. Construction Site Runoff Control
Post-Construction Storm Water
Management in New Development and
Redevelopment
Pollution Prevention/Good
6.
Housekeeping for Municipal Operations
In addition to BMPs, regulated municipalities
will also have to develop Measurable Goals that
will allow both the municipality and the
permitting authority to gauge whether each BMP
was successful. Municipalities also need to
develop a timeline for implementation of each
element of the program and identify the party or
parties responsible.
requirements outlined in the Phase II Storm Water
Rule (Box 1), regulated municipalities must
implement a storm water management program that
includes best management practices (BMPs) and
measurable goals for six minimum control measures.
How Can Your Storm Water Management
Program Be Both Comprehensive and
Cost-Effective?
The National Management Measures to Control
Nonpoint Source Pollution from Urban Areas—Draft
presents a comprehensive process for developing a
program from scratch or from existing programs. The
guidance includes information about establishing
institutional frameworks, securing funding sources,
conducting assessments, working with stakeholders,
and implementing structural and non-structural BMPs.
The process is presented in a stepwise fashion that is
organized by management measures, which each
cover a distinct topic area such as roads and highways,
construction sites, pollution prevention, etc. The
management measures provide a framework for
grouping BMPs based on their role in mitigating the
effects of urban runoff. Storm water managers can
use this organizing framework to ensure that their
program addresses the entire range of pollutants and
sources with a set of BMPs that work together in a
streamlined, cost-effective way.
Each management measure also describes a set of
performance objectives or goals for a specific area of
storm water management. These goals are somewhat
-------
broader in scope than what EPA
intends for measurable goals
under the NPDES Phase II Storm
Water Program, but they can be
adapted for use in the storm
water management program. For
example, the Site Development
Management Measure states the
following:
Plan, design, and develop sites to
— Maintain predevelopment
site hydrology by using site
design techniques that store,
infiltrate, evaporate, or
detain runoff.
— Protect areas that provide
important water quality
benefits or are particularly
susceptible to erosion and
sediment loss.
— Limit increases of impervious areas unless
predevelopment site hydrology is maintained
(Figure 3). Limit land disturbance activities, such
as clearing and grading and cut-and-fill, to reduce
erosion and sediment loss.
— Limit disturbance of natural drainage features and
vegetation.
Some BMPs that were considered appropriate for
meeting this management measure are:
— Promoting the use of cluster and open space
development
— Providing incentives to developers to reduce
impervious areas
— Conducting site assessments to identify
ecologically or historically significant areas for
preservation and locate key opportunities for
storm water management and ground water
recharge.
— Reducing the size of impervious surfaces by using
green roofs or modifying sidewalk, driveway, or
road standards
Some measurable goals that can be derived from this
management measure are as follows:
— Conduct a study, to be completed by the 3rd year
of the 5-year permit, to determine an appropriate
minimum storm water infiltration rate for
practices installed in new development. Also
examine ways that impervious area or density
credits can be offered for innovative and highly
effective storm water management practices.
Describe a protocol for developers to use to
determine the amount of infiltration and detention
practices needed to maintain predevelopment
hydrology and publish this protocol in a report to
be distributed to all developers working within the
NPDES-permitted area or to be included in a local
ordinance.
Conduct a survey to identify areas that provide
water quality benefits (e.g., ground water recharge
areas, areas with steep slopes or highly erodible
soils, ecologically significant areas) in the 1st year
of the permit. Conduct a study that examines
alternatives for protecting the priority lands
identified above by the 3rd permit year.
Incorporate this into guidance provided to the
development community.
the i
Below is a matrix showing how each section of
National Management Measures to Control Nonpoint
Source Pollution from Urban Areas—Draft relates to
the 6 minimum control measures of NPDES Phase II
Storm Water.
-------
How Do the Management Measures Compare to the 6 Minimum Control Measures of
NPDES Phase II?
Public
Education
Public
Involvement
Illicit Discharge
Construction
Site ESC
Post-
construction
Pollution
Prevention
Program Framework and Objectives
Establish Legal Authority
Develop an Institutional Structure
Provide Adequate Funding and Staffing
Foster Input From Technical Experts, Citizens, and Stakeholders
Establish Intergovernmental Coordination
Develop Training and Education Programs and Materials
•/
•/
•/
•/
,/
,/
,/
,/
Watershed Assessment
Characterize Watershed Conditions
Establish a Set of Watershed Indicators
Watershed Protection
Identify Critical Conservation Areas
Preserve Environmentally Significant Areas
Establish and Protect Stream Buffers
Promote Urban Forestry
Encourage Waterbody & Natural Drainage Protection When Siting Developments
,/
•/
•/
•/
•/
Site Development
Site Planning Practices
On-Lot Impervious Surfaces
Residential Street and Right-of-Way Impervious Surfaces
Parking Lot Impervious Surfaces
Xeriscaping Techniques
•/
•/
•/
•/
•/
New Development Runoff Treatment
Detention Ponds or Vaults
Ponds
Wetlands
Infiltration Practices
Filtering Practices
Open Channel Practices
Miscellaneous Practices
-------
Public
Education
Public
Involvement
Illicit Discharge
Construction
Site ESC
Post-
construction
Pollution
Prevention
Pollution Prevention
Household Hazardous Wastes
Lawn, Garden, and Landscape Activities
Commercial Activities
Proper Disposal of Pet Waste
Trash
Nonpoint Source Pollution Education for Citizens
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
•/
Existing Development
Identify, Prioritize, and Schedule Retrofit Opportunities
Implement Retrofit Projects as Scheduled
Restore and Limit the Destruction of Natural Runoff Conveyance Systems
Restore Natural Streams
Preserve, Enhance, or Establish Buffers
Revitalize Urban Areas
•/
•/
•/
•/
•/
•/
Operation and Maintenance
Establishing an Operation and Maintenance Program
Source Control Operation and Maintenance
Treatment Control Operation and Maintenance
•/
•/
•/
•/
•/
•/
Evaluate Program Effectiveness
Assess the Runoff Management Program Framework
Track Management Practice Implementation
Gauge Improvements in Water Quality
References
U.S. Environmental Protection Agency (USEPA).
2000. 'National Water Quality Inventory: 1998 Report
to Congress, http://www.epa.gov/305b/98report. Last
updated and accessed September 26, 2002.
U.S. Census Bureau (USCB). 2002. Urbanized
Areas: Cartographic Boundary Files.
http://www.census.gov/geo/www/cob/ua2000.html.
Last updated May 23, 2002. Accessed September 26,
2002.
For More Information
NPDES Phase II Storm Water Program
http://cfpub.epa.gov/npdes/stormwater/swphase2.cfm
EPA Office of Wetlands, Oceans, and Watersheds
Nonpoint Source Branch
http://www.epa.gov/owow/nps/
How To Obtain A Copy
To obtain a copy of the National Management
Measures to Control Nonpoint Source Pollution
From Urban Areas—Draft, visit www.epa.gov/
owow/nps/urbanmm/index.html to download it in
PDF format or contact Rod Frederick at
U.S. Postal Service Requests:
Assessment and Watershed Protection Division
(4503-T)
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Non-U.S. Postal Service Requests:
Assessment and Watershed Protection Division
U.S. Environmental Protection Agency
EPA West, Room 7417A
1301 Constitution Ave., NW
Washington, DC 20004
Phone:202-566-1197
Fax:202-566-1331
-------
EVALUATING INNOVATIVE STORMWATER TREATMENT
TECHNOLOGIES UNDER THE ENVIRONMENTAL TECHNOLOGY
VERIFICATION (ETV) PROGRAM
Author: Donna B. Hackett, NSF International
Co-Authors: John Schenk, NSF International
and Mary Stinson, USEPA/NRMRL
NSF International and USEPA/NRMRL
Ann Arbor, MI and Edison, NJ respectively
-------
NSF-2144Rev2 8x4 poster 12/16/02 2:07 PM Page 1
ETV Stormwater Devices
Sponsored By
USEPA and NSF International have partnered through the Environmental Technology Verification (ETV) Program's Water Quality Protection Center (WQPC) to
verify the performance of innovative, commercial ready technologies designed to protect ground and surface waters from contamination. The WQPC evaluates
technologies using technically sound protocols and appropriate QA/QC, thereby providing technology users and permitters with independent and credible assess-
ments of technologies. These assessments will be posted on the NSF and EPA ETV Web-sites, www.nsf.org/etv and www.epa.gov/etv Stormwater
treatment technologies are one of five wet weather flow technology areas being addressed in this ETV program. The following devices are either being presently u.s. Environmental Protection Agency NSF i
tested, or have completed testing: Web-slte: www'epa'g°v/av Web-slte:
Stormwater Treatment Technologies
-------
Managing Storm Water in Wisconsin:
A Local Partnership Protects the Kinnickinnic River
D. Kent Johnson and Andy Lamberson
Trout Unlimited, Kiap-TU-Wish Chapter
Setting:
Some of the best trout fishing in the Midwest can be found in St. Croix County, one of
the fastest-growing counties in Wisconsin. The City of River Falls, located on the
southern edge of St. Croix County and in the heart of the Kinnickinnic River Watershed
(Map 1), is home to 12,000 people. Because of its close proximity to the major
metropolitan area of Minneapolis-St. Paul, MN, River Falls is a rapidly growing
community, with a 20% population increase during the past decade. Growth estimates
project a population of 16,500 by the year 2010 (Ayres and Associates, 1987). This
estimate may be conservative, however, since it does not include growth in the
surrounding townships, where agricultural lands are rapidly being converted to rural
residential uses (SEH, 1995).
The Kinnickinnic River, a state "outstanding resource water", flows through River Falls
in west-central Wisconsin. A premiere trout stream, the "Kinni" is renowned for its
dense populations of wild brown trout. Approximately 2,000-8,000 trout per mile reside
in the river, with no stocking needed to sustain this naturally reproducing fishery.
According to fisheries biologists, a trout population of 1,000 fish per mile is considered
excellent.
Scientific Assessment of Local Storm Water Impacts:
The Kinnickinnic River is a valuable cold-water resource representing a major natural
amenity of the River Falls community. Although trout populations in the river are
currently high, the effect of growth in the City of River Falls and surrounding townships
has the potential to degrade the physical, chemical, and biological characteristics of the
Kinnickinnic River and its tributaries. As growth occurs, the creation of impervious
surfaces like roofs, sidewalks, driveways, streets, and parking lots generates a substantial
amount of storm water runoff that can significantly affect a river. Storm water impacts
include higher stream flows, thermal pollution, chemical pollution, and sedimentation
(Schueler, 1994), all of which pose threats to aquatic habitat, trout, and other cold-water
organisms.
Biological and Habitat Impacts
In the early 1990s, the local Kiap-TU-Wish Chapter of Trout Unlimited (Kiap-TU-Wish)
and the Wisconsin Department of Natural Resources (WDNR) began noting differences
in trout populations and habitat quality in the Kinnickinnic River, above and below the
City of River Falls. Likely due to storm water runoff, trout populations were
583
-------
significantly lower and stream bank erosion was increasing downstream from River Falls.
Thermal impacts were also suspected.
Thermal Impacts
In response to the concern about thermal pollution, Kiap-TU-Wish established a
temperature monitoring network in 1992, at four locations on the Kinnickinnic River
(Map 2) and two locations on major tributaries. With funding provided by Kiap-TU-
Wish and the Wisconsin Council of Trout Unlimited, Ryan TempMentor® data-logging
thermometers were purchased and installed at river locations upstream and downstream
from City of River Falls storm water discharges and two local hydropower dams. The
data logging thermometers record river temperatures at 10-minute intervals during the
April-September period, thereby documenting any thermal impacts associated with storm
water runoff during summer rains. Significant thermal impacts have been apparent
downstream from River Falls storm water discharges and hydropower dams. Rapid
increases in river temperature (up to 10 degrees Fahrenheit) are frequently evident at
locations downstream from storm water discharges during summer rainfalls (Figures 1
and 2), and storm water temperatures may exceed 78 degrees Fahrenheit (Figures 3 and
4), the upper lethal limit for brown trout. The thermal impact of the two city hydropower
dams produces downstream temperatures that are at least 3-6 degrees Fahrenheit warmer
than upstream temperatures during the summer months (Figure 5). Conversely,
downstream temperatures are significantly cooler during the winter months, with possible
impacts on incubating eggs in the trout redds.
Sediment and Nutrient Impacts
To evaluate the possible impacts of sediment and other urban pollutants in River Falls
storm water runoff, storm event-based composite sampling of residential, commercial,
and industrial areas of River Falls was conducted in 1992 by Short Elliott Hendrickson
(SEH), a local water resources management firm (SEH, 1995). A comparison of River
Falls monitoring results to EPA (1983) NURP monitoring results (Table 1) indicates that
sediment and nutrients are of particular concern in River Falls storm water runoff, with
total suspended solids, total Kjeldahl nitrogen, and total phosphorus concentrations
substantially higher than the NURP median concentrations.
Using Scientific Assessment Information to Initiate and Support Storm
Water Planning and Management Efforts:
One of the goals of the Kiap-TU-Wish temperature monitoring project was to obtain
sound scientific information on the local impacts of storm water runoff. Using this
monitoring information, Kiap-TU-Wish initiated a discussion with River Falls planners
and policy-makers about the need for storm water management tools that would enable
the city to grow while protecting the Kinni.
584
-------
Leveraging the Ideas and Resources of Local Partners:
City of River Falls Storm Water Management Plan
In 1993, the City of River Falls, through the WDNR, applied for and received federal
205J funding to develop a storm water management plan. Short Elliott Hendrickson
(SEH) was selected by the city to prepare the plan, in partnership with Kiap-TU-Wish,
local townships, the WDNR, the Kinnickinnic River Land Trust, and the University of
Wisconsin-River Falls. The "City of River Falls Water Management Plan for the
Kinnickinnic River and Its Tributaries" (Figure 6) was completed in 1994, at a cost of
$115,000, with a portion of the funding provided by the city and Kiap-TU-Wish. The
plan, adopted by the River Falls City Council in April 1994, provides a "blueprint" for
the city's storm water management efforts to protect the Kinnickinnic River as the city
grows (SEH, 1995).
Shortly after adoption of the storm water management plan, the City of River Falls
established a storm water utility to generate funding for storm water management projects
that protect and enhance the Kinnickinnic River. The storm water utility charges a fee to
city residents and businesses according to the amount of storm water running off a
property. As an incentive to residents and businesses that reduce the amount of storm
water runoff from their properties, the City of River Falls reduces their annual storm
water utility fee proportionately.
In 2002, River Falls adopted a storm water management ordinance (Figure 7). The
ordinance, prepared with input from the partners, is another key element of the city's
storm water management plan, and requires all developers to use storm water
management practices that entirely infiltrate the first 1.5 inches of runoff from all storm
events. Among the options for developers is the low impact development approach,
which uses biotechnology (rain gardens, swales, constructed wetlands, and buffers of
native vegetation) to distribute and infiltrate storm water across the landscape, rather than
concentrating and conveying it to the river with conventional storm water infrastructure
(curb and gutter, storm sewers, and detention ponds).
Kinnickinnic River Priority Watershed Project
In 1995, efforts to protect the Kinnickinnic River expanded watershed-wide when the
WDNR selected the Kinnickinnic River as a part of the state's Priority Watershed
Program. The Priority Watershed Program provides annual funding, over a ten-year
period, for cost-shared projects in both agricultural and urban areas of the watershed that
protect and enhance the quality of the Kinnickinnic River. Prior to receiving state
funding, however, a watershed plan had to be developed so that the state and local cost-
share funding could be appropriately directed to areas of the watershed in greatest need of
agricultural and urban best management practices (BMPs). The WDNR worked in
partnership with Kiap-TU-Wish, two counties, six townships, three cities (including
River Falls), the University of Wisconsin-River Falls, the Kinnickinnic River Land Trust,
and SEH to develop the "Nonpoint Source Control Plan for the Kinnickinnic River
585
-------
Priority Watershed Project" (WDNR, 1999) (Figure 8), which was approved by the
Wisconsin Natural Resources Board in April 1999. The plan is unique in that it is among
the first priority watershed plans in the state to incorporate an urban storm water
management component, applying the approach used in the City of River Falls storm
water management plan to other cities and townships across the watershed. A list of
eligible agricultural and urban BMPs and associated cost-share rates is presented in
Table 2.
Local Environmental Education is Important:
In 1998, recognizing the need for an educational tool that can be used to protect cold-
water resources in urbanizing areas, Kiap-TU-Wish, in partnership with Palisade
Productions of Minneapolis, MN, produced a video entitled: "A Storm on the Horizon"
(Figure 9 and display). Using the Kinnickinnic River as the backdrop, this 15-minute
video describes the value of a cold-water resource, discusses the potential threats posed to
cold-water resources by urban growth, and also describes some tools available to
communities for protecting these resources while accommodating growth. The video
won a Silver Screen Award in the "Environmental Issues and Concerns" category at the
Chicago International Film Festival in 1999. Kiap-TU-Wish members have distributed
nearly 3,000 copies of the video nationwide, to local planners and policy-makers,
engineers, scientists, elementary, middle school, high school, and college educators and
students, nonprofit organizations, and other Trout Unlimited members and chapters.
Translating a Storm Water Plan to Action in River Falls:
In 2000, the City of River Falls and the River Falls School District took advantage of an
opportunity to implement some of the new storm water management techniques
described in the city's storm water management plan. The school district was planning to
build a new high school near the South Fork of the Kinnickinnic River, a tributary to the
main river. After learning that a preliminary site plan had already been designed for the
new high school, several Kiap-TU-Wish members showed "A Storm on the Horizon" to
school officials and city planners, and stressed the need for good storm water
management practices on the site. Kiap-TU-Wish members, the City of River Falls,
SEH, and Kinnickinnic River Priority Watershed Project participants worked with the
school district's landscape architect to redesign the site. A large, expansive parking lot in
the original design was changed to smaller, separated lots buffered with native vegetation
that infiltrates storm water runoff from these impervious surfaces. Native buffers were
also established between the athletic fields, to trap soil and nutrients. Three storm water
detention ponds on the site contain and infiltrate excess runoff, including the runoff from
the building roof. With funding provided by the Priority Watershed Project, an
innovative irrigation system was also installed to pump storm water from the detention
ponds to the athletic fields. As originally designed, the new high school site would have
cost the River Falls School District $8,000 per year in storm water utility fees paid to the
City of River Falls. With the redesign work, it is anticipated that no storm water will
leave the site, saving the school district $8,000 per year while protecting the South Fork
and Kinnickinnic River. With completion of the new high school in the fall of 2001,
586
-------
Kiap-TU-Wish members and Kinnickinnic River Priority Watershed Project participants
plan to help the school district install interpretive signs that explain the various storm
water management components of the site. It is hoped that these components can be
incorporated into the educational curriculum at the high school. Funding for the signage
will also be provided by the Priority Watershed Project.
The Benefits of Effective Storm Water Management:
Trout are an important indicator species of environmental quality, especially in an
urbanizing area. As such, protection of the Kinnickinnic River is critical to help ensure
the environmental, cultural, and economic future of River Falls and surrounding
communities. With nearly 200 members, the Kiap-TU-Wish Chapter of Trout Unlimited
has been instrumental in protecting the Kinnickinnic River during the past decade. The
chapter has raised the awareness of planners, policy-makers, and residents with regard to
storm water issues, and has helped to change the way River Falls manages an outstanding
cold-water resource in Wisconsin, thereby ensuring that the Kinni will be available for
the enjoyment of future generations.
For more information, please contact:
Kent Johnson
Kiap-TU-Wish Chapter, Trout Unlimited
P.O. Box 483
Hudson, WI 54016
Phone: 715-386-5299
FAX: 715-386-6065
E- mail: kentj ohnson@pressenter. com
Kiap-TU-Wish Website: http://www.lambcom.net/kiaptuwish/
587
-------
Mapl Kinnickinnic River Subwatersheds
and Tributaries
/y Ccuniy Etouea*n>'
'.*•_•' Subu-tiershul Bomufat%'
'• j laiominial Stnanu
-------
Map 2
Thermal
Monitoring
Sites
River Falls Subwatershed
LEGEND
B WeOs
* Garni
Sutiwalcndici: Boundary
County Buuoetary
/ ^ /
f\f
/ \/ Intanutlenl Streams
Pot«m*l
Open W
STUDY AREA
589
-------
Map 2 Additional Information on the
Kinnickinnic River Thermal Monitoring Sites:
Quarry Road: The Quarry Road site is located along Quarry Road in the River Falls
Subwatershed, at the upper (ME) River Falls city limit. This upstream location is
unaffected by River Falls storm water discharges and the two city hydropower
impoundments (Lake George and Lake Louise).
Cedar Street: The Cedar Street site is located near the former Cedar Street Bridge in the
River Falls Subwatershed. This urban location is immediately downstream from four
direct storm water discharges draining residential and commercial areas of River Falls.
The site is also immediately upstream from Lake George and Lake Louise.
Upper Glen Park: The Upper Glen Park site is located in the upper part of Glen Park in
the River Falls Subwatershed. This location is approximately 0.1 mile downstream from
a large storm water discharge (Bartosh Canyon) draining a residential area of River Falls.
The site is also 0.1 mile downstream from Lake George and Lake Louise.
Lower Glen Park: The Lower Glen Park site is located in the lower part of Glen Park in
the River Falls Subwatershed, at the lower (WSW) River Falls city limit. This location is
approximately 0.9 mile downstream from Bartosh Canyon and the two impoundments.
The site is also 0.2 mile downstream from the Rocky Branch tributary.
590
-------
Figure 1 Cedar Street Thermograph With
Storm Water-Induced Temperature Spikes (*),
July-August 1993
LU
oc
LU
O.
2
LU
fi
oc
I-
co
72
70
68
66
64
62
60
58
56
54
52
50
07/17/93 ' 07/31/93 f 08/14/93 08/24/93
07/24/93 08/07/93 06/21/93
DATE
Sfream Twnperafyre Summary: Average= 58.8 F Minimum^ 52,3 F Maximum- 69.1
*» Rain Evwil
591
-------
Figure 2 Cedar Street Thermograph With Storm
Water-Induced Temperature Spike July 25, 1993
68
67
t&
€••6
M
63
(1
1 34 !'Wh«J
JiiillilliLilliJ ill"IHItlllillJilll.;l'Jili1tS.llllillllLli:i.il:l!lllllt'lli!lll.llfllllllll 'r'iiltlilj|i:HmilllimiH|{ ' HI 111.1 li'I I
12-00 AM U6:Ul) AM " ? OO f*M 06-00 PM I 12.QOAW
03 OO *M 0900 AM 03;OO PM 0^:00 PW
f'ML
592
-------
Figure 3 Storm Water Temperatures (*)
in a Commercial River Falls Subwatershed,
June 1992
90
£ 80
ULJ
§ 70
mi 60
CL
y so
40
1 06/11/92
M/04/92 06/18/92
DATE
06/25/92
07/01/92
Rain Event
593
-------
Figure 4 Storm Water Temperatures During
Four Rain Events in a Commercial River Falls
Subwatershed, June 1992
84
37
BO
IJAIE
594
-------
Figure 5 Comparison of Quarry Road, Cedar
Street, and Lower Glen Park Thermographs,
July 25, 1993
67
t.6
fill
• I
bl
11
I- Vt IMI lifi
«.
^
111.111.III tllllllUIIIIIIIIIIIIIM
-------
Table 1 River Falls Storm Water Quality (1992)
Compared to NURP Monitoring Results
Residential Subwatershed
Water Quality River Falls NURP
Variable (ma/1) Median Median
TSS (Total
Suspended 240.0
Solids)
TKN (Total
Nitrogen)
TP (Total
Phosphorus) u''°
Cu (Copper) 0.030
Pb(Lead) 0.015
Zn(Zinc) 0.110
101.0
1.90
0.38
0.033
0.144
0.135
Industrial Subwatershed
Water Quality
Variable (ma/I')
TSS (Total Suspended
Solids)
TKN (Total Nitrogen)
TP (Total Phosphorus)
Cu (Copper)
Pb (Lead)
Zn (Zinc)
River Falls
Median
250.0
2.5
0.50
0.030
0.050
0.210
These data represent only one storm event.
No NURP data are available for direct comparison
Commercial
Water Quality
Variable (mg/D
TSS (Total
Suspended
Solids)
TKN (Total
Nitrogen)
TP (Total
Phosphorus)
Cu (Copper)
Pb (Lead)
Zn (Zinc)
Subwatershed
River Falls
Median
150.0
2.1
0.50
0.030
0.080
0.190
NURP
Median
69.0
1.20
0.20
0.029
0.104
0.226
All Subwatersheds
Water Quality
Variable (ma/I)
TSS (Total
Suspended
Solids)
TKN (Total
Nitrogen)
TP (Total
Phosphorus)
Cu (Copper)
Pb (Lead)
River Falls
Median
200.0
2.6
0.50
0.030
0.050
Zn (Zinc) 0.140
*NURP monitoring was completed prior
decrease in leaded gasoline use.
NURP
Median
100.0
1.50
0.38
0.034
0.140*
0.160
to the
596
-------
Figure 6
Figure 7
Figure 8
Ni
-------
Table 2. Eligible Cost-Shared Agricultural and Urban BMPs
Agricultural BMPs
BEST MANAGEMENT PRACTICE
Nutrient and Pesticide Management
Pesticide Handling Spill Control Basins
Livestock Exclusion from Woodlots
Intensive Grazing Management
Manure Storage Facilities
Manure Storage Facility Abandonment
Field Diversions and Terraces
Grassed Waterways
Critical Area Stabilization
Grade Stabilization Structures
Agricultural Sediment Basins
Shoreline and Streambank Stabilization
Shoreline Buffers
Wetland Restoration
Barnyard Runoff Management
Barnyard Abandonment or Relocation
Roofs for Barnyard Runoff Management and Manure
Storage Facilities
Milking Center Waste Control
Cattle Mounds
Land Acquisition
Lake Sediment Treatment
Well Abandonment
STATE COST-SHARE RATE
50%
70%
50%
50%
70% and 50%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
70%
Urban BMPs
BEST MANAGEMENT PRACTICE
Critical Area Stabilization
Grade Stabilization Structures
Streambank Stabilization
Shoreline Buffers
Wetland Restoration
Structural Urban Practices
High Efficiency Street Sweeping
STATE COST-SHARE RATE
70%
70%
70%
70%
70%
70%
50%, 5 years only
598
-------
Figure 9
A Storm on the Horizon
A 1999 Chicago International Film Festival Silver Award Winner
Category: Environmental Issues and Concerns
The purpose of the video is to educate the public about the effects of storm water
on our lakes, streams and rivers. This educational video discusses the issues
surrounding urban development and its impact on water quality. The story of the
Hrmickinnic River in western Wisconsin is told, and the prospect for the river's
long-term health is discussed. The video is a must see for anyone interested in
land use issues and the health of our water resources. The video:
1. Establishes the value of a cold water resource and its importance to the
community.
2. Demonstrates the impact of storm water on water resources.
3. Outlines what can be done to enable development to occur while protecting
water resources.
Professionally produced by Kiap-TU-Wish and Palisade Productions of
Minneapolis, MN, the video is 15 minutes in length and is geared toward
educating the general public, land use planners, and decision makers about the
impacts of storm water on our water resources.
The video is available for a donation of $15, which includes shipping and
handling. To receive the video, please contact us at:
Kent Johnson or Andy Lamberson
Kiap- TU- Wish Chapter of Trout Unlimited
P.O. Box 483
Hudson, WI 54016
Or e-mail us at lamberson@attbi.com
599
-------
References:
Ayres and Associates. 1987. River Falls Master Plan Report: A Policy Guide for
Growth.
EPA. 1983. Results of the Nationwide Urban Runoff Program: Volume 1. Final Report.
U.S. Environmental Protection Agency, Washington, D.C. 197 p.
Schueler. 1994. The Importance of Imperviousness. In: Watershed Protection
Techniques 1 (3): 100-111. Center for Watershed Protection, Silver Spring, MD.
SEH. 1995. City of River Falls Water Management Plan for the Kinnickinnic River and
its Tributaries. Short Elliott Hendrickson, St. Paul, MN. 286 p.
WDNR. 1999. Nonpoint Source Control Plan for the Kinnickinnic River Priority
Watershed Project. Wisconsin Department of Natural Resources, Bureau of Water
Resources Management, Madison, WI. 279 p.
600
-------
Duluth Streams: Community partnerships for understanding urban stormwater and water quality i i i
: ^ ^ .. ^ ^ ^j.j.i^^ i^^^^i _ .c j-i^ ^ ^%—^ ^j. i ^ i_^ ^ \JL LJl JL L41 LJL JL
issues at the head of the Great Lakes
Marion Lonsdale1, Richard Axler2, Cynthia Hagley3, George Host2, Carl Richards3, and Bruce Munson3
1 Duluth Public Works and Utilities, Duluth, MN;2 Natural Resources Research Institute, U. Minnesota-Duluth;3
Minnesota Sea Grant, U. Minnesota-Duluth
streams.or
setting
u nderstanding
the streams
citizen involvement
stormwater manage
T
- . ^g^gjj^., -,^ £~',L .„•
fcSsi?,^-.- "SJ^^wJI^^^rrc:
•"rySs,^^?-.,-"v.:.t;'i"..'.. =iiiiiS
Setting
• 42 named streams; one of the highest densities of stream corridors in any US
metro area
• Urban and rural development impact these streams by increasing water
volume, temperature, suspended sediments, road salts, organic matter and
nutrients
Partnership
• City, UMD researchers, education and outreach professionals, local resource
agencies and other educational institutions
Chief Goal
• Enhance public understanding of aquatic ecosystems and their connections to
watershed land use to provide both economic and environmental sustainability.
Objectives
1. Link real-time remote water quality sensing in 4 urban streams and GIS
technology to current and historic WQ and biological databases using
advanced data visualization tools in a website and information kiosks;
2. Incorporate visually engaging interpretive text, animations and videos into the
website to illustrate the nature and consequences of degraded stormwater and
the real costs to society;
3. Engage the public in the stormwater issue to facilitate development and
implementation of the Duluth Stormwater Management Plan by:
• Establishing high school stewardship of 3 streams
• Adapting the Nonpoint Education for Municipal Officials (NEMO) program
to the greater Duluth Metropolitan Area
• Developing high school and college curricula
• Hosting a Duluth Streams congress as a community forum for presenting
all project results
Average Slope of
Chester Creek
30%
25%
20% -r
15% -
10% -
5%
0%
Duluth Trout Streams
NEMO cutoff for unimpaired water ^H
....••III
* fr
-------
Maximum Utility for Minimum Cost:
Simple Structural Methods for Stormwater Quality Improvement
T. J. Mullen
President, Best Management Products, Inc.
Abstract
Stormwater runoff is characterized by the United States Environmental Protection Agency as 'one
of the greatest remaining sources of water pollution" in America (United States Environmental
Protection Agency, November 1999). Thus, efforts to implement Stormwater quality improvement
regulations are accelerating across the United States, compelling municipalities and land
developers to maximize the usefulness of Stormwater infrastructure as never before. With simple
modifications to current designs, common catch basins and other Stormwater structures can be
more effectively utilized as pollution control devices, rather than merely as a way to move
Stormwater. Future systems must drain runoff which cannot be infiltrated to areas where it can be
appropriately managed, and simultaneously, reduce the environmental impact of the ultimate
discharge to the receiving waters. Adding a deep sump to catch basins, a common feature in
some areas of the country, has been shown to remove some sediments and gross particles. As an
additional benefit, these structures allow the use of an outlet hood or baffle, which can drastically
reduce the discharge of floatable debris and trash, and aid in the removal of free oil and grease.
This "cleaner" runoff can also extend the service life of a traditional Stormwater detention facility,
such as a pond, or a retention facility, such as a groundwater recharge area. This paper
addresses a group of low-cost components which comprise the SNOUT® Stormwater Quality
Control System, manufactured by Best Management Products, Inc. (BMP, Inc.) The applications
include a deep sump catch basin with an outlet hood, a structure with an outlet hood and flow
restrictor, structures configured to bypass high flows, and outlet controls which can accommodate
extreme flow conditions while retaining captured pollutants.
Background
Catch basins, Stormwater inlets, and other specialized structures have a long history of use as part
of municipal separate storm sewer systems (MS4s) in controlling Stormwater runoff. So too have
the many devices used with them to aid in the removal of pollutants, such as grates, traps, hoods,
and sumps. With the aid of new appurtenances based on older concepts, these simple structures
are being more effectively used and maintained as a first line of defense against non-point source
pollution in urbanized areas to improve Stormwater quality.
In its simplest form, a Stormwater inlet's primary function is to intercept sheet flows in order to
prevent the accumulation of Stormwater in an area where flooding could impede traffic or
pedestrians, cause property damage, or otherwise present a nuisance. However, these inlets to
MS4s or combined sewer systems (CSSs) are often the entry point of pollutants from diffuse
sources found in Stormwater runoff. As a result, pollution is often discharged untreated, directly
into our surface waters.
-------
The use of deep sumped structures, as part of the stormwater system along with simple
components essentially creates numerous "micro-detention" nodes throughout a stormwater
conveyance, and allows moderate levels of pollutants to be captured in an economical manner.
Typically, this method requires much lower on-going maintenance needs than traditional "tray-type"
catch basin inserts or baskets, and a significantly lower capital cost than most "end-of-pipe"
controls. Further, the ability of sufficiently sumped structures to intercept gross-pollutants and finer
particles such as suspended solids (SS) has been well documented. A recent study in New Jersey
found an average SS capture rate of 32 percent over several storm events (Pitt and Field, 1998).
Capture of trash, floatables and other gross-pollutants have also been widely recognized as a
benefit of an inlet with a hood. A1995 study of New York City catch basins compared the relative
effectiveness of structures with and without hoods. The hooded structures captured 85 percent of
the litter that entered the combined sewer inlets compared to 30 percent for the catch basins
without hoods (New York City Department of Environmental Protection, 1995, cited in EPA Doc.
832-F99-008 September 1999). The nation's first Total Maximum Daily Load (TMDL) for trash,
being established for the Los Angeles River Basin, calls for reductions in gross-pollutant loading.
Other areas of the country are expected to follow this lead (Will Shuck, Long Beach Press-
Telegram, 2001).
Until recently, the devices available for use as hoods or traps were mainly limited to metal hoods,
metal or PVC elbows, and tees. While many of these devices have been in service for decades,
little design effort was given to these appurtenances in terms of pollutant removal performance,
hydraulic efficiency, or ease of installation. That situation has changed with a versatile product line
available from Best Management Products, Inc. of Lyme, CT. Relative to the traditional hoods or
fittings, which lack an oil-proof gasket or an anti-siphon vent, the new design transforms the hood
concept into a higher performance, multi-task stormwater quality and quantity control system. This
system, the SNOUT® Stormwater Quality Control System (US Patent # 6126817), uses vented
plastic-composite hoods and related components to improve water quality and control flow
quantity.
System Advantages
• SNOUT® hoods use an oil tight gasket sealing system around perimeter of unit.
• Anti-siphon vent prevents pollutants from being drawn downstream in full flows.
• Watertight access port allows easy pipe inspection and maintenance.
• Light Weight/High Strength composite construction is durable and easy to install.
• Sizes to fit over outlet pipes up to 96" outside diameter.
• Highly flexible low-cost component system with a variety of accessories including Flow
Restrictors, Oil Absorbents, Flow Deflectors, and Odor Filters.
• SNOUT® components can be used to construct a wide variety of stormwater quality
structures including those with high flow bypass, swirl chambers, and outlet flow control.
• Use of sumps and SNOUT®hoods keeps pipes cleaner, thus reducing pipe maintenance.
Since this system became commercially available in 1999, more than 7,000 SNOUT® hoods have
been installed. Initial results have been quite favorable. SNOUT® systems have been or will be
installed as part of research or monitoring projects in the following locations:
-------
Washington, D.C., Navy Yard, Center for Low Impact Development
Bryn Mawr, PA, Regional Stormwater Facility, Yerkes Associates, Designer
Harvey's Lake Demonstration Project, Harvey's Lake, PA, PA DEP and Princeton Hydro, LLC
Data collected from these and other projects will be incorporated on an on-going basis on the BMP,
Inc. website atwww.bmpinc.com, along with selected case studies and photos from a variety of
projects.
Applications
A variety of applications and SNOUT® system configurations exist in the field. Each has its
advantages and disadvantages, which are outlined below. These systems include:
Catch Basin with an Outlet Hood- This is the most basic application. This system combines a
sumped catch basin with a hood. It is useful for capturing trash and floatables, and modest levels
of free oils, and sediment. These structures can be inlet-only, or in-line with other structures. To
increase oil retention, oil absorbent booms can be placed in the structure. This application has
limitations based primarily on the volume and sump depth of the structure itself. To minimize re-
suspension of finer captured solids, a deep sump, with a minimum depth of 4 feet, or a depth equal
to 3X the outlet pipe inside diameter is recommended, (see Figure 1)
In-line Catch Basin with a Hood and Flow Restrictor- This application is useful for limiting the
discharge rates down stream. A micro-detention node can be created using a flow restrictor,
making use of the storage volume in pipes upstream or ponding areas above the inlet. It is also
used in outlet structures in detention basins. Discharge rates can be accurately controlled by slot
or orifice dimensions in the riser pipe shielded inside a SNOUT® hood, making it difficult to clog
with floating debris. The structure must receive periodic maintenance to ensure that sediment
accumulation does not reach entrance to riser pipe, is designed to provide absolute flow control.
A caution to the designer is that this in-line application does not provide for overflow other than that
which can flow over the open top of the riser pipe. For installations where occasional flooding
cannot be tolerated, the design shown in Figure 5, Outlet Structure with Overflow, should be used.
(see Figure 2)
Structures in Series with Oil Absorbent is and Flow Deflector Plates- This application is intended
for use as a terminal structure on a site where higher than normal pollutant loads may be present.
Stormwater makes a "multiple pass" through deep sump structures with hoods and accessories.
Accessories include oil absorbent booms for increased oil retention, and deflector plates for
increased solids removal. This application is also an excellent pre-treatment design prior to
discharge to a conventional Stormwater BMP. Limitations are based primarily on structure sizes,
whereby larger structures with deeper sumps will yield better removals, (see Figure 3)
Bypass Structure Configuration- This design combines the features of structures in series, but
allows for high flows to be bypassed from the primary treatment structures. All Stormwater
receives some treatment however, as the terminal structure contains a large SNOUT® hood and a
deep sump. Limitations are primarily that multiple structures must be utilized to perform the
-------
bypass, but they can be configured in a wide variety of ways such that hydraulic grade lines are
maintained, (see Figure 4)
Outlet Structure with Overflow- This design combines accurate outlet control with the SNOUT®
flow restrictor as well as an overflow mode that maintains capture of floatable pollutants and trash.
Limitations may be based primarily on the outlet structure size, as to accommodate large flows,
large size SNOUT® hoods must be used which require large structures that can be costly to build.
(see Figure 5)
Cost Savings Note: Structures for all SNOUT® systems are non-proprietary and obtained locally
from pre-casters or built in place by local contactors. SNOUT® components and designs are low-
cost, but are protected by a US Patent with international patents pending. The combination of low-
cost components in non-proprietary structures can reduce overall installed systems costs
dramatically.
Following are application drawings of the systems mentioned above:
-------
Figure 1- Catch Basin with Hood
• OUTLET
OIL & FLOATABLE DEBRIS
ON SURFACE CANNOT EXIT PIPE
•o iitii";1" i . ..1.: r.'nVi'b: r.. •• •%• .i.-iy,i'. i. i .1. j.i
-------
Figure 2- In-line Catch Basin with Hood and Flow Restrictor
SNOUT HOOD WITH
FLOW RESTRICTOR -j
ASSEMBLY /
FLEXIBLE COUPLING
WITH STAIN LESS CLAMP
WATERTIGHT
ACCESS PORT
PVC RISER PIPE
WITH VERTICAL
NOTCH PER
LOCAL REQS.
."1
SECTION
RISER
DETAIL
DIMENSION REQUIREMENTS
ELEVATION A:_
ELEVATION B:_
_ (OVERFLOW)
.(OUTLET INVERT)
RISER DIMENSIONS
D= RISER ID:
W1= SLOT WIDTH:
W2= NOTCH WIDTH:
H1= SLOT LENGTH:
H2= NOTCH LENGTH:
H3= RISER LENGTH:
H4= SUBMERGE DEPTH: (MIN. 18")
H5= DEPTH TO BOTTOM: (MIN. 24")
HOOD SIZE DETERMINED BY MANUFACTURER
BASED ON RISER DIAMETER.
ADDITIONAL SKIRT PIECES AVAILABLE TO
INCREASE HEIGHT OF HOOD.
-------
Figure 3- Structures in Series with Oil Absorbents and Deflector Plates
STAINLESS DEFLECTOR PLATE
BY BMP, INC.
(OPTIONAL)
SNOUT BY BMP,
INC.
DESIGN NOTE: ENSURE THAT SEPARATION BETWEEN STRUCUTRES
IS SUFFICIENT TO ACHIEVE PROPER COMPACTION BETWEEN PIPES TO
ALLOW FOR DIFFERENTIAL SETTLEMENT.
-------
Figure 4- Bypass Structure Configuration
FROM INLETS
UPSTREAM
FULL FLOW
VOLUME
(BY- PASS)
BY-PASS STRUCTURE
DIAMETER D*'OF OUTLET TO TREATMENT
STRUCTURE SIZED ACCORDING TO
DESIGN STORM REQUIREMENTS
INLET FROM
-TREATMENT
STRUCTURE
* HEIGHT OF RAMP DETERMINED BY
DESIGN STORM TREATMENT VOLUME
BY-PASS STRUCTURE
•^L
x
SYSTEM ADVANTAGES:
GROSS POLLUTANTS,
TRASH AND FLOATABLES
STILL CAPTURED UNDER
FULL FLOW OPERATION.
NO FLOW IS BYPASSED
WITHOUT TREATMENT.
CONFIGURATION NOTE:
STRUCTURES CAN BE IN ANY
CONFIGURATION TO MEET
SITE REQUIREMENTS GIVEN
THAT:
1. BYPASS STRUCTURE MUST
FLOW TO TREATMENT
STRUCTURES AND FULL-FLOW
STRUCTURE.
2. TREATMENT STRUCTURES
MUST FLOW TO FULL-FLOW
STRUCTURE.
•THIS SNOUT
CAN BE
FITTED WITH
FLOW
RESTRICTOR
_ FLOW DEFLECTORS
(FOR ENHANCED
SOLIDS REMOVAL)
-------
Figure 5- Outlet Structure with Overflow
SYSTEM ADVANTAGES:
FLOW DISCHARGE RATE
ACCURATELY
CONTROLLED WITHOUT
CLOGGING WITH SNOUT
FLOWRESTRICTOR
OVERFLOW MODE STILL
CONTAINS FLOATING
POLLUTANTS
INFLOW FROM POND OR
STORMWATER DETENTION
FACILITY
DIMENSIONS:
D= PIPE DIAMETER
H= HEIGHT OVER ALL
L= LENGTH INSIDE
L1= LENGTH TO PARTITION
L2= LENGTH TO BACK WALL
W= WIDTH INSIDE
-------
Design and Maintenance Considerations
The SNOUT® system from BMP, Inc. is based on a vented hood that can reduce floatable trash
and debris, free oils, and other solids from stormwater discharges. In its most basic application, a
SNOUT® hood is installed over the outlet pipe of a catch basin or other stormwater quality
structure which incorporates a deep sump. The SNOUT® forms a baffle in the structure which
collects floatables and free oils on the surface of the captured stormwater, while permitting heavier
solids to sink to the bottom of the sump. The clarified intermediate layer is forced out of the
structure through the open bottom of the SNOUT® by displacement from incoming flow. The
resultant discharge contains considerably less unsightly trash and other gross pollutants, and can
also offer modest reductions of free-oils and finer solids.
As with any structural stormwater quality Best Management Practice, design and maintenance
considerations will have a dramatic impact on SNOUT® system performance over the life of the
facility. The most important factor to consider when designing structures which will incorporate a
SNOUT® is the depth of the sump (the sump is defined as the depth from beneath the in vert of the
outlet pipe to the bottom of the structure). Simply put, the deeper the sump, the more effective the
unit will be in terms of removing pollutants, preventing resuspension, and reducing frequency of
maintenance. More volume in a structure means more quiescence, thus allowing the pollutant
constituents a better chance to separate out. Secondly, more volume means fewer cycles between
maintenance operations, because the structure has a greater capacity.
Design Notes:
• As a rule of thumb, BMP, Inc. recommends minimum sump depths based on outlet pipe
inside diameters of 2.5 to 3 times the outlet pipe size.
• Special Note for Smaller Pipes: A minimum sump depth of 36 inches for all pipe sizes 12
inches ID or less, and 48 inches for pipe 15-18 inches ID is required if collection of finer
solids is desired.
• The plan dimension of the structure should optimally be 6 to 7 times the flow area of the
outlet pipe.
Example Calculation:
A SNOUT® equipped structure with a 15 inch ID outlet pipe (1.23 sqft. flow area) will offer an
optimal combination of cost-effectiveness and pollution removal with a minimum plan area of 7.4
sqft. and minimum 48 inch sump. Thus, a readily available 48 inch diameter manhole-type
structure, or a rectangular structure of 2 feet x 4 feet will offer sufficient size when combined with a
sump depth of 48 inches or greater.
Therefore, it follows that larger pipe sizes will require larger structures and/or deeper sumps to
maintain optimal effectiveness.
As for long term structural maintenance practices, BMP, Inc. recommends the following:
-------
• Monthly monitoring for the first year of a new installation after the site has been stabilized.
• Measurements of sediment depth and observations of floating pollution should be taken
after each rain event of .5 inches or more, or monthly, as determined by local weather
conditions.
• Checking sediment depth and noting the surface pollutants in the structure will be helpful
in planning maintenance. The pollutants collected in SNOUT® equipped structures will
consist of floatable debris and oils on the surface of the captured water, and grit and
sediment on the bottom of the structure.
• It is best to schedule maintenance based on the solids collected in the sump. To achieve a
reasonable compromise between practicality and pollution removal effectiveness, the
structure should be cleaned when the sump is half full (e.g. when 2 feet of material collects
in a 4 foot sump, clean it out). The more often it is cleaned, the better the performance will
be as the structure will maintain a greater "effective volume." Of course, depending on
resources available for maintenance, some performance may have to be sacrificed due to
budgetary constraints.
• Structures should also be cleaned if a spill or other incident causes a larger than normal
accumulation of pollutants in a structure.
• Maintenance is best done with a vacuum truck.
• If oil absorbent hydrophobic booms are being used in the structure to enhance
hydrocarbon capture and removals, they should be checked on a monthly basis, and
serviced or replaced when more than 2/3 of the boom is submerged, indicating a nearly
saturated state.
• All collected wastes must be handled and disposed of according to local environmental
requirements.
• To maintain the SNOUT® hoods themselves, an annual inspection of the anti-siphon vent
and access hatch are recommended. A simple flushing of the vent, or gentle rodding with
a flexible wire are all that's typically needed to maintain the anti-siphon properties.
Opening and closing the access hatch once a year assists a lifetime of trouble-free
service.
Further structural design guidelines, maintenance recommendations and site inspection field report
sheets are available from BMP, Inc. Please contact us if we can offer further assistance.
Summary
Municipal engineers and stormwater designers are grappling to adapt a pollution control function to
traditional drainage systems, recognizing that fundamental changes in traditional stormwater
infrastructure design will be required. Presently, the primary function of most MS4s are to
evacuate stormwater from point A to point B as quickly and efficiently as possible, often with
minimal regard of the impact to receiving waters. As such, compliance with the stormwater quality
regulations that are being promulgated across the United States could be difficult for impacted
municipalities. Fortunately, implementation of simple design changes, and low-cost technologies,
such as those manufactured by Best Management Products, Inc., can make complying with new
regulations mandating reductions in the discharge of trash, floatable debris, oil and grease, and
sediment easier. Updated structure designs are particularly easy to implement for new
construction. In areas where catch basins already have sumps, installing an outlet hood is quick
work which can yield substantial benefits. Retrofits to systems without sumped structures,
-------
especially at strategic nodes, are still cost-effective as they make the existing conveyance systems
more efficient, and can extend the service life of traditional stormwater facilities. While the work
remaining to improve our stormwater infrastructure is daunting, the benefits of reducing pollutants
from stormwater runoff will be numerous. Benefits include improved surface water quality, reduced
impacts to wildlife habitat, and a healthier environment for recreation and enjoyment of our natural
resources.
References
Pitt, Robert and Field, Richard. "An Evaluation of Storm Drainage Inlet Devices for Stormwater
Quality Treatment." The University of Alabama at Birmingham, Department of Civil and
Environmental Engineering, United States Environmental Protection Agency, Wet Weather Flow
Research Program, pp. 3-4,1998.
United States Environmental Protection Agency., Environmental News, "EPA expands controls on
polluted runoff to further protect Nation's drinking water and waterways." R-134, p. 1, November,
1999.
HydroQual, Inc., "City-Wide Floatables Study." New York City Department of Environmental
Protection, Bureau of Environmental Engineering, Division of Water Quality Improvement. As
cited in "CSO Technology Fact Sheet, Floatables Control". United States Environmental Protection
Agency, Washington, D. C. Document EPA 832-F99-008, p. 5, September 1999.
Shuck, W., "Waterway trash limits considered", Long Beach Press-Telegram, p. A3, Nov. 29, 2000.
SNOUT® is a registered trademark of Best Management Products, Inc.
53 Mt. Archer Rd., Lyme, CT 06371
Phone:800-504-8008 Fax:860-434-3195 Web Site: www.bmpinc.com
The SNOUT® is protected by US Patent # 6126817, international patents pending.
------- |