&EPA
        United States
        Environmental Protection
        Agency
           Office of Research and
           Development
           Washington, DC 20460,
EPA/625/R-00/001
July 2000
National Conference on
Tools for Urban Water
Resource Management &
Protection

Proceedings
Chicago, IL
February 7-10, 2000

-------
                                   EPA/625/R-00/001
                                   July 2000
National Conference on Tools for
      Urban Water Resource
   Management and Protection

            Proceedings

        February 7-10, 2000
             Chicago, IL
          Technology Transfer and Support Division
         National Risk Management Research Laboratory
            Office of Research and Development
            U.S. Environmental Protection Agency
               Cincinnati, OH 45268

-------
                                              Notice


The views expressed in these  Proceedings are those of the individual authors and do not necessarily reflect
the views and  policies of the  U.S.  Environmental  Protection  Agency (EPA).  Scientists in EPA's Office  of
Research and  Development have prepared the  EPA sections, and those  sections have  been reviewed in
accordance with EPA's  peer and  administrative review policies  and  approved for presentation and publication.

-------
                                            Preface


A wide array of effective water quality management and protection tools has been developed for urban
environments,  but implementation  is hindered by a  shortage of technology transfer opportunities.  Held in
Chicago,  Illinois on February 8-10,2000, the National Conference on  Tools for  Urban Water Resource
Management and Protection was designed to facilitate the educational process and transfer state-of-the-art
information to  state,  regional, and  local urban  water quality practitioners.

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 Office of Wastewater Management and
its Region 5 office, as well as the  Northeastern  Illinois Planning Commission. The conference was  conducted
in cooperation with the Water Environment Federation.  Over 450 attendees  participated, including
representatives  from  Australia,  Brazil,  Canada, Chile, New Zealand, and Turkey.

The timing for this conference coincided well with the  U.S. Environmental Protection Agency's release of the
NPDES Storm Water Phase II Final  Rule in October  1999. The conference provided participants with
practical,  applied information on the most effective tools and technologies for meeting these new  NPDES
permit requirements.  Program topics were  carefully chosen to  reflect the Phase II Program's six priorities:
public education, public  involvement, detection  and  elimination of illicit  discharges,  construction site runoff
control, post-construction storm  water  management,  and pollution  prevention for  municipal operations.

Two  special pre-conference workshops were  held on February 7. Better Site Design  and Storm Wafer
Management Techniques for Phase // Communitiesexplored the benefits  of alternative  urban  site design
approaches, as well as new advances in storm water  management to protect water resources. The workshop
was led by staff from the Center for Watershed Protection. The second pre-conference workshop,
Introduction to Urban TMDLs, examined current and  pending  requirements for total maximum daily load
(TMDL)  programs.  Instructors for this  workshop were staff from Tetra  Tech,  Inc.  Each  of the  workshops
attracted  over  135 participants.

This  Conference Proceedings includes  many  of the papers presented during  the  conference.  All papers
included were  peer  reviewed. Additional copies, in either paper  or  CD-ROM format,  are available free of
charge from the U.S. Environmental  Protection Agency, telephone 800/490-9198, or visit the web site
.

-------
                                           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
implement actions  leading to a compatible balance between human activities and the ability of natural
systems to support  and nurture life. To meet this  mandate, EPA's research program is  providing data and
technical  support for solving environmental problems today  and building a  science knowledge base necessary
to manage our ecological resources wisely, understand  how pollutants  affect our  health,  and  prevent or
reduce environmental 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  prevention and control of pollution to air,
land, water and subsurface  resources; protection  of  water quality in public water systems; remediation of
contaminated sites and ground water; and  prevention and control of indoor air pollution. The goal of this
research  effort is to catalyze development and implementation of innovative,  cost-effective  environmental
technologies;  develop scientific and engineering information needed by EPA to support regulatory and policy
decisions; and provide technical support and information transfer to ensure effective implementation of
environmental regulations  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 EPA's Office of Research and Development to assist the  user  community
and  to link researchers with their clients.
                                                       E.  Timothy Oppelt, Director
                                                      National Risk Management Research  Laboratory
                                                 IV

-------
                                               Contents
Jordan Cove Urban Watershed National Monitoring Project	,	1
      Melville P. Cote, Jr.
      U.S.  Environmental Protection Agency
      Off ice  of Ecosystem Protection
      Boston,  Massachusetts
      Dr. John Clausen
      University of Connecticut
      Bruce  Morton
      Aqua Solutions,  LLC
      Paul Stacey
      Connecticut  Department of  Environmental Protection
      Stan  Zaremba
      Connecticut  Department of  Environmental Protection

Sources of Phosphorous in Stormwater and  Street  Dirt from Two Urban Residential  Basins in
Madison, Wisconsin, 1994-95	9
      R.J.  Waschbusch
      U.S.  Geological  Survey
      Middleton, Wisconsin
      W.R. Selbig
      U.S.  Geological  Survey
      Middleton, Wisconsin
      R.T.  Bannerman
      Wisconsin Department  of  Natural Resources
      Madison, Wisconsin

Using Biological Criteria to Assess and Classify Urban Streams  and Develop
Improved  Landscape  Indicators  	32
      Chris 0. Yoder  and Robert J. Miltner
      Ohio  EPA, Division  of  Surface Water
      Ecological Assessment Unit
      Groveport, Ohio
      Dale White
      Ohio  EPA, Division  of  Surface Water
      Information  Resources  Management Section
      Columbus. Ohio

Getting Past the Obvious	45
      Bobbin B. Sotir
      Robbin B. Sotir  & Associates,  Inc.
      Marietta,  Georgia

-------
Protecting and Enhancing Urban Waters: Using All the Tools Successfully	53
      Eric  H.  Livingston
      Stormwater/NPS Management Section
      Florida Department of Environmental Protection
      Tallahassee, Florida

Structures to Enhance Fish Habitat in Urban Streams	69
      Douglas T. Severn
      URS  Greiner Woodward-Clyde
      Seattle,  Washington

Lessons Learned About Successfully Using Infiltration Practices ,	,	,	81
      Eric  H.  Livingston
      Stormwater/NPS Management Section
      Florida Department of Environmental Protection
      Tallahassee, Florida

Potential  New  Tools for the  Use of Tracers to Indicate Sources of Contaminants to
Storm Drainage Systems	97
      Robert Pitt, Melinda  Lalor, Jennifer Harper, and Christy Nix
      Department of Civil and Environmental  Engineering
      The University  of Alabama at Birmingham
      Birmingham, Alabama
      Donald  Barbe'
      Department of Civil and Environmental  Engineering
      The University  of New Orleans
      New Orleans,  Louisiana

Elimination of Illicit Connections in Coastal New Hampshire Spurs Cooperation and Controversy	110
      Natalie  Landry and Robert  Livingston
      New Hampshire Department  of Environmental Services
      Concord, New Hampshire

Using Collaborative  Problem-Solving to Protect  North Carolina's Coastal  Resources:
The Experience of the White Oak Advisory Board	,	,	117
      Leon  E. Danielson, C. Suzanne Hoover, and Christy A. Perrin
      North  Carolina State  University
      Department of Agricultural  and Resource  Economics
      Raleigh,  North Carolina
      Nancy M. White
      North  Carolina State  University
      School of Design
      Raleigh,  North Carolina
      Ron Elmore
      North  Carolina Department of Transportation
      Raleigh,  North Carolina
      Jennifer  L. Platt
      Town  of Gary
      Cat-y, North Carolina
                                                     VI

-------
Community Responses to Stormwater Pollution: Case Study Findings with  Examples from the Midwest	124
      George P. Aponte Clarke
      Policy Analyst
      Natural Resources  Defense  Council
      New York, New York
      Peter H. Lehner
      Chief, Environmental Protection Bureau
      New York State Attorney General's Office
      New York, New York
      Diane M. Cameron
      President
      Cameron  Associates
      Kensington, Maryland
      Andrew G.  Frank
      Litigation  Associate
      Paul, Weiss, Rifkind, Wharton, and Garrison
      New York, New York

Integrated Urban Stormwater Master Planning	132
      Eric Strecker, VP, and Krista Reininga, Sr.  Project Engineer
      URS Greiner  Woodward-Clyde
      Portland,  Oregon

Conservation Design: Managing Stormwater through Maximizing  Preventive Nonstructural Practices	147
      Wesley R. Horner
      Environmental  Management  Center
      Brandywine Conservancy
      Chadds Ford, Pennsylvania

Low-Impact Development  Design: A New Paradigm for Stormwater Management Mimicking and
Restoring the Natural Hydrologic Regime	,...  158
      Larry S. Coffman
      Associate  Director
      Department  of Environmental Resources
      Prince George's  County,  Maryland

A National Menu of BMPs for the Phase II  NPDES Storm Water Program	168
      James H. Collins
      Tetra Tech, Incorporated
      Fairfax, Virginia
      John A.  Kosco,  P.E.
      USEPA/Office of Wastewater Management
      Washington, D.C.
                                                    VII

-------
Determining Urban Stormwater BMP Effectiveness	175
      Eric Strecker
      URS Greiner Woodward Clyde
      Portland,  Oregon
      Marcus M.  Quigley
      URS Greiner Woodward Clyde
      Boston,  Massachusetts
      Ben R. Urbonas
      Urban  Drainage and Flood Control District
      Denver,  Colorado

Texas Nonpoint SourceBOOK is Now On-Line !	186
      John Promise, P.E.
      Director of  Environmental Resources
      North Central Texas Council of Governments, and
      Chair,  Water Resources Management  Committee
      American  Public  Works Association
      Keith  Kennedy
      North Central Texas Council of Governments
      Robert W. Brashear, Ph.D.
      Camp  Dresser & McKee, Inc.

A Comparison of the Long-Term Hydrological Impacts of Urban Renewal versus Urban Sprawl	192
      Jon Harbor
      Department of  Earth and  Atmospheric  Sciences
      Purdue University
      West Lafayette,  Indiana
      Bernie Engel
      Department of  Agricultural and  Biological  Engineering
      Purdue University
      West Lafayette,  Indiana
      Don Jones
      Department of  Agricultural and  Biological  Engineering
      Purdue University
      West Lafayette,  Indiana
      Shilpam  Pandey
      Department of  Earth and  Atmospheric  Sciences
      Purdue University
      West Lafayette,  Indiana
      K.J. Lim
      Department of  Agricultural and  Biological  Engineering
      Purdue University
      West Lafayette,  Indiana
      Suresh  Muthukrishnan
      Department of  Earth and  Atmospheric  Sciences
      Purdue University
      West Lafayette,  Indiana

Comparative Nutrient  Export and Economic Benefits of Conventional and  Better Site Design Techniques	198
      Jennifer Zielinski, Deb Caracao, and Rich Claytor
      Center for Watershed  Protection
      Ellicott City, Maryland

                                                    viii

-------
Predicting  Erosion Rates on Construction Sites Using the Universal Soil Loss Equation in
DaneCounty.Wisconsin	206
     J.D.  Balousek
     Erosion  Control Engineer
     Dane  County  Land Conservation  Department (LCD)
     A.  Roa-Espinosa
     Urban Conservationist
     Dane  County  LCD
     G.D. Bubenzer
     Professor
     University  of  Wisconsin-Madison
     Madison,  Wisconsin

Public Involvement Programs That Support Water Quality Management	214
     Josephine  Powell
     Wayne County Department of Environment
     Detroit,  Michigan
     Zachare Ball  and Karen Reaume
     Environmental  Consulting & Technology Inc.
     Detroit,  Michigan

The Water-Wise Gardener Program: Teaching Nutrient Management to Homeowners	222
     Marc T. Aveni
     Virginia  Cooperative Extension
     Manassas,  Virginia

Chicago Wilderness;  Toward an Urban Conservation Culture	,	225
     John  D. Rogner
     U.S. Fish and Wildlife Service
     Chicago Field Office
     Barrington,  Illinois

A Survey  of Resident Nutrient  Behavior in the Chesapeake Bay Watershed  	230
     Chris Swann
     Center for Watershed Protection
     Ellicott City, Maryland

Lawn Care and Water Quality: Finding the Balance	238
     Jerry Spetzman
     Minnesota  Department of Agriculture
     St.  Paul, Minnesota

San  Francisco Bay Area's Pesticide Toxicity Reduction Strategy ,	,	243
     Geoff Brosseau
     Bay Area  Stormwater Management Agencies Association (BASMAA)
     Oakland,  California

Administering the NPDES Industrial Storm Water Program at the Municipal Level	249
     Michael J. Pronold
     Bureau  of Environmental Services
     Portland, Oregon

-------
Lessons  Learned From Three Watershed-Sensitive  Development Demonstration Projects  In
The Great Lakes Basin	  256
      Sarah  Bennett  Nerenberg
      The Conservation Fund
      Great  Lakes Office
      Chicago,  Illinois

Continuous Deflection Separation  (CDS) for Sediment Control in Brevard County, Florida	264
      Justin  Strynchuk, John Royal, and Gordon  England,  P.E.
      Brevard County  Surface Water  Improvement
      Viera,   Florida

Use of Automated Technologies in Watershed Management Planning	  272
      Lake County Stormwater Management  Commission  (SMC)
      Libertyville,  Illinois

Sediment and Runoff Control on Construction Sites  Using  Four Application Methods of Polyacrylamide Mix ...  278
      A.  Roa-Espinosa
      Urban  Conservationist
      Dane County  Land Conservation Department,  and
      Assistant  Visiting  Professor
      Biological  Systems Engineering
      University  of  Wisconsin-Madison
      G.D.  Bubenzer
      Professor
      Biological  Systems Engineering
      University  of  Wisconsin-Madison
      E.S. Miyashita
      Former Project Assistant
      Biological  Systems Engineering
      University  of  Wisconsin-Madison

Construction Site Planning and Management Tools for Water Quality Protection	284
      Thomas Mumley
      California  Regional  Water Quality Control  Board
      San Francisco  Bay Region
      Oakland,  California

Regulating Sedimentation and Erosion Control into Streams: What Really Works and Why?  	291
      Seth R. Reice
      Department of Biology
      University of North Carolina at Chapel  Hill
      Chapel  Hill, North Carolina
      JoAnn Carmin
      Department of Public  Policy
      Virginia Polytechnic Institute and State  University
      Blacksburg, Virginia

-------
Effectiveness in Erosion and Sediment Control: New Initiatives in Indianapolis	, 296
      James Hayes  and Marcia Mathieu
      Marion County Soil and  Water Conservation District
      Indianapolis, Indiana
      Greg  Lindsey
      Center for Urban Policy and the Environment
      School of Public and Environmental Affairs
      Indiana  University
      Indianapolis, Indiana

Using Constructed Wetlands to Reduce Nonpoint Source Pollution in Urban Areas	  303
      Jon Harbor
      Department of Earth  and Atmospheric Sciences
      Purdue  University
      West  Lafayette, Indiana
      Susan Tatalovich
      FMSM Engineers, Inc.
      Columbus,  Ohio
      Ron Turco
      Department of Agronomy
      Purdue  University
      West  Lafayette, Indiana
      Zac Reicher
      Department of Agronomy
      Purdue  University
      West  Lafayette, Indiana
      Anne  Spacie
      Department of Forestry and Natural Resources
      Purdue  University
      West  Lafayette, Indiana
      Vickie Poole
      Department of Forestry and Natural Resources
      Purdue  University
      West  Lafayette, Indiana

Advanced Identification (ADID)  Techniques Used to Protect Wetlands and Aquatic Resources in  a
Rapidly Growing County	314
      Dennis W. Dreher
      Northeastern Illinois  Planning  Commission
      Chicago,  Illinois

Local Government Involvement in Mitigation Banking	322
      Lisa T. Morales
      Wetlands  Division
      U.S. Environmental Protection  Agency
      Washington,  D.C.

Massachusetts  Stormwater  Management Policy/Regulations: Development,  Implementation,  and  Refinement .  332
      Bethany Eisenberg
      Vanasse Hangen Brustlin, Inc.
      Watertown,  Massachusetts

-------
Implementation of Michigan's Voluntary Stormwater Permit — A Community Perspective	339
      Kelly A. Cave
      P.E. Director
      Watershed Management Division
      Wayne  County  Department of the  Environment
      Detroit,  Michigan
      Dale S. Bryson
      Camp Dresser & McKee
      Detroit  Michigan
      Kelly C. Kelly
      P.E., Project Engineer
      Canton  Township, Michigan
      Jack D. Bails
      Vice-President
      Public Sector Consultants
      Lansing,  Michigan

California's Model Urban Runoff Program (MURP): Urban Runoff Programs for Small Municipalities	349
      Jennifer  Hays
      Assistant  Engineer
      Public Works Department
      City of  Monterey,  California
      Cy Oggins
      Coastal Program Analyst III
      California  Coastal Commission
      Central  Coast District
      Santa  Cruz,  California

New Stormwater Treatment BMPs: Determining Acceptability to  Local Implementing Agencies	354
      Gary R. Minton
      Resource  Planning Associates
      Seattle,  Washington
      Paul Bucich
      Pierce County Public  Works
      Tacoma,  Washington
      Mark Blosser
      City of  Olympia  Public Works
      Olympia,  Washington
      Bill Leif
      Snohomish  County Surface  Water Management
      Everett,  Washington
      Jim Lenhart
      Stormwater Management,  Inc.
      Portland,  Oregon
      Joseph  Simmler
      Entrance Engineers
      Bellevue,   Washington
      Steven True
      Vortechnics
      Federal Way, Washington
                                                     XII

-------
By Any Measure	362
     Thomas R. Adams, P.E., Vaikko P. Allen, and Andrea Perley
     Vottechnics,  Inc.
      Portland,  Maine

Stormceptor Hydrology and Non-Point Source Pollution Removal Estimates	371
      G.  Bryant
      Stormceptor  Canada
      Etobicoke, Ontario,  Canada
      R.  Grant
      New England Pipe
      Wauregan,  Connecticut
      D.  Weatherbe
      Donald G. Weatherbe Associates
      Mississauga,  Ontario,  Canada
      V.  Berg
      Stormceptor   Corporation
      Rockville,  Maryland

NPDES Phase II Cost Estimates	,	383
      Andrew J.  Reese, P.E.
      Ogden  Environmental and  Energy  Services,  Inc.
      Nashville, Tennessee

The  Stormwater Utility Concept in the Next Decade (Forget the Millenium)   	397
      Hector J. Cyre
      President
      Water Resource Associates, Inc.
      Kirkland,  Washington

Are  Green Lots Worth More Than Brown  Lots? An  Economic Incentive for  Erosion Control on
Residential Developments	404
      Martha  Herzog and Jon Harbor
      Department of Earth and Atmospheric  Sciences
      Purdue  University
      West Lafayette, Indiana
      Keith McClintock
      Geauga Soil  and  Water Conservation District
      Burton,  Ohio
      John Law
      Indiana  Department of Natural  Resources and
      Saint Joseph Soil and Water Conservation District
      South Bend,  Indiana
      Kara Bennett
      Department  of  Statistics
      Purdue  University
      West Lafayette,  Indiana
                                                     XIII

-------
                                    Acknowledgments


The  success  of the conference and the preparation of this document 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 Co-Officer
U.S. Environmental Protection Agency, Region 5,  Chicago, IL

John A. Kosco, Project Co-Officer
U.S. Environmental Protection  Agency, Washington, DC

Deborah S. Caraco, Center for Watershed Protection, Ellicott City, MD
Martin  H.  Kelly, Southwest  Florida Water Management  District, Brooksville, FL
Thomas E. Mumley, California  Regional Water Quality Control Board, Oakland,  CA
Eric W. Strecker,  URS Greiner Woodward Clyde, Portland, OR

Peer Reviewers

Donald Brown,  U.S.  Environmental Protection  Agency, Cincinnati,  OH
Deborah S. Caraco and  Staff, Center for Watershed Protection, Ellicott City,  MD
Alan Everson, U.S. Environmental Protection  Agency, Cincinnati,  OH
Douglas Grosse,  U.S. Environmental  Protection Agency, Cincinnati, OH
Larry Shepard,  U.S. Environmental  Protection  Agency, Region 7,  Kansas City, KS
Steve Walker, Nebraska Department  of Natural Resources, Lincoln, NE
Ruth Wallace and Staff,  Missouri Department  of Natural Resources, Jefferson City, MO
Jennifer Zielinski and Staff,  Center for Watershed  Protection,  Ellicott City, MD

Technical  Editing and  Publishing

Scott Minamyer, Jean Dye, and Steve Wilson of the U.S. Environmental Protection Agency, Cincinnati, OH.
                                               XIV

-------
              Jordan Cove Urban Watershed National Monitoring Project
                                              Melville  P. Cote, Jr.
                                     U.S. Environmental Protection Agency
                                         Office of Ecosystem  Protection
                                            Boston, Massachusetts
                                                    02114

                                               Dr. John Clausen
                                            University  of Connecticut

                                                 Bruce Morton
                                             Aqua Solutions,  LLC

                                         Paul Stacey and  Stan Zaremba
                              Connecticut Department  of  Environmental  Protection
Introduction

  Stormwater runoff from urban and urbanizing areas is widely recognized as a major cause of water pollution in the
United States. The impacts of stormwater  runoff are threefold: (1) chemically, contaminants deposited  on the  land are
carried by runoff and infiltration to surface and groundwater; (2) physically, increases in impervious surfaces raise runoff
rates  which, in turn, increase mass pollutant loadings and contribute to erosion  and sedimentation; and (3) biologically,
the combined chemical and physical alterations of watershed systems degrade aquatic habitat. Research over the  past
20 years consistently shows a strong correlation between the imperviousness of a  drainage basin and the health of its
receiving waters, with stream health  decreasing with  increasing impervious coverage  of the watershed.' The U.S.
Environmental Protection Agency cites urban runoff as the second  leading cause of impairment to estuaries and the fourth
leading cause of impairment to  lakes.*  Increased runoff rates, and the erosion  and sedimentation associated with  new
development  and  construction, also  are significant  sources  of pollution. In the  United  States,  there are an estimated
522,000 construction "starts" each year, with construction activities disturbing an estimated 5 million acres of land
annually.3

  Connecticut communities,  like those in many  urbanized  states, are confronted  with meeting nonpoint source
management  needs that often  conflict  with traditional subdivision regulations  and construction standards. The challenge
of meeting  publicsafetyand maintenance requirements in an environmentally sensitive manner is not currently being met,
as evidenced by continued water quality impairments  associated with  new development. Can impervious  surfaces be
reduced, and curbing and storm drains be  eliminated in a way that will not raise objections from  municipal boards and
commissions? Will  homeowners accept cluster housing,  natural landscaping, and "greener" home and yard maintenance
practices? Most important, will those modifications  make a difference in the quality and quantity of nonpoint source runoff
under widespread application?  Answering these and related questions is the objective of the Jordan Cove Urban
Watershed  National  Monitoring Project.

Project  Overview

  The primary purpose of the Jordan Cove  project is to compare differences in runoff quantity and quality emanating from
traditional  and "environmentally sensitive"  development  sites.  The 18-acre  "Glen  Brook  Green"  subdivision, located  in
the southeastern Connecticut town of Waterford, is being constructed and  monitored  to make  this  comparison. The
subdivision is split into two distinct "neighborhoods": one with building lots arranged in  a traditional R-20 (half-acre)

-------
zoning  pattern (Figure 1); the other,  cluster housing with  a variety  of  best  management  practices (BMPs)  incorporated
into  the design (Figure 2).
 Figure  1. Glen  Brook Green "Traditional" Neighborhood.
  Figure 2. Glen Brook Green "BMP" Neighborhood.

-------
    Stormwater runoff from the traditional section is collected by curbs and  catch basins, then piped through astormwater
treatment system before entering Nevins Brook, a tributary of Jordan Brook and, ultimately,  Jordan Cove and Long Island
Sound. Homeowners will not be subjected to any enhanced  environmental education, or restrictions  on  how they manage
their  properties.

    The  BMP neighborhood will feature grass swales; roof leader "rain gardens;" shared, permeable driveways; small
building "foot-prints;"deed  restrictions on increasing impervious surfaces; "low-mow,""no-mow,"andconservationzones;
a narrower,  permeable road surface (interlocking concrete pavement); and a vegetated infiltration  basin,  or bioretention
area,  located inside a "tear-drop" cul de sac. Several  different driveway surfaces will be utilized,  including interlocking
concrete pavement,  gravel,  concrete tire strips,  and permeable asphalt,  and monitored for their  relative runoff rates.
Homeowners and town  road maintenance crews will be encouraged to adopt pollution prevention  techniques, including
controlled fertilizer and pesticide application,  pet waste management, street sweeping/vacuuming,  and reduced  use of
deicing  agents.

    The  BMP neighborhood is expected to generate less stormwater runoff and pollution. Monitoring conducted  before,
during and after construction will document  actual results.  The Jordan Cove project team comprises a true public/private
partnership,  with  researchers and  educators  from the  University of Connecticut; federal, state, and local government
officials;  private  consulting firms; and the developer.

National Monitoring  Program

    The  Jordan Cove Urban Watershed  National  Monitoring  Project is funded,  in part, through the Connecticut
Department  of Environmental Protection (CT  DEP) by  the  U.S.  Environmental Protection Agency's (EPA)  Section 319
National  Monitoring  Program (NMP). It is one  of 22 such  projects nationwide. The Jordan  Cove project is the only NMP
project studying the effects of residential subdivision development on runoff quality and quantity,  and of BMPs designed
to mitigate those impacts.

    The Section 319 NMP was established pursuant to section 319(1) of the federal Clean Water Act (Nonpoint Source
Management Programs -  Collection of Information). Section 319(1) states that EPA shall  collect information  and make
available:

      (1)  Information concerning the costs  and relative efficiencies of best management practices for reducing nonpoint
          source pollution.

      (2)  Data concerning the  relationship  between water quality and implementation  of various management practices
          to  control nonpoint sources of pollution.

    The objectives of the  Section 319 NMP are twofold:

      (1) To scientifically  evaluate the effectiveness of watershed technologies designed to control nonpoint source
          pollution.

      (2) To improve our  understanding of nonpoint source pollution.

    To achieve these objectives, the NMP has selected watersheds across the country to  be monitored over  a 6-to 10-
year period to evaluate how improved land management and the  application of BMPs reduce water pollution. The results
from these projects will be  used  to assist land  use and natural  resource  managers by providing information  on the relative
effectiveness of BMPs to  control nonpoint source pollution.

Site  Selection

    In 1993,  nonpoint source program staff from EPA and  CT DEP, and  a University  of Connecticut researcher began
efforts to identify a site at  which to  conduct a  nonpoint source monitoring project under the auspices of the NMP.  Initial

-------
site selection  involved three criteria:  (1) an  appropriate hydrologic setting,  with distinct drainage patterns amenable to
monitoring; (2) a willing land owner or developer who would allow 1-1  1/a years of advance monitoring before beginning
construction; and (3) a municipality willing to  adopt innovative site planning and development strategies. Proximity to the
coast was also considered as an important factor because of the need  to reduce nonpoint source pollution loads to Long
Island Sound  and coastal waters in general.

    CT DEP mailed letters soliciting interest to a number of municipalities recognized for either their progressive approach
to land use planning and  management, orforexperiencing high development rates.  After positive responses from several
municipalities,  and numerous field visits, the  "Glen Brook Green" site  in Waterford was selected in  May 1995. The 18-
acre parcel was an active chicken farm, but its owner, who had grown  up on the farm, was planning to develop it into a
residential  subdivision. The property  owner  wanted  to  develop  the  parcel  in an environmentally-sound  manner, was
interested in the NMP solicitation, and  was willing to be flexible with  his construction schedule to facilitate  monitoring.

    The hydrology of the  parcel featured two  distinct drainage areas, an ideal setting for the proposed monitoring design.
Poultry houses and several other buildings  occupied the  area  that would become the traditional neighborhood and an old,
partially mined gravel pit  dominated  the future BMP  neighborhood. Soil tests determined that  the chicken manure had
not elevated nutrient levels significantly  enough to bias  the monitoring. The  town  of Waterford,  and its planning  officials,
had a reputation as being  progressive on land use issues and had  served as one of the pilot communities for the
University of Connecticut  Cooperative Extension System's Nonpoint Education for Municipal Officials (NEMO)  project.
Because waivers from Water-ford's subdivision regulations would  be needed to build the BMP  neighborhood, the town's
cooperation was critical to  the  project's implementation.

Planning

    Proceeding from a conceptual design to actual construction required a concentrated effort by the  project team  working
together toward a common goal. Once an acceptable plan was agreed upon  by the project team and  committed to paper,
the next step was gaining approval from Water-ford's conservation, and planning  and zoning commissions. As is typical
of New England town  governments, both  commissions paid  close  attention to planning decisions at a series of public
meetings at which many development alternatives were reviewed. Volunteer commissioners  and professional  staff raised
numerous concerns  regarding the health,  safety and general welfare  of the town residents, and the social economic,
environmental,  and political viability of the proposed plan. Among their concerns were  road  widths for emergency access,
road surface integrity for plowing and de-icing, traffic, drainage, sidewalks, parking, maintenance of common  areas, and
responsibility should  BMPs fail.  The  rigorous review was enlightening  to the project team  and  commissioners alike. As
the ongoing dialogue between  the various parties  led to further  planning details and innovative solutions  to problems,
enthusiasm and support for the project grew.

    After a series of public meetings in late 1996 and early 1997, the project was approved by  both commissions.
Technical  modifications  of existing standards were handled  in four ways: as waivers, special design/operation controls,
mitigation, or  discretionary actions. Table 1  lists each of these categories with associated comments and concerns
expressed  by  Water-ford's professional staff and  commissions.  In the end, it was the willingness of all parties involved
to work in  concert, reaching compromises, that allowed this innovative project to  advance to the construction phase.

    It  is a generally  accepted axiom that resource-based site planning  can help minimize increases in runoff  and reduce
the potential for erosion and sedimentation problems typically associated with new development. In this project, goals
identified at the outset are helping to direct the choice of practices and strategies for site development toward those that
will reduce adverse impacts on hydrology and water quality. These  goals include: (1) reproducing pre-development
hydrological conditions;  (2)  confining  development and construction activities to  the least  critical areas;  (3) fitting  the
development to the terrain;  (4)  preserving  and utilizing  the natural  drainage system;  and  (5) creating a desirable living
environment.

-------
Table 1. Technical Modifications of Existing Development Standards.
Considerations
waivers needed
special design/
operational control
mitigation required
discretionary
actions
Traditional Design
specified road surface materials
typical road width = 28 feet,
reduced to 24 feet
curbs and storm drains required
90 ft paved cul-de-sac radius
planning and zoning standards
home owner discretion
home owner discretion
home owner discretion
road runoff piped to storm sewer
creation of 13,400 sq ft wetland at
subdivision entrance
R-20 single-family zoning
open space not contiguous with
all lots
a driveway for each home
BMP/Cluster Design
segmental concrete pavers
(permeable)
reduced road width to 20 feet
for travel lane
no curbs; grassed swales
and sheet flow off road
one way cul de sac design to
reduce road width and
turning radius
bioretention "rain gardens"
vegetative maintenance
pesticide management
domestic animal
management


cluster and zero setback
from lot lines
open space layout
combined driveways
Comments
must be approved by public
works; costs more
must be approved by public
works, fire, and police
turf stone installed to
maintain road edge
integrity; costs less
further reduction in width
and less need for snow
plowing
retains roof runoff on-site
reduces fertilizer use; costs
less
reduces pesticide use:
costs less
reduces pathogen runoff
need to manage storm
water entering the site from
adjacent public road
required to mitigate filling
of 5000 sq ft of wetlands
within subdivision
allows more open space
and natural landscaping
compact housing; natural
landscaping
reduces curb cuts and
impervious surface; cost
less
Monitoring Design

    This  study is utilizing the "paired-watershed"  monitoring  design, which requires a minimum of two watersheds (control
and treatment) and two periods  of study (calibration and treatment). This approach assumes that there is a quantifiable
relationship  between paired water quality data for the two watersheds, and that this relationship is valid  until a  major
change is made in one of the watersheds. It  does not  require that the quality and quantity of runoff be statistically the
same for the two  watersheds, but that the relationship between the paired observations of water quality and quantity
remains  the same over time - except for the  influence  of the land use changes in the  treatment watershed.4

    The  control  watershed  accounts  for  annual and/or seasonal climate  variations.  During the calibration  period, no
changes in  land use occur in the watersheds and  paired water quality and quantity data are collected to develop a
baseline. The paired data are  used  to develop regressions for the  control and treatment watersheds. The treatment
period begins when  changes in land use occur in the treatment watershed. A new regression is developed following the

-------
treatment period. Analysis of variance (ANOVA) is used to test the significance of the regressions in each period.
Analysis of covariance (ANCOVA) is used to test the differences between the two regression slopes and intercepts. The
changes between periods  are calculated  based on a  comparison of predicted values, using the calibration regression
equation,  and observed values during the treatment period.5

    For the Jordan Cove project, the treatment period will occur in two phases: (1) during construction of the traditional
and BMP neighborhoods; and (2) after construction when the  BMPs are in effect. The  paired-watershed approach is
being used  to measure the differences in water quality and quantity between  the treatment  areas (traditional and BMP
neighborhoods) and the control area (a nearby 1 0-year old subdivision) caused  by construction in the two treatment  areas
and the application of BMPs in the BMP  neighborhood. Stormwater quality and quantity are measured at the outlets of
each  of the two treatment neighborhoods, and the control  watershed  (Figure 3). Water quality is measured by analyzing
weekly flow-weighted  composite samples for total suspended solids (TSS), total phosphorus (TP),  total Kjeldahl nitrogen
(TKN), ammonia nitrogen (NH,-N), and nitrate+nitrite nitrogen (NO,-N). Grab samples are analyzed for fecal coliform and
BOD,. Monthly  analyses are conducted  for copper,  lead, and  zinc.
          Figure 3. Existing residential (control) watershed.
    The  calibration  period began in January  1996, to establish a baseline for future comparisons. Since the treatment
period began in May 1998, runoff monitoring has focused on the effects of construction, and on the relative effectiveness
of standard erosion and sediment control practices in the traditional neighborhood.  When construction commences in
the BMP neighborhood, the focus will be on the effects of construction and  the relative  effectiveness of enhanced  erosion
and  sediment control practices  (e.g., phased grading,  stockpile seeding,  open space vegetation,  cross  grading, and
detention swales). Post-construction monitoring is scheduled to begin in 2001  and will continue  for 3-5 years.

    Supplemental monitoring  will be  conducted on selected BMPs,  including different driveway  surfaces and  enhanced
turf management in the BMP neighborhood, and a  "state-of-the-art" stormwater treatment device  in the traditional
neighborhood. This  information will be used  to  evaluate the effectiveness of these  specific  practices.

-------
Monitoring  Results

    During the calibration  period, 75 runoff events were sampled for the control watershed and 12 runoff events for the
two treatment watersheds,  in the treatment period to date, 21 and 20 events were sampled for each treatment watershed,
respectively.  Peak discharge values were obtained for nine paired events in the calibration period and 20 pairs for the
treatment period. The total number of  samples analyzed was  less  than the total number of flow observations because
not all the samples contained a sufficient volume  for analysis6.

    Sampling results to date,  as presented in Table 2, indicate  that construction of the traditional neighborhood is causing
significant impacts on runoff quality and quantity, including observed increases in  mean weekly flow volume (99%), runoff
frequency (from 16%  to 95%), and mean weekly peak discharge (79%).7 The conversion  of the watershed's topography
from a "knoll" to a "bowl," combined with an increase in impervious surface, appears to have caused a significant change
in hydrologic responses. Concentrations of NO,-N and Pb in runoff also increased. However, increases in the
concentrations of sediment and sediment-associated nutrients, typical of construction sites, did not occur.  In fact, TKN
concentrations have declined  during construction.  It  is believed  that erosion  and sediment controls  are  responsible  for
TSS concentrations  remaining constant  before  and during construction*.


Table 2. Summary of means and percent increases of flow, Qp, nutrient and metal concentrations for the control and traditional watershed in the
calibration  and treatment periods.
Calibration Period

Parameter
Control
Traditional
Treatment Period

Control
Traditional
Observed
fms/umak\

Flow
113.84
0.14
107.76
1.94


Qp
0.05
3.00E-04
0.04
1.00E-03
	 	 	 	 	 „_ fmn/l ^ 	
	 i"ig"-;
TSS
NO3-N
NH3
TKN
TP
31
0.5
0.15
1.3
0.159
132
0.3
0.08
4.0
1.009
28
0.4
0.31
1.8
0.127
106
0.8
0.18
2.1
0.481
/nr,/! \ 	
	 ^g/i_; 	
Cu
Pb
Zn
8
6
58
8
11
65
14
6
79
21
17
126
Predicted


% Change

0.02
99"*

3.00E-04
79***

121
0.3
0.17
4.5
0.758
-15
62"
2
-113"
-58

13
10
98
38
42*
22
* P value < 0.05
" P value < 0.01
. ** P value < 0.001
    Coinciding with the increases in  pollutant concentration and  flow, the mass export of NO.-N and Pb increased as well,
as did the  mass exports of TP, TSS, Cu,  and Zn. These increases appear to be attributable to increased stormwater
                                                        7

-------
runoff volumes. The preliminary results from this study suggest that increased runoff, rather than erosion, is the  cause
of increased  pollutant export from this  construction  site. Traditionally,  erosion and sediment controls and  stormwater
management plans focus on the prevention of sediment and, occasionally, peakflow impacts on downstream areas. The
preservation of pre-development  hydrologic conditions within the watershed where construction is occurring is typically
ignored.
    Excess runoff, which is the driving force behind nonpoint source pollution, will transport pollutants into waterways and
contribute to their degradation. Preliminary  monitoring  results  demonstrate that  erosion  and sediment controls  can reduce
sediment and  sediment-associated pollutants in construction site runoff.  However,  current erosion  and sediment control
practices do not address the increase in runoff from development sites. Consequently, these practices fail  at  reducing
pollutant loads.9
Next Steps
    By the end of 2000, this combination of traditional and "green" designs for residential subdivisions should be fully
constructed.  Monitoring  of stormwater quality and quantity will be conducted  for several years after  build-out to determine
the overall efficiency of the  design. It should demonstrate that careful planning, landscaping, and use of vegetative  BMPs
can help protect and  enhance  the environment, while addressing other concerns that local  planning and zoning
commissions  face. Lessons learned from this project  have already been, and  will continue to be,  passed along to other
communities through ongoing technical assistance and training programs administered  by  the  CT DEP, the University
of Connecticut Cooperative  Extension System, and  other agencies and organizations.
References
1. Arnold, C.L. and C.J. Gibbons.  1996. Impervious Surface  Coverage: The  Emergence of a Key Environmental
    Indicator.  American  Planning  Association Journal. 62:2. Chicago, IL.
2.  USEPA.  1996.  National Water Quality Inventory. Washington, D.C. 20460
3.  US  Bureau of the  Census. 1996.1992 Census of Construction Industries. Manufacturing  and Construction  Division.
    Washington,  D.C.  20460
4.  Clausen, J.C. and J. Spooner.  1993.  Paired Watershed  Study Design. United  States Environmental Protection
    Agency. USEPA 841 -F-93-009.
5.  Engdahl,  J. 1999. Impacts of Residential Construction on  Water Quality and  Quantity  in Connecticut. University of
    Connecticut.  Storrs,  CT.
6.  Ibid.
7.  Ibid.
8.  Ibid.
9.  Ibid.

-------
      Sources of Phosphorus in Stormwater and  Street  Dirt from Two Urban

                   Residential Basins  in Madison, Wisconsin, 1994-95


                                      R. J. Waschbusch and W. R.  Selbig
                                           U.S. Geological  Survey
                                                Middleton,  Wl

                                               R.T. Bannerman
                                 Wisconsin Department of  Natural  Resources
                                                Madison, Wl


Abstract

    Eutrophication is a common problem  for lakes in agricultural and urban areas, such as Lakes Wingra and  Mendota
in Madison,  Wisconsin. This report describes a study to  estimate the  sources of phosphorus, a major contributor to
eutrophication, to Lakes Wingra and  Mendota from two small  urban residential  drainage basins. The Monroe Basin
empties  into  Lake Wingra, and the Harper Basin into Lake Mendota. Phosphorus data were collected from streets, lawns,
roofs, driveways, and parking  lots (source areas) within these two  basins and were used to estimate loads from each
area. In  addition to the samples collected  from these source areas,  flow-composite samples were collected at monitoring
stations  located at the watershed outfalls (storm sewers); discharge  and rainfall also were measured. Resulting data were
then used to calibrate the Source Loading  and Management Model (SLAMM, version 6.3, copyright 1993, Pitt & Vorhees)
for conditions in the city of Madison and determine within these basins which of the source areas are contributing the most
phosphorus.

    Water volumes in the calibrated model were  calculated to within 23% and 24% of those measured at the outfalls of
each of  the  basins. These water volumes  were applied to the suspended-solids and phosphorus  concentrations that were
used to  calibrate SLAMM for suspended-solids and phosphorus loads.  Suspended-solids loads were calculated to  be
within 4% and 17%, total-phosphorus loads within 24% and 28%, and dissolved- phosphorus loads within 9% and 10%
of those measured at the storm-sewer outfall  at the Monroe and Harper basins,  respectively.

    Lawns  and streets are the largest sources  of total and dissolved phosphorus in the basins. Their combined
contribution  was approximately 80%, with  lawns contributing more  than the  streets. Streets were the largest source of
suspended   solids.

    Street-dirt samples were collected  using industrial vacuum equipment. Leaves  in these samples were separated out
and the  remaining sediment was sieved into >250 um, 250-63 urn, 63-25 um, <25 urn size fractions and were analyzed
for total  phosphorus. Approximately 75%  of the sediment mass resides in the >250 um size fractions. Less  than 5% of
the mass can be found in the particle  sizes less  than 63 um. The >250  um size fraction also contributed nearly 50% of
the total-phosphorus mass. The leaf fraction contributed  an  additional 30%. In each particle size, approximately 25% of
the total-phosphorus mass is derived  from leaves or other vegetation.

Introduction

    Eutrophication is a common problem  for lakes in agricultural and urban areas, such as Lakes Wingra and  Mendota
in Madison,  Wl. Primary  productivity in northern temperate  lakes is most often limited by phosphorus  (Schindler 1974;
1977). Urban runoff has  been noted to contain high phosphorus concentrations (U.S. Environmental  Protection  Agency,
1983) that may be increasing the  eutrophication. The focus of the study described  in this  report was to estimate the
sources  of phosphorus to Lakes Wingraand Mendotafrom two small urban residential watersheds in Madison, Wl (Figure
1).  This  study was done in  cooperation with the city of Madison and the Wisconsin  Department of Natural Resources.

-------
        EXPLANATION

     (1) Monroe Basin
     (2; Harper Basin
     CJ Lakeland Basin Outfall
                                                                                         Madison  /
Figure 1. Lakes Wingra and Mendota in Madison, Wl.

Lake Mendota and Lake Wingra are both part of the Wisconsin Department of Natural Resources (WDNR) Priority
Watershed  Program (Betz and  others, 1997). State funding is available to help pay for management aimed at  reducing
the amounts of phosphorus and other pollutants discharged to the  lakes. The goal of the Lake Mendota Priority
Watershed  Project is to  reduce the frequency of nuisance algae blooms in the lake from one out of every two days to one
out of every five days. To accomplish  this goal, it is estimated that a 50% reduction  is needed in the amount of
phosphorus entering the lake. To help reach this target, the Nonpoint Source  Control Plan for the Lake Mendota Priority
Watershed  Project (Betz and others, 1997) set an objective  of reducing phosphorus loading to the lake by20% from urban
areas. The remaining 30%  reduction is  intended to  come from  rural  phosphorus management.

    For this study, phosphorus data were collected from five source areas-streets, lawns, roofs, driveways, and parking
lots-within the two drainage  basins from urban  residential and commercial areas to estimate loads from  each source
area (Table 1).  Resulting data were used to calibrate  the Source Loading and Management Model (SLAMM, version 6.3,
copyright 1993, Pitt & Vorhees)  for conditions in the  city of Madison and determine which  source areas are contributing
the most phosphorus within these  basins. The city is planning to use SLAMM to target specific source areas for
management efforts to meet the 20%  phosphorus-reduction objective of the priority watershed  project and to  meet
requirements of its Wisconsin  Pollutant  Discharge Elimination System stormwater permit.
                                                      10

-------
Table 1. Land-use Characteristics of the Monroe Basin, 1994, and Harper Basin, 1995, in Madison, Wl.
       [--, source not present]

             Monroe Basin                                         Harper Basin
Source area           Residential                Commercial                  Residential

               Acreage   Percent of basin   Acreage   Percent of basin    Acreage   Percent of basin
Lawn             119.8          51.5          ..            ..            23.6          57.4
Roof             26.5          11.4          2.3           1.0             5.4          13.2
Street            30.5          13.1           1.6           .7             5.3          13.0
Woodlot            --            --            ..            ..             3.0          7.3
Driveway          10.6          4.6           --            --             2.1          5.1
Parkland          19.3          8.3           -            ..             .7           1.7
Sidewalk          12.5          5.4           -            ..             .7           1.7
Parking lot          .4            .2           3.4           1.5             .3           .7
Railroad bed       5.3          2.3
Total             224.9          96.8          7.3           3.2            41.1          100


    Stormwater runoff samples from source areas and the basin outfall were collected  from a medium-density residential
watershed  draining to  Lake  Wingra from  May  to  November,  1994  and  from a  medium-density residential  watershed
draining to Lake  Mendota  from June to November 1995. These  runoff samples were  used to  estimate the phosphorus
and  suspended-solids load that  each of these source areas  and  basins contributes.  In addition,  a  third basin, the Lakeland
Basin that drains into  Lake Monona, was  monitored  for lawn  runoff in 1995.  This basin, which  encompasses an older
section of Madison, was sampled  in an  attempt to  determine whether any difference exists between this section of the
city (which has older, smaller lawns) and other areas of the city.

Study-Area  Description

    The Monroe Basin, monitored  during 1994, drains into  Lake Wingra. The basin is 232.2 acres, of which 224.9 acres
are residential and 7.3 acres are commercial (Figure 2). Lake Wingra has a surface area of 338.9 acres (1.37 km2) and
a drainage area of 3,889 acres (15.74  km*). About  75% of the  Lake Wingra drainage basin is urbanized and  about 25%
is  composed of forest,  prairie,  and marsh  within the  University of Wisconsin Arboretum (Oakes  and  others  1975).

    The Harper  Basin, monitored during 1995, drains into  Lake Mendota.  The Harper  Basin is  41 .1 acres, all of which
is  residential land use  (Figure 2). Lake Mendota has a surface area of 9,859 acres (39.9 km2)  and has a  drainage area
of 138,823 acres (561.8 km2) (Lathrop and others, 1992). Approximately 20% of the Lake Mendota drainage area is
urban,  57%  is agricultural,  and  the  remaining 23% is grassland/woodland/marsh/open-water area  (Betz and  others, 1997).
The Lakeland Basin is approximately  3 miles southeast of the Harper Basin.

    In addition to the samples  collected  from source areas, a  monitoring station was located  at each  basin storm-sewer
outfall to collect flow-composite samples and to monitor discharge and rainfall. The total rainfall amounts for the  months
of June through October 1994 at Monroe and July through October 1995 at Harper were 9.24 and 10.67 in., respectively.
These amounts are 64%  and 71% of the average from  1961 to  1990 (Brian Hahn, National Weather Service, oral
communication,   1997).

Acknowledgments

    We thank all of the volunteers  that allowed  us  to install our sampling  equipment in their yards, the City of Madison
Engineering Division and  Department  of Public  Health for their efforts in  making  this  study  possible,  Jeff Beck of the
USGS for his exceptional field efforts,  and Holly Ray and  Dr.  Bob Pitt at the University of Alabama at Birmingham for
their work analyzing  the street-dirt  samples. Also, we thank Mary Anne  Lowndes of the WDNR and
                                                        11

-------
                            Troy Or
                                                                    t>.05      0.10 MtlE
                                                                       	r
                                                                       0.10 KILOMETER
       EOTANA
        j  Residential roof-
          Commercial  roof
         Lawn-
        1 Park-
        1 fctfdlot-
         Sidewalk
         Driveway
         Parking lot
        Railroad'
        High-traffic road
        Medium-traffic road
         Low-traffic road-
         with grass ditch
         Low-traffic road-
         without curb and gutter-
  f llJij Low-traffic road-
        with curb and gutter
            0      0.1      0.2    KILOMETER
Figure 2. Harper and Monroe Basins.
Huron Hill
                                                                   12

-------
Steve Corsi of the USGS for insightful comments that have greatly improved the report. Lastly, we thank Gail Moede and
Aaron Konkol  at the USGS for their help with the report  editing and preparation of illustrations.

Data-Collection  Equipment and  Methods

    Runoff samples  were collected from  each source area by  use of sampling equipment slightly modified  from that
described by Bannerman and others (1993).  Brief descriptions  of the sampling  equipment follow.

    Street samplers. The street samplers were grouted into the street approximately 5 ft from the curb (Figure 3). The
sample  bottle was covered with  a  6-in. concave polycarbonate cap, set flush  with the street surface, with a  center drain
hole. The bottle and  cap were placed into a 6-in. diameter polyvinyl chloride (PVC) sleeve. Water flowed over the top of
the cap and drained through the center hole into a collection bottle.  The drain hole could be constricted by a  set screw
that controlled the flow  rate into the sample bottle.
          3/8-inch thread
        for removing cap
             Concave •
          polycarbonate1
               cap
(PVC) coupling

Polycarbonate set •
screw used to adjust1
size of drain hole

1.5-liter sample  bottle

 &inch-diameter
PVC sleeve

Buick-set grout
                                                                                    Undisturbed asphalt
                                                                                    Asphalt subgrade •
                                                                                    material
Figure 3. Schematic of street samplers.

    Driveway samplers.  Runoff water from driveways was  diverted into a sampler by means of a flat  piece of clear
plastic, 1/4 in. high by 1 in. wide by 3 ft long, glued to the surface of the driveway. The sampler consisted of a 1.5-L glass
bottle placed in a 10-in.-diameter protective PVC sleeve set into the ground next to the driveway. A 1/2-in.-diametersilicon
tube carried the runoff through a plastic cap covering the PVC sleeve and into the  sampler. During the 1994 field season,
the tubing emptied directly into the sample bottle, causing several sample bottles  to overfill. To alleviate this problem, in
1995, the tube emptied onto a polycarbonate cap  like  those  used with the street  samplers, so that the volume of water
entering  the  sample bottle could be controlled.

    Lawn samplers. Lawn sample bottles received runoff through two 5-ft pieces of 1/2-in.-diameter PVC pipe  placed
flush with the surface of the  ground, on  a sloping surface, with an  angle of about 150 degrees between  the two pipes
(Figure 4).  Runoff entered  the pipes through two  slits cut the entire length of pipe.  Each pipe was wrapped  with  fiberglass
                                                        13

-------
     1/2-inch
     WC cap
                                                                                                 Wnch-thick 4-1/3-inch
                                                                                                 (diameter gray PVC cap
                                                                                                 with V-notch cut out to
                                                                                                 accommodate the
                                                                                                 tygon tube.
                                                   Direction of water flow
1/2-inch PVC tubing  TOD Vi 6W
wrapped in nylon
screening
              Two rows of 2.5-mcn
              slots plunge cut
              with a table saw to
              allow the water to
              enter tube.
                                                                                                Tygon tubing
                                                                                                stretched over
                                                                                                the 1/2-inch
                                                                                                PVC tube to
                                                                                                carry water to
                                                                                                the bottle.
                                                                                      4-inch Schedule
                                                                                      40 PVC
                                                                                           iSide Vievx
Figure 4. Schematic of lawn samplers.

screen to prevent insects  and large debris from  entering.  Wooden  clothespins with  small  pieces  of nylon rope  held the
pipes in  place. Water from the pipes flowed into  a sampler through a notched cap. The sampler was  a 1- qt glass bottle
placed in a 4-in.-diameter protective PVC sleeve. The  cap had a  notch to accommodate  silicon tubing, which ran from
the end  of the PVC  collector pipes to the sample bottle.

    Roof samplers. Roof  samplers were designed to divert a small  portion of the  water in the gutter  downspout to a
sample bottle.  A 1/4-in.-diameter vinyl tube was attached to the inside of the downspout  by means of plastic wire ties.
Each tube went into a 1.5-L glass  sample bottle that was placed  in  a covered 10-in.-diameter protective PVC sleeve.
Because of problems  with overfilled  sample bottles, the design  was changed in  the same manner as the driveway
samplers so that the volume of water entering the sample bottle  could be controlled.

    Parking  Lot sampler.  The parking lot sampler collected runoff entering a storm-sewer inlet grate.  A small portion of
the inlet flow was diverted to a sample bottle by  means of a 6-in. trough made of 1/2-in.-diameter PVC pipe cut length-
wise  and held  in place with stainless steel  hose clamps attached to  the inlet grating. Water drained  from the trough
through a tube to a 2.5-gal glass sample bottle hanging from the inlet grate. No samples were collected from parking lots
during 1995.
                                                          14

-------
    Basin storm-sewer outfall samplers. An  automated sampling  station was placed  at the storm-sewer outfall of both
the Monroe and Harper Basins. In 1994, water level in the basin storm- sewer outfall pipe was measured with a  pressure
transducer as water drained into a detention pond. Velocity was measured with  an  electromagnetic velocity meter.  In
1995, stormwater-runoff volumes were computed using a modified Palmer-Bowlus flume design (Kilpatrick and others,
1985). The water level was measured one pipe diameter (36 in.) upstream from the  entrance to the flume using a
pressure transducer connected to a nitrogen bubble  system.  This water level was  used in the following equation to
calculate the total discharge through the flume:

          Q = a[Ha /D] "D2 5,

where

    Q is discharge,  in cubic feet per second,
    a is a constant, 3.685,
    b is a constant, 1.868,
    Ha is the water  level above  the upstream lip
    of the flume, at a  distance of one pipe
    diameter upstream from the flume
    entrance, in feet,  and
    D is  pipe diameter, in feet.

    Flow-composite  water quality samples were collected using programmable, refrigerated automaticsamplerswith 3/8-
in.-diameter Teflon-lined sample  tubing.  Rainfall was measured using a tipping-bucket rain gage and  was  recorded by
a digital data logger.

Stormwater Sample Collection and Processing Protocols

    Sample bottles were placed  in the source-area  samplers as close to the start of each rain event as possible. As the
bottles were being deployed, the  sampling equipment was rinsed with deionized water to remove any accumulated surface
dirt. Before the lawn sampler pipes were  rinsed, they were cleaned with a small test tube brush. As soon as possible after
runoff had stopped, the sample bottles were collected and the approximate volume of water in each bottle was recorded.

    All the bottles from a given source area were composited by pouring water from each bottle into  a  5-gal or l-gal
stainless steel, Teflon-coated churn splitter modified  from the type described in Ward and Harr (1990). The City of
Madison  Department of Public Health Laboratory analyzed a small subsample taken from the churn for suspended solids
and  phosphorus.

Street Dirt Collection  and Analysis

    In addition to stormwater-runoff samples, samples of street dirt  were collected with a 9-gal wet-dry shop-vacuum using
a 6-in.-wide wand.  A section  of the street was vacuumed from curb  to curb, 10 times across each of 3 streets  in the basin,
similar to the technique described by Pitt (1979) and Bannerman (1983). Monroe Street, Glenway Street, and Seneca
Place/Spring Trail/Huron Hill (the latter three are considered one residential street) were sampled during  1994 (Figure
2). Woodward Drive, Nova Way, and Luster Avenue were sampled during 1995. Woodward Drive did not have curbs, so
the sample was collected by vacuuming  between  wooden 4-by 4-in. blocks  placed at the edge of the asphalt on each side
of the street. During the fall, leaves on the street would  often plug the vacuum hose. To alleviate  this problem, a 6-in. by
2-ft wooden frame was placed with the 6-in. side abutting the curb. Before vacuuming the inside of the frame, the leaves
inside it were collected by hand  and placed in the vacuum collection bag. Then the street was vacuumed in the normal
manner.

    The dirt samples were dried  at 105°C, sieved into size fractions of >250 jjm, 250-63 pm, 63- 25 pm, and <25 urn and
weighed. The sieved samples were  sent to the University of Alabama at  Birmingham (UAB) for phosphorus analysis.  In
addition to total phosphorus, samples collected from the Monroe Basin were analyzed for percentage of vegetative
material. Two independent methods were used to determine the percentage  of vegetative material:  thermal

                                                     15

-------
chromatography and microscopic examination. In thermal chromatography,  dirt  samples  were placed in ovens at
increasing temperatures and the mass that was lost to incineration was determined after each increase in temperature.
The  mass  loss was compared to  the standard  temperatures where various substances like  leaves,  rubber, and paper
burned  off. The sample  mass  lost at the  temperature  range corresponding to  the  leaf standards was assumed to be
vegetation. In microscopic examination, samples of the dirt were compared to microscopic pictures of vegetation.  More
details of these  methods can be found in  Ray (1997).

Sources of Phosphorus

Measured Concentrations  in Runoff from Source Areas

    A total of 25 runoff events were monitored at each basin. Runoff samples were  collected  from May to November of
1994 at Monroe  and from June to  November of  1995 at Harper and  Lakeland.  Driveway samples  collected from the
Monroe Basin were excluded because of problems with the sample bottle overfilling  (discussed in the methods section).
Summary statistics  are listed in Table 2.

Table 2. Concentrations for Suspended Solids, Total  Phosphorus, and Dissolved Phosphorus at the Monroe Basin and Harper Basin, 1994-95
[--, concentrations were not used because of problems with the samplers; -, source area not present in basin; mg/L, milligrams per liter]
     Statistic
                  Lawns
   Geometric mean     59
 Coeff. of variation      .55
      Mean          85
     Median          75

   Geometric mean    0.79
Coeff. of variation       .62
Mean               1.03
Median               .99

   Geometric mean    0.37
 Coeff. of variation      .62
      Mean          .52
     Median          .61
   Geometric mean     122
Coeff. of variation       .37
Mean                132
Median               154

   Geometric mean     1.61
Coeff. of variation       1.12
Mean               2.34
Median               1.54

   Geometric mean    0.77
Coeff. of variation       1.51
Mean                1.54
Median               .81

Feeder
Street


68
1.17
99
60

0.40
1.24
.75
.31

Collector
street


51
.97
67
46

0.22
1.23
.36
.16
Source area
Arterial Driveways
street
MONROE BASIN
Suspended solids (mg/L)
65
.92
83
64
Total phosphorus (mg/L)
0.18
1.15
.24 ~
.17 ~

Parking
lots


51
1.27
82
44

0.10
1.04
.14
.09

Pitched
roofs


15
.95
85
18

0.07
.76
.09
.06

Flat
roofs


18
1.21
35
20

0.13
.96
.2
.12
0.16
1.72
.40
.14
 69
 .68
 98
 88

0.24
 .75
 .31
 .22

0.08
 .98
 .12
 .08
0.05
1.47
.14
.04
Dissolved phosphorus (mg/L)
    0.03
    1.20
    .05         '.'.
    .03
     HARPER  BASIN
  Suspended solids (mg/L)
                34
                .93
                57
                31
  Total phosphorus (mg/L)
               0.18
                .80
                .24
                .20
Dissolved phosphorus (mg/L)
               0.07
                1.0
                .11
                .07
0.02
1.24
.04
.02
0.02
1.22
.03
.02
                                             17
                                             .96
                                             25
                                             17

                                            0.15
                                             .68
                                             .19
                                             .15

                                            0.08
                                             .83
                                             .11
                                             .07
0.02
1.24
.04
.02
    The  concentration  data from the Monroe  and Harper Basins seem to exhibit  log-normal distributions that are
consistent with  urban-runoff concentration datacollected  during the Nationwide Urban Runoff Project (U.S. Environmental
Protection Agency,  1983).  In such cases, the geometric mean is a better estimate  of the central tendency than the
                                                        16

-------
arithmetic mean because the  geometric  mean  gives less weight to extremes (Helsel,  1992). Several of the coefficients
of variation in Table 2  have a  value greaterthan 1, indicating substantial variability in concentrations within a source area.

   The large variation  seen in the source-area concentration data could cast doubt on the predictability of the data. For
lawn runoff, the difference between geometric  mean phosphorus concentrations from 1994 to 1995 was greater than a
factor of 2; however, the lawn-runoff data collected from the Lakeland Basin are remarkably similar to data collected at
Harper (Table 3),  indicating that the variation in lawn-runoff phosphorus concentrations is not random. Primary sources
of phosphorus,  such as tree canopy, also could have a  large effect on the  source-area concentrations measured between
basins.  Figure 5 illustrates a trend  between the concentration of phosphorus  and the  percentage of overhead tree canopy
on streets for the  Monroe and Harper Basins. Canopy in the Monroe Basin tends to be less than 35%, whereas the
percentage of canopy in the  Harper Basin ranges from 5 to 78%. Variation also could be caused by meteorological
factors  like rain  depth, intensity, or inter-event period or by seasonal variables.

    Roof runoff had the lowest geometric mean concentrations of  suspended solids, and  lawn  runoff had  the  highest total
and dissolved phosphorus concentrations in both the Monroe and Harper Basins (Figures 6 and 7). In addition, patterns
in geometric  mean  concentrations between  source  areas within  each basin  were similar;  however, their magnitudes  were
very different. The geometric mean concentration  of phosphorus for low-traffic streets  in the Monroe Basin were  about
twice those at the  Harper Basin. Conversely, geometric  mean phosphorus concentrations for  lawn and roof runoff in the
Harper  Basin were more than twice  as  high as those in the  Monroe Basin. The beginning  of the sampling periods differed
by one month between basins (Monroe Basin in May and  Harper Basin in June),  and this difference may  have  caused
some of the  variability.

    Concentration  results for suspended solids  and phosphorus from earlier source-area studies  in Madison,  Wl,
Marquette, Ml, and  Birmingham, AL (Bannerman and others, 1993; Steuer and others, 1997; Pitt and others, 1995), were
compared to  the concentration results from this study. Suspended-solids concentrations in street-runoff samples collected
during  the other studies were considerably higher than those in samples collected forthis study. Sandier soils are present
in Marquette that could partially  account  for this difference. Furthermore, some of the same  lawns in the Monroe  Basin
were monitored for phosphorus concentrations  in the  previous Madison study (Bannerman and others, 1993), and  both
the dissolved and  total phosphorus geometric means calculated for that study were more than three times higher than
those in 1994. Because phosphorus concentrations varied  highly  from  Monroe and Harper Basins, did not closely  agree
with each other, and did  not agree well  with previous  studies, the geometric mean of the  combined data collected at
Monroe and  Harper Basins was used  for the modeling  phase of this  study.

Tables.  Rainfall Amountsand Intensities and Total-phosphorus Concentrations from Lawn-runoff Samples for Harperand Lakeland Basins, Madison, Wl,
1995
                   HARPER
                                                           LAKELAND
Start of rain  Rainfall   Intensity  Total-phosphorus    Rainfall    Intensity  Total-phosphorus
event
(date)
06/26/95
07/05/95
07/15/95
07/22/95
08/16/95
08/16/95
08/28/95
10/19/95
(inches)
0.26
.36
.50
.79
.61
.38
.80
.32
(in/hr)
0.12
.62
.12
.10
.43
.29
.19
.07
concentration
(mg/L)
10.72
1.32
3.61
1.08
1.82
.60
1.39
2.24
(inches)
0.31
.10
.80
.80
.55
.55
.67
.33
(inhr)
0.17
.16
.10
.09
.94
.49
.15
.06
concentration
(mg/L)
9.05
2.06
2.99
1.35
2.48
.58
1.58
2.59
                                                        17

-------
                                                  DISSOLVED PHOSPHORUS
              0.35
               0.9
               0.6
               0.7
                                                      TOTAL  PHOSPHORUS
             I 0.4
             z
             tn

             o 0.3
             i
             o
             £0.2
                 0
                0
                                                               linear regression
                                                                                              R  =0.94
10
             20
                         30
                                     40
                                                  50
                                                               60
                                                                           70
                                                                                        80
Figure 5. Trend between concentration of phosphorous and percentage of overhead tree canopy on streets for Monroe and Harper Basin.
                                                                 18

-------
                                                              Monroe Basin
                           DISSOLVED PHOSPHORUS
                                                                                         DISSOLVED  PHOSPHORUS
3
2.5
2

1.5
1

0.5
0
I i I I I i I





1 SK 1
Lawns Low-Traffic* Med. -Traffic
0.25
£j 0-2
§ g
g £ 0.15
z 5
UJ <
ll 0.1
o z!
s
^ 0.05
0
1 1 1 1 i i i 1 1
-

-
-

1 1 1 1 1 1 1 1 1
High-Traffic Parking Pitched Fiat
                                     Streets      Streets
                                 SOURCE AREA
  Streets    Iot5     Roofs     Roofs
            SOURCE AREA
                              TOTAL PHOSPHORUS



LU
^5
g|

UJ -^
§1
s
s



3.5
3
2.5

2
1.5

1

0.5
n
I I I I I I I
-
-
- -

r ~
-
^
-
Js
« %
'• I I I I I I I
                                                                                           TOTAL  PHOSPHORUS
       8
                           Lawns     Low-Traffic.   Med.-Traffic-
                                      Streets      Streets
                                 SOURCE AREA
£^
LlJ
-V t
\ — UJ
z 2
Uj <"
O Q£
8d
*

1.0
1.4 ,
OS-
0.6

0.4

o :

i i i i i i i i i





*
i i i i i i i i i
High-Traffic  Parking'    Pitched      Flat
  Streets    lots      Roofs     Roofs
            SOURCE AREA
600
500
E£ 400
£ 300
s
| 200
,_j
1 100
0
SUSPENDED SOLIDS
I I I I i I i I I I I I I
-
_

-
« * a « «
i I I i I I I I I i I SS I % \
iauns Low-Traffic* Med-Traffic- High-Traffic Parking Pitched Flat
Streets Streets Streets Iot5 Roofs Roofs

EXPLANATION
Source-area
concentration
S3 Geometric mean


                                                    SOURCE AREA
Figure  6. Dissolved phosphorous concentrations in Monroe Basin.
                                                                    19

-------
      DISSOLVED PHOSPHORUS


      9
                                   *
                                 Lawns

                            SOURCE ARM
                                                            Harper  Basin
0.45



 0.4



0.35



 0.3



0.25



 0.2



0.15


 01



0.05


   0
                                                                                   DISSOLVED PHOSPHORUS
                                                                                         !
                                                                 Streets      Driveways


                                                                         SOURCE AREA
                                                                                                            Roofs
                         TOTAL PHOSPHORUS
                                                                                     TOTAL PHOSPHORUS
              z 2
              UJ <
              O fi/
                    12
                     10
                              I
                                  Lawns

                             SOURCE AKEA
       200



 oi
 UJ

:5     150
              §1    i°°
              " c£
              o§
              u d
                1     50
                                                            2
                                                            z
                                                   1


                                                  0.9


                                                  0.6


                                                  0.7


                                                  0.6


                                                  0.5


                                                  0.4


                                                  0.3


                                                  0.2


                                                  0.1
                                                  SUSPENDED SOLIDS
                                                                                         !
                                                                 Streets      Driveways


                                                                          SOURCE ARE^A
                                                            i
                                     Lawns        Streets     Driveways


                                                     SOURCE AREA
                                                                             Roofs
                                                                                                            Roofs
                                                                                                    EXPLANATION
                                                                                                         Source-area •

                                                                                                         concentration


                                                                                                       Geometric  mean
Figure 7. Dissolved phosphorous concentrations in Harper Basin
                                                                   20

-------
Calibration of the Source Loading and  Management Model (SLAMM)

    A concentration data base to simulate stormwater quality and theoretical runoff coefficients to simulate runoff volumes
is used in the Source Loading  and  Management Model  (SLAMM). Because large amounts of concentration data and
runoff information were collected during this study, it was an opportunity to calibrate the model's concentration data base
and improve the runoff coefficients with data collected from the Monroe  and Harper Basins.

    Calibrating SLAMM with concentration  and water-volume data was a three-step process. First, the runoff volume
generated by each source area was calibrated  (a critical step because an accurate water volume is essential for
estimating all pollutant loads). Second, sediment was calibrated because sediment concentrations and loads are used
in SLAMM to estimate phosphorus loads. The final step  was to calibrate the model for phosphorus concentrations.

    A systematic procedure was used to calibrate suspended-solids and  phosphorus concentrations in SLAMM. First,
a mass-balance approach compared total measured loads from source areas  summed over 25 events to the loads
measured at the outfall. Monitored loads from source areas were calculated using SLAMM-generated water volumes.
Individual source areas were not equipped to measure runoff volumes during an event. Therefore, the accuracy of source-
area volumes, as assigned by SLAMM, was subject to agreement with the actual  volumes  measured at the outfall. If the
sum of all source-area volumes closely matched what was measured at  the outfall, the individual source-area volumes
assigned by SLAMM were assumed to be correct.  Second,  source-area concentrations were adjusted to optimize the
mass balance. SLAMM was adjusted after  agreement between  measured source-area and  outfall loads was achieved.

Water-Volume  Calibration

  Water-runoff volume from each source  area for  each rain event is calculated with the model. These calculations are
based on the amount of rainfall and a runoff coefficient developed  for various rainfall depths for each source area. Source-
area characteristics such as imperviousness, connectedness (amount of impervious area directly connected to the storm
sewer), and infiltration rates on pervious areas were used to  develop runoff coefficients (Pitt, 1987). The volumetric dis-
charges for each source area are then summed for each  event.  The total runoff volume can be decreased in the model
by using control  measures,  such as infiltration devices.

    SLAMM was used to  estimate runoff volume for the 25 storm events  from all source areas in each basin. The sum
of the volumes from all of these source areas was compared  to the volume measured at the basin storm-sewer outfalls
for these 25 events.  Initially, the model overpredicted the water volumes measured at Monroe by a total of 55% (over the
entire study period), whereas it  underpredicted those measured at Harper by only 2%. To obtain a  balance of
overprediction and  underprediction between the basins, the  runoff coefficients were  adjusted. Historically, more
measurements have been made for runoff from impervious surfaces (Pitt, 1987)  and more than 50% of the area within
each basin is pervious, mostly because  of residential lawns. Therefore,  it was  decided that the runoff coefficients  for
pervious  areas were more uncertain and model  calibration could  benefit from minor adjustments.

    Two  sets of runoff coefficients are available for pervious areas; one is designed to represent clayey soils, and the
other represents sandy soils. The predicted water volumes  mentioned above were  determined  using the runoff coefficients
for clayey soils (based on soil  maps). Changing the pervious classification from clayey to sandy resulted in SLAMM
underprediction of water volumes; approximately a 4% and a42% underprediction at the Monroe and Harper Basin storm-
sewer outfalls, respectively. A much better agreement was achieved at Monroe by assuming that the original soil
classification was incorrect. Sandy and clayey runoff coefficients, available to  the model, probably represented two
extremes, and more  realistic runoff coefficients fell  somewhere  between these two coefficients.

    Lawn-runoff data collected from Monroe and Harper Basins were  used to create  runoff coefficients that more
accurately represent the  pervious conditions found in Madison. First, the rainfall depth  sufficient to initiate runoff in
SLAMM was changed using data on the  amount of stormwater in  the lawn-sample bottles after each event. For  rainfall
amounts less than approximately 0.3 in.,  the bottles were less than 10% filled. From this  observation,  0.3 in. was
established as the minimum precipitation  required  to initiate runoff. However, the  runoff coefficient table for  clayey soils
                                                     21

-------
used in SLAMM resulted in 10% runoff for a rainfall depth of 0.2 in. Hence, SLAMM was changed to initiate runoff at 0.3
in. rather than 0.2 in. of precipitation.

    In addition to the change described above, a trial-and-error approach was used to change the coefficients  until
optimum agreement was  reached  between water volumes predicted in SLAMM and those measured at the Monroe and
Harper storm-sewer outfalls.  The resulting coefficients were  between those  for sandy and clayey soils and  were
approximately two-thirds the value for clayey soils. Figure 8 shows how the  new "Madison" runoff coefficients compare
to the  sandy and clayey coefficients.
             0.45
              0.4
             0.35
              0.3
             0.25
          8   0-2
          o
          I  0.15
              0.1
             0.05
                         0.5
1.5       2      25       3

       PRECIPITATION, IN INCHES
                                                                           3.5
Figure 8. Madison runoff coefficients compared to sandy and clayey coefficients.

    Based on the  revised Madison runoff coefficients, SLAMM overpredicted storm-sewer outfall volumes at Monroe by
23% and  underpredicted  Harper storm-sewer outfall volumes by 24% (Table 4). The Madison coefficients also produced
consistent lawn-runoff contributions in both  basins, approximately 20%  of the total volume. It is expected that the
percentage of lawn contribution  would  be similar for both basins because they have nearly the same percentage of lawn
area.
                                                       22

-------
Table 4. Percentage Difference in Cumulative Modeled Water Volumes Compared with Measured Outfall Water Volumes Using Three Soil Types,
Madison,  Wi.

  Basin     Sandy soils    Clayey soils    Madison soils

  Monroe        -4            55            23
  Harper        -42           -2            -24


Sediment  Calibration

    Once the runoff volumes were calibrated, SLAMM was used to estimate  sediment loads for the 25 events in  each
basin. With the exception of streets, a data base of sediment concentrations for each source area is used in SLAMM and
these concentrations are applied to  the water volumes to derive a load. Sediment concentrations from streets are
computed by a wash-off function  that is related to a street-dirt accumulation  rate.

    A mass-balance approach was used to  test the source-area concentrations  within each basin with those measured
at the storm-sewer outfall.  Source-area loads were computed by multiplying the water volumes produced from SLAMM
by the concentrations measured at the source areas for each event and  then summing these event loads. Sidewalks and
woodlots were two of the larger unmonitored source areas in each basin, accounting for 12% and 1% of the water volume
produced at  Monroe and 7% and 2% at Harper, respectively. To add  sidewalks to  the load estimates, concentrations
measured at  driveways were  applied to estimates at sidewalks to create a sidewalk load. Woodlot concentrations  were
estimated by  use of data collected  in an undeveloped urban site near Superior, WI (Steuer  and others, 1997). The source-
area loads were 39% lower and 60% lower than the measured load at the storm- sewer outfall in the Monroe and Harper
Basins,  respectively. This difference between  source-area and storm-sewer outfall  loads indicates  that one or more source
areas within  each basin  were not effectively monitored.

    Streets were the most  likely source area to  be ineffectively  monitored. Street samplers were placed  approximately
5 ft away from the curb to  prevent gutter flow into the  sampler because gutter flow usually contains a mixture of  water
from several  source areas.  Other  street studies (Pitt, 1979) have estimated that  90%  of the dirt on residential streets in
good condition with little to  no parking accumulates within 3 ft of the curb. A larger amount of dirt can  sometimes collect
along the curb itself rather than  in the driving lane. Some of this dirt could have been  deposited on the driving lane, and
turbulence from passing vehicles and wind may have moved  it to the curb. Most street dir-t falls within 1  to 2 ft of the curb
if the driving  lane  is next to the curb  (Pitt, 1979).  This information suggests that the street samplers in the Monroe and
Harper Basins were too far from  the  curb (5 ft) to representatively collect the particulate dirt from the streets.

    For the reasons previously described, a  trial-and-error approach  was used to select a street- sediment concentration
that more accurately reflected the  street sediment entering the storm sewer. The final suspended-solids concentrations
for streets were increased  by a factor of 5.  Applying this  factor  to the  simulated street  suspended-solids concentration
during each storm event allowed the sum of  source-area loads to be within 7% and  9% of the storm- sewer outfall loads
at Monroe and Harper, respectively (Table 5). The  geometric means for the revised street suspended-solids
concentrations were 340 and 325 mg/L for  low-  and high-traffic streets,  respectively. These values were within 5% of
those measured at both  Marquette, Ml (Steuer and others, 1997), and  Madison, WI (Bannerman and others,  1993).

    The  geometric means of the  observed suspended-solids  concentrations,  excluding streets,  for the 25 storm events
at Monroe and Harper Basins were placed into the SLAMM data base  (Table 6). The suspended-solids concentrations
for streets were not as  easily altered because they are determined  by dirt accumulation and wash-off  functions in the
model. Entering the geometric means enabled the model to more  accurately  predict  the loads measured at the storm-
sewer outfall. After summing the 25 events,  sediment loads predicted by SLAMM were 17% lower at Monroe  and 32%
lower at Harper compared  to the  measured storm-sewer  outfall  loads.

    To improve the match  between  measured and simulated storm-sewer outfall loads, the delivery coefficients  were
removed from SLAMM calculations, essentially assuming  100%  delivery from source area to  storm-sewer outfall. The
delivery  coefficients had been added  in a  previous calibration  study to  force a match to the storm-sewer outfall


                                                       23

-------
numbers. This adjustment resulted  in a  4%  and 17%  undersimulation  between storm-sewer outfall loads  of sus-
pended  solids calculated  by SLAMM  and those  measured at the storm-sewer outfall for Monroe and Harper,
respectively  (Table 5).

Phosphorus  Calibration

    One objective  in calibrating  phosphorus concentrations in  SLAMM was to ensure that the Monroe and Harper Basins
were accurately  represented by the monitored source areas.  The sum of source-area loads for total phosphorus for the
25

Table 5. Percentage Difference in Cumulative Source-area Sediment Loads, Before and  After Sediment Adjustment, and Modeled and Measured
Sediment  Loads at the Basin Outfall, Madison, WI. [Loads computed as suspended solids; %, percent]

        Cumulative source area compared
             with storm-sewer outfall
  Basin    Suspended      Suspended       Modeled
          solids before     solids after    compared with
          adjustment      adjustment      measured

Monroe        -39             -7              -4
Harper        -60             -9             -17


Table 6. Suspended-solids Concentrations Used to Calibrate the Source Loading and Management Model for Basins in Madison, Wi. [mg/L, milligrams
per liter: —, a series  of algorithms and coefficients are used in the model to calculate a suspended-solids concentration for street runoff]

         Suspended solids (mg/L)

Source area    Residential    Commercial
Driveways          34           34
Lawns             84           84
Parking  lots         51            51
Streets
Woodlots          15           15
Roofs              16           18
Sidewalks          34           34


storm events was nearly identical to the storm-sewer outfall load (Table 7). The difference  was larger for dissolved
phosphorus,  but  no  information was available to determine  what adjustments  should have been  made to reduce  the
difference. For this reason,  the unadjusted concentrations were entered into the SLAMM  data base.

Table?. Percentage  Difference Between Cumulative Source Area Versus Outfall Loads and Modeled Results Versus Outfall Loads, after Calibration of
the Source Loading  and Management Model for Basins in Madison,  Wi. [%, percent]

         Cumulative source  area      Model results after
                versus          calibration versus storm-
          storm-sewer outfall          sewer outfall
  Basin                  Phosphorus load

         Dissolved     Total      Dissolved      Total

Monroe       39          -1           -9          -24
Harper       35          4          -10          -28


    The  model simulates  total phosphorus loads  by  adding the dissolved  phosphorus  and particulate phosphorus loads.
For all source areas except streets, particulate-phosphorus  concentrations were calculated using total- and dissolved-
phosphorus and  sediment concentrations measured at the Monroe and  Harper  Basins. To  be consistent with calibration


                                                          24

-------
procedures, particulate-phosphorus  concentration for street runoff was calculated using the adjusted value for sediment
(an increase by a factor of 5). Changing the phosphorus concentrations resulted in SLAMM undersimulation of dissolved
and total  phosphorus by 9% and 24% at the Monroe storm-sewer outfall and 10% and 28% at the Harper storm-sewer
outfall,  respectively.

 The dissolved- and particulate-phosphorus concentrations entered  into the SLAMM data base are listed  in Tables 8 and
9. Significant  changes  in  dissolved-phos-phorus concentrations (Table 8) were observed  for lawns (from 0.22  to  0.53
mg/L), streets (from 0.39 to 0.12 mg/L), woodlots (from 0.25 to 0.01 mg/L), and sidewalks (from 0.60 to 0.07 mg/L).  With
the exception of streets,  where  the  particulate-phosphorus concentrations  in  runoff decreased,  particulate-phosphorus
concentrations  increased  significantly  (Table  9).

Table 8.  Dissolved-phosphorus Concentrations Used to Calibrate the Source Loading and Management Model for Basins in Madison, Wi.
[mg/L, milligrams per liter]

           Dissolved  phosphorus (mg/L)

Source-area          Residential      Commercial
Driveways                0.07             0.07
Lawns                   .53              .53
Parking lots               .02              .02
Streets                   .12              .03
Woodlots                .01              .01
Roofs                    .04              .02
Sidewalks                .07              .07


Table 9.  Particulate-phosphorus Concentrations Used to Calibrate the Source Loading and Management Model for Basins in Madison, Wi.
[mg/kg, milligrams per kilogram]

            Particulate phosphorus (mg/kg)
Source area          Residential       Commercial
Driveways             2,649            2,649
Lawns                4,943            4,943
Parking lots            1,467            1,467
Streets                 569             409
Woodlots             5,000            5,000
Roofs                 3,777            7,946
Sidewalks             2,649            2,649



Distribution  of  Source-Area Loads

    The distribution  of suspended-solids  and total- and  dissolved-phosphorus loads for source areas in the Monroe and
Harper  Basins  using measured source-area concentrations multiplied  by SLAMM-generated water volumes  is shown in
Table 10.  The distribution of water volumes is  nearly identical at Monroe and  Harper Basins. The percentage of the total
basin represented by each source area is similar for both basins (Table 1);  thus, one should expect to see similar  relative
volumes of water calculated from both basins. Streets  contributed most of the suspended-solids  loads  at both Monroe
and Harper Basins,  generating 81%  and 73%, respectively. Lawns contributed more than 10% of the solids loads  at both
basins.  The phosphorus loading, however, was quite different. Lawns  in the  Harper Basin generate more than two-thirds
of the phosphorus loads, whereas phosphorus in the  Monroe  Basin  is more evenly distributed between lawns  and  streets.
These differences in load distribution are the result of the  measured phosphorus  concentrations, especially for  lawns,
which are  much higher for the Harper Basin (Table 2).
                                                         25

-------
Table 10. Distribution of Loads Based on Measured Values at Monroe and Harper Basins, Madison, Wi, and Incorporating the Suspended-solids Adjustment
[N/A, source area not present; %, percent abundance; --, value less than 0.5 percent]
 Source area
 Water
volume'
  HARPER
Suspended
 solids (%)
Lawns
Streets
Driveways
Sidewalks
Parking lots
Roofs
Parks
Woodlots
Other
Total
21
37
18
7
3
11

3
N/A
311,122
15
732
7
3


--
--
N/A
3,598
                                     Total      Dissolved
                                   phosphorus  phosphorus
                                      67
                                      14
                                       9
                                       4
                                       1
                                       3
                                       2

                                      N/A
                                       9
71
11
 9
 3

 4
 2

N/A
 5
                          MONROE
           Water   Suspended    Total      Dissolved
         volume (%) solids (%)  phosphorus  phosphorus
                                                                    44          45
                                                                    37          39
                                                                     5           4
                                                                     6           4
                                                                     1
                                                                     1
                                                                     7           7
                                                                    N/A         N/A

                                                                    70          33
20
38
12
14
6
7
3
N/A
1
2,417,341
10
281
2
3
2
	
2
N/A
..
26,045
 'Water volume totals expressed in cubic feet; all other totals expressed in pounds.
2Street-runoff concentrations multiplied by 5.


    The suspended-solids load distribution  at the Monroe Basin in 1994 is similar to the distribution  observed in a 1991
study (Bannerman and others, 1993). Streets contributed 80% of the total basin suspended-solids load in 1991 and 81%
in 1994.  Lawns also were comparable, contributing 7%  and 10% of the total  basin suspended-solids load in 1991  and
1994,  respectively. Total and dissolved phosphorus, however, were very different. During the 1991 study,  the proportion
of the  total-phosphorus load from streets (58%) outweighed that for lawns (14%). The same was true in 1991 for dissolved
phosphorus, where streets produced 46% and lawns22% of the basin load. However, most total- and dissolved-phospho-
rus loading in 1994 was attributed to lawns rather than streets.  Streets and lawns,  in 1994, generated 37% and 44%  of
the total-phosphorus load  and 39% and 45%  of the dissolved- phosphorus load  in the Monroe Basin. The difference  in
distributions between the two studies is possibly due to  differences in sampling methodology. The street-sampler design
was modified for the 1994 study to eliminate a first-flush  effect, where the sample bottle would quickly fill with stormwater
and act as a sediment trap for the remaining duration of the storm  event. Also,  during the 1994 study, 25 events were
monitored, whereas only 10 events were monitored in 1991. This larger sample size in 1994 improves confidence in the
loading-distribution   estimates.

    Distribution of suspended-solids, total-phosphorus, and  dissolved-phosphorus loads estimated by SLAMM are given
in Table  11. The distribution of loads is consistent with the distribution of measured loads shown  in Table  10. For each
constituent, slightly less total  load was simulated with SLAMM than calculated  using the measured concentrations (other
than suspended solids from streets), yet the distributions of each constituent were  similar. Streets and lawns contribute
nearly all  of the suspended-solids load for the entire basin. Streets alone contribute more than 75%  of the suspended
solids  at both Monroe and Harper Basins. Additionally, the significance of lawns as generators of phosphorus is again
noted  in  SLAMM  simulations. Lawns in the  Harper Basin  contribute  52% and 61% of total and dissolved phosphorus
loads,  and lawns in the Monroe Basin contribute 49% and 57%. Streets contribute  the second largest phosphorus loads
(about 25%), whereas  driveways and sidewalks  combined contribute approximately  10%.

    The distribution of suspended solids and total  and dissolved phosphorus for source areas in the Monroe and Harper
Basins using measured source-area  concentrations multiplied by  SLAMM-generated water volumes is shown in Table
12. Only loads  for the source areas measured are shown in Table 12; concentrations of suspended solids in street runoff
have not been  adjusted. These source areas accounted for 82% and 90% of the total water volume from the Monroe and
Harper Basins, respectively.
                                                        26

-------
Table 11. Distribution of Loads from Model Simulation Results at Monroe and Harper Basins, Madison, Wi. [N/A, source area not present; %, percent
abundance; -, value less than 0.5 percent]
                      HARPER
Source area   Water    Suspended     Total     Dissolved
             volume   solids (%) phosphorus  phosphorus
                                      52
                                      32
                                      8
                                      3
                                      1
                                      3
                                      2
                                      1
                                      N/A
                                      7
Lawns
Streets
Driveways
Sidewalks
Parking lots
Roofs
Parks
Woodlots
Other
Total
21
37
18
7
3
11

3
N/A
311,122
11
81
4
1
1
1
	
--
N/A
3,170
                                                                         MONROE
                                                          Water   Suspended    Total      Dissolved
                                                        volume (%) solids (%) phosphorus  phosphorus
                                                                             49          57
                                                                             26          22
                                                                              5          4
                                                                              6          5
                                                                              1
                                                                              2
                                                                              8          9
                                                                             N/A         N/A
                                                                              2          2
                                                                             60          29
61
24
7
3

2
2
..
N/A
4
20
38
12
14
6
7
3
N/A
1
2,417,341
12
77
3
3
2
1
2
N/A
1
20,814
1'Water volume totals expressed in cubic feet; all other totals expressed in pounds.

Table 12. Distribution of Loads from Monitored Source Areas Only, Based on Unadjusted Concentrations at the Monroe And Harper Basins, Madison, Wi.
[ %, percent abundance; percentage columns may not add up to 100% because of independent rounding]
Source area
Lawns
Streets
Driveways
Parking lots
Roofs
Water
volume
  (%)
  23
  41
  20
   4
  12
HARPER
Suspended
  solids
   (%)
    41
    43
    10
    3
    3
                                   Total     Dissolved
                                phosphorus phosphorus
                                    70
                                    20
                                     7
                                     1
                                     3
                                              75
                                              15
                                               6
                                               0
                                               4
                                                          Water
                                                         volume
                                                           (%)
                                                           24
                                                           46
                                                           14
                                                            7
                                                            9
                                                                         MONROE
                                                                 Suspended     Total      Dissolved
                                                                   solids    phosphorus phosphorus
28
53
9
7
3
56
33
7
2
2
69
21
 8
 1
    The suspended-solids distributions shown in Table 12 differ from Tables 10 and 11, in that the significance of lawns
as a source increases and the significance of streets as a source decreases. Streets  are still the largest source of
suspended solids in  both  basins. The phosphorus  distributions also change, but  not as much  as  the suspended solids
because the measured phosphorus concentrations were used in  all three tables  (Tables 1 0,  1 1, and 12). The significance
of lawns as a source increases slightly, and streets are a slightly  larger source in  some cases, and in others, are slightly
smaller sources. Results  shown in Table 12  indicate that the  adjustments  made to suspended-solids concentrations in
street runoff do  not  greatly affect the phosphorus  distributions in the  basins.

Sediment and Phosphorus Mass  in Street-Dirt Samples

    Approximately 75% of the  total sediment mass in  the street-dirt samples originated in the >250 urn size fraction,
whereas the smaller fractions (<63 Lim) made up less than 5%.  Material composed of leaves,  twigs, and other organic
debris also were measured, contributing, on average, less than 10% of the total sediment mass of the sample (Figure
9).
Like sediment  mass, the largest amount of total phosphorus was found  in  the >250 pm size fraction (nearly
50%) (Figure  9). Combining this size fraction with  the  leaf fraction, about 80% of the total  phosphorus is
accounted for.  The contribution of total  phosphorus mass decreased  as  the size fraction  decreased.  Other
studies  have shown that large phosphorus concentrations correspond  with  small particle sizes because of the
high surface area to mass ratio for small particles (Sartor and Boyd,  1972).
                                                       27

-------
                       <25
25-63            63-250

    PARTICLE SIZE, IN MICROMETERS
>250
                Leaves
Figure 9. Relations between sediment and total-phosphorus mass from street-dirt samples for five particle-size fractions for basins in Madison, Wl
However, the bulk of the phosphorus phosphorus load results from the greater particle-size fractions (Ray, 1997).
Approximately 25% of the total phosphorus mass in each size fraction can be attributed to leaves (Ray, 1997).

    A recent study of particle-size distribution in stormwater at the Monroe storm-sewer outfall demonstrated that most
of the solids are in the particle sizes <63 ?m (Greb and Bannerman, 1997). This distribution is the opposite of the  particle-
size distribution observed for the street dirt collected in this study. These results indicate either a loss of the larger
particles somewhere between the street and the outfall or a problem in collecting  larger particles in the runoff samples.
Most of the larger particles (>63 ?m) might settle out before reaching the storm-sewer outfall. Street sweeping,
resuspension onto street terraces,  and catch basins can remove these particles from the streets  before they reach the
storm sewer. Also, the transport of large particles in a storm sewer is not as efficient as the transport of smaller, more
mobile particles.  Large sediment particles may become trapped or part of the bedload before reaching the sampler.
Bedload  is  not sampled efficiently by the  automatic samplers described earlier in this report.

Summary and Conclusions

    Concentrations of suspended solids,  total phosphorus, and dissolved  phosphorus were collected from various source
areas at  two urban residential  basins in Madison, Wl. To  represent a range of  source-area concentrations for  urban
residential basins in  Madison,  the geometric means of the combined  concentration data from the Monroe and Harper
Basins were incorporated into the urban-runoff model, SLAMM.

    Source-area suspended  solids and phosphorus loads from the Monroe and Harper Basins were determined  based
on measured concentrations that were multiplied  by water volume  estimated  by  use of SLAMM. Collected  data were  used
to calibrate  and increase confidence  in water volumes, suspended solids, and phosphorus source-area  loads estimated
by SLAMM.  The calibrated model calculated water volumes  to within 23% and  24% of those measured at the outfalls of
                                                      28

-------
the Monroe and Harper Basins. These calibrated water volumes were  then  applied to  the calibrated suspended-solids
and  phosphorus concentrations entered into the SLAMM data  bases.  Suspended-solids loads were  estimated by the
calibrated SLAMM to be within 4% and 17%, total-phosphorus loads within  24% and 28%, and dissolved-phosphorus
loads within 9% and 10% of those measured at the storm-sewer outfall  to the Monroe and  Harper Basins, respectively.

    Streets and  lawns are the largest  contributors of suspended-solids, total-phosphorus,  and dissolved-phosphorus  loads
in a residential  urban  basin.  Lawns are the largest contributors of total and  dissolved phosphorus; however, streets
contributed nearly 40% of the basin load, as seen  in the Monroe Basin. Streets were found to be the largest source of
suspended  solids.

    There was a large difference between  geometric mean concentrations of phosphorus  in lawn runoff from 1994  to
1995. Phosphorus data collected from lawns  in the Harper and Lakeland Basins during 1995 are remarkably similar, which
suggests that  the phosphorus concentration in lawn  runoff is affected  by some  variable or variables that are  not yet
understood.

    Street-dirt samples indicate that approximately 75% of the sediment mass resides in the >250 pm particle-size
fraction. Less than 5% of the mass can be found in the particle sizes less than 63 pm. The >250 urn particle-size fraction
also contributed nearly 50% of the total- phosphorus mass,  and the leaf fraction contributed an additional 30%.  In each
particle-size fraction, approximately 25% of the total-phosphorus mass  is derived from  leaves or other vegetation.

    A possible limitation of this study may be that in order for the sum of the source-area loads to match the basin-outfall
loads, it was assumed that the concentrations of suspended solids in street  runoff were about 5-times higher than the
concentrations measured.  However, the analysis  of load distributions based  onlyon unadjusted monitoredconcentration
data shows  little change in the distributions. In addition,  samples from more rain events were collected in this study than
previous source-area studies. Also,  improved  data-collection  equipment were used during this study. Both of these factors
lead to greater confidence in  the study results.

    Most  of the measured suspended-solids concentrations were  lower than those  measured from  previous studies.
However,  when  comparing concentration results in  this  study to results  from  earlier studies, it is important to note that
with the exception of the Marquette study, previous studies  used earlier generation source-area sampling equipment.  In
Marquette, Ml,  the soils are considerably more sandythan those  in Madison,  which may explain why the suspended-solids
concentrations determined for the Marquette study  are higher than those from Madison  even though  both  studies used
the same  sample-collection equipment.

    The recalibration of the SLAMM model results in an improved model that should more accurately simulate phosphorus
and sediment  runoff loads in Wisconsin than the  earlier version of the model. The newly created lawn-runoff coefficients
for Madison  represent a  compromise between the  two previous soil-type options available  for model input,  which probably
represented runoff extremes.  The runoff coefficients calculated for Madison should probably be applied to most urban
lawns in  Wisconsin  unless soils are known to be either sandy  or clayey. The  phosphorus- and sediment-concentration
data bases created for this study are the largest to date using the most  advanced  source-area  sample  collection
technology available.

References

Andren, A.W.,  and  Stolzenburg, T.R.,  1979,  Atmospheric deposition  of lead and  phosphorus on the Menomonee  River
watershed-Volume 8,  The IJC Menomonee River Watershed Study: U.S.  Environmental  Protection Agency,
EPA-905/4-79-029-H, 98 p.

Bannerman, R.T., Baun,  K., Bohn, M., Hughes,  P.E., and  Graczyk, D.A.,  1983, Evaluation of urban nonpoint source
pollution  management in  Milwaukee  County, Wisconsin-Volume I,  Urban stormwater characteristics,  pollutant sources
and management by street sweeping: U.S.  Environmental Protection Agency, Water Planning Division,  PB-84-114164,
191 p.
                                                      29

-------
Bannerman, R.T., Owens, D.W., Dodds, R.B., and  Hornewer, N.J., 1993, Sources of pollutants in Wisconsin stormwater:
Water  Science  and Technology, v. 28, no. 3-5, p. 241-259.

Betz, CR.,  Lowndes, M.A., and Porter, S., 1997, Nonpoint source control plan for the  Lake Mendota priority watershed
project-Project  summary: Wisconsin Department of Natural Resources, WT-481-97, 16 p.

Cullen, R.S., and  Huff,  D.D.,  1972, Determination of land use categories in the  Lake Wingra Basin:  US-International
Biological Program, Eastern  Deciduous Forest  Biome Program,  Memo Report 72-43,  17 p.

Dong,  A., Simsiman, G.V., and Chester, G., 1983, Particle size distribution and phosphorus levels in soils, sediment and
urban  dust and dirt samples from  the Menomonee River watershed, Wisconsin, USA: Water Research, v. 17, no. 5, p.
569-577.

Greb,  S.R., and Bannerman,  R.T., 1997, Influence of particle  size on wet pond effectiveness: Water Environment
Research, v. 69, no. 6,  p. 1134-1  138.

Helsel, D.R., and Hirsch, R.M., 1992, Statistical methods in water resources: Amsterdam and  New York, Elsevier, 529
P-

Kilpatrick, FA,  Kaehrle,  W.R.,  Hardee, J., Cordes, E.H., and Landers,  M.N.,  1985, Development and testing of highway
storm-sewer flow measurement and  recording  system:

U.S. Geological  Survey  Water-Resources Investigations Report 85-4111, 98 p.

Kunz,  K.W., 1980,  Atmospheric bulk precipitation in the  Great Lakes basin: Environment Canada, Scientific Series 115,
28 p.

Lathrop, R.C.,  Nehls, S.B., Brynildson,  CL, and Plass, K.R.,  1992, The  fishery of the Yahara Lakes: Wisconsin
Department of Natural Resources,  Technical Bulletin  181,  214 p.

Legg, A.D.,  Bannerman,  R.T., and Panuska, J., 1996, Variation in the relation of rainfall to runoff from residential lawns
in Madison, Wisconsin, July and August 1995:  U.S. Geological Survey Water-Resources Investigations Report 96-4196,
11 p.

Murphy,  T.J., 1974,  Sources of phosphorus inputs from  the atmosphere and their significance  to oligotrophic lakes:
University of Illinois Water Resources Center,  Research Report 92,  41 p.

Oakes, E.L., Hendrickson, G.E., and Zuehls, E.E., 1975, Hydrology  of the Lake Wingra Basin,

Dane  County, Wisconsin: U.S. Geological Survey Water-Resources Investigations  17-75,  31  p.

Pitt,  R., 1979, Demonstration of nonpoint pollution abatement through improved street and sewerage cleaning: Cincinnati,
Ohio, U.S.  Environmental Protection Agency, EPA/600/2-79/161, 270 p.

-1987, Small storm urban flow and particulate washoff contributions to storm-sewer  outfall discharges: University of
Wisconsin-Madison, Dept. of Civil  and  Environmental Engineering,  513 p.

Pitt,  R.,  Field,  R., Lalor,  M.  and Brown,  B.  1995, Urban  stormwater toxic pollutants-Assessment,  sources, and
treatability:  Water Environment Research,  v. 67, no. 3,  p.  260-275.

Prentki, R.T., Rogers, D.S.,  Watson, V.J.,  Weiler, P.R., and Loucks, O.L., 1977, Summary tables  of Lake Wingra  basin
data: University of Wisconsin-Madison,  Institute for Environmental Studies, IES Report 85, 89 p.

Ray, H,  1997, Street dirt as a  phosphorus source  in urban stormwater:  University of Alabama at  Birmingham, Department
of Civil Engineering,  125 p.


                                                      30

-------
Ryden, J., Syers, J. and Harris, R., 1972, Nutrient enrichment of runoff waters by soils, phase  I- Phosphorus enrichment
potential  of urban soils in the city of Madison: University of Wisconsin-Madison,  Department of Soil Science, Technical
Report OWRR-A-038-WIS, 70 p.

Sartor, J.D.,  and Boyd, G.B., 1972, Water pollution aspects  of street surface contaminants: U.S. Environmental Protection
Agency, EPA-R2-72-081.237 p.

Schindler, D.W., 1974, Eutrophication and recovery in  experimental lakes-Implications  for lake management: Science,
v.  184, p. 897-899.

-1977, Evolution of phosphorus  limitation in lakes: Science, v. 195, p. 260-262.

Steuer, J.J.,  Selbig,  W.R., and Hornewer, N.J., 1997, Sources of contamination in an urban basin in Marquette, Michigan
and an analysis  of concentrations, loads, and data quality: U.S.  Geological Survey Water-Resources Investigation Report
97-4242, 25 p.

U.S. Environmental  Protection Agency, 1983,  Results of the Nationwide  Urban Runoff Program, volume l-Final report:
Water Planning  Division, Washington, D.C., NTIS PB#84-185552, 200 p.

Ward, J.R., and  Harr,  C.A.,  eds, 1990, Methods for collection and processing  of surface-water  and  bed-material  samples
for physical  and chemical analyses:  U.S.Geological Survey Open-File Report 90-140, 71 p.
                                                      31

-------
  Using Biological Criteria to Assess and Classify Urban Streams and Develop
                                Improved Landscape Indicators
                                     Chris 0. Yoder and Robert J. Miltner
                                     Ohio EPA, Division of Surface Water
                                         Ecological Assessment Unit
                                           Groveport, Ohio 43125
                                                Dale White
                                     Ohio EPA, Division  of Surface Water
                                 Information Resources Management Section
                                        Columbus, Ohio 43216-0149
Abstract

    This study consisted of a quantitative analysis of the relationship between the Index of Biotic Integrity (IBI), an
indicator of urban land use,  and a qualitative analysis of overlying stressors in six of the major metropolitan areas of Ohio.
A database consisting of 267 sampling locations was extracted from the Ohio EPA statewide biological  and  habitat
database.  Most of these sites were sampled between 1990 and 1998 and contained watershed areas less than 50 mi.2,
with most draining less than 20 mi.2. A negative relationship between IBI and urban land use was observed in four of the
six areas, whereas  little or no relationship was seen in two areas. For each area, the highest percentage of urban land
use that corresponded  to minimum attainment of the applicable warmwater habitat IBI biocriterion  ranged from  1%
(Cleveland/Akron) to 12% (Dayton) for the regression line, and  15% (Cleveland/Akron) to 58% (Columbus)  as the  highest
%urban land use where the IBI biocriterion was attained at any given site. No significant linear relationship was found
in either the Toledo or Youngstown areas, and only a weak relationship was visually apparent for the Toledo streams.
The lack of association was due to  the strong presence of overlying stressors (e.g.,  legacy  pollutants, sewage discharges,
combined sewer overflows) that resulted in very low IBI values at sites with lower levels of urbanization. The percentage
of urban land use explained approximately 35% of the variation in IBI scores in the regression model when these impact
types were excluded (compared to  11% when included). The maximum %urban land use that commonly corresponded
to attainment of the warmwater habitat IBI biocriterion based on inspection of the scatter plot was approximately 26%.
Only a very few sites exhibited attainment at  urban land uses  between 40-60% and none occurred above 60%. These
former sites had either an intact, wooded riparian zone, a continuous influx of groundwater, and/or the relatively recent
onset of urbanization. These results indicate that it might be possible to mitigate the negative effects of urbanization by
preserving or enhancing near and instream habitats, particularly the quality of the riparian buffer zone. The  results also
suggest that there is a threshold of watershed urbanization (e.g., >60%) beyond which attainment of warmwater habitat
is unlikely. This threshold is not the same in all watersheds and it can occupy a rather wide range. It is affected  by co-
factors such as pollutant loadings, watershed development history, chemical stressors, and watershed scale influences
such as the quality  of the riparian buffer and  the mosaic of different types of land use. Thus, single-dimension urban land
use indicators, such as watershed  imperviousness, are not sufficiently precise  or robust as a single indicator  of use
attainability. The further development and refinement of multiple indicators of watershed  urbanization has merit from a
management and decision-making  standpoint. We suggest that co-factors, in addition to more refined urban land use
indicators, be  better developed.   More precise definitions  of different urban  land uses are also needed to  better
understand and respond to the water quality management challenges posed in existing and developing urban  areas.
                                                    32

-------
 Introduction

    The health and well-being of the aquatic biota in surface waters is an important barometer of how effectively we are
 achieving the goals of the Clean Water Act (CWA); namely the maintenance and restoration of biological integrity, and
 the basic  intent of water quality standards. States designate water bodies for beneficial uses (termed designated uses)
 that,  along  with  chemical,  physical,  and  biological criteria,  assure the protection  and restoration  of aquatic life,
 recreational,  and water supply functions and attributes. Biological criteria are the principal tool for determining impairment
 of designated aquatic life uses as defined  by the Ohio  WQS (Ohio  Administrative Code  3745-1). As  such,
 bioassessments play a central role in the Ohio Nonpoint Source Assessment (Ohio EPA 1990; 1991), the biennial Ohio
 Water Resource  Inventory (305b  Report; Ohio  EPA 1998), and watershed-specific assessments, of which Ohio EPA
 completes between 6  and 12 each year. Biological criteria represent a measurable and tangible goal, against which the
 effectiveness  of  pollution control and other  water  quality management efforts can  be  judged.  However, biological
 assessments must be  accompanied by appropriate chemical/physical  measures, land use  characterization, and  pollution
 source information necessary to establish linkages between stressors and the biological responses (Yoder and Rankin
 1998). Biological criteria in the Ohio WQS also supports the determination of appropriate aquatic life use designations
 for individual water bodies, provides for a "reality check" on the application of surrogate indicators, assesses cumulative
 impacts, extends anti-degradation concerns to nonpoint sources and habitat influences, defines high quality waters, and
 serves as a  meaningful indicator in the management of regulatory programs for environmental results.  This provides a
 means to incorporate the broader concept of water resource integrity (Karr et al. 1986) in policy and planning  while
 preserving the  appropriate  roles  of the traditional chemical/physical and toxicological approaches  developed over the past
 three decades.

    We, and others at Ohio EPA, have previously described the status of Ohio's streams and  rivers as affected by
 watershed urbanization (Yoder et al.  1999; Yoder and Rankin 1997; Yoder 1995). Small watersheds are especially
 impacted, as  illustrated by Yoder and  Rankin (1997), where no headwater streams in established urban settings
 throughout Ohio attained the minimum CWA benchmark use designation of warmwater habitat. This finding has led to
 the perception that the impairment of beneficial aquatic life uses in these small watersheds is intractable, at  least within
 the constraints of current land use policies, restoration technologies, and funding levels. Together, these factors present
 potentially  significant barriers to the objective of fully restoring degraded watersheds or upgrading urban streams that are
 presently  designated for less than fishable  and  swimmable uses.

    Headwater streams are critical to watershed functioning in that they serve as the principal interface between  runoff
 from land  use and receiving streams. The ability of a headwater stream to physically filter and biologically assimilate the
 primary and secondary effects of pollutants is a function of habitat quality and the structure of the biological system. A
 healthy headwater stream ecosystem is characterized  by good  habitat  and  a well balanced  assemblage of aquatic
 organismsand plants, one which  processes external inputs in a manner which  promotes  high quality  downstream exports.
 These exports include good quality  water and high value  biomass, both of which positively  impact the ability of
 downstream waters to deliver quality goods  and services (e.g., water supply,  recreation, waste  assimilation,  water
 retention,  ecological values). A degraded headwater  stream ecosystem is characterized by poor  habitat and an
 assemblage of aquatic organisms and plants that processes external inputs in a manner which  promotes low quality
 downstream  exports. Thus  in this latter scenario,  the effects from  urban runoff can accumulate in a downstream  direction
 and adversely  affect water quality and ecosystem goods and services in  larger water bodies.  In Ohio,  more than 78%
 of stream  miles drain less than 20 mi.2 and are classified as headwater streams. While these may individually seem less
 significant  than larger water bodies, they are collectively the most numerous and perhaps important  stream type.  In  many
ways, and in a collective sense, headwater streams are analogous to the capillaries of the human circulatory system
where essential product transport and waste assimilation functions are accomplished.  Certainly the finding that a high
 proportion of headwater streams fail to meet CWA goals in Ohio urban areas translates to the potential for  undesirable
 impacts in downstream waters and obvious consequences for the overall health of the "patient".

    There is concern  that the  attainment of CWA  goal uses  (e.g., warmwater habitat  in Ohio) within small urban
watersheds may  be precluded by the legacy  of urbanization.  If this is true, how is this determined and what are the
 protection endpoints to guide water quality management? Federal water quality standards regulations (40CFR, Part 131)
                                                      33

-------
allow for the establishment of a use that is less than the CWA fishable and swimmable goals when it is precluded by the
following:

    1)   the degraded conditions are naturally occurring;

    2)   restoring the degraded conditions would result in widespread adverse socioeconomic impacts;

    3)   the degraded conditions are irretrievable and human induced.

    Such uses are established on a waterbody-specific basis and are supported by a use attainability analysis. In Ohio,
such analyses are routinely conducted as a result of the five-year basin approach to monitoring, assessment, and water
quality management. One purpose of this  paper is to advance the development of the tools and indicators needed to
make use attainability decisions in urban watersheds.

    Ohio EPA routinely conducts biological and water quality surveys, or "biosurveys", on a systematic basis statewide.
A biosurvey is an interdisciplinary monitoring effort coordinated on a waterbody-specific or watershed scale. Such efforts
may be relatively simple, focusing on one or two small streams, one  or two principal stressors, and a handful of sampling
sites; or much more complex, including entire drainage basins, multiple and overlapping  stressors, and  tens of sites. Each
year, Ohio EPA conducts biosurveys in 1 O-l 5 different study areas with an aggregate total of 350-450 sampling sites.
Biological, chemical, and physical monitoring and assessment techniques are employed in biosurveys in order to meet
three major objectives: 1) determine the extent to which use designations assigned in the Ohio Water Quality Standards
(WQS) are either attained or not attained; 2) determine if use designations assigned to a given water body are appropriate
and attainable; and  3) determine if any changes in key ambient biological, chemical, or physical  indicators have taken
place over time,  particularly before and after the implementation of point source pollution controls or best management
practices. The data  gathered by a biosurvey is processed, evaluated, and synthesized in a biological and water quality
report. The findings and conclusions of each biological and water quality study may factor into  regulatory actions taken
by  Ohio EPA and are incorporated into Water Quality Permit Support Documents (WQPSDs), State Water Quality
Management Plans, the Ohio Nonpoint Source Assessment, and the Ohio Water Resource Inventory (305[b] Report).

    In  1990, the Ohio EPA initiated an organized, sequential approach to monitoring and assessment, termed the Five-
Year Basin Approach. One of the principal objectives of this new approach was to better coordinate the collection of
ambient monitoring data so that information and reports would be available in time to support water quality management
activities such as the reissuance of NPDES permits and periodic revision of the Ohio  Water Quality  Standards (WQS).
Ohio EPA's  approach to surface water monitoring and  water quality management via the Five-Year Basin Approach
essentially serves as an environmental feedback process taking "cues" from environmental indicators to effect needed
changes or  adjustments within water quality  management.  The  environmental indicators used in this process  are
categorized as stressor,  exposure,  and response indicators (Yoder and Rankin 1998).  Stressor indicators generally
include activities that impact, but which may or may not degrade the environment. This includes point and nonpoint
source loadings, land use changes,  and other broad-scale influences that generally result from anthropogenicactivities.
Exposure indicators include chemical-specific, whole effluent toxicity,  tissue residues, and biomarkers, each of which
suggests or provides evidence of biological exposure to stressor agents. Responseindicators include the direct measures
of the status  of use designations. For aquatic life uses, the community  and population response parameters that  are
represented  by the biological indices that comprise Ohio EPA's biological criteria are the principal response indicators.

    Previously, our  analyses examined the water quality and biological assessment database from  watersheds  in and
near existing and developing urban and suburban  areas of Ohio. Yoder and Rankin (1997) compiled their analyses  based
on sampling conducted at more than 100 stream sampling locations. Yoder et al. (1999) examined more detailed land
use and stressor relationships with the Index of Biotic  Integrity (IBI), based on fish assemblage data, and the Invertebrate
Community  Index  (ICI),  based  on macroinvertebrate  assemblage  data,  within two major Ohio urban  areas
(Akron/Cleveland and Columbus). This study consisted of a quantitative analysis of the relationship between the  IBI, an
indicator of urban land use, and a qualitative analysis of other stressors influencing this relationship using available data
from all six  of the  major metropolitan  areas within  Ohio.  The importance  of understanding these  relationships is
heightened by contemporary water quality management issues such as combined sewers and stormwater permitting.
One challenge we faced was in attempting to separate the influences of  these multiple stressors on  aquatic life attainment


                                                     34

-------
status.  Could we  sufficiently understand the baseline influence of urbanization apart from these other and better
understood  stressors?

    The principal analysis conducted in this study examined the relationship between urban land cover and the IBI, both
visually and  by statistical  analysis. Some goals were to determine the extent to which biological performance  (as
expressed by the IBI) was correlated with urban land use, thresholds at which this occurred, and the overlying effects of
other stressors.

Met hods

    A database consisting of 267 sampling locations from the six major metropolitan areas of Ohio was extracted from
the Ohio EPA statewide biological and habitat database. Most of these sites were sampled between 1990 and 1998 and
contained watershed areas less than 50 mi.2, with most draining less than 20 mi.2. As such, the database represents a
collection of discrete watershed units where land  uses may have a significant effect on the composition and quality of the
instream habitat and biological communities. Urban land  use effects  have been much more apparent in these smaller
watersheds as evidenced by the  higher proportion of impaired stream miles compared to  larger streams and  rivers in Ohio
(Yoder  1995; Yoder and Rankin 1997).

    Fish communities were sampled  using generator-powered, pulsed D.C. electrofishing units and a  standardized
methodology (Ohio EPA 1987a,b, 1989a,b;  Yoder and Smith 1999).  Fish community  attributes were  collectively
expressed by the Index of Biotic Integrity (IBI; Karr 1981; Karr et al. 1986), as modified for Ohio streams and rivers (Yoder
and Rankin 1995;  Ohio EPA 1987b, 1989b). Habitat was assessed at all fish sampling locations using the Qualitative
Habitat  Evaluation  Index (QHEI; Rankin 1989, 1995). The QHEI is a qualitative,  visual assessment of the functional
aspects of stream  macrohabitats (e.g., amount  and type of cover, substrate quality and  condition, riparian quality and
width, siltation, channel morphology, etc.). Ohio EPA also collected macroinvertebrate assemblage data at some of these
sites, but it was not included in this study because of the partial coverage and the extensive use of the qualitative method
was not always compatible with regression analysis. Some of the analyses in our earlier studies (Yoder et al. 1999)
included macroinvertebrate data.

    The urban land use indicator was derived from Landsat Thematic Mapper satellite imagery of  land cover classification
(September 1994)  provided by the Ohio Department of Natural Resources. The percentage of land use in the urban
classification was calculated for the subwatershed  upstream from each fish sampling location  to the boundary of the
watershed.  Because many of the sites included  in the statewide data set are subjected to a variety of stressors, each
site was qualitatively classified by predominant impact type. Impact types included least impacted sites, estate sites (i.e.,
subwatersheds with large lot sizes or green space provided by parks), sites reflecting gross instream habitat alterations
(i.e., channel modifications or impoundment), sites impacted  directly by discharges from combined sewer overflows
(CSOs), sites impacted by wastewater treatment  plant discharges, sites impacted by instream sewer line placement and
construction (Cincinnati area only), sites with  evidence of impacts by  legacy pollutants, or sites affected  by  general
urbanization only. This latter category included urban land uses not containing any of the other impact types and usually
consisted of residential development.

Results

    The relationship between the IBI and urban land use was initially  characterized  by regressing IBI scores against
percent urban land use (log,, transformed) and QHEI scores  using a database  of 267 sites  for all of the six major
metropolitan  areas  of Ohio (Figure  1). Diagnostic plots (e.g., residuals, normal  probability)  indicated nonconstancyof error
variance. To provide insights into whether the results varied substantially between each metropolitan area, scatter plots
of the relationship between urban land use and IBI in each of the six metro areas were also made (Figure 2). A negative
relationship between IBI and urban  land use was observed in four of the six areas, whereas little or no relationship was
seen in  two areas.  For each area, the  highest  percentage of urban land  use that corresponded  to minimum attainment
of the WWH  IBI biocriterion was determined by inspection of the scatter plot and the intersection of the regression line
and the WWH IBI biocriterion were determined (Figure 2). This ranged from 1%  (Cleveland/Akron) to 12%  (Dayton) for
the regression line, and 15% (Cleveland/Akron) to 58% (Columbus) as the highest %urban land use where WWH was
attained in each area  at a given sampling location. No significant linear relationship was found in either the Toledo or

                                                      35

-------
          CO
                                                                                     IMPACT  TYPES
                             EWH  Biocriterion
                               (IB I = 501
                                                                       WWH
                                                                     Biocriterion
                                                                    /(IBI = 40)
                                                  •   Habitat
                                                  ^   Background
                                                  a   Urban
                                                  0   WWTP
                                                  *    Legacy
                                                  ±   CSO/SSO
                                                  Q   Estate
                                                  ^   Sewer Line
                                                      •95%
                             4J* _     al •§    «U- — ^                  _X  '      '               	
                                £ S - »»o  - " --—'	Jl ' ~ -   •  • ' —  "• *    .    	95%
                             IS.TI-'  •   -r o   •  "-       ""*-*».^ » o  '    ' "  - ,-    	75%
                              . JB"»--.  * f             "       ""v	    *    '	1
                                V  O   l*-^p                 A
                               •  T   •np      "••-.        o       o                """--•.
                             •       AV      GO     "*••-...
          GO
        OMB     • •    •
T  •• HT         mm
O   OT   Ai     *
                                                                               O A   «
                                                                    • •""••-.. .A
                                                                          _      ' • .,
    •SB    T  •• BBT          mm       mm   ••-...A
 TB> • O    O    OT   Ai     *    A     B,       B
A3BB*                AlfT      B           	
L    TB B     A \   <3k        A*    *  B     A        ^
A             *     0» *                        (*     "^

3            °0       "
                        A A A
                                          AA    A
                                                                       T  A
                           10.0      20.0      30.0     40.0      50.0      60.0
                                                Percent  Urban Landuse
                                               70.0
                                                         80.0
Figure 1. Scatter plot of Index of Biotic Integrity (IBI) scores against percentage of watershed upstream from the site in urban land use at 267 small (<50
mi.2) sampling sites.

Youngstown areas, and only a weak relationship was visually  apparent for the Toledo streams. All of the sites in  the
Youngstown area were impaired, and so severely that no land use relationship was  evident (Figure 2). In the Toledo area,
the highest urban  land  use corresponding  to WWH attainment was 28%. However, the WWH IBI biocriterion in  the
Huron/Erie Lake Plain ecoregion is the lowest in the state and almost all of the small streams in the Toledo area have
been channel modified to some degree.  It was apparent that the lack of a stronger association  between IBI and urban
land use was due to overlying stressors (e.g., legacy pollutants, WWTPs, CSOs/SSOs), particularly those that resulted
in very low IBI values at sites with low levels of urbanization. While some threshold relationships were evident in these
results, the resulting variability in IBI scores led to only weak or non-existent linear relationships.

    Some of the impact types had a strong effect on the IBI regardless of the effect of urban land  use. The IBI results
were examined by  impact type across all six metro areas (Figure 3). The legacy, CSO/SSO, habitat, and WWTP impact
types had the strongest negative effects on the IBI, respectively, and this was independent of the urban land use  indicator.
While these impact types are common  to  urban areas, they  were  removed from the remaining  statistical analyses
(elimination of  these impact types  reduced the sample size to  123 sites) to better develop the  IBI/urban land  use
relationship. The entire Toledo and Youngstown datasets were  also removed since they are comprised entirely of these
impact types.  This  resulted in a better regression model fit, and  diagnostics consistent with regression model assumptions
(Neter et al. 1990). This also allowed us to  discern the threshold of urbanization at which WWH attainment is lost with
greater precision and in the absence of potentially confounding impacts, which was a major objective of our study. The
relationship between different levels of urbanization and biotic integrity was further quantified with an  analysis of variance
model where quartiles of percent urban land use determined factor level (e.g., all sites within the 1 st quartile of percent
urban land use were coded as factor level 1).  Similarly, an analysis of covariance model using QHEI as the covariate
was employed  to test for further refinements. Multiple comparisons  of factor level mean differences were made using
Tukey's method (Neter et al., 1990).
                                                      36

-------
                        50
                      IBI
                         10
                                  Regression Line
                                  Intersects W W H
                                 	(11%)
                           Maximum Observed
                           %Urban That
                           Meets WWH
                           127%)	
       Cincinnati Area 3(ream 4

                 .     i*    i
                                                         1H--I--
                        60
                        50  -
                        40  -
                     IBI
                        30


                        20
                                            10               100
                               Percent  Urban Landuse
                                                     Maximum Observed
                                       Regression Line    %Urban That
                                       Intersects WWH     Meets WWH
                              	(3%)	158%!.
                            '. Columbus Area ^Streams
                                            I
                                       I           10         100
                               Percent Urban  Landuse
                                                 Maximum Observed
                                                    %Urban That
                                                    Meets WWH
                                                       (28%)
  4O
IBI
  30


  20
                         10
                              Toledo Area Streams
                           0.1          1           10         100
                               Percent  Urban  Landuse
                        Maximum Observed
          Regression Line    %Urban That
          Intersects WWH    Meets WWH
              (1%)	(35%)
                                                                           10
                                                                              T^
                                                      O.I          I            10
                                                          Percent  Urban Landuse
                                                                           50
                                                                           40
                                                                        IBI
                                                                           30
                                                                        IBI
 50


 40
I
 30


 20


 10
                                                                              t   Youngstown   Area Streams
                                                             Regression Line
                                                             Intersects WWH
                                                                      (12%)
                          Maximum Observed
                          %Urban That
                          Meets WWH
                          (22%)
                                                                                 Dayton Area Stffam s
                                                      1                10
                                                          Percent U rban Landuse
                                                                     Maximum Observed
                                                                        %Urban That
                                                                        Meets WWH
                                                          	(0»)
                                                            r-  i  i  i i i i i|     i   i   i .
                                                      1                10               100
                                                          Percent  Urban Landuse
Figure 2. Scatter plots of Index of Biotic Integrity (IBI) scores against percentage of watershed upstream from the site in urban land use at small stream
(<50 mi.2) sampling sites in six of the major metropolitan areas of Ohio. Predominant impact types are indicated for each site  (see Figure 1) along with the
regression line. The warmwater habitat and exceptional warmwater habitat biological criteria  for the IBI are  also indicated.
                                                                    37

-------
                      60
                 IBI
                                 1
I
1
                                             I Maximum %Urban
                               O °          i   That Meets WWH
                                       O    c|   (exploding outliers)
                                           cc^o
                                                     o
                o
                        o o   o   o   o
                            i i  i iiiiiTTii  iii i  ii\\iiirniii \iirn

                          0     10    20    30     40     50     60    70    80

                                         Percent Urban Landuse
                 IBI
                      60
                      50
                     40
                      30
                      20
                      12
I
022.6 [
BO O30.2CJ18.4
MS, O O O ' /
¥XJ O ' ./
JljCL <$ir} 03 '^
i 0 00§ 0 « ',

-0 (P O i
- OO O i
a o i CD
- 0 O 0(33
o o
oa
I
-Avg. Age = 28.2 ' o
-Med. Age = 28.5 ]
i i i i i i i i i i i i i i i i i

Maximum %Urban
That Meets WWH
/- (excluding outliers)
019.8

	 O 37.8 	 "iiiji • O'"29.9 " — '{'WHV
nn- °17'5 •"'""


°8 Q
QD CD
O. O O
00 R °
u O
O
o
I I I I I I I I I I I I I I I I I I I I
_
_
-
_
-
_
_
-
—
-
-
-



1 1
                          0      10    20     30     40     50     60    70    80
                                         Percent Urban Landuse
    Figure 3. Box-and-whisker plots of Index of Biotic Integrity (IBI) scores by each of the major impact types used in Figures 1 and 2. The warmwater
habitat and exceptional warmwater habitat biological criteria for the IBI are indicated.
                                                           38

-------
    The percentage of urban land use explained approximately 35% of the variation in IBI scores in the regression model
when the other impact types were excluded. Local habitat quality (as measured by the QHEI) explained an additional
7% of the variation (Table  1). The ANOVA model showed  that there were significant differences in  mean IBI scores
between quartile level of percent urbanization, with sites exceeding 29 % urban land cover having lower IBI scores on
average than sites with less urban land  cover (Figure 4). Sites characterized by less than 4%  urban land cover had
higher IBI scores than sites with urban land use exceeding 15%  The ANCOVA model provided a slightly better fit, but
the additional variation explained was marginal (Table 2), and the results of pairwise comparisons were similar between
models (Figure 4).

Table 1. Regression Results for the Model Ibl = Log10 (Percent Urban Land  Use+l)+qhei for All  Sites and the Removal of Selected Impact Types.

Effect                   Coefficient           SE               t             P(2 Tail)         Adj.  R-Squared
All Sites
CONSTANT
Urban
QHEI
Impact Types Removed
CONSTANT
Urban
QHEI

21 .9333
-6.8323
0.2676

32.4069
-11.1496
0.2390

2.9450
1.1370
0.0418

4.2184
1.3102
0.0605

7.4477
-6.0092
6.4102

7.6822
-8.5096
3.9493

0.0000
0.0000
0.0001

0.0000
0.0000
0.0001


0.1179
0.2388


0.3500
0.4199
               CD   60
               oZ   50
               O
               UJ
O

O
CO
                    40
                    30
               O   20
               X
               LJU   -10
               Q    '^
                                                                            EWH Statewide
                                                                              Biocriterion
                                                                  WWH Ecoregion
                                                                     Biocriteria
            Minimum IBI = 12

             n = 8    n = 45   n = 1B   n = 94   n = 41    n = 26   n = 37   n=11
                             UJ
                                      o>
                                      o
                                      CO
                                      CD
                                 a)
                                (/>
                                                                    O
Figure 4. Distributions of Index of Biotic Integrity (IBI) from small streams (<50 mi.2) in the six major metropolitan areas of Ohio plotted by quartiles of
percent of urbanization upstream from sampling locations. Horizontal lines spanning adjacent box plots indicate similar means. Levels of percent of
urbanization corresponding to the 25th, 50th and 75th percentile are indicated. The shaded areas indicate the applicable warmwater habitat biological
criterion and the range of insignificant departure for the IBI.
                                                         39

-------
Table 2. Analysis of Variance Results for the Anova Model, and the Ancova Model Using Qhei as a Covariate.
ANOVA
Source
Urban
Error
ANCOVA
Source
Urban
QHEI
Error
Sum-of-Squares
4248.53
6129.10
Sum-of-Squares
4020.66
640.71
5488.39
df
3
119
df
3
1
118
Mean-Square
1416.18
51.51
Mean-Square
1340.22
640.71
46.51
F -ratio
27.50
F-ratio
28.81
13.78
P
0.0000
P
0.0000
0.0003
R-Squared
0.4094

0.4711
    In an attempt to better visualize where attainment of warmwater habitat occurs along the urban land use gradient,
the IBI results were plotted against percent of urban  land use for all sites used in this study and with the other impact
types excluded  (Figure 5). The elimination of the other impact types provided for a more precise statistical relationship
between urban  land use and the IBI (i.e., lower error  of regression estimates). For example, the R2 was higher with the
removal of the other impact types and the  slope of the regression was steeper,  both  of which suggest a more  meaningful
relationship between the IBI and urban land use (Table 1).   However, the percent of urban land use that corresponded
to attainment of the warmwater habitat IBI biocriterion  based  on  inspection  of  the scatter  plot was  the same
(approximately  26%) in both plots.
                  m
                        50
                        40
                        30
                        20
                        10
                           W W H Biocriteria
                               (IBI = 40)

                                        O

                                    -  • ,O
                                                                                   \
                                   1 st
                              <4.2%
        2nd
4.3-I  4.6%
       3rd
14.7-29.3%
     4th
>29.3%
                                  Quartile  Level  of  Percent  Urbanization
Figure 5. Scatter plots of Index of Biotic Integrity (IBI) scores against percentage of watershed upstream from the site in urban land use at small stream
(<50 mi.2) sampling sites in six of the major metropolitan areas of Ohio for all sites (upper) and a subset with non-urban impact types removed (lower). The
age of the urbanized area is indicated for selected sites and the mean and median age for entire dataset. The warmwater habitat biological criterion and
the range of insignificant departure for the IBI are indicated.
                                                        40

-------
    Also apparent in both plots was the occurrence of "outliers" where IBI scores above the warmwater habitat biocriterion
occurred at sites with 40% to 60% urban land use. These sites had either an intact, wooded riparian zone, a continuous
influx of groundwater, and/or the relatively recent onset of urbanization. Intact riparian buffers can mitigate the effects
of urban land use up to a point (Steedman 1988; Horner et al. 1997) and local hydrology can strongly influence the quality
of the fish assemblage (Poff and Allen 1995).  The three sites with the relatively  recent onset of urbanization (all <20
years) may not yet have accrued the types of negative effects that are  readily apparent in some of the  older urbanized
areas of Ohio.

Discussion

    Threshold levels of urbanization beyond which biological communities are likely to be impaired have previously been
identified in the range of 8% to 20% impervious cover within a watershed (Schuler  1994). Our previous analyses (Yoder
et al. 1999)  produced results of approximately 8% and 33% urban land use cover for the Cuyahoga River basin and
Columbus area streams, as identified by analysis of variance. We also concluded that the threshold level identified by
regression for the Cuyahoga River basin was lowered by the presence of other stressors (e.g., CSOs, point sources,
legacy pollutants). The elimination of those sites impacted by these other stressorsfrom the regression analysis resulted
in a higher threshold of urbanization. Our expanded study seemed to confirm this phenomenon, as the elimination of the
other impact types  helped  clarify the urban land use/IBI relationship in a broader array of urban  influenced  streams
throughout Ohio (Figure  5). The upper threshold of urbanization which corresponded to a loss of warmwater habitat
attainment was in the 25-30% range. However, our results show that non-attainment also occurs at lower thresholds of
urbanization  (Figure 5) due primarily to the co-occurrence of other stressors. This makes  both the linear and visual
derivation of sufficiently  precise indicator thresholds such as percentage of impervious surfaces more difficult.

    In terms of understanding the potential effect of urbanization on aquatic life use attainment, the most meaningful
results of our analyses are  the upper thresholds at which attainment of CWA goal uses are mostly lost (e.g., 25%) and
that  beyond which it never occurs (>60%). Only a very few sites exhibited full attainment of the warmwater habitat
biocriteria at urban  land  uses between 40-60%  (Figure  5). A closer examination of these sites and  the watersheds
showed  the presence of  high quality  riparian zones,  an  influx of flow augmenting groundwater,  and/or development of the
urban land use occurring within the past 20 years. For the latter, we hypothesized that the full effect of negative impacts
in an urban  setting may take time  to accumulate and may not be  immediately  manifest  in the form of instream
impairments. This could account  for the higher-than-expected  urban land  use (i.e., 40-60%)  correlating  with full
attainment of the  biocriteria. If this is true, then we might expect these sites to exhibit declines in IBI scores over the next
one  or two decades.  It  also suggests that newly urbanizing watersheds  should  be  developed with an emphasis on
determining which attributes (e.g., riparian zones, wetlands, flow regime) need to be  maintained and preserved in  order
to protect and maintain instream habitat and biological quality.

    The results of this study indicate that it might be possible to partially mitigate the negative  effects of urbanization by
preserving or enhancing near and instream habitats, particularly the quality of the riparian buffer zone. The "outlier" sites
that exhibited full  attainment of the warmwater  habitat  biocriteria had more extensive and higher quality riparian zones
and good to excellent instream habitat quality.  Some streams were nestled in small valleys which were not amenable
to development and the accompanying encroachment of urban land uses. This generally agrees with the findings of
Steedman (1988) who demonstrated a co-relationship between riparian zone quality and land use in terms of how each
affected the  fish communities and IBI values of Toronto area streams. Horner et al. (1997) also found that the negative
effects of urban land use were mitigated by riparian protection and other management interventions. However, in both
studies the quality and extent of the riparian zone ceased to be effective above 45-60% impervious land cover, which
generally corresponds to the thresholds identified by our study. Until we better understand the effect of the "age" of the
urban effect, it seems prudent to advocate policies that preserve existing riparian zones rather than responding with post-
urbanization  retrofits.

    Yoder et al. (1999) discussed the implications of their findings on the designation  of aquatic life  uses in state water
quality standards,  particularly to the  use  attainability  analysis process. Uses designated for specific water  bodies are done
so with  the  expectation  that the criteria associated with the use are  reasonably attainable.  If CWA goal uses  (e.g.,
warmwater habitat in Ohio)  are found to  be unattainable, then lower quality uses may be established and  assigned on
a case-by-case basis (40CFR, Part 131 .10[g]). Recently, the imperviousness of the watershed has  been suggested as

                                                     41

-------
an  indicator that is  correlated with  use attainability. If the frequently  cited threshold  of 25%  impermeability is used,
streams in watersheds with greater than this value could be considered  unlikely to ever attain a beneficial use regardless
of site- and  reach-specific factors. This assumes that the negative effects of urbanization cannot be remediated, which
has yet to be extensively tested. However, the results of our study suggest that there is a threshold of watershed
urbanization (e.g., >60%) beyond which attainment of the WWH use becomes increasingly unlikely, at least as affected
by  contemporary practices. This threshold is  not the same in all watersheds, as evidenced by the results from  the six
Ohio metropolitan areas, and it can occupy  a rather wide range. In  addition, co-factors such as pollutant loadings,
watershed development history, chemical stressors, and watershed scale influences such as the quality of the riparian
buffer and the mosaic of different types of land use,  also act singly and in  combination to determine to the resultant
biological quality in  the receiving streams. Thus, single  dimensional  urban land use indicators, such  as  watershed
imperviousness, is not sufficiently precise or reliable as  a single indicator of use attainability.

    The further development and  refinement of multiple indicators of watershed urbanization has  merit from a
management  and decision-making  standpoint. Because of the  many co-factors involved  (e.g.,  water quality,  habitat
quality, hydrologic regime, etc.), some of which are controllable and amenable to reasonable remediation, this will be a
complex undertaking. We suggest  that these co-factors, in  addition  to more refined urban land use  indicators, be
developed and tested using datasets from broad geographic areas spanning the extremes of the urbanization gradient.
Urban land  use and its analogs (e.g., %  imperviousness) are coarse approximations of the  cumulative effect of all
negative influences within a watershed. Thus  co-factors and more precise definitions of different urban land  uses need
to be defined in order to better understand and respond to the water quality management challenges posed in existing
and developing  urban areas.

    A management outgrowth of such an effort could be the development of an urban stream habitat use designation.
Yoder et al. (1999) previously indicated where the biological criteria for this potential new use designation might occur
compared to the already existing hierarchy of aquatic life  uses in the Ohio WQS (Figure 6). This designated use would
                              Max.
Quality Gradient of Aquatic Life Uses and Narrative
  Descriptions of Biological Community Condition
                             Exceptional
                             Warmwater
                            Habitat (EWH)
                             Index
                             Value
                            (IBI.ICI)
                              Min.
                                                   Warmwater
                                                 Habitat (VWVH)
            Modified
           Warmwater
          Habitat (MWH)
     Limited
    Resource
     Waters
     Exceptional"
"Very Good"
                                            "Poor"
                                      '"Very Poor"
                                  LOW-
                                               • BIOLOGICAL INTEGRITY^
                                                                           HIGH
Figure 6. Relationship between the tiered aquatic life uses in the Ohio WQS and narrative evaluations of biological community performance and how this
corresponds to a qualitative scale of biological integrity and of the biological indices that comprise the Ohio biological criteria. The position of a potential
new Urban Stream Habitat (USH) use designation is indicated (after Yoder et al. 1999).
                                                       42

-------
satisfy the desire to afford urban streams the maximum protection practicable, while recognizing the inherent limitations
that the irretrievable effects of urbanization may impose on stream quality. In the meantime, simplistic regulatory and
management approaches should be avoided, particularly in those watersheds where uncertainty about the attainability
of CWA goal uses (i.e., WWH and higher) exists. For example, a single indicator of urban development (e.g., proportion
of impermeable surfaces) is alone insufficient to drive this process. We envision that more refined, multiple indicators
of urban development will provide the necessary sophistication to more  appropriately define when this less than CWA
goal  use should be applied. In the meantime, management strategies such as the nine minimum controls for CSOs seem
reasonable analogies for the  management of urban watersheds and stormwater runoff. However, proceeding beyond
such minimum requirements with long-term remediation plans should be done with deference to the use attainability
issues and with the aid of sufficiently robust before-and-after biological  and water quality assessments.

References

Horner, R.R., D.B.  Booth, A. Azous, and C.W. May. 1997. Watershed determinants of ecosystem functioning,  pp. 251-
274. In  C.  Roesner (ed.).  Effects of Watershed Development and Management on Aquatic Ecosystems, American
Society  of Civil Engineers,  New York, NY.

Karr, J. R.  1981. Assessment of biotic integrity using fish communities. Fisheries 6(6):  21-27.

Karr, J.  R.,  K. D. Fausch, P. L. Angermier, P. R.  Yant, and I. J.  Schlosser. 1986. Assessing biological integrity in running
waters:  a method and its rationale. Illinois  Natural History Survey Special Publication 5: 28  pp.

Neter J., Wasserman W., and Kutner M. H. 1990. Applied Linear Statistical Models: Regression, Analysis of Variance,
and  Experimental Designs. Irwin, Homewood,  Illinois.

Ohio Environmental Protection Agency. 1998.  1998 Ohio water resource  inventory - volume I addendum, summary,
status, and trends. Rankin, E.T., Yoder, C.O., and Mishne, D.A. (eds.),  Ohio EPA Tech.  Bull. MAS/1998-6-1, Division
of Surface Water, Columbus, Ohio. 190 pp.

Ohio Environmental Protection Agency. 1990. Ohio's nonpoint source pollution assessment.  Division  of Water Quality
Planning and Assessment.. Columbus, Ohio.

Ohio Environmental Protection Agency. 1991.  1991 Ohio nonpoint source assessment. Ohio EPA. Division  of Water
Quality  Planning  and Assessment.. Columbus,  Ohio.

Ohio  Environmental  Protection Agency.   1989a. Biological  criteria for  the  protection of  aquatic  life,  volume  III:
standardized biological field sampling and laboratory  methods for assessing fish and macroinvertebrate communities,
Division of Water Quality Monitoring and Assessment, Columbus,  Ohio.

Ohio Environmental Protection Agency. 1989b. Addendum to biological criteria for the  protection of aquatic life, volume
II: users manual for biological field assessment  of Ohio  surface waters,    Division of Water Quality  Planning and
Assessment, Surface Water Section,  Columbus, Ohio.

Ohio Environmental Protection Agency. 1987a.  Biological criteria for the protection of aquatic life: Volume I.  The role
of biological data in water quality assessment.  Division  of Water Quality Monitoring and Assessment, Surface Water
Section, Columbus, Ohio.

Ohio Environmental  Protection  Agency. 1987b.   Biological criteria for the protection of aquatic life:  volume  II. users
manual  for  biological field assessment of Ohio surface waters, Division  of Water Quality Monitoring and Assessment,
Surface Water Section, Columbus,  Ohio.

Poff, N.L. And J.D. Allen. 1995. Functional organization of stream fish assemblages in relation to hydrological variability.
Ecology 76(2): 606-627.
                                                     43

-------
Rankin, E. T. 1995. The use of habitat assessments in water resource management programs, pp. 181-208. in W. Davis
and T. Simon (eds.). Biological Assessment and Criteria: Tools for Water Resource Planning and Decision Making.
Lewis Publishers, Boca Raton, FL.

Rankin, E. T. 1989. The qualitative habitat evaluation  index (QHEI), rationale, methods, and application, Ohio EPA,
Division of Water Quality Planning and Assessment, Ecological Assessment Section, Columbus, Ohio.

Schuler, 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. Can. J. Fish. Aq.  Sci. 45: 492-501.

Yoder,  C.O. 1995. Incorporating ecological concepts and biological criteria  in the assessment and management of urban
nonpoint source pollution, pp. 183-197. in  D.  Murray (ed.).  National Conference on Urban Runoff Management:
Enhancing Urban Watershed Management at the  Local,  County, and State Levels. EPA/625/R-95/003.

Yoder, C.O. and E.T. Rankin. 1998. The role of biological indicators in a state water quality management process. J.
Env. Mon. Assess. 51(1-2): 61-88.

Yoder, C.O. and E.T. Rankin. 1997. Assessing the condition and status  of aquatic life  designated uses in urban and
suburban watersheds, pp.  201-227. in Roesner, L.A. (ed.). Effects of Watershed Development and Management on
Aquatic Ecosystems, American Society of Civil Engineers, New York, NY.

Yoder, C.O. and E.T. Rankin. 1995. Biological criteria program development and implementation in Ohio, pp. 109-144.
in W.  Davis and T. Simon (eds.). Biological Assessment and Criteria: Tools for Water Resource Planning and Decision
Making. Lewis Publishers,  Boca Raton, FL.

Yoder, C.O.  and M.A. Smith. 1999. Using fish  assemblages in a  state biological assessment and criteria program:
essential concepts and considerations, pp. 17-56. in T. Simon (ed.). Assessing the Sustainability and Biological Integrity
of Water Resources Using Fish Communities. CRC  Press, Boca Raton, FL.

Yoder, C.O., Miltner, R.J., and D. White.  1999. Assessing the status of aquatic life designated uses in urban  and
suburban watersheds, pp. 16-28.  in Everson, A. et a/, (eds.). National Conference on Retrofit Opportunities for Water
Resource Protection in Urban Environments, Chicago, IL. EPA/625/R-99/002.
                                                    44

-------
                                     Getting Past the Obvious
                                           Robbin B. Sotir, President
                                       Bobbin B. Sotir & Associates, Inc.
                                             Marietta, GA 30064
Abstract

    Long Leaf Creek is located within an urbanized watershed along coastal North Carolina. The specific stream reach
addressed  is located  in a  residential  subdivision. Conditions had dramatically changed there  due to the continued
development of the watershed. The  stream had deepened and widened as  result  of increased  runoff and high
concentration events, including hurricanes. This increased the loss of aesthetic value, riparian corridor vegetation, and
aquatic and terrestrial habitats. Water quality was also degraded.

    Before they could decide how best to control flooding and stabilize and restore Long Leaf Creek to a naturally
functioning  channel within its changed watershed conditions, citizens had to be educated about natural stabilization and
restoration  technologies and specificmethods that would work, including conventional options. Soil  bioengineering was
agreed  upon, with numerous modifications to meet specific needs.

    This paper  is presented from both  the client's and consultant's perspectives. It identifies what worked, what did  not
work, and what was necessary to improve the process for successful, long-term results. We present the lessons learned
from criteria issue development and understanding, educational process alternatives preparation, design, construction,
and project results since construction.

Paper

    Long Leaf Creek is located within an urbanized watershed along coastal North Carolina. The specific stream reach
addressed  is approximately 2000 feet in  length and is located in a residential subdivision. It is a highly sensitive project
with a variety of multi-objective goals specific to its location and function and typical to urbanizing areas. The watershed
includes residential, office, institutional, and commercial properties, including 25 homes that line the creek in this area.
Residents living along the creek described the former Long Leaf, as a  small, picturesque stream that pleasantly flowed
through their neighborhood—a stream that could be jumped across. It was enjoyed by many people. The conditions had
dramatically changed, however, due to the continued development of the watershed, especially a new road and large
shopping center immediately upstream. Residents  have seen their stream deepen by almost ten feet and widen by 40
feet in areas, a  result of increased runoff and high concentration events, including  hurricanes. In many areas, the banks
were vertical. Large, woody  debris filled much of the channel, and many people were now using the "ditch" as a yard and
construction waste dump (see  Figure  1). This has resulted in increased the loss of aesthetic value, riparian corridor
vegetation,  and aquatic and terrestrial habitats. Water  quality has also been degraded.

    Many people had already lost property due to stream widening  and were unwilling to lose more. Flooding was a major
problem in  the downstream end, while erosion was occurring throughout the project reach. The City of Wilmington was
interested in exploring a natural approach to solving the problem. After assessing the site and conditions, and listening
to the residents=  concerns  and desired solutions,  it was clear that a strong, continual working  relationship had to  be
formed with the neighborhood to ensure project success.


                                                     45

-------
Figure  1.  Pre-Construction Conditions
                                                                  46

-------
    Before they could decide how best to control flooding and stabilize and restore  Long Leaf Creek to a naturally
functioning channel within its changed watershed conditions, citizens had to be educated about natural stabilization and
restoration technologies and the specific methods that would work, including comparative conventional options. To "get
past the obvious," it was clear that almost everyone would have to give up some land  and existing trees to solve their
continued land loss and flooding problems and to improve the environmental and aesthetic values of Long Leaf Creek.
How much land and how many trees they would lose would ultimately depend on their  selected restoration alternative.
A matrix was developed using critical issues and matching these to possible alternatives (Table 1). Soil bioengineering
was agreed upon, with numerous modifications to meet specific needs.

    Bobbin B. Sotir & Associates, Inc., (RBSA), served as the soil bioengineering consultant to the prime, the Kimley-
Horn's interdisciplinary team, developed the geotechnical design and  hydraulic eff iciencies of a soil bioengineering solution
to address the desired goals and critical  engineering, environmental, and aesthetic issues.

    Alternatives were compared with such critical issues as erosion control, streambank  stabilization, safer and healthier
environment,  flood control,  timely project completion, environmental and  aesthetic  improvement,  property  loss
minimization, hydraulic efficiency, and cost feasibility.

    After an initial investigation, an alternative analysis was produced in the summer of 1997. This alternative analysis
explored numerous approaches to solving  each of the project goals, with cost and risk factors  assigned to each
alternative. Several alternatives were  considered, such as box culverts, 3:1 (horizontal :  vertical) grassed slopes, 2:1
riprap  rock, 2:1 concrete lining, and  soil bioengineered slope systems. With  input from the  residents and permit
authorities, the City selected the soil bioengineering approach and commissioned a design team to produce  plan and
specification documents, including construction cost estimates.

    The selected systems employed the use  of live fascines, brushlayer/live fascines,  joint  planting and vegetated
geogrids (see  Figures 2 through 5).

    The majority of the improvement was done using vegetated geogrid, due to its soil reinforcing capabilities and ability
to reduce  land losses (see Figure 6).

    Pre-bid, preconstruction,  and permit  application services were provided to support the project. Construction was
completed by the spring of 1999.

    The project has performed well from a biological perspective. Willow, baccharis, and myrtle installed as cuttings in
the lower layer had a survival rate of approximately 80%. The rooted stock installed in the upper two layers comprised
of spirea bush, Carolina allspice, serviceberry, and viburnum, were  less successful, with a survival rate of approximately
60% due to an insect infestation (see Figure  7). Hydraulically,  we  have had some bed scour, accompanied by toe
erosion.

    The survival rate of the rooted stock would have been higher had the watering maintenance program been followed.
It is also possible that the insect infestation would not have occurred if the plants had been  kept healthier by better
maintenance practices. The rooted plants will  be replaced by the contractor under the maintenance agreement. The
contractor is also  responsible for taking  care  of the insect infestation. The bed scour in  the upper level caused  by
Hurricane Floyd is being  handled with check dams to stop the bank from lowering and to control the toe scour.
                                                      47

-------
Table 1. Long Leaf Hills/Hewletts Creek Alternatives and Critical Issues

                                                                       Alternatives

Critical                   Intermediate       3:1 Side Slopes    2:1 Side  Slopes         2:1  Side Slopes         Reinforced            Soil
Issues                     Action          Grass Lining'        Riprap Rock'      with Concrete Lining*      Box Concrete      Bioengineering

Stop Erosion                                     •                  •                      •                    n/a               •
&  Stabilize
Banks

Clean Out                   •                  •                  •                      •                    •                •
Trash &
Debris,  Remove
Fallen Trees

Safer &                                          •                                                                                 •
Healthier
Area

Control                                           •                                         •                    •                •
Flooding

Timely                       •                  •                                         •                    •                •
Project
Completion

Environmental                                    •                                                               •                •
Improvement

Aesthetically                                     •                                                              n/a                •
Enhancing

Meets  Hydraulic                                  •                                         •                   n/a                •
COE and •
Environmental
Permits
Approval
Probability
Minimize •
Property
Loss
Preliminary $250,000 $640,000 $900,000
Cost Estimate to to to
Range $400,000 $800,000 $1,400,000
$785,000 $1,750,000 $1 ,000,000
to to to
$1,200,000 $2,300,000 $1,300,000
'Does not address geotechnical issues of sandy bank material stability and major land loss requirements.
*Does not address increased safety concerns or reduction in property values.
                                                                    48

-------
Figure 2. Live Fascine.
 Figure 3. Brushlayer/Live Fascine.
                                                                      LIVE CUT
                                                                      BRANCHES
                                                           LIVE STAKE
                                                                 JUTVE COT BRANCHES
                                 EXCAVATED TERRACE '
                                                                       LIVE FASCINE
                                                          49

-------
                                                                                                 EXISTING
                                                                                                 RIPRAP
                                                                                        EXISTING
                                                                                      LIVE STAKE
Figure 4.  Joint Planting.
                 HLL MATERIAL
                 LIVE BRANCHES
                                                                                              EXISTING
                                      - COMPETENT FOUNDATION -
    Figure 5. Vegetated Geogri
                                                                    50

-------
Figure 6. During Construction.
Figure  7. Three  Months After Construction.
                                                                  51

-------
    The project is functioning well from the bank stability and flood control aspects and the stream is operating within the
parameters of the new watershed conditions.  It is aesthetically attractive and,  over time, should develop some ecological
diversity. In summary, it is clear that the soil bioengineering approach is succeeding. The most important lessons learned
were as follows:

    .  Learn more about the bed conditions in areas that have had high deposits of mobile materials

    .  Employ sophisticated grade control structures

    •  Ensure installation procedures are followed correctly and that materials are not changed

    .  Keep tabs on the contractor's maintenance schedule

    •  There is no substitute for communication and cooperation

Reference

Sotir,  R.B.; "Brushing Up On Erosion," American  Cities and Counties, February, 1998.
                                                      52

-------
                          Protecting and Enhancing Urban Waters:
                               Using All the Tools Successfully

                                             Eric H. Livingston
                                   Stormwater/NPS Management Section
                               Florida  Department of Environmental Protection
                                      Tallahassee, Florida 32399-2400
Abstract
    Reducing the hydrologic effects and pollutant loadings from urban drainage systems and restoring aquatic habitats
to improve the  health  of our aquatic ecosystems presents many unique challenges. These challenges can  be categorized
as technical, institutional, financial, and cultural. This paper will examine each of the challenges and the tools that have
and are being developed to overcome them. A case study on how the tools are being used in Florida to enhance the
health of the Tampa Bay aquatic ecosystem will be presented.

Introduction

    Research conducted in Florida during the late 1970s characterized stormwater pollutants, provided cost and benefit
information on  many types of stormwater treatment practices, and determined the importance of stormwater discharges
as a major source of pollution. As a result,  in 1979, the Florida Environmental Regulation Commission adopted the state's
first stormwater treatment  requirements.  In 1982, the state's stormwater rule  was  fully adopted, requiring all new
development and redevelopment projects to include site appropriate BMPs to treat stormwater. This technology-based
program establishes a performance standard of removing at least 80% of the average annual post-development loading
of total suspended solids for stormwater discharged to most waters. Stormwater discharges to the state's most pristine
waters,  known  as Outstanding Florida Waters, are required to reduce pollutant loading by 95%.

    Florida's stormwater treatment  program, in combination with the  state's wetlands protection, land acquisition, and
growth management programs, has greatly minimized the effects of Florida's rapid growth on its water bodies. However,
land uses and  hydrologic alterations that occurred before the mid-l 980s has continued to adversely affect the state's
vulnerable and  valuable  aquatic ecosystems.  Accordingly, the focus of  Florida's watershed  management  program shifted
to cleaning  up  "older sources" such as existing land uses, whether  urban or agricultural, and to  integrating program
components to eliminate duplication and improve efficiency and effectiveness. This has led to greater emphasis on more
holistic approaches to address cumulative effects of land use activities within a watershed and to a greater emphasis on
regional structural controls and the purchase or restoration  of environmentally sensitive lands. The key institutional
components of this watershed approach have been described in detail (WMI, 1997; Livingston et al., 1995).

Development of  Florida's Watershed Assessment Tools

    Florida's Water Implementation Rule, Chapter 62-40, F.A.C.,  establishes a performance standard for reducing, on
a watershed basis, the pollutant loading from older stormwater systems. The goal is to protect, maintain or restore the
beneficial uses  of the  receiving water body. The amount  of needed  pollutant load reduction is  known  as a "Pollutant Load
Reduction Goal or PLRG". The rule further specifies that PLRGs be established as part of the states's priority watershed
program, Surface Water Improvement and Management (SWIM), which is implemented by the  state's five water
management districts. Consequently,  stormwater PLRGs have been established for several water bodies leading to the
development and implementation of watershed plans to reduce pollutants from urban and agricultural stormwater
discharges. For example, farms within the Everglades Agricultural  Area have implemented BMPs to reduce phosphorus
loadings by 40%. Additionally, the federal and state government, SFWMD, and landowners  within the EAA are sharing
the cost of constructing  tens of thousands of acres of wetlands (Stormwater Treatment Areas) to provide for additional
reduction of phosphorus.

                                                    53

-------
    Having a sound institutional framework, however, is only one of the tools needed to successfully reduce stormwater
pollutant loadings from existing land uses. Equally important are funding and public education to promote the cultural
change necessary to reduce "Pointless Personal Pollution". A final cornerstone is good science -to establish ecologically
meaningful watershed management goals, evaluate the effectiveness of BMPs and management programs, and assess
the cumulative effects of wet weather discharges.

    Unlike traditional point sources of pollution, the effluent quality and environmental effects of stormwater and other
nonpoint sources of pollution  are highly variable  because of their intermittent,  diffuse, land use-specific nature. Of
particular environmental concern is the  cumulative impact on a water body from the numerous stormwater/nonpoint
sources within a watershed. Consequently, traditional water quality monitoring and management efforts used for point
discharges generally suffer from several deficiencies when trying to understand and manage stormwater/NPS pollution.
These deficiencies include difficulty in:

    1. Assessing intermittent, shock loadings of pollutants.

    2. Assessing cumulative impacts of multiple sources.

    3. Comparing water bodies and establishing priorities  for management actions.

    4. Assessing hydrologic, geomorphological, and habitat alterations within a watershed.

    5. Distinguishing actual or potential  problems from perceived problems.

    6. Discriminating anthropogenic loadings from natural watershed loadings of metals and nutrients.

    7. Establishing cost-effective ways to assess pollution sources and trends on a watershed  basis.

    To overcome these problems, the Florida  Department of Environmental Protection (DEP) has developed  cost-effective
sediment and biological monitoring tools that are much  better suited for assessing cumulative effects than traditional water
chemistry monitoring (Livingston et. al,  1995; McCarron et al., 1997). Most stormwater pollutants accumulate overtime
in sediments, not the water column. Therefore, the sediments and the organisms that reside in them offer an in-situ
monitoring opportunity to determine the cumulative effects  of watershed stormwater/NPS pollution sources on aquatic
systems or to evaluate the effectiveness of management programs.

    Sediment quality is a sensitive indicator of overall environmental quality. Sediments influence the environmental fate
of many toxic and  bioaccumulative substances in  aquatic  ecosystems (Long and Morgan, 1990). Sediments tend to
integrate contaminant concentrations over time and may  represent long-term sources of contamination. Specifically,
sediment quality is important because many toxic contaminants found in only trace amounts in water can accumulate to
elevated  levels in  sediments.  Sediment-associated  contaminants  can also directly affect benthic and  other
sediment-associated organisms since sediments provide benthic and pelagic communities suitable habitats for essential
biological processes (e.g. spawning, incubation, rearing, etc.).

    Sediments  provide  an  essential  link between chemical and biological  processes. By  understanding  this link,
environmental scientists can develop assessment tools and  conduct monitoring programs to more rapidly and accurately
evaluate the health  of aquatic systems  (Pardue  et  al.,  1992). Therefore,  sediment quality  data  provide  essential
information for evaluating ambient environmental quality conditions in water bodies. Additionally, information about the
amount and quality of sediments within stormwater systems, stormsewers and other stormwater conveyances can help
trace  pollution sources, prioritize  areas  for implementing  control measures, and help to assure proper disposal of
accumulated sediments.

    Inclusion  of biological community monitoring allows a more holistic,  systems approach  that greatly  enhances surface
water quality assessment and management (Yoder, 1989; Yoder and Rankin, 1997). In particular, it allows assessment
of the degradation  of habitat (e.g., channel and bank erosion)  and  siltation within water bodies, neither of which are


                                                     54

-------
detected by water chemistry sampling, and both of which are typically associated with wet weather discharges. While
chemical data reflect short-term conditions that exist when a particular sample is collected, biological communities
accurately indicate overall environmental health because they continuously inhabit receiving waters where they integrate
a variety of environmental influences - chemical, physical and biological.

    Biological assessment involves an integrated  analysis of functional and structural components of aquaticcommunities
(Karr and Dudley, 1981; Karr, 1991). Bioassessments are best used to detect aquatic life impairments and assess their
relative severity. Once  an impairment is detected, additional chemical and biological toxicity testing  can  identify the
causative agent and  its  source.  Both biological and chemical  methods play critical roles in successful pollution control and
environmental management programs. They  are complementary, not  mutually exclusive,  approaches that  enhance overall
program effectiveness.

    A fundamental part  of bioassessments is "metrics"  (Karr and Chu, 1998). Just as  a doctor uses metrics such as blood
pressure and  heart rate to assess human health, biological  metrics allow the ecologist to  use meaningful indicator
attributes to assess  the status of communities in response to perturbation. The definition of a  metric is  a characteristic
of the biota that changes in some predictable way with increased human influence (Barbour et al., in review). By using
multiple metrics to assess biological condition, the information available about the  elements  and processes of aquatic
communities is maximized.  The validity of an integrated assessment using  multiple metrics is supported by the use of
measurements of biological attributes firmly rooted in sound ecological principles (Fausch et  al. 1990; Lyons 1992).

    In  1983, the DEP began developing assessment tools that could  be  used to assess  stormwater and NPS  effects and
the  effectiveness of management programs,  practices, and activities.  The  first efforts focused on  estuarine sediment
assessment tools (FDER, 1988; Schropp et al., 1989,1990; MacDonald, 1994). In 1989, efforts began to modify national
bioassessment protocols (EPA, 1989) to develop quantitative  bioassessment  protocols for Florida (Griffith et al., 1994;
Barbour et  al., 1996, FDEP, 1996). The tools that have been developed and that are under development (noted by a *)
include:

    A.  Sediment assessment  tools

        1.  Standardized sediment collection and  analysis  protocols.

        2.  Estuarine normalization  of metal concentrations to aluminum concentrations ratio.

        3.  Estuarine sediment quality assessment  guidelines

        4.  Freshwater normalization of metal concentrations to aluminum concentrations ratio(*).

    5.   Freshwater  sediment quality assessment guidelines(*).

    B.   Bioassessment   tools:

        1.  Stream  Condition Index

        2.  Lake Condition Index (* - nearly completed)

        3.  Wetland bioassessment methods (* - in  early stages of development)

        4.  Canal bioassessment methods (* - in middle stages of development)

        5.  Estuarine bioassessment methods  (* - in early  stages of development)

Using the Tools for Watershed Management

    Sediment assessment, together with watershed characterization, future land use plans, stormwater master plans,
and mapping of pollution sources, can be used to screen watersheds and sub-basins to determine potential  "hot spots".

                                                     55

-------
Bioassessment and water chemistry sampling can then be done to assess the actual health of the aquatic system in these
locations. An important component of the bioassessment is habitat quality, especially since urbanization often leads to
dramatic changes in stream hydrology, geomorphology, riparian zones, habitats, and ultimately biological communities.
Possible outcomes of the bioassessment are: (1) no biological effects; (2) effects due to habitat degradation; (3)  effects
due to sediment or water quality; or (4) effects due to a combination of sediment, water quality and habitat degradation.
Bioasssessments also allow the establishment of  an ecologically-based aquatic resource goal, rather than one based
solely on traditional water chemistry standards to which the applicability to wet weather discharges is questionable. Once
an aquatic resource goal is established, then  relationships between the  needs of the biological community and the
chemical, sediment, and habitat influences on the community can be established. The next step is to then use all of this
information to develop a watershed management plan that includes specific projects, schedules, and funding sources that
will ultimately  result in the acheivement of the desired aquatic ecological goal. Much of this effort can  be coordinated and
implemented  using the rotating basin  approach  that  many states are now using in combination with the NPDES
stormwater permitting  program.

Putting the Puzzle Pieces Together:  Tampa Bay Case Study

    Puzzle Piece 1: The Assessment- As part of the development of the Tampa Bay SWIM Plan (SWFWMD, 1992) and
the  Tampa Bay  NEP  Comprehensive  Conservation and Management Plan  (TBNEP, 1996), existing environmental
information was assessed to determine the ecological health of the bay system. Major findings of these assessments are
summarized below.

Habitats

    1.   Since 1950, about  half of the bay's natural shoreline and 40% of its seagrasses have been destroyed.

   2.   In 1950, the bay's shallow shelf supported  about 40,000 acres of seagrasses. By 1982,  only 21,600 acres
        remained and virtually all of Hillsborough Bay's 2,700 acres were gone. Seagrass decline is due to dredging and
        filling for residential development, turbidity caused by dredging of the main shipping channel, and reduced light
        penetration caused by shading by algae fueled by excess nutrient discharges.

   3.   Since the  early 1900s approximately 13,200 acres  of bay bottom (3.6% of the bay's surface area) were filled, with
        over 90% of the activity occurring along the bay's  shallow shelf where seagrasses once thrived. The surface area
        of Hillsborough Bay has been reduced by  14%.

   4.   Upgrading sewage plants in  the 1980s to  provide advanced wastewater treatment reduced  nitrogen loadings,
        leading to a decline in phytoplankton, an increase in water clarity,  and greater light penetration. Consequently,
        between  1982 and  1992 seagrass coverage increased  by about 4,000 acres  (18.5%) raising the bay's total
        acreage to over 25,600  acres.

   5.   About 43% (9,700 acres) of Tampa Bay's original saltwater wetlands were lost between 1950  and 1990, primarily
        because of dredging and filling for waterfront development. However, as many as 5,900 acres of new wetlands
        formed along causeways and other emergent land created by dredged spoil material  during this period. Recent
        estimates  of wetland habitat  in Tampa  Bay  indicate that about 18,000 acres of mangroves  and saltmarsh  remain
        but many thousands of acres are damaged by invasion by exotic plants such as Brazilian pepper.

Fish and Wildlife

    1.   Between 1966 and 1990, the harvest of 11 commercial species offish declined by 24%, primarily because of
        smaller catches of  mullet and sea trout. Each of these species is  dependent on seagrass habitats.

   2.   Harvest of spotted sea trout declined by 86% between 1950 and 1990, from 487,000 pounds to 67,000 pounds.
        Similarly,  red drum harvests  plummeted from 80,000 pounds in 1950 to 15,000 pounds in 1986.
                                                    56

-------
    3.   Tampa Bay's once-thriving shellfish industry has virtually collapsed, except for bait shrimping. Harvests of clams
        or  oysters are restricted or prohibited throughout the bay because of high  bacterial levels associated with
        stormwater discharges and  septic tanks.  The  bay scallop, a  highly pollution  sensitive  organism,  all but
        disappeared from the bay in the 1960s.

 Water and Sediment Quality

    1.   While water quality  has improved over the past ten years, primarily as a result of better wastewater treatment,
        water clarity, nutrients, and toxics continue to be a problem.

    2.   Because of natural  circulation and flushing from the Gulf of Mexico water clarity is greatest in the lower part of
        Tampa Bay (2.5 m), and naturally decreases moving up the bay, dropping to an average of 2 meters (6.6 feet)
        in Middle Tampa Bay and Old Tampa Bay. The lowest average water clarity is in Hillsborough Bay (1.5 m) which
        has poor circulation and receives a larger share of nutrients and sediments from major rivers.

    3.   Excessive amounts of nitrogen continue to accelerate algal growth which subsequently reduces light penetration
        to seagrasses and contributes to oxygen depletion. The bay's total annual nitrogen load was estimated to be 2.5
        times greater in 1976 than the load computed for 1985 to 1991  (Figure 1).

    4.   Recent studies by NOAA, in cooperation with FDEP,  provide an excellent overview of the levels and distribution
        of toxics in bay sediments (Long, et. al. 1991, 1994;  FDEP, 1994). Compared to other urban estuaries, Tampa
        Bay has low-to-moderate levels of most toxic  parameters. Contamination  is centered around large urban  centers,
        ports and marinas,  and concentrations generally decrease from the top of the bay toward the Gulf of Mexico.

    5.  Generally,  the highest  levels of sediment toxic  contamination  occur in  Hillsborough Bay, the  bay's  most
        industrialized area  and home to the state's busiest port. Upper Hillsborough Bay  has the highest levels  of
        cadmium, copper, mercury, zinc, and lead, as well as the pesticide DDT. Concentrations in sediments at a site
        in northern Hillsborough Bay were the highest  of any toxic pollutant  measured  in  Tampa Bay. Two other bays with
        heavily  urbanized watersheds,  Boca  Ciega  Bay and Bayboro Harbor, also can be considered  as hot spots of toxic
        contamination.

    6.   Figure 2 shows sites in Tampa Bay  where concentrations of toxic contaminants in sediments exceeded Florida's
        Probable  Effects Level (PEL) and No Observable Effects  Level (NOEL). Sites above the  PEL indicate a  high
        probability for biological impact to marine organisms while those above the NOEL are considered "at risk" to
        biological  impact (MacDonald, 1994).

    Puzzle Piece 2: The Goal: A critical component of watershed management is using biological living resources as a
measure of a water body's health, with far less emphasis on traditional laboratory-based water quality standards.  This
approach addresses critical  ecological effects that are not seen by water chemistry standards, allows greater flexibility
to achieve the desired ecological goals, and provides taxpayers with a  better benchmark to judge the return on  their
expenditures. Through the SWIM Plan and the Tampa Bay NEP CCMP, the primary overall goals have been established
for the restoration and protection of Tampa Bay:

    1.   To reverse the environmental degradation of the Tampa Bay estuarine system.

    2.   To  optimize water quality and  other  habitat values, thereby  promoting the sustained  existence or  re-establishment
        of thriving, integrated,  biological communities.

    3.   To ensure the maintenance of a productive, balanced ecosystem complimentary with human needs and uses of
       the resources.
                                                     57

-------
                               Figure 1. Tampa Bay Total Nitrogen  Loading  and Sources
                                                                            Hilltborough Bay
                                                                            121 tons/
                                                                         S§322 tons/  17.3%
                                                                            236 tons/  12.7%
                                                                            195 tons/  10.5%
                                                                            Total 1,856 tons
  Old Tampa Bay
  172 tons/ 34.5%
  250 tons/ 50.1%
  77  tons/ 15.4%
      ton/  0%
  Total 499 tana
                      Middle Tampa Bay
                      340 tons/ 46.2%
                      339 tons/ 46.1%
                      26  tons/  3.7%
                          ton/    0%
                      Total 70S ton*
                     Boca Clega Bay
: 83 tons/ 43.2%
: 106 tons/ 55.2%
 3  tons/  1.6%
 Total 192 tons
                                                                                Middle Tampa Bay
                                                   ^J  BaeaCiega
                                                          Boy
                                                                                      • 10 tons/  29.4%
                                                                                      Li 20 tons/  58.6%
                                                                                      H 4 ton*/11
                                                                                          Total 34 ton*
  Lower Tampa Bay
  1    ton/  0.3%
  302 tons/ 82.7%
  36 tons/  9.9%
  26 ton*/  7.1%
      ton/   0%
  Total 365 tona
                                                       tmw Tampa Bay ;
                                                       r *•?.    "
                         •  Stormwater Runoff
                         U  Atmospnaric Deposition
                             Industrial & Municipal Point Sources
                             Fertilizer Material Losses
                             Springs & Grouodwater
                                                                  364 tons/  63.3%
                                                                  53  tons^   8.7%
                                                                  170 tons/  26.0%
                                                                  Total 607 tons
Figure 1. Tampa Bay Total Nitrogen Loading and Sources.
                                                                     58

-------
                               Figure 2. Sediment Toxic Hot Spots in Tampa Bay



            Figure 2 shows sites in Tampa Bay where concentrations of toxic

            contaminants  in sediments have exceeded Florida's Probable Effects

            Level (PEL) and No Observable Effects Level (NOEL) for biological

            impact. Sites registering above the PEL indicate that some biological

            impact to marine organisms is likely. Sites registering above the NOEL

            are "at risk" to biological impact.
                               = -A-
                                    1
                                         A
                                                                                Tampa       -^j
                                 St. Petersburg
                                                                      Tampa
                            4
                            «
                            A
*%••
                                 , A 1
                                    Sites exceeding
                                    Probable Effects Level

                                    Sites exceeding No
                                    Observable Effects
                                    Level
Figure 2. Sediment Toxic Hot Spots in Tampa Bay.
                                                       59

-------
    To achieve these overall goals the following specific goals have been established:

    1.   The overall goal is to restore seagrasses to 1950s levels. This will lead to restoration of commercially important
        species such as the bay scallop, mullet, sea trout, and red drum.

    2.   To restore seagrasses to 14,000 acres of the bay. The ability of seagrasses to recolonize the bay depends on
        the amount of sunlight the grass species require, as well as shading factors such as the amount of drift macro-
        algae and  attached algal growth on grass blades. For most seagrasses in the bay, an estimated 20% to 25% of
        the light striking the bay's surface must penetrate to target depths to allow seagrass regrowth.  Reducing nitrogen
        loadings will  reduce chlorophyll a concentrations thereby increasing the depth of sunlight penetration.

    3.   As many as 12,000 acres of seagrass can be recovered by maintaining recent water quality conditions. This will
        require local communities to reduce their nitrogen  loadings to the bay  by about 10% by the year 2010 to
        compensate  for increases in  nitrogen loadings associated with the watershed's population growth.

    4.   A coastal habitat master plan has been developed for the watershed that will help to coordinate and prioritize
        existing state, regional, and local restoration programs. The long term goal is to recover 1,800 acres of low-
        salinity tidal marshes while maintaining  and enhancing salt marshes and mangroves  at existing levels. A minimum
        goal is to restore 100 acres of tidal marsh habitat every five years.

    5.   Reduce sediment toxicity to minimize risks to marine life and humans. Using three tests -evaluation of sediment
        chemistry,  sediment toxicity, and benthic community health - bay sediments will be characterized and prioritized
        for management.

    6.   Reduce bacterial contamination to levels safe for swimming and shellfish harvesting.

    Puzzle Piece 3:  The PLRG: To achieve the ecological goal,  nitrogen loadings would need to  be  held to those
occurring in 1992-94 meaning that the 17 tons of nitrogen loading that would accompany  projected growth within the
watershed would have to  be compensated.  Reduction  of nitrogen would reduce chlorophyll a levels which would increase
how deep the minimum levels of light needed for seagrass growth would penetrate the water column.

    Puzzle Piece 4: Quantifying  Pollution Sources:  Before load  reductions can be achieved, it is essential that the sources
and  relative contribution of the sources be quantified.

    Figure 3 summarizes sources of nonpoint nitrogen loadings to Tampa Bay.

    1.   Stormwater runoff from the Tampa Bay watershed contributes about 47% of the bay's total annual nitrogen load
        with urban runoff accounting for about 16%, or 680 tons. Residential areas, the watershed's largest land use, is
        responsible for over half of the nitrogen loading while commercial/industrial sites account for about 20%.

    2.   About 28% of the bay's total nitrogen loadings, or 1,200 tons, are from atmospheric pollutants falling directly on
        the water.  An additional 7,500 tons fall in the watershed, although no one can determine how much enters the
        bay via stormwater. EPA estimates that as much as 67% of the bay's  total  nitrogen load  may be from the
        atmosphere.

    3.   Stationary  sources, primarily power plants, contribute an estimated  50% of the anthropogenic  NO, emissions as
        compared  to 35% from  motor vehicles.

    4.   Domestic wastewater discharges still discharge about 8% (340 tons) of the bay's total  annual nitrogen loadings,
        even though  all plants provide AWT. Hillsborough Bay receives about two-thirds of the cumulative nitrogen load
        from the 36 billion gallons of effluent discharged to Tampa Bay each day.

    5.   Industrial  wastewater discharges, primarily  fertilizer manufacturing  and shipping facilities, are responsible for
        about 6% of  the bay's total annual nitrogen loadings.
                                                     60

-------
                                     Figure 3. Total Nitrogen Loadings
Per-Acre Nitrogen Loadings
from Non-Point Sources

Ruidential
Commercial Industrial/Institutional
Mining
Range and Pasture
Intensive Agriculture
Undeveloped Land
% Loading
11
5
4
13
6
8
* Watershed
15.5
6.4
3.2
20.4
6.5
39.93
Yield Iba/oc/yr
4.52
5.26
4.97
2.81
5.63
1.15
                        Total nitrogen Loadings to Tampa Bay (1882-1894 average)
                                     6%
                                  Industrial
                                 Wastewater
                                                          26%
                                                       Atmospheric
                                                        Deposition
                               8%
                            Municipal
                           Wastewater
                              6%
                           Accidental
                        Fertilizer Losses

                               4%
                            Groundwater
                              4%
                             Mining
    47%
Stormwater
   Runoff
                      5%
                   Commercial/
                 Industrial Runoff
                                    8%
                                Undeveloped
                                    Land
                6%
             Intensive
            Agriculture
                                                       13%
                                                  Pasture/Range
                                                      Lands
Figure 3. Total Nitrogen Loadings.

    6.   Septic tanks,  which  serve about 20% of the watershed's population, are another important source of nitrogen  and
        pathogen loadings, especially in  some areas such as Allen's Creek and tributaries to McKay  Bay.

    7.   Another 7% of the bay's total nitrogen  loadings had been attributed to losses of fertilizer during ship loading and
        en route to port. However, this  figure has declined substantially  since 1991  as source  control BMPs were
        implemented at the port.
                                                        61

-------
    8.   More than 60% of the bay's annual loadings of chromium, zinc, mercury and lead,  as well as significant amounts
        of petroleum hydrocarbons and pesticides are conveyed by stormwater.

    9.   Atmospheric deposition also is a major source of toxic substances accounting for 44% of the bay's total cadmium
        loading and about 17%  of the copper and lead loadings. PAHs also enter the bay from the atmosphere.

    10.  Industrial  and domestic point sources also  contribute  about 30% of the  bay's total loadings of aresenic, cadmium,
        chromium, and copper.

    Puzzle Piece 5:  The Watershed Management Plan: The  state's Surface Water Improvement and Management
(SWIM) program  was established by the legislature in  1987. Tampa Bay was named in the SWIM Act as a priority
waterbody within the Southwest Florida Water Management District and a SWIM Plan was adopted in  1992. In 1990,
Tampa Bay was adopted into the National Estuary Program  by EPA leading to the development of Charting the Course,
a Comprehensive Conservation  and Management Plan (CCMP) for the bay. Community participation was an essential
component of the development of both watershed plans. In particular, many of the agencies, citizen groups, and others
long active in the  restoration and management of Tampa Bay participated in the development of the CCMP. The CCMP
built on  many of the region's ongoing environmental programs, from land acquisition to urban stormwater retrofitting to
habitat restoration. It also identified where unneccesary duplication existed  in current environmental  programs and
provided recommendations to ensure that limited  public funds are spent in the most environmentally effective manner.

    Puzzle Piece 6:Setting Priorities:!'o assure that limited  financial resources were used judiciously, restoration projects
and  programs were prioritized and targeted to specific subbasins. This was done by combining GIS analysis with
watershed modeling and characterization with the results  of the  sediment, biological,  habitat,  and  chemical assessments.
For example, an essential early  activity of the SWIM program was a watershed-wide assessment of pollution sources,
especially stormwater,  to identify "hot spots" -  subbasins with high  stormwater loadings - and to prioritize  urban
stormwater retrofitting projects (SWFWMD, 1990). Similarly, priority  habitat restoration sites were also selected as were
projects to reduce overall nitrogen loadings to the bay, including those from  atmospheric deposition.

    fuzz/e Piece 7: Action Hans: A successful watershed management plan must  include a specific set of actions that
will be taken within a specified time. Charting the Courseincludes 41 specific actions that are needed to achieve the plan's
goals. These  include the construction of numerous urban stormwater treatment  and  habitat restoration projects that have
been built, are underway,  or are  planned in priority subbasins (Figure 4). Since vacant land in the highly urbanized area
is scarce or extremely expensive, many of these projects are being conducted  on existing public lands providing multiple
benefits  including regional stormwater management, open space,  and recreation. Public education is a frequent
component of these projects with the placement of signs depicting the  effects of urbanization, the  importance of wetlands
and  riparian vegetation, and the need  for stormwater treatment and habitat restoration.

    Programs and actions that rely upon nonstructural BMPs are  also being  used to reduce "Pointless Personal Pollution"
at  its source and increase the effectiveness of existing programs. For example, since surveys showed that up to 70% of
the stormwater BMPs serving new development are  not being  properly maintained, assuring their long-term operation
and  maintenance can greatly reduce stormwater pollution.  Maintenance and operation of  BMPs typically is the legal
responsibility of private land  owners  and property  owner associations. Unfortunately,  DEP,  SWFWMD,  and local
governments do not have enough staff to conduct  regular inspections. To improve this deficiency,  DEP is implementing,
in  cooperation with local governments and the WMDs, a training and certification program for public and private sector
individuals involved in  erosion,  sediment,  and stormwater inspections. Local governments also  are  encouraged  to
implement Stormwater Operating  Permit systems which require  an annual inspection  and certification that the  stormwater
system has been maintained and is properly operating. As an economic incentive, some local stormwater utilities provide
credits for individuals served  by a properly maintained and operating system. Additionally,  Hillsborough County has
implemented the "Adopt a Pond" program to help educate stormwater system owners on how to maintain their systems.

    Other nonstructural efforts include assisting businesses in developing and implementing pollution prevention plans
and the  continued implementation of the Florida Yards and  Neighborhoods Program. This program is being expanded
to  help  develop model landscaping guidelines for commercial landscapes,  and  promote  the  incorporation  of FYN


                                                     62

-------
                         Figure 4. Active SWIM Habitat Restoration and Stormwater

                                   Retrofitting Projects
Figure 4. Active SWIM Habitat Restoration and Stormwater Retrofitting Projects.

landscaping guidelines into local government site review processes for new development. The region's continuing rapid
growth provides opportunities through local government comprehensive  plans and land  development regulations to
promote compact development and to reduce impervious surfaces, especially parking lots at commercial developments.

    Puzzle Piece 8:  Assuring  Implementation: An  important aspect of any watershed management program is an
institutional framework that assures that all  of the partners will implement their responsibilities. In 1998, an Interlocal
Agreement was signed  by  the Tampa Bay NEP's local  government  and regulatory implementation partners. The
agreement requires the  partners to submit detailed plans describing how they will fulfill their responsibitlities for bay
restoration and protection. Additionally, all of the local governments within the watershed have been issued, either as
individual permittees or co-permittees, NDPES muncipal Stormwater permits. These permits  include specific requirements
that are identified in the  CCMP's action plans.
                                                     63

-------
    Active Habitat Restoration Projects
                               Active Stormwater Retrofitting Projects
1.  Lowry Park
2.  NE McKay Bay
3.  Delaney Creek
4.  Simmons Park
5.  Hendry Fill
6.  Peanut Lake
7.  Bayshore Blvd.
8.  Gandy Park
9. Cabbage Head Bayou
10.  Boca Ciega
11.  Cargil S. parcel
12.  Mangrove  Bay
13.  Cockroach Bay
14.  Little Bayou
15.  MacDillAFB
16.  Picnic Island
1.  Lowry Park
2.  Horizon Park
3.  Old Coachman
4.  S. Pasadena
5.  Jungle Lake
6.  Pinellas Park
7.  N. Redington
8.  EMS  Site
9.  St. Pete/Clearwater Airport
10. Brushy Creek
11. Safety Harbor
12. 102nd Avenue
13. 94th Avenue
14. Lake Carroll
15. Delaney  Creek
16. Haynsworth
17. 141st Avenue
    Puzzle Piece 9: Funding Implementation: Partners in the implementation of the CCMP include EPA, Florida DEP,
SWFWMD, local cities and counties, and the private sector. Each of the partners has made a substantial commitment
of financial resources since 1995 to accomplish the desired aquatic ecological goals. Primary funding sources have
included the P2000 (state and local land acquisition funds), the SWIM Program (state and SWFWMD funds), SWFWMD
Basin Boards, the private sector, and local stormwater utility fees. In many cases, funding  for projects is from a
combination of sources that often allow the leveraging of other funds needing nonfederal matching funds such as those
from the Section 319 nonpoint source implementation grant program.

Costs associated with the  individual actions presented in the Tampa  Bay SWIM  Plan and the NEP  CCMP  are
considerable. However, these should not automatically be construed  as requirements  for new sources of revenues, since
some of these initiatives can and are being accomplished with existing resources or by redirecting  current funding
allocations to better address the bay's needs. A number of actions seek to improve coordination, cooperation, and
planning among  state and local governments, and the private sector.  These may actually result in cost savings for
currently funded  activities.

    A 1994 survey by the Tampa Bay NEP attempted to quantify how much money is spent to manage  and monitor bay
quality and administer environmental programs. Based on FY94-95 budgets, the study indicates that over $260 million
is spent annually by federal, state, and local agencies on the restoration and management of Tampa Bay. As seen in
Figure 5, the largest par-t of the funds (65%  or $170 million) are spent on wastewater collection, treatment, and reuse.
                                      Figure 5. Funds Spent
                                                             e.r%
                                                          Habitat Restoration
                                                           4 Msnag«m*nt

                                                                3.7*
                                                             Land Acquisition
                                                                S.2%
                                                               RsguMlon t
                                                               Enforctmsnt
                                                              0.5%
                                                           Public Avnrann»
                                                              O.D%
                                                            MnHnlMntlon
Figure 5. Funds Spent.
                                                    64

-------
    Approximately $35 million (13%) is spent by local governments and the SWFWMD on stormwater management.
Habitat restoration and land acquisition, two long favored and implemented environmental programs in the Tampa Bay
region, account for over $27 million in  expenditures.

    Puzzle Piece 10: Monitoring Implementation: Since the adoption of Charting the Course in 1996, numerous projects
and  programs have been implemented. The Tampa Bay  NEP  recently issued the  first official  progress  report on
implementation (TBNEP, 1999).  It shows that the program's partners are on or ahead of schedule in achieving most of
the priority goals for bay improvement. Highlights of implementation activities that occurred between 1995-99 include:

    .  Goal:  Recover an additional 12,350 acres of seagrassover 1992 levels, while preserving the bay's existing 25,600
            acres.

      Status:   Since 1988, seagrass acreage is increasing  at about 500 acres per year meaning the goal will be
               reached in 25 years.

    •  Goal:  Restoring and  protecting  bay  habitats

      Status:   A total of 250 acres of low-salinity habitat will  be restored in all bay segments, exceeding the five year
               target by 150 acres. Additionally,  a total of 1,340 acres  of mangrove and saltmarsh habitat have been
               restored. All 28 priority sites identified in the habitat master plan have been given the highest priority for
               acquisition under the  state's land-buying  programs.

    .  Goal:  "Hold the  line" at nitrogen loadings estimated from 1992-94.

      Status:   When fully implemented, the 105 projects constructed, underway, or planned will reduce the amount of
               nitrogen entering the bay by an average of 134 tons per year,  exceeding the target by  60%.

    In addition to progress on the above  goals, very good progress has been made on other goals. These include:

    .  The goal of protecting the endangered manatee population in Tampa Bay. The Manatee Awareness Coalition has
      implemented "Manatee Watch" where trained volunteers help educate and  encourage boaters to go slow in  waters
      frequented by manatees.

    .  The goal of  returning bay scallop  to Tampa  Bay. Stocking  programs  are adding bay scallops and citizen volunteers
      are measuring the effectiveness  of these efforts through the Great Bay Scallop Search.

    •  The goal of reducing atmospheric deposition into and onto Tampa Bay. To better understand the linkage between
      air pollution and water quality, eight research and monitoring  programs  addressing atmospheric deposition are
      underway.

    . The goal of making Tampa Bay safe for shellfish harvesting  and swimming. As part of the "Healthy Beaches
      Project", research is underway to identify and test better indicators of microbial contamination, the prevalence of
      the indicators at bay and gulf beaches, and probable sources  of the contamination.

    .  The goal of  providing flows  necessary to support plant and animal communities below the dam  on the Hillsborough
      River.  In February 1999, the SWFWMD approved a draft minimum flow rule for the Hillsborough River that hopefully
      will provide a basis for resolving  conflicts over competing uses of the river.

    .  The goal of developing a long term dredge material management plan for Tampa Bay. A Dredged Material Advisory
      Committee is being organized in partnership with the  U. S. Army Corps  of Engineers.

Discussion and  Recommendations

    Florida has established a wide variety of laws, regulations and programs  at the state, regional and local level to
protect,  manage and  restore the state's incredibly valuable  yet vulnerable  natural  resources,  especially its water


                                                     65

-------
resources. There is no doubt that these programs have been effective in helping to reduce adverse impacts on natural
resources resulting from the state's rapid and continuing growth over the past twenty years. However, even with the
implementation of these programs, many of Florida's natural resources have been severely strained or degraded. Some
of these adverse effects can be attributed to activities that occurred before the implementation of modern watershed
management programs such as the channelization of the Kissimmee River and the creation of the vast drainage canal
network south of Lake Okeechobee both of which are  contributing to the decline of Lake Okeechobee, the Everglades
and Florida Bay. Other adverse impacts, though, are directly related to the state's rapid growth and development during
the last twenty years.  These include water supply problems, water quality problems, declining habitat and impacts on
endangered species such as the manatee and the Florida panther.

    Why are these adverse impacts still occurring given the wide range of watershed management programs that have
been implemented in Florida? What  could be done to reduce these effects and  possibly restore already degraded areas?
The continuing evolution of Florida's land and water management programs into a more holistic approach which seek to
manage cumulative effects can help to overcome many  of the current program deficiencies. Cooperative efforts and
partnerships, together with citizen  education  and involvement to improve the stewardship  ethic of all Floridians is
essential. With increased  support and  participation by all Floridians, the effectiveness of the  state's programs can be
improved helping to assure that our natural resources will be able to be enjoyed by future generations.

References

Barbour, M.T., J. Gerritsen,  and J.S. White. 1996. Development of a Stream Condition Index (SCI) for Florida. Prepared
for Stormwater/NPS Management Section, FDEP, Tallahassee,  Florida.

Barbour, M.T.,  J.B. Stribling, and  J.R.  Karr,  (in press). The Multimetric Approach  for Establishing Biocriteria  and
Measuring Biological Condition. |n  W. Davis,  T. Simon  (eds.), Biological Assessment and Criteria: Tools for Water
Resource Planning and Decision Making. Lewis Publishers.

Barbour, M.T.,  J.B. Stribling, and J.R. Karr,  (in review). Biological Criteria:  Technical Guidance for Streams. U.S.
Environmental Protection Agency, Office of Science and Technology, Health and Ecological  Criteria Division,  Washington,
DC.

EPA.  1989. Rapid bioassessment protocols  for use in streams and rivers: Benthic macroinvertebrates and  fish.
EPA/444/4-89/001. Washington D.C.

Fausch, K.D., J. Lyons, J.R. Karr, P.L. Angermeier. 1990.  Fish communities as  indicators of environmental degradation.
Am. Soc. Symp. 8:123-44.

FDER.  1988. A guide to  the  interpretation of  metal concentrations in estuarine  sediments. Florida  Dept. of Env.
Regulation. Office of Coastal Zone  Management. Tallahassee, Fl.

FDEP.  1994. Florida  coastal sediment  contaminants atlas and technical  volume. Florida  Dept.  of Environmental
Protection. Sediment Research Group. Tallahassee, Fl.

FDEP. 1994. Florida Coastal Sediment Contamination Atlas. Office of Water Policy, Tallahassee, Florida.

FDEP. 1996. Standard Operating Procedures for Biological Assessments. Bureau of Laboratories, Tallahassee, Florida.

Griffith, G, J. M. Omernik,  C. Rohmand, S. Pierson. 1994.  Florida regionalization  project. USEPA Environmental Research
Laboratory. Corvallis, Oregon. Final Report Prepared for  Fla. Dept. Environ. Protection. Tallahassee, Fl.

Karr, J.R. 1991. Biological  integrity: A long-neglected aspect of water resource  management.  Ecological Applications 1:66-
84.

Karr, J. and D.  Dudley. 1981. Ecological perspective on  water quality goals. Environ. Mgmt. 5(1):  55-68.
                                                     66

-------
Karr, J.R., and E.W.  Chu. 1998. Restoring  Life in  Running Waters: Better Biological Monitoring. Island Press,  Washington,
D.C.

Livingston, E.H. 1995. The evolution of Florida's stormwater/watershed management program. In Proceedings of the
National Conference  on Urban  Runoff Management:  Enhancing Urban Watershed Management at the Local,  County,  and
State Levels. EPA 625/R-95/003. Cincinnati, Ohio.

Livingston, E.H., E. McCarron, T. Seal, and G. Sloane. 1995. Use of sediment and biological monitoring. In Stormwater
NPDES Related Monitoring Needs,  Proceedings of an Engineering Foundation Conference. ASCE, New York.

Long, E.R. and Morgan, L.G. 1990.  The potential for biological effects of sediment-sorbed contaminants tested in the
National Status and  Trends Program. NOAA Tech. Memo NOS OMA 52.

Long, E. R., D. MacDonald, and C. Cairncross. 1991.  Status and Trends of Toxicants and the Potential for Their Biological
Effects in  Tampa Bay, Florida. NOAA Technical  Memorandum NOS OMA 58.

Long, E.  R., D.A. Wolfe, R.S. Carr,  K.J. Scott, G.B. Thursby, H.L.  Windom, R. Lee, F.D. Calder,  G.M. Sloane, and T.
Seal. 1994. Magnitude and extent of  sediment toxicity in Tampa Bay, Florida. NOAA Technical Memorandum NOS ORCA
78.

Lyons, J.  1992. Using the index of biotic integrity  (IBI) to  measure environmental  quality in warmwater streams of
Wisconsin. General Technical  Report, NC-I 49. U.S. Department of Agriculture, Forest Service. St. Paul, Mn.

MacDonald, D. 1993. Development of an approach to the  assessment of sediment quality  in Florida coastal waters. Final
report submitted to FDER/SRG. Tallahassee, Fl.

MacDonald, D. 1994. Approach to the Assessment of Sediment Quality in Florida Coastal Waters. Final report submitted
to the Florida Department of Environmental Protection, Tallahassee, Florida.

McCarron, M. E., E.  H. Livingston, and R. Frydenborg. 1997. Using bioassessments to evaluate cumulative effects. In
Effects of Watershed Development and Management on  Aquatic Systems, Proceedings of an Engineering Foundation
Conference. ASCE,  New York.

Pardue, J.,  R. DeLaune,  and W.  Patrick, Jr. 1992. Metal to aluminum correlation in Louisiana coastal wetlands:
Identification of elevated metal concentrations. J. Environmental Quality 21: 539-545.

Schropp, S., F.Calder, L. Burney, and H. Windom. 1989. A practical approach for assessing metals contamination in
coastal sediments - An example in Tampa Bay. In:

Proceedings of the  Sixth Symposium on Coastal and Ocean Management. July 11-1 4,  1989. Published by ASCE.
Charleston, S.C.

Schropp, S., F. Lewis, H. Windom, J. Ryan, F. Calder, and  L. Burney. 1990. Interpretation  of metal concentrations in
estuarine sediments of Florida using aluminum as a reference element. Esturaries 13(3): 227-235.

Southwest Florida Water Management District. 1990. Urban stormwater analysis and  Improvements for the Tampa Bay
watershed. Brooksville, Florida.

Southwest Florida Water Management District. 1992. Tampa Bay Surface Water Improvement and Management Plan.
Brooksville,  Florida.

Tampa Bay National Estuary Program. 1996. Charting the Course for Tampa Bay: Final Comprehensive Conservation
and Management  Plan, St. Petersburg, Florida.

Tampa Bay National  Estuary Program. 1999.  Bay  Guardian: News of the Tampa Bay National Estuary Program.  Summer
1999. St.  Petersburg, Florida.

                                                   67

-------
Yoder, C. 0.1989. The development and use of biological criteria for Ohio surface waters. Pp 139-1 46. In: Water Quality
Standards for the 21 st Century.

Yoder, C. 0. and E. T. Rankin. 1997. Assessing the condition and status of aquatic life designated uses in urban and
suburban watersheds. In Effects of Watershed Development and Management on Aquatic Systems, Proceedings of an
Engineering Foundation Conference. ASCE,  New York.
                                                   68

-------
       Urban Stream Structure And Selection Of Structures To Build Habitat

                               To Support Wild  Fish Populations


                                              Douglas T. Severn
                                        URS  Greiner Woodward-Clyde
                                             Seattle, Washington


Abstract

    This  paper gives a brief overview of the physical  impacts of urbanization on streams and examines the selection of
in-stream methods,  tools, and devices for stabilizing streams  and creating habitat to support native fish species. Although
the paper discusses salmonid  species in the Pacific  Northwest in particular, the methodologies and tools employed to
evaluate  and support fish habitat can be generally applied to streams and watersheds in other regions.

    The effects of urbanization, such as decreased pervious area and vegetative cover and increased stormwater runoff
and  erosion, destabilize watersheds  and  streambeds and  destroy  aquatic  habitats. Stable stream environments are
necessary if biological systems including fish  and  their supporting food web  are to flourish. Changes to urban streams
and  watersheds may be so significant, however, that decades may pass before they reach stability. Even then, the
resulting Astable environment© might not provide the type of habitat needed to support species from the natural
environments.

    The evaluation  of channel erosion and sedimentation in urban streams provides one measure to assess the relative
stability of streams and, thus, their ability to  support fish and  amphibian  species.  For most streams, an  evaluation of
relative streambed stability can be completed through  a visual examination  of streambed morphology  and minimal
supporting calculations. Tools  for performing  these analyses will be presented in this paper.

    Rehabilitation of streams in urban and  heavily logged watersheds requires establishing a stream structure that will
maintain streambed stability and create the different types of habitats needed to support desired fish species. One size
or type of in-stream device cannot  meet all stabilization and habitat requirements. The selection of devices should
correspond to the relative stability of the  individual stream  reach. Devices  for maintaining streambed  stability and creating
habitat, as well as the procedures for selecting  them, will be discussed  in this  paper.

Introduction

    Salmon  populations in  the  Pacific Northwest are  dwindling. One significant cause of this reduction in  population is
the destruction of small stream (1s1 to 4th order) habitats.

    In order to respond to the destruction of fish habitat in small streams, we must have an understanding of watershed
processes and natural stream morphology.  Although this  paper concerns western Washington streams, which are
surrounded by heavy forests and fed by rain and groundwater, the general  principles  for stabilizing streams discussed
here can be applied to most natural small stream systems.

Natural  Stream  Morphology

    Streambed gradients gradually decrease from the upper reaches of  a  watershed to the outlet of a  stream because
flow  rates increase in the lower parts of the watershed (Leopold, Wolman, and Miller, 1964).

   The streambed gradients  create different types of fish habitat features (Rosgen, 1994).  Figure 1  shows the
relationship of habitat type to streambed slope in western Washington streams.  Pool/drop habitat is dominant in reaches
with  smaller  flows and steeper valley gradients. The pools are formed by large organic debris  or rock formations. As the
stream flow  rates increase,  valley gradients tend to decrease and the streams are dominated by  pool/riffle habitats.


                                                     69

-------
                       POOL RIFFLE
                       Stop* Range2

                       TYPICAL SPECIES
                       Chum
                                        PLANEBED
                                        GLIDES
                                        POOL/DP
                                        PCQL/RIFFLE
TRANSITION
Slope Ranga =2-4%

TYPICAL SPECIES
                                        Goto
                                                          POOL DROP
                                                          Slope Range =4-10%
                                                          TYPICAL SPECIES
                               Qfturo (tKtwm ftM Qrfef (Hfe ftafe)
Figure 1. Relationship of Stream Habitat to Streambed Slope in Western Washington Streams.

Pools and riffles form alternately on the outside of stream bends. These alternating pools and riffles are present in
practically all perennial  channels. In straight or meandering  streams, pools and riffles generally form every 5 to 7 channel
widths. As the  stream widths  increase,  however, the number of pools decrease  (Leopold, Wolman, Miller, 1964).

    In each non-rigid, natural stream, a dominant channel is formed by the stream=s dominant discharge. This  dominant
discharge  channel  is a component of most fish  habitats.  The dominant  discharge has  a recurrence period of
approximately once each  1.5 years in natural systems (Simons, Senturk, 1991). In urban  systems, the bank-full discharge
has a recurrence period  of about one year (MacCrae, 1996).

    Western Washington Natural Stream Characteristics The small  streams of western Washington are fed by rain and
groundwater and are found  in steep-sided  canyons with fairly straight valley bottoms.  In  their natural (forested) state,
small streams have  a slightly meandering,  low-flow channel in a narrow valley bottom. The vegetation  along the stream
banks is often dense and provides shade, channel stability, and cover. Debris jams are common and act to slow  stream
flows during storm events. Western  Washington soils are products of glacial activity and consist of smooth cobbles and
stones,  as well  as fine materials.  Clayey bank materials,  heavy root structures  along  the banks, and steady base flows
create channels with nearly vertical  sides and small widths (3 to  6 feet wide, and sometimes  as small as 1 foot wide)
relative to the width of the valley bottom. Typical old-growth forest streams have low nutrient levels and low annual
sediment  yields.
                                                       70

-------
    Most natural Pacific Northwest streams can be described as sediment starved. Natural watersheds are heavily
forested and act as a sponge for rainfall.  In natural watersheds, the storm runoff response is slow and the flow rates are
low. Because the ratio  of the 1.5-year storm to the base flow is quite low (often less than 5), these streams have small
width to depth ratios, with most of the dominant discharge channel  acting as an  aquatic habitat channel. The width of the
dominant  discharge channel is coincident with the width of the aquatic habitat  channel. The aquatic habitat channel  is
the normally wetted, low-flow part of the streambed  (Seattle,  1997).  Most small, natural Western  Washington streams
are dominated  by  pools and drops formed by large  organic debris (Gustav, Severn, Washington,  1993)

    The aquatic biological community in western Washington may include as many as 250 plant and animal species.  The
aquatic biological community depends  on a stable aquatic habitat associated  with old growth, coniferous  watersheds.
Much of this aquatic community functions as a  food web, with fish  populations representing the mega fauna. Some
salmon and trout species  are the  top aquatic predators (Severn,  Washington,  1996).

    Salmon and trout in western Washington evolved to take  advantage of these  regional  stream  conditions. Figure  1
lists different species found in the various regions of the watershed. More athletic fish species like coho, steelhead,  and
cutthroat trout occupy the  upper  reaches (steeper gradient) of the watershed. The young coho  and steelhead reside  in
the stream for a year before migrating  to saltwater. Cutthroat may reside in fresh water for two years before migrating
to saltwater.  Less  athletic species, such as chum and pink salmon, can not migrate through the  steeper gradients of the
upper watershed to spawn.  The fry of these species occur in the lower regions of the watershed and migrate to saltwater
shortly after emergence. Young salmon, as well as young and adult trout, will utilize any part  of the watershed that meets
their habitat requirements. For example, if fish  habitat in the upper reaches of a watershed is  unsuitable, young coho may
seek  winter refuge  in the  lower regions of a  watershed (Severn, Washington,  1996).

    Five general categories of habitat  occur  in  natural western Washington streams (Severn, Washington,  1996):

    .  Estuaries/Deltas

    . Passage

    . Refuge

    . Rearing

    .  Spawning and  Incubation

      Pacific Northwest fish derive most of their  food from organisms (benthos) that live in, or on,  the substrate of the
stream. Most food  production occurs in the same  stream areas that provide spawning and incubation habitat for fish. An
annual surplus of approximately 10 pounds of biomass  is  required to support one  pound of fish in the  stream (Washington,
1999).

Assessing  Stream Deterioration

      Visual inspection  of stream morphology can provide rapid and relatively accurate assessments  of astream=s ability
to support fish populations. Practical  experience working in western Washington streams has shown that visual
streambed assessments correlate well with benthic sampling (Seattle, 1999). Although benthic  sampling is necessary,
visual inspections can reduce the amount  of  benthicsampling required when quickassessments  are needed orextensive
benthic testing  is  cost-prohibitive.

    Natural streams are generally non-rigid.  Their cross-sections vary with changes in flow rates and yearly rainfall
volumes.  Stream systems  will generally aggrade during low flow periods, and degrade during high flow periods. To
assess the deterioration of non-rigid streams, it  is  necessary to understand the following three concepts:

    *Sediment  transport
                                                      71

-------
    *Effects of urbanization on stream stability

    *Effects of urbanization on fish habitat

    Sediment Transport  Concepts Understanding the dynamics  of sediment transport is useful for predicting  hydraulic
equilibrium conditions in  a  stream. Any stream will  respond to imposed changes. Six basic  relationships  exist  between
discharge levels and  channel  form, regardless of stream size (Simons and Senturk, 1991):

    Depth of flow in  the dominant discharge  channel is directly proportional to discharge.

    . Width of the dominant discharge channel is  directly proportional to water discharge and sediment discharge.

    . Dominant discharge  channel  shape is directly related  to  sediment  discharge.

    . Channel gradient is inversely proportional to water discharge and directly proportional  to sediment discharge and
      grain  size.

    . Sinuosity is  directly proportional  to valley gradient and inversely  proportional to sediment discharge (larger valley
      gradient causes greater meander, larger sediment  discharge  causes less  meander).

    . Transport of bed materials is directly related to flow velocity  and concentration  of fine material,  and inversely
      proportional  to the fall diameter  of the bed material  (greater depths and higher velocities  cause larger bed  load
      volume in transport,  sediments  shaped  like kites fall slower than  round-shaped sediments).

    A stable channel exists when a stream has the bed  slope  and cross-section  which allow its channel to transport water
and  sediment from upstream  without  aggradation,  deposition, or streambank  erosion (Simons and Senturk, 1991).

    When natural flow rates  are  exceeded,  sedimentation and erosion can be  a dominant limiting factor for fish
populations. The exaggerated volumes and rates  of stormwater runoff in urban areas increase both the rate of erosion
and volume of sediments generated from  upland  and riparian areas  in the watershed. Soil erosion  can lead to excess
streambed erosion and sedimentation  and destroy  redds, fish rearing  habitats, and food  production areas.

    Effects  of Urbanization on Stream Stability Urbanization permanently alters the  hydrologic balance  within stream in
the following ways:

    .  Total water  passing through urban  streams increases.

    .  Stormwater runoff  rates and volumes increase.

    .  Increased impervious surface areas prevent  groundwater recharge; as a result,  base flow  rates during summer and
      fall  are often less than natural flow  rates were.

    .  Increased stormwater runoff causes  erosion and transports  significant amounts of  sediments and  pollutants,
      including  oil, grease, and polluted fine sediments from streets  and  parking  lots, into urban streams,

    In urban areas, the ratio of the 1  -year  storm to the stream=s base flow  (dominant discharge)  is  large,  sometimes
greater than 100. In  many  watersheds, terrestrial  sediment volumes  are dramatically increased by  urbanization. Excess
sediment  increases the width  of the dominant discharge channel.

Stable urban streambeds have 182% gradients, compared to the 2810%  gradients that support anadromous species in
natural watersheds.  To reach a stable gradient, the streambed can lower several feet, causing significant bed load
sediments from shallow  landslides.   Measurements taken from several western Washington streams  show  that
streambeds will flatten from a 4% gradient to a 1% gradient as a result of urbanization. A change in streambed  gradient
from 4% to 1% over the distance of 1,000 feet can  result in streambed  erosion and an elevation difference of 30 feet at
the upper  end of the  reach. During the transition from steep to flat gradients, fish habitat is in a perpetual state of change
(Severn, Washington, 1996). Unfortunately,  it  can take decades before stability is  again  reached.

                                                       72

-------
In unstable  streams, braiding occurs at crossover  points between  bends where stream gradients are  steep. At normal
stage, a braided section  has a divided flow with  small, mid-channel  bars and a single large channel composed of
subordinate  channels. The  base flow channel often  changes  location within the bottom  of the  dominant discharge channel
(Severn, Washington,  1996).

The erosion of the channel bottom along the basin  (often  called head-cutting)  indicates a  readjustment of the basin's
gradient, the stream discharge, and the sediment  load.  (Simons and SentCirk, 1991).

Effects  of Urbanization  on Habitat Stable  habitat  conditions within the channel have stringent requirements in  urban
streams, including a sediment-starved condition and  minimal movement of spawning-sized  gravel (3-inch and  smaller)
during most storm runoff  events.

In urban streams, the benthic system can be limited both by erosion (which  provides conditions of constant change) and
by sedimentation (which  smothers redds and food  production areas, and fills rearing habitats with  silt). In addition,
streams in  urban or deforested watersheds experience significant habitat loss and  are unable to support the biological
diversity that fish species  depend  upon.  In contrast  to  natural watersheds, where  250 plant and animal species may
comprise the aquatic habitat, urban watersheds may have  fewer than  50  plant and animal species.

    Living systems do not adapt to constantly changing environmental conditions.  The changes in aquatic  habitats
caused  by urbanization decrease food production and destroy spawning and incubation areas (Bell, 1990). As flow rates
and volumes increase, streambeds become unstable. When streambeds become unstable,  the aquatic habitat channel
may retain a small width to depth ratio, but it will  be substantially less than the width of the dominant discharge channel.
In addition,  streambed instability causes a constant shifting of the aquatic habitat channel and this limits development of
the benthic  community and  destroys  redds.

    As an  urban stream  approaches stability, the resulting aquatic habitat channel will  be too wide,  shallow, and
homogeneous to support fish populations. Streams naturally deposit bed  load on the  inside of bends and form point bars.
Because natural sinuosity  is low in western Washington  streams, point bars form infrequently or  incompletely leaving a
wide,  shallow, cross section  during base  flows.   Under these conditions,  the flow depths  of most  urban  streams  are
insufficient to submerge returning adult fish. Because small-grained sediments settle as flow rates decrease, redds and
food production areas are smothered with silt and  pool  habitats are filled with sediment.

    Perhaps the greatest general impact  is the  permanent loss of habitat types that sustain coho and steelhead
populations.  These species prefer pool/drop habitat. Coho, in particular, require quiet pools  (Seattle, 1997). Examination
of Figure 1   shows that as  the streambed  gradient lessons, the  habitat type shifts from pool/drop to pool/riffle. Pool/riffle
habitat does not provide  sufficient pool depth for normal  fish rearing or urban storm refuge. Because large storms
frequently occur after the  fry emerge  from  the streambed gravels, the need for urban storm refuge  habitat is critical.
Juvenile fish cannot maintain their position in high velocity reaches. In fact, normal urban storm flows often wash  juvenile
fish into larger bodies of water  (salt water or  streams) where they  cannot  survive.

Tools  and Methodologies

The methodologies available to  assess existing stream conditions and predict future conditions include the use of visual
streambed assessments and analytical tools  such  as simple hydraulic mathematics.

Streambed  Assessment

General indicators of habitat degradation  in urban  streams  are visually apparent and  include the following  elements:

    .  Dominant  discharge channel wider than in natural conditions

    .  Reduced pool frequency  and less  diverse habitat

    .  Increased  sediment from terrestrial  sources
                                                      73

-------
     . Reduced large woody debris

     . Drastically reduced aquatic community diversity

       Both  natural and urban  streams  fluctuate between stability, degrading,  or aggrading. Compared to  natural streams,
 however,  streams in urban watersheds  exhibit extreme  traits  of aggradation,  degradation, or  instability.  Urban  stream
 deterioration  indicators differ according to the condition of the specific reach (stable, degrading, or aggrading).  In a
 degrading streambed,  there is a  progressive lowering  of the  channel due to scour. In an  aggrading streambed, there is
 a progressive buildup or raising of the channel due to sediment deposition.  Both degradation and aggradation are
 indicators that a change  in the stream=s discharge and sediment load  is taking place  (Simons and  Senturk,  1991).

       Urban stream deterioration  indicators differ  according to the condition  of the specific reach. Table  1  (the following
 bulleted paragraphs) describes  the deterioration  indicators  in  a stable,  degrading,  or aggrading stream.

 Table 1. Urban Stream Deterioration Assessment Indicators


 Stable B If an urban stream reach is stable and terrestrial sediment loads are low, the reach may be able to support species that reside temporarily  in the
 stream before moving to salt water. The wide, shallow channel of a stable stream provides little protection from predation, however, and also lacks resting
 pools. The following are indicators of stable urban streams:
     . Apparent changes in channel shape and configuration afler large storms are small.
     . Width of aquatic habitat channel coincident with width of dominant discharge channel B shallow flow depth, prevents fish passage.
     .  Head-cutting and nick points are  absent or nearly absent.
     . Substrate  stability:
      .  Periphyton stays  on streambed after significant storms B streambed is stable.
      .  Streambed gravel lack periphyton B gravel is being moved during significant runoff and replaced when the storm flow recedes,
         Small-grained sediments settle when storm flows recede, smothering redds and food production areas.
         Pools (on-stream  or off-stream) that can retain newly hatched fry  are generally not present.
 Degrading B In a degrading urban stream the dominant discharge channel is as wide as in a stable stream, but base flow rarely covers  the bottom of the
 channel except at crossover points between bends. When streams begin to  unravel due to degradation, the effects do not appear  instantly. Head-cuts
 move through a stream until it reaches a vertical drop. When enough head-cuts accumulate, the vertical drop will be undercut, releasing large amounts
 of bed load type sediments.  Occasional pools will develop that may support anadromous fish,  however, they are often inaccessible.
     .  Streambed gravel sizes are larger than stable sections of the stream, but are mostly bare of periphyton or other aquatic growth.
     . Channel braiding occurs at crossovers between bends.
     . The substrate of the  base flow channel is not coated with periphyton.
     . The base flow channel and dominant discharge channel lack large woody debris.
     .  Large woody debris  that spans the banks of the dominant discharge channel may indicate a recent streambed elevation and  may illustrate the
      amount of degradation that has occurred.
     . Stream banks are bare and often  nearly vertical.
Aggrading B Aggrading stream reaches may be visually similar to stable reaches. Generally, aggrading streams will have a flatter streambed gradient
and accumulate more fine sediment. Disturbing the bed of an aggrading reach usually results in long periods of murky water flow.  Aggradation will occur
 locally in pools, which reduces habitat value, but has less impact on habitat than an entire aggrading stream reach would.
     .  Head-cuts and nick  points do not exist.
     . Deltas  may be visible at the top of the reach.
     . Periphyton covers the substrate.
     . Large woody debris  in streambed  is partially buried.
     .  Pool/riffle and  pool/drop habitat can occur as isolated conditions.
     . Substrate surface aravel sizes are small.


     The visual indicators described in Table 1  (combined with assessments of fish passage problems and periodic benthic
 population checks) can be used to describe the potential  habitat capacity for a stream reach or,  collectively, for an  entire
 stream.  Needless to say, habitats  for fish species  that reside  in  the stream for one year or longer must  be able to support
the  full life cycle of the species.   In addition, fish  need to have full access  to these habitats.

     A methodology based on visual streambed assessment was developed  for rapid stream assessments  in Longfellow
 Creek  and  Pipers Creek  in  Seattle,  Washington  (it was  also used within  six  watersheds in Snohomish County,
Washington). The streambed assessments provided an accurate  measurement  of the ability of the watershed  or stream
 in question  to  support  fish. An example  of the summary rating for Longfellow Creek is shown  in Table 2.
                                                             74

-------
Table 2. Summary Rating for Longfellow Creek

    Creek                                                  Map                     Bank      Sediments       Other

   Segment                   Description                   Sheet      Habitat'      Erosion'      Sources*     Pollutants*     Total
       1        Pipe from outfall to Andover Street              1,2,3          5           3            3             2          13
       2        Open channel from Andover Street to             4,5          10           1             1              2          14
               Genessee Street
       3        Open channel from Genessee Street to          5,6,7          10           2            2             2          16
               confluence of unnamed tributary in West
               Seattle Golf Course
       4        Unnamed tributary in West Seattle Golf           6,7          6           2            2             212
               Course
       5        Open channel from confluence with West         7,8,9          91             1              213
               Seattle Golf Course  tributary to  Brandon
               Street
       6        Open channel from Brandon Street to Findlay       9            9           2            2             215
               Street. Contains confluence of Juneau  Street
               bypass via "biochannel."
       7        Open channel from Findlay Street to Juneau      9,10          822             2          14
               Street. Also contains piped high flow bypass
               starting at Juneau Street and rejoining Creek
               in Segment 6.
       8        Open channel from Juneau Street to Graham      10,11          11           2            3             2          18
               Street
       9        Open channel from Graham Street to Willow      11,12          8           2            1              2          13
               Street
      10        Open channel from Willow Street to Myrtle        12,13          822             2          14
               Street
      11        Piped channel Webster Basin; open channel      13,14          722             2          13
               to Holden Street. Contains "K-Mart bypass."
      12        Open channel from Holden Street to Thistle     14,15,16        821              2          13
               Street
      13        Pipe from Thistle  Street to head of basin at         16           6           3            1              2          12
               Roxbury  Street
                                                            Notes:

Ranking Codes:
      1 = Poor condition
      2 = Moderate
      3 =   Relatively good
      • B   The value in the "habitat" column is  the result of another ranking process. The total "habitat" volume for each Creek segment is
      transferred into this  table to  complete the ranking  process. The habitat rank has a range of O-18 which is developed from evaluating bed
      erosion, fine sediment accumulation, gravels (clean/stable), benthic  (quantity, quality), habitat structure, and riparian vegetation. All other
      columns in this table are ranked from 1 (poor condition) to 3 (relatively good condition).

           Analytical  Tools  Simple hydraulic  mathematical  tools can be  applied to analyze existing conditions and are
      needed to check passage conditions for  channel rehabilitation  projects. In  addition to depth  and velocity, the amount
      of turbulence in  a stream has significant impact on the amount of sediment  that can be  moved. Greater turbulence
      increases  the amount  and size  of sediments that can  be moved.   Maximum turbulence  occurs  when  the Froude
      number  is equal to 1  .0. The  Froude number is defined by  Equation  1:
                        FN= V/(gY)
                                    1/2
(Equation  1)
Where
    FN =  Froude  number

    V = average  velocity in the cross-section

    g = acceleration force due to gravity

    Y =  the hydraulic depth, which  is the  cross-sectional  area divided  by the  top width
                                                              75

-------
    The Froude number should not exceed 0.8 for large storms (like the 25 or 100-year event) except at channel drops.
This criterion is the same for rigid, grass,  and dirt lined channels in most stormwater manuals.  For frequently occurring
storm flows (1 to 2 year events), the Froude number has to be  much less in order to meet the criteria for larger flows.
Spawning-size gravel will  generally be stable  if this Froude number criterion is met.

    In the  author's experience, backwater  analysis (HEC-2 or HEC-RAS) may not be strictly applicable to analyze natural
streams, but  is a useful tool to  analyze deteriorated  channels and to model  proposed improvements  for stream
rehabilitation. For stream rehabilitation backwater analysis  is  done for large and small events (base flow, 1 -year flow, and
one point  in between) to ensure passage of juveniles throughout their life-cycle  habitat within the stream.  To provide
accurate results, more cross-sections are  needed to conduct backwater analysis on non-rigid,  urban streams compared
to  conventional  backwater  analysis.

Selecting  Rehabilitation  Devices

    Unless excess stormwater from both frequent and  rare storm runoff events can be eliminated, it is the author=s
opinion that some form  of structural intervention is needed to create fish  habitat  in urban streams. In western
Washington, streams become unstable and significant  fish habitat is  lost when the impervious area reaches 10 to 15
percent (Booth, 1996). The  dominant discharge channel will respond to hydrologicchanges (Simons and Senturk, 1991).
As  a  result, modifying  how land  is urbanized can reduce the effects of urbanization, but it will  not obviate the hydraulic
impact  on  the dominant discharge channel.  Stream rehabilitation  measures  will still work best where there are  fewer
disturbances to the  watershed  and where wide riparian corridors are maintained.

    The goal of urban stream rehabilitation is to stabilize the streambed with devices that also create a habitat for fish
populations. The stream must develop sufficient food mass and diversity to support desired fish species. Quality salmon
and trout habitat can exist in urban streams when  hydraulic/habitat criteria are met, the streambed is  stable, and the base
flow channel is  confined  (Severn  and  Washington, 1996).

    Establishing stable streams in urban watersheds is often a  moving target. As  more urbanization occurs,  hydrologic
and biological  changes  accumulate. The extent that the dominant discharge  channel spreads is a direct function of the
amount of  pavement in a watershed.

    The New Urban  Stream. Restoring an  urban stream to pre-development conditions is  not possible (National Research
Council,  1992). It is  the author's opinion that too many  stream rehabilitation projects emphasize stream bank rehabilitation
rather than focusing on the root causes  of stream habitat destruction.  Often, that root cause is streambed instability,
which is a  natural response to  increased  flow rates and  volumes in the stream.

    To  maintain pre-development species in  urban and deforested streams,  a  "new  urban stream" is needed that can
provide a variety of  fish habitats  including pools. Without intervention, the urban streams will  convert pool/drop habitat
into pool/riffle habitat, eliminating the diversity of habitat required to support a variety of fish species (Severn  and
Washington,  1996).

    The goal  is to stabilize the stream and return it  to its original sediment-starved  condition. While bed load and
suspended  sediment are readily available,  the object is to sculpt sediment deposition to form bars and banks and confine
the  aquatic habitat channel  within  a single location. Concentrating flows in the aquatic  habitat channel  helps keep the
substrate size optimal  and clears the stream of fines.

    Urban  stream rehabilitation must focus  on historic conditions that can  be recreated, rather than on the conditions
that cannot be meaningfully restored.  Except  for the  following  three exceptions,  historic environmental conditions can
be  recreated in urban streams:

    .  The  width of the dominant discharge channel will  always be greater in urban  streams.

    .  The  banks of  the aquatic  habitat channel cannot be coincident  with the dominant discharge  channel (in pool/riffle
      habitat).

    .  High  flood flows,  deep flow  depths, and large velocities are more frequent.

                                                       76

-------
    In form,  the resulting channels resemble snowmelt type streams, where flow rates significantly exceed base flows
for several weeks each  year. Most  urban streams lack  large woody debris that can  be incorporated into  the stream's
structure when a reach attains stability. Most reaches will take decades to reach stability. Just because the channel will
not look "natural", doesn't mean that we will have failed or that anadromous species  cannot be supported. Many eastern
Washington snowmelt type  streams support anadromous species.

    A rehabilitated  stream has five  primary needs:

    . A dominant discharge channel sized  to carry the 1 -yearto 1 S-year storm (depending on  the degree of  urbanization)
      at full bank.  It is important to recognize this need because the stream will reshape the dominant discharge channel
      and  may undo much of the rehabilitation effort.

    . Within  the dominant discharge channel, hydraulic conditions must provide biologic and  stream stability (keep most
      of the  spawning sized  gravel  and  rock from  moving during frequent storms).

    . Within  the dominant discharge channel,  habitat must  be provided for the entire life cycle of the desired species.

    . Because fine-grained sediment falls out last and needs to be kept  in transport,  the base flow channel has  to be
      narrow, deep, and stable.

    . Stream banks need to  be stable to support  riparian vegetation. Bio-stabilization  techniques can help reduce the
      width of the  dominant discharge channel.

    Ideally, long reaches of  unstable streams  will be stabilized.  Near the spanning  structures,  aggradation will replace
degradation (as long as  a sediment supply is  available).  If only short reaches are  stabilized, large storms will deposit
substantial sediments within the stabilized  reach, particularly at the upstream end  of the reach. The sediment sizes most
likely to accumulate during the stabilization of a reach  are  the larger sizes  that are moved  as bed  load. Once stable
streambed gradients are attained, the  amount  of sediment that can move and cause aggradation  is finite.  After a few
larger storms, bed  load  movement will be minimal.

    Rehabilitation Devices A variety  of devices may be used to confine the base flow channel and provide streambed
stability. The  author has  experience  with several types of bed  control structures,  glides, lunkers, and confining devices.
In small streams, these devices may  be a well-placed piece of timber, a boulder, or randomly placed stones and rootwads
(to increase  roughness).  In larger streams, stabilization  and confining devices are  much more  complex.

    Selection and siting  of devices is dependent on the stream condition  in a specific area and the type of habitat that
is to  be  provided.  It is important to  realize that one type of structure is not suitable for all  applications. Selection of a
stream rehabilitation device depends on the type of habitat needed  and the device's  perceived hydraulic attributes. Both
non-rigid channel design  and biologic  skills are required to be  successful.

    This section will address  issues related  to the selection of three types of structures: timber stepdown structures,
boulder bed  control structures,  and  deflectors.

    Timber Stepdown stepdown. Sometimes referred to as  a  "K-structure".  A form  of timber is shown in Figure 2. The
logs,  which are set  at 45 degrees to  the channel, are called weir logs and  create pools  during  high storm flows. The weir
logs need to  have a steep pitch (the  bank  end higher than the center of the stream). In  dense,  urban western Washington
watersheds, storms with  high  yearly  return frequencies produce flows 20  to  50 times  the base flow. Timber stepdown
structures form large, quiet pools during  storms, allowing newly emerged fish juveniles to find  refuge.  In
                                                       77

-------
Figure 2. Timber Stepdown K-Structure (shown with weir logs).
addition, the  K-type stepdown creates rollers. Rollers assist fish  passage upstream for less athletic adults and juveniles
during higher flow rates (Kerr Wood Leidal Associates, LTD,  1980).

    Common variations of the K-structure include a straight timber stepdown (no weir logs) and a vortex structure (with
weir logs). In creating a diverse habitat environment,  both types can be used.

    While the straight timberstepdown does not form an upstream pool, the substrate above the stepdown is turned-over,
even  for frequently occurring  storm  flows. The Washington  Department of Fisheries and  Wildlife has successfully installed
many straight timber stepdowns in  logged watersheds and those with  minimum to moderate urbanization. For lower flow
rates, the substrate  above the stepdown  remains stable  and provides excellent spawning  habitat.

    The vortex type timber stepdown does  not form storm refuge  pools as well as the k-structure and the log configuration
stymies formation of a roller.  The structure can help develop confined low flows, however, and creates good fish-rearing
habitat for many northwest species (not Coho).

    Boulder Bed Control Structure. Figure  3  shows a  boulder bed control structure.  Like the K-structure,  the  boulders
that form  the  structure are on the  upstream side, but these boulders could  also  be on the downstream side to form  a
vortex-type  structure.  In the author's experience, changing the  configuration  and the angles of the boulders  provides
slightly different habitat characters, all of which are acceptable. While wood is usually preferred in small streams, boulder
bed control structures are flexible and can adjust to channel degradation in  unstable streams. Boulder bed control
structures are most beneficial at the downstream  end of a  reach  where no streambed control is established in the reach
below.

Deflectors. Deflectors can be made of either wood or boulders. Figure  4 shows a timber deflector with a bank log on the
opposite bank. Point bars  form on  the insides of bends. Deflectors should be installed on the insides  of bends to help
build larger point bars and to  confine the base flow channel. The reach  in Figure 4 is nearly straight (sinuosity about 1  .0).
To develop point bars in straight reaches, several deflectors  may be needed and could  be on one side of the  channel,
or on alternating sides. The reach shown in Figure 4 has a large  bed load, and  point bar formation would have occurred
if the deflector installations were correctly located and implemented. Without modification, however, the deflector shown
in  Figure 4 will not form a point bar.
                                                        78

-------
Figure 3. Boulder Bed Control Structure.

                                             3-^
Figure 4. Deflector.

Conclusions

    Streambed assessment based on sediment transport principles can be  a  useful tool to  rapidly determine a stream's
capability to support fish populations.  Standard engineering analytical tools  can be used with streambed  assessment to
support hydraulic design for stream rehabilitation  projects. Although the stream will not have the same appearance as
a natural stream, stream rehabilitation can be successful  and urban streams can  support  anadromous fish populations
in western Washington. Hydraulic design is needed to develop the dominant discharge  channel and properly place
structures to attain the desired  habitat conditions. Selection of the rehabilitation devices must consider design  needs of
non-rigid channels, as well as habitat  requirements. Finally, one size or type of habitat rehabilitation device cannot serve
                                                       79

-------
all purposes.  Selection of habitat rehabilitation devices should achieve specific habitat requirements and provide habitat
diversity.

References

Bell, Milo  C., Fisheries Handbook of Engineering Requirements and Biological Criteria,  US Army Corps of Engineers,
Portland, Oregon, January, 1990.

Booth, Derek B.; Montgomery, David R.; Bethel, John; "Large Woody Debris in Urban Streams of the Pacific Northwest;"
Proceedings of an Engineering  Foundation Conference, Edited  by Larry Roesner, Snowbird, Utah, August 1996.

Gustav, Richard S.;  Severn,  Douglas T.; Washington, Percy M.; "If You Build It They Will Come: A Maintenance
Approach to Restoring Fish Habitat in  Urban Streams Heavily  Influenced by Stormwater Flows;"  Proceedings of the Water
Environment Federation  66th Annual  Conference & Exposition,  Anaheim California,  October  1993.

Kerr Wood Leidal Associates LTD., D.B. Lister & Associates LTD, Stream  Enhancement Guide, Province of British
Columbia,  Vancouver, British  Columbia, 1980.

Leopold, Luna  B., Wolman, Gordon M., Miller, John P., Fluvial Processes  in Geomorphology, W.  H. Freeman and
Company,  San  Francisco and  London,  1964.

MacCrae,  C. R., "Experience From Morphological Research on Canadian Streams: Is Control of the Two-Year Frequency
Runoff Event the Best Basis for Stream Channel Protection?" Proceedings of an Engineering Foundation  Conference,
Edited by  Larry Roesner, Snowbird, Utah, August  1996.

National Research Council, Restoration of Aquatic  Ecosystems, Washington, D.C.,  1992.

Rosgen, David L., "A  Classification of Natural Rivers,"Catena, ElsiverScience, B.V., Amsterdam, The Netherlands, 1994.

Seattle, City of, Longfellow Creek Habitat Restoration Master Plan, Seattle,  Washington, 1999.

Seattle, City of,  Pipers Creek Rehabilitation,  Erosion and Sedimentation Management Program and Design Manual,  1997.

Simons, Daryl B. and Senturk, Fuat, Sediment Transport Technology, Water and Sediment Dynamics, Water Resources
Publications,  Littleton, Colorado, 1991.

Severn, Douglas T.  and Washington, Percy M., "Effects of Urban Growth on  Stream Habitat," Proceedings  of an
Engineering  Foundation  Conference,  Edited  by Larry Roesner,  Snowbird, Utah, August 1996.

Washington, Percy M.,  Conservation, Seattle, Washington, August 1, 1999.
                                                     80

-------
            Lessons  Learned  About Successfully  Using Infiltration  Practices

                                                 Eric H. Livingston
                                        Stormwater/NPS Management Section
                                   Florida Department of Environmental Protection
                                          2600 Blair Stone Road (MS3570)
                                          Tallahassee, Florida 32399-2400


Abstract

    Infiltration  practices are one  of the most  valuable urban  stormwater BMPs  because they not only help to  reduce
stormwater pollutants but,  more importantly, help to reduce  stormwater volume.   Unfortunately, infiltration practices  have
gotten a bad reputation over the past 20 years because of their potential to fail.  This paper will review the successes and
failures of the  use of infiltration  practices in the United States.   It will summarize  the lessons that have been learned about
the use of  infiltration  practices. This will  include recommendations on when  they  should  be  used and  how,   and
recommendations on when they  should not be used.  Finally,  the paper will discuss what can be done to reduce stormwater
volume when infiltration  practices cannot  be used successfully.

Introduction

    To achieve the desired objectives  of flood control, water  quality  protection,  erosion control, improved aesthetics,  and
recreation,  a stormwater management system must be an integral part  of the site planning  for every  development site.
Although the basic principles of stormwater management remain the same, each individual site and each specific  project
presents  unique challenges, obstacles, and opportunities. The many variations in climate, soils,  geology,  groundwater,
topography, vegetation,  and planned land use require site-specific design.

    The natural attributes  of each site will strongly influence the type and configuration of  the stormwater system.  These
features will suggest which particular combination  of Best Management Practices (BMPs)  can  be successfully integrated
into an effective system.   Whenever site  conditions allow,  the stormwater management system  should  be  designed to
achieve maximum on-site  storage of runoff by incorporating  infiltration  practices throughout the remaining natural  and
landscaped  areas of a site. A  stormwater management  system should be viewed  as a  "treatment train" in which  the
individual  BMPs are the cars. Generally, the more BMPs that are incorporated  into the  system, the better the  performance
of the treatment train.  Inclusion of infiltrative practices as one  of the  cars should be a primary  goal of stormwater  system
designers, even on sites where detention practices will be the primary BMP.

Infiltration  practices (or retention  practices)  retain stormwater on-site, allowing it to  soak into the ground or evaporate. There
are a  number  of different infiltration practices that have been  widely  used throughout the  United States, including  basins,
trenches,  dry wells,  pervious pavements, and swales.  Often infiltration practices include vegetation with a wide variety of
trees  and shrubs.  In 1987,  Prince  George's County,  Maryland, began evolving  this type  of  infiltration  practice  into
"bioretention,"  which is a   BMP that uses the  soil  matrix  and  vegetation to  sequester  pollutants within the terrestrial
environment (PGC, 1993, 1997)

    Infiltration  practices are one of the  few BMPs that can help to assure that all four stormwater  characteristics (volume,
rate, timing,  and pollutant load)  after development closely approximate the conditions  which occurred before development.
That is because infiltration  practices help  to maintain pre-development  site perviousness  and vegetative cover,  thereby
reducing stormwater volume and discharge rate, which further promotes infiltration and filtering of the runoff.

    The benefits  of infiltration  include  reducing stormwater volume and peak runoff  rate;  recharging groundwater, which
helps  to replenish wetlands, creeks, rivers, lakes, and estuaries; augmenting base flow in  streams, especially during low
flow times; settling and  filtering of pollutants as they move through the system's vegetation  and  surficial  soils;  lowering the
probability of downstream flooding, stream erosion, and sedimentation; and providing water for other beneficial uses.

                                                       81

-------
    Another benefit of infiltration practices is their ability to serve multiple uses since they are temporary storage basins.
Recreational areas (e.g., ballfields, tennis courts, volleyball courts), greenbelt areas, neighborhood parks, and even parking
facilities provide  excellent settings for the temporary storage of stormwater.   Such areas  are  not  usually in  use during
periods of precipitation and the ponding of stormwater for short durations does not seriously impede their primary functions.

Longevity of Infiltration Systems

    One  of the problems with infiltration BMPs that has  been consistently identified, either quantitatively or qualitatively,
is their high rate of failure.  Maryland's Stormwater Program  produced one of the most comprehensive quantitative reviews
of the longevity of infiltration  systems (Pensyl  and Clement, 1987; Lindsey et  al,  1992).  This information is summarized
in Table 1, where it can be seen that the overall condition and functioning of  infiltration systems declined over time.  In
1986, about two-thirds of all  surveyed facilities were functioning as designed, while in 1990, only about half were.  Only 42%
of the facilities were functioning as designed in both 1986 and  1990,  while about 27% were not functioning as designed in
both years.  About 24% of the systems were functioning in  1986, but not in  1990; while only 7% of those not  working in
1986 were working in 1990.  Maintenance was needed at more facilities  in 1990 (66%) than in 1986 (45%).  Additionally,
38% of facilities that needed maintenance in 1986,  still needed maintenance  in 1990, while 32% of the  facilities that did
not  need maintenance in 1986, did need it in 1990.  Only 10% of the systems that needed maintenance in 1986 did not need
maintenance  in 1990.   These data suggest that  little effort is  expended on maintaining the operational capabilities of
stormwater management systems.

    Additional quantitative information on the success and failure of infiltration systems was collected in the Puget Sound,
Washington, area (Klochak,  1992;  Gaus, 1993; Hilding,   1993; Jacobson and Horner,  1993).   Of 23 infiltration basins
evaluated, 12 did not comply with the region's guidelines for either infiltration rate or time for the basin  to recover its storage
volume.   Interestingly, the authors found no relationship between  lack of  basin maintenance and  failure, with examples of
basins with and  without maintenance that did  not  function properly.  Some basins were functioning properly even though
they had never been maintained, while 43% of the 23 basins had been scarified to enhance performance.

    The above data,  when combined with  qualitative information from  Florida and Delaware  (Baldwin, 1999, personal
communication), seem to indicate that infiltration  basin failures are associated with:

    1.  Inaccurate estimation of infiltration rates

    2.  Inaccurate estimation of the seasonal high water table

    3.  Excessive compaction during the construction process

    4.  Excessive  sediment  loadings either from improper  erosion and sediment control during  the construction  process
       or a lack of pretreatment BMPs

    5.  Lack of maintenance

Factors Influencing Successful Use of Infiltration Systems

    Factors that influence the successful use  of any stormwater  BMP can be categorized as institutional, technical,  and
implementational.  This section of the paper will examine the  components of each of these categories that  must be included
in a stormwater program if the causes of infiltration system failure are to be minimized.

Institutional Components

    The "BMP Golden Rule" states that stormwater  BMPs should never be required unless the stormwater  program includes
components that  will  assure that the BMPs  are  correctly  designed,  approved,  constructed, inspected, maintained,
                                                       82

-------
Table 1. Results of Maryland Infiltration Practices Surveys
Basins

# facilities
(% of total)
1986
63
30%
1990
48
27%
Trenches
1986
94
45%
1990
88
30%
Dry Wells
1986
30
14%
1990
25
14%
Perv. Paving
1986
14
7%
1990
13
7%
Veg. Swale
1986
6
3%
1990
3
2%
Facility Evaluations
Functioning
OM Needed
30
48%
41
65%
18
38%
39
81%
75
80%
28
30%
47
53%
64
73%
23
77%
9
30%
18
72%
4
16%
7
50%
10
71%
2
15%
8
62%
3
50%
6
100%
2
67%
3
67%
Performance and Maintenance Criteria
Buffer strip
inadequate
Stabilization
needed
Sediment
entry
Inappropriate
ponding
No observed
well
20 4 65
32% 8A 69%
12 23 11
19% 48% 12%
24 28 32
38% 58% 34%
41 25 25
65% 52% 27%
na na 45
45%
35
39%
13
15%
58
66%
20
23%
58
56%
24 0 14 3
80% 100% 23%
13 11
3% 12% 7% 8%
02 99
8% 64% 64%
93 74
30% 12% 50% 31%
47 10 11
13% 28% 71% 85%
1 0
17%
3 1
50% 33%
4 1
67% 33%
4 0
67%
na na

and  operated (Livingston, 1997). Specifically,  the program must have  stormwater treatment  and management  goals,
performance standards, education, and an institutional framework that includes plan approval, inspections during and after
construction,  legal  operation  and  maintenance  entity requirements,  effective  compliance  mechanisms, and dedicated
funding mechanisms.

                                                         83

-------
    Program  goals: Experience has shown that stormwater  programs need  to have  multiple  objectives  that are important
to the general public in  order to gain the support of the community.  Typically, these will include flood protection,  erosion
and sediment control during construction, water quality  and habitat protection, and  open space and recreation.  Infiltration
systems can help achieve all of these goals.

    Performance standards: Nearly all stormwater treatment programs in the United States are  technology-based, relying
upon  a  performance standard (minimum level  of treatment) and design criteria for various  BMPs that assure that they
achieve the desired treatment goal. A  review of 32 stormwater  programs around the country showed that the most common
performance  standard is removing at  least 80% of the average annual loading  of total suspended solids (WMI,  1997a).
Some programs require higher levels  of  treatment for  stormwater discharges to  sensitive waters,  such  as Florida's
requirement that discharges to Outstanding Florida Waters remove at  least 95% of the  average  annual pollutant  load.
Technology-based  performance  standards such as these provide water quality goals  for nonpoint sources that create equity
with the  minimum treatment requirements for domestic wastewater point sources (Livingston, 1988).

    Institutional framework:  The stormwater program must have a strong institutional  framework that assures that all BMPs
are (1) properly designed, (2) reviewed and approved,  (3) inspected during and after construction, and (4)  operated and
maintained. The components of this institutional framework are set forth in Figure 1. One of the most important components
especially for infiltration  practices, is  a  feedback  mechanism  among system  inspectors,  plan  reviewers,  and designers
about what is working and what is not.  This information can then be used to revise the design criteria for infiltration BMPs
and improve their potential for long-term success.

Technical Components:

    Successful implementation of  any  BMP depends on a thorough  understanding  of the factors  that determine the BMP's
treatment effectiveness, a strong scientific basis for the BMP's  design  criteria, and  an understanding of the site conditions
that are  required or that limit the utility  of a specific BMP.  Infiltration practices are also commonly called retention practices
because they retain  the  runoff  on-site and are designed to infiltrate a design volume  (treatment volume) of stormwater.
Factors  that influence the treatment effectiveness and feasibility of infiltration practices include (1) precipitation patterns,
(2) whether the system  is designed as an on-line or off-line system, (3) whether pretreatment via the  BMP treatment train
is provided,  and (4) site characterisitcs such  as land use, soil type,  geology,  water table elevation,  topography,  and
vegetation.

    Infiltration areas, especially off-line ones, can be incorporated easily into landscaping  or  open space  areas of a site.
These can include  natural or excavated grassed or landscaped depressions and recreational areas.  Parking lots, with their
landscape islands,  offer an  excellent opportunity for the use of this concept since  even the infiltration  of a quarter inch  of
runoff will greatly  reduce sediments,  metals,  oils and greases.  Placing storm sewer inlets within recessed  parking lot
landscape areas, raising the  inlet  a few inches above the bottom,  and using curb cuts to allow runoff to enter this area
represents a  highly effective treatment train. If site conditions prevent the exclusive  use of infiltration, then off-line retention
areas should be used as pretreatment practices in a stormwater treatment train.  This  is especially true if detention lakes
are the primary component  of the stormwater system  and the lakes are intended to serve  as a  focal  point of the
development.

Pollutant removal efficiency factors:  Average annual pollutant removal efficiency is  calculated  considering the  annual mass
of pollutants available for discharge and the annual mass removed. The primary removal mechanism for infiltration practices
is the volume  of  stormwater that is  infiltrated,  since this eliminates  the  discharge  of  stormwater and its  associated
pollutants. In addition, the system's vegetation and the surficial soils play an important  role in  binding and transforming
pollutants as  the water infiltrates.   As with any type of stormwater management practice,  actual field efficiency will  depend
on a large number of factors. For infiltration practices, such factors include:

    1. Long-term precipitation characteristics;  such as  mean number  of storms per  year,  their  intensity and volume,  and
       average inter-event time.
                                                        84

-------
         Fig yre 1, Storm wife r Program Institutional Framework Components
 Inspections
  Erosion/sediment controls
  Storm water system construction
  Storm water system operation

 Plan review and approval
  Site plans
  Erosion control plans
  Storm water plans
  Feedback evaluation process
          Storm water system operation/maintenance
              Public facilities
              Private facilities
              Adopt a pond program

                             Education programs
                                 General public
                                 Elected officials
                                 School curriculum
                                 Designers
Structural BMPs
  Design criteria
  Re s ea rch/perfo rma n ce
  Proper construction
  Proper operation
  Maintenance

Performance standards
  Peak discharge rate
  Volume
  Treatment

Nonstmctural BMPs
  Site planning
  Source controls
  Land acquisition
  Street sweeping

Local Tand use plan
  Administration
     Lead agency?
     Separate agency?
  Staffing
     Engineers
     Inspectors
     Planners
     Scientists
     Maintenance
     Clerical
Program Evaluation
  Citizen surveys
  Bldg, community surveys
  BMP monitoring
  Water body monitoring
   Developers
   Builders
   Practitioners
   Inspectors

Comp F ia nce/enfo rcerrta nt
   Stop-wo/k orders
   Fines
   Civil or criminal
Storm water Retrofitting
   Watershed goal
   Targ eting/pri o ri tizati on
   Capital improvements
   Regional BMPs
  Storm water master planning   Watershed planning
           Integration wfth other federal, state, regional, and local programs
    Adopt prog ram lawsfregulations       Adopt starmwater utility ordinance/fees
                                                               Govt. Roles
                                                              Responsibilities
                                      85

-------
    2.  The  occurrence of first  flush,  which  is  related to the amount  of  directly connected impervious  area;  type of
       stormwater conveyance system;  and the pollutant of interest

    3.  Whether the system is an "on-line" or an "off-line" design

    4.  Cumulatively, the above  three  factors  determine the minimum treatment volume and  maximum  storage recovery
       time

    The  U. S. Weather Bureau  has  measured weather statistics at many  locations around  the  country.   Long-term
precipitation records, including  such information  as  day and duration of  event, intensity, volume,  etc.,  are  available from
either the Federal  government or private vendors. Statistical analysis of these records can develop probability frequencies
for storm characteristics such as the mean storm volume and the mean inter-event period between storms.

    "First flush" describes the washing  action that stormwater has on accumulated pollutants in the watershed. In the early
stages of runoff, the land surfaces, especially impervious  ones like  streets and parking areas, are flushed clean by the
stormwater.  This  flushing  creates a shock loading  of  pollutants.   However, the occurrence and  prevalence of first flush
depends largely on precipitation patterns, the  degree of imperviousness of the contributing drainage area, the size  of the
drainage area, and the type of stormwater conveyance  system.  Florida studies have determined that for highly impervious
urban land  uses with  drainage areas  under 100 acres, there  is a first flush for many pollutants,  especially particulates
(Yousef et al.,  1985;  Miller,  1985). In areas  such as  Oregon  and Washington, however, where rainfall consists of low
intensity, long-duration "events,"first flush is not very prevalent.

    On-line stormwater practices store runoff temporarily before  most of the volume  is discharged to surface waters. These
systems capture all of  the runoff from  a design storm.   This mixes  all  stormwater within the system, thereby masking first
flush and reducing pollutant removal.   They primarily provide flood  control  benefits, with water quality benefits usually
secondary—although on-line wet detention systems provide both.

    Off-line practices are  designed to divert the more  polluted first flush stormwater  for water quality treatment,  isolating
it from the  remaining  stormwater  that is  managed for  flood control.   In  infiltration  systems, the  diverted first flush  is not
discharged to  surface  waters, but  is stored until it is  gradually  removed by  infiltration,  evaporation, and evapotranspiration.
Vegetation,  such as grass in the bottom  and sides of infiltration areas, helps to trap stormwater pollutants and reduce the
potential for transfer of these pollutants to groundwater.  Off-line retention practices are the most effective for water quality
enhancement of stormwater.

    Since an  off-line retention area primarily  provides for stormwater treatment, it  must be combined with other BMPs for
flood protection to  form a  comprehensive stormwater management system.   Figure  2  is a schematic of  an off-line system,
commonly  referred to  as a  "dual  pond system."  In these systems, a  smart  weir directs the first flush stormwater into the
infiltration area until it is filled, with the remaining runoff being routed to the detention facility for flood control.

    A more recent investigation of the  influence of long-term rainfall characteristics  on the  efficiency of  retention practices
included inter-event dry periods, leading to the development of diversion volume curves for inter-event dry  periods of varying
length (Wanielista  et al., 1991 a).   Figure 3 shows an example diversion volume curve  for the Orlando area.  It is important
to note  that first flush  is not considered in these curves.  If a  first flush effect does exist, the  design curves would be
conservative in that the percent treatment efficiency of the infiltration system would  increase. Furthermore, these curves
are based on  precipitation-interevent-frequency (PIF) curves that also include consideration of the probability that a storm
greater than the design storm will  occur. The  PIF analysis looked at exceedance probabilities storms with a return period
of 2, 3,  4, or 6 months, representing a chance that the  storm will exceed the design volume 6, 4, 3, or  2 times per year.
Other useful products  from this  project  are  Rate-Efficiency-Volume charts  and curves that  can  be used  to design wet
detention  ponds that reuse runoff and  thereby help to balance  pre- and post-development volume (Livingston et  al., 1999).
                                                         86

-------
                                        GH-Une Tiwafmenf System
                            DfflfSiSM'l Ct-vil t
                            "Smart in*'
                                  K
                                                  Dntot!:••, Pjri
                                                                     "Smart" Box Bchemntle
                                             OutlBt 5VrjOfUK3
                                              or PIB-PK*.
Figure 2. Schematic of an Off-line Stormwater System.
                                        C.S     1
                                                     Dvcia'on Volunc
                                          4  faur/E  24ta.rLE
Figure 3.  Diversion Volume Curve for Orlando, Florida.
                                                             87

-------
 Site Characteristics:  The suitability of a site for using infiltration practices will depend on a careful evaluation of the site's
natural  attributes.   Proposed  infiltration areas should  be evaluated  for feasibility on  any particular site  or project by
examining the following:

    SOILS -  Must  be suitable for infiltration.  Nationally, most states  recommend that soil textures should not have more
than 30% clay  content or 40% silt content.  Most importantly,  they need to  be able to percolate the diverted volume to
infiltrate within 72 hours, or within 24-36 hours for infiltration areas that are planted with grasses.  Therefore, soils that have
been classified by the NRCS as HSG A are recommended for infiltration practices,  although they can be successfully used
with HSG B soil types.

    INFILTRATION  RATES -  In  recent  years, the  minimum  permeability rate recommended for  infiltration  practices has
been  raised  by implementing agencies.   Shaver (1986) recommended a  minimum rate  of 0.25 inches  per hour,  but
Maryland's regulations now recommend 0.52 inches  per hour.  One of the  most difficult aspects of  designing infiltration
practices is obtaining reliable  information about the actual infiltration rate of the soil where the practice will be  constructed.
Unfortunately, such information is not easily obtainable.  Avellaneda (1985) conducted 20 hydrologic studies of vegetated
swales constructed  on sandy soils with  a water table at least one foot below the bottom during dry conditions. Infiltration
rates were measured using laboratory permeability tests, double  ring infiltrometers, and field  mass balance experiments.
The field mass balance method measured a minimum infiltration rate of 2-3 inches/hour. This measured rate was much less
than lab permeabilities, rates measured by double ring infiltrometer tests (5-20 in/hr), or rates published in the Detailed Soil
Survey.

    The following should be considered for determining the infiltration rate for retention practices:

    1.  Since the infiltration  rate is  the  key to designing any  retention practices,  conservative estimates should be  used,
       and safety factors incorporated into the design to ensure that the design volume will  actually be percolated  into the
       soil and not discharged downstream.

    2.  It is  important that on-site  infiltration measurements be taken at the locations  where  retention  practices will be
       located.  More importantly,  since soil characteristics and  infiltration  rate change with  depth,  it is crucial that the
       measurements be made at the depth of the design  elevation of the bottom of the retention practice.

    3.  Infiltration rates should be  determined by mass  balance field tests if  possible.   They  provide the most realistic,
       accurate estimate of the percolation rate.  If field tests are  not  possible,  infiltrometer tests should be  used, with lab
       permeability tests a third  option.  In either of these latter two tests, the design infiltration rate should be half of the
       lowest measured rate.  As a last  resort,  information from Detailed  Soil  Surveys can  be used  to  estimated  the
       infiltration rate.  However, the lowest rate should be used-as should a safety factor of two.

    A recent assessment of infiltration practices in Carroll County, Maryland, quantified  the infiltration rates for six basins
and six  trenches of differing ages (Nelson et al., 1999).  They found that 64% of the systems  had an average infiltration
rate  below the  state's minimum  recommended  rate.  However,  70% of the  practices were still recovering their  storage
volume within the required 72  hours. Interestingly, for some facilities (mainly trenches), the infiltration rate met or exceeded
the minimum  state rate for a large percentage of the volume of water infiltrated, while the remaining water persisted for much
longer periods of time before infiltrating.  This may indicate that (1) the infiltration rate is related to the hydraulic head, where
the higher depth of the stormwater in the BMP creates  a higher pressure pushing water into the  ground, or (2) the bottoms
of the systems accumulate fines  that impede percolation, while the sides of the systems are still infiltrating runoff  rapidly.

    WATER TABLE - The seasonal high water table should be at least three feet beneath the bottom of the infiltration area
to assure that stormwater  pollutants are removed  by the vegetation, soil, and  microbes before contacting  the groundwater.
Jacobson  and Horner (1993)  recommend a  minimum of five feet if the seasonal high water table cannot be estimated
accurately.  When  considering the groundwater elevation, it is important to remember that the retention area can cause a
mounding effect on the water table, thereby raising it above  the predevelopment level.   The Southwest  Florida Water
Management District (SWFWMD) has developed a model  that can be used to  more accurately determine the  seasonal high
water table and the effects of mounding (SWFWMD, 1998).


                                                        88

-------
    GEOLOGY -  Bedrock should be at least three feet beneath the bottom of the infiltration  area.  In those parts  of the
country where limestone  is at or near the land surface, special precautions must be taken when using  infiltration practices.
The potential for groundwater contamination in such areas is quite high, especially in  "Karst Sensitive Areas" (KSA)  where
sinkhole  formation is  common.   In  KSAs, solution pipe sinkholes  may form  in the bottom  of  infiltration areas creating a
direct conduit for stormwater pollutants to enter the groundwater. Solution pipes often open in the bottom of retention areas
because  the natural soil plug  capping the solution pipe  is thinned by partial excavation to create the retention area  and
because the stormwater creates a hydraulic pressure which can wash out the plug.

    In KSAs,  a site-specific hydrogeologic  investigation  should be  undertaken that  includes  geologic borings wherever
infiltration areas are proposed and mapping limerock outcroppings and sinkholes  on site.   Infiltration systems in  KSAs
should (1) include several small  off-site areas, (2)  use swale conveyances for pretreatment,  (3) be as shallow as possible,
(4) be vegetated with a permanent cover such as sodded grasses, and (5) have flat bottoms to keep the  stormwater spread
out across the entire area.

    TOPOGRAPHY -  Infiltration practices  should not be  located on  areas with slopes over 20%  to minimize the chance
of downstream  water seepage from the subgrade.  Sloping sites often  require extensive cut and fill operations.  Infiltration
practices should  not be  sited  on fill material, since fill  areas are very susceptible to slope failure,  especially when the
interface of the fill/natural soil becomes saturated.

    VEGETATION -  To reduce the potential for  stormwater  pollutants to enter groundwater,  and to help maintain the soil's
capacity to absorb water, infiltration practices should  be vegetated with appropriate native vegetation, especially grasses.
However, this type of vegetation  cannot tolerate long-term inundation, so the retention area must be capable of infiltrating
all of  its runoff  within  a relatively short time period  (i.e., 24 to 36 hours).  The design of "bioretention" systems incorporates
soils  and vegetation that are proficient in trapping stormwater pollutants within them and  takes advantage  of microbial
processes that help transform and trap pollutants in the terrestrial environment.

    SET  BACKS - Infiltration areas should  be located at least 100 feet from any water supply well and at least 12 feet  down-
gradient  from any building  foundations.  Additionally,  they should be  set  back at  least 50 feet from on-site wastewater
systems,  especially drain fields.

    LAND USE RESTRICTIONS  - Certain infiltration practices can only be applied to  particular land uses.  For example,
some sites are so small  or intensively developed that space is insufficient for surficial practices (e.g.,  retention basin), but
they may allow for infiltration or exfiltration trenches if pretreatment can be provided. A concern with any infiltration practice
is the potential for hazardous or toxic wastes to enter the system and migrate into the groundwater.  Land uses where such
substances  are used should implement  comprehensive  pollution prevention, spill  containment, and emergency response
plans  that will prevent dangerous materials from getting into the infiltration system.

    POTENTIAL  FOR  GROUNDWATER  POLLUTION -  A  possible concern  about infiltration practices  is whether they
simply are transferring the stormwater pollution problem from surface waters to groundwaters.   Stormwater pollutants,
especially heavy  metals, quickly bind to  soil particles and vegetation is effective in filtering pollutants, thereby minimizing
the  risk of groundwater contamination (Harper, 1985; Yousef et al, 1985b).  However, groundwater beneath  swales,  and
retention  areas located in highly sandy soils with low organic content, did show elevated levels of heavy metals down to
depths of 20 feet (Harper,  1988).

Design Criteria

    Once all of the above factors have been quantified using state, regional, or local data as appropriate, specific design
criteria can be  established.  Table  2 summarizes the design criteria for  infiltration systems set forth in  Florida's stormwater
regulations. St. John's River Water Management District (SJRWMD, 1992).
                                                        89

-------
Table 2.  Florida's Design Criteria for Infiltration Practices

BMP                           80% Treatment Effectiveness Diversion Volume
Swales                         Infiltrate 80% of the runoff from a 3-yr, 1-hr storm (2.5 inches)
Retention - Off-line                Infiltrate the larger of 0.5 inches of runoff or 1.25" X % impervious
Retention - On-line                Infiltrate an additional 0.5 inches of runoff
DESIGN FACTOR	CRITERIA	
Soil type                         HSG A or B with < 30% clay or < 40% silt/clay
Treatment volume recovery time      72 hours, 24 to 36 hours if grassed
Water table or bedrock             At least 3 feet beneath bottom after mounding
Topography                     On slopes < 20%, not on fill soils
Vegetation                       Recommended to reduce potential for groundwater pollution and to maintain soil permeability
Land use	May not be appropriate at sites where hazardous materials spills may occur	
    Swales:  Traditionally, swales are used primarily for stormwater conveyance and, as such, are considered  an on-line
practice.  The  removal of stormwater pollutants by swales can occur  by  infiltration or vegetative filtration.  Investigations
in Florida (Yousef et al., 1985a;  Harper, 1985)  have concluded that swale treatment efficiency  largely depends on the
volume of stormwater that can be infiltrated through the filtering vegetation and into the soil.

    Avellaneda (1985) developed the following equation  for a triangular-shaped  swale to  estimate  the length of swale
necessary to infiltrate the design stormwater treatment volume:
                                                  5/8 c 3/16
                                                  n3/8i
where:
  L = swale length (m)                                    n = Mannings roughness coefficient
  Q = average runoff flow rate (m3/S)                       i = infiltration rate (cm/hr)
  S = longitudinal slope (m/m)
  K = constant which is a function of side slope that varies from 4,722 to 10,516

    For most residential,  commercial, and  highway  projects,  the length of swales necessary to percolate the stormwater
needed to achieve the 80% performance standard  was found to  be  excessive,  or  at  least twice the distance available.
Thus, some  type  of  swale block (berm) or  on-line detention/retention  may  be  more  helpful.   Swales  make excellent
pretreatment  practices by providing for the infiltration of some stormwater and for some vegetative filtration.  By using a
raised storm sewer  inlet, swales can  provide water  quality enhancement  via   retention  and  still serve as effective
conveyances for flood protection.  Swales can incorporate retention by using  swale blocks,  small  check dams,  or elevated
driveway  culverts  to  create  storage; thereby  reducing  runoff  velocity, reducing erosion,  and promoting  infiltration.   In
highway designs for  high speed situations, safety must be considered; thus,  a maximum depth of water equal to  1.5  feet
and flow line slopes  on the berms of 1  vertical/20 horizontal  are recommended.  Along lower speed highways or in some
residential/commercial urban settings, steeperflow line berm slopes (1 on 6) are acceptable (Wanielista et.al., 1986).

    Unlike Florida, investigations in  Washington  State  (Horner, 1988; WPCD, 1992)  indicate that swales can  also act as
a biofilter, with  removal of particulate pollutants without infiltration of stormwater. The following recommendations were made
to improve their water quality benefits:

    1. Maximum design velocity should not exceed 27 cm per second.
                                                          90

-------
    2.  A hydraulic residence time of at least 9 minutes is  recommended for removal of about 80% of the total suspended
       solids.  Longer residence times will  provide higher removal effectiveness.

    3.  Swale width should be limited  to 6 to 8 feet, unless special measures are provided to assure a level swale  bottom,
       uniform flow spreading, and management of flows to prevent formation of low-flow channels.

    4.  Swale slopes should be between 2 and 4%.

    5.  Water depth should be limited  to no greater than one half the height of the  grass, up to  a maximum of 3 inches of
       water depth.

    6.  Swale length will  be  a function of the hydraulic residence time, swale width,  and stormwater volume and velocity.

Implementation Components

    Even  if effective design  criteria have been  established for an infiltration system,  the  design has been reviewed  and
approved, and the institutional framework to assure performance has been  established, an infiltration system  may still not
work correctly.  In fact, assessments of the success or failure of infiltration systems have determined that poor construction
is a major factor in system failure (Pensyl and Clement, 1987;  Lindsey et al., 1991).  We will discuss five considerations
that are essential to proper implementation of infiltration practices including (1) education, (2) erosion and sediment control,
(3) construction, (4) inspections, and (5) maintenance.

    Education:  The stormwater program needs to  include an extensive education program that targets BMP  designers,
plan reviewers, inspectors, contractors, and  maintenance  personnel.  Each of these  practitioners  is an important  part of
the  stormwater team.  They  must each understand their role in BMP design and implementation, as well as the technical
factors  discussed  above.   Additionally,  a  communication mechanism  needs  to be  established  among  all  of these
practitioners so that in-the-field  knowledge of what works, and what does  not work, is transferred back to all other team
members.  With respect to BMP installation, the  plan reviewers and inspectors should  meet with  the project engineers  and
contractors on-site  to review the site plan, construction sequencing, erosion  and sediment control plan  and details, and the
infiltration system's detailed standards and specifications.

    Erosion and sediment control:  Infiltration  practices must  be  protected from sediment loadings,  especially  during the
project's construction  phases.  Infiltration practices should never  be used as  part of the erosion and sediment  control
system, nor should they be put  into operation  until all  contributing drainage areas are fully stabilized. Although sediment
loads drop sharply  after  construction  is completed, gradual clogging of  infiltration practices can  still  occur.  Pre-treatment
practices such  as  swale conveyances or vegetated  buffer strips can help to filter out sediments and  extend  the life of
retention practices.  Do not forget the treatment train concept.

    Construction considerations: To prevent clogging  of infiltration areas, special  precautions  must be taken  during the
entire construction  phase of  a project to prevent  reduction  of the system's infiltration  capacity. In  particular, two areas need
to be stressed, including  preventing sedimentation during construction and  preventing compaction  of the soil.  Areas  that
are  selected for infiltration use should  be well marked during site  surveying and protected during construction.  Specifical
construction recommendations are as follows (WMI, 1997b):

    1.   If possible, schedule construction  so it does not occur during the  rainy season but does occur during the  vegetation
        growing season.   For example, in Auckland, New Zealand, construction sites are shut down during the winter when
        long, prolonged rains make erosion and sediment controls ineffective and when vegetation does not

        grow well.  Unfortunately, in  the United States, these seasons  often overlap and the economics  of  development
        dictate the time frame for starting construction.

    2.   Before the development site is graded, areas  planned for use as infiltration  systems should be well marked during
        site surveying, and all traffic and heavy  equipment kept away from the area to prevent compacting the  underlying
        soils.

                                                        91

-------
3.   Construction  should be overseen by someone who is  trained and  experienced in the installation of infiltration
    practices, and who is knowledgeable about their purpose and operation.

4.   The design team  should inspect the exposed soil after excavation  to confirm that soil conditions are as expected
    and are suitable for the approved design.  If they are not, work should not proceed and the situation should be
    analyzed to determine whether or not design or construction changes should be made to the approved design.

5.   Construction  of the infiltration system should not begin until after the site has been completely stabilized.  If this
    is not possible, then:

    a.  Diversion  berms should be placed around the perimeter of the infiltration area during all phases of construction
       to divert runoff and sediment away from it.

    b.  Sediment  and  erosion control plans for the site should be oriented to keep sediment and runoff completely away
       from the area.

    c.  The facility should not be excavated to final grade until after the contributing drainage area is stablilized.  Leave
       at least two feet of native soil during the initial excavation.

6.   Infiltration areas  should never   be used  as  a temporary sediment  basins  during  the  construction  phase.
    Unfortunately, it is  common  for infiltration  areas,  especially basins,  to  be  used  as a  sediment  trap,  with  initial
    excavation  to within two feet of the final design elevation of the  basin  floor.   If the facility  is to be used during
    construction,  this  soil  can be  removed in  layers  as it  clogs.   Once construction  is  completed,  sediment that
    accumulated  during the construction phase can then be removed when the basin undergoes  final  excavation to its
    design elevation.   However, recent  experience indicates that even  with this type of construction practice, infiltration
    areas used as sediment traps have a higher rate of failure.

7.   Infiltration areas/basins should  be excavated using light earth-moving  equipment with tracks or  over-sized tires.
    Normal rubber tires should be avoided since they compact the  subsoil and reduce its infiltration  capabilities.  For
    the same reason, the use of bulldozers or front end loaders  should be avoided.

8.   During construction, place excavated material at least 10-15 feet away from the infiltration area.

9.   Since some compaction of the underlying soils is still likely to occur during excavation, the floor of the basin should
    be deeply tilled with a rotary tiller or disc harrow at the end of the excavation process.

10.  Rock  used in infiltration or exfiltration trenches should be washed clean of sediments.   Rock should be placed in
    lifts and compacted after each lift.

11.  Trenches should be clear of any protruding objects and carefully inspected before installing geotextile fabrics.  The
    fabrics should have the proper permeability and be installed with at least a 12-inch overlap, in a shingle-like manner

12.  Trenches should  be covered  and not put into operation until the contributing drainage area is completely stabilized
    and all pretreatment BMPs installed.

13.  Pervious  asphalt  and concrete should be installed only by certified personnel  who are specifically trained in their
    batching, pouring, and finishing.

14.  The basin should be stabilized with vegetation within a week after construction.  Use of low maintenance, rapidly
    germinating native  grasses are  recommended.  The condition  of  the newly established vegetation  should be
    checked  several  times over  the  first two  months,  and  any necessary remedial actions  taken  (e.g.,  reseeding,
    fertilization, and irrigation).
                                                    92

-------
    Inspections:  Like all stormwater treatment practices, infiltration systems need to be inspected during construction and
on a regular basis after construction.  Inspections during construction are needed to assure that the infiltration system  is
built in accordance with  the approved design and standards and specifications.  Five inspections are recommended: (1)
pre-construction,  (2) during excavation, (3) during construction of the embankment (if applicable), (4) after final  excavation,
and (5) after construction is completed.  During this final inspection,  the  inspector should have a copy of the  "As-built or
record drawings."  In addition,  infiltration systems  should be inspected semi-annually after construction (before  and after
wet seasons) to  ensure that they continue to function.   Two site trips are recommended:  one during or immediately after
a rainfall,  so  that conditions during operation can  be observed, and  a second from 24  to 72 hours  after the rainfall, to
determine whether the system is recovering  its storage volume as designed.  Inspection  forms are highly recommended.
Examples  of  inspection  forms  (WMI,   1997b)   can  be   downloaded  from   the  EPA  web  site  located  at
http://www.epa.gov/owow/NPS/orderform.html.

    Maintenance:  All  infiltration practices will require  regular and  non-routine maintenance to  maintain their  ability to
infiltrate stormwater.   The frequency  and  need for maintenance will depend primarily  on the loading of  particulates and
whether pre-treatment practices have been employed.  Routine maintenance includes revegetating eroding areas,  removing
materials that  accumulate in pretreatment BMPs, and removing materials from  inlets and outlets. Non-routine, restorative
maintenance  activities  should  be conducted whenever inspections reveal that stormwater remains in the system beyond
the designed  time.  These may include structural repairs to the inlets  or outlets and restoration of the infiltration  capability
of the system.

Additional Concerns and Recommendations

    Concerns  with Pervious Pavement:  Local land development codes typically  specify the type of material for  a parking
lot (i.e., paved,  grass,  gravel)   and  determine the  number and size of  parking  spaces within  a  parking lot.  These
requirements  should  be reviewed carefully to  ensure  that they  are necessary (is  paving really required in every case) and
that the number  of spaces is related to actual traffic demands.  After these requirements have been reviewed and verified,
the use of pervious  pavement within  a parking lot should be examined.  Pervious pavement materials  include pervious
asphalt, pervious concrete, turf blocks, and even Geoweb covered with sod.

    Overall, experiences with  pervious pavements  have  not been very good.  Pervious  pavements  have  been  prone to
clogging.   Causes include  poor erosion and sediment control  during  construction,  unstabilized drainage  areas  after
construction, improper  mixing  and  finishing of the pavement, and  poor maintenance.   However, field  investigations of
pervious concrete that has been in use for up  to 15  years in Florida indicate that these parking lots can continue to infiltrate
rainfall and runoff if they were installed and maintained properly (FCMA, undated). Pervious concrete not only helps reduce
site imperviousness, but also reduces hydroplaning and road noise.

    Recommendations: Specific recommendations and  other important information  about infiltration systems that will help
increase their  successful implementation are  summarized in Table 3.  This table includes essential information about the
advantages, disadvantages, maintenance,  and  other  aspects of successfully  using  infiltration  practice. To   improve
evaluation of  site conditions for the  suitability of infiltration practices, Jacobson and Horner (1993) recommended a  quan
titative rating  system.   The factors  used in the system included: (1)  soil  till layer (presence and location), (2) location of
seasonal high water  table, (3)  removal efficiency of the pretreatment  BMPs, (4) degree  of siltation protection, (5)  soil  type,
and (6) infiltration rate.  Different degrees of acceptability are  possible:  (1)  disqualifying (characteristics that eliminate design
or location from  further consideration), (2) passable  (characteristics that allow  consideration but  not  ideal), and  (3)  ideal
(optimum characteristics for design or siting of facility). Table 4 illustrates the  proposed rating system factors.
                                                        93

-------
Table 3.  Additional Information About Infiltration Systems to Enhance Successful Implementation

Infiltration
Bmp Type
Surface Basin
(Typically recessed
areas or, in Mid-Atlantic
States, rock filled)

Infiltration Trench
(Typically a rock filled
trench)
Exfiltration Trench
(Typically a perforated
pipe with a gravel
envelope)
Advantages
•Integrate into land-
scaping, open space,
parking lot islands
•Use for recreation
•Easier inspection and
maintenance
•Can be used where
land or space is limited

•Can be used where
land or space is
limited
Disadvantages
•Land area required
•Potential mosquito
problem if not designed
or maintained properly

•Easily clogged
•Difficult to unclog
•Difficult to monitor
performance
•Easily clogged
•Difficult to unclog
•Difficult to monitor
performance
Maintenance
•Vegetated basins should be
mowed and clippings removed
•Remove sediments when dry
and cracked
•Non-vegetated basins require
annual disking
•Removing sediments that
accumulate in rocks

•Remove materials that enter
pipe
•High pressure wash perforated
pipe
Comments
•Can serve larger drainage
3:1 or flatter side slopes, flat
bottom
•Bottom and side slope vege-
tation recommended
•Pretreatment essential
•Use observation well
•Keep covered until drainage
area stabilized
•Pretreatment essential
•Source controls useful
•Geotextile is limiting infiltra-
tion factor
Pervious Pavement
Swale (Typically
a shallow, grassed
conveyance system)
•Reduces impervious-
ness
•Reduces hydroplaning
and highway noise
•Higher recharge rates


•Can be incorporated
into site 's landscaping/
open areas
•Great car in BMP Train
                                                  •Easily clogged
                                                  •Lack of trained
                                                  practitioners
                                                  •Anaerobic conditions
                                                  may develop in soils
•Not for flood control
•May "disappear" from
residential back yards
•May become depository
fnr traQh  varH waQtAQ	
•Removing sediments that
Accumulate in rocks, replace
rocks

•Regular vacuum street
sweeping
•High pressure cleaning
•Drilling holes to restore
infiltration
•Replacement

•Mow and remove grass
clippings
•Hydroscope accumulated
sediments and resod
•Inf. rate 0.5'Vhr if use sides
and bottom
•Inf. rate 1"/hr if use bottom

•Proper batching and place-
ment is crucial
•Education programs needed
for practitioners
•Post signs to inform users
and keep dir and mud out

•Wet swales (wetland plants)
work great too
•Use swale blocks, raised
driveway culverts to retain
Table 4.  Possible Rating System to Evaluate the Suitability of Infiltration BMPs

                              Disqualifying                           Passable
                                                                                    Ideal
Factor

Soil till layer
   Characteristic

 Impenetrable, thick layer
 near surface
                Characteristic

           Layer present but at >5' depth, or
           easily penetrable
                             Characteristic

                             No till layer present
Seasonal High Water
Table
 Close to surface, within 5'
           At intermediate depth, at least 5' below
           BMP bottom
                             Very deep, well below
                             BMP bottom
Pretreatment
                           None provided
                                   Some,minimum 50% TSS removal
                                                       Pretreatment provides >80%
                                                       TSS removal
Siltation
Protection
 None provided
           Any silt or construction sediment
           removed before final BMP construction
                             Fully protected from silt during
                             construction
Soils
                           Saturated or with >30% clay
                           or >40% silt/clay content
                                   Coarse, highly infiltrative soil that can be
                                   modified to produce proper inf. rate
                                                       Loam or loamy sand
Infiltration Rate
                           <0.57hr
                                                             >2.5"/hr but with very deep water table
                                                             or modified to slower rate
                                                                                                         0.5 to 2.57hr
                                                                  94

-------
References

Avellaneda,  Eduardo.   Hydrologic Design of Swales.  M.S. Thesis,  University of Central  Florida, Orlando,  FL  December
1985.

Florida Concrete Manufacturers Association.  Portland Cement Pervious  Pavement  Manual.  FCPMA, 649  Vassar Street,
Orlando, FL, 32804. No date.

Gaus, J. J.   Soils of Stormwater Infiltration Basins in  the Puget Sound Region:  Trace Metal  Form and Concentration  and
Comparison to Washington State Department of Ecology Guidelines.  M. S. Thesis.  College of Forest Resources, U. of
Washington, Seattle, WA 1993.

Harper, Harvey H.  Effects of Stormwater Management Systems on Ground Water Quality.   Final  Report, Contract WM-190,
Submitted to Florida Dept. of Environmental Regulation, Tallahassee, FL 1988.

Harper, Harvey H.  Fate of Heavy Metals from  Highway Runoff in Stormwater Management Systems.  Ph.D.  Dissertation,
University of Central Florida, Orlando, FL 1985.

Hilding, K.  A Survey of Infiltration  Basins in the Puget Sound Area. Non-thesis  paper, Dept.  of  Biological and Agricultural
Engineering, University of California, Davis, CA 1993.

Horner, R.R.  Biofiltration Systems for  Storm Runoff Water Quality Control.  Final Report Submitted to City  of Seattle,  WA
1988.

Jacobson,  M.A.  and  R.R.  Horner.    Summary  and Conclusions  from Applied  Research  on  Infiltration  Basins  and
Recommendations for Modifying the Washington  Department of Ecology Stormwater Management Manual for the Puget
Sound Basin. Submitted to Urban Nonpoint Management Unit, WDOE, Seattle, WA 1993

Klochak,  J.R.  An Investigation  of the Effectiveness of Infiltration  Systems  in Treating  Urban  Runoff.  M.S.E.  Thesis,
Department of Civil Engineering,  University of Washington, Seattle, WA 1992.

Lindsey, G.,  L. Roberts and W. Page.   Inspection  and Maintenance of  Infiltration Facilities.  J. Soil Water Conserv. 47 (6):
481-186.  1992.

Livingston,  E.H.   State Perspectives  on  Water Quality Criteria.   In:  Design of Urban  Runoff  Controls, an Engineering
Foundation Conference, Potosi, MO 1988.

Livingston, E.H.  Successful Stormwater Management: Selecting and Putting the Puzzle Pieces Together.  In:  Proc. Future
Directions for Australian Soil and Water Management, September 9-12, 1997, Brisbane, Australia 1997.

Livingston,  E.H.   M.P. Wanielista, J.  Bradner.   Stormwater  Reuse: An  Added Benefit of Wet Detention Systems.   In:
Proceedings of the First South  Pacific Conference on Comprehensive  Stormwater and Aquatic  Ecosystem  Management,
Auckland, New Zealand, 1999.

Miller, R.A.  Percentage Entrainment of Constituent Loads in Urban Runoff, South Florida.  USGS WRI 84-4329. 1985.

Nelson, S.,  T.  Devilbiss, and E. Bolton.  Assessment of Stormwater Management Infiltration  Practices  in Carroll County,
MD Final Report. March 15, 1999.

Pensyl, L.K. and P.F.  Clement.   Results of  the State  of Maryland Infiltration Practices Survey.  Presented at the State of
Maryland Sediment and Stormwater Management Conference, Washington College, Chestertown, MD August 1987.

Prince George's County, Maryland.  Bioretention Manual. Dept. of Env. Resources.  1993.


                                                       95

-------
Prince George's County, Maryland. Low-Impact Design Manual.  Dept. of Env. Res. 1997.

St. Johns River Water Management District. Rule 40C-42 and associated Applicant's Handbook. Palatka, FL 1992.

Shaver,  H. Earl.  Infiltration  as a  Stormwater Management Component.  In: Proceedings of the  Urban Runoff Technology,
Engineering Foundation Conference, Henniker, NH  1986,

Southwest Florida Water Management District.  Methods for Identifying Soils and Determining Seasonal High Ground Water
Elevations. Handout Literature for Technical Training Workshops. Brooksville, FL. 1998.

U.S. EPA, "Results of the Nationwide Urban Runoff Program: Volume 2: Final Report", Washington, DC, 1983.

Wanielista, M.P.  Quality Considerations in the Design of Holding Ponds, Stormwater Retention/Detention  Basins Seminar,
University of Central Florida, Orlando, FL 1977.

Wanielista, M.P., Y.A. Yousef, L. M.  Van DeGroaff and S.H. Rehmann-Koo. Best  Management Practices - Enhanced
Erosion  and Sediment Control Using Swale Blocks. Final  Report,  FL-ER-35-87, Submitted to Florida Dept. of Transportation,
Tallahassee,  FL 1986b.

Wanielista, M.P., M.J. Gauthier, and D.L. Evans. Design  and Performance of Exfiltration Systems.  Final Report,  Contract
C3331, Submitted to the Florida Dept. of Transportation, Tallahassee, FL 1991 a.

Wanielista, M.P., Y.A. Yousef, G.M. Harper, T.R. Lineback,  and  L. Dansereau.  Precipitation,  Inter-Event Dry Periods, and
Reuse  Design  Curves for  Selected  Areas  of Florida.   Final  Report, Contract WM-   Submitted  to the Florida  Dept.
Environmental Regulation, Tallahassee, FL 1991b.

Water Pollution  Control Dept.  Biofiltration Swale Performance, Recommendations, and Design Considerations.  Publication
657, Seattle, WA 1992.

Watershed  Management  Institute,  Inc.   Institutional  Aspects of  Urban  Runoff Management: A   Guide  for  Program
Development  and  Implementation.  Cooperative  Agreement with the  Nonpoint  Source  Control  Branch,  U.S.   EPA,
Washington, D.C. 1997a.

Watershed Management Institute, Inc. Operation,  Maintenance,  and Management of Completed Stormwater Management
Systems. Cooperative Agreement with the  Nonpoint Source Control Branch, U.S. EPA, Washington,  D.C.  1997b.

Yousef,  Y.A.,  M.P. Wanielista, H.H.  Harper, D.B. Pearce  and  R.D. Tolbert.  Best  Management Practices -  Removal  of
Highway Contaminants by Roadside Swales.   Final  Report,  Submitted to the Florida Dept. of Transportation, Tallahassee,
FL  1985a.

Yousef,  Y.A.,   M.P  Wanielista,   and  H.  Harper.    Fate  of Pollutants in  Retention/Detention  Ponds.   In:  Stormwater
Management:  An Update.  Publication 85-1, University of Central Florida, Orlando, FL. 1985b.
                                                      96

-------
          Potential  New Tools for the Use of Tracers to Indicate Sources of

                         Contaminants to  Storm  Drainage Systems

                           Robert Pitt, Melinda Lalor, Jennifer Harper,  and Christy  Nix
                              Department of Civil  and Environmental  Engineering
                                   The  University  of Alabama  at Birmingham

                                               Donald Barbe'
                              Department of Civil  and Environmental  Engineering
                                        The University of New Orleans


Abstract

    This paper is a description of previously developed methods used to identify sources of contaminants in storm
drainage systems, plus a review of emerging techniques that may also  be useful. The original methods, along with
selected new  procedures, were tested  using almost 700  stormwater samples collected from  telecommunication manholes
from throughout the U.S. About 10% of the  samples were estimated to be contaminated with  sanitary sewage,  using these
methods, similar to what is expected  for most stormwater systems. The original methods are still recommended as the
most useful procedure for identifying contamination of storm drainage systems,  with the possible  addition of specific tests
for E. co//and enterococci and UV absorbance at 228 nm. Most newly emerging methods require exotic equipment and
unusual expertise  and are  therefore not very available,  especially at low  cost and with fast turn-around times for the
analyses. These emerging methods may therefore be more useful for special research  projects than for routine screening
of storm drainage systems.

Introduction

    Urban stormwater runoff includes  waters from many  other sources which find their way  into storm drainage systems,
besides from  precipitation. There are  cases where pollutant levels in storm drainage  are much higher than they  would
otherwise  be  because of excessive amounts of contaminants  that are introduced into the storm drainage system by
various  non-stormwater  discharges. Additionally, baseflows  (during dry weather) are also  common  in storm drainage
systems. Dry-weather flows and wet-weather flows have been  monitored during numerous urban runoff studies. These
studies have found that  discharges observed at outfalls  during  dry weather were significantly different from wet-weather
discharges and may account for the  majority of the annual discharges for some  pollutants of  concern from the  storm
drainage system.

    There have been numerous methods used  to investigate inappropriate discharges to storm drainage systems. Pitt,
ef a/. (1993) and Lalor (1994) reviewed many of these procedures  and  developed a system that municipalities could use
for screening outfalls in residential and commercial areas. They are currently updating these earlier methods under
funding from the U.S. EPA and the University of New Orleans. In these  areas, sewage contamination, along with low-rate
discharges from small businesses  (especially laundries, vehicle repair shops, plating shops,  etc.) are of primary concern.
One of the earliest methods used to  identify sewage contamination utilized the ratio of fecal coliform to fecal strep.
bacteria. This method is still in use, but unfortunately  has proven inaccurate in most urban  stormwater applications The
following discussion reviews the methodology developed by Pitt,  etal. (1993) and Lalor (1994), and some new
approaches that were investigated.

Use of Tracers to Identify  Sources of Contamination in Urban Drainage Systems

This research investigated inappropriate discharges into storm drainage systems. It was  of most concern to identify toxic
or pathogenic sources of water, typically  raw sewage  or industrial  wastewaters, that were being discharged  accidentally
into the storm drainage system.
                                                    97

-------
    Investigations designed to determine the contribution of urban stormwater runoff to receiving water quality  problems
have  led to a continuing interest in  inappropriate connections to storm drainage systems. Urban stormwater  runoff is
traditionally defined  as that portion of precipitation which drains from city surfaces  and flows  via natural or man-made
drainage systems into receiving waters. In fact, urban stormwater runoff also includes waters from many other sources
which find their way into storm drainage systems. Sources  of some of this water can be  identified and accounted for by
examining  current National  Pollutant Discharge Elimination System (NPDES) permit records for permitted  industrial
wastewaters that can be legally discharged to the storm drainage system. However, most of the water comes from other
sources, including illicit and/or inappropriate entries to the storm drainage system. These entries can account for a
significant amount of the pollutants discharged from storm sewerage systems (Pitt  and  McLean  1986).

    Permits for municipal separate storm  sewers include a requirement  to effectively prohibit problematic non-stormwater
discharges, thereby placing  emphasis on  the  elimination  of inappropriate connections to  urban  storm drains. Section
122.26 (d)(l )(iv)(D) of the rule specifically requires an initial screening program to provide means for detecting high  levels
of pollutants in dry weatherflows, which should serve  as an  indicator of illicit connections to the storm sewers. To facilitate
the application  of this  rule,  the  EPA's Office  of Research and  Development's Storm and Combined Sewer  Pollution
Control Program and the Environmental Engineering & Technology Demonstration Branch, along with the  Office of
Water's Nonpoint Source Branch, supported  research for the investigation  of inappropriate  entries to storm  drainage
systems  (Pitt,  et a/. 1993).  The approach  presented in this research was based on the  identification  and  quantification
of clean  baseflow and the  contaminated  components during dry weather. If the relative  amounts of these components
are known, the importance  of the dry weather discharge can be determined.

    The  ideal  tracer to identify major flow sources should have  the following characteristics:

    .  Significant difference in concentrations  between possible  pollutant sources;

    .  Small variations in concentrations within  each likely  pollutant source category;

    .  A conservative behavior (i.e., no significant concentration  change due to physical, chemical or

      biological processes);   and,

    .  Ease of measurement  with adequate  detection  limits, good  sensitivity and  repeatability.

    In order to identify tracers meeting the above criteria, literature characterizing potential  inappropriate entries into  storm
drainage systems was examined.  Several case studies that identified  procedures  used by individual municipalities or
regional  agencies were also examined.

Parameters  Suitable for  Indicators of Contamination by Sanitary Sewage

    Tracer Characteristics of Local Source Flows. Table  1  is a summary of tracer parameter measurements for
Birmingham, Alabama.  This  table is  a summary of the "library"  that describes the tracer  conditions for each  potential
source category. The important information shown on this  table includes the median and coefficient of variation  (COV)
values for each tracer parameter for  each  source category. Appropriate tracers are  characterized  by having significantly
different concentrations in flow categories that need to be distinguished.  In addition, effective tracers  also need low COV
values within  each flow category.  The study  indicated that  the COV values  were quite low for each category,  with the
exception of chlorine,  which  had  much greater COV values. Chlorine  is therefore  not recommended as a quantitative
tracer to  estimate the flow components. Similar data must  be collected  in each community  where  these procedures are
to be used. Recommended field observations include color, odor, clarity,  presence  of floatables and deposits,  and rate
of flow, in  addition to the selected chemical measurements.

Simple Data Evaluation Methods to  Indicate Sources of Contamination

    Indicators Implying Contamination. Indicators of contamination (negative indicators)  are clearly apparent  visual or
physical  parameters indicating obvious problems and are readily observable at  the outfall during the field screening


                                                       98

-------
Table 1. Tracer Concentrations Found in Birmingham, Alabama, Waters (Mean, Standard Deviation and Coefficient of Variation)

Fluorescence
(% scale)

Potassium
(mg/L)

Ammonia
(mg/L)

Ammonia/Potassium
(ratio)

Fluoride
(mg/L)

Toxicity
(% light decrease after
25 minutes, I26)
Surfactants
(mg/LasMBAS)

Hardness
(mg/L)

PH
(pH units)

Color
(color units)

Chlorine
(mg/L)

Specific conductivity
(pS/cm)

Number of samples
Spring
water
6.8
2.9
0.43
0.73
0.070
0.10
0.009
0.016
1.7
0.011
0.02
2.0
0.031
0.027
0.87
<5
n/a
n/a
<0.5
n/a
n/a
240
7.8
0.03
7.0
0.05
0.01
<1
n/a
n/a
0.003
0.005
1.6
300
12
0.04
10
Treated
potable water
4.6
0.35
0.08
1.6
0.059
0.04
0.028
0.006
0.23
0.018
0.006
0.35
0.97
0.014
0.02
47
20
0.44
<0.5
n/a
n/a
49
1.4
0.03
6.9
0.29
0.04
cl
n/a
n/a
0.88
0.60
0.68
110
1.1
0.01
10
Laundry
wastewater
1020
125
0.12
3.5
0.38
0.11
0.82
0.12
0.14
0.24
0.050
0.21
33
13
0.38
99.9
<1
n/a
27
6.7
0.25
14
8.0
0.57
9.1
0.35
0.04
47
12
0.27
0.40
0.10
0.26
560
120
0.21
10
Sanitary
wastewater
250
50
0.20
6.0
1.4
0.23
10
3.3
0.34
1.7
0.52
0.31
0.77
0.17
0.23
43
26
0.59
1.5
1.2
0.82
140
15
0.11
7.1
0.13
0.02
38
21
0.55
0.014
0.020
1.4
420
55
0.13
36
Septic
tank
effluent
430
100
0.23
20
9.5
0.47
90
40
0.44
5.2
3.7
0.71
0.99
0.33
0.33
99.9
<1
n/a
3.1
4.8
1.5
235
150
0.64
6.8
0.34
0.05
59
25
0.41
0.013
0.013
1.0
430
311
0.72
9
Car wash
water
1200
130
0.11
43
16
0.37
0.24
0.066
0.28
0.006
0.005
0.86
12
2.4
0.20
99.9
<1
n/a
49
5.1
0.11
160
9.2
0.06
6.7
0.22
0.03
220
78
0.35
0.070
0.080
1.1
485
29
0.06
10
Radiator
water
22,000
950
0.04
2800
375
0.13
0.03
0.01
0.3
0.011
0.011
1.0
150
24
0.16
99.9
<1
n/a
15
1.6
0.11
50
1.5
0.03
7.0
0.39
0.06
3000
44
0.02
0.03
0.016
0.52
3300
700
0.22
10
activities. These observations are very important during the field survey because they are the simplest method of
identifying  grossly contaminated dry-weather outfall  flows. The direct examination  of  outfall characteristics for unusual
conditions of flow, odor, color, turbidity,  floatables,  deposits/stains, vegetation conditions, and damage to  drainage
structures is therefore an important part of these investigations. Table 2  presents a summary of these indicators, along
with narratives of the descriptors to be selected in the field.
                                                         99

-------
Table 2. Interpretations of Physical Observation Parameters and Likely Associated Flow Sources
Odor - Most strong odors, especially gasoline, oils, and solvents, are likely associated with high responses on the toxicity screening test. Typical obvious
     odors include: gasoline, oil, sanitary wastewater, industrial chemicals, decomposing organic wastes,  etc.
     sewage: smell associated with stale sanitary wastewater, especially in pools near outfall.
     sulfur ("rotten eggs"): industries that discharge sulfide compounds or organics (meat packers, canneries, dairies, etc.).
          oil and gas: petroleum refineries or many facilities associated with vehicle maintenance or petroleum product storage.
     rancid-sour, food preparation facilities (restaurants, hotels, etc.).

Color - Important indicatorof inappropriate industrial sources. Industrial dry-weather discharges may be of any color, but darkcolors, such as brown, gray,
     or black, are usually of most common.
     yellow: chemical plants, textile and tanning plants.
     brown: meat packers, printing plants, metal works, stone and concrete, fertilizers, and petroleum refining facilities.
     green: chemical plants, textile facilities.
     red: meat packers.
     gray: dairies.

Turbidity - Often affected by the degree of gross contamination. Dry-weather industrial flows with moderate turbidity can be cloudy, while highly turbid
     flows can be opaque.  High turbidity is often a characteristic of undiluted dry-weather industrial  discharges.
     cloudy: sanitary  wastewater, concrete  or stone operations, fertilizer facilities, automotive dealers.
     opaque: food processors, lumber mills, metal operations, pigment plants.

Floatable Matter - A contaminated flow may contain floating solids or liquids directly related to industrial or sanitary wastewater pollution. Floatables of
     industrial origin may include animal fats,  spoiled food, oils,  solvents, sawdust, foams, packing materials, or fuel.
     oil sheen: petroleum refineries or storage facilities and  vehicle service facilities.
     sewage:  sanitary wastewater.

Deposits and Stains- Refer to any type of coating near the outfall and are usually of a dark color. Deposits and stains often will contain fragments of
     floatable substances. These situations are illustrated by the  grayish-black deposits that contain fragments of animal flesh and hair which  often are
     produced by  leather tanneries, or the white crystalline powder which commonly coats outfalls due to  nitrogenous fertilizer wastes.
     sediment construction  site erosion.
     oily petroleum refineries or storage facilities and vehicle service facilities.

Vegetation -Vegetation surrounding an outfall  may  show the  effects of industrial  pollutants. Decaying  organic materials coming  from various food  product
     wastes would cause an increase in plant life, while the dischargeof chemical dyes and inorganicpigmentsfrom textile mills could noticeably decrease
     vegetation. It is important not to confuse the adverse effects of high stormwater flows on vegetation with highly toxic dry-weather intermittent flows.
     excessive growth: food product facilities.
     inhibited growth: high stormwater flows, beverage facilities, printing plants, metal product facilities, drug manufacturing, petroleum facilities, vehicle
         service  facilities and automobile dealers.

Damage to Outfall  Structures - Another  readily visible indication of industrial contamination. Cracking, deterioration,  and spelling of  concrete or peeling
     of surface paint, occurring at an outfall are usually caused by severely contaminated discharges, usually of industrial origin. These contaminants are
     usuallyveryacidicor basic in nature. Primary metal industries have a strong  potential  forcausing outfall structural damage because their batch  dumps
     are highly acidic. Poor construction, hydraulic  scour, and old age may also adversely affect the condition of the outfall structure.
     concrete cracking: industrial flows
     concrete spalling: industrial flows
     peeling paint industrial flows
     metal corrosion: industrial flows
     Correlation tests were conducted to  identify relationships  between  outfalls that  were  known to have  severe
contamination problems and the negative indicators (Lalor  1994).  Pearson  correlation tests indicated that  high turbidity
and obvious  odors  appeared to be the most useful  physical indicators of contamination when contamination was defined
by  toxicity and  the presence of detergents. High turbidity was noted  in 74% of the contaminated source flow samples, This
represented a  26%  false  negative rate (indication of no contamination when contamination actually  exists), if one  relied
on  turbidity alone  as an  indicator of contamination. High  turbidity was noted in only 5%  of the uncontaminated  source flow
samples.  This represents  the rate of false positives (indication of contamination when  none actually  exists) when relying
on  turbidity alone. Noticeable odor was indicated in 67%  of flow samples from contaminated sources, but in none of the
flow samples from uncontaminated  sources.  This  translates  to  37%  false  negatives,  but no  false  positives.  Obvious  odors
identified  included gasoline, oil,  sewage,  industrial chemicals or  detergents,  decomposing  organic wastes,  etc.
                                                                 100

-------
    False negatives are  more of a concern than  a reasonable  number of false positives when working with a screening
methodology. Screening  methodologies are used  to direct further,  more detailed investigations. False positives would be
discarded after further investigation. However, a  false negative during a screening investigation  results in the dismissal
of a problem outfall for at least the near future. Missed contributors to stream contamination may result in unsatisfactory
in-stream results following the application of costly corrective measures elsewhere.

    The method  of  using physical characteristics to  indicate  contamination in outfall  flows does not allow  quantifiable
estimates of the flow components and, if used alone,  will likely result in many incorrect determinations, especially false
negatives. These simple characteristics are  most useful for identifying  gross contamination: only the most  significantly
contaminated outfalls  and drainage areas would  therefore  be recognized using this method.

    Detergents as Indicators of Contamination.  Results from the Mann-Whitney U tests (Lalor 1994)  indicated that
samples from  any of the dry-weather flow sources could be correctly  classified as clean or contaminated  based only on
the measured  value of any one of the following parameters:  detergents,  color, or conductivity. Color and high  conductivity
were present in samples from clean sources as well as contaminated  sources, but their levels of occurrence were
significantly  different between the two groups. If samples from  only one source were expected to  make up outfall flows,
the level of color or conductivity  could be  used to distinguish contaminated outfalls from clean  outfalls. However, since
multi-source flows occur, measured levels  of color or conductivity  could fall  within acceptable levels  because of dilution,
even though a contaminating source was contributing to the flow. Detergents,  on  the other hand,  can be used to
distinguish between clean and contaminated outfalls simply  by  their presence or absence,  using a detection limit of 0.06
mg/L All samples analyzed from  contaminated sources contained detergents  in excess of this amount  (with the exception
of three septage  samples collected from homes  discharging only  toilet  flushing water). No clean source  samples were
found to contain  detergents.  Contaminated sources would  be  detected in mixtures with  uncontaminated waters if they
made up at least 10% of the  mixture.

    Flow Chart for Most Significant Flow Component Identification. A further refinement is the flow chart shown on Figure
1. This  flow chart describes an analysis strategy which may be used to  identify the major component  of dry-weather  flow
samples in  residential  and commercial areas. This method  does not  attempt to distinguish among all potential sources
of dry-weather flows identified earlier, but rather the following four major groups of flow are identified:  (1)  tap waters
(including domestic tap water, irrigation  water and rinse water),  (2)  natural waters (spring water and shallow ground
water),  (3) sanitary wastewaters (sanitary sewage and septic tank discharge), and  (4)  wash waters (commercial laundry
waters,  commercial car wash  waters,  radiator flushing wastes, and plating  bath wastewaters). The  use  of this method
would not only allow outfall flows  to be categorized as  contaminated or uncontaminated, but would allow outfalls carrying
sanitary wastewaters to  be identified.  These outfalls could  then  receive highest priority for further investigation leading
to source control. This flow chart was designed for use in  residential and/or commercial areas only.

In residential and/or commercial  areas, all outfalls should be located and examined. The first  indicator is the  presence
or absence  of dry-weather flow.  If no dry-weather flow exists at an outfall, then  indications of intermittent  flows must be
investigated, Specifically, stains,  deposits,  odors,  unusual  stream-side vegetation  conditions,  and  damage to outfall
structures can all indicate intermittent non-stormwater flows. However, frequent visits to outfalls over long time  periods,
or the use of other monitoring  techniques, may be needed to confirm that only stormwater flows occur. If intermittent flow
is not indicated, then the outfall probably does not have a contaminated non-stormwater source. The  other points on the
flow chart serve to indicate if a  major contaminating source is present, or  if the water is uncontaminated. Component
contributions cannot be quantified using  this method, and only the "most contaminated" type of source present will be
identified.

    If dry-weather flow exists  at  an outfall, the flow should  be sampled and tested for detergents. If detergents are not
present, the flow  is probably from a non-contaminated non-stormwater source. The lower limit of detection for detergent
should  be about 0.06 mg/L.
                                                       101

-------
                            Start
Residential or
Commercial
^
•j
*
Land Use
In Area

fe

Industrial or
Industrial/
Commercial
                                               Sanitary
                                            Wastewater or
                                              Washwater
     Not a
 Contaminated
Non-Stormwatrr
    Source
                                                                   Likely
                                                                 Sanitary
                                                                Wastewater
                                                                   Source
                                                                   Likely
                                                                Wastewater
                                                                   Source
No
fc
ifylng
(Lalor 1994).
Natural
Water
Source
Flow Chart Methodology for Idrntifylng
Most Significant Flow Component (Lalor 1994).

   Figure 1. Simple Flow Chart Method to Identify Significant Contaminating Sources
                                102

-------
    If detergents are not present, fluoride levels can be used to distinguish between flows with treated watersources and
flows with natural sources in communities where water supplies are fluoridated and natural fluoride  levels are low.  In the
absence of detergents, high fluoride levels would indicate a potable water line leak, irrigation water, or wash/rinse  water.
Low fluoride  levels would indicate waters originating from  springs or shallow groundwater. Based on the flow source
samples tested in  this research  (Table 1), fluoride levels above 0.13 mg/L would most likely indicate that a tap water
source was  contributing to the dry-weather flow  in the Birmingham, Alabama, study area.

    If detergents  are  present, the flow is probably from a contaminated non-stormwater source, as indicated on  Table
1. The  ratio  of ammonia to potassium can  be used to indicate whether or not the source is sanitary wastewater.
Ammonia/potassium ratios greater than 0.60 would indicate likelysanitarywastewater contamination. Ammonia/potassium
ratios were above  0.9 for all septage and sewage samples collected in Birmingham (values ranged from 0.97 to 15.37,
averaging 2.55). Ammonia/potassium ratios for all othersamplescontaining  detergents were below 0.7, ranging from 0.00
to 0.65, averaging 0.11.  One radiator waste sample  had an ammonia/potassium ratio of 0.65.

    Non-contaminated samples  collected  in Birmingham had ammonia/potassium ratios ranging from 0.00 to 0.41, with
a  mean value of 0.06 and a median  value of 0.03.  Using the mean  values for non-contaminated samples (0.06) and
sanitary wastewaters (2.55), flows comprised  of mixtures containing at least 25% sanitary wastes with  the remainder of
the flow from uncontaminated sources would likely be identified as  sanitary wastewaters using  this method. Flows
containing smaller  percent contributions from sanitary wastewaters might  be  identified as  having a wash water source,
but would not be  identified  as uncontaminated.

Emerging  Tools for Identifying Sources of Discharges

    Coprostanol and Other Fecal Sterol Compounds Utilized as Tracers of Contamination by Sanitary Sewage. A  more
likely indicator of human  wastes  than fecal coliforms and other "indicator" bacteria may  be the use of certain molecular
markers, specifically the fecal sterols, such as coprostanol and epicoprostanol (Eaganhouse, eta/, 1988). However,  these
compounds are also discharged by other carnivores in a drainage (especially dogs). A number  of research projects have
used these compounds to investigate  the presence of sanitary sewage contamination. The most successful application
may be associated with sediment analyses instead of water analyses.  As an example, water analyses of coprostanol are
difficult due to the typically very low concentrations found, although the concentrations in many sediments are quite high
and much easier to quantify. Unfortunately, the long persistence of these compounds in the environment easily confuses
recent  contamination  with  historical or intermittent  contamination.

    Particulates and sediments collected  from coastal areas in Spain and Cuba receiving  municipal sewage loads were
analyzed by Grimalt, ef a/. (1990) to determine the utility of coprostanol as a  chemical marker  of sewage contamination.
Coprostanol  can not by itself be  attributed to  fecal matter inputs. However, relative contributions of steroid components
can be  a useful indicator. When  the relative  concentrations of coprostanol and coprostanone are  higher than  their 5?
epimers, or more  realistically, other sterol components of background  or natural  occurrence, it can provide useful
information.

    Sediment cores from Santa  Monica Basin, CA, and effluent from two local municipal wastewater  discharges were
analyzed by  Venkatesan and Kaplan (1990) for  coprostanol to determine the degree of sewage addition  to  sediment.
Coprostanols were distributed throughout the basin sediments in association with fine particles. Some stations contained
elevated levels, either  due to their proximity to outfalls or because of preferential advection of  fine-grained sediments. A
noted decline of coprostanols relative to total sterols from outfalls seaward indicated dilution of sewage by biogenic
sterols.

    Other chemical compounds have been utilized for sewage tracer work. Saturated hydrocarbons with 16-1 8 carbons,
and saturated  hydrocarbons with  16-21 carbons, in addition to coprostanol,  were chosen as markers for sewage in  water,
particulate, and sediment samples near the Cocoa, FL, domestic wastewater treatment plant (Holm  and Windsor, 1990).
The concentration of the  markers was highest at points close to the outfall pipe and diminished with distance. However
the concentration of Cl 6-C21 compounds was  high at  a  site  800 m from the outfall indicating that these compounds were
                                                      103

-------
 unsuitable markers for locating areas exposed to the sewage plume. The concentrations for the other markers were very
 low at this station.

    The range of concentrations of coprostanol found in sediments and  mussels of Venice;  Italy, were reported by
 Sherwin, ef al, (1993). Raw sewage is  still discharged directly into the Venice lagoon. Coprostanol concentrations were
 determined in sediment and mussel  samples from the lagoon  using gas chromatography/mass spectroscopy.  Samples
 were  collected in  interior canals  and  compared to open-bay concentrations. Sediment concentrations  ranged from  0.2-41 .0
 ug/g  (dry weight). Interior canal sediment samples averaged 16 ug/g compared  to 2 |jg/g found in open bay sediment
 samples. Total coprostanol concentrations in  mussels ranged from 80 to 620 ng/g (wet weight).  No mussels were found
 in the four most  polluted interior canal  sites.

    Nichols, etal. (1996) also examined coprostanol in stormwater and the sea-surface microlayer to distinguish human
 versus nonhuman sources of contamination. Other  steroid compounds in sewage effluent were investigated by Routledge,
 ef al. (1998) and  Desbrow, ef al. (1998) who both examined estrogenic chemicals. The most common found were 173-
 Estradiol and estrone which were detected at concentrations in the tens of nanograms per liter range.  These were
 identified as  estrogenic through a toxicity identification and evaluation  approach, where sequential separations and
 analyses identified the sample fractions causing estrogenic activity  using a yeast-based  estrogen screen.  GC/MS was
 then  used to identify  the specific compounds.

    Estimating Potential  Sanitary Sewage Discharges into  Storm Drainage and Receiving  Wafers using Detergent Tracer
 Compounds. As  described above, detergent  measurements  (using methylene blue active  substance,  MBAS, test
 methods) were the most successful individual tracer to  indicate contaminated water in storm sewerage  dry-weather flows.
 Unfortunately,  the MBAS  method uses hazardous chloroform  for an extraction  step. Different detergent  components,
 especially linearalkylbenzene sulphonates (LAS) and linear alkylbenzenes (LAB),  have also been tried to indicatesewage
 dispersal patterns in  receiving waters. Boron, a major  historical ingredient of laundry chemicals, can also potentially  be
 used. Boron has the  great advantage of being relatively easy to analyze using portable field test kits, while LAS  requires
 chromatographic equipment. LAS can be measured using HPLC with fluorescent detection, after solid phase extraction,
 to very low levels. Fujita, ef a/. (1998) developed an efficient enzyme-linked immunosorbent assay (ELISA)  for detecting
 LAS at levels from 20 to 500 ng/L.

    LAS from synthetic surfactants (Terzic and Ahel 1993) which degrade rapidly,  as well as nonionic detergents (Zoller,
 ef a/.  1991) which do not degrade rapidly, have been  utilized as sanitary sewage markers. LAS was quickly dispersed
 from  wastewater outfalls except in areas where wind was calm. In these areas LAS  concentrations increased in
 freshwater but were unaffected in saline water.  After time,  the lower alkyl  groups were mostly found, possibly as a result
 of degradation or settling of longer alkyl chain compounds with sediments. Chung, ef  a/ (1995) also describe the
 distribution and fate  of LAS in an urban stream in  Korea. They  examined different LAS compounds having  carbon ratios
 of Cl  2 and Cl 3 compared to Cl 0 and  Cl 1,  plus  ratios of phosphates to MBAS  and the internal to external isomer ratio
 (I/E) as part of their research. Gonalez-Mazo, etal. (1998)  examined LAS  in  the Bay of Cadiz off the southwest of Spain.
 They  found that  LAS  degrades  rapidly  (Fujita,  ef a/., 1998, found that complete biodegradation of LAS requires several
 days), and is also strongly sorbed to particulates. In areas close to shore  and near the untreated wastewater discharges,
there  is significant vertical stratification  of LAS: the top 3 to 5 mm of water had LAS concentrations about 100 times
greater than found at 0.5 m.

    Zeng and Vista  (1997) and  Zeng, ef al, (1997) describe  a study off of San Diego where LAB was measured, along
with polycyclic aromatic hydrocarbons (PAHs) and aliphatic  hydrocarbons (AHs) to indicate the  relative  pollutant
 contributions of wastewater from sanitary sewage, nonpoint  sources,  and hydrocarbon combustion sources. They
 developed and tested several indicator ratios  (alkyl homologue  distributions and  parent  compound  distributions) and
 examined the ratio of various  PAHs (such  as phenanthrene to anthracene, methylphenanthrene to phenanthrene,
fluoranthene to pyrene, and benzo(a)anthracene to chrysene) as tools for distinguishing these sources. They concluded
that LABs are useful  tracers of domestic waste  inputs to the  environment due to their limited sources. They also describe
the use of the internal to external isomer ratio  (I/E) to indicate the amount of biodegradation that may  have occurred to
the LABs. They observed concentrations of total LABs in sewage  effluent  of about 3 ug/L, although previous researchers
 have  seen concentrations of about 150 ug/L  in  sewage effluent from the same area.
                                                      104

-------
    The fluorescent properties of detergents have also been used as a tracer by investigating the fluorescent whitening
agents (FWAs), as described by Poiger, ef a/. (1996) and Kramer, ef a/. (1996). HPLC with fluorescence detection was
used  in these studies to quantify very low concentrations of FWAs. The two most frequently used FWAs in household
detergents  (DSBP and DAS  1) were found at 7 to 21 ug/L in primary sewage effluent and at 3 to 9 ug/L in secondary
effluent. Raw sewage contains about 10 to 20 ug/L FWAs. The removal mechanisms in sewage treatment processes is
by adsorption to activated sludge. The type of FWAs varies from laundry applications to textile finishing and paper
production, making it possible to identify sewage sources. The FWAs were found in river water at 0.04 to 0.6 ug/L. The
FWAs are  not easily  biodegradable  but they are  readily photodegraded. Photodegradation  rates have been  reported to
be about 7% for DSBP and 71% for DAS 1 in river water exposed to  natural sunlight, after one hour exposure.
Subsequent photodegradation  is quite slow.

    Other Compounds Found in Sanitary Sewage that may be used for Identifying Contamination by Sewage. Halling-
Sorensen, eta/. (1998) detected numerous pharmaceutical substances in sewage effluents and  in receiving waters. Their
work addressed human health concerns of these low  level compounds that can enter downstream  drinking water supplies.
However, the information can also be possibly used to help identify sewage contamination. Most of the research has
focused on clofibric acid, a chemical  used in cholesterol lowering drugs. It has been found in concentrations ranging from
10 to 165 ng/L in Berlin drinking water samples. Other drugs commonly  found  include aspirin, caffeine, and ibuprofen.
Current FDA guidance mandates that  the maximum  concentration of a substance or  its active metabolites at the point
of entry into the aquatic environment be less than 1  ug/L (Hun  1998).

    Caffeine has been used as an indicator of sewage contamination by several investigators (Shuman and Strand 1996).
The King County, WA, Water Quality Assessment  Project is  examining the impacts of CSOs on the Duwamish River and
Elliott Bay. They are using  both caffeine (representing dissolved  CSO constituents) and coprostanol (representing
particulate bound CSO constituents), in conjunction with heavy metals  and  conventional analyses, to help determine the
contribution of CSOs to the river. The caffeine is  unique to sewage, while coprostanol is from both humans and
carnivorous animals and is therefore  also in stormwater. They sampled upstream of all CSOs, but with some  stormwater
influences,  100 m  upstream  of the primary CSO discharge (but downstream of other CSOs), within the primary CSO
discharge line, and 100  m downriver of the CSO  discharge location. The  relationship  between caffeine and coprostanol
was fairly consistent for the four sites (coprostanol was about 0.5 to  1.5 ug/L higher than caffeine). Similar patterns were
found among metals; chromium was always the lowest and zinc was the highest.  King Co. is also  using clean  transported
mussels placed in the Duwamish River to measure the bioconcentration potential of metal and organic toxicants and the
effects of the CSOs on mussel growth rates (after 6  weekexposure periods). Paired reference locations are available near
the areas of deployment, but outside the areas of immediate CSO  influence. US  Water News (1998) also described a
study in Boston Harbor that found caffeine at levels of about 7 ug/L in the harbor water. The caffeine content of regular
coffee is about 700 mg/L, in contrast.

    Kratch  (1997) summarized several  investigations on cataloging the DMA of  £ co/i to identify their source in water.
This  rapidly emerging technique seems to have great  promise in addressing a number of nonpoint source water pollution
issues. The procedure, developed at the Virginia Polytechnic Institute  and State University,  has been used in Chesapeake
Bay.  In one example, it was possible to identify a large wild animal population as the source of fecal coliform
contamination of a shellfish bed, instead of suspected failing  septic tanks. DMA patterns in fecal coliforms vary among
animals and birds, and it is relatively easy to distinguish  between human and non-human sources of the bacteria.
However, some wild animals  have  DMA patterns that  are not easily distinguishable. Some researchers question the  value
of £  coll DMA fingerprinting believing  that  there is little  direct relationship between £  co/i and  human pathogens.
However, this method should be useful to identify the presence of sewage  contamination in stormwater or in  a  receiving
water.

   One application of the technique,  as described by Krane, ef a/  (1999) of Wright State University, used randomly
amplified polymorphic DMA polymerase  chain reaction (RAPD-PCR) generated  profiles of naturally  occurring  crayfish.
They  found that changes in  the underlying genetic diversity of these  populations  were significantly correlated  with the
extent to which they have been exposed to anthropogenic stressors. They concluded that this rapid and relatively simple
technique can be used to develop a sensitive means of directly  assessing the impact of stressors upon ecosystems.
These Wright State University researchers have also  used the RAPD-PCR techniques on populations of snails, pill bugs,


                                                     105

-------
violets, spiders, earthworms, herring,  and some benthic macroinvertebrates, finding relatively few obstacles in its use for
different organisms. As noted above,  other researchers have used DMA profiling techniques to identify sources of £ coli
bacteria found in coastal waterways.  It is possible that these techniques can be expanded to enable rapid detection of
many different types of pathogens in receiving waters, and the most likely sources  of these pathogens.

    Other Tracer Methods for Identifying Sources of Water. Stable isotopes  had been recommended  as an efficient
indicator of illicit connections to storm sewerage. A demonstration was conducted in Detroit as part of the Rouge River
project to identify sources of dry weather flows in storm sewerage (Sangal, ef a/. 1996). Naturally occurring stable
isotopes of oxygen and hydrogen can be used  to identify waters originating from different geographical sources  (especially
along a north-south  gradient). Ma and Spalding (1996) discuss this approach  by  using stable isotopes to investigate
recharge of groundwaters  by  surface waters.  During  water vapor transport from equatorial source  regions to higher
latitudes, depletion of heavy isotopes  occurs with rain. Deviation from a standard  relationship between deuterium and 180
for a specific area indicates that the water has undergone additional evaporation. The ratio  is also affected by seasonal
changes. As discussed  by Ma and Spalding (1996), the Platte River water is normally derived in part from snowmelt from
the Rocky Mountains, while the groundwater in  parts  of Nebraska is mainly contributed from the Gulf air stream. The
origins of these waters are sufficiently different and allow good measurements of the recharge rate of the surface water
to the groundwater. In  Detroit, Sangal, ef a/. (1996) used differences in origin between the  domestic water supply,  local
surface waters,  and the local groundwater to identify potential sanitary sewage contributions to the separate storm
sewerage. Rieley, eta/,  (1997)  used stable isotopes of carbon  in marine organisms to distinguish the  primary source of
carbon being consumed  (sewage sludge vs. natural carbon  sources) in two  deep sea  sewage sludge disposal areas.

    Stable isotope analyses would not be able  to distinguish between sanitary sewage, industrial discharges, washwaters,
and  domestic water, as they all have the  same origin, nor  would it be possible  to distinguish sewage from  local
groundwaters if the domestic water supply was  from the same local aquifer. This method works best  for situations where
the water supply  is from  a  distant source and where separation of waters into separate flow components is not  needed.
It may be an excellent tool to study the effects of deep well injection of stormwater on deep aquifers having distant
recharge sources  (such  as in the Phoenix area). Few laboratories can analyze for these stable isotopes,  requiring shipping
and a long wait for the analytical results. Sangal, etal. (1995) used Geochron Laboratories, in Cambridge, Massachusetts.

    Dating of sediments using 137Cs was described by Davis, et a/. (1997). Arsenic contaminated sediments in the
Hylebos Waterway in Tacoma,  WA, could have  originated  from numerous sources, including a pesticide manufacturing
facility, a rock-wool plant, steel slags, powdered metal  plant, shipbuilding facilities, marinas and arsenic boat paints, and
the Tacoma Smelter. Dating the sediments,  combined with knowing the history of potential discharges and conducting
optical and electron microscopic studies of the  sediments, was found to be a powerful tool to differentiate between the
different metal sources to the sediments.

Conclusions

    Recent tests examined several  potential  tracer parameters during a project characterizing stormwater that had
collected in telecommunication manholes, funded by Tecordia (previously Bellcore), AT&T, and  eight regional  telephone
companies  throughout  the  country (Pitt and Clark 1999).  Numerous  conventional constituents,  plus major  ions, and
toxicants were measured, along with  candidate tracers  to indicate sewage contamination of this water. Boron, caffeine,
coprostanol,  £ coll, enterococci, fluorescence (using specific wavelengths for detergents), and a simpler test for
detergents were evaluated,  along with the use of fluoride, ammonia, potassium, and obvious odors and color. About 700
water samples were evaluated for  all of these parameters, with the exception of bacteria and boron (about 250 samples),
and only infrequent samples were analyzed for  fluorescence. Coprostanol was found in about 25% of the water samples
(and in about 75% of the 350 sediment samples analyzed). Caffeine was found in very few samples, while elevated E.
co//and enterococci (using IDEXX tests) were observed in about 10% of the samples. Strong sewage odors in water and
sediment samples were also detected in about 10% of the samples. Detergents  and fluoride (at >0.3 mg/L) were found
in about 40% of  the samples and are expected to have been contaminated with industrial activities (lubricants and
cleansers) and not sewerage. Overall, about 10% of the samples were therefore expected to have been  contaminated
with  sanitary  sewage, about the same rate  previously  estimated for stormwater systems.
                                                      106

-------
    Additional laboratory tests, funded by the University of New Orleans and the EPA, were conducted using many
sewage and laundry detergent samples and found that the boron test was a poor indicator of sewage, possibly due to
changes in formulations in modern laundry detergents. Other laboratory tests found that fluorescence was an excellent
indicator of sewage, especially when  using specialized "detergent whitener"  filter sets, but was not very repeatable. We
also examined several  UV absorbance wavelengths as sewage indicators and found excellent correlations with 228 nm,
a wavelength having very little  background  absorbance in local spring waters, but with a strong  response  factor with
increasing strengths of sewage. We recommend that our originally developed and tested protocol still be used  as the most
efficient routine  indicator of sewage contamination of stormwater drainage systems, with  the possible addition of specific
E. co//and enterococci measurements and UV absorbance at 228 nm. The numerous exotic tests requiring specialized
instrumentation  and expertise reviewed in this  paper do  not appear to warrant their expense and  long analytical turn-
around  times,  except  in  specialized  research  situations.

References

Chung,  K.H., K.S. Ro, and S.U. Hong. "Syntheticdetergent in an urban stream in Korea." Proceedings of the 65th Water
Environment Federation Technical  Exhibition and Conference,  Miami Beach, Florida. Volume 4,  Surface Water Quality
and Ecology, pp. 429 - 434. Alexandria, VA. 1995.

Davis, A., P. DeCurnou, and I.E.  Eary. "Discriminating between sources of arsenic in the sediments of a tidal waterway,
Tacoma, Washington." Environmental Science & Technology. Vol. 31, no. 7, pg. 1985.  1997.

Desbrow, C., E.J. Routledge, G.C. Brighty, J.P. Sumpter,  and M. Waldock. "Identification of estrogenicchemicals in STW
effluent. 1.  Chemical fractionation  and in vitro biological  screening."  Environmental Science & Technology. Vol. 32, no.
11, pg.  1549. 1998.

Eganhouse, R.P., D.P.  Olaguer, B.R.  Gould and C.S.  Phinney.  "Use of Molecular Markers for the Detection of Municipal
Sewage Sludge at Sea." Marine  Environmental Research. Volume 25, No.1. pp. I-22.  1988.

Fujita, M., M. Ike,  Y. Goda, S. Fujimoto, Y. Toyoda, and K-l.  Miyagawa. "An  enzyme-linked immunosorbent assay for
detection of linear alkylbenzene  sulfonate: development and  field studies." Environmental Science & Technology. Vol.
32,  no.  8, pp. 1143-1 146. 1998.

Gonalez-Mazo, E., J.M. Forja, and A. Gomez-Parra. "Fate and  distribution of linear alkylbenzene sulfonates in the littoral
environment."  Environmental Science  & Technology.  Vol. 32, no. 11, pp.  1636-1 641. 1998.

Grimalt, J.  0.,  ef a/. "Assessment  of fecal sterols and ketones as indicators of urban sewage inputs to coastal waters."
Env. Science &  Technology, Vol. 24, no. 3, p357. March 1990.

Halling-Sorensen, B., S. Nors Nielsen, P.F. Lanzky,  F. Ingerslev, H.C. Molten Lutzhoft, andS.E. Jorgensen. "Occurrence,
fate and effects  of pharmaceutical substances  in  the environment -A  review." Chemosphere.  Vol.  36, no. 2, pp. 357 -
393.1998.

Holm, S.E., J.G. Windsor. "Exposure assessment of sewage  treatment plant  effluent  by  a selected chemical marker
method." Archives Env. Contam. & Tox. Vol.  19, no. 5, p674. Sep-Ott 1990.

Hun,  T. "NewsWatch: Water Quality,  Studies indicate  drugs in water may  come from effluent discharges." Water
Environment &  Technology, pp. 17 -  22. July 1998.

Kramer, J.B., S. Canonica, J. Hoige, and J. Kaschig. "Degradation of fluorescent whitening agents in sunlit natural
waters." Environmental  Science & Technology. Vol. 30, no. 7, pp.  2227 - 2234. 1996.

Krane, D. E., D.  C. Sternberg, and G.A. Burton. "Randomly amplified polymorphic DMA profile-based measures  of genetic
diversity in  crayfish correlated with environmental  impacts." Environmental Toxicology and Chemistry. Vol.  18, no. 3, pp.
504 - 508. March  1999.
                                                     107

-------
Kratch, K. "NewsWatch, Water Quality: Cataloging  DMA of E. coli sources pinpoints contamination causes."  Water
Environment  & Technology. Vol. 9, no. 8, pp. 24 - 26. August 1997.

Lalor, M. An Assessment of Non-Stormwater Discharges toStorm Drainage Systemsin Residential and Commercial Land
Use  Areas.  Ph.D. Dissertation. Department of Civil and Environmental  Engineering. Vanderbilt  University.  1994.

Ma,  L. and  R.F. Spalding.  "Stable  isotopes characterization  of the impacts of  artificial ground water recharge." Water
Resources Bulletin. Vol. 32, no. 6, pp. 1273 - 1282. December 1996.

Nichols, P.O., R. Leeming. M.S. Rayner, and V. Latham. "Use of capillary gas-chromatography for measuring fecal-
derived sterols application to  stormwater, the sea-surface microlayer, beach greases, regional studies, and distinguishing
algal blooms and human and nonhuman sources of sewage  pollution." Journal of Chromatography. Vol.  733,  no. 1 - 2,
pp. 497-509. May 10, 1996.

Poiger, T., J.A.  Field, T.M. Field, and W. Giger. "Occurrence of fluorescent whitening agents in sewage and river
determined by solid-phase extraction and high-performance liquid Chromatography." Environmental Science &
Technology.  Vol.  30, no.  7, pp. 2220 - 2226. 1996.

Pitt,  R. and S. Clark.  Communication Manhole  Water Study: Characteristics of Water Found in Communications
Manholes. Final Draft. Office  of Water, U.S. Environmental Protection Agency.  Washington, D.C. July  1999.

Pitt,  R. and  J. McLean.  Humber River Pilot Watershed Project, Ontario Ministry of the Environment, Toronto, Canada.
483  pgs. June 1986.

Pitt,  R., M. Lalor, R. Field, D.D. Adrian, and D.Barbe'. A User's Guide forthe Assessment ofNon-Stormwater Discharges
into Separate Storm  Drainage Systems. Jointly published  by the Center of  Environmental  Research Information, US  EPA,
and the Urban Waste Management & Research Center (UWM&RC). EPA/600/R-92/238. PB93-131472. Cincinnati, Ohio.
January 1993.

Rieley, G., C.L. VanDiver, and G. Eglinton. "Fatty acids as sensitive tracers of sewage sludge carbon in a deep-sea
ecosystem." Environmental Science  & Technology. Vol. 31,  no. 4, pg. 1018. 1997.

Routledge, E.J.,  D. Sheahan, C. Desbrow, CC. Brighty,  M.  Waldock, and J.P. Sumpter. "Identification of estrogenic
chemicals  in STW  effluent. 2. In vivo responses in  trout and roach." Environmental Science & Technology. Vol. 32, no.
11, pg. 1559. 1998.

Sangal, S., P.K.  Aggarwal, and D. Tuomari. "Identification  of illicit  connections in storm  sewers: An innovative approach
using stable  isotopes." Rouge  River Studies.  Detroit,  Ml.  1996.

Sherwin, M.R., et al. "Coprostanol (5?-cholestan-3?-ol) in lagoonal sediments and mussels of Venice, Italy." Mar. Potlut.
Bull.  Vol. 26, no. 9, p501. September 1993.

Shuman, R.  and  J.  Strand.  "King County water quality assessment: CSO  discharges, biological impacts being assessed."
Wet  Weatherx,  Water Environment Research  Foundation. Fairfax, VA. Vol. 1, no. 3, pp. 10  - 14. Fall  1996.

Terzic, S., and M.  Ahel.  "Determination of alkylbenzene sulphonates in the  Krka River Estuary." Bull. Environ. Contam.
Toxicol. Vol. 50,  no. 2, p241. February 1993.

US Water News.  "Scientists say Boston Harbor water full of caffeine." Pg. 3. July 1998.

Venkatesan, M.I. and I.R.Kaplan. "Sedimentary coprostanol  as an index of sewage addition in Santa Monica Basin,
Southern California." Env.  Science & Technology, Vol. 24, no. 2, p208. February 1990.
                                                    108

-------
Zeng, E. and C.L. Vista.  "Organic pollutants in the coastal environment off San Diego,  California. 1. Source identification
and  assessment  by  compositional  indices of polycyclic aromatic hydrocarbons." Environmental Toxicology and Chemistry,
Vol.  16, no. 2, pp. 179- 188.  1997.

Zeng, E., A.M. Khan, and K. Iran. "Organic pollutants in the coastal environment off San Diego, California. 2. Using linear
alkylbenzenes to trace sewage-derived  organic materials." Environmental Toxicologyand Chemistry. Vol.  16, no. 2,  pp.
196-201.  1997.

Zoller,  U.,  et al.  "Nonionic detergents as tracers  of groundwater pollution caused by municipal sewage." Chemistry for
the Protection of the Environment (Plenum). p225. 1991.
                                                      109

-------
 Elimination of Illicit Connections in Coastal  New Hampshire Spurs Cooperation
                                          and Controversy
                                    Natalie Landry and Robert Livingston
                            New Hampshire Department of Environmental Services
                                         Concord, New Hampshire
Introduction

    Discharging stormwater runoff into our waterways has long been an accepted practice. In theory, storm drainage
pipes should only discharge during  and  after storm events unless the source is groundwater or surface water piped
underground. Therefore,  the  dry weather discharge  should  be relatively  free of contaminants. However, many
communities across the country are finding out this is not always true. Some cities and towns are discovering illegal
connections of residential  and  commercial sewer lines to storm water collection systems. Illicit connections have been
identified by the New Hampshire Department of Environmental Services (DES) as the point source of high fecal coliform
levels in the New Hampshire coastal  basin (Jones,  1995). These  illegal connections pose a health riskto those recreating
in the coastal waters and  have forced the closure of shellfish growing waters to harvesting.

Goals of the Coastal Investigations Programs

    Determining the extent of dry weather contamination  in  storm drainage systems  is the first step an investigator should
take when  researching stormwater  pollution. Dry  weather flows in  storm drainage systems  are often  the result of
groundwater infiltration, but can also result from inappropriate connections from  commercial, industrial, or residential
buildings. In 1996, the New Hampshire DES published  the Coastal Basin Nonpoint Source Pollution Assessment and
Abatement  Plan (NHDES,  1996) that directed coastal investigations of each community's storm drainage system during
dry weather. This decision to conduct dry weather investigations in the coast was made after 300 illicit connections were
identified in  the northern New Hampshire city of Berlin. State environmental officials were convinced that illicit connections
were always present in storm drainage systems that were once considered a pollution threat only during wet weather.

    DES initiated a multi-year effort that focused on identifying and abating the sources of the bacterial violations found
in the state's coastal waters with the goal of opening shellfish growing waters during dry weather (Landry, 1997). About
the  same time,  the New Hampshire Estuaries Project (NHEP) began a  three-year process of developing a comprehensive
management plan aimed  at restoring, protecting and enhancing the water quality and living resources of the state's
estuaries. The major goal of the NHEP was to address the sources  of pollution currently impacting the estuaries  and
prevent future problems through effective land use planning and shoreline protection of the coastal resources (NHEP,
1996). To accomplish this goal, part  of the NHEP strategy was to locate and remediate the sources of the water quality
violations,  primarily  bacterial violations, found in  the estuaries and  coastal  waters (Landry, 1997). DES and NHEP
combined resources and developed an investigation strategy with the  overall goal of improving and protecting estuarine
water quality.

Specific  Program Objectives

    The main objectives of the investigation strategy were to identify inappropriate connections in the storm drainage
systems of  urban, coastal communities and  to eliminate the illicit connections through the available means,  which include
voluntary compliance and enforcement.

A Brief Look at  Coastal  New Hampshire

    The eighteen  miles of New  Hampshire coastline do  not begin  to tell the story  of the state's abundant marine
resources. The relatively modest coastline is only a small part of the coastal basin. The estuarine resources include the
Great Bay  Estuary and seven associated tidal rivers, Hampton Harbor, and  Rye Harbor. These waters are used by
residents and many visitors for swimming, boat touring, shellfish harvesting, surfing, and angling. Forty-two communities


                                                   110

-------
comprise the coastal basin watershed, with a population density just under 300 residents/mile2(Jones, In Review). The
upper watershed is generally undeveloped and forested while the more urban centers are situated in the lower watershed
as the rivers approach the coast.

    Ten wastewater treatment facilities are situated on the tributaries of Great Bay and Hampton Harbor and two facilities
discharge directly into the Atlantic Ocean. Coastal communities are working diligently to upgrade the  wastewater
treatment facilities and sewage collection systems. Inflow/infiltration problems and undersized pump stations plague the
treatment facilities and have resulted in financial hardshipsforaffected municipalities. Shellfish growing waters have been
temporarily  closed after  heavy rainstorms when bacteria levels  rise due to sewage  by-passes. Sewage is a well-
recognized threat to the marine environment because it often contains harmful chemicals, disease-causing bacteria and
viruses, dissolved  material and solid matter. Pathogens can cause avariety of illnesses and humans are exposed to these
organisms through contaminated water, shellfish, and fish (Sea Grant, 1999).

Investigating  Illicit  Connections

    Recently, more and more watershed studies are investigating  inappropriate discharges in storm drainage systems.
This pollution source originates  from an identifiable point and flows through the storm drainage system  to the outfall pipe.
For example, instead of connecting to the sewer system, a direct connection of sewer service discharges into the  storm
drainage system. Other inappropriate sources include floor drains and laundry pipes. These inappropriate connections
are also referred to as illicit or illegal cross connections. The health threat and the potential to interfere with stormwater
contamination assessments elevate illicit connections to priority status for watershed managers to investigate.

    Pitt et  al. (1993), in  cooperation with the Center of Environmental  Research  Information,  U.S. Environmental
Protection Agency, published a user's guide for conducting investigations of illicit connections. Several of the methods
suggested  in this guide  were implemented  during the New  Hampshire  coastal investigations.  Detailed  surveys to
determine the extent of contamination through specific water quality monitoring and careful observation of storm drainage
outfalls are recommended for each type of land use in the watershed. Pitt recommends an initial phase of investigative
protocol that includes the initial mapping and field surveys. The initial activities are followed by more detailed watershed
surveys to locate and correct the sources of the contamination in the identified problem areas. After corrective action has
been taken, repeated  outfall field surveys are required to ensure that the outfalls remain uncontaminated.
                                                      111

-------
Surveys of Storm Drainage Systems

    Over the course of the investigations, several methods were used,  ranging from the initial screening process of
surveying storm drainage discharges to dye testing the indoor plumbing of suspected sources. Steps between the initial
survey and the final determination of the source, included analyzing the discharge for water quality; visual  and  odor
observations at  outfalls, manholes,  and catchbasins; smoke testing;  and video inspection of the storm drainage and sewer
systems.

    Tidal rivers  and  coastal waters were divided into study sections by community.  The urban, downtown centers of these
communities were targeted based on the existence of the storm drainage infrastructure. The investigators compiled  maps
and  as-built drawings of the storm drainage and sewer infrastructure. If the maps were inaccurate,  insufficient, or
unavailable, information  on the storm drainage system was developed based on field investigations by the staff, typically
with  the assistance  of public works employees.

    Communities with maps based in a  geographic information system (CIS) saved staff time and were generally more
accurate than  record drawings that  are not updated regularly (Landry,  1997). Tuomari (1996)  applied the Rouge
Watershed Geographic Information System to the Wayne County Illicit Connections Detection Program and concluded
that the new GIS  strategy eliminated the need to use maps and graphics from disparate reports and sources, significantly
reducing the time and effort once spent on research, field data acquisition, and interpretation.

    Beginning in the summer of 1996, the coastal shorelines were surveyed at low tide, on foot or by canoe, depending
on access, for potential pollution sources. All pipes, seeps, streams, and swales with flow were sampled for bacteria. In
addition, temperature was measured and observations  relating to the condition of the pipe (stained  or structurally
damaged),  odor, evidence of untreated wastewater (toilet paper, etc.), turbidity, color, debris, estimated flow, and any
other observations were noted. Dry pipes were rechecked on several occasions for intermittent flow. Evidence indicating
the presence of wastewater and/or elevated  bacteria levels prompted further  investigation of these locations.

    Upstream catchbasins and manholes associated with the outfall pipes that were identified in the screening process
previously described, were surveyed for evidence of wastewater and sampled for bacteria.  Smoke testing (using non-toxic
smoke  blown into  catch basins)  was then  used to identify buildings connected to the storm drainage system by canvassing
the neighborhood for vents emitting smoke. Final confirmation of an illicit connection from the buildings that emitted smoke
was  accomplished with dye testing of indoor plumbing and observing the storm drainage and sewer systems for the
presence/absence of the dye.

    Feeder streams were surveyed for outfall pipes with dry weather flow. Other potential bacteriological sources (e.g.,
pigeon roosting sites on bridges) were bracketed with water quality sampling stations. Where contaminated seeps and
swales were suspected, the drainage area was surveyed for  potential sources such as broken sewer mains.

Water Quality Results

    Bacteria data (1997/98) from outfall pipes with confirmed cross connections ranged from 1,700 - >1,000,000 E. coli
counts/I  00 ml during dry weather in Dover, New Hampshire. Many outfall pipes with cross connections had a gray biomat
comprised of filamentous bacteria coating the inside of the pipe and, often, the rocks or sediment below. These biomats
were used as a  wastewater indicator based on  the presence of these mats at more than 50% of the outfalls with  confirmed
cross connections.

    Dr.  Stephen Jones of the University of New  Hampshire Jackson Estuarine  Laboratory conducted a twelve-month
study that examined the significance of all flow coming from urban storm drainage systems in the downtown Dover
watershed of the Cocheco River (Jones, 1998). Jones found that storm drains were consistent sources of relatively high
concentrations of bacterial indicators and  pathogens at concentrations that exceeded  state standards for recreational and
shellfish-growing waters during  both dry and wet weather.

    Flow from a damaged stormwater outfall pipe was determined to have a geometric mean £. coli concentration of
1,047,199cfu/100 ml and a dissolved inorganic nitrogen (DIN) concentration of 22.4 mg/l.The data were brought to the
attention of DES and an investigation revealed a cross connection from a commercial building. Dr. Jones continued to
monitor the quality of flow after the cross connection was eliminated and the results show a significant decline in both
bacteria and DIN  geometric means. The post-repair results were 93 E. coli cfu/100 ml and 7.2 DIN mg/l.


                                                     112

-------
    Although no public health problems were known to have occurred as a result of exposure to bacterial pathogens in
the Cocheco River, the contamination may be a significant contribution to the fecal-borne bacteria that are presently the
reason for closing the area's shellfish growing waters in New Hampshire and restricting harvests in Maine (Jones, 1998).

Remedial Actions

    Once confirmed, illicit connections in coastal New Hampshire  have been eliminated in different ways.  The most
desired course of action from the  regulatory perspective is  voluntary compliance and many fixes  have  been accomplished
through this  process.  Economics and prioritization of the many demands on public works  departments sometimes compel
the state and federal environmental agencies to initiate regulatory action to eliminate raw wastewater  discharges into
surface waters. Lawsuits, although not common, have been filed against municipalities after cross connections were
discovered.

Voluntary Compliance: Town of Exeter Case Study

    In 1994, researchers at the UNH Jackson Estuarine Laboratory (Jones and Langan, 1995) reported elevated dry-
weather bacterial levels  collected in Morris Brook,  a tributary to the Squamscott River in Exeter, New Hampshire. In  1996,
DES  collected bacteria samples  at various locations on Norris Brook (NHDES,  1997) and found relatively low E. coil
concentrations  of <150  counts/100 ml.  In  1998, an  Exeter  official urged DES  to  investigate the  watershed for
contamination based on the 1994 data that showed a fecal coliform concentration of 600 counts/I 00 ml. In April of 1998,
DES and a town official conducted a survey of the lower watershed and discovered a storm drainage  outfall discharging
a large volume of flow even though the weather had been dry. Upon closer inspection, toilet paper was observed in the
outfall pipe and the  immediate area.

    The town public works department was notified of the survey results and, following  reminders by DES, began
investigations to determine the sources of untreated wastewater.  Progress was slowed because several of the residences
were  rental  properties which  involved  contacting the owner, who was in some  cases  from out-of-state, and gaining
permission to access the building for dye testing. By November, the town reported that  a few of the cross connections
still remained. DES considered enforcement action and an administrative fine but did not take that action  to maintain the
spirit of cooperation. In January of 1999, the town reported that the owner of the last remaining property to be dye tested
was not responding to requests  for access. More prodding by DES followed and  in February 1999, DES received
notification that the  cross connections were eliminated.  A follow up inspection confirmed the absence of  untreated
wastewater in the storm drainage outfall.

    A lesson learned from this experience is that a persistent, local advocate is often the key to maintaining attention on
a local water quality  problem. In addition, local advocates, whether a conservation commissioner, selectmen, or citizen,
often  have detailed knowledge of the complaint and the locale, which  provides valuable  and time-saving information to
the state investigators.

    Time and resource demands on local officials as well as state investigators can cause this process to  be distressingly
slow.  Budgeting for 2-3  cross connection investigations and fixes per year is recommended at approximately $6,000 per
fix, to help alleviate  the unexpected financial burdens on  urban communities when illicit connections are found.

    Bacteria alone should not be  the determining factor of the presence or absence of an illicit connection for a variety
of reasons. Chlorine or other toxins in untreated wastewater may depress bacteria levels and bacteria lack conservative
behavior, which deem it a poor indicator (Pitt, 1993). Investigators have found that a careful and thorough outfall survey
is  usually  more informative than just collecting water samples.

Enforcement Action: City of Newmarket Case Study

    Enforcement  is another tool available  to DES to achieve compliance. For example, setting timetables  for compliance
milestones in a legal document is a method that, while typically thought of as a  burden to a community, may actually
provide the impetus  for action in a positive way. Public works departments of New Hampshire coastal communities are
not equipped with large, discretionary budgets to address unplanned remediation of illicit connections. When faced with
this dilemma, an enforcement action against the community  provides public works departments with the validation to
support a  request for additional funding from the officials who approve the allocation of funds.
                                                    113

-------
    As the mill Town of Newmarket, New Hampshire, developed over the years, a small watercourse named Moonlight
Brook was built over and culverted in the center of downtown. In 1996,  DES investigated Moonlight Brook based on
historic elevated bacteria levels. The DES investigation revealed dry weather E. coll concentrations as high as 41,600
counts/I 00ml in the brook. DES encouraged the town to initiate dye testing of the structures in the vicinity of Moonlight
Brook but, at that time, the town was reluctant to allocate staff and funding for clean up efforts (NHDES,  1997).

    An administrative order was issued  by the US  Environmental Protection Agency for various violations of permitted
effluent limitations in October, 1997 and  included a requirement that the town eliminate the raw sewage discharges from
the storm sewer system (USEPA Docket No. 97-78). The order required a plan and schedule for eliminating  any pollutants
discharging during dry weather. The order also specified sampling of each active dry weather discharge that remained
following elimination of the illicit connections to the system identified by the town's fieldwork.

    In response to the order, the town  hired  an environmental consultant to address the problems at the wastewater
treatment facility and the illicit connections. During the summer of 1997, the consultant and the town performed a dye
study of the subdrainage area that the town suspected was the likely source of bacterial contamination identified at the
discharge. The dye study resulted in the  identification of a total  of  four untreated discharges to the storm drainage system
from three properties. A subsequent video inspection of the sewer lines adjacent to these properties revealed that the
sewer service connections from these  properties might have been installed at  the time the original  sewer was constructed.
The consultant then concluded that this would indicate the sources  of sewage discharging to the storm drainage are
broken sewer service connections rather than direct connections. The town stated that the  remedial work would be
completed by June 1998 (Plante, 1998).

    Another storm discharge pipe servicing this area was separated in 1985 and  1986, at which time dye testing  was
performed to identify sanitary services that were connected to the sewer. The consultant determined that it would be
unlikely that direct sanitary service connections to the storm drain were present in this area,  however, broken service
connections could result in sewage entering the storm drain culvert along Main Street. A dye study was planned for May
1998.

    A total of 59 properties were included in the dye-testing program. Four of the properties were confirmed to be cross-
connected to the storm drainage system. Two of the four were the result of direct connections of sewer laterals to the
drainage system. The remaining two were a result of exfiltration from the sewer lateral through the ground to the drain
line (Town of Newmarket, 1999). The town reports a 90% reduction in the E. co//counts following the elimination of the
illicit discharges.

Legal Action: City of Dover Case Study

    In the 1970's, the City of Dover, New Hampshire, constructed a new sewage collection system and treatment facility.
In 1997,  DES  investigators began surveying the storm drainage  outfalls for contamination. Around this same time,
University of New Hampshire researcher Dr. Stephen Jones initiated a study in  Dover to determine the significance of
flow (both dry and wet weather) coming from urban storm drainage systems (Jones, 1998). Jones identified a source of
bacterial contamination to be a cross connection later confirmed by DES  and the City of  Dover  Public Works Department.
The city fixed the illicit connection by connecting the sanitary service into the sewer main, at no charge to the building
owner, while noting substantial flow from this  service due to a hair salon in the building.

    After learning about the existence of the cross connection, the building owners made an unsuccessful request to the
city for an abatement of the sewer fees  they had paid since 1981 and initiated legal proceedings. The city alleged that
the case  law mandated  a decision in its favor  and filed a motion for summary judgement (Strafford Superior Court, Order
#98-C-207).  In a responding order from the judge, the case law was said to illustrate that the Court had  considered a
variety of factors in related cases including (1) whether the new and old system were integral to one another, (2) whether
the benefit provided to the plaintiff under the new and old systems was comparable, and (3) whether the property owner
had access to the new system. The motion for summary judgment was denied because the Court found that these were
issues for a jury and that summary judgment  at that stage would be premature.

    A trial date was  set. One week before the trial was to occur, the two parties settled out of court. The terms of the
settlement were confidential. If the property owners were successful in seeking a tax abatement and damages for unjust
enrichment, implied contract, and negligent misrepresentation, as sought, the pollution investigations could have been
                                                    114

-------
in jeopardy of becoming ineffective. Such a precedent could have led to other similar suits and would effectively remove
the incentive for municipalities to be proactive in fixing cross connections.

Conclusions

    The Department  of Environmental Services,  in conjunction with the New Hampshire  Estuaries Project,  has
systematically identified illicit connections in the urban communities of coastal New Hampshire.  Applying both voluntary
compliance and enforcement has resulted in the removal of cross connections to the storm drainage systems and a
decrease in the contamination reaching the  coastal surface waters. DES is currently monitoring the shellfish growing
waters to determine the extent of water quality improvement resulting from the removal of illicit connections.

References

Hamilton R. Krans,  and Allan B. Krans v. City of Dover, No. 98-C-207 (Strafford, SS. N.H. Dec. 28,1998) (order denying
defendant's motion for summary judgment.)

Jones, S.H. and R. Langan. 1995. Assessment of Nonpoint Source Pollution in Tributaries to the Great Bay Estuary. Final
Report. New Hampshire Office of State Planning/Coastal Program. Concord, New  Hampshire.

Jones, Stephen,  Ph.D. In  Review.  A Technical Characterization of Estuarine and Coastal  New  Hampshire. New
Hampshire Estuaries Project, Office of State Planning, Portsmouth,  New Hampshire.

Jones, Stephen,  Ph.D. 1998. Stormwater contamination of New Hampshire Coastal Surface Waters. Final Report
Submitted  to the  New  Hampshire Coastal Program/Office of State Planning, Concord, New  Hampshire.

Jones, Stephen, Ph.D.  1999. Public Health Significance of Stormwater-Borne Microorganisms. Final Report to the New
Hampshire Department of Environmental  Services, Concord,  New Hampshire.

Landry, Natalie.  1997.  An  Investigations  of Water Quality in New Hampshire Estuaries.  Final Report  to the New
Hampshire Estuaries Project. New Hampshire Department of Environmental Services, Concord, New Hampshire.

New Hampshire Department of Environmental Services (NHDES). 1996.1996 Coastal Basin Nonpoint Source Pollution
Assessment and  Abatement Plan. Concord, New Hampshire.

New Hampshire Department of Environmental Services (NHDES). 1997. 1996 Nonpoint Source Coastal Assessment
Report. Concord, New Hampshire.

New Hampshire Estuaries Project  (NHEP).  1996.  New  Hampshire Estuaries Project National Estuary Program: EPA/State
Management Conference Agreement. New Hampshire Office of State Planning, Concord, New Hampshire.

Pitt, Robert, Melinda Lalor, Richard Field, Donald Dean Adrian, and Donald  Barbe. 1993.  Investigation of Inappropriate
Pollutant Entries into  Storm  Drainage  Systems, A Users Guide. United  States Environmental Protection  Agency.
EPA/600/r-92/238.

Plante, Thomas R., P.E. 1998. Report to the US Environmental Protection Agency Regarding Administrative Order No.
97-87, Item 6-Dry  Weather Discharges From Stormwater System. New  Hampshire Department of Environmental
Services, National Point Discharge Elimination System files, Concord, New Hampshire.

Sea Grant. 1999. Sewage: Wastewater Major Pollutant Challenge. Sea Grant National Media  Relations Office,  National
Sea Grant College  Program, http://www.seagrantnews.org/news/sewage.html.

Town of Newmarket. 1999. Storm Drain/Sewer Cross Connections, Identification and Correction. Final report to the New
Hampshire Estuaries Project. Prepared by Underwood Engineers, Portsmouth,  New Hampshire.

Tuomari, Dean. 1996. Rouge River Watershed Illicit Sewer Connection Detection Program:  A GIS Application. Watershed
96. Wayne County  Department of Environment, Detroit, Michigan.
                                                   115

-------
United States Environmental Protection Agency. 1997. Administrative Order No. 97-78, Findings of Violation and Order
for Compliance In the Matter of Newmarket, New Hampshire, NPDES Permit No. NH01 00196.
                                                  116

-------
     Using Collaborative Problem-Solving to Protect  North Carolina's Coastal

        Resources: The Experience of the White Oak River Advisory  Board*


                           Leon E. Danielson, C. Suzanne Hoover, Christy A. Perrin
                                       North Carolina State University
                              Department of Agricultural and Resource Economics
                                                Raleigh, NC

                                              Nancy M. White
                                        North Carolina State University
                                             School of Design
                                                Raleigh, NC

                                                Ron Elmore
                                 North Carolina Department of Transportation
                                                Raleigh, NC

                                             Jennifer L. Platt
                                                Town of Gary
                                                 Gary, NC


Introduction

    In North Carolina coastal estuarine systems, land use change  has been implicated as a significant cause of water
quality impairment  (NC Department of  Environment and Natural Resources, 1997; White,  et al., 1998).  Such  development
processes change surface hydrology, pollutant  delivery, and,  as  a consequence, adjacent water quality.  Decisions
regarding  placement, density, and type of development are controlled by policy  implementation  at the local level.
Furthermore, while the degree of impact may vary with each location, it is the cumulative effects throughout a watershed
that can be most damaging to water quality. Hence,  there is a need to develop and enact policy locally, but on a multi-
jurisdictional, watershed  basis.

    Increasingly,  local communities  and governments are showing  interest in  playing a role in developing  and
implementing solutions to water quality problems (NC Department of Environment and Natural Resources,  1997).
However, logistical complications arise upon implementation of this  concept. First, a mechanism for effectively involving
local citizen  stakeholders in the policymaking process may not exist and/or is difficult to establish (Danielson, 1998).
Second, technical  data needed to address local issues and concerns are often not readily available, or are in a form not
easily understood. Third,  programs for  addressing water quality problems on a watershed-basis may not exist, suggesting
a need to develop, coordinate, and deliver multi-jurisdictional education on water quality issues and  policy  alternatives.
Through a project entitled  Watershed Education for Communities and Local Officials (WECO), the North Carolina
Cooperative  Extension Service has worked with  a number of state and federal agencies, along with citizens and local
governments within a coastal watershed to address these needs.

    The goal of this project is to improve water quality in all of the White Oak River Watershed through involvement and
education of citizens and  government officials who live and work in the watershed. The project's main thrusts are : 1) the
delivery of technical information and educational material on water quality, management strategies,  and policy options
that support watershed-based planning; 2) the empowerment of local citizens by facilitating collaborative partnerships
between communities, local officials, and state agencies within the watershed;  and 3) the facilitation of the development
• Funding for this project is provided by USDA-CSREES under project number 97-EWQI-1-0150.


                                                   117

-------
of local stakeholder - driven policy recommendations for the entire watershed. This paper discusses the application of
these concepts to an issue of critical importance to local citizens.

Background

    The White Oak River watershed is one of four rivers in  the White Oak River Basin (Figure 1). It is 48 miles long and
encompasses 320 mi2 .  The watershed begins in freshwater creeks and swamps of Jones County, NC, and contains
portions of three other counties—Craven, Onslow, and Carteret. Along its route to Bogue Sound and the Atlantic Ocean,
the river traverses between 30 ft. banks, which are relics of ancient dune ridges. This river is home to five threatened
or endangered  organisms,  including alligators; loggerhead, green, and leatherback turtles, and the Croatan crayfish. The
river and  its estuarine waters have extensive primary nursery waters and provide habitat for several anadromous species--
herring, shad,  striped bass,  and sturgeon.  The majority of the river is classified as SA, or saltwater suitable for
commercial shellfish harvesting.
                             White  Oak   River  Watershed
                                            North  Carolina
                                                                     •aven  County
Figure 1. General map of White Oak River Watershed in North Carolina.

    The White Oak River watershed has six major land cover/land use classes with wetlands encompassing the largest
single type at 52% of the total. Forests are the second largest land cover type constituting the majority of the headwaters
in the Croatan National and Hoffman State Forests (22%). A very small portion of the watershed is  urban (2%) and
agricultural  (11%) (NC Department of Environment and Natural Resources, 1997).

    Despite the low level of urbanization,  the North Carolina Division of Water Quality's basinwide management plan notes
an increase in shellfish closures in the river (North Carolina Division of Environmental and Natural Resources, 1997). At
state-sponsored public meetings, over 100 citizens expressed  concern and called for more public education on water
quality.
                                                    118

-------
    Recognizing the interest of their constituencies in water quality education, local NC Cooperative Extension Service
 leaders assembled a project team involving members from the North Carolina Division of Water Quality, North Carolina
 Division of Coastal Management,  North Carolina Division of Environmental Health - Shellfish Sanitation Branch, North
 Carolina Cooperative Extension, and 25 citizens who comprise the stakeholder-based Advisory Board for the White Oak
 River Watershed. This group includes  crop farmers, livestock farmers, fisherpersons, developers, foresters, tourism
 directors,  teachers,  scientists, and local government officials from the watershed (see Table 1). The citizen advisory
 board is the decision-making entity.  The government  agency representatives  and Cooperative Extension  personnel
 function as support  staff to the Board. Support staff provide resources, perform research and reviews, make reports,
 serve as technical advisors, and provide formal facilitation and  consensus-building  services.

 The White Oak Advisory  Board's Primary Issue of Concern

    The Board began meeting in August  of 1996. Their first task was to prioritize water quality issues upon which to focus
 their efforts,  Board members expressed concern that past bridge and road construction  across the mouth of the  river had
 contributed to a decline in water quality.  Furthermore, this road,  Highway 24, was slated for expansion, and they were
 concerned that this would exacerbate the problems. The Board acknowledged the need for expansion of the road, but
 recognized a unique opportunity to mitigate its impact if they  could move quickly  to work with the  North  Carolina
 Department of Transportation (NC-DOT)

    At the time that the Board was convened and identified the highway and its expansion as an issue, NC-DOT was in
 the process of conducting an Environmental Assessment of the project and were anticipating a Finding of No Significant
 Impact.  During a meeting between the Extension Project Team and NC-DOT, DOT representatives were made aware
 of the Board's concerns and expressed an interest in  working with the  Board to address those concerns. However, timing
 was an issue because in several months, NC-DOT was planning  right of way acquisition to begin the expansion project.
 Because of the urgency of the matter,  the Board resolved to meet twice monthly and work to develop their comments and
 recommendations.

 Technical Information  Gathered by the Board

    In response to the Board's inquiry, the Project Team reviewed, summarized, and presented scientificstudies that had
 been conducted on the river that related to sedimentation  and flow  patterns in the river and the possible effects of highway
 construction over the mouth of the river.  Results from the following four studies were especially useful in understanding
the science behind this policy issue.

Table 1.  White Oak River Watershed Advisory Board -  Stakeholder Composition.
Stakeholder Groups
Fishing, Commercial
Fishing, Recreational
Real Estate or Development
Environment/ Conservation
Farming, Crop
Farming, Livestock
Forestry, Private
Business & Industry
Local Government
Academia/ Public Schools
Travel & Tourism
NC Shellfish Sanitation
Soil & Water Cons.
Public Forestry
|Totals
Carteret














12
Jones














5
Onslow














5
At Large














3
Total
4
1
1
1
2
1
1
3
3
4

1
1
2
25
                                                    119

-------
    One study (Martens and Goldhaber, 1978) determined that the metabolic pathways by which bacteria degrades
organic matter  in the sediments  differ depending  on whether the overlying aquatic environment is  salty  or fresh.
Chemistry analyses done on soil cores taken  at various locations in the river found framboidal  pyrite at upstream samples,
which indicated that saltwater wedges had previously penetrated further upstream than current patterns in soil chemistry
showed. These results provided evidence to the Board that saltwater flows in the river had changed  over time.

    Adams, Benniger, Hosier, Overton, and Reed (1982) studied water circulation and sedimentation patterns in the White
Oak Estuary and found that sedimentation in the estuary varies from 0.3 cm/yr to 5 cm/yr. approximating the annual rate
of submergence along the Atlantic coast. Their study confirmed for the Board that the system is a flood tide dominated
system with sediment transport primarily occurring during  storms with strong on-shore winds. The study also noted that
the construction of the Intracoastal Waterway (ICWW) in 1930-32 in conjunction with the construction  of Highway 24 in
1933 altered  channel flow from one channel  (adjacent to Muggins Island) to another (adjacent to the mainland near
Highway 24).  Spoil deposition from dredging  operations may also be responsible for decreased channel flow in the west
channel of the Inlet. The authors noted no evidence of a  declining fishery based on the fact that it was comparable to
other fisheries in the area and in line with historical production rates for the estuary. This study also quantified the extent
of fill and alteration  to the estuary caused by  the original construction of the ICWW and the road  in 1932 and 1933,
respectively.  Historical maps, when compared to current data, showed that two inlets were closed and that more than
80% of the river was obstructed by these projects.

    Benniger and Martens (1983) investigated the  age and the sources of organic matter in  the estuary. This study
characterized the organic matter degradation rates, which is important  in understanding the estuary's capability to  process
organic inputs. The  researchers determined that the upstream organic matter inputs were primarily terrestrial and the
downstream organic inputs were primarily marine. However, they found that microbial processes acted preferentially to
remove recently produced organic matter. This implies that recently produced or partially treated organic matter could
substantially increase sediment  oxygen demand and the rate  of  nutrient  regeneration.   This would  increase the
vulnerability of the estuaries to anoxia and algal blooms.

    Kelley, Martens, and  Chanton (1990),  by collecting and analyzing sediment  cores,  characterized the  relative
remineralization rates of sedimentary  carbon for the fresh and saltwater environments in the river. They found that the
upstream environment, which is dominated by  terrestrial inputs and the process of methane reduction, remineralized at
a rate three times faster than the downstream  site, which is dominated by marine inputs and uses sulfate reduction as
the energy pathway for organic matter remineralization. As indicated in the previously described paper by Martens and
Goldhaber (1978), saltwater circulation patterns, as well  as freshwater inputs, appear to have changed such that the
estuarine ecology has shifted towards a more  freshwater system. Since  freshwater facilitates rapid remineralization of
organic carbons, this, over time, can reduce the river's buffering capacity and result in nutrient enrichment.

Initial  Conclusions of Board

    Based  on these and other  related  studies, the Board  concluded the following:

    . Salt wedges that used to extend upstream have not occurred in  recent history.

    . Organic inputs upstream  are from terrestrial sources and downstream are from marine sources.

    . Salinity regimes in the river are highly variable seasonally and spatially.

    . Salinity helps buffer the river from nutrient inputs.

    . Sedimentation at the mouth of the estuary was considered normal for coastal estuary systems.

    .  There was no evidence to support a perceived decline in the fishery.

    .  Increased fresh water inputs from the expanded impervious surface area related to the highway  expansion may
      have a  negative impact.
                                                     120

-------
    .  Higher salinity reduces concentrations of fecal coliform bacteria.

    .  Ditching and other means of moving water faster off the land causes problems with increased freshwater to the river
      as well as increased bacterial contamination in shellfish beds.

    .  There is a significant shellfish resource at the mouth of the river that has historically remained open.

    Next, the Advisory  Board convened a  panel  of specialists to discuss this  information  and potential mitigation
strategies.  The panel participants included:

    .  Dr. Larry K. Benninger, Geologist, University of North Carolina  at Chapel Hill (UNC-CH);

    .  Archie Hankins, Biologist,  NC-DOT;

    .  Tom Jarrett, Hydraulic Modeling, United States Army Corps  of Engineers (USCOE);

    .  Dr. Chris Martens, Marine Sciences,  UNC-CH;

    .  Dr. Paul Hosier, Biologist, UNC-CH;

    .  Dr. Rick Leuttich,  Sedimentary Geologist, UNC-CH;  and

    .  Howard Varnam,  Hydrologist, USCOE.

    The panel reviewed the scientific information presented  to the Board, and they agreed that there has been an impact
on the circulation and flushing of the White Oak River since the  construction of the causeway and the ICWW, but
quantification of those effects would require  intensive modeling that would take  a minimum of 1.5 to 2 years. The panel
felt that any action to increase circulation and salt water inputs to the river would have an overall positive effect on water
quality.  However,  the  best manner in which to  accomplish those  goals and  the  particular  effects  on fisheries,
sedimentation, or other water resource values would be difficult without modeling studies. The panel concluded that due
to changes in land use, hydrology in the watershed had been altered. As a result, runoff volume during storm flows has
increased.  This increases pollutant loading and increases erosion processes  during storms.  The panel noted that a
reduction in freshwater runoff would not have any significant effect  on  the diversity and density of species, but on their
distribution. This would have little effect on flora and fauna in the river,  but might improve water quality.  The  panel also
noted that the most effective strategy for protecting water quality is to involve all  of the communities impacting the system
and to implement  overall  land use planning in the watershed. It was  suggested that the group needed to define their water
quality goals and how they want to manage the river and watershed to achieve those goals.   Individual actions for
localized effects would require some  additional modeling and research to  determine the best  options.

    The  panel concluded with a list of mitigation recommendations listed below:

    .  Pursue a study of the river to determine what,  if any, actions  should be taken to improve circulation up and
      downstream of the highway.

    .  Examine options to manage stormwater in new and  existing developments.

    .  Pursue the maintenance of buffers along creeks and streams.

    .  Pursue stricter enforcement of sediment  and erosion control at  construction sites.

    .  Endorse, encourage, and facilitate the use of BMPs in forests and farms.

    .  Work to develop a  mechanism for watershed -  based or coordinated land-use planning to address all  of the
      suggestions.

    .  Explore alternative waste management strategies for both  single users and municipalities to reduce nutrients.


                                                     121

-------
The Board's Recommendations

    Over the next several meetings, and as a consequence of these findings, the Board recommended the following
actions;

    1.  To reduce freshwater inputs to the estuary and possible negative impacts of highway runoff on water quality, the
       Advisory Board recommended storm water runoff from bridge and highway expansion not be discharged into the
       river and that the Department of Transportation (DOT) explore options to eliminate discharge into the waterways.
       At a minimum, discharge from Highway 24 should be directed south (downstream) of the causeway to prevent
       impacts to shellfish. In addition, it was recommended that amelioration of the velocity, volume, and quality of that
       runoff be implemented, if feasible.

    2.  Historic maps showed that,  prior to the 1930's, the mouth of the White Oak River was open and unrestricted,
    allowing free tidal flow. In 1932 and 1933, Department of Transportation and US Army Corps of Engineers (ACOE)
    projects closed approximately 80% of the mouth of the river and altered physical processes. The Advisory Board
    recommended that to restore salinity  regimes, increase tidal  circulation,  and  reduce sedimentation,  DOT take actions
    to reopen the mouth of the river to the maximum extent possible. One option would be the creation of a north-south
    channel connecting the  estuary with the sound near the current location of the Flying  Bridge Restaurant on the
    Carteret  County  side of the  river  spanned by a bridge or  connected  by a culvert. Additionally,  the Board
    recommended that DOT  and ACOE  access ACOE  ecological restoration funds and collaborate with each other to
    mitigate the impacts of this expansion and past actions.

    3.  Since efforts to open the  channel would not remain  effective unless the State  of North Carolina initiates an ongoing
       maintenance program, the  Advisory Board recommended that a long-term maintenance  program supporting
       improved circulation,  reduced sedimentation, and  restored salinity  regimes be developed and implemented  by
       responsible  agencies.

    These recommendations were presented to and adopted by commissioners for Carteret and Jones Counties in May
and June  of 1997. In addition, the White Oak  River Watershed  Advisory Commission of Onslow  County (a group
appointed by the Onslow County Board of Commissioners to address water quality issues in Onslow County) endorsed
the recommendations of the Board at their May of 1997, meeting. This collaborative, consistent, watershed-based policy
statement became part of the  public record for the NC-DOT hearings in May  of 1997, and a preliminary draft was included
in the NC  Division of Water Quality's Basinwide Water Quality Management Plan for the White Oak River Basin (North
Carolina Division of Environmental and Natural Resources, 1997).

Response to the Recommendations of the  Board

    At  a joint meeting that included representatives of the  NC-DOT, the White Oak River Advisory Board, the  Extension
Project Team, and USCOE, the DOT agreed to support the Board's recommendations and revise their  stormwater plans
to direct runoff away from  the shellfish resource in the river.

    Blueprints for construction were redrawn reflecting the following features.  In the vicinity of the bridges  carrying
Highway 24  over the White Oak River at Swansboro, NC-DOT agreed, to the extent possible, to direct the stormwater
runoff from the roadway to the Bogue Sound side of Highway 24  and away from the river.  In  Swansboro (west of the
island causeway),  the existing stormwater collection system (which has outfalls on both the river and sound side  of
Highway 24) will continue to be used for the runoff following  roadway expansion, thus preventing the need for additional
outfalls in the river.

    From the island causeway  eastward for approximately  2.5 miles, the  stormwater runoff from the highway will be
collected and piped to outfalls on the sound side of Highway 24. Also, NC-DOT has designed special channelization
islands for commercial driveways to accommodate some of the stormwater runoff from the highway and bridges. These
water quality islands are depressed inside the curb to allow the first inch of highway runoff to pond within these islands
and filter through the grassed areas located there. The  filtered runoff from these islands is then collected and piped to
outfalls on the sound side of Highway 24.

                                                    122

-------
    In addition to the stormwater design changes, it was agreed that DOT would cooperate with other state and federal
agencies in any efforts to improve circulation and tidal flushing. Currently, the Board is continuing to work on adding a
section to the Congressional Water Resources Development Act that would authorize the USCOE to conduct the study
necessary to determine what,  if any, actions could be taken to improve flushing in the river.

Conclusions

    Local stakeholder-based citizen groups can impact  policies that affect their environment. Support for gathering,
summarizing,  and delivering technical  information  to local citizens and governments is an  important aspect of the success
of these processes.  Knowing who to  approach for answers to specific questions, and where to look  for scientific
information is an important function of the group's technical support. In addition, translating the information gathered into
digestible and usable material is also critical.

References

Adams,  David A.; Larry K. Benniger; Paul E. Hosier; Margery F. Over-ton; and James P. Reed. 1982. White  Oak River
Management  Project.  Office of Water Resources, NC Dept. of Natural Resources and Community Development, Raleigh,
NC.

Benniger, Larry K. and Christopher S. Martens. 1983. Sources and fates of sedimentary organic matter in the White Oak
and Neuse River estuaries. WRRI Rept. # 194.

Danielson,  L. E.,   1998. Are We Really  Prepared to be  Honest Brokers  and Conflict Resolvers in Controversial
Situations? Organized Symposia in Dealing with Controversy in Natural  Resource Issues, at AAEA Annual Conference,
Aug 4, 1998. Salt Lake City, Utah.

Kelley, Cheryl A.; Christopher S.  Martens; and Jeffery  P.  Chanton.  1990.  Variations in  the  sedimentary carbons
remineralization rates in the White Oak River estuary, North Carolina.

Martens, Christopher S. and Martin B. Goldhaber.   1978. Early diagenesis in transitional sedimentary environments of
the White Oak River Estuary,  North  Carolina. Limnol.  Oceanogr., 23(3), 428-441.

North Carolina Department of Environment and Natural Resources,  1997. White Oak  River Basinwide Water Quality
Management Plan. North Carolina Division of Water Quality, Raleigh, North  Carolina,  227  pps.

White, N. M., D. E. Line, J.D. Potts, William Kirby-Smith, Barbara Doll and W.F. Hunt. 1999. Jump Run Creek Shellfish
Restoration Project.  In review, Journal of Shellfish Research.
                                                    123

-------
                     Community Responses  to Stormwater  Pollution:
                 Case Study Findings with  Examples from  the Midwest
                                   George P. Aponte Clarke, Policy Analyst
                                     Natural Resources Defense Council
                                               New York, NY

                                           Peter H. Lehner1, Chief
                                      Environmental Protection Bureau
                                   New York State Attorney General's Office
                                               New York, NY

                                       Diane M. Cameron', President
                                            Cameron Associates
                                             Kensington, MD

                                    Andrew G. Frank3, Litigation Associate
                                 Paul, Weiss, Rifkind, Wharton, and Garrison
                                               New York, NY
    Stormwater runoff threatens the nation's waterways and public health, and costs Americans hundreds of millions of
dollars each year. Concerns about urban runoff and interest in proposed new federal Stormwater regulations prompted
the Natural Resources Defense Council (NRDC) to document existing, effective Stormwater strategies. Our report aims
to encourage municipal action and empower communities to address this critical issue. More than 150 case studies from
across the nation were compiled  and evaluated to highlight effective pollution prevention, administrative, and financing
strategies for addressing Stormwater runoff. The case studies show, on a practical level, that Stormwater management
can be environmentally effective,  economically  advantageous, and politicallyfeasible. The reportalsoforms  the foundation
of a comprehensive outreach effort. Together, they help guide communities as they implement or improve Stormwater
management programs by providing  detailed examples of proven tools and approaches  used to prevent Stormwater
pollution. Collectively, the case  studies offer an outline for further successful  Stormwater management strategies.
Elements critical to  the effectiveness of these programs include: a  pollution prevention  emphasis with structural treatment
measures when needed; a focus on preserving natural features and processes;  programs that inform and involve the
public; a framework that creates and maintains accountability; a dedicated and equitable funding source to ensure long-
term viability; strong leadership; and effective administration.  These  broad themes translate into a set of nine local actions
for addressing the technical, social, and political issues associated with Stormwater runoff. The case studies show that
following these actions will help communities form a  sound Stormwater policy.

    Key Terms: urban Stormwater runoff, impervious surfaces, pollution prevention, best management practices, diffuse
pollution, accountability.
'formerly Senior Attorney and Clean Water Project Director, Natural Resources Defense Council
2 formerly Senior Scientist, Natural Resources Defense Council
3 formerly Policy Analyst, Natural Resources Defense Council

                                                    124

-------
Introduction

   Currently, there is substantial concern about the impacts of urban and suburban runoff.  Pollution from diffuse sources,
including urban stormwater, is the leading source of contamination in the nation's waters (U.S. Environmental Protection
Agency, 1997a). Stormwater  runoff pollution is a particularly important issue since most of the population of the United
States lives in  urban and coastal areas. Water resources in urban and coastal areas are  highly vulnerable to and are often
severely degraded by stormwater runoff. Specifically, urban and suburban runoff is the second most prevalent source
of water quality impairment in the nation's estuaries after industrial discharges (U.S. Environmental Protection Agency,
1998b).

   Economic impacts are an important aspect of this concern. Even  a partial accounting shows that hundreds of millions
of dollars are lost each year through added government  expenditures, illness, or loss  in economic output due to urban
runoff pollution and damages  (U.S.  Environmental Protection Agency, 1998a). The ecological damage  is also  severe and
is at least as significant. In particular,  uncontrolled urban runoff contributes to hydrologic and habitat modification, two
important sources of river impairment identified by the U.S. Environmental Protection Agency (EPA).

   The polluted stormwater runoff problem has two main components: the increased volume and rate  of runoff from
impervious surfaces and the concentration of pollutants in  the runoff.  Both components are closely related to development
in urban and urbanizing areas (Booth and  Reinelt, 1993;  Schueler, 1994; U.S. Environmental Protection Agency, 1997b).
When impervious cover (roads, highways, parking lots, and rooftops) reaches between  10 and 20 percent  of the area
of a watershed, ecological stress becomes clearly apparent (Klein, 1979; Booth and Reinelt, 1993; Schueler, 1994).
Everyday activities can deposit  on  these surfaces a coating of various harmful materials. When it  rains or when snows
melts,  many of these pollutants are washed into receiving waters, often without any treatment.

   The deposition of pollutants and the increased velocity and volume of runoff together cause dramatic changes  in
hydrology and  water quality (Klein, 1979;  Jones and Clark, 1987; Booth, 1990; Galli, 1990; U.S.  Environmental Protection
Agency, 1997b). These changes affect  ecosystem  functions, biological diversity, public health, recreation, economic
activity, and general community well-being  (Bannerman ef a/., 1993; Novotny and  Olem,  1994; Haile et a/., 1996;
Carpenter et a/.,  1998). Urban stormwater is not alone in polluting the nation's waters. Industrial and agricultural runoff
are often equal or greater contributors. But the environmental, aesthetic, and public health impacts of diffuse pollution
will not be eliminated until urban stormwater pollution is controlled.

   While urban and suburban runoff continues to be a critical issue, there  is substantial evidence that the problems are
not intractable. Increasingly, communities  are recognizing the causes  and consequences of uncontrolled  urban runoff and
taking  action to control and prevent runoff pollution, often without any mandate. These  innovative communities are
realizing the environmental, economic, and  social  benefits of preventing stormwater pollution. However,  neither the  extent
of these efforts nor the specific actions being taken  have been well documented.

   There is also a growing interest  in proposed new federal stormwater regulations. Comprehensive stormwater regulation
is required under Section 402(p) of the Clean  Water Act. Since 1992, cities with  populations over 100,000, certain
industries, and construction sites over 5 acres have been required  to develop and implement stormwater plans under
Phase I of the National Pollutant Discharge Elimination System (NPDES) stormwater regulations  (U.S.  Environmental
Protection Agency, 1990). In October 1999, EPA is expected  to promulgate a  new  rule  requiring  municipalities  with
populations fewer than 100,000 people  located in "urbanized areas"  (where population density is greater than 1,000
persons per square mile) to develop stormwater plans. Under what is known as the "Phase M" rule, the EPA and states
will develop "tool boxes" from which the smaller local  governments can  choose particular stormwater strategies to develop
their stormwater plans  (U.S. Environmental  Protection Agency, 1998a).

   To address all  of these issues and concerns, the authors developed a study to examine, document, and disseminate
information  on environmentally effective and  economically advantageous  stormwater  pollution  prevention strategies.  The
study resulted in a report, Stormwater Strategies: Community Responses fo Runoff Pollution, that highlights some of the

                                                      125

-------
most  effective existing stormwater strategies from around the country (Lehner et a/, 1999). The report provides
substantial evidence that such programs exist and highlights a variety of innovative strategies actually being used. The
report also aims to provide guidance to communities addressing stormwater issues, encourage municipal action, and help
empower  communities to be involved in this critical issue. This paper summarizes the study and presents its primary
findings and recommendations.

Study Design  and Approach

   The study was exploratory in  nature,  with the  intent  of presenting  information on existing effective stormwater
management programs.  To achieve this goal, we collected examples of environmentally beneficial and cost-effective
stormwater programs from across the country. We compiled this information into the case-study-based report described
above. This information and report have become the basis for a comprehensive  outreach effort.

   The first step was to gather information on programs and projects by examining existing programs (several begun
under Phase I  as well as many that  started earlier),  reviewing literature,  contacting regional and local stormwater
management experts  and  researchers, and interviewing representatives from stormwater management or other local
government agencies. We gathered information on over 250 programs.  The information was then examined in detail and
narrowed  down to a set of case studies that demonstrated elements of success. Three fundamental criteria for selection
were used: environmental gains, economicadvantages, and  community benefits. Environmental gains  included  biological,
hydrological, or chemical improvements resulting from stormwater management. Economic advantages included cost
savings to the municipality or developers,  or increases in property values related to the pollution prevention  measure.
Community benefits  included  aesthetic or recreational enhancement,  administrative or institutional successes, or
community relations improvements.

   Seventy-seven programs and projects were selected as case studies for the final report. Another 88 programs were
annotated to  provide  additional programs/locations not fully evaluated for  the  report. The  case studies  represent
communities of all sizes, types, and  regions throughout the United States. To help ensure  accuracy, local experts or
people familiar with the program, called  "groundtruthers,"were contacted  to review the case studies and add information
from their own knowledge  and experience.

   The case studies were first organized geographically by dividing the United States into six regions based in part on
general rainfall patterns. Within each of the regions, case  studies were then further subdivided into five categories of
stormwater management measures including, (1) addressing stormwater in new  development  and redevelopment, (2)
promoting  public education  and  participation, (3) controlling construction site runoff,  (4) detecting  and  eliminating improper
or illegal  connections and discharges, (5) and  implementing pollution  prevention for municipal  operations. These
categories roughly parallel those measures that large municipalities address under existing Federal regulations (40 CFR
parts 122.26 and 123.25) and  small municipalities will address under pending Federal regulations (U.S. Environmental
Protection Agency, 1998a).

Case Study  Findings

   Through reporting over 150  examples of actual programs, the full report provides substantial evidence that stormwater
pollution can be reduced or prevented with proper planning and implementation in growing or re-developing areas. The
examples  presented in  the  report also demonstrate that if some  communities can measurably  and cost-effectively reduce
stormwater pollution, so can other communities and states (Lehner, et al., 1999).

The Five Categories of Stormwater Management Measures

   Individually, the case  studies  provide  detailed examples of substantial  water  quality improvement,  effective or
innovative stormwater control strategies  to protect the  natural environment, significant cost-savings, and  important
ancillary benefits to the community. The programs and strategies highlighted come from communities of all sizes, types,

                                                    126

-------
and regions. They include efforts by municipal agencies, developers, and community groups. In many cases, several of
these groups  worked together to create  win-win  outcomes. The case studies highlight  a variety of strategies for
addressing the five categories of stormwater management measures previously enumerated, and are described in more
detail as follows.

   Addressing Sformwaferin New Development and Redevelopment. By far the most important category of stormwater
strategies focuses on land  use and development. It encompasses a wide range of measures including  regional or
watershed planning,  buffers  and open space preservation, infill development, conservation design,  and the use of site-
specific structural and nonstructural treatment measures.  One of the best strategies a municipality or developer can
employ is to minimize the aggregate amount of new impervious surfaces.  For example, developers of the Prairie Crossing
project in Grayslake, Illinois, prevented runoff pollution and saved  money by using conservation design strategies. The
developers first reduced impervious cover by clustering  317  residences on  only 132 acres of the site, which left 80 percent
as open space. They then designed the developed area around a natural  drainage system consisting  of vegetated swales,
restored  prairie, and wetlands. Modeling indicates that this stormwater treatment drain system will remove approximately
85% of nutrients, metals, and suspended sediments and reduce peak flows by 68%. Eliminating curbs and gutters
resulted in savings of $1.6 to$2.7 million. The development is also  very appealing to homebuyers, with sales  comparable
to or better than conventional developments in the area  (see Lehner et al., 1999, p. 224).

   Promoting  Public Education and Participation. Individuals play  a key role in reducing stormwater  impacts both in their
own day-to-day activities and in showing support for municipal programs and ordinances. The  most successful highlighted
programs accomplished three goals: they educated the public about the nature of the problem, they  informed  the people
about what they can do to  solve the problem, and they involved citizens in hands-on activities to achieve pollutant
reduction or restoration  targets. One example of this success is in Minneapolis, Minnesota, where a decline in water
quality motivated the Lake Harriet Watershed Awareness Project. Monitoring revealed that lawn-care chemicals were a
significant contributor to the problem, which suggested focused education efforts.  In turn, the  project developed two
approaches: a volunteer master gardener program and the distribution of educational materials. Evaluation showed that
67% of watershed residents reported using the information presented and 30% reported a change in behavior. As a result,
concentrations of lawn-care pesticides have dropped by 50% or more since the program began (see Lehner et al., 1999,
p. 231).

   Controlling Construction  Site Runoff. The case studies demonstrate that effective construction site pollution prevention
is politically and economically feasible and can dramatically reduce pollution. The most effective programs rest on four
cornerstones laid in pairs: enforcement and education; erosion prevention and sediment control. However, the first and
over-arching necessity is a clear set of requirements.  For example, Herzog et a/. (1998) found that in Geauga County,
Ohio,  and St. Joseph County, Indiana, aggressive, widespread seeding and mulching reduced construction site erosion
by up to 86% and reduced phosphorus loadings by 80%. These measures can also benefit developers financially. They
found that homebuyers perceive these "green" lots to  be worth $750 more than comparable "brown" lots (see Lehner et
a/., 1999, p. 236). While existing programs employ a wide variety of erosion and sediment control practices, virtually all
successful strategies require proper planning and phasing of construction activities to minimize land disturbance.

   Defecting  and Eliminating Improper or Illegal Connections and Discharges.  Local governments have found that
identifying and eliminating illicit connections and discharges is a remarkably simple and cost-effective way to address
some of the worst stormwater pollution. The case studies show that two factors are critical to success of this  element of
stormwater programs: finding illicit connections and discharges, and enforcement. In Washtenaw County, Michigan, the
Huron River Pollution Abatement project resulted  in a 75% reduction in the river's fecal coliform levels in just 4 years. The
project focused on eliminating existing illicit connections and preventing future incidents through chemical storage surveys,
industrial  inspections,  water-quality  monitoring, public education, and  complaint and spill response. Over a six-year period,
the program dye-tested more than 3,800 facilities, after which 328 of the 450 illicit connections found were removed (see
Lehner et al.,  1999, p. 239).
                                                      127

-------
   Implementing Pollution Prevention for Municipal Operations. A wide range of municipal operations can affect
stormwater quantity and quality. The case studies reveal that some local governments have been able to manage their
municipal operations to reduce stormwater pollution. The municipalities highlighted have done so in a variety of ways
including reducing the use of harmful chemicals in the maintenance of municipal properties and vehicles, improving the
maintenance and cleaning of roads and stormwater infrastructure, and training staff in pollution prevention practices.
Several municipalities have taken these steps at their golf courses. For example, the Village Links Golf Course in Glen
Ellyn, Illinois, is  preventing runoff pollution by  incorporating  integrated pest  management, water conservation, stormwater
detention, native planting, recycling, and public outreach into its day-to-day  management. The golf course relies on both
mechanical  and biological pest controls and has  significantly increased natural areas. The course collects runoff from
nearby streets and neighborhoods in its system of ponds and spillways. These ponds provide approximately 60% of the
course's irrigation water, and the course itself  passively treats and filters all excess runoff from irrigation (see Lehner et
a/., 1999, p. 243).

Themes Common to Success Stories

   Collectively, over 150 case  studies present a clear model for success. Evaluation of the case studies revealed several
common  elements among the highlighted programs. We distilled those elements into the seven  broad themes listed  below
to help guide communities as they develop or improve stormwater programs. Since they are based on actual programs,
these themes form a solid foundation for successful programs.

   Preventing pollution is high/y effective and saves money. Pollution prevention measures dramatically and cost-
effectively reduce the quantity and concentration of pollutants "winding up"  in stormwater. Common pollution prevention
measures include reducing or eliminating the use of harmful products,  preventing erosion, reducing the amount of
pavement in new developments, and  changing  maintenance  practices.  In  highly urbanized areas, however, such
measures may  be difficult. In such cases, several communities have found treatment of runoff with structural measures
or retrofitting existing structures to  be  effective alternatives.

   Preserving and utilizing natural features andprocesses have many benefits.  Many communities and developers have
found strategies that rely on natural processes to be highly effective and economically advantageous. Undeveloped
landscapes absorb large quantities of rainfall  and snowmelt and vegetation  helps to filter out pollutants from stormwater.
Buffer zones, conservation-designed development, sensitive area protection,  or encouragement of infill development all
enhance natural processes.

   Educafingandinforming the general public and municipal staff improves program effectiveness. Providing information
and training to the general public and local businesses is a key component to many of the highlighted programs. Since
many sources of stormwater  pollution are derived from individual activities such as driving and maintaining homes,
educating the public goes a long way to reducing stormwater pollution. Several communities involve the public in civic
activities, such as monitoring water quality or  stenciling storm drains, which not only provide educational opportunities
but also save the municipality money.

   Strong incentives,  routine moniforing, and consisfenf enforcement establish  accountability.  Enforcement, or more
broadly  accountability, is a key element to improving water quality.  All actors need a clear statement of performance
goals, and they need to be held accountable by others for accomplishing these goals. We found that programs with high
accountability were the most effective, often achieving pollutant reductions of 50% or greater.

   Financial stability helps ensure effective programs. Effective stormwater programs are financially viable and affordable.
Dedicated funding sources, such as stormwater utilities or environmental fees are equitable ways to build  stability into
stormwater programs. Stability and equity are  also important in gaining public support. Nearly 200 communities across
the nation are already realizing the benefits of implementing stormwater  utilities  as dedicated and  equitable funding
sources.
                                                     128

-------
   Strong leadership is often a catalyst for success. Success, at least initially, often requires an individual to champion
the project and make it happen.

   Effective administration is critical. Regardless of which strategies  a community chooses, those programs with  clear
goals and objectives  are the most  successful. Such clarity enhances accountability,  responsibility, and trust.  Furthermore,
an established and understood institutional framework often improves administration by fostering collaboration among
different  parts and levels of government, neighboring communities, and  local citizens. Effective  administration allows
implementation of broad-based, multi-faceted programs, which are often the most effective at controlling the diffuse
problem  of stormwater pollution.

Authors' Recommendations for Local Action

   To further guide communities addressing stormwater runoff issues, we translated the broad themes presented above
into an action  plan based on nine key recommendations. These  actions  roughly parallel the broad themes presented
above. The case studies demonstrated that following the nine local actions outlined below will help build  a strong
framework for effective, efficient, and successful stormwater  management over the long term.

    1) P/an in advance and set clear goals. Carefully plan programs, as opposed to simply reacting to provided
       opportunities, crises, or transient pressures.  Planning allows development of more effective and cost-effective
       actions. An essential outcome of planning is addressing the issues and concerns of all stakeholders involved.

    2) Encourage and facilitate broadparficipation.  Program  planning, development, and implementation should involve
       multiple  levels  of government,  key members of the  community, and professionals from a variety of related
       disciplines. A key to success is the public's understanding  of the issue, how it relates to them, and what they can
       do about it.

    3) Promote public education opportunities. Implement broad-based  programs that reach a range of audiences and
       solicit different levels of public involvement. Remain committed to the education program and take advantage of
       existing community organizations to enhance participation.

    4)  Work to prevent pollution first; rely on  structural treatment on/y when necessary. Focus on prevention-based
       approaches,  through regional and  watershed planning,  local zoning ordinances, preservation of natural areas,
       stormwater-sensitive site design, and erosion prevention as these are significantly more effective than treatment
       of polluted runoff.

    5) Establish and maintain accountability. Essential components of this process are setting clear standards, creating
       strong incentives and disincentives, conducting routine monitoring and inspections, keeping the  public informed,
       promoting public availability of stormwater plans and  permits,  and consistently enforcing laws and  regulations.
       Strong enforcement is often key to significant water quality improvements.

    6) Secure  financial resources. Consider establishing  a dedicated funding  source such  as a stormwater utility.
       Combine with it budget-saving measures such as creative  staffing, public-public and public-private collaboration,
       and building  off existing programs.

    7) Tailor strategies to the region and setting. Recognizing that every case will be different, consider strategies that
       are particularly tailored to the region, the specific audience, and the problem.

    8) Evaluate and allow for evolution ofprograms.  Set clear goals and priorities, and allow programs to develop over
       time. Establish clear ways to check and see  that goals and objectives are being met. This opens opportunities
       for improvements and helps ensure long-term success.
                                                     129

-------
    9)  Recognize the importance of associated community benefits. Stormwater  pollution  prevention measures usually
        offer ancillary quality-of-life benefits in addition to targeted improvements. For example, preserved areas offer
        parks, ponds offer beauty and habitat, clean streets are more attractive, education helps empower people, and
        sediment control improves fisheries and prevents flooding.

Conclusion

    Many fine handbooks provide theoretical and technical guidance concerning the design and implementation  of
effective stormwater pollution prevention and control measures. This study took a different approach and focused on
existing  effective programs  in  a variety of settings.  In  doing  so,  it accomplished two  key goals. First, the study
demonstrates that stormwater management is quite possible. The case studies show on a practical level that stormwater
management can be environmentally effective, economically advantageous, and politically feasible. Second, the case
studies enable communities developing or improving stormwater programs to  learn from their peers. In doing so,  the case
studies offer an  outline for future successful stormwater management strategies.

Acknowledgments

    The authors extend their appreciation to all the communities, organizations, agencies, and individuals who  provided
information for this study; the technical consultants, peer reviewers, and groundtruthers for their helpful comments; and
the volunteers and Natural Resources Defense Council staff who assisted with this study. The study was supported  by
a grant from the EPA under Section 104(b)(3) of the Clean Water Act.

References

Bannerman, R. T., D. W. Owens, R. B. Dodds, and N. J. Hornewer.  1993. Sources of Pollution in Wisconsin Stormwater. Water
Science and Technology, 28(3-5): 241-259.

Booth, D. B.  1990. Stream-Channel Incision Following Drainage-Basin  Urbanization. Water Resources Bulletin, 26(3): 407-417.

Booth, D. B. and L. E. Reinelt. 1993. Consequences of Urbanization on Aquatic Systems-Measured Effects, Degradation
Thresholds, and  Corrective Strategies.  Proceedings, Watershed '93: A National  Conference  on Watershed Management,
Alexandria VA, pp.  545-550.

Carpenter, S.  R, N.  F. Caraco,  D. L. Correll, R. W. Howarth, A. N. Sharpley, and  V. H.  Smith.  1998. Nonpoint Pollution of
Surface Waters with Phosphorus  and Nitrogen. Ecological Applications, 8(3): 559-567.

Galli, J. 1990. Thermal Impacts Associated with Urbanization and Stormwater Management Best  Management Practices: Final
Report. Metropolitan Washington Council of Governments, Washington, DC, 157  pp.

Halle, R. W., James Alamillo, Kevin Barrett, Ron Cressey, John Dermond, Carolyn Ervin, Alice Glasser, Nina Harawa, Patricia
Harmon, Janice Harper,  Charles McGee, Robert C. Millikan, Mitchell Nides, John S. Witte, 1996.  An Epidemiological Study  of
Possible Adverse Health Effects of Swimming  in Santa Monica Bay. Santa Monica Bay Restoration Project, Santa  Monica, CA,
70 PP-

Herzog, M., J. Harbor, K. McClintock, J. Law, and K. Goranson. 1998. Are Green Lots Worth  More than Brown Lots? An
Economic Incentive for Erosion Control on Residential  Developments. Journal of Soil and Water Conservation,  Accepted.

Jones, R. C. and C. C.  Clark. 1987.  Impacts of Watershed Urbanization on Stream Insect Communities. Water Resources
Bulletin. 23(6): 1047-1 055.

Klein, R.  D. 1979. Urbanization  and  Stream Quality Impairment. Water Resources Bulletin, 15(4): 948-963.


                                                     130

-------
Lehner, P. H, G.  P. Aponte Clarke, D. M. Cameron,  and A. G. Frank.  1999. Stormwater Strategies: Community Responses
to Runoff Pollution. Natural Resources Defense Council,  New York, N.Y., 269 pp.

Novotny, V.  H. and Harvey Olem.  1994. Water Quality: Prevention, Identification, and Management of Diffuse Pollution. Van
Nostrand Reinhold, New York,  1054 pp.

Schueler, T. R. 1994.  The Importance of  Imperviousness. Watershed Protection Techniques l(3): 100-1  11.

U.S. Environmental Protection Agency. 1990.40 CFR Parts 122,123, and 124, National Pollutant Discharge Elimination System
Permit Application Regulations  for Storm Water Discharges; Final  Rule. Federal  Register, U.S. Government  Printing Office,
Washington, DC 55(222): 47992-47993.

U.S. Environmental Protection Agency, 1997a. Nonpoint Source Pollution: The Nation's Largest Water Quality Problem.
Accessed December 21, 1998, at URL http://www.epa.gov/OWOW/ NPS/facts/point1 .htm.

U.S. Environmental Protection Agency. 1997b. Urbanization and  Streams: Studies of  Hydrologic Impacts, Office  of Water,
Washington, D.C., 841  R-97-009, 15 pp.

US.  Environmental Protection Agency.  1998a. 40 CFR Parts 122  and  123 Part  II,  National Pollutant  Discharge Elimination
System-Proposed  Regulations for  Revision of the Water Pollution Control Program Addressing  Storm  Water Discharged;
Proposed Rule, Federal Register, U.S. Government Printing Office, Washington,  D.C.,  63(6): 1536-1 643.

U.S. Environmental Protection Agency. 1998b. National Water Quality Inventory: 1996 Report to Congress,  EPA841 -R-97-008,
521 pp.
                                                       131

-------
                       Integrated Urban Stormwater Master Planning

                                         Eric Strecker, Vice President
                                        URS Greiner Woodward-Clyde
                                             Portland, OR 97201

                                                    and

                                     Krista Reininga, Sr.  Project Engineer
                                        URS Greiner Woodward-Clyde
                                             Portland, OR 97201
Abstract
    Urban stormwater management agencies are increasingly being called upon to address water quality and natural
resources issues in addition to their traditional focus on flood conveyance. In response to this, stormwater drainage
master plans have been increasingly addressing stormwater quality and, in limited cases, natural resources and habitat.
This paper will describe some  of the problems with traditional stormwater master planning approaches, including those
where water quality and natural resources have been included as "add-ons," and the urban stormwater problems we are
now trying to address which have resulted from these approaches. A framework for how communities can develop
integrated stormwater master plans that address  multiple objectives,  as increasingly  mandated  by public concern as well
as by regulations,  will be  presented. Given that  the tools available for master planning are not equivalent in their
numerical evaluations, new procedures and  project approaches are required.   Especially important is  how the
hydrology/hydraulic  methods are performed, including both flood evaluations and evaluation of the smaller channel-
forming  storms.

    Communities are often not institutionally organized to address multiple objectives. Master planning has traditionally
been led and performed by engineers trained in hydrology and hydraulics, and they are usually in different departments
from those who are responsible for other environmental aspects of the drainage system. This paper will focus on the
technical, institutional, and process-oriented aspects of how master planning can be  improved. Several case studies from
the Pacific Northwest of the United States will be discussed.

Introduction

    The purpose of this paper  is to discuss some of the attributes of urban stormwater master planning and how those
master plans can be improved to more fully address issues besides conveyance capacity and flood control. Stormwater
master plans go by a number of names, including storm drain master plans, stormwater infrastructure plans, and urban
catchment management plans. These plans are usually very  focused on flood control and, until just recently, address
water quality minimally. This paper will discuss some of the attributes of traditional urban stormwater master planning
and its results,  regulatory programs (which in the US and New Zealand are requiring adifferent  approach), how integrated
master  planning can  be accomplished, and institutional barriers  which often prevent  integrated master planning from  being
accomplished.  In this paper, an Integrated Stormwater Master Plan is an infrastructure and management plan that not
only addresses flood control and property protection issues, but also considers stream stability and habitat, along with
water quality and aesthetics.

Urban Stormwater Drainage  Problems

    It has long been recognized that in urban areas, unplanned stormwater management systems result in damage to
property and sometimes people. As it will be well demonstrated by other papers in these proceedings, urbanization of
watersheds and the resulting impervious areas also cause  changes to the  hydrology  and water quality of receiving waters
which ultimately result in other impacts to aquatic life and humans. Even some of our measures to control impacts can
have unplanned detrimental effects. Especially sensitive to these changes are stream systems and coastal embayments


                                                    132

-------
that are not well flushed. Almost always there are also direct impacts to stream riparian areas which also increase these
changes through canopy removal and channel modifications.

    Urbanization usually includes impervious areas directly connected to efficient stormwater conveyance systems
(including roof drains and driveways connected to streets and  curbs to inlets to pipes) which then are discharged to
streams directly or through engineered channels. This  has resulted in stormwater being conveyed as fast as possible
to receiving waters (and away from properties). Increasingly, it is being recognized that because stormwater is drained
to streams in this manner, small storm  hydrological changes that  result in increased  runoff flows can significantly increase
the frequency and duration of elevated flows. This energy change within the normal wetted channel often results in
channel cutting, widening, and/or sedimentation, which in turn can  cause severe habitat and water quality degradation
(MacRae 1996; Severn and Washington, 1996). Often to "fix" these  channel problems, streams are enclosed, hardened,
and/or straightened. Even without considering the water quality of stormwater, our stormwater systems are severely
impacted from a physical habitat standpoint, including habitat loss, higher velocities, and temperature changes. Figure
1 shows an example of how stream runoff can change with urbanization, including much higher and peaky flows as well
as increased volumes of runoff.
                                      Rainfall and  Runoff
0.9
0.8
0.7 •
£0.6
!rb.5 •
|o.4 •
CC
0.3 •
0.2
0.1
n n .
* MB Rainfall (in/hr)
I |\ — • Post Dev (c=.95)
j« \\ - - Pre Dev (c=.30)
n P ji 	 Post Dev (with discharge control)
I |^ t h t i
i PN\
i \/\; \
I i^l u ^
' i / * ]i \
-. \ ' •• v i
;:\/-A- *. '- •' •.' l \
• '. ' ' V v . • \ , I ^v
•^ '. I \
0.09 -L \ \
JL.
^^^^^g^^^^^^^^^^^^^^^^^^ 1 1 1 t 1 1 1 1 1 1 1 1 1 1 	 Ill
0.08
0.07
0.06
0.05
0.04 -g
0.03 0
E
0.02
0.01
0
-0.01
                                                             00
                                                             CM
             CM  •* (O 00  O
             CO  CO CO CO  "*t

Hours
Figure 1.  Example Schematic of Changing Rainfall/Runoff Relationships with Development.

        With  urbanization also comes a dramatic change in water quality. Urban stormwater systems are the efficient
conveyance system of urban pollutants, both those discharged during storm events and those occurring during dry-
weather discharges. There are numerous ways that pollutants enter stormwater from those in the rainfall itself to
commonly thought of sources such as street dirt and car drippings. Stormwater often exceeds US EPA water  quality
criteria.  Figure 2 is a graph of the frequency that stormwater runoff from identified land uses in Oregon exceed US EPA
acute dissolved metals water quality criteria in runoff from identified land uses (Strecker et al., 1997).  It should be noted
that most of this runoff was measured in pipes, while the criteria are meant to apply to receiving waters. A data set of
flow-weighted  composite samples (representing average storm concentrations) from over 40 land use stations from
various  areas  of the Willamette Valley was  utilized to  develop the information displayed in the figure. The stations
included an open land use station in an urban area (Forest Park in Portland) for comparison. Note that dissolved copper
and zinc in developed land uses exceeded criteria for 30 to 65% of the storm events. Similar findings have been found
in other programs, including the San Francisco Bay area programs (Cooke and Lee, 1993).
                                                     133

-------
                                                                         •   Open
                                                                         •  Residential
                                                                         •   Commercial
                                                                         D  Transportation
                                                                         U   Industrial
                    Cadmium  Copper      Lead
Zinc
                       Based  upon Oregon  NPDES Stormwater Monitoring  Data
                       Compiled by ACWA.  Developed areas:  27 to 67 storm
                       events; Open space: 9 storm events

Figure 2. Frequency that Flow-weighted Composite Urban Stormwater Runoff Samples Metals Concentrations Exceeded US EPAs Acute Criteria for
Aquatic Life.

    The water quality impacts together with the physical hydrology changes described  above have caused our urban
stream systems to become severely degraded.  Our traditional systems have not protected the resources nearly as well
as they have protected property. As many have now recognized, at about 10 to 25% imperviousness, the health of the
aquatic system is severely degraded (May et al.,  1997; Schueler,  1994). In many cases because of the  longer-term
channel stresses, property has been damaged as well, including under cutting of headwalls, etc. The plans typically only
identified solutions that solved large flooding  problems, sometimes just temporally  until what has  been  considered
"maintenance" problems such as  head  wall failures, occur.

Environmental  Concerns and Regulatory  Requirements

    In the US, Congress has recognized that urban Stormwater plays a major role in affecting receiving waters when it
mandated in the revised Clean Water Act that urban Stormwater water quality be addressed through  a permitting
(consent) program. New Zealand has similar requirements through its Resources  Management Act of 1991. Both these
programs are still evolving. The Stormwater permit program in the US specifically requires that larger cities (over 100,000)
and soon smaller cities address Stormwater quality issues as they conduct flood control projects.  New Zealand's  program
also requires that municipalities obtain  consents for Stormwater discharges.

    Under the overall program, one area that  has been slow to change is how urban Stormwater master plans are
developed  and implemented.  Although there are  requirements to consider water quality  in  conducting flood control efforts,
for a number of reasons (including  institutional inertia) agencies have  been somewhat slow in actually giving water quality
and habitat protection equal weight with flood control in master planning.  Some of this is due to the fact that Stormwater
master plans are typically the  responsibility of engineers who are experienced in hydraulics, but that often  lack experience
and knowledge in other aspects of environmental Stormwater management. To be fair, engineers have been told to plan
for managing Stormwater based upon land-use zoning that was selected without considering Stormwater issues. Another
major issue is the resources allocated to conduct integrated planning efforts which are more expensive; often agencies
do not recognize the value of better up-front planning compared to capital and maintenance costs.

    Increasingly though,  the  public has started to demand that more environmentally sound and/or aesthetically pleasing
Stormwater management approaches be utilized.  For example, with the endangered species act (ESA) listings and
proposed listings of salmon  and  trout  species in the US Pacific Northwest,  many  neighborhood organizations are
                                                    134

-------
pressuring municipal agencies to change their stormwater management approaches. Some of these efforts are having
more success than the regulatory programs.

Stormwater Management  Agency Functions

    Understanding a stormwater  management agency's function and history is important to understanding its approach
to stormwater management.  Stormwater management agencies typically fulfill the following roles:

    1. Stormwater System Maintenance

    2. Development Standard

    3. Stormwater Master Planning

    4. CIP Design and Construction

    5. Funding-Utilities/System Development Charges

    6. Stormwater System Permitting and Environmental  Impact Minimization

    7. Education

    The last two elements are the most recent. Many agencies began by responding to emergencies and problems, and
were then tasked to develop onsite design conveyance standards. Stormwater master plans for the most par-t were
developed in response to problems that arose after watersheds were developing with little or no stormwater planning.
They also were typically  focused on just flood control and property protection. Most often they focused on the piped
systems and road culverts. Often creeks away from culverts were not evaluated unless there had been a particular
problem  identified. In the  US, the  Federal Emergency Management  Agency (FEMA) had  separately developed  flood plain
maps for larger systems,  which communities relied upon for protecting structures from larger river and stream flooding.
This was done to meet requirements for participation in FEMA's flood insurance program. Therefore, flood  plains and
the creeks themselves have not been a focus of master plan (e.g., creek sections were typically not evaluated to a great
extent).

Stormwater Drainage Master Plan Goals  and  Results

    The traditional purposes of the Stormwater  Drainage Master Plan were to:

    . Guide a city's stormwater drainage system capital improvement project (CIP) program, (e.g., identify, select, cost,
     and prioritize stormwater system  construction projects.)

    .  Establish a maintenance program for the stormwater system (recommended stormwater system maintenance
     practices and frequencies)

    • Establish onsite conveyance requirements (design standards for level of peak flow conveyance by an engineered
     stormwater system and, sometimes, requirements  for street conveyance  of  stormwater  beyond the onsite
     requirements)

    Master plans seldom included requirements  for development with regard to  stormwater system impacts (e.g.,
downstream flow and/or water quality impacts). Master plans were sometimes utilized  to  assess potential future problems
as well as to fix existing  problems. Often systems were evaluated under current conditions and future planned zoning
to be able to assess costs to current rate/tax payers or new developments. Because master plans were not usually
completed prior to some significant  level of development, attributing these costs was important to the development
community as well as to  the residents.
                                                   135

-------
    The traditional approach to stormwater master  planning  has  been to focus on  hydrology and hydraulics of the existing
stormwater systems, and proposed larger trunk systems to determine whether there is enough capacity. This is usually
is accomplished by the following steps:

    . Route a designated large  storm through system, assume worst case conditions (saturated,  etc.) and determine
      capacity  deficiencies

    •  Develop an enlarged (or more efficient) system to handle larger flows or, when necessary, reduce peak flows by
      detention (if the cost of detention is less than a conveyance upgrade)

    .  Sometimes consider water quality as a "add-on" (e.g., if detention is required, claim a water quality  benefit)

    This approach has certainly significantly  reduced property damage (sometimes only for short-term), but has  led to
more  damage in streams. The damage  has  been a result of a significant increase and duration in small  storm  runoff
flows. The result of not planning for this  increased energy, which is primarily contained within the stream channel, has
often  been an increase  in maintenance  and  property damage. For example, channel cutting that occurs  upstream of
culverts often causes headwall and culvert failures. In other areas where channel cut sediments settle out (often in over-
designed or poorly designed culverts), areas are filled in with sediments. When this occurs (especially in a culvert), it can
lead to flooding. These problems (headwall failures, culverts filled  in, etc.) are often called maintenance issues,  when
they are in fact really failures of the master plan to adequately address stream impacts of development.

    Typically smaller urban stormwater systems (e.g., 10 to 50 acre catchments)  are dominated from a flooding
standpoint by shorter-duration, more-intense  storms (thunderstorms), whereas, the larger urban watersheds are often
impacted by larger,  but less-intense storms of longer duration.  Master plans typically utilize a  single large design  storm
event based upon a rainfall depth  (mm of rain over a watershed) for a specified duration and  return period. This depth
is then assigned a conservative  shape such as the SCS type IA shape shown in Figure 3). The storm shown is the 25-
year, 24-hour storm depth for Eugene, Oregon, with the  SCS distribution applied to it. As an  example of how  overly
conservative the peak of
                                   25-Year SCS Type lASynthetic Design Storm
                                                              Totd Rartai Volume = 5.28 in
                                                             Max Hourly Ranai Intensity = Q9 irVhr
                                                      Time (hi)
Figure 3. 25-Year, 24-Hour SCS Type 1A Synthetic Design Storm for Eugene, Oregon.
                                                     136

-------
the "design" hydrograph is, Figure 4 shows an actual 25-year storm hydrograph (based upon analysis of the Eugene
Airport rain gage). This storm  was confirmed by  long-term simulation modeling to have caused approximately the 25-year
return-period flows in the larger stormwater systems in the city.  In reality, the 25-year  return-period storm depth seldom
if ever arrives with the peaky "shape" given it in master plans.
                         National Weather Service Rainfall Data for February 5-8,1996
                                                              Total Rainfall Volume = 7.3 in
                                                            Max. Rainfall Intensity = 0.66 in/hr
                                                   Time (hours}
Figure 4. Rainfall Event that was Considered to Cause the Approximately 25-Year Return Period Peak Runoff Flows in Eugene, Oregon.

    Many have justified this shape as being one that will also allow flood control effects of smaller thunderstorms on the
smaller stormwater systems to be  adequately evaluated. When the peak is modeled in this fashion on a larger watershed
during an already large rainfall, the peak may greatly affect the larger system design. This conservative design approach
we  believe has  led many  communities to determine that streams are undersized and must be widened, channeled, and/or
piped.

    Of course in communities where there is the potential for combined phenomenon to cause severe flooding (e.g.,
snowmelt and frozen ground combined with  a hard rain), there may be good reason to over size facilities. However, in
most cases, it may be more appropriate to utilize methods that account for this and to  strive to preserve open channels
in more natural ways  (e.g., larger stream buffers) to the extent possible.

    Another assumption  that is often  made is that  the watershed is saturated  before the design storm arrives. This
assumption is made to be "conservative." However, it results in an uneven level of conservatism. This assumption would
tend to lead to the most over-designed conveyance systems in the  least paved areas. That is, the saturation assumption
would  tend to  make  systems most over-designed  in low-density, single-family areas vs.  less over-designed  in the
downtown core area. The point here is that the levels of over-design are not consistent, nor targeted to the areas where
the  greatest level of protection is  desired (highest property value).
                                                      137

-------
    Finally, water quality, and sometimes habitat, is only now being considered in master planning. This is most often
accomplished by adding water quality to a detention feature or specifically selecting and locating several demonstration
water quality projects. Some communities have chosen to not emphasize habitat by engineering their streams with the
purpose of providing flood conveyance as recreational amenities. The Denver area is a good example of this type of
design. In arid areas, where streams are seasonal or even just storm driven, this may be a good choice for communities.
However, some communities are considering the value of seasonal  streams play for downstream  resources from a
biological  and water quality perspective. For example  in Eugene, Oregon, the city has determined that  seasonal streams
contain a  rich fauna of aquatic invertebrates (WCC,  1995) which likely would benefit the health of downstream systems.

    There are a number of reasons why the above approaches have continued to be employed. First, planning that
considers multiple objectives is much  more difficult to accomplish, from the technical approaches, due to the need to
involve more parties in decision making.  The traditional technical approach described above is straightforward, while
design of more natural systems is not (e.g., pipe flow equations are much easier to utilize then open channel flow in
natural streams). In addition, there are many more people to involve in making  decisions than dealing  only with
engineered physical structures within the stormwater system.  Second,  most municipalities are not organized well for the
purpose of urban watershed planning. The City of Portland, Oregon (which has been very progressive in  many ways)
still has four separate  departments (all  in one bureau) that do:  1) facilities planning (stormwater  system  master planning),
2) site stormwater standards, 3) stormwater quality (permit compliance), and 4) watershed management. Each of these
groups has developed  its own  plans and programs that have understandably not been  very well-coordinated or integrated.
Finally, and probably most important, is that integrated planning studies cost significantly more (on the order of 2 to 4
times as much).

Integrated  Storm Drainage Master Plans - Approach

    The new approach to stormwater master plans is the integration  of flood control, water quality, natural resources, and
aesthetics of stormwater systems. This approach requires significantly  more effort and should be thought of as one that
will entail  adaptive management. That is, the master plan  must include components that allow for changing conditions
as development occurs and the downstream systems react.

    In completing a stormwater master plan, it is difficult to achieve "maximums" of flood control, water quality, natural
aquatic habitat, and aesthetics.  It is somewhat analogous to the rule that it is hard to get a  cheap price, good service,
and high  quality. It is our belief that one of the problems with  master plans has been a lack of recognition that streams
will change and that the plans should  be developed to manage change in  a positive fashion.

    One of the keys to successful integrated master planning  is that the planning approach places the proper emphasis
on the technical and decision-making processes employed. As  mentioned above, master plans typically have been driven
by the hydrologic/hydraulic modeling of large storm(s) and  usually begin with model data collection and analysis. Figure
5 presents a suggested flow diagram for an alternative way of sequencing  the development  of a master plan.  It begins
by conducting an  inventory of  all aspects of the stormwater  system, including all attributes related to  the multiple
objectives mentioned above. The approach suggests utilizing  multi-disciplinary teams to review conditions in the field to
look for opportunities for meeting objectives, as well as reviewing existing and suspected future problems. Next, before
any modeling  is done, the project team  and decision-makers should utilize the collected information to develop goals  and
objectives  for the  plan. Then additional  technical analyses, including where and what type  of detailed hydrologic/hydraulic
modeling is appropriate, can be  decided upon based upon these objectives. We have found this approach sharpens the
focus of modeling so that the model is not "driving" the master plan into solutions that focus primarily on  conveyance
upgrades.

    In developing an integrated master plan, it is generally understood that the right mix of multi-disciplinary  technical
specialists should be involved. In addition, it is important to involve the "right" decision-makers and stakeholders early
in the process.  It is also important to  agree up-front upon the decision-making  process  that will be  utilized. We have
found that utilizing an agreed upon set of factors to evaluate, select, and rank projects is very useful not only for guiding
the process more objectively, but also  to serve as a history of why certain projects were recommended and why others
were not. This is very useful for future  decision-makers for two reasons. First, when questioned by others, there will be


                                                     138

-------
        , Flooding
      Water Quality
         Natural
        Resources
                                                                     NPDES
                                                                Natural Resources
                                                                    Permitting
  Problem/
  Opportunity
Identification
   Team
                     Evaluate
                     Flooding
                    Water Quality
                  Natural Resources
                    Alternatives
                    Evaluation and
                    Selection
                      Surface  Water
                     Master  Plan/CIP
                         Program
- updated
 inventory  of
 problems  and
 opportunities
-  problem
 definition
                                                       modeling
      -  gather/assess
      information
-  goals, objectives,   . fina| criteria
  and  criteria
 options for
 resolving problems
 -  options for
  enhancements
- apply criteria
 to alternatives
                                                            - selected  Capitol
                                                            Improvement
                                                            Projects  (CIPs)
                                                            - Multiple and
                                                            single  objective
                                                            solutions
Figure 5. Suggested Integrated Stormwater Master Plan Project Approach.
some backing for why certain decisions were made. Second, as conditions that affect the factors change, selection of
projects can also change in a  logical fashion.

    The approach we recommend is to evaluate solutions that are primarily single objective rather than those that are
multiple objective. That is, a two-stage decision process is employed to make sure that good single-objective solutions
are not ignored because of the multi-objective nature of the factors. The factors employed include:
    . Addresses flooding problems

    • Addresses water quality pollutants of concern

    . Meets  community  amenity objectives

    • Habitat value

    . Life-cycle costs
                                   Meets regulatory requirements

                                   Implementability

                                    Reliability/sustainability

                                   Other environmental impacts

                                   Equability
Integrated Storm Drainage Master Plans - Hydrology

    Integrated stormwater master planning includes evaluating  and considering  smaller storm hydrological impacts.
Figure 6 presents a storm-depth frequency curve for Portland, Oregon. The figure  demonstrates that storms of a depth
of 1.5 inches and less dominate both the number of storms (more than 95%) and the volume of runoff (over 90%). It is
the smaller storms of about 0.3 to 0.8 inches in depth that change the most in their characteristics. In natural areas of
the Northwest, these often  did not result in  appreciable runoff or  resulted only in slightly elevated flows for a long duration.
However, after urbanization, these storms are causing severe and  rapid changes in flow levels with  each storm. This kind
of analysis can be used to assist decision makers  in deciding what level of water quantity and water quality control is
going to be the most cost-effective in  reducing the impacts of urbanization.

    The best hydrologic and hydraulic modeling approach for assessing and designing stormwater systems is likely the
use of continuous simulation models using long-term rainfall records to evaluate a system under a wide range of varying
hydrologic conditions. However, this is quite expensive. One of the approaches that we have been taking is to utilize
long-term simulations of stormwater systems to select design storms. We believe  that this improves the consistency in
providing design storms that are closer  to the level of protection that is being "advertised," without having to run long-term
simulations. This approach involves using real rainfall data with continuous simulation models  (e.g., SWMM) to define
the resulting return frequency of runoff peaks in various parts of the stormwater system. Then, real storms are selected
                                                      139

-------
                                     Portland Rainfall Statistics
                     
                     o
                     .C1.5

                     Q)

                     3
                     0>
                     00.5
                     E
                     a
                     V)
                   95% of ev<
                                                                   90*/o of
70%of,
    Fits
                         0%    10%   20%   30%    40%   50%   60%   70%
                                          %  of annual runoff volume
                     80%
9 0 %  1C?
Figure 6. Cumulative Storm Event Rainfall Depth Analysis for Portland, OR Airport

that resulted in the return period of interest (using a partial duration frequency analysis). These "real" storms are then
utilized to design the system. Figure 7 presents an example  of partial-duration frequency evaluation of peak flows in one
of the basins in the Eugene, Oregon  area. From this frequency distribution, the storm that was closest to the intended
design level (25-year)  was selected for design analysis of the system.  Figure 8 shows a similar analysis for another basin
in Eugene, along with the design flows from an earlier master plan (which utilized the traditional SCS storm method with
saturated conditions).

    What Figure 8 demonstrates is that in this basin, the more traditional approach would have resulted in what is likely
a significant over-design of the system. In  most  basins, this was found to be the  case. However, there were several
basins that were close and a few where the real  storm  approach resulted in larger designs. Figure 9 and 10 compare
the resulting designs in the Flat Creek basin in Eugene.  Note that the real storm approach resulted in fewer and smaller
projects in this basin. This means that the city can utilize more of its scarce resources to complete other types of multi-
objective projects. One of the advantages of the use of real storms is that the concept is very easy to communicate to
citizens.  In  addition, the city has found that some of its channels are over-designed compared to the stated level of
protection, and that they may be able to relax vegetation maintenance requirements to allow for more natural channels.
Overall, the city is finding that allocating sufficient  resources to conduct an integrated plan will likely lead to a more cost-
effective program overall, in terms of multiple benefits.

Integrated Storm Drainage Master Plans - Water Quality

    There are a number of stormwater quality models and  approaches (Donigian  and  Huber, 1991). Some  are quite
simple and straightforward,  while others are much  more complex. In general, water quality models currently cannot
accurately predict  how pollutants get into stormwater.  Although some researchers have  made great strides  in establishing
sources of pollutants in the urban environment  (Pitt,  1993), there still are numerous pollutant sources that are not fully
understood.  Most models rely on either some land-use-based concentrations to drive water quality predictions or they
use a build-up/wash-off function to describe pollutant concentrations  (Donigian and Huber, 1991).
                                                     140

-------
                                         Exanpte Frequency Distribution of Peak Flows
                                           Bethel Danebo Basinette (Area=687 Acres)
                              1000
                                                                                            Upper 95% Orf tierce Lint
                                                                                            Special Fbak Row (cfs)
                                                                                            Lover 95% Cbrfidence Lint
                                                          10
                                                     FHlFNPBKr)(yr)
Figure 7. Example Frequency Distribution of Peak  Flows in Eugene, OR.
1000-,
PEAK FLOW (cfs)
5 8
Peak Row Comparison
Santa Clara - River Road Basinette (Area = 301 Acres)









«












hr^r- •










-_ r~* j










)
^A




L J J .1




,
<-"^



-


1 •



~>->






— .^ . ^
( 	
-j^i^^^TTTT


. , ,








. — X
*- * ^" _












_












TT^





_ i. J J J






, . , .
'T^Tvr

• i i i
, , , ,
, , , ,
, , , ,


^«-ny e-j • •
	 Lpper 95% Confidence umt
— • — Bcpected Fbak Row (cfs)
, 	 Lower 95% Confidence Unit
_-*-- 1990 Master Ran

1 10 100
RETURN PEROD(yr)
Figure 8. Peak Flow Comparison in Urban  Runoff from Eugene, OR.
                                                                141

-------
Flat Creek Drainage System
Capital Improvements Proposed in 1990 Master Plan
Using lo-year SCS Synthetic Design Storm
City of Eugene


	 Open Channel
— Pipe or Cukert
• Transition Between Section

CIP Proposed in 1990 Master Plan
® 30" RCCP replaced by 4'x3' RCBC
® open channel replaced by 2-36" RCCP
® 30" RCCP replaced by 2-3'x3' RCBC
0 existing channel expansion
(S) existing channel expansion
® existing channel expansion
® 3-43"x27" CMP replaced by 2-6'x3' RCBC
® 2-36" RCCP replaced by 2-4'x3' RCBC
® 58"x3B" and 24" CMP replaced by 2-5'x3' RCBC
Total
tU r^
£y \
rJ14h,,f™-l/^Sl ' \
Enlarged Area
Cost
$27,300
$175,400
$31,800
$9,100
(3,900
121,000
$87,100
$81,100

$553,800 c; 	 n
i iguit ;
Figure 9. Proposed Conveyance System Improvements Utilizing the SCS Type 1A Synthetic Design Storm and Assuming Saturated Conditions.

    The build-up/wash-off of suspended solids (TSS)  is modeled  and  then TSS concentrations are utilized to predict other
concentrations for such parameters as phosphorus  and heavy metals. The first problem with this approach is that it
assumes that the build-up/wash-off of TSS is much greater than any other source pathway. This has not been found to
be the case (Pitt, 1993).  When build-up based/wash-off models  are calibrated to real  data, the  build-up/wash-off function
must be set to be much larger than it really is in order to match actual data.  When a source control such as  street
sweeping  is applied, the  model will then significantly overestimate its effectiveness no matter what the assumed street
sweepingefficiency  is.   This  may explain why street sweeping has seldom if ever been found to be as effective as
predicted. The second problem with these models is the assumption that other constituent concentrations can be related
                                                      142

-------
                          Flat Creek  Drainage  System

            Capital  Improvements  Proposed by  Woodward-Clyde

             Using lo-year  Design Storm  of November 23,  1960
                                                              City of Eugene
                                                               Enlarged Area
                                            37cfs  f s
          Open Channel
          Pipe or Culvert
          Transition Between  Section
                  
-------
    There have been a number of attempts to develop a better understanding of how stormwater quality BMPs work and
why (Strecker, 1992; Brown and Schueler, 1997). However, what we know about the effectiveness of stormwater best
management practices in improving water quality and ultimately aquatic health has been questioned with good reason
(Strecker 1994; Urbonas  1995; Maxted and Shaver,  1996). Some of the questions arise from the actual studies and how
they have not been completed as to be very useful in assessing effectiveness. In addition, there have been suggestions
that pollutant removal efficiencies may not be the best way to assess effectiveness (irreducible concentrations, etc.).

    Finally, there have been some studies that have shown that downstream of some BMPs (e.g., detention systems),
aquatic invertebrate populations are no different from systems that do not have such in-stream ponds (Maxted and
Shaver, 1996).

    What we know is the application of BMPs is an evolving science and that the exact cause and effect relationships
are not well  known. However, we do know that BMPs have been effective at reducing concentrations. In cases where
there has been no downstream improvement in aquatic invertebrate health from BMPs, we should ascertain what the
limiting factors are and whether the BMP was able to mitigate some if not all of them before we dismiss a BMP.  In
addition, we  need to understand whether other attributes of the BMP may be contributing to downstream problems such
as demonstrated downstream temperature impacts  (Galli, 1991) of on-line ponds, as well as the interruption of drift of
aquatic invertebrates downstream.   It is becoming increasingly clear that within-stream detention systems need to  be
very carefully evaluated before they are selected as BMPs. What we will need to do  in master planning is to make good
subjective decisions regarding  the appropriate application of BMPs for water quality. We do not have the data and models
to do otherwise.

Integrated Storm Drainage  Master Plans  - Stream  Stability/Habitat

    Unless a watershed has a great ability to infiltrate stormwater or evaporation is a viable technique, stream hydrology
will change (increased runoff) with development. While there are some great techniques to reduce the changes (e.g.,
Prince  Georges Department  of Environmental  Resources. 1997), in many  cases these techniques will not be able to
reduce the increased energy within a  stream  enough to stop channel cutting and downstream sedimentation  from
occurring. A technique that has been employed in an attempt to prevent downstream damage is the requirement that
new development controls runoff from a one- or  two-year  event such that pre- and post- development peak flows for that
event are equaled. MacRae  (1996) has demonstrated that this approach  may actually cause more problems then it
solves.  It usually leads to shifting over-bank flow energy to the wetted channel, further exacerbating channel down
cutting. Figure 1 demonstrates this. Suppose that the peak in hour 14 was the one-year pre-development flow for this
creek.  Maintaining post-development flows to this level would significantly lengthen  the time the creek is subject to this
channel-forming flow condition, while reducing over-bank flows.  Even setting  post-development peak runoff rates to  one-
half pre-development, results in significant extended energy in the channel. One would likely have to set a requirement
that the flow rate be one-fourth or one-fifth to have  a positive effect. This would require very large detention areas.

    In  many, if not most cases,  we believe that the master plan  must include within-stream structures to assist it in
changing with development (Severn, 1996). That is, the plan must move beyond just getting runoff to a stream and
making sure any culverts in the stream are "right-sized."  Master plans should include a component to design in-stream
structures (habitat friendly ones,  of course) and have an adaptive management program for them. This approach has
been successfully  applied  to the  Pipers Creek and Thorton Creek watersheds  in Seattle,  both  heavily urbanized
watersheds.  What this can accomplish is much faster and more postive equilibrium for the stream system (e.g., the
increased energy can be utilized to create deeper pools and increased spawning gravels in the pool tailway).

Integrated Storm Drainage  Master Plans - Capital Improve  Projects (CIPs)

    The above integrated stormwater master planning elements will result in changing the traditional definition of what
a CIP is. Traditionally it has been structural controls located within the municipally owned stormwater systems (e.g., the
streets and street drainage structures and at creek crossings, etc.) Now CIPs can include property or property rights
acquisition, buffer areas, protection and enhancement of natural  resource sites and preservation of the open channel
drainage system.


                                                    144

-------
Integrated Storm Drainage Master Plans - Public vs.  Private Solutions

    Another element of master planning can include the evaluation of the trade-off of requiring private solutions (e.g.,
on-site design requirements) versus implementing public stormwater system measures. Figure 11 shows schematically
that on a watershed basis, one could  employ a combination of both to achieve the overall most cost-effective system.
This can be addressed in  modeling  and cost-estimation for both approaches and  then one or some combination
employed.
            w
            o
           o
           +*
            v
            o
            O)
            n
            
-------
Metropolitan Washington Council of Governments.  Maryland  Department of Environment. Washington D.C. 188pp.

MacRae, C. 1996.  Experience  from  Morphological Research  on Canadian  Streams:  Is Control of the  Two-year
Frequency  Event the Best Basis for Stream  Channel Protection?   In: Effects of Watershed  Development and
Management on Aquatic Ecosystems. American Society of Civil  Engineers. Edited by  Larry Roesner. Snowbird, Utah.
Pp 144-1 62.

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.

Maxted, J.,  and E. Shaver. 1996. The Use of Retention Basins to  Mitigate Stormwater Impacts on Aquatic Life. In: Effects
of Watershed Development and Management on Aquatic Ecosystems. American  Society of Civil Engineers.  Edited by
Larry  Roesner. Snowbird, Utah. pp. 494-512.

Prince Georges Department of Environmental Resources. 1997. Low Impact Development Design Manual.  Department
of Environmental Resources, Prince Georges County, Maryland.

Pitt, R. 1993. Source loading and management model (SLAMM). National Conference on Urban Runoff Management,
March 30-April 2, 1993. Chicago, IL.

Schueler, T. 1994. The Importance of Imperviousness. In: Watershed Protection Techniques, l(3): 100-1 11.

Severn,  D.,  and  P.  Washington.   1996.  Effects of Urban Growth  on Stream Habitat. IN: Effects of Watershed
Development and Management on Aquatic  Ecosystems. American Society of Civil Engineers. Edited by Larry Roesner.
Snowbird,  Utah.  pp. 163-1 77

Strecker, E.,  1994. Constituents and Methods for Assessing BMPs. Proceedings of the Engineering Foundation Conference
on Stormwater  Related  Monitoring Needs. Aug. 7-12, Crested Butte, Colorado. ASCE.

Strecker, E.W., Kersnar, J.M., Driscoll, E.D., and Horner, R.R.  1992. The Use of Wetlands for Controlling Storm Water
Pollution. The Terrene  Institute.  Washington, D.C.

Strecker, E.,  M. lannelli, and B.  Wu. 1997. Analysis of Oregon Urban Runoff Water Quality  Monitoring Data Collected from
1990  to  1996.  Prepared by Woodward-Clyde for the Association of Clean  Water  Agencies.

Urbonas, B.R.,  1995. "Recommended Parameters to Report with BMP Monitoring  Data. J. Water Resources Planning and
Management, ASCE. 121(1),  23-34.

Woodward-Clyde, 1997. Willow Creek Physical,  Biological, and  Chemical Baseline Report. Prepared  by Woodward-
Clyde, for the City of Eugene.
                                                   146

-------
  Conservation Design: Managing Stormwater through Maximizing Preventive

                                     Nonstructural Practices

                                             Wesley R. Horner
                                     Environmental Management  Center
                                         Brandywine  Conservancy
                                        Chadds  Ford,  Pennsylvania


Abstract

    Unlike conventional methods of stormwater management that prioritize peak rate control to mitigate post-development
downstream flooding effects, Conservation Design  first aims to prevent or minimize the creation of stormwater from the
outset.  Preventive Conservation Design methods are defined in this paper as those that integrate stormwater
management into the initial stages of project design,  instead of waiting to  consider them in the final steps of the site
planning process. Mitigative Conservation  Design techniques will  be explored that use natural  processes performed by
vegetation and soil to mitigate unavoidable stormwater runoff impacts  once prevention has  been maximized to the
greatest extent possible.  Underlying these techniques-whether preventive  or mitigative  in  nature-is a  comprehensive
perspective of water resources that views stormwater as an asset to be  managed, not a waste for disposal.

    This paper summarizes a recent project which  the Brandywine Conservancy  undertook for the Delaware Department
of Natural Resources and Environmental Control, with support from USEPA  Section 319 funding. For interested readers,
Conservation Design for  Stormwater  Management: A Design Approach  to Reduce Stormwater Impacts from land
Development  (Delaware Department  of Natural Resources  and Environmental  Control with  Brandywine Conservancy,
1997) further details all  aspects  of the Conservation Design program described  here. This manual is referenced
throughout this paper and is available  by contacting DNREC at 302-739-4411 in Dover  DE.

Introduction

    Most Stormwater management programs place a heavy reliance on implementation of structural  stormwater
managementfacilities:  detention basins,  conveyance piping  and  inlet/outlet  structures. These  facilities-though created
to mitigate negative stormwater impacts by controlling flooding-cannot in  and of themselves  eliminate adverse impacts
of urban development throughout a watershed.  In fact,  because these systems fail to  acknowledge  and  plan for critical
system-wide water cycle processes, stormwater management itself can become  a problem, rather than a solution. This
is especially true when  conventional  stormwater management  systems  are  combined with conventional large-lot
subdivision  designs.

    The negative effects of this type of development and conventional stormwater management have been described in
a variety of recent studies and reports, including  the Pennsylvania Handbook of Best Management Practices for
Developing Areas (CH2MHJII, 1998) and a variety of other state stormwater manuals; Center for Watershed Protection
publications  such as Better Site Design: A Handbook for Changing Development Rules in Your Community  (Center for
Watershed Protection, 1998) and Planning for Urban  Stream Protection (Schueler,  1995); the Northeastern  Illinois
Planning Commission's  Reducing the  Impacts of  Urban Runoff: The Advantages of Alternative Site Design Approaches
(Northeastern Illinois Planning Commission, 1997), and Urban Stormwater Best Management Practices for Northeastern
Illinois (Northeastern  Illinois Planning  Commission, 1993).  These  effects include:

    . Altered site  hydrology and  reduced groundwater recharge

    . Reduced stream base flows

    . Altered stream  geomorphology (resulting in damaged  aquatic habitat)
                                                   147

-------
      Loss of site area for other uses (e.g.; recreation)

    •  Single  purpose:  disregards  site resource conservation  benefits

    .  Lack of attention to water quality

    .  High construction costs

    .  Maintenance burdens and costs

    .  Negative  visual appearance (e.g.,  basins often fenced off)

    .  Limited number of stormwater discharge points

    • Less flexibility in design

    Conservation  Design reflects a totally different philosophy toward land development that integrates stormwater
management into the very core of site design, as opposed to considering it a problem to be resolved after the design has
been completed. This philosophy regards stormwater as a key component of the hydrologic cycle and critical to
maintaining the water balance-and groundwater reserves-for a particular watershed.

    Recently  we have come to realize that  land development's impacts to water  resources are  not one-dimensional.  They
include,  in addition  to flooding, the multiple concerns of water quality,  groundwater quantity, stream  and wetland
characteristics,  in-stream habitat, and  biodiversity.  Therefore, stormwater management and  site design must be
approached  much  more comprehensively. At the foundation of this comprehensive approach  lies an understanding of
the relationship between land development and our water  resources.  In  order to better comprehend this relationship, we
must understand  the water cycle itself-the amount of rainfall, evapo-transpiration,  groundwater  infiltration, and
runoff-and how this cycle is affected by the characteristics of an individual site such as soil types, topography, and
vegetation.

The Water Cycle and Landscape Dynamics

    Appreciation  of the water cycle  is especially important to achieve successful, comprehensive  stormwater management
(Figure 1). In fact,  only through understanding full water cycle dynamics, can we hope to achieve some sort of system
balance  and minimize negative stormwater impacts. Figure 2 displays a generic flow chart of the water cycle that
highlights the various components  of this cycle  and  how they are interconnected (Conservation Design for Stormwater
Management,  1997).  It is important to appreciate that the system itself is a closed loop: what goes in, must come out.
If inputs  to infiltration are decreased by 10 inches, then inputs to surface runoff and/or depression storage must be
increased by this same amount. Furthermore,  infiltration outputs must  also be decreased: following  along on the flow
diagram,  the  groundwater reservoir, evapo-transpiration and soil moisture elements together will be reduced by this 10
inches, which will  reduce stream  baseflows.

    The  logical  first step in any  discussion of the water cycle is precipitation-h all its various forms.  In southeast
Pennsylvania, and indeed throughout much of the Mid-Atlantic states, the climate is relatively humid (Conservation
Design for Stormwater Management 1997, based  on Hydrosphere 1992 database). Substantial  precipitation tends
to be distributed throughout the year in frequent events of modest size. This consistency in  rainfall throughout the year
indicates that this  region does not have a defined wet or dry season  as do other areas of the country.  This rainfall
potential  throughout the year has significant implications for consideration  of stormwater runoff.  For example,  having
rainfall throughout the year  indicates that sediment  laden runoff can  occur at any time; therefore, it is  important to
establish  some  sort of erosion-controlling groundcover during all seasons  of the  year.

    Also  important  is the distribution of rainfall by size of event. Based  on analysis of 35 years of data from a Wilmington,
Delaware rain gage  (Conservation Design for Stormwater Management 1997), it is clear that the precipitation occurs
mostly in small "events" or storm  intensities. Ninety-eight  percent of the total number of events during this extended
period were classified in the  "less  than 2 inches" category. Even more important from a water cycle perspective,  96%
of the average annual rainfall volume occurred  in storms of less than 3 inches (which  is less than the  2-year, 24 hour

                                                     148

-------
               Shallow Infiltration
                  Deep Infiltration
                                       Groundwater discharge to lakes, streams and oceans
Figure 1. The Water Cycle.
storm). This understanding  of storm size distribution is  critical for a variety of reasons in stormwater management. For
example, if our concern  is keeping the water cycle in relative balance, capturing and recharging the 1 - or 2-year storm
as the basis for design will encompass the vast bulk of precipitation and stormwater runoff volumes  in the average year
and provide adequate water cycle balance. This leads to very different design criteria than if flooding (peak runoff rates)
is the  only concern addressed.

    Another key component of the water cycle is the linkage between stormwater infiltration, groundwater recharge  and
stream baseflow.  As  land  is developed and  impervious coverage  increased, less water is recharged to groundwater
aquifers  (Thomas  Dunne and Luna Leopold's  Wafer in Environmental Planning [Dunne and  Luna, 1978] is an
excellent background text in  addition  to the above referenced  reports). As these subtractions continue acre-by-
acre,  development-by-development, their  cumulative effects grow larger. Also,  as development occurs,  more water  is
often  withdrawn from  the underground  reserves for  drinking, irrigation, or commercial uses, As subtractions are made
from the  groundwater reservoir flow, the impact will  be  seen in the form of a lowered water table and reduced  stream
baseflow discharge. Headwater springs and first-order  streams-the lifeblood  of our stream systems-may even dry up.
The baseflow from  headwater zones is critical to maintaining a diversity of aquatic plant and animal life, as well as
terrestrial animals dependent on certain aquatic species for survival.  In  some cases the groundwater reservoir does not
discharge to a stream, but ratherto a wetland. In these instances, reduced infiltration and a lowered water table ultimately
translate  into a loss of wetlands themselves, and an elimination of their rich and vibrant ecological function.

    A final component of the water cycle that must be addressed  is  overland runoff. This is the component most
frequently addressed in conventional stormwater management approaches, for it is the cause of increased downstream
flooding.  Three major elements determine the volume and character of stormwater runoff for a given  storm intensity:  soil
type, land cover (including vegetation and debris), and slopes. Soils vary  widely in their ability to infiltrate stormwater and
                                                      149

-------
Figure 2. Water Cycle System Flow Chart.

minimize runoff and are classified accordingly  by the  USDA Natural Resources Conservation Service  (NRCS) into four
categories  based on their  permeability  rates (Hydrologic  Soil Groups A through D, with A having best permeability).

    Land cover greatly affects the rate and volume of stormwater runoff and has significant water quality impacts as well.
Obviously,  the  landcover of greatest concern for stormwater management is impervious coverage created through the
development process.  Interestingly,  compacted lawns  and cultivated fields can have  significant  runoff rates as well,
especially when no crop covers the bare soil.  The landcover in this region best suited to retard stormwater runoff and
assist in its infiltration is the natural one: the piedmont forest. A mature forest can absorb  much more water than an
equivalent area of turf grass due to the presence of an organic litter layer and herbaceous and woody plant material. The
organic litter layer on the forest floor provides a physical barrier to sediments, maintains surface soil porosity, and  assists
in  denitrification and other water quality functions. The vegetation, both herbaceous and woody, physically retards runoff
and erosion with its spreading root mats and also assists in  maintaining soil permeability and water quality by taking up
nutrients through its root  systems.

    Finally, slopes are another critical component of the stormwater runoff equation. Steeper  slopescan accelerate runoff
and increase the erosive force of the water.  Therefore,  removing vegetation on steeper slopes  can have dramatic impacts
on downslope  aquatic systems.
                                                       150

-------
    As seen above, the water cycle and the implications for stormwater management are complex and comprehensive.
The process  of urbanization  dramatically impacts the functioning of this water cycle.  Conservation Design  has been
developed to address the issues of comprehensive stormwater management and to address the land use patterns that
impact it.

Land Use  and Site Development Impacts

    Throughout much of the United  States, farmland and natural areas are converted to suburban development at an ever
accelerating pace. In fact there is hardly a city in America that does not  occupy at least two to three times more land area
than in 1970, even if population has not increased  proportionately. This history of land use change is certainly true of the
Mid-Atlantic states, where  communities continue to grapple with the effects of unmitigated suburban sprawl.

    The dynamic nature of wet-weather flow regimes  and  landscape ecology make it difficult to assess the impact  of
urbanization on aspects of the water cycle such  as groundwater  reserves and  aquatic habitat. However,  studies have
indicated that the biological  community in urban streams isfundamentallychanged to a lower ecological quality than what
was there  before development occurred.  In one study in Delaware, approximately 70% of the  macroinvertebrate
community found in streams of undeveloped forested watersheds was comprised of pollution  sensitive mayflies,
stoneflies, and caddisflies, compared to 20% for urbanized watersheds (Maxted and Shaver  1996). Other studies suggest
that the decay in stream quality is very rapid in the  early stages of watershed  urbanization; watersheds with less than 10%
impervious cover are the most susceptible to the adverse effects of  urbanization. Therefore early intervention as a
watershed begins to develop  is critical,  and furthermore, this  intervention  should  include  measures to address stormwater
management  and land use in a connected, comprehensive manner.

    In addition to in-stream  habitat impacts,  the issue of land development and water resources also has great
implications for our human communities well beyond the issue of flooding.  Reduced stream baseflows and groundwater
resources means decreased availability of drinking water supplies.  Also, reduced baseflows  result in less available water
for diluting the pollution output from industrial or municipal waste systems. As stormwater runoff increases, water  quality
can be greatly impacted by stream bank erosion, resuspension of sediment,  runoff of chemicals  and fertilizers from lawns
and fields,  and increased stream temperatures. Stormwater-linked  pollutants vary with type of  land use and intensity  of
use and have been shown to include bacteria, suspended solids, nutrients,  hydrocarbons, metals, herbicides and
pesticides, toxins and organic matter. Not  only are these pollutants increased,  but the  landscape's natural capacity for
filtering and chemical uptake through vegetation is decreased as land is cleared and paved. All of these pollutants can
impact both drinking  water supplies and natural aquatic systems.

    Thus it becomes evident that if the  negative effects  of land development on our  water resources are to  be minimized,
we must find alternatives to the  conventional structural approach  to stormwater management. Moreover,  these
alternatives must address the issue  of land  use and patterns of development  in a comprehensive fashion, one that  strives
to maintain a  hydrologic balance  on site and  replicate the pre-development hydrologic regime to the greatest  extent
possible. One approach-or collection  of approaches-that can accomplish these goals is  Conservation Design.

Conservation Design Principles

    Stormwater management throughout the Commonwealth (and  elsewhere)  can be markedly  improved by approaching
stormwater differently than has been the  practice in the past, where  "stormwater management" has been defined  largely
as stormwater disposal. This different perspective challenges us  to  maximize prevention, even before stormwater
becomes a problem, and to avoid highly  engineered structural solutions  that are expensive to build and maintain. In their
place, Conservation  Design focuses on utilization of natural  systems and processes to achieve stormwater management
objectives where feasible. At the same time, this new approach is intended to work with  site resources-woodlands, soils,
wetlands, etc.-to enhance their stormwater functions. The end  result is a site  design which minimizes stormwater
generation  and then  mitigates the  remaining stormwater in a low-impact  manner, with an emphasis  on groundwater
recharge. Conservation Design is not so much a  singular approach or solution as it is a collection of approaches and
practices that are flexible enough to effectively address any given  site and development program. Common to all these
approaches and practices  are several basic principles.

                                                     151

-------
    Achieve multiple objectives. Stormwater management should  be comprehensive in scope, with techniques
designed to achieve multiple stormwater objectives. These objectives include both peak rate and total  volume control
(i.e., balance with the hydrologic cycle), as well as water quality control  and temperature maintenance. These objectives
should include  maintaining or improving the pre-development hydrologic regime.

    Integrate stormwater management early into the site  design process. Stormwater management tacked on at
the end  of the site  design  process  almost  invariably is flawed.  To optimize comprehensive stormwater  management
objectives,  stormwater management must be integrated into  the first stages of  the site planning. Stormwater  impacts may
even  be  a  factor in determining type  of use, extent of use, and location of the development on a site.

    Prevent first, mitigate second. Approaches to site design which can reduce stormwater generation from the outset
are the most effective approach to stormwater management. For example, effective clustering of units significantly
reduces length of roads when compared  to conventional development. Reduction in  street width and driveway  length  can
minimize impervious  coverage. These type  of approaches  are rarely thought of as stormwater management practices,
yet they  achieve  powerful stormwater quality and quantity benefits.

    Manage stormwater as close  to  the source of generation as possible. From both an environmental  and
economic perspective, redirecting runoff back into the ground as close to the  point  of origin as possible,  is  preferable to
constructing elaborate conveyance  systems that increase flows and  suffer from failures over  time. Avoid concentrating
stormwater. Disconnect, rather than  connect, where  feasible.

    Engage natural processes in  soil  mantle and plant communities. The soil mantle  offers critical  groundwater
recharge conveyance and  pollutant  removal functions through physical filtration,  biological action,  and chemical
processing.  Understanding  how much of what  type of soil is in place on any given site is  essential when assessing
stormwater   management/water  quality  impacts  and  opportunities.  Vegetation similarly  provides  substantial  pollutant
uptake/removal potential and can assist in infiltration by  maintaining  soil porosity and retarding runoff.  In addition,
naturally vegetated areas improve their stormwater functions  over time as leaf litter and  debris  builds a richer organic soil
layer. Areas of good soil permeabilities (A and B  soils) and intact vegetative communities should be  prioritized in
prevention  strategies.

A Conservation  Design Procedure

    The  Conservation Design principles  outlined above, though  greatly simplified,  can offer valuable guidance  when
approaching a particular land  development project.  In fact, these five principles form the basis  for a Conservation Design
Procedure. This  Design  Procedure  incorporates both Preventive Approaches and  Mitigative Practices. Preventive
Approaches tend to be broader in geographic scope than  other techniques and typically may influence some  of the major
decisions regarding a particular development project. Approaches  may even transcend the site itself, involving an  entire
planning jurisdiction or area, or even an entire region. Also, Preventive Approaches attempt to reduce impervious
coverage or minimally disturb  the existing vegetation and soils in prime recharge areas. For example, a reduction in road
width  from  30 feet to 18 feet  means an immediate  33.3% reduction in roadway imperviousness, which typically comprises
a large  portion  of site imperviousness.

    Mitigative Practices  include mitigative techniques which are often  more structural in  nature. These practices
encompass a rapidly growing array of biofiltration and bioretention methods that maximize the stormwater management
potential  of soils  and vegetation. Mitigative  Practices include vegetated swales for stormwater conveyance, vegetated
filter strips  and riparian buffers, grading,  berming, terrraforming, and level spreading stormwater in  natural areas. These
practices  should mitigate as close to the  source as possible and achieve multiple objectives.  For example, a berm, which
is used to retain stormwater runoff on  a forested slope, can  double as a walking trail, thus decreasing the  expense of two
separate  individual systems.

    Figure  3 graphically displays the Design  Procedure as  a flow diagram. The procedure itself can be thought of as a
series of questions which  must be asked as Conservation  Design  is applied to each site. If site designers rigorously
address all of these questions, the "answers"-the Conservation DesignPreventive Approaches and  Mitigative


                                                     152

-------
                             Conservation Design Procedure
                             Prevention
                              Building
                              Program
                                Lot
                           Configuration
                            Impervious
                             Coverage
                             Minimum
                            Disturbance
                                                       Conceptual
                                                       Stormwater
                                                      Management
                                                          Plan
                                                 Calculations
                                                  (peak and
                                                   volume)
 Sizing
   of
Controls
Figure 3. A Conservation Design Procedure.
                                          153

-------
Practices-will successfully be identified for each site. The overriding objective ultimately is to achieve a new way of
thinking about site design. The  procedure  begins with an  effective and complete  Site Analysis, which  can  help identify
both areas of concern and resourcesforopportunity in regard to stormwater management. The procedure then flows from
macro, larger-scale preventive questions (i.e., how can the design be clustered to reduce  site disturbance) to micro, small-
scale mitigative questions (i.e., can stormwater be infiltrated in  bioretention areas?). Probably the most  important aspect
of the procedure in Figure 3 is its  positioning of the Conceptual Stormwater Management Plan as a concurrent task with
the entire site design process. This reinforces the notion that stormwater  management should be an integral part of the
entire design process, including the site analysis.
    In  order to  better understand the Conservation Design  Procedure, each of its components  (the Preventive Approaches
and  Mitigative Practices)  is discussed  in more detail below.
Site Analysis
Three major aspects need to be addressed in the  Site Analysis process:
               Site  Background and  Context
               What is the  surrounding context?
               What is its location in the watershed?
               In which geologic/geographic region  is it  located?
               What is the site size?
               What are adjacent uses and landcover?
               Critical  Natural  Features
               Existing   hydrology?
               Wetlands? Floodplains? Riparian  buffers?
               Steep slopes?  Special habitat areas?
               Stormwater  Opportunity Areas
               Where are soils that are best suited for stormwater recharge? Worst?
               Where  is existing landcover optimal  to prevent stormwater?
               What opportunities exist to use vegetation  and soils in  mitigation?
               On what soils and slopes  is this vegetation?
               What is  depth to bedrock or water table?
        Preventive  Approaches
        The Preventive  Approaches include a range of hierarchical questioning:
               Building Program
               What is  the current zoning and density for this tract?
                                                       154

-------
        Is there currently an open space design option for the site?
        Can  the proposed  building program be reduced in terms of density?
        Can  the type of unit or lot size be modified to promote open space?
        What are the possibilities for water and sewer supply?
        Lot  Configuration
        Have lots  been reduced and open space been maximized?
        Have lots  been clustered to avoid critical  areas of recharge?
        Have lots  been configured to take advantage of mitigative  practices?
        Impervious Coverage
        Has  development been  clustered  to reduce  impervious surfaces?
        Have road  widths  been  minimized?
        Have building setbacks  been  minimized to reduce driveway lengths?
        Have parking ratios and needs  been carefully examined?
        Have needs and sizes of walkways been examined?
        Minimum  Disturbance
        Has  maximum  total  site area, including soils and vegetation,  been  protected from  clearing and
        disturbance?
        Are  zones of undisturbed open  space maximized?
        Have buildings been  sited carefully to reduce vegetation removal?
        Can  no-disturbance buffers be installed to limit zones of soil compaction?
Mitigative  Practices
The  Mitigative Practices include  a tool box of  options that promote  groundwater recharge and  improve water
quality. These practices have been  assigned to  several groupings, although in many cases the practicesoverlap.
Virtually all of these  techniques make  maximum use  of vegetation and soil functions,  so although they are all
technically structures, they are  of lower complexity  and more rooted  in natural process than conventional
approaches.
        Vegetated  Swales
        Vegetated  swales  are effective  means of stormwater conveyance. At low slopes, they  can  recharge
        modest amounts  of stormwater,  filter it through vegetative processes, and slow  it down.
                                              155

-------
                Terraforming

                Terraforming comes in a variety of techniques. These include constructing subtle berms  along contour
                below undisturbed areas. The berms act as modest "dams" retaining  the water for up-slope recharge.
                Also, subtle grading of depression areas promote  retention and recharge  throughout a site.

                Level  Spreading/Natural Areas

                With a  level spreader,  stormwater spills over the lip of a long trench or berm,  creating sheet flow across
                a broad area. The level spreaders slow down the intensity  of runoff and discharge it over a large,
                adjoining vegetated area with good soils, which in turn filter it  and assist in groundwater recharge. Filter
                strips are  planted vegetated  strips through which runoff passes that filter it and slow it down.  Riparian
                buffers are vegetated zones along stream corridors  that filter the stormwater passing through it and help
                minimize  erosion. These techniques are  most valuable when used in conjunction with preventive
                strategies  that leave larger natural areas  undisturbed in order to handle these additional  stormwater
                inputs.

                B ioretention/b iof i Itration

                Bioretention is a  popular name given to just about any type  of device that utilizes vegetation and soil to
                manage stormwater flows. They can  be subtle depressions that exist  naturally and  receive stormwater
                or depending on  soil conditions, they  may be  physically constructed "pits" that are filled with permeable
                soils and planted  with  native vegetation that adapt  to both wet and dry conditions.  These systems can
                either be "on-line"  (part of the stormwater conveyance flow) or "off-line" (separate from the  rest of the
                stormwater management/mitigation system). In either case,  they have modest ponding  storage that is
                recharged over the course of time.

                Other  mitigative  devices

                Not all  of the required  volume storage to meet peak rate requirements for a given site may be attained
                through the practices outlined above.  At times, it may be necessary to put in "structural" systems such
                as  in-ground  infiltration trenches, infiltration  pipes,  or stormwater wetlands.  However,  these systems
                should  be  explored only after  both Preventive Approaches and  Mitigative  Practices  of  Conservation
                Design have  been maximized  to  the greatest  extent possible.
Conclusion
    The Conservation Design Procedure is perhaps best characterized as a "check list" or protocol of questioning during
the site design  process.  The key to this approach is its range of innovative, yet effective options, not afforded in
conventional systems which tend to be standardized irrespective of the particular site. With Conservation  Design,  the
approaches and practices can be combined in a variety of ways  to minimize the  impacts of development on the water
cycle and  still  meet  regulatory stormwater  management criteria  such as peak rate control.  Often, because these
approaches and practices tend to favor multiple objectives and nonstructural techniques,  Conservation  Design can be
less expensive to install and maintain than conventional systems. Also, because they are largely based  on soil and
vegetative processes, conservation design techniques tend to improve in function  overtime,  while conventional detention
basin systems tend to diminish in function over time. In terms of water quality, Conservation Design Approaches and
Practices can  outperform  conventional systems.  For example, filter strips and biofiltration areas  can remove over  90%
of the suspended solids, 40% of the phosphorous, and 20% of the nitrates (Dillaha et al. 1986 and 1989; Yu  et al. 1993).
In addition,  reduced yard areas and  increased forested  zones prevent chemical  runoff from lawns-a great contributor
to non-point source pollution-at the  outset.

Conservation Design is  limited only by the creativity of the designer and the flexibility of the developer and  regulatory
agencies. It must be emphasized that the  Conservation Design  approach will not eliminate a need for structural systems
in all cases; however, more often than not, Conservation  Design can replace or reduce the  need for structural practices

                                                       156

-------
while  providing attractive site amenities. And in the process,  the water cycle  will be  balanced, and forests and  other
sensitive  resources will be preserved. In short, Conservation  Design  can  do more with less, and more for less, than
conventional approaches  to  stormwater management.

References

Center for Watershed Protection, 1998; Better Site Design:  A Handbook for Changing Development Rules in  Your
Community;  Elliot City, MD.

CH2MHNI, 1998; for the Pennsylvania Association of Conservation Districts, Inc., and other agencies; Pennsylvania
Handbook of Best Management  Practices for Developing  Areas; Harrisburg, PA.

Delaware  Dept. of Natural  Resources and Environmental Control with  the Brandywine Conservancy, 1997; Conservation
Design for Stormwater Management:  A Design  Approach to Reduce  Stormwater Impacts from  Land  Development; Dover,
DE.

Dillaha, T.A., J.H. Sherrard,  and D. Lee, 1986;  Long-term Effectiveness and  Maintenance of  Vegetative Filter Strips;
Virginia Water Resources  Research Center  Bulletin No. 153; Blacksburg, VA.

Dunne, Thomas  and  Luna Leopold,  1978; Water in Environmental  Planning; W.H. Freeman  and Company; New  York,
NY.

Northeastern Illinois Planning  Commission,  1993; Urban Stormwater  Best Management Practices  for Northeastern Illinois;
Chicago,  IL

Northeastern Illinois Planning Commission, 1997;  Reducing the Impacts of  Urban Runoff: The Advantages of Alternative
Site Design  Approaches;  Chicago,  IL.

Schueler,  Tom, 1995;  Center for Watershed Protection; Site Planning for  Urban Stream Protection;  Silver Spring, MD
(relocated to Ellicott City).

Yu, S., S. Barnes, and V.  Gerde; 1993; Testing of Best Management Practices for Controlling Highway Runoff; Virginia
Transportation Research Coundil; FHWA/VA 93-R16.
                                                     157

-------
         Low-Impact Development Design: A New Paradigm for Stormwater
       Management Mimicking and Restoring the Natural Hydrologic Regime
                  An Alternative Stormwater Management Technology
                                     Larry S. Coffman, Associate Director
                                   Department  of  Environmental Resources
                                      Prince  George's  County,  Maryland
Abstract
    Whether complying with federal  or state regulations or addressing  local  vital watershed  protection/restoration
objectives, local jurisdictions are confronted  with the daunting task of developing,  administering and funding complex
effective multi-objective Stormwater management programs,  Today's comprehensive Stormwater program  not only has
to deal with  runoff quantity and quality control but, may  also have to  address such complicated  issues as ecosystem
restoration, combined sewer overflow  reduction, fisheries protection, potable surface/ground water source  protection,  and
wetland,  riparian  buffer and stream  protection. As our understanding of the technical and practical limitations of
conventional  Stormwater management technology has  increased over the past two decades,  and as watershed  protection
objectives  have changed, many jurisdictions have begun  to question the efficacy and cost-effectiveness  of conventional
Stormwater approaches in  meeting today's complex  environmental/water resources  objectives. Older communities with
existing extensive  Stormwater management infrastructures are also struggling with the  economic reality of funding the
high costs of maintenance, inspection, enforcement and public outreach necessary to support an expanding and aging
infrastructure. Still  more challenging are the  exceptionally high costs  of retrofitting existing urban development using
conventional  Stormwater management end-of-pipe practices to restore and  protect receiving waters and living resources.

    With growing  concerns about the  limitations of conventional technology and  to  address the changing  objectives of
watershed protection, in 1990 Prince George's County's Department of Environmental Resources (PGDER) began
exploring alternative Stormwater management practices and strategies.  The development of bioretention or "Rain
Gardens" (using the green space to manage  runoff within small depressed landscaped  areas) led to an understanding
of how to  optimize and  engineer the landscape to  restore hydrologic functions by uniformly integrating micro-scale
management practices and impact-minimization  measures into the development landscape. In 1997 PGDER released
the Low Impact Development (LID) Design  Manual demonstrating the  principles and practices  of LID to create a
hydrologically functional  landscape  (PGDER, 1997).

    LID Stormwater management technology  can maintain or restore a watershed's  hydrologic regime by fundamentally
changing conventional site design to create an environmentally and  hydrologically functional  landscape that mimics natural
hydrologic  functions (volume,  frequency,  recharge and discharge).  This  is accomplished in four ways. First:  minimizing
impacts to the extent practicable  by  reducing  imperviousness,  conserving natural  resources and ecosystems, maintaining
natural drainage courses, and  reducing the use of pipes and minimizing clearing/grading. Second:  recreating detention
and  retention storage dispersed  and  evenly  distributed throughout a  site with the  use of open swales, flatter slopes,
depression storage, rain gardens  (bioretention), water use (rain barrels), etc. Third: maintaining the predevelopment"time
of concentration" by strategically  routing  flows to maintain travel time.  Fourth: providing effective public education and
socioeconomic incentives to  ensure property owners  use effective  pollution prevention measures and maintain
management measures. With LID, every site  feature  is multifunctional (green space, landscaping,  grading, streetscapes,
roads and parking  lots) and helps to reduce Stormwater impacts or provide/maintain beneficial hydrologic functions. The
cumulative beneficial impact of using the wide array of distributed LID techniques allows the  site designer to maintain or
restore watershed's natural relationship  between rainfall,  runoff,  infiltration and  evaporation.

    The effective use of LID site design techniques can significantly reduce the cost of providing Stormwater management.
Savings are  achieved by eliminating the  use  of Stormwater management ponds,  using less  pipe,  inlet structures, curbs

                                                    158

-------
and gutters, less roadway paving, less grading and clearing. Where LID techniques are applicable, and depending on the
type of development and  site  constraints,  stormwater and site development design construction and maintenance costs
can be reduced by 25 % to 30% compared  to conventional  approaches.

    The creation of LID's  wide array of micro-scale management principles and practices has led to the development of
new tools to retrofit  existing urban development.  Micro-scale management  practices  that filter, retain and detain runoff
can be easily integrated into the existing green space and streetscapes as part of the routine maintenance and repair of
urban infrastructure. LID retrofit techniques may lead to drastic reductions in the cost of retrofitting  existing urban
development. Reducing urban retrofit costs will increase the  ability of cities to implement effective  retrofit programs to
reduce the frequency and  improve the quality of CSOs and improve the quality  of urban runoff to protect receiving waters.
LID represents a radically different approach to controlling stormwater runoff that provides effective tools to  restore or
maintain a watershed's hydrologic functions for new or  existing development.

    In 1998 EPA provided grant funding to  assist PGDER in their efforts to  develop a  general manual  describing LID
principles and  practices,  and share this technology with other local governments throughout the nation. Efforts are
currently underway with EPA  to  further advance  LID technology by improving the sensitivity of current  hydrology and
hydraulic  analytical  models for application with small watersheds and sites and to develop new micro-scale control
approaches and practices for  urban retrofit. Additional efforts are  also underway to  demonstrate how LID micro-scale
management and  multifunctional  infrastructure  principles and practices can  be  used to control highway runoff within
existing rights-of-way. It is hoped that the LID national manual will  help to stimulate  debate on the state of current
stormwater,  watershed protection and  restoration technology and  its  future direction. The lessons learned  about LID
planning, principles,  practices and research are described in  detail in  the reference documents listed at the end of this
paper. Copies of these reference documents can  be obtained  by calling the Prince  George's County's  Department  of
Environmental  Resources at (301) 883-5832.

Background

    Typically, adverse stormwater impacts are mitigated through  conservation of natural resources (forests,  streams,
floodplains and wetlands); zoning restrictions  to direct densities and increase open space; and the  use  of structural or
non-structural control technologies (best management practices - BMP's) to treat and manage runoff quantity and quality.
Many conventional stormwater  mitigation approaches,  such as  management ponds,  exhibit a number of inherent practical,
environmental  and economic  limitations including  inability  to replicate predevelopment  watershed hydrology,  elevated
water temperatures,  costly maintenance burdens,  and accelerated stream erosion due  to the increased duration  and
frequency  of runoff events.  Furthermore, because current mitigation  practices only lessen development impacts, there
is concern about the cumulative impacts of the widespread use  of conventional mitigation practices that may
fundamentally  alter  a watershed's hydrologic  regime and water  quality,  adversely affecting  receiving waters and the
integrity of their ecosystems. Many highly  urbanized jurisdictions are beginning to question the efficacy  of current
technology  and are finding it harder to  ensure, enforce or fund stormwater programs and maintain the  massive
infrastructure created by  conventional approaches.

    Currently every site is designed with one  basic  overriding goal - to achieve good drainage.  As we  develop a site
reshaping  the  landscape  inch  by  inch,  its  hydrologicfunctionsare  altered on  a  micro-scale level. The cumulative impacts
of micro-scale changes to the  landscape drastically alter watershed hydrology. If sites can be designed to achieve good
drainage,  destroying natural  hydrologic  functions, why not design  sites with the opposite objective to maintain
predevelopment hydrologic functions? If inch by inch,  sites are carefully and intelligently engineered to maintain  hydrologic
functions,  would the  cumulative beneficial  affects  result in the preservation  of a watershed's  hydrology? Can a site be
designed  in a way to remain  as a functional part of a watershed's hydrological regime? To achieve a  hydrologically
functional development there must be a radical change in our thinking. We  must not  think in terms of impact mitigation
as the stormwater management objective,  but  rather preservation of hydrologic and environmental functions. We should
design  sites to maintain  hydrologic functions not just to mitigate  impacts. Can our current stormwater management
technology adequately meet our  regulatory objectives and water resources/ecosystem  protection needs? No  one can
answer that question for sure. However, it has  not been shown that conventional ponds replicate predevelopment
hydrology nor is there any  evidence to suggest that conventional technology can  ensure the  ecological integrity of

                                                      159

-------
ecosystems. In fact,  recent studies suggest that conventional approaches can not meet our water/natural resources and
ecological  objectives.

Introduction

    With growing concerns  about the economicsand  efficacy of conventional technology,  in 1990  Prince George's County
Maryland's Department of Environmental  Resources began exploring  alternative stormwater  management practices. The
success that was achieved through the development and use of bioretention (filtering or infiltration runoff in small
depressed landscaped areas) led us to understand that perhaps changing the form and function of the developed
landscape could be  important in mitigating  urban stormwater impacts. Later it was realized that through intelligent site
design and uniform distribution of LID micro-scale  management controls it was possible to maintain or restore hydrologic
functions in a developed watershed. What is not known is how much of a watershed's hydrologic functions can  be
maintained or  restored within  a given development type  (residential, commercial or industrial)? The one limiting factor to
maintaining/restoring  the  hydrologic regime for highly urbanized development is the lack  of available micro-management
tools. Much of the current research underway is to expand the  number of practices applicable in  highly urbanized areas.

    LID's objective is to  preserve the natural predevelopment hydrologic  regime. If predevelopment hydrology and water
quality can be maintained, this would provide the  best level of protection possible to receiving waters and aquatic living
resources. Experience  over the last  20 years  has demonstrated that maximizing the  efficiency of conventional
conservation measures and the use of conventional end-of-pipe  stormwater  management practices can  not reasonably
be used to restore watershed functions. What is needed is a new philosophical approach to site development,  an
approach that  will  allow the designer to  retain a  site's hydrologic functions.

    The approach used in LID designs  is really an old one. LID borrows  its basic principles from nature - uniform
distribution  of  micro-management controls.  In a  natural setting,  stormwater is controlled by a  variety of  mechanisms
(interception by vegetation,  small depression  storage,  channel  storage,  infiltration and evaporation)  uniformly distributed
throughout the landscape. LID  mimics these mechanisms by uniformly distributing  small  infiltration, storage, and retention
and detention  measures  throughout the developed landscape.  What we soon began to  see is that  every development
feature (green  space, landscaping, grading, streetscapes, roads, and parking lots)  can be  designed to provide some type
of  beneficial  hydrologic function.

Low - Impact Development General

    LID controls stormwater at the source creating  a hydrologically functional landscape  that mimics natural watershed
hydrology.  Low impact development (LID) achieves stormwater management controls by fundamentally changing
conventional site design  to create an  environmentally functional  landscape that  mimics natural watershed  hydrologic
functions (volume, frequency,  recharge and discharge). LID uses four  basic management  planning and design principles.
First: minimize impacts to the  extent practicable by  reducing imperviousness,  conserving natural resources/ecosystems,
maintaining natural drainage courses, reducing use of pipes and minimizing clearing and grading. Second: provide runoff
storage  measures dispersed uniformly throughout the landscape with the use of a variety of small  decentralized detention,
retention and runoff practices such as bioretention, open swales and  flatter grades. Third: maintain the predevelopment
time of concentration by strategically routing flows to maintain travel time and  control  discharge.  Fourth: implement
effective public education and incentive programs to encourage property owners to use pollution prevention measures
and maintain on-lot  landscape management  practices. A developed  site can be  designed  to become  a hydrologically
functional part of the watershed with comprehensive and intelligent use of LID practices and principles.

LID Basic Site Planning Strategies

    The goal of LID  is to design the site in a way that  mimics hydrologic functions. The first step is to minimize the
generation of runoff (reduce the change in the runoff curve number (CM)). In many respects, this step is very similar to
traditional techniques of maximizing natural resource conservation,  limiting disturbance and  reducing impervious areas.
The major difference  with LID is you must carefully consider how best  to make use of the  hydrologic soil groups and site
topography to  help reduce  and control  runoff. These considerations would include how  to:


                                                      160

-------
    1. maintain natural drainage patterns, topography  and depressions,

    2. preserve as much  existing vegetation  as  possible in  pervious soils;  hydrologic soil groups A and B,

    3. locate BMP's in pervious soils;  hydrologic soil groups A and B,

    4. where feasible construct impervious areas on less pervious soil groups C and D,

    5. disconnect impervious surfaces,

    6. direct and disburse runoff to soil groups A and B,

    7. flatten slopes within cleared  areas to  facilitate on-lot storage  and infiltration  and

    8. re-vegetate cleared and graded areas.

    Where  ground water recharge is  particularly important (to protect well, spring, stream and  wetland flows) it is
important to understand the source and mechanisms for ground water recharge. When using the  LID design  concepts
to mimic the hydrologic regime you must determine how and where  ground water on the site is recharged and where
necessary, protect and utilize the recharge areas in the site.

LID  Hydrologic  Analysis/Response

    The  objective  of LID site design is to  minimize, detain  and retain the post  development runoff volumes  uniformly
throughout the  site close to the  source to simulate predevelopment hydrologic functions. Widespread  use  and uniform
dispersion of on-lot small retention and/or detention practices to control both  runoff discharge volume and rate  is key to
better replicating predevelopment  hydrology. Using LID  practices also produces runoff frequencies that are  much closer
to existing conditions than  can be achieved  by  typical application  of conventional BMP's. Management of both  runoff
volume and peak runoff rate is included in the design. This is in contrast to conventional  end-of-pipe treatment that
completely alters the watershed  hydrology regime.

    The  LID site analysis and design  approach  focuses on four major hydrologically based planning  elements. These
fundamental factors affect hydrologic and are introduced  below.

    1.  Curve Number (CN) - A factor that accounts for the effects of soils and land cover on amount  of runoff generated.
       Minimizing the change in the post development CN by reducing  impervious areas and preserving more trees and
       meadows to reduce runoff storage requirements,  all  to  maintain the predevelopment runoff volume.

    2. Time of Concentration (Tc) - This is related to the time runoff travels through the watershed. Maintaining the
       predevelopment Tc reduces peak runoff rates after development by  lengthening flow paths and reducing the use
       of pipe  conveyance systems.

    3.  Permanent  storage areas (Retention) - Retention storage is needed for volume  and peak control, water quality
       control and to maintain the same CN  as the  predevelopment condition.

    4.  Temporary  storage areas (Detention) - Detention storage may  be needed to maintain the  peak runoff rate and/or
       prevent  flooding.

Minimizing the Change  in  CN

    Reducing the change  in CN will reduce both the post development peak discharge  rate and volume. Calculation of
the LID CN  is based on a detailed evaluation  of the existing and proposed land cover so that an  accurate representation
of the potential  for runoff can be obtained. This calculation requires the engineer/planner to investigate  the following key
parameters  associated  with LID  including:


                                                      161

-------
    1. land cover type,
    2. percentage of and connectivity of impervious cover,
    3. hydrologic soils group  (HSG), and
    4. hydrologic conditions (average  moisture  or  runoff conditions).
    The following are  some of  the LID site planning practices that can be utilized to achieve a substantial reduction in the
change of the calculated CN:
    1. narrower driveways and  roads (minimizing impervious areas),
    2. maximized tree  preservation and/or afforestation,
    3. site finger-printing (carefully siting  lots/roadways  to avoid disturbance of streams, wetlands and  other resources),
       greater use  of open  drainage swales,
    4. preservation  of soils with  high infiltration  rates to reduce  CN,
    5. location of BMP's on high-infiltration soils and,
    6. construction  of impervious features  on  soils  with low infiltration rates.
Maintaining the Predevelopment Time of Concentration Tc
    The LID hydrologic evaluation requires that the post development Tc be close to the predevelopment Tc. Minimizing
the change in pre and post Tc will help maintain the same frequency of runoff discharges, assuming there is  uniform
distributed  micro-scale retention and detention  of LID practices. The following are some of the site planning techniques
can be used  to  maintain the existing Tc:
    1. maintain  predevelopment  flow path length by dispersing and redirecting flows  using open swales and vegetated
       drainage   patterns,
    2. increase  surface  roughness (e.g., preserving woodlands, vegetated  swales),
    3. detain flows  (e.g., open swales, rain gardens, rain barrels etc.),
    4. minimize  disturbances  (minimizing soil  compaction  and changes  to  existing  vegetation /drainage  patterns),
    5. flatten grades  in impacted areas,
    6. disconnect impervious  areas (e.g.,  eliminating  curb/gutter and redirecting down  spouts) and,
    7. connect pervious  areas to vegetated areas  (e.g., reforestation, afforestation).
    The combined use of all these techniques results in  cumulative impacts that modify  runoff characteristics  to effectively
shift the post development peak runoff time and frequencies to  that of the predevelopment condition, and  lower the peak
runoff rate.
Maintaining the Redevelopment Curve Number and Runoff Volume
    Once  the post development Tc  is maintained at the predevelopment conditions and the impact of  CN  is minimized,
any additional reductions in runoff volume must be accomplished through distributed  micro-scale  on-site stormwater
management techniques. The goal is to select the appropriate  combination of management techniques  that simulate the
hydrologic functions of the predevelopment condition to maintain  the existing CN and corresponding runoff  volume.  The
target design volume is equal to the  initial abstraction of rainfall  that would  have occurred in the predevelopment condition.
                                                      162

-------
LID site designs maximize the use of small retention practices distributed throughout the site at the source to provide the
required volume storage. The required storage volume will be reduced when the change in the pre and post CN is
minimized.

    Retention storage allows for a reduction  in the post development volume and the  peak runoff rate. The increased
storage and  infiltration capacity of retention LID BMP's allow the predevelopment volume to be maintained. The most
appropriate  retention  BMP's include:

    1. bioretention  cells (rain gardens),

    2. infiltration trenches,

    3. water use storage  (rain  barrels  and gray water uses) and,

    4. roof top storage.

    Other possible retention BMP's include retention ponds, cisterns and irrigation  ponds but it may be difficult to distribute
these types of controls  throughout a development site.

    As retention storage volume is increased there is a  corresponding decrease  in the  peak runoff rate, in addition to
runoff volume reduction. If a  sufficient  amount of runoff is stored, the peak runoff rate may be reduced to a level at or
below the predevelopment  runoff rate. This storage may be all that is necessary to control the  peak runoff rate when there
is a small change in CN. However, when there is a large change in CN, it may be less  practical to achieve flow  control
using volume control  only.

Potential Requirement for  Additional Detention Storage

    In cases where very large changes in CN cannot  be  avoided, retention storage  practices alone may be either
insufficient to maintain the predevelopment runoff volume  or peakdischarge rates or require too much space to represent
a viable solution.  In these  cases, additional  detention storage will be needed to maintain the predevelopment peak runoff
rates. A number of traditional detention  storage techniques are available that can be integrated into the site planning and
design process for a  LID  site. These techniques include:

    1. swales with  check dams,  restricted drainage pipes,  and inlet/entrance controls,

    2. wide,  low gradient swales,

    3. rain   barrels/cisterns,

    4. rooftop storage and

    5. shallow parking  lot/road storage.

Determination of Design Storm  Event

    The hydrologic approach  of LID is  to  retain the same amount of rainfall within the development site as was retained
prior to any development (e.g., woods  or meadow in good condition) and then release  runoff as the woods or meadow
would have. By doing so,  it  is possible to mimic,  to the greatest extent practical, the predevelopment hydrologic regime
to maximize  protection of  receiving waters, aquatic ecosystems and  ground water recharge. This approach allows the
determination of a design storm volume that is  tailored to the unique soils, vegetation and topographic characteristics of
the watershed. This approach is particularly important in watersheds that are critical for ground water recharge to protect
stream/wetland base flow  and ground  or surface water supplies.
                                                      163

-------
 LID  BMP's

    Site design techniques and  BMP's  can be organized into  three  major categories as follows;  1)  runoff prevention
 measures designed to minimize impacts and changes in  predevelopment CN and Tc, 2) retention  facilities that store runoff
for infiltration, exfiltration or evaporation and  3) detention facilities that temporarily store runoff and  release through a
 measured outlet. Table 1, below, lists some of a wide array of LID BMP's and their primary functions. Placing these  BMP's
 in series and uniformly dispersing  them throughout the site provides the maximum  benefits for  hydrologic controls.

Table 1. Examples of LID BMP's and Primary Functions
BMP

Bioretention
Infiltration Trench
Dry Wells
Roof Top Storage
Vegetative Filter Strips
Rain Barrels
Swale and Small Culverts
Swales
Infiltration Swale
Reduce  Imperviousness
Strategic Clearing / Grading
Engineered Landscape
Eliminate Curb and Gutter
Vegetative Buffers
Runoff
Prevention
     X
     X
     X
     X
     X
                                                  Detention
                   X
                   X

                   X
                   X
                   X
                   X
Retention

    X
    X
    X
    X
Conveyance
                   X
                   X
                   X
Water
Quality
    X
    X

    X
    X

    X
    X
    X
                                X
                                X
Water  Quality

    LID  maximizes the use of the developed landscape to treat stormwater runoff. Not only can the landscape be used
to store,   infiltrate  and detain  runoff,  the  unique  physical,  chemical and  biological pollutant  removal/
transformation/immobilization/detoxification  capabilities of the  soil,  soil microbes and plants can be  used to  remove
pollutants from runoff.  For example,  bioretention basins or rain gardens are designed  to  use the  upland soil/microbe/plant
complex to  remove pollutants from runoff.  Rain gardens which look and function like  any other garden except they treat
runoff are designed with a layer of 2-3  inches of mulch, 2-3  feet of planting soil and vegetation (trees shrubs and flowers).
Figure 1 shows a parking lot landscape island  rain garden (bioretention  practice) that uses a high rate filter media with
plants to filter and treat 90%  of the annual volume  of runoff from the parking lot.
Figure 1. Parking Lot Rain Garden.
                                                         164

-------
    Studies conducted  by the University of Maryland have shown rain gardens to  be very effective in removing  pollutants.
The percent pollutant  removal  of various contaminants is  shown  below in  Table  2.  The results shown represent  the
average  removal rates under a wide variety of flow rates  and pollutant concentrations.

Table 2. Percent Pollutant Removal by Rain Gardens

       cu           Pb          Zn           P          TKN        NH4+          NO3-        TN*
       93           99           99           81          68          79           23          43

.  Removal varied as a function of depth in the soil. Percent removal shown is at a depth of approximately 3 feet.
Testing Conducted by the University of Maryland, Department of Engineering


    The  variety of physical,  chemical and biological pollutant  removal  mechanisms  available in the  complex rain garden
system is staggering. A description or explanation in any detail of these  mechanisms is beyond the scope of this paper.
A more detailed  description can  be found in the 1998 "Optimization of Bioretention Design" study conducted by the
University of Maryland. Mulch has been found to be very effective  in removing heavy metals through organic complexing
with the  hydroxyl and  carboxyl sites on  the organic molecules. Soil bacteria can  metabolize (use as a carbon energy
source) oil, grease and gasoline into C02 and water in the presence  of adequate nutrients and  oxygen. Soil bacteria have
been used for years for the remediation  of contaminated soils. Plants are known to uptake,  transpire,  accumulate and
detoxify heavy metals and many other toxic compounds.  The  physiologic and metabolic processes of plants are  used  to
clean contaminated soils through phytoremediation. A goal of LID is to maximize the  use of upland landscape with its soil/
microbes/ plant complex to treat runoff.  Using  upland systems to  trap and remove  pollutants  allows one to more easily
control the fate of contaminants and prevent them from entering the water column  where they are  almost  impossible to
contain  and remove.

Public  Outreach  and Pollution  Prevention

    Pollution prevention  and maintenance  of on-lot LID BMP's are two key  elements  in  a  comprehensive  approach.
Effective  pollution prevention measures can reduce  the  introduction  of pollutants to the environment and extend the life
of LID treatment BMP's.  Public education is essential to successful pollution prevention and  BMP maintenance. Not only
will effective public education complement and enhance BMP effectiveness, it can  also be used as a marketing tool  to
attract environmentally conscious  buyers,  promote  citizen stewardship,  awareness and  participation in environmental
protection programs and  help to build a greater sense of community  based  on common environmental objectives  and the
unique character  of LID designs.

Costs

    LID case studies and pilot programs  show that  at least a 25% reduction in  both site development and  maintenance
costs can be  achieved by reducing grading and the use  of pipes,  ponds, curbs and  paving.  In one subdivision called
Somerset which used the rain  garden LID technique for water quality  controls, the  developer  saved $4,500 per  lot or a
total of $900,000  by eliminating the need for curbs,  ponds and drainage  structures.  Maintenance costs are also reduced
in scale  and magnitude by using the small LID practices. LID site designs require only routine  landscape care and
maintenance of the vegetation, This eliminates the high costs  of pond  maintenance associated with dam  repairs and
dredging.

Road Blocks  to LID

    In the development and  acceptance  of the LID site  planning approach,  a number of roadblocks  had to be overcome.
Regulating agencies, the development community and the public all  had  concerns about the use of  new technology. The
                                                      165

-------
LID design manual  represents the culmination of four years of work to address all of these concerns and issues. Some
of the major components  of the LID approach, which addressed the many concerns,  include:

    1. develop an hydrologic  analytical  methodology to demonstrates the equivalence of LID to conventional  approaches,

    2. develop new road  standards which  allow for narrow  roads, open  drainage and cluster  techniques,

    3. streamline the review process for innovative LID designs which allow easy modification of site, subdivision, road
       and  stormwater requirements,

    4. develop a public education process which informs  property owners on  how to prevent pollution and  maintain on
       lot BMP,

    5. develop legal and  educational  mechanisms to  ensure BMP's  are  maintained,

    6. demonstrate the marketability of green  development,

    7. demonstrate the cost benefits of the LID approach,

    8. provide  training for regulators,  consultants, public  and political leaders  and,

    9. conduct research to demonstrate the effectiveness of bioretention BMP's.

Summary

    LID is a viable economically sustainable alternative  approach  to stormwater  management and the  protection of natural
resources. LID provides tangible incentives to a developer to save natural areas and  reduce stormwater and roadway
infrastructure costs.  LID can achieve greater natural conservation by using conservation as a stormwater  BMP to reduce
the change in CN. As more natural areas are saved,  less runoff is generated and stormwater management  costs are
reduced. This allows  multiple use and  benefits (environmental and economical) of the resource.

    Additionally, developers have incentives to reduce  infrastructure costs by reducing impervious areas,  and eliminating
curbs/gutters  and stormwater ponds to achieve  LID stormwater controls.  Reduction of the infrastructure also reduces
infrastructure maintenance burdens making LID designs more economically sustainable.  Superior protection of aquatic
and riparian  ecosystems can be achieved since a LID developed watershed functions in a hydrologically similar manner
as the predevelopment conditions. Recreating  the predevelopment hydrological regime is  a  better  way to protect the
receiving waters  than  the conventional end-of-pipe  mitigation approaches.

    LID promotes public awareness, education and participation in environmental protection. As every property owner's
landscape functions as part of the watershed, they must be  educated on the benefits and the need for maintenance  of
the landscape and pollution prevention. LID developments can be designed  in  a very environmentally sensitive manner
to protect streams,  wetlands,  forest  habitat, save energy, etc. The unique character of a LID green development can
create a greater  sense of community pride based on environmental  stewardship.

References

Prince George's County, Maryland. Department  of Environmental Resources. Low-Impact Development Design Manual.
(PGC, 1997)

Prince George's County,  Maryland. Department of Environmental Resources. Low-Impact Development Guidance
Manual. (PGC, 1997).

Prince George's County, Maryland.  Design Manual for use  of Bioretention  in stormwater Management. (PGC, 1993).
                                                      166

-------
Prince George's County, Maryland. Optimization of Bioretention Design for WaterQualityand HydrologicCharacteristics.
(Davis, et al.  1998).
                                                     167

-------
     A National Menu of BMPs for the Phase II NPDES Storm Water Program


                                             James H. Collins
                                          Tetra Tech,  Incorporated
                                                Fairfax, VA

                                            John A. Kosco, P.E.
                                  USEPA/Off ice of Wastewater Management
                                             Washington,  DC


Introduction

    Implementation of the Phase II NPDES Storm Water Program, which is presently "proposed," will be enhanced by
the development of several tools used to help reduce discharges of pollutants from regulated small construction activities
and from regulated small municipal separate storm sewer systems (MS4s). One of the tools being  developed by EPA
is a national menu of best management practices (BMPs), from which regulated  Phase  II municipalities can select as they
develop their stormwater management  programs.'  The purpose of the national menu of BMPs is to provide a list of
options available  to regulated Phase  II  municipalities as they develop  a stormwater  programs.  The national  menu of
BMPs may be adopted or modified by each  NPDES permitting authority, or  the permitting authority may develop its own
menu of BMPs for use  by the Phase II municipalities under its jurisdiction. This paper describes the process of developing
the national menu of BMPs and measurable goals for each of the six minimum measures required to be in the stormwater
management  programs.

Process  of Development

    The process used in developing the menu was to first list appropriate BMPs for each minimum control measure, with
subcategories under certain control  measures.' Then, a basicformat for presenting the  information about each BMP was
established. Information being provided for each  BMP  in the menu consists of BMP  name, description,  an illustration,
applicability and design considerations, limitations, operation  and  maintenance,  effectiveness, cost, and references. The
menu is being prepared by  EPA, with support from Tetra Tech, Incorporated, and the Center for Watershed Protection.
A peer  involvement/peer review group has been selected  and will provide review and input to the process of developing
the menu over the course of the next year.  The  menu is currently being reviewed  and developed as a traditional hard
copy document.  Following development, there are plans to make the menu available as an interactive Web-based tool.
The menu of BMPs is  scheduled to be released  by October, 2000.

Descriptions of Six Minimum Control Measures, with Lists of BMPs

1) Public Education and Outreach

    This measure in the proposed rule calls for the creation  of a public  education program to inform citizens about the
impacts  that stormwater  runoff can have  on water quality.  It includes  the preparation and distribution  of educational
materials to the community, describing these impacts and steps that can be taken to reduce pollution from discharges
of stormwater. Examples of such steps include  proper septic system  maintenance, limitations on use and runoff of
household and garden  chemicals,  proper disposal of used motor oil or  household  hazardous wastes, and  involvement
in local stream restoration activities. The following BMPs have been  identified (in the Draft Rule) under four  major
subcategory groups in  the menu for the  Public Education and Outreach minimum measure:

    Public outreach/education for homeowners

    . Lawn and garden activities, including  proper pesticide  use  and disposal practices


                                                   168

-------
    .  Water  conservation practices
    .  Proper disposal of used materials or household  hazardous wastes
    .  Pet waste  management
    Targeting public outreach/education
    •  Education/outreach for  commercial activities  (parking lots,  gas stations, etc.)
    .  Tailoring  outreach programs to  communities, including minority and  disadvantaged, as well as children
    .  Classroom  education on stormwater
    .  Distributing  stormwater  educational materials (how,  to whom)
    Public Outreach Programs for New Development
    .  Low-impact development (includes buyer awareness,  legal documents, and settlement documents)
    Pollution Prevention Programs for Existing Development
    .  Educational display, pamphlet,  booklet,  utility stuffer
    .  Using the media  (includes  newspaper, magazine, radio, television,  public service announcements,  and Internet
      messages)
    .  Promotional giveaway
2) Public Participa tion/lnvolvemen t
    This  measure in the proposed rule includes  compliance  with state  and local public notice requirements,  but goes
beyond that to  encourage municipalities to seek  public  involvement in the  development and review of their stormwater
programs. Opportunities  for members  of the public to participate in the development of their municipality's stormwater
management program may include serving as citizen representatives on a local stormwater management panel,  attending
public hearings,  working  as citizen  volunteers to educate  others about the program, assisting in program coordination with
other  pre-existing  programs,  or participating  in volunteer  monitoring efforts. The following BMPs have  been identified
under two major subcategory groups of practices in  the  menu for  the Public Participation/Involvement minimum  measure:
    Activities/Public  Participation
    .  Storm drain stenciling
    .  Stream cleanup
    .  Volunteer monitoring
    .  Reforestation program
    .  Wetland  plantings
    .  Adopt-A-Stream  program
    Involvement/Public  Opinion
    .  Watershed  organization
                                                      169

-------
    .  Stakeholder meetings (includes local  stormwater  management  ponds)

    .  Attitude surveys

    .  Community  hotlines

3) Illicit Discharge Detection and Elimination

    This measure envisions the creation of an illicit discharge detection and elimination program. Specific program
elements include developing a demonstrated knowledge  base of the MS4, using maps or other documents to identify
major outfalls and pipe networks on a topographic basis; developing a plan to address illicit discharges into the MS4,
including appropriate enforcement procedures to the extent allowable by law; and developing a process for informing the
public about the hazards associated with illicit  discharges and  the improper disposal of waste.  For example, recycling
programs and other public outreach activities could be developed to address sources of illicit  discharges,  including used
motor oil, antifreeze,  pesticides,  herbicides, and fertilizers. The following BMPs have  been identified for the menu for the
Illicit Discharge Detection  and  Elimination minimum  measure:

    .  Failing septic  systems

    .  Industrial  connections

    •  Recreational  sewage

    . Sanitary sewer overflows

    .  Identifying  illicit  connections

    .  Wastewater connections to the storm drain system

4) Construction Runoff Control

    This measure provides for the enforcement  of a program to  reduce  pollutants in storm water runoff from construction
activities resulting in the disturbance of one acre to five acres of land. The  program would apply to the individuals
responsible for activities at construction sites and should  include an  ordinance to control sediment and erosion; a
mechanism to ensure control of other wastes at construction sites, such as discarded building materials,  concrete truck
washout, and sanitary waste that could impact  water quality;  requirements for the implementation of appropriate BMPs,
such as silt fences,  temporary detention  ponds  and hay bales; provisions for preconstruction review of site management
plans; procedures for receipt and consideration of comments and other information  provided by the public; regular
inspections during construction;  and penalties to ensure compliance. The following BMPs have been identified  under 11
major subcategory groups of practices in the menu  for the Construction Runoff Control minimum  measure:

    Minimize  Clearing

    .  Land grading

    .  Permanent diversion

    .  Preservation of natural vegetation (includes tree  preservation  and protection)

    Stabilize Exposed Soils

    .  Chemical  stabilization

    .  Mulching
                                                      170

-------
.  Permanent  seeding
. Sodding
. Soil roughening
Protect Steep Slopes
. Geotextiles
.  Gradient terraces
.  Soil  retention  (includes slope stabilization,  retaining wall,  reinforcement)
. Temporary slope  drain (a.k.a. - pipe slope drain)
.  Temporary storm drain diversion
Stabilize Drainage  Ways
.  Check dam  (a.k.a.  grade stabilization  structure)
.  Filter  berm
.  Grass-lined  channel
. Riprap
Protect  Waterways
.  Temporary  diversion
.  Temporary stream  crossing  (bridge, culvert)
.  Vegetated buffer
Phase  Construction
.  Construction  sequencing
. Dust control
Install  Perimeter Controls
.  Temporary diversion dikes,  earth dikes,  and interceptor dikes (includes temporary fill diversions)
.  Sand fence  and wind  fences
. Silt fence
.  Brush barriers
 Install  Sediment Trapping  Devices
.  Sediment basin/rock dam
.  Sediment filters and sediment chambers
                                                   171

-------
    . Sediment traps
    Inlet Protection
    . Stabilized  construction  entrances
    . Storm drain inlet protection (includes block and gravel, excavated drop, fabric drop, and sod drop inlet protection)
    Education  and Awareness
    • Contractor certification and inspector training
    . BMP inspection and  maintenance
5)  Post-Construction  Runoff Control
    This measure uses post-construction  controls as  part of a  program  to address stormwater runoff from  new
development  and  redevelopment projects using appropriate structural and  non-structural BMPs. Non-structural BMPs
are preventive actions using management and source controls, such as  policies and ordinances that  result in protection
of natural resources  and prevention of runoff.  Non-structural BMPs might include requirements  that encourage growth
in identified areas while protecting sensitive areas such as wetlands and riparian zones,  minimizing impervious surfaces,
maintaining open  space, and  minimizing  clearing,  grading, or other disturbance  of soils and  vegetation.  Some of the
typical structural BMPs  include storage practices (wet ponds, extended-detention dry ponds, or other storage  facilities
with outlets);  infiltration  practices  (infiltration basins, infiltration  trenches, and  porous pavement);  and filtration practices
(grassed swales, sand filters, and vegetated filter strips). This measure  should also ensure effective and  reliable
performance by providing for the  long-term operation and maintenance of the selected BMPs. The following BMPs have
been identified  under eight  major subcategory groups of practices for the  Post-Construction  Runoff Control minimum
measure:
    Ponds
    . Extended detention dry basin or pond  (with or without permanent pools or shallow marshes near the  outlet),
      includes  tank  storage
    • Wet pond
    Infiltration  Practices
    • Infiltration basin  (a.k.a. - recharge basin)
    . Infiltration trench  (a.k.a. - infiltration galley)
    . Porous  pavement
    Filtration Practices
    .  Bioretention
    . Filters,  including organic media filter (peat sand or compost-type), sand  filter, multichamber treatment train  (MCTT)
      system,  and inlet filtration  systems
     Vege ta five Practices
    . Constructed wetland,  shallow marsh
    . Grassed  swale
                                                      172

-------
    . Vegetative filter strip
    Runoff Pretreatment Practices
    . Catch  basin
    . In-line  storage,  includes flow  regulator  information
    . Manufactured  systems for water quality inlets
    Experimental  Practices
    • Alum injection  system
    On-lot Treatment
    . On-lot  treatment includes information on dry wells, roof downspout systems, rain barrels, exfiltration storage
      systems, french drains, and dutch drains
    Better Site Design
    . Conservation  easements
    . Infrastructure  planning
    . Buffer  zones/setbacks
    . Open  space development
    . Narrow streets
    . Curb elimination
    . Green  parking  lot
    .  Alternative turn-around
    . Urban  forestry
    . Alternative pavers
6)  Pollution  Prevention/Good Housekeeping
    This  measure in the proposed  rule envisions the  creation of an  operation and maintenance/training program to prevent
or reduce pollutant runoff from municipal  operations. The program should  include training for municipal staff to address
prevention measures in government operations, such as park and  open space maintenance, fleet maintenance, planning,
building  oversight and stormwater  collection system maintenance. Other  possible pollution prevention activities that might
be relevant include controls  for reducing or eliminating the discharge of pollutants from streets, roads, highways, municipal
parking  lots,  maintenance and storage yards, and waste transfer  stations;  programs to promote recycling and pesticide
use information; procedures for proper disposal  of waste  removed from municipal  systems and public areas (such  as
streets)  including dredge  spoil, accumulated sediments, floatables,  and other debris;  and new flood  management projects
to assess the impacts on water quality and examine exiting projects to determine  if they need additional water quality
protection devices or practices. The following  BMPs  have been identified  under two major subcategory groups of
practices in the menu for  the Pollution  Prevention/Good  Housekeeping minimum  measure:
                                                      173

-------
    Source con trols
    .  Animal waste collection
    .  Automobile  maintenance
    .  Car washing
    .  Illegal dumping  control
    .  Landscaping and lawn care
    .  Pest control
    .  Parking lot and  street cleaning
    .  Roadway and  bridge maintenance
    .  Septic system  controls
    Materials Management
    .  Alternative  products
    .  Hazardous  materials  storage
    .  Household  hazardous waste collection
    .  Road  salt  application  and storage
    .  Spill response  and prevention
    .  Used oil recycling
Conclusion
    As part of the Stormwater Phase II Tool Box, the Menu of BMPs should help municipalities develop, implement, and
enforce the Phase  II  program. The menu will be available in  time for regulated municipalities to use in  complying with
stormwater management program requirements  under  Phase 2 permits  and might  also benefit other jurisdictions and
individuals.
References
USEPA, 1999. Storm Water Phase II Fact Sheet Series 1-4. April 1999. EPA 833-F-99-001 through 015.
USEPA, 1998. 40  CFR Parts  122 and 123; National  Pollutant Discharge Elimination System-Proposed  Regulations  for
Revision of the Water Pollution Control  Program Addressing  Storm Water Discharges; Proposed Rule.  Friday January
9, 1998.63  FR 1536.
                                                    174

-------
                    Determining Urban Stormwater BMP Effectiveness

                                                 Eric Strecker
                                         URS Greiner Woodward Clyde
                                                 Portland,  OR

                                              Marcus M. Quigley
                                         URS Greiner Woodward Clyde
                                                  Boston, MA

                                               Ben R.  Urbonas
                                   Urban Drainage and  Flood Control  District
                                                  Denver, CO


Abstract

    The overall purpose this US EPA funded cooperative research program with the American Society of Civil Engineers
(ASCE) is to develop a  more useful set of data on the effectiveness of individual best management practices (BMPs) used
to reduce pollutant discharges from urban development. BMP performance data gathered at a particular site should not
only be useful for that site, but also be useful for comparing studies of similar and different types of BMPs in other
locations. Almost all BMP effectiveness  studies in the past  have  provided very limited data that is useful for comparing
BMP design and selection among individual BMP types (e.g. sand filters).  This paper overviews some  of the  problems
of past BMP effectiveness studies from  the  perspective of comparability between studies. It suggests some of the ways
that data  should be collected to make it  more useful for assessing factors (such as settling characteristics of inflow solids
and  physical features  of the BMP) that might  have led to the  performance levels achieved. It briefly presents the
database that has been developed by this project, which not only  serves as a tool for storing data from  existing studies,
but as a tool for entering and  storing data  collected  from future  studies. Discussed are considerations that affect data
transferability, such as  effectiveness  estimations, statistical testing, etc. It overviews the efforts to establish and analyze
the data base for existing studies and overviews  proposed analyses for the  future, when more studies that  have followed
the protocols are available. The database has specifically pointed out the need for  additional BMP performance studies,
as the current  data is very sparse in terms  of studies that have recorded enough information to be  useful in assessing
BMP  type  performance.

Introduction

    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. However, in
reviewing and  summarizing the  information gathered from  these individual  BMP evaluations,  it is apparent that
inconsistent study methods and reporting  make wider-scale assessments difficult, if not impossible. For example,
individual studies often  included the analysis  of different constituents and utilized different  methods for data  collection and
analysis,  as well as varying degrees of information on  BMP design and inflow characteristics. Just the  differences in
monitoring strategies and  data evaluation alone contribute significantly to the range  of BMP "effectiveness"that has been
reported.  These differences make combining these individual studies almost impossible to assess what design factors
may have contributed to  the variation in performance  (Strecker  et al.,  1992). Urbonas  (1994 and  1995) and  Strecker
(1994) summarized information that should  be recorded  regarding the physical, climatic,  and geological  parameters that
likely  affect the performance of a BMP and considerations regarding sampling and analysis methods.

Efficiency, Effectivness, and Performance

     In order to better  clarify the terminology used to describe the level  of treatment achieved and  how well a device,
system, or practice meets  its goals, definitions of  some terms often used loosely in the literature are provided here. These
terms help to better specify the scope of monitoring  studies and related analyses:

                                                     175

-------
    . Best Management Practice (BMP) -A device,  practice, or method for removing, reducing,  retarding, or preventing
      targeted stormwater  runoff constituents, pollutants,  and contaminants from  reaching  receiving  waters

    . BMP System - A BMP system includes the BMP and any related bypass or overflow. For example, the efficiency
      (see below) can be determined for an offline retention (Wet) Pond either by itself (as a BMP) or for the BMP system
      (BMP including  bypass)

    . Performance - measure of how well a BMP meets its goals for stormwater that is treated by the BMP

    . Effectiveness - measure of how well a BMP system meets its goals  in relation to all stormwater flows

    . Efficiency - measure  of how well a  BMP or BMP system removes  pollutants

    The ASCE project team is working with available data to determine efficiency of BMPs and BMP systems.    In
addition, effectiveness and  performance are being evaluated, acknowledging the limitations of existing information about
the goals  of specific BMP  projects.  Quantification of efficiency only evaluates a portion  of the overall performance  or
effectiveness  of a BMP or BMP system.  Calculation of the efficiency helps to  determine additional measures  of
performance and effectiveness, for example the ability of a BMP to meet any regulatory goals. A list of typical goals and
the current ability of the ASCE/EPA  project to help  evaluate them is shown in Table 1.

Problem: BMP Performance Study  Inconsistencies

Studies of BMP effectiveness have  utilized significantly different:

    . sample  collection techniques  (e.g., from sample collection types-grab, composite, etc., flow  measurement
      techniques, to how the sample was composited, etc.);

    . water quality  constituents,  including: chemical species,  methods (detection limits), form (e.g., dissolved vs. total,
      vs.  total recoverable,  etc.), and treatment  potential;

    . data reporting on  tributary watershed and BMP design characteristics ( e.g., tributary area or watershed attributes
      such as percent impervious, land use  categories, rainfall statistics,  etc.);

    . effectiveness estimation techniques (there are  at  least four common techniques that have  been utilized to assess
      effectiveness that can cause significant differences in pollutant removal reporting, with the  same set of data), and
      potential alternatives  to reporting just  concentration/loading  reductions;  and

    . statistical validation of results (typical lack of statistical tests to determine if the reported removal efficiency can in
      fact be shown to  be  statistically different than zero).

      Monitoring  strategies  that could be  employed to  monitor BMP effectiveness  include:

    . New BMP installation with new development- input/output (e.g.,  monitor  new  detention  pond of newly developed
      watershed  and evaluate inflow concentrations/loads vs.  outflows)  or conduct a "control"  watershed  comparison

    . Retrofit of  existing or  new single BMP within  existing watershed-input/output,  and/or, before/after (e.g.,  retrofit of
      an existing flood control basin for water quality)

    . Watershed-wide  new structural or non-structural-"control" watershed comparison (e.g., new BMP catch basins
      in developing area)

    . Watershed-wide  structural retrofit or application of non-structural  - before/after,  and/or, "control" watershed
      comparison (e.g., catch basin retrofit  on watershed scale)

    Input/output monitoring  is the typical approach  utilized. However,  control watersheds and before/after  approaches
have also  been employed. All of the  other potential factors that could be contributing to differences in performance must
identified and accounted. On the other hand, it is beneficial to be able  to show that  a watershed-wide difference is or  is


                                                      176

-------
   Table 1. Goals of BMP Projects and the Ability of the National Stormwater BMP Database to Provide Information Useful for Determining
   Performance and Effectiveness
   Goals of BMP Projects
                                                                                                               Ability to Evaluate
                                                                                                            Performance and Effective
        Category
    Hydraulics

    Hydrology

    Water Quality
    (Efficiency)
   Toxicity


   Regulatory


   Implementation  Feasibility


   Cost

"Imfftesttiappearance of site

   Maintenance


  •    Uangg* functionality

   Resources



   Safety, Risk and Liability


   Public Perception
'Improve flow characteristics upstream and/or downstream of BMP

'Flood mitigation, improve runoff characteristics (peak shaving)

'Reduce downstream pollutant loads and concentrations of pollutants
'Improve/minimize downstream  temperature  impact
•Removal of litter and debris

'Reduce acute toxicity of runoff
•Reduce chronic toxicity of runoff

'Compliance with NPDES permit
*Meet local, state, or federal water quality criteria

*For non-structural  BMPs, ability to function within management and oversight
structure

'Capital, operation,  and maintenance  costs
'Operate within maintenance, and repair schedule and requirements
'Ability of system to be retrofit, modified or expanded
'Improve downstream aquatic environment/erosion control
'Improve wildlife habitat
'Multiple use functionality

'Function without significant risk or liability
•Ability to function with minimal environmental risk downstream

'Information is available to clarify public understanding of runoff quality, quantity
and impacts on receiving waters
/
/
/'

/1
   / can be evaluated using the ASCE/EPA Database as information source
   /1 will be able to be evaluated using the database as primary source of information after enough studies have been submitted
   /2 can be evaluated using the database the primary source of information combined with a secondary source of comparative data
    not being  detected  with BMP  implementation. These  differences in monitoring approach certainly  effect  the ability to
    compare  studies.

        Any of the above topics would require an in-depth discussion  beyond the scope of this paper. Therefore, this paper
    will present only  a brief overview  of  each  and some potential solutions for improving  how data is collected. The ASCE
    project team has  developed a set of protocols and  a database on  BMP  performance studies with the purpose of improving
    the consistency  of  BMP  monitoring  information. The  project includes:

        .  Developing protocols  for  BMP monitoring  and  reporting

        .  Developing a  database on  BMP performance studies

        .  Conducting an  evaluation of existing information to assist EPA in  providing guidance to the regulated  community
                                                                  177

-------
    The database specifies a chosen  set of reporting  information, but does not guide  users on  how to develop such
information. For example, it does not specify in detail what a flow-weighted composite sample is and how it should be
collected. The next step beyond the EPA protocols and database effort should be a guidance document on monitoring
data collection strategies and techniques to improve their consistency and  ultimate transferability. A few of the issues
related to proper guidance are discussed in the next two sections. It should  be recognized that, with the development of
the database  and the protocols, it will be a number of years (5 to 10) before a significant number of new studies on BMPs
are conducted utilizing the protocols. Therefore, a  rigorous evaluation of BMP selection  and design  factors  will need to
take place in the  long-term future.

Recommended  Parameters for Assessing  BMP Performance

    In  developing a method  for quantifying BMP performance, it is helpful to look at the objectives of previous studies
seeking such  a goal.  BMP performance studies usually  are conducted to obtain information regarding one or more of the
following  objectives:

    .  What degree  of pollution  control does the BMP  provide  under typical operating conditions?

    .  How  does performance vary from pollutant  to pollutant?

    .  How  does performance vary with  various input  concentrations?

    .  How  does performance vary with large  or small  storm events?

    .  How  does performance vary with  rainfall intensity?

    .  How  do design variables affect performance?

    .  How  does performance vary with different  operational and/or maintenance approaches?

    . Does  performance improve,  decay, or remain  the stable  over time?

    . How does the BMP's performance compare relative to other BMPs?

    .  Does the BMP reduce  toxicity to acceptable levels?

    .  Does the BMP cause  an improvement  in downstream biotic communities?

    .  Does the BMP have potential downstream  negative  impacts?

    The monitoring  efforts implemented most typically seek to answer a subset of the above questions. This often leaves
larger questions about the performance of the BMP, and the  relationship between  design and performance, unanswered.
Standardization of BMP data collection and evaluation methods (i.e.,  guidance and the ASCE/EPA database) allows this
broader set of questions to be examined.

    There has been a  very  wide  variety of pollutants analyzed in BMP and characterization studies.  The protocols
established  under the EPA-funded  cooperative research  program recommend a  standard set of  constituents for  BMP
testing  programs. Table  2  presents the recommended  constituents developed  from the  review of previous studies with
an  understanding of costs and likelihood of providing meaningful results.  A discussion  of how these constituents  were
selected and  a detailed description  of each can be found in  Strecker (1994).

     There are some practical and technical considerations  regarding data  reporting which would  facilitate  data
usefulness,  including consistent formatting of data,  the clear indication of QA/QC results, standard  comparisons to water
quality  criteria, reporting of tributary watershed characteristics,  and BMP design information. The  last  two items are
considered  critical for evaluation of what contributed to BMP effectiveness  in  one location over  another.

    Data Reporting,  It is recommended  that  all  constituent concentration data be reported as event mean  concentrations
(EMCs). These statistics should be based on use of the lognormal distribution. The NURP and  FHWA studies (EPA,

                                                     178

-------
Table 2. Recommended Standard Analytical Tests for Urban Stormwater BMP Assessment

                                                     EPA Method                                           Detection
Lab Analysis	Number	Limit (mg/l)
Conventional
TSS                                                 EPA 160.2                                              1
TDS                                                 EPA 160.1                                              1
TOO                                                 EPA 415.1                                              3
COD                                                EPA 410.4                                              1
Total Hardness                                        SM314-A                                              1

Nutrients
(NH3-N)                                             SM417-AD                                             0.1
Total phosphorus (as P)                                  SM 424-CE                                             0.005
Ortho-phosphate (as P)                                  SM 424-E                                              0.05
Nitrate + nitrite (NO, + NO2 - N)                            EPA 353.1 or .2                                         0.05

Total Metals
Cd (cadmium)                                         EPA 7131                                              0.0002
Pb (lead)                                             EPA 7421                                              0.0003
Cu (copper)                                           EPA 6010                                              0.001
Zn(zinc)                                              EPA 6010                                              0.001

Dissolved Metals
Cd (cadmium)                                         EPA 7131                                              0.0002
Pb (lead                                              EPA 7421                                              0.0003
Cu (copper)                                           EPA 6010                                              0.001
Zn (zinc)	EPA 6010	0.001

1983a; Driscoll et  al., 1990) identified that the lognormal distribution is suitable for characterizing EMC distributions. The
high degree of variability is why proper statistical techniques should be employed to evaluate whether a measured
difference between BMP before/after  or input/output is significant. The recommended inclusion of outlet data as  a part
of any paper or  report will allow comparisons of typical outlet concentrations and may allow the determination of the  lowest
or average expected  concentration from  a particular type of BMP.  For example, it may be that wet ponds may only be
able to treat  to some minimum concentration range at the outlet  and the "effectiveness" is greatly impacted by the inlet
concentrations.

  Quality Assurance/Quality Control (QA/QC). All monitoring studies should include a QA/QC program. The details and
results of the QA/QC program should be reported  in monitoring study reports and summarized in papers. It is especially
important to  discuss when data are characterized as estimates due  to QA/QC results and when detection  limits were
affected.  Too often this information  is not included.

    Comparisons to Water Quality Criteria. A method to gage effectiveness could be to monitor how the BMP affects
the number of times (frequency) that EPA water quality criteria are exceeded in both the inflow and the outflow, to assess
how the  BMP reduces  (or does not reduce) the frequency  of potentially toxic events.  For heavy metals analyses, it  is
recommended that hardness be collected for all  storms  monitored and that comparisons to criteria be made utilizing the
dissolved  fraction with the  computed  aquatic criteria as modified by EPA (1993b).

     Watershed and BMP Design Parameters. Table 3 presents  a summary  of these  parameters.  These parameters
(more detailed parameter lists are available on ASCE's Web page at http://www.asce.org/peta/tech/ nsbdOl .html) have
been selected with the purpose of being able to utilize this information to evaluate what BMP design attributes and
tributary watershed characteristics can be linked to BMP effectiveness  information.

    The  primary goals of the ASCE/EPA database development process were to facilitate efficient  data  entry, provide
useful queries of stored data, and output information in  a comprehensive and applicable  manner through  a  user-friendly
interface.  The database was written in Microsoft© Access incorporating Access relational database engine and features
and the Visual Basic®for Applications programming language for  customization of the functional aspects of the front end.
Distribution will take place initially as  an Access run-time version on  CD-ROM, but will be available in the nearfuture over
the Internet.


                                                       179

-------
Table 3. Parameters to Report with Water Quality Data for Various BMPs

3arameter
Type
ibutary
atershed







eneral
ydrology












later








Parameter
Tributary watershed area, average slope,
average runoff coefficient, length, soil types,
vegetation types
Total tributary watershed impervious percentage
md percent hydraulically connected
Details about gutter, swer, swale, ditches,
larking, roads in watershed
•and use types (res., comm., ind., open) and
acreage
)ate and start/stop times for monitored storms
lunoff volumes for monitored storms
>eak 1 -hr intensity
)esign storm/flood recurrence intervals and
nagnitude
3eak flow rate, depth, and Manning's roughness
loefficient for the 2-year storm
)epth to seasonal high
jroundwater/impermeable layer
Saturated hydraulic conductivity, infiltration rate,
soil group
\verage annual values for number of storms,
>recipitation, snowfall, minVmax. Temp., from
appropriate weather station
Mkalinity, hardness and pH for each monitored
storm
A/ater temperature
Sediment settling velocity distribution, when
ivailable
:acility on- or off-line?
Bypassed flows during events

Retention (Wet)
Pond























t


.



Extended
Detention (Dry)
Basin























*


*



Wetland
Pond
Basin























<


<




Grass Swale/
Wetland Channel























<


<




Sand/Leaf
Compost Filter























<


<



Oil & Sand Trap/
Hydro-dynamic
Device























<


<




Infiltration and
Perc..























<


.



                                                                                    180

-------
eneral
rcility
ret Pool
etention
olume
retreatment
lant
Zetland
lant
Type and frequency of maintenance
Types and location of monitoring instruments
nlet and outlet dimensions, details, and number
Media or granular material depth, type, storage
volume, and porosity
/olume of permanent pool
.ength of permanent pool
Dermanent pool surface area
Jttoral zone surface area
Jolar radiation, days of sunshine, wind speed,
>an evaporation, from appropriate weather
station
detention (or surcharge) and flood control
volumes
Detention basin's surface area and length
Srimful and half-brimful emptying time
Sottom stage/infiltrating surface area and type
rorebay volume, surface area
Relationship to other BMP's upstream
/Vetland/swale type, surface area, and length,
side slope Ibottom width for swales and
;hannels)
Percent of wetland surface between 0-12", 12-
>4", and 24-48"
'lant species and age of facility






•












•







•







181

-------
Estimation of BMP Pollutant Removal Effectiveness and Effectiveness
    BMP pollutant removal effectiveness estimations are not straightforward and  a wide variety  of  methods  have  been
employed.  Martin and Smoot (1986) discussed three types  of methods to compute  efficiencies, including an  efficiency
ratio,  sum  of loads,  and regression  of loads.  Many  researchers  have utilized an efficiency measure based  upon storm
pollutant loads  into and out of the BMP  on a storm-by-storm basis.  This weights the  effectiveness considering that all
storms are  "equal" in computing the  average removal.  However, it  is readily apparent that all storm volumes and their
associated concentrations are not equal.

    One factor  that complicates the estimation of the  effectiveness is that for wet ponds and wetlands, (and  other BMPs
where there is a permanent pool), comparing effectiveness on a storm-by-storm basis neglects the fact that the outflow for
a particular event  being measured may have little or  no relationship to the inflow for that same event.   Based upon a
national characterization of rainfall (Driscoll, et.  al., 1989), if a basin were sized to have a permanent  pool equal to the
average storm,  about 60 to 70% of the storms would  be less than this volume.  Therefore, in  many cases, flows leaving
may have little  or  no relationship to  flows  entering the pond. Storm-to-storm  comparisons are probably not  valid.   It is
probably more  appropriate to  utilize statistical  characterizations of the inflow  and  outflow concentrations  to  evaluate
effectiveness  or, if enough samples are collected (i.e., almost all  storms monitored), to  utilize  total  loads into and out  of
the  BMP.

    Table 4  compares three of the  methods, including  percent  removal by storm with a statistical  characterization  of
inflow/outflow concentration and a simple comparison of total loads in and out for the sampled storms for an  example site.
The removals estimated differ by up to 18 percentage points.  In this record, there are several storm events where inflow
concentrations were  relatively low and therefore  the system was  not "efficient."  However, it was effective at  maintaining
the  effluent  quality.

Table 4. Comparison of BMP Pollutant Removal Efficiency Techniques


Storm
1
2
3
4
5
Volume of
Flow (ft3)
lnflow= Outflow
445,300
649,800
456,100
348,111
730,261
Concentration
In Out
(mg/l)
352 24
30 25
99 83
433 141
115 63
Load
In Out
(Ibs)
9780 670
1220 1010
2820 2360
9410 3060
5240 2870

% Removal
by storm
93%
17%
16%
67%
45%
Med 139 65 A
Cov 1.48 .86 28,470 9,970 V 48%
Mean 249 85 G
                                           Cone 66%
                                                                   Loads 65%
note: 1 lbm = 2.2046 kg and 1 ft3 = 0.028317 m3

    Based  on these factors, it is  recommended that a  statistical characterization of inflows vs. outflows  be utilized.   Use
of the log-transform  of EMCs is  recommended. Tests of the applicability of a log transform should be made to support the
transform of data when sufficient data is available.   Standard descriptive  statistics,  box-and-whisker plots,  and normal
probability plots of the transformed data for both the inflow and outflow should be employed to clearly demonstrate not only
the  differences  in the mean EMCs, but also  the effectiveness  of the BMP throughout the range of influent and  effluent
EMCs.  This approach  provides  the  ability  to  determine  whether any  apparent  differences in  inflow and outflow  EMC
populations are statistically greater than zero.  If enough  data  on storms is collected,  (e.g., continuous  samples over an
                                                       182

-------
extended period including base flow measurements where significant), the total loads in and out may also be an acceptable
method.  A graphical look  at  the distribution of contributing  storms will  often provide insight into the applicability of the
method, (e.g., do a small number of large storms dominate the resulting effectiveness value).

    The variability in runoff concentrations from  event to event is  large.   In attempting to statistically characterize a BMP
influent concentration (and  outflow), the more data the better. As  mentioned above, there  are a  number of types of BMP
evaluations  that can be conducted: (1) standard evaluation  of a single  BMP, testing  input and  output, (2) evaluation  of
multiple BMPs within a basin (before/after or control basin), and (3) evaluation of a BMP with multiple inlets (where it might
be very difficult and expensive  to  evaluate the BMP utilizing input/output). All  methods should require that  a rigorous
statistical approach be applied  in selecting the number of samples to be collected to help assure detection of a given level
of change.

    As an example of the number of samples required to detect  a  "true" difference, Table 5 presents an analysis of two  of
the Portland monitoring stations (WCC, 1993) where 10 flow-weighted composite samples were  collected. The Fanno Creek
station is a large (about 1,200 acres) residential catchment that is in an open channel, while the M1 station is a smaller
(about 100 acres) mixed land use  station that  is in a pipe.   An analysis  of a variance-based test was  utilized with the
existing data to determine how many samples are estimated to be  needed to detect a  5%, 20%,  and 50% change in the
mean concentration at the station. The test was performed considering an 80% probability that the difference will be found
to be significant, with a 5% level of significance (Sokal and Rohlf,  1969). This analysis does not consider potential seasonal
effects on  the  collection of data as a factor.  Even so, quite a large number of samples would be required to detect a 5%
to 20%  difference in concentrations.  In  many locations, given that  there may be only  10 to 20 storm  events per year that
are  large  enough to monitor,  it  would take a number of  years  of sampling all  storm events to  be  able to  detect small
differences.

Table 5.  Analysis of Sample Sizes  Needed to Statistically Detect Changes in Mean Pollutant Concentrations from 2 Stations in Portland, Oregon
     Monitoring Site
    R1 - Fanno Creek
      Residential

     M1 - NE 122nd
       Columbia
    Slough Mixed Use
 Parameter
   TSS
 Copper
Phosphorus
   TSS
 Copper
Phosphorus
   Number of Samples Required to Detect the Indicated %
          Reduction in Site Mean Concentration*
 5%                     20%                    50%
202                     14                      4
442                     29                      6
224                     16                      4
 61                      5                      2
226                     15                      4
105                      8                      3
*80% certain of detecting the indicated % reduction in mean of the EMCs.

    There are numerous examples in the literature where small differences (2 to 5%) have been reported based upon fewer
samples than  indicated  by this analysis.   This highlights  the  need to be  more rigorous with regard to statistical  testing of
reported effectiveness estimates.   To detect larger changes, the number of samples becomes reasonable.  The mixed  land
use  catchment in Portland is currently being studied for  the  effectiveness of the implementation  of a number  of source
controls and  other controls that do not lend themselves  to input/output testing.  Examples include maintenance changes
(catch  basin  cleaning,  street sweeping);  education (business and residences);  tree planting,  and others.    Post-BMP
monitoring will be conducted along with qualitative evaluations.

    Another approach that this study will be evaluating is  the use of effluent data to compare to design criteria.  It has been
suggested by some researchers that BMPs may be  able  to treat only to  a given concentration and therefore, if relatively
clean water is  entering  a BMP, performance based upon efficiency may not fully  characterize whether  a BMP is well-
designed.  An example  of this is based upon Rushton et  al. (1997). The  pond was located  at the Southwest Florida Water
Management District service  office in Tampa.  The drainage basin is 6.5 acres with about  30%  of the watershed covered
by rooftops and asphalt parking lots, 6%  by a crushed limestone  storage compound and the remaining 64% as a grassed
storage area.  The pond was  modified twice after initial construction; therefore, there are three periods of performance  data
for three different designs.  The first pond had an average  retention time of 2 days, the second 5 days and the third 15 days.
The second design added wetland features, while the third utilized deeper and larger pools.
                                                        183

-------
    Figure  1 shows the input and output median  concentrations in  log based  10 scale as well as the 95% confidence
limits. The study reported that performance of  the pond (defined  as  removal efficiency) decreased after the first
modification. What  appears to be evident is that the average inflow concentrations were much lower during the second
period, while the outflow concentrations were about equal  (less, but not statistically different from the first design). It
appears that with the original  and  first modified designs that the effluent  level  was not  decreased. However, one could
not say that the BMP was any poorer in efficiency. The last design appears to have lowered the potential effluent
concentration, but the  major difference in efficiency came from the significantly higher  inflow concentrations during the
sampling period. This  example points out the need to  carefully think about whether pollutant removal efficiency is an
accurate representation of how well a BMP works or does not.

    In many cases, there is a need to conduct dry weather analyses between  storms on BMPs with dry weather flows.
It may be that  pollutants captured during storms  are slowly released during dry weather discharges.

    Biological  and  downstream  physical habitat assessments such as aquatic invertebrate sampling  and habitat
classification should be explored as an alternative to just utilizing chemical measures of effectiveness (Maxted, 1999).
Long-term trends in receiving  water quality, coupled with  biological assessments, would likely be a much better gage of
the  success of  the implementation of BMPs, especially on  an area-wide  basis.
                    2.60
                    2.00
                           Median=
                         |  21mg/l
                                          81mg/l

                                           I191
                  . 1.50 i
                  O      !
                  I      I
                    1.00
                    0.50
i 1.32
                   llmg/l
    1.03      T 1-05


llmg/l
                                lo.96

                                9mg/l
                                                                                     ).86 i
                                                                                 5mg/lj
                         Inflow 1990    Outfbw    Inflow  1993-*  Outflow   Inflow 1994-    Outflow
                                       1900       1004     1993-1OS4     1995     1994-1995
Figure 1. Inflow and Outflow Log Mean TSS Concentrations (mg/l) and 95% Confidence Limits for Different Designs of a Wet Pond Located at SWFWMD
Service Office in Tampa, Florida.

Summary  and Recommendations

    There is a great need to have consistency with the constituents and methods utilized for assessing BMP
effectiveness. This paper  has presented only some of the consistency issues.  It is recommended that researchers who
undertake BMP  effectiveness studies  consider the  recommendations suggested  here,  by Urbonas (1995) and other
recommendations based upon further analysis of this subject.  It is the authors' opinion that EPA should require studies
receiving  federal  funding to conduct BMP effectiveness studies  that utilize standard methods as suggested here, together
with  much  still-needed  detailed guidance on data collection and sampling methods to improve data transferability.
                                                     184

-------
Acknowledgments

    The authors wish to thank Gene Driscoll, Jonathan Jones, Bob  Pitt,  Bill Snodgrass, and Larry Roesner for their helpful
discussions and comments on the subject. Jonathon Jones, Jane  Clary, and John O'Brien from Wright Water Engineers
developed the Database Structure and Platform as  members of the ASCE project  team.  The Urban Water Resources
Research  Council has provided very  beneficial support and feedback. Finally, the  assistance by EPA in funding this
cooperative agreement is acknowledged as well as the helpful participation of Eric Strassler and Jesse Pritts in  reviewing
our work.

References

Driscoll, E.D., P.E. Shelley, and E.W. Strecker. 1990. Pollutant Loadings and Impacts from Storm Water Runoff, Volume
III:  Analytical Investigation and Research  Report.  FHWA-RD-88-008, Federal  Highway  Administration.

Driscoll, E., G. Palhegyi, E. Strecker, and P. Shelley, 1989. Analysis of Storm Event Characteristics for Selected Rainfall
Gages Throughout the United States.  Draft Report. Prepared by Woodward-Clyde for the U.S. Environmental  Protection
Agency. 43pp.

Martin, E.H. and J.L. Smoot.  1986. Constituent-Load Changes in Urban Stormwater  Runoff Routed Through a  Detention
Pond-Wetland  System in Central  Florida. U.S. Geological Survey Water  Resources Investigation Report 85-4310.

Maxted, J. 1999. Proceedings  of the First International South  Pacific Conference  on  Urban Stormwater

Ruston, B.T., Miller,  C.H., Hull, H.C., and  J.  Cunningham,  1997. Three  Design Alternatives for Stormwater Detention
Ponds. Southwest Florida Water Management District. Brooksville,  FL.

Sokal,  R.R.  and F. James Rohlf. 1969. Biometry: The Principles and Practice of Statistics  in Biological  Research. W.
H. Freeman and  Company.  San Francisco, CA.

Strecker,  E., 1994.  Constituents and Methods for Assessing BMPs.  Proceedings of the Engineering Foundation
Conference on  Stormwater Related Monitoring Needs. Aug. 7-12, Crested Butte, CO. ASCE.

Strecker, E.W.,  Kersnar,  J.M., Driscoll, E.D., and Horner, R.R. 1992. The Use of Wetlands for Controlling Storm Water
Pollution.  The Terrene Institute.  Washington, D.C.

Strecker, E. and B.  Urbonas, 1995. Monitoring of Best Management Practices. Proceedings  of the 22nd Annual Water
Reources Planning  and  Management Division Conference, ASCE,  New York.  pp.  48-51.

U.S. Environmental  Protection Agency. 1983a. Final Report on the National Urban  Runoff Program. Water Planning
Division,  U.S.  EPA.  Prepared  by Woodward-Clyde Consultants.

U.S. Environmental Protection Agency.  1993b. Memorandum.  Office of Water Policy and Technical  Guidance on
Interpretation and Implementation of Aquatic  Life Metals  Criteria.

Urbonas,  B.R.  1994. Parameters to Report with BMP Monitoring Data. Proceedings of the Engineering Foundation
Conference on  Storm Water  Monitoring Related  Monitoring  Needs.  August 7-12, Crested Butte,  CO. ASCE.

Urbonas, B.R.,  1995. "Recommended Parameters to Report with BMP Monitoring Data. J.  Water Resources Planning
and Management, ASCE. 121 (1), 23-34.

Woodward-Clyde Consultants. 1993.  Final Data Report: Data  from  Storm Monitored  between May  1991 and January
1993. Submitted to  Bureau of  Environmental  Services, City of Portland,  OR.
                                                    185

-------
                     Texas Nonpoint SourceBOOK Is Now On-Line!
                                   http://www.txnpsbook.org

                                           John  Promise, P.E.
                                   Director  of Environmental Resources
                              North Central Texas  Council  of Governments,  and
              Chair, Water  Resources Management Committee,  American  Public Works Association

                                             Keith Kennedy
                                North Central Texas Council of  Governments

                                       Robert W. Brashear, Ph.D.,
                                       Camp Dresser & McKee Inc.
Summary
           Nonpoi
         urce BOti
           \
The Texas Nonpoint SourceBOOKis an internet-based resource that has been developed
to assist public works officials across Texas with storm water management. The
SourceBOOK'provides basic information about storm water quantity and quality impacts,
outlines how to develop and implement a local storm water management program, identifies
localized water quality issues, and provides an interactive database of more than 100 Best
Management Practices (BMP's) to use in a variety of situations.
                          The Texas Nonpoint SourceBOOK provides information for the novice as well as the
                          experienced storm water manager.  The project  was funded by the Environmental
Protection Agency and matching funds from 20 local governments across Texas.  The North Central Texas Council  of
Governments (NCTCOG) served as project administrator.  The SourceBOOK was developed by a consulting team lead
by Camp Dresser & McKee  Inc. (COM). A Project Management Committee of  local governments provided project
oversight. After extensive review via the Internet, the SourceBOOK'was officially endorsed  by the Executive Committee
of the Texas Chapter - American Public Works Association  (APWA).  Five training workshops were conducted across
the state. The SourceBOOK is intended to be a living resource, with additions and changes occurring continually in
response to input from users. A feedback page allows direct input from the Internet.

Why a Texas Nonpoint SourceBOOK?

    Recognizing the  need for improved communication, cooperation, and education statewide on stormwater issues, a
Statewide Storm Water Quality Task Force was established  by the  Executive  Committee of the Texas Chapter -
American Public Works Association.  At an organizational meeting in February 1994,  a Steering Committee and
subcommittees were  formed.  The various subcommittees  immediately tackled  the task of identifying current issues and
needs regarding storm water quality and nonpoint source  pollution, particularly  with  respect  to the needs of public works
officials  across Texas.

   Already known was that nonpoint sources, including stormwater, contribute to  water pollution problems. The Water
Quality Subcommittee began to review data from the Texas Clean Rivers Program,  available nonpoint source monitoring
data, and the State's Nonpoint Source Water Pollution Assessment Report.  They presented this assessment at
subsequent  meetings of the Task Force. Water quality problems were known, but not how best to address them.
                                                  186

-------
    What was not known was the applicability and cost-effectiveness of Best Management Practices (BMPs) for
addressing many of the typical water quality pollutants: bacteria, pesticides, nutrients, metals, toxic chemicals, and
others. The Best Management Practices (BMP) Subcommittee surveyed  local  governments across Texas on BMP
implementation but found little technical data.  It was evident that until questions such  as applicability and cost-
effectiveness could be answered, local governments would not invest limited public funds on storm water controls.

    A project  was formulated that would provide the assistance local governments needed by developing an internet-
based resource of storm water management information. At the time it was a striking idea, since the Internet was very
new and few local governments  had any "on-line" experience. Using the emerging  Internet would provide ready electronic
access and would allow for the use of new technologies in communication. This  resource was to be  called the Texas
Nonpoint SourceBOOK, and would be developed in both "hardcover" and electronic form. A grant application, submitted
to the Texas Natural Resource Conservation Commission under the Section 319(h) Nonpoint Source Program, was
awarded in the spring of 1996.  Work on the project began in September, 1996.

How Was the Texas Nonpoint SourceBOOK Developed?

    The North Central Texas Council of Governments provided staff support and general administrative oversight. To
guide the development of the SourceBOOK, a Project Management Committee was established from the Texas Chapter-
APWA membership. Among  its first tasks was issuing a Request for Proposals for professional consultant assistance,
and  selecting  the  consultant finalists.  From the finalists the Committee selected a consultant team led by the firm Camp
Dresser & McKee Inc., in association with  Espey Huston & Associates, Inc.; Center for Watershed Protection; Booth, Ahrens
& Werkenthin, P.C.; Carter Burgess; and Pavlik & Associates. Together, the committee and consultants used the State's
Nonpoint Source Water Assessment Report and supporting information to identify particular pollutants from priority
watersheds  and related pollution prevention BMPs.

    During FY97, the Project Management Committee worked with the consultant to establish the format of the Texas
Nonpoint SourceBOOK on the Internet. Presentations on local BMP experiences were made at the  TX-APWA Short
Course at Texas A&M in February, 1997.  Initial consultant materials were  reviewed by the TX-APWA general
membership at its summer,  1997, Annual Meeting. A draft of the Texas Nonpoint SourceBOOK was  presented to the
TX-APWA  general membership  at the February 1998 Short Course, and local government comments were solicited.

    The TX-APWA Executive Committee endorsed the  Texas Nonpoint SourceBOOKin February, 1999. It is available
through the Internet and on CD-ROM for use by local governments across Texas. The Committee and consultant
conducted technology transfer and training workshops on storm water management  and the Texas Nonpoint SourceBOOK
at five regional  one-day workshops across Texas during February and March of 1999.

How is the Texas Nonpoint SourceBOOK Organized?

    The SourceBOOK is designed to make use of the capabilities of the Internet. This includes the ability to organize and
present textual and graphical information through common browser formats,  as well as providing active links to related sites.
The design of the content of the SourceBOOK maximized the use of existing web sources wherever possible.

    The content of the SourceBOOK consists of a set of modules:

    Introduction  and  Overview

    • About This Site

    • Frequently Asked Questions  (FAQs)

    • Related  Links

    . Nonpoint Source News


                                                   187

-------
. Post Your Feedback
Module 1 - Nonpoint Source Management 101
. History of Nonpoint Source Management
. Urban Nonpoint Source Primer
. Controlling Urban Runoff-Guidance for Beginners
. Selecting the Right BMP - Guidance for Beginners
. Planning Your Stormwater Management  Program - Guidance for Beginners
• Glossary
Module 2 - Urban  Runoff Management Programs
.  Introduction
. The Planning and  Goal Setting Process
. Planning and Program Approaches
. Funding Mechanisms
. Measuring  Effectiveness of  Management Programs
. Implementation Strategies
. Case Studies
.  Bibliography
. Additional Resources
Module  3  - Characterizing  Urban Waterways
. Urban Runoff Flow and Water Quality
• Assessing Urban  Waterways
. Water Quality and Other Watershed  Physical Characteristics  in Texas
Module 4 - Runoff Quality  Best  Management Practices
• Selecting  Management Practices
. Housekeeping Practices
• Source Control  Practices
. Treatment Control Practices
. Interactive BMP Selector
                                               188

-------
What  Does  Each  Module  Provide in the Texas Nonpoint SourceBOOK?

    Module 1,  "Nonpoint Source Management 101," is a primer for beginners on urban stormwater management. It
quickly  establishes that storm water quality £0° quantity  management need to be addressed as one integrated program
within a local government. It provides guidance  on regulatory issues, basic axioms of runoff control, and the use of
pollution prevention, source and  treatment  controls.

    Module 2,  "Urban Runoff Management  Programs,"  describes the process to be used to manage urban runoff within
the overall framework of the  city, county, or special district. Particular  attention is placed on  the  key institutional and
financial components necessary  for a successful  ongoing program.

    Module 3, "Characterizing Urban Waterways,"  begins with a generic  discussion of urban runoff flow and water quality
relationships. Considerable attention is then  given to proper techniques  for monitoring urban waterways and stormwater
runoff. The  majority  of the module focuses on Texas-specific information.  Descriptions of known water quality problems
can be  accessed for the entire state. Each  regional planning area and basin has specific information on water bodies,
 B* &*_
 Urban Nonpoint Source Management 1Q1
 Ghaplrcr Contents  Home
     Examples of site controls: an extended detention basin with permanent pool (right) and an infiltration basin
 Regional Controls serve a multi- acre drainage area, usually greater than 10 acres. Regional facilities can be
                                    h^nor^r r»n,nn«l rnnfrnlyrg h*»»Pi
                                   Urban Runoff Management Programs
                                   Chapjar Cnntants  Home
                                     - monitor program effectiveness.
                                              *         figure M

                                                  THE URMP PLANNING PROCESS

                                              Program Activities         Technical Activities

                                                        INVOLVE Ml
                                                        ,,^.,
                                              _       __   _,
                                                                                E
odule 1
                                                                               $CW£-"
                                                       189

-------
          tfc E* Y»» ao  frnto H*
         ,*•••**OJ3<3
          Characterizing Urban Waterways

         5.0  Other Watershed Physical Characteristics by Region

          Texas State River Basins, Council of Government Boundaries, and Urbanized Areas


                       (click on your «re« of Interest)             ;
         tie™
watershed  characteristics, annual precipitation and  runoff, major soil types, and the like. There are many "hot"  links to
real-time gauging  stations, local programs,  and state/federal sites, such as EPA's Surf Your  Watershed.

    Module 4,  "Runoff Quality Best Management Practices,"  provides guidance  on the selection of Best Management
Practices for pollution prevention,  source control, and treatment control.   Considerable effort was placed on  gathering
the most current information on more than 100 BMPs and  review  by the Project Management Committee of local
governments. Each BMP  includes detailed information, such as performance data, photographs, and relevant  reference
citations. An innovative BMP Interactive Selector was developed for the SourceBOOK. It enables the user to  peruse
BMP's in each category, or to input several characteristics  specific to their situation and request a set of the most
applicable  BMPs.
                                                       190

-------
Lunoff Quality Best Management  Practices
•Mptmr qof*ent*   tUMD*    S^P Cftflsnt?   tl«CCb



Source Controls



UBS suras.  Source controls appropriate for construction sites  art designated as 'CS-' and source controls for perminaent us» are designated as
PS-'. All source controls are rated for their suitability on Residemial/CQrtvnarcial,  Indus t rial/Comma rtcial, or Construction applications,
i  Residential/Commercial applications include residential developments as i

  commericial and/or industrial).

i  Indus I rat,'Commercial applications are focused an individual sites whose activity

  with storm water regulations or who have activities that could pollute ttormwattr runoff,

i  Construction applications are those practices required during tne construction of residential,
                                                                 larger developments that involve mixed land use (residential,


                                                                 es are industrial or commercial in nature and who must comply


                                                                                ercial, or industrial facilities.
Clicking once on the column headings in the table below will sort th« display in descending order  Clicking again on that same column heading will
sort the table in reverse order.   Click an the Number column to preview the description of the practice, or click on the BMP ID to view the BMP.
                                       !>• Prrt HIM  faum fim**  Bat- R»"

                                 Sort»4 by [R»WConB. Uujndti] Rx [I U>35] of 15 Gi
            ^1   CS-EC 15 Er°*°n Control - Channel StabfliMQ


             1 ,   PS SW1 Swale


             1 f PS-SW.O Hter Stop (cwpS-3)


             i i' PS-IHO  .Mto-ationBarinCcwpM)


             i ,E' PS-EC.5 i&owio Control - Flow Controls
                                                                                                                    radices
                                                                                                                       in for the disturbed ci
                                                                                                                                                                   J
                                                                  Portion of Sediment Captured

                                                                 (Sediment Rettlned» A- A  x F
                                                                                                           Undliturbed andJor
                                                                                                                Offsrte Area
                                                               Sit.R.tinj-
                                                                              Totil Stdimtnt R«t«in«d On SIM

                                                                             Total S«dim«nt Loidlng Fr«m SIM
                                                                                                                             AI
                                                                                                                   ment conttDl plans
  Runoff Quality Best Management Practices
     Jar CoiUantt   Home    ilXii-l(KOlflI3£i   Bq-mfl
   Treatment Control Practice Overview
                                                                                                                                                           m
   91 "on doaign alotm t
   100 year »tor») .  Dnt
   •od vhan properly daeigned .
                                       a? ii dDwnctrBB> ch«n
                                                                                      191

-------
               A Comparison  of the  Long-Term  Hydrological  Impacts of
                              Urban  Renewal versus Urban Sprawl
                            Jon Harbor, Suresh Muthukrishnan,  and Shilpam Pandey
                                Departments of Earth and Atmospheric Sciences
                                              Purdue University
                                           West Lafayette, Indiana

                                     Bernie Engel, Don Jones, and K.J. Lim
                             Departments  of Agricultural  and  Biological Engineering
                                              Purdue University
                                           West Lafayette, Indiana
Abstract
    Recent concern over environmental and economic impacts of urban sprawl  has focused  renewed attention  on the
importance of making full use of existing  urban  areas. Revitalizing former industrial, commercial, and  residential areas
often involves changes in land  use type or intensity of use. It is  important to have the ability to evaluate the long-term
hydrological impacts of such changes. These impacts can then be placed within the context of impacts that similar land
uses would have if a decision were made to place them in the urban fringe (urban sprawl) rather than in existing urban
areas (urban  renewal).

    In this study, we illustrate how the Long-Term Hydrological Impact Analysis  (L-THIA) tool can be used to compare
the hydrological impacts of land  use change in existing urban  areas versus change  in the urban fringe. L-THIA is a simple,
comparative tool that requires the  user to provide information  on  land use  and soil type for existing and future/planned
conditions. The tool combines this  information with local rainfall data to calculate long-term average annual surface runoff
under existing and future/planned conditions. L-THIA analyses can be run directly  at our web site for locations throughout
the U.S. where the curve number technique is already routinely used (http://danpatch.ecn.purdue.edu/-sprawl/L-THIA).
By performing analyses of renewal versus conversion  of agricultural  land at the urban fringe, it is possible to provide a
comparative assessment of impacts. This initial comparison  can be helpful  in  educating the general publicand decision-
makers, thereby raising awareness of this element of the set of variables that are considered in land  use decisions.

Introduction

    Because almost every major North American city had been founded by 1900, the dominant form of urban
development  during the 20th Century has been  growth on  the outer edges of existing cities, or just beyond city limits
(Orum, 1995). With improvements in transportation  and communications, the need for  people to be clustered in high-
density central  areas has decreased (Chinitz, 1991),  encouraging decentralization, suburbanization, and sprawl.  In the
United  States, 87% of the population now lives  in metropolitan areas and  their hinterlands (Angotti, 1995), and  steady
infilling between urban areas has resulted in the development of megalopolises such as the Philadelphia - Boston -
Washington DC - New York urban  corridor. Even metropolitan areas which are stagnating or declining  in terms of total
population are still growing in terms of total built area because of low-density suburban  growth (Johnston, 1982).

    Decentralization  and suburbanization  have  changed the relative  importance of the core areas of cities (Richardson,
1982). Although these central areas were the sites of initial city growth and development,  many cities are now faced with
the challenge  of revitalizing these  once vibrant central industrial, commercial, and residential  areas that have been in
decline in recent decades:. The following  quote  reflects efforts to  slow the tide of migration from urban  centers.
                                                     192

-------
    "To combat the number of people fleeing [Chicago] for the suburbs, developers have lured middle-class home
    buyers back with promises of safe neighborhoods and affordable homes. Chicago also leads the nation in
    converting  office and warehouse property into  residential space such as condominiums and  rental  units, often
    targeted  to low- and  moderate-income  buyers." (Heavens,  1999)

    At the same time that city administrations have been  coping with the challenges of urban core renewal, suburban and
rural communities have become increasingly concerned about the environmental, economic, social,  and aesthetic  impacts
of continued urban growth at the fringes of developed areas (these later concerns are often grouped under the  term urban
sprawl). Preservation of prime farmland and  protection  of rural  areas  have become important concerns, alongside a
growing emphasis on combating the impacts of continued sprawl on  flooding, groundwater recharge, air pollution,  climate,
ecology, and habitat fragmentation (Schueler,  1994). Although there is considerable interests in revitalizing urban cores,
especially if this reduces urban sprawl, to accomplish  this  requires that the decision-making process for urban and
suburban planning include consideration of the  environmental as well as the  economic aspects of land use.

    Land use decisions are highly complex, involving consideration of economics, infrastructure,  politics, labor and
population dynamics, and the environment. The planning process requires collection and comparison  of a wide array of
data,  usually with the goal of providing  a planned solution that meets goals based on sustainable growth in industry and
commerce. However, increasing public and political concern over the environmental aspects of urban development has
raised  the profile of efforts to develop efficient and environmentally sustainable  urban environments. The key components
of environmentally sound urban  development include  land use patterns that  minimize  environmental impacts  (Arendt,
1996), efficient automobile and pedestrian traffic,  and the use of energy saving  and environmentally sound building
designs. When attempting to balance economic and environmental concerns,  it is important to quantify the differential
environmental  impacts of alternate land-use scenarios. Objective  measures of differential impacts provide  a rational basis
for decision-making. In addition, they can  be used to educate the public and key decision-makers in government  and the
private sector about the level of environmental  benefit that can be gained  from alternative  land-use decisions.

    The aim of the work presented here is to demonstrate the application of an impact assessment tool in evaluating the
long-term hydrologic impact of development consistent with urban renewal versus the impact of an identical development
located at the  urban fringe. Although the general outcome of such a comparison is unlikely to  surprise anyone, the
advantage of quantifying differential impacts is in providing an objective numeric measure that is much easier to  include
in decision-making  than vague subjective assessments of environmental  benefits.

Long-Term  Hydrologic Impact Assessment (L-THIA)

    In  response to concerns  from local planners  that they had no simple, objective way to assess the impacts of alternate
development  plans on surface water  runoff and  groundwater recharge, a Long-Term Hydrologic Impact Assessment tool
(L-THIA) has been developed (Harbor,  1994; McClintock et al.,  1995; Ogden,  1996; Grove,  1997; Bhaduri et al., 1997;
Bhaduri, 1998; Minner, 1998; Minner et al. 1998; Lim  et al., 1999; Leitch  and Harbor, in press). L-THIA uses readily
available data on soils, climate, and land use to  estimate  long-term surface water runoff. By running the model  for current
conditions, and then with changed land uses, the user can simulate the potentialimpact of land use change. The method,
initially developed as a simple spreadsheet application  (Harbor,  1994), is based on the U.S.  Department of Agriculture's
curve number (CN) method  for relating precipitation and runoff as  a function  of land use and soil  type (USDA, 1983,
1986). The CN method was selected because  it forms the basis of other commonly used hydrologic models, thus the
data required for its use is readily available in most planning settings.  Because of the reliance on the CN method,  L-THIA
applies directly to those  areas where the  CN  method is  routinely used.  Subsequent development  of the L-THIA  method
has included provision  of a  Geographic Information System  (CIS) version (Grove, 1997), addition of nonpoint source
pollution loadings to land  uses (Bhaduri, 1998),  and  development of an Internet-accessible version of the  method (Lim
et al., 1999).

    In the curve number technique,  the land use  and hydrologic soil type of an area are used to derive a CN value (values
typically range from 30 to 98).  For any given  daily precipitation, surface runoff is then computed from empirically based
relationships between rainfall, CN, and runoff. Although most commonly used to estimate runoff for extreme storm
events, in L-THIA the CN technique is used to determine daily runoff for a 30-year time series  of daily precipitation values.


                                                      193

-------
Average annual runoff is calculated for each CN to provide a measure of long-term average impact, rather than simply
impact  on isolated  extreme storm events.  To compare different land use change options,  pre-development and post-
development average annual runoff can be calculated for each scenario.  The L-THIA method is freely available at
http://danpatch.ecn.purdue.edu/~sprawl/LTHIA. This site includes information on the technique and its application, as
well as access to US climate and soils data necessary to  run analyses. Users can submit land use and soil information
through a spreadsheet-style interface (Figure  1). Analyses are performed on a server at Purdue University  and results
are delivered  back to the  user in the form of tables and graphs.
 fleets or Land Use Change on Hydrology and NFS pollution - Netscape
                                        vvEPA
                                             United States
                                             Environmental Protection
                                             Agency
                                              LTHIA
                          (Long Term Hydrologic Impact Assessment) WWW_

                                               Input
                                                               | Illimii
                                              Kvn.son.cRon             AM*
                                                            YEAKl    Wail    YEAJI3
                                                nru
                                                nm
                                                                                      Bte
         * lament: DoneUi  :T7ST    *           " "":,   '  ~*** ""7"*";'  !';/'°"  t

Figure 1. L-THIA WWW Input Screen at http://danpatch.ecn.purdue.edu/~sprawl/LTHIA.

A Comparison of Core Renewal versus Fringe Development

Study Scenario

    The L-THIA tool  can be used to examine the relative  impact of land  use change in the form of an urban renewal
project; replacing underused or abandoned  commercial, residential, and  industrial buildings  in  an  urban core  region;
versus an urban sprawl project;  replacing agricultural  land  at the edge of a city. For the  sake of  illustration, consider
planning a 70 Ha  major commercial development with urban core  and urban fringe location  alternatives. Although the
location decision-making process will be  driven  by economic and infrastructure concerns, also assume that differential
environmental impact is important  in decision-making, perhaps as a result of political  or regulatory  pressure.  In the
context of improving  urban  environments  then, an important question  is the  extent to which placing  this development in
an urban core region would have different hydrologic impacts than  placing it at the city fringe.

    To simulate this situation, consider two possible sites in  the Chicago area.  The  first  is in the urban core, and currently
consists of a mix of residential, industrial, and  commercial properties that are unused or underused (Figure 1). The
second possible site  is on the urban fringe, and currently is used for agriculture. For simplicity we assume that both  sites
are on the same type of soil (from a hydrologic perspective), although in a real world example this might not be the case.
In each case, we use the L-THIA web  tool to analyze how average annual  runoff will change  if the site is converted to
                                                      194

-------
solely  commercial use (Figure 1). In the L-THIA input and output, the urban  core site is labeled "YEAR 1 ", the urban
fringe  agricultural site is  labeled "YEAR  2" and the commercial land use for both sites is labeled "YEAR 3." The L-THIA
web tool uses the "YEAR" designation for different scenarios because  analyses are typically for land use changes over
time.
 Results

    For the example described here,  placing a commercial development in an  urban core region, replacing an  existing
 mix of urban  land uses, increases average  annual runoff by 58% compared to the initial  situation (Table 1 and; Figure
 2). Note that the levels of impact given in Table 1  do not depend on the size of the commercial development; the same
 percent increase applies  regardless of area. Runoff increases because land uses with less  impervious  cover, such  as
 residential, are replaced by  commercial land use  that has a  higher percentage of impervious area,  in contrast, for the
 urban fringe location, replacing agriculture with commercial  use increases runoff by 670% (Table 1 and Figure 2), a ten-
 times greater  impact. Runoff increases so dramatically because agricultural use  on  relatively  permeable soil is replaced
 by very  extensive  impervious  surfaces.

 Table 1. Averaae annual runoff deoths and chanae for commercial development (post-development) in the urban core versus the urban fringe.  Results
 are for the specific example described in the text."
Urban Core
Urban Fringe
Pre  Development
Average Annual  Runoff
(mm)
          81.8
          16.8
Post  Development
Average Annual  Runoff
(mm)
         129.3
         129.3
                                                                               Increase in Runoff (%)
 58
670
100000 -
g o^ 75000 -
1 "
> | 50000 -
1 1
§»
EC •§ 25000
^o,
A

Low Density
Residential

-*, High Density, .
- ' Ccwmeriaall^-
/ 	


«HV*
                                          Core
                                 Fringe      Commercial
Figure 2. Average annual runoff volumes for commercial development, the urban core mixed-use, and the urban fringe agricultural use. The much larger
difference between the fringe location runoff volume and the commercial case indicates that fringe development will have the largest hydologic impact.
Note that the runoff volume is simple the average annual runoff depth (Table 1) multiplied by the site area. Results are for the specific example described
in the text.
                                                        195

-------
Discussion  and Conclusions

    The straightforward example presented  here  indicates that developing a  commercial site in  an urban  core versus and
urban fringe location can have  a  very significant impact on the level of disturbance of the hydrologic  regime.  For the
Chicago example presented  here, the urban  fringe location produces an approximately ten times larger  impact than the
urban core location.  Clearly, from a solely hydrological standpoint, the urban core location is a  better  choice than the
fringe location. Although this  is a hypothetical example, it illustrates the relative ease of use of the L-THIA tool,  and more
importantly demonstrates an  accessible way to provide  a  quantitative estimate of the relative impacts of different land use
decisions. More complex land use mixes and soil types  can  be run  on  the L-THIA web tool,  either in  the spreadsheet
version or in a CIS  version also available at the web site.  Thus, more sophisticated comparative analyses can be
performed.

In most cases, an L-THIA analysis provides a result that  shows that renewal of existing areas has less hydrologic  impact
than development of  an area with rural use. This is not a  surprise, rather the value of the tool is that  it provides  a context
for understanding and  considering the  magnitude of this difference  in the decision-making process. For areas  where
problems such as groundwater supply and  downstream flooding  are important,  the scale and magnitude  of the  hydrologic
impact can be of considerable importance  and can be considered alongside other  concerns, such as infrastructure and
economic viability. We  suggest  use of tools  such as L-THIA as  part of the planning process, to ensure that land use
decisions are made after consideration of a full range of concerns, including  environmental  parameters as well as
economic,  infrastructure, and political issues.

Acknowledgements

    Development of  the L-THIA tool is being supported by grants from the Environmental  Protection Agency,  Region 5.

References

Angotti,  T., 1995. The  Metropolis  Revisited.  Futures, 27: 627-639.

Arendt, R., 1996. Conservation  design for  subdivisions, a practical guide to creating  open space networks. Washington
D.C., Island Press.

Bhaduri, B. 1998. A  geographic  information  system based model of the long-term impact of land use  change on  non-point
source pollution at a  watershed  scale. Unpublished Ph.D. dissertation, Purdue  University, West Lafayette, Indiana 47907-
1397.

Bhaduri, B., M. Grove,  C. Lowry,  and J. Harbor. 1997. Assessing the long-term  hydrologic impact of land use change.
Journal of the American Water  Works Association. 89:94-1 06.

Chinitz,  B.  1991. A Framework  for Speculating about Future  Urban Growth Patterns  in the US.  Urban Studies, 28:939-
959.

Grove, M. 1997. Development and application of a CIS-based model for assessing the long-term hydrologic impacts of
land-use  change. Unpublished  MS Thesis, Purdue  University,  West Lafayette,  IN 47907-I 397.

Harbor, J.  1994. A  practical method for estimating the impact of land use change on surface runoff, groundwater
recharge and wetland  hydrology. Journal  of American Planning Association, 60:91-104.

Heavens, A.J., 1999, Building Anew in the City. Philadelphia Inquirer (09/12/99) L1.

Johnston,  R., 1982.  The American Urban  System: A Geographical Perspective.  New York: St.  Martin's Press.

Leitch, C. and Harbor, J., (In  Press) Impacts of land use change on freshwater runoff into the near-coastal  zone, Holetown
watershed, Barbados: Comparisons of  long-term to single-storm effects. Journal of Soil and Water Conservation.
                                                      196

-------
Lim, K.J., Engel, B., Kim, Y., Bhaduri,  B., and J. Harbor.  1999. Development of the long term hydrologic impact
assessment (L-THIA) WWW systems.  Paper No 992009, ASAE,  St.  Joseph,  Ml.

McClintock, K., J. Harbor, and T. Wilson. 1995.  Assessing the hydrologic impact of land use change in wetland
watersheds, a case study  from northern Ohio, USA. In: McGregor, D.  and Thompson, (Eds.) Geomorphology and  Land
Management in a Changing Environment, pp.1  07-I 19.

Minner,  M., 1998.  Sensitivity analysis  and  advanced applications of L-THIA. Unpublished MS Thesis.  West Lafayette,
IN:  Purdue University, Department of  Earth and Atmospheric Sciences, West Lafayette, IN 47907-I  397.

Minner,  M., Harbor J., Happold S., and  P.  Michael-Butler, 1998. Cost Apportionment for a Storm-Water Management
System, Applied Geographic Studies, 2:247-260.

Ogden,  M., 1996. Development and  Land Use Changes in Holetown,  Barbados:  Hydrologic  Implications for Town
Planning and Coastal Zone Management.  Unpublished  M.U.P.  Thesis. Montreal: McGill University, School of Urban
Planning.

Orum, A., 1995. City-building in America. Westview Press,  Boulder,  CO.

Richardson, J., 1982. The Evolving Dynamics of American Urban  Development, in: Cities in the  21st Century. Gapped,
G. and Knight, R.  (Eds.) Urban Affairs Annual Reviews,  23.

Schueler, T.,  1994.  The importance of imperviousness.  Watershed  Protection Techniques,  1:100-1 11.

U.S. Department of Agriculture, Soil Conservation Service, 1983. Computer programs for project formulation - hydrology.
Technical  Release  20.

U.S. Department of Agriculture, Soil Conservation Service, 1986. Urban hydrology for small watersheds. Technical
Release 55.
                                                    197

-------
                Comparative  Nutrient Export And  Economic Benefits of

                     Conventional And Better Site Design  Techniques


                                Jennifer Zielinski, Deb Caraco, and Rich  Claytor
                                       Center for  Watershed Protection
                                            Ellicott City, Maryland


    Better site  design describes  a fundamentally  different approach to the design of residential  and commercial
development projects. It seeks to accomplish  three goals at every development site: to reduce the amount of impervious
cover, to increase the amount of natural  land set aside  for conservation, and to  use pervious areas for more effective
stormwater  treatment.

    When designing new residential developments, planners have  the opportunity to reduce stormwater runoff and
pollutant export  through better site design techniques. The better site design  techniques applied to these developments
are referred to  here as "open space design," and present an alternative to  conventional  residential subdivisions. Also
known as cluster development, open space design concentrates density on one portion of a site in order to conserve open
space elsewhere by relaxing  lot sizes, setbacks, frontages and road section and other lot geometry.  Open space design
also consists  of:

    •    installing narrower streets and shorter  driveways

    •    spreading stormwater runoff over pervious  areas

        using open  channels rather than curb and gutter

    •    clustering development to conserve  forests  and natural  areas

        reducing the area devoted to turf

    . protecting stream buffers

        enhancing the quality of septic system effluent in areas where sewage is disposed  of on-site

    When these techniques are applied together, the cumulative benefits  of better site design can be impressive.
Documenting the precise benefits is difficult, however, since few developments incorporating better site design techniques
have been built, let alone monitored.

    As most better site design techniques are non-structural in nature, the achievable benefits will vary depending  on the
unique  characteristics of each development  site and the actual site  planning practices applied. Also, since better site
design  techniques are commonly  applied  together,  it  has been difficult to accurately quantify their individual nutrient
removal benefits. Many local governments, consultants,  and developers have expressed a strong desire for clear
documentation of these  presumed  benefits.

    To help meet this need,  the Center for  Watershed Protection (CWP) recently completed a study  to document the
comparative nutrient export and economic  benefits of conventional and better site design techniques. The  simple
assessment methodology analyzed  both the residential  and  commercial  environment through four real-world development
case studies in  the  Chesapeake  Bay watershed.  This paper presents  the results of the residential component  of that
project, including the incorporation of open space design techniques into the redesign of two residential  case studies; the
resultant hydrologic,  nutrient  export, and economic benefits; and finally, the implications of our findings for the watershed
manager.
                                                     198

-------
Methodology

    The basic method used in the Nutrient Loading from Conventional and Innovative Site Developmentproject (Caraco,
et al., 1998) conducted by CWP is a redesign  analysis that compares conventional and better  site  design at actual  project
sites using a simplified  model.

    CWP  first assembled plans of previously developed sites representative of typical development scenarios across the
Chesapeake Bay, including a  medium-density residential development from Virginia's Piedmont,  a  large-lot single family
residential subdivision from  Maryland's Eastern Shore, a retail strip  mall  from Frederick County, Maryland, and a
commercial office park located outside of the District of Columbia in suburban  Maryland. Each site was then "redesigned"
using better site  design  techniques.

    The Simplified Urban Nutrient Output Model (SUNOM) was then used to  compare each  conventionally designed site
to the  redesign. SUNOM is a spreadsheet model that  computes the hydrologic budget,  infrastructure cost and  nutrient
export  from any site,  using common  site planning variables. The  model provides watershed practitioners with  a  simple
tool to  compare the  costs and  benefits of better site design.  It is not meant to  be used as a method  for determining actual
stormwater runoff and nutrient loading from a development site.  To  obtain accurate numbers for  this, a  more detailed
model  should  be used or on-site monitoring should be conducted.

    Model input  includes basic site  planning  variables that  can be  directly obtained or measured from a  typical
development submittal to a land  use  authority, including total drainage area,  length  of sidewalks, total impervious cover,
linear feet of roads, lawn cover, utilities (length and type), forest cover, size, type, and length of stormwater conveyance,
riparian forest cover, size and type of stormwater practices, soil  type(s), and method of wastewater treatment.  Default
data are  provided  for many  parameters and many of these assumptions can be changed  based on  site specific
information.

    SUNOM is governed by  the principles of a simplified water balance. In addition to annual runoff and infiltration,
SUNOM computes the annual nutrient load from each  development site in pounds.  In brief, the surface  nutrient export
from each site is  estimated using the  Simple Method (Schueler, 1987).  This export  is then adjusted to reflect the mean
removal capability of stormwater BMPs where present (Schueler,  1997).  The subsurface component of the model utilizes
annual subsurface recharge rates (based on the site's  prevailing hydrologicsoil group)  and  monitored baseflow  nutrient
concentrations in the receiving water  to estimate the annual subsurface  nutrient export from  urban areas. These  values
are then adjusted for the area of the site that cannot recharge (i.e., impervious cover) or are  hindered  from infiltrating by
other conditions (e.g., compacted  urban turf). The  model also calculates the cost of development  utilizing previously
published  or user-specified unit costs  and predictive equations  for infrastructure, stormwater management,  landscaping,
and  septic systems.

    For each case study, SUNOM was used to compare the annual hydrologic budget and annual  nutrient export under
five development scenarios:  pre-developed  conditions,  conventional  design without stormwater practices (uncontrolled),
conventional design with stormwater  practices (controlled), design incorporating better site design  techniques  without
stormwater practices (uncontrolled), and design incorporating  better  site design techniques with  stormwater  practices
(controlled). The cost of development associated  with  each design was  also  estimated.

Case Study #1: Duck Crossing, A Low Density Residential Subdivision

    Duck  Crossing, a large-lot residential development, is located in Wicomico County  on  Maryland's  Eastern  Shore.
Prior to development, the parcel was representative of the  typical terrain on  Maryland's coastal  plain,  with very little
gradient. The site  contained tidal and non-tidal  wetlands,  natural  forest, meadow,  the 1 00-year floodplain, as well as three
existing dwellings with on-site sewage  disposal.

    The large-lot subdivision of single family homes, constructed in the  1990's, (Figure 1) contains eight new residential
lots, each  of which  are 3 to 5 acres in size with houses set far back from the  street. The  street is wide  given the few
homes that are served,  ends  in a large cul-de-sac, and is lined with  a  sidewalk. Each lot has an on-site private septic


                                                      199

-------
Figure 1. The conventional design of Duck Crossing, a low density residential subdivision on Maryland's Eastern Shore.

system, with  a septic reserve field of about 10,000 square feet. Individual home  property lines extend to the  protected
tidal marsh, which is the only common open space on  the site. Stormwater management consists of street runoff
conveyed by curb and gutter to a storm drain system that discharges to a small  wet  pond.

    The major better site design techniques  applied when redesigning  this  site (Figure 2) included:

        conservation of tidal  and  non-tidal wetlands and  forested areas

    •   a 100-foot buffer along tidal  and non-tidal  wetlands

        clustering  development to provide additional  open space

    •   identification of potential  development and  open  space areas based  on location of sensitive areas,  100-year
        floodplain, and  potential  septic field  areas

    .   distribution of stormwater treatment  practices throughout the site

        use of a  narrower access road; shorter, shared driveways; and wood chip paths through community open space
        instead of sidewalks  along the road

    •   use of shared septic systems  utilizing more  advanced  re-circulating sand  filter technology

     The open space design  resulted in reduced impervious cover,  reduced  stormwater runoff, increased stormwater
infiltration, and reduced infrastructure cost over  the conventional design.

Case  Study #2: Stonehill  Estates, A Medium-Density  Residential Subdivision

    Stonehill  Estates is located in Stafford County  just north of Fredericksburg,  Virginia.  The original site was  almost
entirely  forested in a mix of mature deciduous hardwoods, with perennial  and  intermittent streams,  and non-tidal  wetlands.
An existing network of public water and  sewer lines serves the site and road  access to the subdivision  is by two existing
streets.
                                                      200

-------
Figure 2. The open space design of Duck Crossing.

    The conventional design produced a total of 108  house lots,  each of which are about 9000  square  feet in size (Figure
3).  The subdivision  is quite typical of a medium-density residential subdivision developed in the last two decades in the
Mid
Figure 3. The conventional design of Stonehill Estates, a medium density residential subdivision in Stafford County, Virginia.

Atlantic with uniform lot sizes  and shapes, and generous  front  setbacks. The streets were 34 and 26 feet wide,  numerous
cul-de-sacs were used as turnarounds, and sidewalks were generally installed on both  sides of the street. With the
exception  of a small  tot-lot,  the majority of the open space is unbuildable land,  such as floodplains, steep slopes,
wetlands, and stormwater management areas. Street runoff is conveyed by curb and gutter to a storm drain system that
discharges to the intermittent stream channel.  It then travels to a dry extended detention  pond, which is primarily used
to control flooding, but also provides limited removal  of stormwater pollutants.
                                                        201

-------
    The open space design also results in 108 lots, but these were slightly smaller with an average size of 6,300 square
feet. The design also incorporates many  techniques of open space design as advocated by Arendt (1994). The design
techniques  employed in the redesigned site (Figure  4)  include:
Figure 4. The open space design of Stonehill Estates.

    •   identify sensitive natural features, including mature forest and wetland, to be protected
        incorporate a minimum 1 00-foot buffer along all perennial and  intermittent streams
        maximize the amount of community open  space and preservation of natural areas
        maintain the same number of lots as  the  conventional  design
    .   provide open  space adjacent to as many lots as possible
        incorporate stormwater management attenuation and treatment throughout the site
        use narrower streets,  loop  roads, shorter  driveways, and fewer sidewalks
        allow for  irregular shaped  lots and shared driveways
        manage stormwater in a "treatment train" with bioretention facilities that discharge to a small but more effective
        wet pond
    The open space  design resulted in  reduced impervious cover,  reduced stormwater runoff,  increased  stormwater
infiltration, and  reduced infrastructure cost over the medium density  subdivision  conventional design (Table 1).
The Benefits of Open Space Design
    For both of these  case  studies, application of the open space design techniques resulted in reduced impervious cover,
which translates  directly to reduced stormwater  runoff. Other "redesign"  studies recently conducted in Delaware,
Maryland, and Virginia have provided similar results. These combined results  consistently demonstrate that better site
design can reduce impervious cover by 25 to nearly 60% and stormwater runoff by 4 to over 60% for  a range of
subdivisions (Table 1).
                                                      202

-------
Table 1. Redesign Analyses Comparing Impervious Cover and Stormwater Runoff from Conventional and Open Space Subdivisions
Residential
Subdivision
Duck Crossing
Stonehill Estates
Remlik Hall 1
Tharpe Knoll 2
Chapel Run 2
Pleasant Hill 2
Prairie Crossing 3
Buckingham Greene
Belle-Hall "
Conventional Zoning for
Subdivision
3-5 acre lots
1/3 acre lots
5 acre lots
1 acre lots
1/2 acre lots
Va acre lots
Vz - 1/3 acre lots
1 /8 acre lots
High Density
Impervious Cover at the Site
Conventional
Design
8%
27%
5.4%
13%
29%
26%
20%
23%
35%
Open Space
Design
5%
21%
3.7%
7%
17%
11%
18%
21%
Net Change
-35%
-24%
% Reduction in
Stormwater
Runoff
23%
24%
-31 % I 20%
-46%
-41%
-58%
-20%
-7%
4%
31%
54%
66%
8%
20% - 43% 31 %
sources: 1 Maurer, 1996; * DE DNREC, 1997; 3Dreher, 1994; and "SCCCL, 1995.

    For both Duck Crossing and Stonehill Estates, the conventional design results in the highest annual volume of runoff
and the lowest volume of infiltration, as was expected.  Of particular interest is the fact that the controlled  conventional
design results in a higher annual runoff volume and a lower infiltration rate than the uncontrolled open space  design. This,
however,  should not imply that better site design alone, without structural Stormwater management,  is sufficient in
controlling Stormwater runoff from this site since the open space designs do not come close to replicating pre-developed
hydrology.

    Less  impervious cover and  Stormwater runoff, in turn, translates directly to smaller pollutant loads. Reducing the
impervious cover, preserving  natural  areas, and providing  multiple Stormwater practices in series reduced nutrient export
for both case studies. However,  neither open space design meets pre-development nutrient loads.

    One area  of particular interest for Duck Crossing is the implication of on-site sewage disposal  systems. The
conventional design included  astandard septictankand field for each lot, which resulted in phosphorus and nitrogen loads
that far exceeded pre-development levels. Recirculating sand filters were  used in the  open  space design,  instead of
conventional  septic systems,  because they yield better  nitrogen removal  efficiencies  and are actually  less expensive to
construct.  This resulted in a  much lower nutrient  output from the  entire site.  However, even in the  open space  design,
the septic systems are the predominant  source  of nutrients.

    For both case studies, the total infrastructure costs include the sum of the estimated costs of  Stormwater
management, storm drainage, paving, sidewalk, curb and  gutter, landscaping and reforestation, water, sewer  and septic
systems. In both  cases, the  open space design  resulted in a cost savings.  Costs associated  with grading, erosion  and
sediment control,  building  construction  and other  incidental costs  associated with land development were not analyzed.
In general, these  costs should be comparable between the two development options.  If anything, the  grading  and  erosion
and sediment control  costs should be  lower with  the open space design since less  land is  disturbed.

    Several other studies have also shown that open space development can be significantly less expensive  to  build than
conventional  subdivision developments. Most of the cost  savings are  due to savings in road building  and  Stormwater
management conveyance costs.  The use of open space design techniques at a  residential development in Davis,
California  provided an estimated infrastructure construction costs  savings of $800 per home (Liptan and Brown, 1996).
Other examples demonstrate infrastructure costs  savings  ranging  from 11 to 66%. Table 2  lists some of the projected
construction cost  savings  generated  by the  use  of open space redesign at several  residential sites.
                                                      203

-------
Table 2. Projected Construction Cost Savings for Open Space Designs from Redesign Analyses
Residential Subdivision
Duck Crossing
Stonehill Estates
Remlik Hall 1
% Construction
Savings
12%
20%
52%
Tharpe Knoll 2 56%
Chapel Run 2
Pleasant Hill 2
Buckingham Greene 2
64%
43%
63%
Notes
Includes roads, stormwater management, and
Includes roads, stormwater management, and
reforestation
reforestation
Includes costs for engineering, road construction, and obtaining water
and sewer permits
Includes roads and stormwater management
Includes roads, stormwater management, and
Includes roads, stormwater management, and
reforestation
reforestation
Includes roads and stormwater management
Sources:' Maurer, 1996; 2DE DNREC, 1997.
Implications for the Watershed Manager

    Better  site design reduces impervious cover, conserves  larger contiguous  natural areas,  and incorporates  more
advanced stormwater treatment, which results in  reduced stormwater runoff, increased infiltration, and  reduced nutrient
export.  Hopefully, the results of this study, as well as other redesign analyses, will answer some of the questions of local
governments,  consultants, and developers as to the benefits of better site design.

    However,  there may still be difficulties to overcome before better site design becomes a reality and common practice
in many communities. Once there is a willingness to  incorporate better site design techniques into new developments,
many communities may find that their existing development codes and ordinances are in conflict with the goals of  better
site design. For example, many  local  codes and  ordinances  require excessive  impervious  cover in the form of wide
streets, expansive  parking lots, and large-lot subdivisions. In addition, there are generally few, if any, incentives or
requirements for developers to conserve natural areas. When obstacles to better site design are present, it is a sign that
a community may want to reevaluate and consider changing some of its local codes and ordinances.

    In 1997, CWP convened a  national site planning roundtable to address this  very issue. During the 18-month
consensus-building process,  a diverse  cross section of national  planning, environmental,  home builder,  fire  and safety,
and public works organizations (as well  as local  planning officials) crafted 22 model development principles to help further
better site design at the local level. This national  roundtable is serving as a model for local government implementation
of better site design principles.

    Recently, Frederick County, Maryland, initiated a local  roundtable to take a  critical look at its own development  rules.
Members of the development  community in partnership with local  planning and zoning  and  public works staff are meeting
to identify and  overcome impediments  to better site design  that are embedded in the county's codes  and  ordinances.
The outcome of the consensus process should  be  development rules that encourage rather than  discourage the
application  of better site design techniques.

    Changing  local  development rules is not easy. Progress toward better site development will require more and  more
local governments to examine their current practices in the context of a broad range of concerns, such as how changes
will affect development costs,  local liability, property values, public safety,  and a host of other factors. Advocates of  better
site design will have to answer some difficult questions from  fire  chiefs, lawyers, traffic engineers, developers, and  many
others in the community.  Will a  proposed change make it more difficult to park? Lengthen response times for emergency
vehicles? Increase risks to the  community's children? True  change occurs only  when  the community addresses  these
and other questions to the satisfaction of all  interests.
                                                      204

-------
References

Arendt, Randall.  1994. Designing Open Space Subdivisions. Natural Lands Trust, Media, PA.

Caraco,  D., R. Claytor,  and J. Zielinski. 1998. Nutrient Loading  from  Conventional  and  Innovative Site  Development.
Center for Watershed Protection.  Ellicott City, MD.

Delaware Department  of  Natural Resources and  Environmental  Conservation. 1997.  Conservation  Design for
Stormwater Management.  Delaware Department  of  Natural Resources and Environmental  Conservation,  Dover, DE.

Dreher, D.W. and T.H. Price. 1994. Reducing the  Impacts of Urban Runoff: The Advantages of Alternative  Site Design
Approaches.  Northeastern  Illinois  Planning Commission, Chicago  IL.

Maurer, George. 1996. A Better Way to Grow:  For More  Livable Communities and a Healthier Chesapeake Bay.
Chesapeake  Bay Foundation, Annapolis MD.

Schueler, T. 1987. Controlling Urban Runoff: A Practical Manual for Planning and Designing Urban BMPs. Metropolitan
Washington Council  of  Governments.  Washington,  DC.

Schueler, T. 1997. Comparative Pollutant Removal Capability of Urban BMPs: A Reanalysis. Watershed Protection
Techniques, 2(4): 539-542. Center for Watershed Protection. Ellicott City, MD.

South Carolina  Coastal Conservation League (SCCCL). Fall 1995. Getting a Rein on Runoff:  How Sprawl and Traditional
Town Compare in SCCCL Land  Development Bulletin (Number 7). South Carolina  Coastal Conservation League,
Charleston,  SC.
                                                   205

-------
  Predicting Erosion Rates  on Construction Sites  Using the Universal  Soil  Loss
                              Equation in  Dane  County, Wisconsin


                                J.D.  Balousek,  A. Roa-Espinosa, G.D. Bubenzer
 Erosion Control  Engineer, Dane County Land  Conservation Department  (LCD),  Urban Conservationist, Dane  County
                                LCD, Professor, University  of Wisconsin-Madison
                                              Madison,  Wisconsin

                          Tools for Urban Water Resources  Management and Protection
                                Sponsored  by Urban  Water resource  Conference
                                   Northeastern Illinois  Planning Commission
                                         The Westin, Michigan Avenue
                                                Chicago,  Illinois
                                             February 7 to 10, 2000

Abstract

    The Universal Soil Loss Equation (USLE) was developed for estimating sheet and  rill erosion from agricultural fields
 under  specific conditions. Parameters  used  to  estimate erosion include  rainfall energy, soil  credibility,  slope  length,
steepness, surface  cover, and management practices. Traditionally, urban conservation planners have not  used the USLE
for estimating soil loss and evaluating conservation measures and have relied on intuition alone to locate erosion  control
practices on  constructions sites. The results of this process are often  subjective and may vary with the skill of the planner.
A USLE-based equation would provide a valuable, objective method for all planners, regardless of skill, to  tailor specific
construction  site  practices to existing conditions.  A method to predict soil loss from construction sites was developed by
adapting existing data for USLE  erosion calculations to construction  site conditions.  In addition, the  construction site
procedure was used to  create a user-friendly computer-based  program to  assist  planners in developing erosion  control
plans.  The computer program was distributed  to  engineers responsible for erosion control  planning  in Dane County,
Wisconsin. Implementation of the  USLE-based equation has proven to be a valuable tool for assessing alternatives for
site management and erosion  control. Planners are able to  uniformly implement the equation on construction sites
throughout the county, decrease the time necessary to complete a USLE calculation,  and reduce human error.

Keywords:

Universal Soil Loss  Equation (USLE), urban erosion control.

 Background

    Soil erosion, detachment of soil particles from  the soil surface, results when soil is exposed to the power of rainfall
energy and flowing water. Soil erosion causes a  loss of productivity in the land, delivers millions of tons of sediment into
waterways, and  provides a substrate  for toxic chemicals which are carried  into receiving waters.  Construction site erosion
has been identified as a significant source of suspended solids  in runoff in many parts of the United States (Hagman, et
al.,  1980; Yorke  and Herb,  1976; Becker, et al., 1974). In  the  State of Wisconsin, sediment is the largest pollutant by
volume  (Wisconsin Department  of Natural Resources,  1994). When erosion is compared on a  rate basis, construction
site erosion generates more erosion in ashortperiod of time than any other land disturbing activity (Johnson  and  Juengst,
 1997). While it is not possible to urbanize a watershed without exposing soil to erosive forces, it is possible to plan
construction  to control the production of sediment through  the  use  of erosion prevention and  reduction practices.

    The Universal Soil Loss Equation (USLE) (Equation 1) was  developed by the United States Department  of Agriculture
(USDA) for  estimating sheet and rill  erosion from agricultural  fields under specific conditions (Wischmeier and  Smith,
 1978).  The USLE enables planners to predict the  average  annual rate of soil  erosion  for combinations  of seeding and
management practices in association with  a specified soil  type, rainfall pattern, and topography. The equation groups


                                                     206

-------
interrelated  physical and  management parameters  influencing  erosion rate into  six  major factors whose site-specific
values can be expressed numerically.  More than  a half century of erosion research  in many states has  supplied
information  from  which the USLE factors were  determined.

The Universal Soil Loss  Equation.

                                         A = RxKx(LS)xCxP                                     (Equation 1)

    Where:

          A = average annual soil loss
          R = rainfall  and runoff factor
          K = soil credibility factor
          L = slope length
          S = steepness  factor
          C = cover and management factor
          P = support practice factor

    A     The computed  soil  loss in tons/acre/year.

    R     The rainfall and runoff factor is the number of erosion-index units in an average year's rain. The erosion index
          is the storm energy in  hundreds of foot tons times the 30  minute storm intensity.

    K     The soil credibility factor is the soil loss rate (tons per acre) of a specific soil type and  horizon as measured on
          a standard plot of land.

    L     The slope/length factor is the ratio  of soil loss from the actual land slope  length to that from a standard  plot (726
          feet in  length) of land. Slope length is defined as the distance from the point of origin of overland flow to  the
          point where either the slope gradient  decreases enough that deposition  begins or runoff water enters a well
          defined channel that may be part of a drainage network or a constructed structure.

    S     The slope/steepness factor is the  ratio of soil loss from the actual land slope gradient to that  from a  standard
          plot of land  (9%).

    C     The cover and  management factor is the ratio of soil loss from an area with specified cover and management
          to the  corresponding loss from a clean-tilled, continuously  fallow condition.

    P     The ratio of soil loss with a support  practice such as contouring, stripcropping, or implementing  terraces
          compared to up and down the slope cultivation. The support practice factor does not usually apply to soil loss
          on construction  sites.

    Soil losses computed  with the USLE are best available estimates, not absolutes. The USLE will generally be most
accurate for medium-textured soils, slope lengths of less than 400 feet, gradients of 3 to 18%, and consistent seeding
and management  systems  represented  in  the USDA erosion  studies. The USDA research shows  that  in  comparing actual
soil loss to  computed soil loss, 84% of the  differences in long-time average soil  losses were  less than 2 tons/acre/year
(Wischmeier and Smith,  1978). The accuracy of a predicted soil  loss depends on how accurately physical  and
management conditions on the particular  site are described by the parameter values. Large-scale averaging of parameter
values on  mixed  drainage areas reduces accuracy.

    Traditionally,  urban conservation planners have not widely used  an equation  similar to the  USLE for estimating  soil
loss and evaluating conservation  measures. They have relied on intuition alone to locate  erosion control practices on
construction sites. A USLE-based equation provides a valuable, objective method  for all planners, regardless of skill, to
tailor specific construction site practices to existing  conditions. Erosion control is  more efficient  when it focuses erosion
control practices in areas  on the site identified by the USLE as being the  most susceptible to erosion.


                                                       207

-------
    The objectives of this project were to: 1) develop a method to predict soil loss from construction sites by adapting
existing data for USLE erosion estimation to construction site conditions  and 2) create a  user-friendly computer-based
program to assist planners  in  developing  construction  site erosion control plans with  the  USLE.

Implementation Area

    The project was conducted in Dane County, located in south-central Wisconsin. Dane County has extremely diverse
and vast water resources with 475 miles of rivers and streams and 37 lakes, but these resources are threatened by rapid
urban growth. Within the next twenty years, it is conservatively estimated that an additional 72,000 people will live in the
county.  Residents recognize  how impacts to water quality affect their standard  of living, and are interested in protecting
water  resources.

    Due to the value that the citizens of Dane County place on water quality, a very restrictive erosion  control ordinance
was adopted in 1995. Any land disturbance  greater than 4000  square feet must comply with the Dane County Erosion
Control  Ordinance (Dane County, 1999). As part of this ordinance, applicants must prove that the erosion rate on  their
project will not exceed 15 tons per acre over the construction period for non-sensitive areas. In sensitive areas, including
sites adjacent to, or directly draining  to, lakes, streams,  and wetlands, the soil loss is limited to 7.5 tons per acre over the
construction period. In order to prove the  soil loss  rate is below the county standard, applicants  need to calculate the
USLE for their site from the start  of construction  until the site is  stabilized. The Dane County Land  Conservation
Department reviews erosion  control  plans  for accuracy of the plan and  compliance with the ordinance.

Met hods

Adapting USLE to  construction site  conditions

    Our first objective was to develop a method of predicting soil loss from erosion on construction sites  based  on the
guidelines given by the USDA for the USLE. In order to  adapt the  USLE to urban conditions,  each variable  in the equation
was  examined (see Equation  1).

    The rainfall  factor, R, is the first  factor  modified. Published R values represent erosivity during an average year. Most
construction sites do not remain disturbed for exactly one year. In addition, the time of year that the site is open is critical
in determining the amount of rainfall energy that will occur. In the Midwest, over half of this  rainfall energy occurs during
July, August, and September. Projects that  take  place in the  summer will experience higher  intensity storms than projects
constructed in the winter. For these reasons, the R factor needs  to be adapted to the construction schedule of the  project
(Table 1).

Table 1. Percent of R occurring after January T'for Dane County, Wisconsin.

             January         February        March           April            May            June
      1st     0              0              2              4              9              20
      15'"    0              1               3              6              14             28

             July            August          September       October         November       December
      1"     39             63             80             91              97             99
      15'h    59             72             87             94             98             100


    Once the percent R is calculated for the  interval of time that the land will be open, it is multiplied by  the annual R
factor for Dane County (150).

                                      R = (%  of R to  date) x (Annual R factor)

    The soil credibility factor, K, represents a soil's ability to  resist erosion. The  factor  is  determined  by documenting
erosion  of a soil in a bare condition on a unit test plot. The  higher the erosion rate, the higher the K factor. On construction
sites, the subsoil K factor is  often used because the topsoil is usually stripped.  Subsoil K factors can be found in USDA

                                                        208

-------
Soil  Interpretation Records. The  soil  properties that affect credibility include:  soil structure, soil particle size  distribution,
permeability,  organic matter content, and iron content.

    The slope length/steepness factor, LS, relates the length  and steepness  of the slope (Equation 2). The  rate of erosion
increases exponentially as the  length of the slope becomes longer. Erosion rates rise  even more drastically as the
steepness of the slope increases. The percent slope is a representative portion of the disturbed area, representing
overland flow, not channel flow. The slope length is measured along the flow path from the top to the bottom  of the slope
of the disturbed area.

Formula used to calculate  the LS factor.

                                   LS = (L/76.6)M(65.41 Sin28+4.56Sin6+0.065)
                                                                                      (Equation 2)
Where:   L = slope length in feet

    6 = angle of slope (in degrees)
    M = 0.2 for slopes < 1%
    M = 0.3 for slopes 1 .0 to 3.0%
    M = 0.4 for slopes 3.0  to 4.5%
    M = 0.5 for slopes > 4.5%

    The cover and management factor, C, is based on the type and condition of the  cover on the soil surface. In
construction site erosion control,  the  cover is extremely important. The vegetative cover provides protection from rainfall
impact and runoff water. If  the condition of the cover is poor, the C factor  will be high. Conversely, when the vegetation
is well established, the erosion and C factor will  be reduced. C factors for  construction sites can be found in Predicting
Rainfall Erosion Losses (Wischmeier  and  Smith, 1978). The C factors for seeding, seeding and mulching, and sod
represent the average cover over the establishment period.  Once the site is seeded or sod is installed, a period  of sixty
days during the growing season is automatically assumed for cover establishment. If the end of the sixty-day cover
establishment period falls  after the  recommended  seeding  dates, the calculation  must be carried out to the following
spring to allow for adequate growth.

Commonly Used C Factors:   Bare  ground                       1 .00
                             Seeding                          0.40
                             Seeding  and Mulching             0.12
                             Sod                               0.01

    The support practice factor,  P, is not used to calculate soil loss on  construction sites.

    The product of the R, K, LS, and C factors equals the  computed soil loss per acre over the construction period. In
Dane County, if this number is greater than the  required standard, the  project must reduce erosion below the standard
by using erosion control practices or by  changing  the management schedule. This assumes  that  100% of soil  loss is
transported and  deposited  off-site for relatively small  areas of less than 40 acres with no intervening obstructions or
flattening of the  land slope.

Developing the Spreadsheet to  Calculate the  USLE

    Implementation of the USLE  in erosion control plans was required for all land-disturbing activities greater than  20,000
square feet  in Dane County after January,  1995. The calculation of soil loss was difficult for the  consulting engineers
responsible for submitting plans.  In addition, the  USLE calculations were often done incorrectly or  the wrong data were
used  as inputs.  For these reasons,  a user-friendly computer-based  program was developed to assist erosion  control
planners with the USLE calculation.  The program  uses Microsoft Excel  97*, a spreadsheet program that is commonly
used  among the engineering  community.
                                                      209

-------
     The worksheet (Figure 1) uses the following variables and  inputs (Table 2) which are either entered by the user or
 automatically calculated in the  non-shaded rows.
                                Universal Soil Loss Equation for Construction Sites
                                            Dane County Land Conservation Department
                  Developer
                  Construction Project
Run Dale
 6/7/99
L*nd DIMurbtag
Activity

•»

<*

•»•

T

•w

*•

D»t»













%Rto
p*t»













Period %R













Annual R
Factor













SoilM*p
UnK













$0(1 ErotHblHty
K Factor













Hop»*
S













Slop* Lttigth
(f"tl L













Land DMurblng Actlvlll«i Input definition
bare ground acbvily v*itch loaves the ground devoid of veget
mulch with seed application of straw at 1 5 tony acre mlh or with
seeding temporary or permanent seeding without the us
sod instaiation ol sod
end end ot 60 day cover establishment (required inp
LS Fatter










i >



Land C«v«f C










'< -

. **•
TOTALS
Soil lot*
A-%RxR*KxLS)iC
(ten*)»tr»)


*"H 	 H

1 \- * * *





! ! : ! 1

' >• i'.

Porcont Reduction Roqulrod
t« ra**t Ordlnane*
7M tot>*(a«r«





^ 5



•H-S'.,.

if"
NONG
«0»n,,,,,.

? S f

i . -^ <



. H



'' \ ;
NONE
Won
out seeding
of mulch ng rnatenats
ul)
Figure 1.  Screen-capture of spreadsheet.

Table 2. Variables used in the spreadsheet.
Column #
1
2
3
4
5
6
7
8
9
10
11
12
13
Variable
Land Disturbing Activity
Date
% R to Date
Period % R
Annual R Factor
Soil Map Unit
Soil Erodibility K Factor
Slope % S
Slope Length L
LS Factor
Land Cover C Factor
Soil Loss
Percent Reduction to Meet Ordinance
Type
entered by user
entered by user
automatically calculated
automatically calculated
automatically calculated
entered by user
automatically determined
entered by user
entered by user
automatically calculated
automatically determined
automatically calculated
automatically calculated
Variable/Input  Descriptions:

Land-  Disturbing Activity

      The land-disturbing activity relates to the type of disturbance that is occurring on the ground and must be selected
by using a pull down  menu. Activity Inputs:

*Use of the commercial product name is for the convenience of the reader and does not imply endorsement of the product by either the Dane County Land Conservation
Department or the University of Wisconsin.
                                                          210

-------
      bare ground       Usually the initial disturbance and occurs when the ground is left bare due to stripping
                         vegetation, grading, or other actions that  leave the ground  devoid  of vegetation.
      seeding           The application of permanent ortemporaryseeding without the use of mulch. Seeding requires
                         that the user allows 60-days during the growing  season for cover  establishment.
      mulch with seed    The application of a minimum of 1.5 tons/acre straw or other comparable mulching. This input
                         is entered if the seeding  and mulching are done at the same time. It is not necessary to also
                         enter
                          seeding if this input  is used. This input  also requires a 60 day cover establishment period
                         during the growing season.
      sod               The installation of sod for cover  establishment.
      end               End is a required input at the end of the 60-day cover establishment period. If the site is
                         stabilized  by  a method other than vegetative cover, end should also be entered.
Date
      The date the planned land disturbing activity begins, e.g. 5/1 5/99. The activity is assumed to continue until the next
activity is  entered. When seeding dates are later than the dates recommended for permanent cover establishment, the
end date  must be carried out to the  next spring, rather than 60 days.
% R to Date
      The percentage of the  annual R factor from January 1 st to the entered date.
Period % R
      The percentage of the annual  R factor calculated for the period  from  one land disturbing activity to the next.
Annual R Factor
      The rainfall factor, R, is the number of erosion-index units in a  normal year's rain. The erosion index is a measure
of the erosive force of a specific rainfall. In  Dane County, Wisconsin the rainfall factor  is 150.
Soil Map Unit
      The soil map unit for the predominant soil type in the area of  the land disturbing activity.
Soil Erodibility K Factor
      The erosiveness factor of the subsoil  for the specified soil map unit.
Slope % S
      The percentage  slope  for the representative portion of the disturbed area, representing  overland flow  and not
channel flow.
Slope Length L
      Slope length (in feet) is measured  along the overland flow path from the top to the bottom of the slope of the
representative disturbed  area.
                                                      211

-------
LS Factor

      The LS factor  is calculated using the equation for LS described  previously (see Equation 2).

Land Cover C factor

      The cover and management factor  is the ratio of soil loss from an area with  a specified cover and management
practice to that  of a  unit plot of bare land. The input for the land disturbing activity  corresponds to this factor.

Soil Loss

      The predicted value of soil loss (tons/acre) which corresponds to the time period of each land  disturbing  activity.
This value is calculated  using the equation:

                                             A = %R  x R x K x (LS) x C

Percent Reduction  Required to Meet Ordinance

      The percentage value in the total row corresponds to the reduction of soil  loss necessary to comply with the Dane
County Erosion Control  Ordinance. It is  required that the cumulative  soil loss rate not  exceed 15 tons/acre for non-
sensitive areas  and 7.5 tons/acre for sites that  are  located adjacent to or directly drain to sensitive areas.

Typical Spreadsheet Example for Dane  County, Wisconsin

      Figure 2 shows a sample USLE  calculation using  the spreadsheet. The assumptions  are that construction will begin
on July 17, 1999, and  the site will be seeded and  mulched on October 31, 1999. The representative pre-existing slope
is 10% over 100 feet and the slope after grading will  be  5% over 250 feet. The soil type is Dresden Silt  loam (DsC2). The
estimated soil loss rate for this site is 15.9  tons/acre. If this site is located near a sensitive  area, the soil loss  must be
reduced by  53% to comply with the 7.5 tons/acre standard; on the other  hand, if the site was  not located near a sensitive
area,  the  soil loss only needs  to be  reduced by 6% (15 tons/acre standard).
                                Universal Soil Loss Equation for Construction Sites
                                            Dane County Land Conservation Department
                                                                        PRINT'(
                                   HEU>'",
Developer Sxarrfse csiojiatcr Ph««t — euecT ftk^e \ "un Date
Construction Project 	 SHEET I ,-?£¥* i 7/7/39
version 1 0
Lind Dltturblng
Activity

bareg^xnd •*•

™th w«h soeti *r

end -w

w

•v

••

D)i.

T 7's.S
To
3,-1'w-
TO
TV. ;







URto
Dat*

533%

79.5%

96.8%









262%

173%

32%







Annual R

150

150

150







SoHMap

DSC:

!>' 1

n=< i







Sell Erodlbilty

028

028

0 28







Slop* %

"i j%

r; '-%

') I,1*,







Slop. Lwiflth



;!o

/: )









138

085
,
OB5


' it





UndCov^rC

1 00

0.12









TOTALS
5«H fat*
A-ttRxRxKxL$xC
(WnTfierM

152

0 7


, n f t






15. B
Pm»nt Reduction R«qulr«d
to MMtOrdkiini*
T.B toniJacrt







'. *



••? * '* *
53%
15.D tantfecr*





< -1

'' ]e '-' l



* <\\
6%
                  Land Disturbing Actlvltl**
                                    bare ground
                                    mulch with seed
                                    seeding
diflnttlor
activity which leaves the ground devoid of vegetsbon
application of slraw at 1 5 tons/acre with or without seeding
temporary or permanent seeding without the use of mulching materials
installation of sod
end of 60 day cover establishment (required input)
Figure 2. Sample USLE calculation.
                                                        212

-------
Results and  Discussion

      There are several advantages to using the adapted USLE for erosion control  planning on construction sites. One
advantage is being able to locate areas with the highest erosion rates, which results in  more effective erosion control. If
one portion of a construction site is predicted to have a higher erosion rate, more or larger erosion control practices may
be targeted  in that area, while less intensive practices may  be required  elsewhere on the site. The adapted USLE also
facilitates the design of sediment ponds and other erosion  control practices. The predicted amount of soil  loss exceeding
the standard can be used to calculate the  percent reduction necessary to comply with the ordinance.

      Another advantage is that the adapted USLE  brings in the important element of time. In Wisconsin, the majority
of the year's rainfall erosion occurs during the summer months. Summer is also the time of year that most construction
is occurring. The USLE accounts for the date and  duration the  development project occurs and predicts the soil's
vulnerability to erosion at that time. The USLE may show that staging the construction project will help to  reduce the soil
loss on the site.

      The spreadsheet program has proven to be a  valuable  tool  for calculating the soil loss. The program has been
distributed for more than a year, free  of charge, to the planners and consultants in Dane County. The County's review
of the calculation in the erosion control plans has become easier and quicker by having a printout that  summarizes
the variables used. An advantage of having tables  and formulas included in the spreadsheet,  is the consistency that is
achieved by everyone using  the same parameters.  Not only have  the calculations of soil loss  been more precise and
time schedules more realistic, but planners  and  consultants  have stated that it  has  saved them time and simplified the
calculation   process.

References

Becker,  B.C.,  Nawroki, M.A.,  and Sitek, G.M., (1974), An Executive Summary  of Three EPA  Demonstration Programs
in Erosion and Sediment Control, Hittman and Associates,  Inc., Columbia, MD.

Dane  County Board of Supervisors, Dane County Code of Ordinances (1999), Chapter 14, subchapter II, Erosion Control
System, Dane County  Board of Supervisors, Madison, Wl.

Hagman, B.S., Konrad, J.G., and Madison, F.W. (1980), Methods for Controlling Erosion and Sedimentation From
Residential Construction Activities,  Wisconsin Department of  Natural Resources, Madison,  Wl.

Johnson, C.D. and Juengst, D.,  (1997), Polluted  Urban Runoff: A Source of Concern, University of Wisconsin-Extension,
Madison, Wl.

Wischmeier, W.H.  and  Smith,  D.D.,  (1978), Predicting Rainfall Erosion Losses-  A Guide to Conservation  Planning,
United States Department of Agriculture, Washington, D.C.

Wisconsin Department of Natural  Resources, (1994), The Wisconsin Water Quality Assessment Report to Congress,
PUBL-WR 254-94-REV, Wisconsin  Department of Natural  Resources,  Madison, Wl.

Yorke, T.H.  and Herb,  W.J., (1976), Urban Area Sediment  Yield Affects of Construction Site Conditions and Sediment
Control  Methods, United States Geological Survey,  Parkville, MD.
                                                     213

-------
                        Public Involvement Programs That Support

                                   Water Quality Management


                                             Josephine Powell
                                 Wayne  County Department of Environment
                                              Detroit,  Michigan

                                               Zachare Ball
                                               Karen  Reaume
                                 Environmental Consulting & Technology, Inc.
                                              Detroit,  Michigan


    The Rouge River, a tributary to the  Detroit River, in southeast Michigan,  has been documented as a significant source
of pollution to the Great Lakes System. The Rouge River Watershed spans approximately 438 square miles in 48
communities and three counties and is  home to  over 1.5 million residents. The eastern  portion of the watershed contains
much of the older, industrial areas of Detroit and Dearborn. The western and  northern portions contains newer suburban
communities and areas under heavy  development  pressure.

    This paper discusses the  programs used  by the Public Involvement Team of Wayne County's Rouge River National
Wet Weather Demonstration Project (Rouge Project) to (1) increase watershed awareness of Rouge River Watershed
residents and business owners, (2) educate them about pollution sources  to the Rouge River and  (3) involve  them in
restoration of the Rouge River by showing them that small changes in their daily activities can help improve water quality
and  restore the  river.

    The Rouge Friendly Neighborhood  Program was piloted over a two year period in watershed neighborhoods in three
distinctly different areas of the watershed. All neighborhoods were surveyed to determine the initial  level  of knowledge
about water quality  issues, lawn care maintenance, and  pollution  prevention practices.  Survey  results were  used to
fashion a neighborhood program for each area. All three neighborhoods  received Rouge Friendly brochures, newsletter
articles, and other materials.

    The Rouge Friendly Business Program, a  companion program to the neighborhood effort, sought to educate small-to-
mid-sized businesses about how they can positively impact the Rouge River by making small changes to daily business
practices.  Since auto-related businesses are very common in the Rouge River Watershed, an automotive services
roundtable was  convened. The partners included representatives of automotive service associations,  the local chamber
of commerce, and businessmen who met periodically for a year  to review  draft materials, make suggestions about the
program's promotion, and to help mold the program before it was implemented. Once  implemented, the industry
representatives promoted the  program  in their publications and  recruited  businesses to participate in the program.

    This paper will describe both of these pollution prevention programs and  discuss how the Brightmoor  neighborhood
in the  Rouge River  Watershed was impacted by the Rouge Friendly Neighborhood and the Rouge Friendly Business
Programs.

The Rouge Friendly Neighborhood  Program

    The Rouge Friendly Neighborhood Program was designed to be  carried out by responsible neighborhood
organizations. Preferred  prerequisites  were:

    1.  The group participating in the program must represent a defined area or neighborhood.
                                                   214

-------
    2. The group would participate in the Friends of the Rouge River Watch  Program. The river system need not pass
       directly through the neighborhood for  participation.  An assigned segment could be  identified for the group by
       Friends of the Rouge.

    3.  The group would participate in the Friends of the Rouge Storm Drain Stenciling Program. The stenciling of storm
       drains should include, but  is not limited to, all the storm drains within their designated neighborhood or  area.

    4.  The group should  actively participate and/or  encourage proper household hazardous waste management. This
       could  occur  through:

            Reduced purchasing  of hazardous house  chemicals

            Proper  use of household   hazardous chemicals

            Proper  disposal of hazardous  household  chemicals

            Use of less-toxic alternatives to  household hazardous chemicals

        The group can accomplish this requirement  by distributing information concerning  proper household  hazardous
        waste management to  their designated  neighborhood.

    5. The group would facilitate education of residents regarding  non-point source pollution. Information would be
       provided  by  the Rouge Project  Team  for distribution to the designated  neighborhoods.

    6. Submittal  of semi-annual reports discussing the  activities that have  been  taking place could  be a requirement to
       maintain  Rouge  Friendly  Neighborhood  status.

    Three Rouge Project area  neighborhoods  representing different demographics and development  history were  chosen
as pilots for the Rouge Friendly Neighborhood Program.  They were the (1)  Brightmoor area of Detroit, an older, developed
area of the watershed along the Main  Branch  of the  Rouge River;  (2) Golfview  Manor subdivision in Dearborn Heights,
a newer subdivision  along  the Middle Branch of the Rouge River;  and (3) West  Bloomfield  Place, a subdivision in West
Bloomfield  Township,  a developing area along the  Upper  Branch  of the Rouge River.  These three pilots represented
communities with diverse  demographics and  concerns.

    The  Brightmoor  neighborhood was a  deteriorating area with  strong  community activism regarding  neighborhood
problems and  concerns. The neighborhood also  showed strong stewardship for the Rouge  River, which serves  as  a
western boundary to the neighborhood and flows through a nearby park. Golfview Manor in Dearborn  Heights was  a more
upscale, manicured neighborhood that was very active through its subdivision association, but did  not have a real
connection  to the River. West Bloomfield Place in West Bloomfield Township was  an  upper income,  less urban  area
bounded by a wetland.

    Meetings were  held with a core  group of representatives from each neighborhood to garner support from the
neighborhoods' leadership  and to  discuss what the program was and what the expected  outcomes were.

    The residents of all three pilot areas were sent a survey, distributed by mail  or door-to-door, to document their
knowledge  of Rouge  River water quality, storm  water issues, and household  hazardous waste disposal.  In addition,
respondents were asked for demographic information.  The survey information was  used to determine  what the Rouge
Friendly Neighborhood Program should focus on in each  particular neighborhood.

    Educational  materials  that had been developed  about storm water pollution,  household  hazardous  waste (and its
disposal), and watershed  awareness  were reviewed by each group. Each core group helped develop the particular
program that would  be implemented in  their  neighborhood, because Rouge Project staff knew that no program would be
a success without the core groups' support and endorsement. These core group members were  relied on to explain the
program  at neighborhood  meetings.


                                                     215

-------
The  Rouge Friendly Business Program

    The Rouge Friendly Business Program was developed as a partnership between Wayne County, local government,
and the business community to restore and protect the Rouge River. To accomplish this goal, information and assistance
are provided to small business owners to teach them how they can positively affect the water quality of the Rouge River
by changing some of their everyday practices. The  education process is not about major contaminants, but those little
things that slip the mind, such as keeping the dumpster lid closed and storing materials under cover. These simple  actions
can affect water  quality because they stop pollutants from entering the storm system. As  an incentive to participate,
Wayne County embraced the concept that businesses  in the watershed  that demonstrate  stewardship and  a  strong
environmental  ethic should  be recognized  by the community  for their voluntary participation. As such, these businesses
should enjoy greater name recognition through the efforts  of the local and regional media as well as specific program
materials,  such as decals and magnets  that identify  the business as Rouge Friendly.

    The Rouge  River Watershed has approximately  42,000  businesses in its  48 communities and three  counties. To
design program  materials that would have the greatest impact, three criteria were developed to target business  types.
They  are:

    .  The business has a  high incidence of illicit connections to storm drains

    .  The  business conducts a significant number  of pollutant-generating  activities outdoors

    .  The business is found  in large  numbers in  the watershed

    Using these criteria, six types of businesses were selected and specific activities  identified. They are:

    .  Vehicle  Service  Industry

    .  Food  producers, grocers, and  eating  establishments

    .  Metal Machining

    .  Earth Disturbing  Construction

    .  Remodeling and  Repair Contractors

    .  General  Business

    Pollution control criteria  were established for each kind of business. These criteria were used to create a self-
assessment form to  be used by business owners to evaluate  how "Rouge  Friendly" their businesses are. Best
Management Practices (BMPs) were also written that correspond to each activity and this information was put into the
booklet along  with a self-assessment form for distribution.

    Representatives of various trade organizations were  invited to participate in a Vehicle Service Industry Roundtable.
The roundtable was asked  to review and  comment  on the educational  materials, the self-assessment form, the  BMPs,
and the best way to conduct program outreach. Rouge Project staff sought to engage businesses in an ongoing dialogue
to determine what approach would work  best, with a secondary goal of determining how to get businesses to  participate.
Feedback from this group resulted in a name change from the "Clean Business Program" to the "Rouge Friendly Business
Program." This was not a quick process, but took approximately six months  of meetings to (1) form a Vehicle Service
Industry roundtable, (2)  explain the  purpose of the Rouge Friendly Business Program, and  (3) refine the  program and
products. The  Rouge Friendly Business Program elements were  finalized  as follows:

    .  Self-assessment form and action  plan

    .  Best Management Practices


                                                      216

-------
    . Site visits by technical staff

    . Recognition  materials for  participating  businesses  (stickers and  magnets)

    . Business  pledge  and  newspaper recognition

    After these materials were finalized, Wayne County and Rouge Project staff promoted the Rouge Friendly Business
on  a pilot basis and recruited businesses through the  following mechanisms:

    . Business  Roundtable  contacts

    . Letter and telephone  contacts

    . Door to door contact  with  businesses

    . Contact  through   homeowner/neighborhood  associations

    . Integration of Business and Residential Programs

    By the end  of the  pilot  period, the Wayne County Department of Environment had recognized  20 businesses as
Rouge Friendly.

The Brightmoor Community  Pilot

    The Brighmoor area  of Detroit was developed in the 1920s as a neighborhood  forworking-class families. Most  houses
are frame, with the newer areas  of the neighborhood (1940s and 1950s) of brick construction. Over the past 15 years,
the Brightmoor area has deteriorated. Its once vibrant  business strip is dotted with boarded, vacant buildings,  graffiti,
trash, and debris.  Whole blocks of residential land are vacant and  overgrown  and illegal dumping  is abundant.
Environmental  abuses ranged from a  myriad of abandoned vehicles to illegal car repair businesses  on  residential  streets.
The Rouge Friendly Neighborhood Survey (Attachment A), distributed in Brightmoor in 1996, showed that the top two
environmental  concerns  in the area were  illegal dumping and abandoned housing.

    Despite  these  challenges, the Brightmoor  neighborhood  had  two characteristics that made  it  a viable pilot  for the
Rouge-Friendly Programs. One, Eliza Howell Park, located on its western edge, was traversed by two branches of the
Rouge River. Second, Brightmoor had a wealth of grassroots  organizations who were working to  make the neighborhood
better. Some annually removed  log jams and other debris from the Rouge River in Eliza  Howell Park during  Rouge
Rescue, sponsored by Friends of the Rouge,  a grassroots organization serving the whole watershed.

    Initial contact was made with the  Brightmoor Concerned Citizens and other neighborhood representatives  in January,
1996. The group agreed  that they would like to participate as a Rouge  Friendly Neighborhood pilot. A month later, the
same group met again  with Rouge Project staff. This time, city parks staff were  present. They were told about the
possibility  of grant  funding for storm water projects by the Rouge Project.  The group brainstormed the kinds of things
they would like to  see happen at the  park, which had suffered from spotty maintenance. They agreed  that  they would
like to see wildflowers  and  prairie grass  planted, nature  trails restored,  and a community  garden created. The parks
department later applied for  and  was awarded a $180,000 grant  to plant wildflowers and prairie grasses and to install
nature trails  in the  lower  end of the park, near the Rouge River.

    The next step in the program was to survey residents about their knowledge of pollution entering the river and
household hazardous waste  disposal, their neighborhood environmental concerns,  and demographic information.  The
survey was created with  input by the core neighborhood group. The major data extracted from the survey were:

    .  78% thought  the Rouge River was polluted  or very polluted, and 20% thought the river was getting worse.

    .  38% did not  know that the storm drains lead directly to the Rouge River. However, 56% understood that sanitary
      sewers go to the  wastewater treatment  plant.

                                                      217

-------
    .  18% correctly answered that industry pollutes the Rouge River the least and 66% thought stormwater pollutes the
      least.
    . 87%  maintained their own lawn.  There was an even distribution among those who never fertilize their lawns and
     those who fertilize I-2 times per year.
    . 75%  did not  change their own motor oil.
    .  80% took their cars to a car wash instead of washing it themselves.
    .  92% claimed indicated that they know what  household hazardous waste is, and  73% correctly identified motor oil
      as a household hazardous waste. However,  54% did not properly dispose  of their wastes.
    . 83%  said they were committed/very committed to make small  changes to prevent  pollution.
    Following are neighborhood issues,  in order of importance:
    1.  Abandoned buildlings
    2.  Illegal dumping
    3.  Household  hazardous material  disposal
    4.  Infrequency  of street sweeping and storm  drain  cleaning
    5.  Recycling
    6.  Do-it-yourself  car  repair/illegal car lots on  residential streets
    7.  Overuse of garden/lawn pesticides
    8.  Overuse of fertilizer
    9.  Composting
    Wayne County Rouge Project staff,  Friends of the Rouge,  and Brightmoor Concerned Citizens leadership made a
presentation, including  survey results,  to the general membership in May, 1996.  The general membership was
enthusiastic about the program.  The annual  Rouge Rescue held in Eliza  Howell Park on  June 1,1996, was expanded
to include other activities, including storm drain stenciling, a tour of a newly constructed combined sewer overflow basin,
and children's games.
    Subsequent meetings with the Brightmoor group were used to brainstorm  what  the specific  program  elements should
be and what outcomes were expected. The  following elements were supported by the core group:
    .  Urban gardens  on vacant lots
    .   Composting  education
    . Attempting to get rid of the massive log jam  at the confluence of the Upper and Main Rouge River in Eliza Howell
      Park
    . A tour of the area for the Detroit Environmental Court judge
    .  Lawn signs that read "I  support the Rouge Friendly Neighborhood  Program"
    .  Early recognition of well-maintained lawns  and gardens; Brightmoor's "Resident  of the  Month"
                                                     218

-------
    .  A  Brightmoor  Rouge-Friendly  Business  Program

    .  A renters' workshop to educate tenants about their rights and responsibilities and the responsibilities of landlords

    .  Educational  materials in  the  various neighborhood  newsletters

    All of the  activities were implemented except the lawn signs, the recognition  of well-maintained gardens,  and the
renters'workshop. By the fall of 1996, the focus had shifted to conducting a monthly combined resident/business owners'
meeting to include businesses, which were primarily vehicle service oriented,  into the Rouge-Friendly  initiative.

Results and Outcomes

    Rouge-Friendly Neighborhood Program: Because of the enthusiasm and commitment of Brightmoor residents and
business owners, many activities were  conducted. They were:

    .  Thousands of educational brochures and children's materials were distributed in community centers, businesses,
      schools,  and  newsletters.

    .  A local business owner successfully sued a  public utility that was pumping hundreds of gallons of polluted
      stormwater into the  Rouge River.

    .  The  local community  organization  not only enthusiastically participated in the annual Rouge Rescue event, but
      conducted another such event on its own.

    .  Through  a partnership with the city  parks department,  the Greening of Detroit, and  the Brightmoor Concerned
      Citizens,  100 trees were planted in Eliza  Howell park by  200 local elementary  school students.

    .  City officials agreed  to conduct an environmental ticket  blitz in the  neighborhood, which  resulted in the following
      tickets being written:  179 parking  tickets, 71  abandoned  cars tagged for  removal, 8 stolen cars being towed, 15
      public works tickets  for bulk garbage being put at the curb too early, 2 environmental protection tickets, and 47
      tickets for inoperable  vehicles.

    .  A monthly meeting that included  neighborhood residents, business owners,  non-profit organizations, city officials
      (including  police commanders),  and county officials  focused on  environmental issues.

    Rouge-Friendly Business Program: While many outstanding initiatives were accomplished by meeting with the
Brightmoor stakeholders monthly, only one Brightmoor business  was recognized as Rouge Friendly  after ten months of
monthly meetings. The meetings were well-attended and business owners felt comfortable discussing their environmental
concerns. In June, of 1997, the approach for recruiting Brightmoor businesses as Rouge Friendly was changed. A
community leader was paired with a technical staff member and they proceeded to visit neighborhood businesses.  They
visited 14  neighborhood businesses  several times over a two-month period. Information about the program was left with
the business owners, as well as an offer from the technical staff member to  help the business owner with the self-
assessment form. Through this effort,  Wayne  County recognized  six additional Brightmoor businesses as Rouge-
Friendly.  This was a  successful  (43% participation),  but labor  intensive, method of recruiting  businesses.
                                                      219

-------
              ATTACHMENT A
                       ME ROUGE RIVER PROJECT
                       A WORLD CLASS EFFORT
                       BRINGINGOlMIVEFBACKTaifE
                           Are you willing to prevent pollution
                           in your neighborhood?

                           Start today by filling  out this questionnaire.
               We are working with the Rouge River National Wet Weather Demonstration Project to make our
               subdivision a more attractive place to jive. We have been chosen as one of three pilot neighborhoods in the
               Rouge River Watersheds to participate in a pollution prevention program that may be used as a model for other
               urban watersheds across the country. In order to design a program that best fits our needs, we need you to
               answer a few questions. The following survey is voluntary and confidential. Use the enclosed  pre-stamped
               envelope to return the questionnaire by April 22, 1996.
                2.
                3.
                                     Very polluted     Somewhat polluted      Not  polluted
Do you think the Rouge River is polluted?       54321

                                    Getting cleaner     Staying the same      Getting worse
Do you think the Rouge River Is getting         5        4         3         2         f
cleaner.staying the same, or getting worse?

                                          Q  To a storage tank under the ground
                                          rj To the Rouge River
                                          rj  To the waste water  treatment plant
                                          n Don't know
Where does water go when it enters an
outside storm drain in your neighborhood?
                    Where does water go when it
                    is flushed down the toilet or sink?
                5.   What pollutes the Rouge River the LEAST?
                6.   How do you maintain your lawn?
                7.   How often is your lawn fertilized?
                a.   Where do you change the oil
                    in your car?

                9.   Do you usually wash your car
                    or take it to a car wash?

               10.   If you change your oil at home,
                    how do you dispose of it?
                11.   Do you know what household
                    hazardous materials are?

                12.   Which of the following is a
                    household hazardous material?
                                           rj  To a storage tank under the ground
                                             BTo the Rouge River
                                              To the waste water treatment plant
                                             Don't know
                                           a

                                           a
                                           a
                                           a
                                           B
                                               Combined sewer overflows
                                              (a mixture of sewage and stormwater that
                                              flows into the river when it rains.)
                                              Stormwater (water that runs off
                                              the grvund and enters the river)
                                               Industry

                                               Paid professional company
                                             Paid neighbor
                                               Someone in the household maintains it
                                           U  1 to 2 times per year
                                           rj  3 to 4 times per year
                                           rj  More than 4 times per year
                                           rj  Never
B
                                             Auto Repair Shop/Quick Oil Chnage
                                               Yard rj Street Q   Driveway
                                          n Wash it myself
                                          Q   Carwash
                                          d In the garbage
a                                             In the sewer
                                               Don't know

                                          n Yes
                                          n  NO
                                          Q  Baking soda
                                          rj Motor Oil
                                                                  rj On the ground
                                                                  fj Take to facility that
                                                                     accepts used oil
                                                                  rj Lemon oil
                                                                  Q Vinegar
                                                              220

-------
13.   How do you dispose of your                    0  Put it in the trash       Q   Dump it down the sink
     household's hazardous materials?              rj Dump it on the ground Q Don't know
                                                  fj  Take  it to a Household Hazardous Drop-off Area

14.   Indicate whether the following environmental issues are
     very important, important, or not important to your subdivision.
Very important Important
Overuse of fertilizer 5432
Composting 5432
Abandoned buildings 5432
Frequency of street sweeping 5432
and storm drain cleaning
Overuse of garden/lawn pesticides 5432
Recycling 5432
Household hazardous waste 5432
Illegal dumping 5432
Do-it-yourself car repair / Illegal 5432
car lots on residential streets
Other (soecifv: ) 5432
15. You can make small changes to prevent pollution (i.e. the type of fertilizer you purchase,
pose of your motor oil, etc.). What is your level of commitment to make these changes?
Very committed Somewhat committed
5432
16. How many people, including yourself, Q ^ Q 5
live at this address? Q Q 6
Q 3 n More
D 4
17. How many of these are children? D 0 D 3
D 1 D 4
rj 2 a More
18. How many pets do you own? Ej Q 3
n o n 4
Q 2 Q More
19. What is your gender? rj Female • : Male
20. What is your age group? Q Under 18 years n 46-60
Not important
f
f
1
1

1
1
1
1
1

1
how you dis-

Not interested
I


than 6


than 4

than 4

years
rj 18-30 years Q Above 60 years
O 3 145 years
21. What was the last grade rj Some high school
you completed in school? rj Completed high school
Q Post-high school training
Q Some college
Q Completed college
l_In Graduate or professional school
Thank you for doing your part in cleaning up our subdivision! Remember, return
in the pre-paid envelope by April 22, 1996.
Any questions should go to John or Shelley Mlynarczyk at 533*3453







the questionnaire


                                                 221

-------
                              The  Water-Wise Gardener Program:
                      Teaching Nutrient  Management to  Homeowners
                                                Marc T. Aveni
                                        Virginia  Cooperative  Extension
                                             Manassas,  Virginia
Introduction

    The Water-wise Gardener program was developed by Virginia  Cooperative Extension (VCE) seven years ago with
funding from the Cooperative State Research, Education, and  Extension Service (CSREES)  of the U.S. Department of
Agriculture (USDA). It is an educational program aimed at reducing non-point source pollution from suburban residential
areas. The educational focus is  upon nutrients,  especially nitrogen and phosphorus from lawn fertilizer over-application
or misuse. The program seeks to  reduce such nutrient pollution to Virginia waterways, and eventually the Chesapeake
Bay, through the recruitment  of homeowner participants from impaired watersheds.  Participating homeowners attend
educational seminars on lawn best management practices, are partnered with a Master Gardener volunteer,  and are
expected to keep accurate records and  implement recommended practices. The program, which is currently being
implemented in 12 urban/suburban Virginia counties, is supported by a combination of local county  funds, grants from
the Virginia Water  Quality Improvement Act, and funds from USDA. Cooperative Extension  Units in North and South
Carolina have  replicated the program.

How the Program Works

    The Water-wise Gardener program begins  by  recruiting  homeowners from watersheds with  impaired streams or
other identified problems to participate in  a year-long lawn  care educational program. The most successful recruitment
method to-date has been to conduct a "reverse search" on the  Internet by street name. Once names and addresses are
identified, a recruitment letter is sent personalized for the watershed; e.g., "Dear Resident of the Bull Run Watershed."
The letter invites the homeowner to participate in the program and lists the benefits of participation, such as free
seasonal seminars  with regional  experts, visits from a Master Gardener (volunteers trained by VCE in various aspects
of horticulture), a free soil  test, and Virginia  Tech publications.  In order to be enrolled in the  program, the  homeowners
must return a completed pre-survey and a signed agreement form that details their  obligation to the  program. The pre-
survey asks questions about their lawn care  practices and attitudes before program involvement, as well as demographic
information  such as race, gender,  income,  and education  levels. A  stamped, self-addressed envelope is included  for
ease of return. For every  100  letters sent out, between 20-30 are typically returned. A simple database  program keeps
track of participants and  their lawn care data. A reporting system on the Virginia Tech Intranet  is currently being
designed to record this information on  a statewide  basis by hydrologic unit.

    Once enrolled,  participants are assigned  a  personal Master Gardener. The Master Gardener schedules a visit with
the homeowner to discuss  his or her lawn. All Master Gardeners are instructed to stay outside on the lawn and not to go
inside anyone's  home. Some choose to  bring along a spouse, friend, or another  Master Gardener. At this visit, the
Master Gardener works with the  homeowner to correctly  measure the square footage of lawn area, determine  the type
and variety of grass, collect a soil sample, and ascertain  previous fertilization practices and amounts previously applied,
if known. This  information  forms the basis of a personalized lawn care plan for the homeowner. Master Gardeners also
answer other questions the participant  may have; common questions include weed  and  pest identification, what plant
grows best where,  and why certain plants are not thriving. The Master Gardener leaves a business  card with  a phone
number or e-mail where  he  or she can  be reached for further questions throughout the  program year. All Master
Gardeners receive 50 hours of classroom training as well as supplemental field training  before being assigned to
homeowners. Typically, a  Master Gardener  will  be  assigned to between 5-I 0 homeowners.


                                                    222

-------
    In addition to the one-on-one visits from Master Gardeners, homeowners attend seasonal seminars on timely topics
of interest to those  with lawns.  In Northern Virginia, where cool-season grasses like Fescue and Bluegrass predominate,
fall topics include soil testing, fertilization, core aeration, and over seeding. Spring topics include  mowing  and pruning,
integrated pest management,  and proper watering and  planting. Popular locations for seminars include parks with
covered  pavilions, school auditoriums, county buildings, and  libraries.  Any easily accessible public  location large  enough
to hold 50 to  100 people  comfortably, and accessible to wheel chairs, will work. If held inside, cold temperatures, rain,
or wind  are not a  problem; however, an outside area for demonstration purposes is essential.  State and regional
Cooperative Extension experts  are recruited for the seminars to answer questions. Master Gardeners  are also present,
with various displays, to answer questions and to meet with their assigned participants.

    A professional-quality newsletter is  sent to all participants approximately six times per year. A  grant-paid editor
solicits articles that reinforce or complement topics taught at the seminars.  Articles  on various  aspects of watershed
management are also introduced. The newsletter is made available  electronically to  other Extension Agents for editing
and  reproduction elsewhere.

    After participants  have attended fall  and spring seminars, they are visited again by their Master Gardener to collect
final lawn data and  conduct a post-survey of practices and attitudes. The most important piece of data collected is the
amount of fertilizer  now being applied. Square footage of turf can be re-checked, if  needed, and questions answered.

    The  homeowner may  chose to  participate again the following year, or to offer their lawn as a demonstration lawn,
and erect a sign in  their yard to promote  the program  in the  community. The post-surveys and data sheets are collected
from all participants annually. Results are compiled and analyzed by a grant-paid technician  and  a final report generated
for each Cooperative  Extension unit as well as an overall report  for statewide efforts.

Results

    Data for the period March  1998 to  June  1999 for the Virginia counties of Arlington, Loudoun,  and  Prince  William
shows 326 individual homeowner participants.  These 326 homeowners managed  57.1  acres of turf  in 11 different
hydrologic  units in  the  Northern Virginia area. Between  100-200 additional individuals  attended  seminars  but  did not
participate  in the pre- and post-survey and data  collection.

    Accurate information on amounts of nitrogen and phosphorus applied  by participants before program  involvement
is difficult to get. Most did not remember how much  fertilizer they had applied in  the previous year.  Many stated  the
reason they joined the program was in order to understand how much fertilizer to apply. A total of 72 participants
reported  pre-program  fertilizer application of 1,062 pounds of nitrogen. The same 72  participants  reported 762  pounds
of nitrogen applied  after program involvement, or a reduction of 300 pounds. Information on pre- and post-phosphorus
was not  collected.

    Pre-surveys indicated that only 12%  of all participants had  soil tested for their lawns prior to  applying fertilizer.
Homeowners not testing soil are more likely to apply excess fertilizer. For this reason, Virginia  Tech  recommends  soil
testing as a nutrient management  practice for home  lawns. Post surveys show 95% of participants returning surveys
tested soils after program involvement.  Another  important  nutrient  management practice for homeowners with cool-
season turf is to fertilize in the fall, when uptake by roots occurs best.  Pre-surveys indicated that only 32% were fertilizing
at this time of year,  while post surveys indicated that 64% were fertilizing in the  fall. Similar increases were also  observed
for recommended practices such as aeration (from 34% to 83%),  and over-seeding (from 35% to  76%). An increase in
the number of  participants not watering the lawn at all in the summer also  increased (from 18%  to 44%) (Figure 1).

Demographics from the program  indicate that 72% of participants were male and 28% female. Participants were
overwhelmingly white (89%); followed by black (7%), Asian (4%), and  Hispanic (1%). The majority (42%) had a four-year
college degree  and  a gross family income of over  $70,000  a year (54%). More than  one-third of the lawns were between
5,000 and  10,000 square  feet  (35%).
                                                      223

-------
                                             Pre-program D Post-program
                       w  100  	p=r
                    c  |  80
                    o  o  60
                    CD  "•=  AC\
                    fT  CO  ^«
                    °-^  20

                       °   0
                                  Figure 1.  Pre and Post Program Practice Adoption
Figure 1. Pre and post program practice adoption.

Conclusions and Lessons Learned

    The Water-wise Gardener  Program was successful at reaching the intended  audience  and achieving adoption of
nutrient  management practices.  Based on the success of the program, it appears that  suburban homeowners can be
recruited to maintain their lawns according to recommended practices.  Homeowners are willing,  with the  help of Master
Gardener volunteers in some cases, to keep records on  their nutrient  use as a part of program participation. Although
326 individuals and 57 acres of turf may seem low for an area like Northern Virginia, it is significant for a populace  that
does not traditionally participate  in water quality educational  programs. Considering that most lawns in  suburban
subdivisions have a turf area of around 5,000 square  feet, clearly many individuals will need to be enrolled to reach
meaningful  numbers.

    The study showed that it is difficult to  obtain  information on  pre-program  nutrient use for  most participants. Most
homeowners cannot provide accurate nutrient use data from the previous 6-12 months. They simply do  not  remember
how much nitrogen and phosphorous was in the fertilizer bag applied last spring  or fall. However, after program
involvement, they do appear to  understand how  much  nitrogen they applied and the square footage of their  turf. From
a water quality public policy perspective, it may be preferable to record nutrient use after program involvement and
consider participants' turf square footage as the  urban nutrient management measurement. In this way, the focus could
be upon recruiting more  and more individuals to participate in nutrient management educational programs like the Water-
wise Gardener,  thus  increasing the number of acres addressed by urban nutrient management efforts. Such an
approach could easily be integrated  into local Geographic Information  Systems, providing localities a simple method of
accounting for and reporting on urban nutrient  management. Localities interested in a program like the Water-wise
Gardener should contact their local Cooperative  Extension office to see if a similar project is already occurring  or could
be developed. As this program is  being continued and  expanded in the 1999-2001 time frame, the opportunity  to better
define what is realistic as an urban nutrient management measurement for homeowners  will hopefully occur.
                                                     224

-------
             Chicago Wilderness; Toward an  Urban Conservation Culture
                                          John D. Rogner
                                    U.S.  Fish and  Wildlife Service
                                        Barrington, IL  60010
Chicago Wilderness - Origin and Purpose

    We in the Chicago conservation  community have been using the word "wilderness" in a highly unconventional  context
since 1996. We have  coined the term "Chicago Wilderness" to refer to the rich biodiversity which resides in and  around
this  huge, sprawling  metropolitan area, extending from southeast Wisconsin, through the six-county metropolitan area
in Illinois,  and around  Lake Michigan to northwest Indiana. This is a  region which most people think of as anything but
"untrammeled by man, where man  is a visitor who does not remain," in the words of the Wilderness Act, which has
defined our  modern concept  of  wilderness.

    This is an area that  is associated with-indeed, defined by-humans and their  cultural footprint.  Although the
"wilderness" is scattered throughout the region, mostly in parcels that would be considered  slivers of land by conventional
wilderness standards,  it totals over 200,000 acres of land protected within a complex of national tallgrass prairie, national
lakeshore, county forest preserves, city and township parks, and similar preserved  public lands. Its  protected lands and
waters range from half-acre remnants to the 15,000-acre Midewin National Tallgrass  Prairie,

    Within this system of preserves can be found some of the largest  and  best woodlands, wetlands, and  prairies in the
Midwest. These lands are set in a  much larger matrix of public and private, developed and undeveloped lands that
support nature and the region's  8 million people.

    We have  called these lands "wilderness,"in part to draw the attention of people who  are focused on  Chicago's  cultural
attractions to the existence of these lands in their own metropolis, and in part to deliberately blur the distinction, or
conversely, emphasize the connections between  formal wilderness in  remote and inaccessible places and wild lands in
the places where people live and work. The biotic connections  exist on the  land, and they  ought to exist  in people's
minds, as well.

    The boundaries of the Chicago Wilderness region do,  in fact, capture a spectacular  concentration of rare ecosystem
types.  These  ecosystems harbor a high  diversity of species, including a large  number of those listed  as  threatened or
endangered in the states  of Illinois, Indiana,  and Wisconsin.  Outside of the metropolitan  area, particularly in rural  Indiana
and  Illinois, diversity decreases  sharply  as  agriculture dominates  the landscape.

    "Chicago  Wilderness" is also the name we have given the collaboration  of over 90 organizations in the Chicago region
that  have banded together to  better protect,  restore, celebrate, promote,  and publicize  our rich biodiversity. An
unfortunate and perhaps somewhat inevitable consequence of urban life  is a detachment from the land; thus, a principal
goal  of the partnership is  to reconnect a  landless urban population, in  Aldo Leopold's words, to the "raw material out of
which  we  have hammered that  artifact called civilization."

    Despite the richness of nature and opportunity for conservation in  the region, evidence suggests the Chicago region
is experiencing a decline in  native species and  communities.  Prior to protection, much of the region's current base of
protected land was subject to agriculture, drainage,  and other human  influences which reduced or eliminated  native plant
and animal communities. These areas are often fragmented and isolated  from healthy lands which could otherwise serve
as immigration  sources for native species.
                                                     225

-------
    Small fragments are also subject to influences beyond their boundaries, such as urban runoff.  Keystone species like
wolves and mountain lions, predators which formerly kept prey species like whitetailed deer in  check, have  been
eliminated, and deer now threaten to destroy some of our finest lands through overbrowsing. Many of our protected  sites
are too small to sustain populations of area-sensitive species, or to retain their full complement of species in the face of
random population  processes like immigration and emigration. Exotic plant and animal  species  pose major threats to
nearly all  of our native  communities. Landscape level processes, like fire, that shaped the fundamental character of our
ecosystems do not  occur with the frequency  or to the extent they once did, resulting in shifts in community composition
that  usually result in a  decline in biodiversity.

    Chicago area residents are the beneficiaries  of farsighted leaders early in the 20th century who established a tradition
of setting  aside natural  land in the urban matrix  for the public good,  a tradition that  our forest preserve districts continue
today.  The early model  was not  based on sophisticated concepts of biodiversity conservation, or of ecological processes,
but on the museum approach of setting nature aside and not meddling. We now are the beneficiaries  of the science of
ecology, which begins to tell us how the land mechanism is constructed and how  it operates.  It is dynamic, not static,
and  changes occur when landscape processes  are interrupted. The  science of ecology also reinforces the connections
between humans and the rest of nature.

    This allows us  to reexamine the old model of setting nature aside and leaving it alone. That  removes  the  most
immediate threat of development, but  it  does not  address the aforementioned degenerative loss  of biodiversity due to
fragmentation  and alteration of landscape processes. These processes clearly must be reintroduced into our preserves
if biodiversity is to  be preserved or restored. Prescribed fire must  be intelligently applied, invasive  species must be
controlled,  plant and animal  species must be reintroduced where they have been eliminated, hydrology must be restored
where altered,  and  science  must be improved where our understanding of ecosystem processes is deficient.  Perhaps
most fundamentally, the people who must  support the  greatly increased levels of land management and  research
necessary to restore and maintain our public  lands  in a healthy condition must have  a  basic understanding of land  health
and  the value system to commit public resources toward  attaining it. Chicago Wilderness, the coalition, is committed to
working on all of these  fronts.

    Quite  understandably, the Chicago region's  system of public lands was, and perhaps still  is, the core of Chicago
Wilderness, the initiative. It is what members  rallied around during the coalition's formation in 1996.  But  the vision quickly
expanded beyond public lands,  for two  reasons.

    First,  our public lands do not exist in isolation. They are  part of a much larger land base, and the protected 200,000
acres are affected  by what happens  on the remaining 6 million acres of the Chicago metropolitan land area. The
preserves form the  core, but they cannot preserve  all the  biological parts by themselves since much  biodiversity resides
on private  unprotected land and  because they are  subject to outside influences.

    Biodiversity considerations need to infuse all of the region's  land  use decisions  much more extensively that they do
now. Private lands work  either in harmony or discordantly with our network of preserves. The link  between the two is most
apparent in the case of wetland or aquatic habitats,  which  in  many cases are sustained or impacted  by runoff from distant
areas. Streams, rivers,  lakes, and wetlands  defy the "protect by  fencing" approach.  Overall watershed characteristics
determine aquatic and wetland habitat  quality quite independently of  whether the habitat  is in a formal preserve or not.

    Second, many high-quality, biologically rich  pieces of nature  persist outside of our preserve  system  and are
threatened by development, along with other stresses like lack of management. Identifying these biologically  important
areas within proposed developments,  redflagging  them, designing  development with their sustainability  in mind, and doing
all this with equity for the landowner,  is one  of our greatest challenges. Nature in the  places where we live contributes
so much to quality  of life, yet maintaining  it through the  development process resists standard  regulatory  approaches.
There are questions now asked routinely  in  the subdivision design process:  does the plan conform to drainage code, are
storm water basins sized properly, is it consistent with surrounding  development, does it have proper standards of
landscaping? A standard question should be:  does it leave the  land biologically richer or poorer? We are not yet routinely
asking this question, although there  are  development approaches available that can  allow us to  answer this  question
affirmatively.


                                                       226

-------
    This question could properly  be asked for aesthetic  reasons alone, but there are  practical reasons for doing  so.
Native landscapes hold enormous  potential for managing storm water and preventing flooding. They  hold  enormous
potential for cleaning up  surface  waters so that urban waters  become fishable and  swimmable,  instead of  the
neighborhood joke or eyesore. Finding and applying the template for development that preserves and restores biological
diversity, and which  serves both aesthetic and utilitarian  purposes, is one of  the  objectives of Chicago Wilderness.

Chicago Wilderness  - Structure and Function

    Chicago Wilderness formally began  as an initiative with the signing of a memorandum of understanding  (MOU) by
34 founding members. Members included landowners and land managers; local, state, and  federal  agencies; centers for
research and education;  and conservation  organizations,  among others. These institutions pooled their resources  and
strengths to form the Chicago Region Biodiversity  Council, which has grown  to include nearly 100 members.

    By  signing  the  MOU, the members  of this innovative partnership have pledged a commitment to the protection,
restoration, and management of biodiversity in the Chicago region. Four teams focus on central lines of action: science,
land management, policy and strategy, and education and communication. The teams attract the participation of many
non-member institutions, which adds to the scope and strength of the coalition. Chairs of the teams and other  member
organization staff form the  nucleus  of a coordinating group that  develops central  strategies and maintains momentum.
A steering committee of executives oversees the direction of the overall initiatives.  Despite  this organizational structure,
Chicago Wilderness  has  not  become  legally  incorporated under  state law,  but remains a  loose partnership bound  by
common goals  and  objectives.

    The  potential for Chicago Wilderness to  serve  as a  model for  urban  conservation attracted the  early attention of
several  federal  agencies, including  the  U.S.  Forest Service,  U.S.  Fish and Wildlife Service,  and U.S.  Environmental
Protection Agency, who have provided significant operating grants. State and private grants have supplemented federal
dollars.  Direct grants have totaled  over 4 million dollars since 1996. This total does not include members' matching funds
or funds attracted by members for  projects catalyzed, but not directly supported by, Chicago Wilderness.

Chicago Wilderness  Accomplishments

    The Chicago  Region  Biodiversity Council  funds  projects on an annual cycle. The Council's four teams  set priorities
for these projects; core staff  ensure broad participation from  team members.  Reviewed and approved  by  a proposals
committee, funded projects  result from  collaboration between member  institutions and address critical conservation  needs
in the region. Since its launch in April  1996, Chicago Wilderness  has  funded over 130 collaborative projects. In addition
to projects funded directly by the Council, the work  of our individual member organizations in their own initiatives is central
to  the  success  of Chicago Wilderness.  Projects completed or underway fall  into six categories:  characterization and
information management; ecological inventory and  monitoring;  ecological  restoration; planning  and policy; education,
outreach,  and public participation; and  communications  and  publications.

    Individual projects have included  a  NASA-supported  land cover mapping project; development  of  models of pre-
settlement savannas,  woodlands, and forests to guide restoration;  assessment of restoration  effects on bird communities;
a vegetation monitoring workshop; assessment of garlic mustard impacts on native woodland ground flora; development
of  model restoration  interpretive programs; a biodiversity educators workshop;  and creation of a Chicago Wilderness Atlas
of Biodiversity. An  early pilot project supported by Chicago Wilderness was the  launching of Chicago  Wilderness
magazine, a glossy, popular publication on nature in the Chicago area  which since has been  incorporated as  a 501  (c)(3)
and has over 7,000  paid subscribers.

Chicago Wilderness  Biodiversity  Recovery  Plan

    In 1909, the Commercial Club of Chicago  released the "Burnham  Plan," a landmark of urban planning that proposed,
among other things, a network of public parklands to be set  aside for nature and passive recreation. This led to the
legislative establishment of a system  of such publicly owned preserves for the  Chicago  region which  has continually
expanded, and  now  forms  the core of the protected lands that currently comprise Chicago Wilderness.

                                                      227

-------
    The Biodiversity  Recovery Plan, completed in late 1999, takes the open space component of the Burnham Plan to
the next step by creating  a  vision  of sustainability, not only for the core of protected land,  but for all of nature and its
human inhabitants in the  urban area. The  recovery plan is a  comprehensive statement of  what  Chicago Wilderness is
about, and it is clearly the most ambitious  and significant accomplishment of the  coalition to  date.

    This plan is the result  of three years of assessment and planning by representatives of the Chicago Region
Biodiversity Council. The plan identifies the ecological communities of the greater Chicago region, assesses their
condition,  identifies major factors affecting them, and provides recommendations for actions needed to  restore and protect
them into  the future  in a  sustainable condition.   In short, the recovery plan  outlines the steps necessary to achieve the
overall goal of the Chicago Wilderness collaboration, which is to protect the natural communities of the Chicago Region
and to restore them to long-term  viability, in order to enrich the quality of life of its citizens and to contribute to the
preservation of global  biodiversity.

    To achieve this goal,  the recovery  plan identifies the following  objectives: 1) involve the citizens, organizations, and
agencies of the region in  efforts to conserve biodiversity; 2) improve the scientific basis of ecological management; 3)
protect globally and  regionally important natural communities;  4)  restore natural  communities to ecological  health; 5)
manage natural  communities  to sustain  native  biodiversity;  6) develop citizen  awareness  and understanding of local
biodiversity to  ensure support and participation;  7) foster a sustainable  relationship  between  society and nature  in the
region; and 8) enrich the  quality of the lives of the region's citizens.

    The plan has many recommendations, some specific and some general,  and identifies roles and  specific actions for
Chicago Wilderness members and the greater  public that must  be  engaged to help implement the plan. The plan's
intended intended audiences include the  many staff members and  general members of Chicago Wilderness institutions,
publicagencydecision-makers, large landowners, and all concerned  and active citizens who vote and otherwise influence
biodiversity conservation  in  the region.

    The recovery plan is both a plan and  a  process guided by its many sponsors. It is intended as  a living document that
will continue to evolve as  new ideas and  information arise. It is intended to complement the many other planning efforts
completed  or underway in the Chicago metropolitan area that  are guiding the region to a  better and  more productive
future. Its   ultimate success  probably rests  on its successful  integration into a  broader,  mainstream regional planning
framework  that has  economic, cultural, social,  and environmental components.

Strategic Visioning

    After the second year of operation,  the Biodiversity Council saw the opportunity to step back and evaluate the
structure  and function of the coalition during  the first two years,  consider expectations of  members at the outset and
evaluate to what extent they were met, and reprioritize its work for the next  two years.  This  process consisted of
development of a member questionnaire, convening of a focus group representing a cross-section of members, and a
weekend  retreat  by Chicago Wilderness  Steering Committee members  and other leaders. It culminated with the
development of six  priority functions for the next two years, and  associated budget requirements.

    Some  of these functions represent an intensification and refinement of activities the Council is already involved  in;
in  other cases, they represent new  endeavors. They include 1)facilitate networking among Chicago Wilderness members,
including  new  orientation materials, workshops, symposia,  and lectures; 2) establish  an integrated information
clearinghouse,  including  the  development of regionwide resource databases,  enhancement of the  existing web site, and
development of more communication resources;  3)  increase publicity and outreach to  broader audiences; 4) influence
key  actors outside  Chicago  Wilderness,  including the establishment of a  Conservation  Policy Committee to develop
position statements  on regional issues; 5)  develop and  implement a funding strategy, focusing  on large grants  from
foundations; and 6) implement, promote,  and monitor the Recovery  Plan.
                                                      228

-------
The  Urban Conservation Culture

    Conservation efforts in  urban  areas are often frustrated by the complexity  of  land use  issues, countless  players,
tangled politics, ecologically  wrecked land, and a public dispossessed of nature. Yet it is crucial that we focus on urban
areas because of the  strong  political forces concentrated in urban centers that need  to  be engaged in national
conservation decision-making, and because  there is no other way to engage the great majority of people  other than to
take the messages to them. Moreover, urban residents are still plain members of Leopold's land community, regardless
of how obscure the connections, and these connections are best illustrated in the  places  where they live. Fortunately,
the Chicago region has an added  bonus  of harboring world-class  biodiversity, which creates  a local, immediately
compelling reason  for public involvement and action.

    Some writers have  argued  that the American ideal of wilderness has tended to shape  our dominant view of nature
itself as a place that can only be corrupted by human influence.  In urban areas, this has created an assumption that "real"
nature cannot exist in these places and it tends to absolve urban residents from  local responsibility. Thus,  it seems that
Chicagoans are much more aware of the plight of Brazilian rain forests that they are of the plight of oak savannas, a
globally rare community, in local  forest preserves. In remote areas, the standard approach has been to specifically
designate  areas as  wilderness, and then maintain as complete a separation between  people  and these areas as possible.
Chicago Wilderness proposes to redefine wilderness to include local plant and animal communities, which can  only  be
sustained through direct, creative human intervention. A  premise  of the recovery plan  is that if we do not adequately enlist
people to  directly or indirectly support management and restoration of our lands, they will not become or remain  healthy.


    It is appropriate to recognize that humans in the Midwest always have influenced landscapes, for better or worse, and
that people can be  a positive force in maintaining ecosystem health.  It may be that by calling a 200-acre patch of prairie
in a sea of development wilderness,  and by  involving people in  its stewardship, we can promote a correct sense of unity
between the places that we  live and  remote  places we  may never see except as pictures on calendars. Restoration and
stewardship can be  the antidote to dualistic  thinking. Remote  wilderness and Chicago Wilderness  can perhaps  then  be
seen  as simply examples of nature,  as part of a single system that includes people.

    From  a  relatively straightforward  beginning that focused on  public land management issues, this,  I think, has  become
the broader goal of Chicago Wilderness-to reconnect people with nature and to make a societal commitment to sustain
and nurture nature-for utility, for aesthetics,  for spirituality, for  all of the equally valid  reasons  for doing  it,  on all of our
urban lands and in  all of our land-use decisions.  It  begins with  a process of educating the public about  the natural wealth
in the  Chicago area, and  hopefully ends sometime  in the future with the development of an  urban conservation culture
of concern and personal responsibility for the health of all of our lands, both public  and private.
                                                      229

-------
                       A Survey of Resident Nutrient  Behavior  in the

                                   Chesapeake Bay  Watershed


                                                Chris Swann
                                       Center for  Watershed Protection
                                            Ellicott City, Maryland


    In  recent years a handful of communities  have attempted to craft education programs to influence  our watershed
behaviors. These initial efforts have gone by a  confusing assortment of names, such as public outreach, source control,
watershed awareness, pollution prevention, citizen  involvement, and stewardship, but they all  have  a common theme -
educating  residents  on how to live within their watershed.

    Many communities will need to develop watershed education programs in the coming years to  comply with pending
EPA municipal stormwater NPDES regulations.  Indeed,  half of the six minimum management measures prescribed under
these regulations directly deal with watershed  education  - pollution  prevention, public outreach  and public involvement.
Yet,  many communities have no idea what  kind of message  to send, or what media to use.

    In  the following  presentation, we review the prospects for changing our behaviors to better  protect watersheds. We
begin by outlining some of the daunting challenges that face educators who seek to influence deeply rooted public
attitudes. Next, we  profile research on the outreach techniques that appear most effective in influencing watershed
behavior. Special emphasis is placed on media campaignsand intensive training programs. Lastly, recommendations are
made to enhance the effectiveness  of watershed  education  programs.

Challenges in Watershed Education

    Watershed managers  face several daunting challenges when they attempt to influence watershed behaviors. Some
of those challenges include:

A lot  of minds  to change

    The most pressing challenge is that there are simply a lot of minds to change. Some notion of the selling job at hand
can be grasped from Table 1, which contains  provisional, but conservative, estimates of potential residential  "polluters"
in the United States by various categories. It is clear that we are attempting to change deeply rooted attitudes held by
millions of people. While most people profess  to support the environment,  only a  fraction  actually practice much  of a
watershed ethic on the small parcels of the  environment where  they live.
Table 1. Provisional Estimates of Potential Residential Polluters in the United States
Watershed Behavior Prevalence
Population
in Overall Estimates of Potential Residential Polluters
Over-Fertilizers 35% 38 million
Bad Dog Walkers 1 5 % 16 million
Chronic Car washers 25% 27 million
Septic Slackers 15% 16 million
Bad Mechanics 1 to 5% 3 million
Pesticide Sprayers 40% 43 million
Hosers 15% 16 million
 Notes: estimates are based on 1999 U.S. population of 270 million, 2.5 persons per household, and average behavior prevalence rates based on
 numerous market surveys (See references).
                                                    230

-------
Most Residents  are Only Dimly Aware of the Watershed Concept.
    It stands to reason that if citizens are asked to practice a watershed ethic, they will need to know what a watershed
 is. Surveys indicate, however,  that the average citizen is unaware of the watershed concept in general, and does not fully
 understand the hydrologic connection between their yard, the  street, the storm sewer, and (finally) local streams. Resident
 surveys also continue to show limited or incomplete understanding of terms such as "watershed", "stormwater quality"
 or "runoff pollution". For example,  a  recent Roper survey found that only41  %  of Americans had any  idea of what the term
 watershed meant (NEETF, 1999). The same survey found that just 22% of Americans know that stormwater runoff is the
 most  common source of pollution of streams, rivers, and oceans.

    At the same time, most of us  claim to be very environmentally aware. For example, a Chesapeake Bay survey
 reported that 69% of respondents professed to be  very active or at least somewhat active in helping to reduce pollution
 in the environment (SRC, 1994).

 Resources Devoted to  Watershed Education are Inadequate.

    In recent years,  several  communities have developed  education programs to influence the  watershed  behaviors
 practiced by their residents. Most of these  efforts, however, are run on  a shoestring.  For example,  CWP recently surveyed
 50 local programs that have tried to influence lawn care, septic cleaning and  pet waste behaviors (Swann, 1999). These
 education  programs are typically  run  by  the cooperative extension services, local recycling or stormwater agencies,  or
 urban soil and water  conservation districts. Most are poorly staffed (0.1 to 0.5 staff years), relatively new (within last five
 years), and  have  tiny annual budgets ($2,000 to  $25,000). Given  these limited  resources,  most watershed  education
 programs have no choice but to  practice retail, rather than wholesale, outreach techniques.  Consequently, most
 watershed educators rely heavily on  low-cost techniques  such  as brochures, posters,  workshops,  and  demonstration
 projects to  send their message out.

 The  Marketing Techniques  We Can Afford Don't Reach Many People

    Watershed managers need to  send  a clear and simple educational  message that can  attract the attention of the
 average citizen who is simultaneously bombarded  by dozens of competing messages every  day. A number of surveys
 have asked  residents which  outreach techniques are most  influential  in attracting their attention (Table 2).  Messages sent
through television,  radio and  local  newspapers are consistently more influential  in reaching residents than  any other
technique,  with up to  30% recall rates  by  the watershed population for each technique. By contrast,  messages transmitted
through meetings,  brochures, local cable and videos tend to be recalled by only a very small segment of  the watershed
 population.
Table 2. Most Influential Methods of Getting Messages to Citizens, in 8 Citizen Surveys


TV
TV ad
Newspaper
Local paper
Video
Brochure
Local cable
Meeting
WA
(Elgin, 1996)
TV ad
TV
Newspaper
Radio Ad
Brochure
Radio news
Paper Ad
Billboard
OR
(AMR.1997)
Direct Mail
TV ad
Newspaper
Radio
TV
Bill Insert
Newsletter
Local paper
CA
(As-sing, 1994)
TV Ad
Stencils
Billboard
Local paper
Brochure
Radio Ad
Bus Sign
Direct Mail
CA
(Pellegrin,
1998)
TV
Paper
Radio
Magazine
Neighbors
School
Billboard
Brochure
Ml
(PSC, 1994)
TV
Paper
Cable TV
Local paper
News-letter
Video
Meetings
Brochure
Wl
(Simpson,
1994)
TV
Paper
Newsletter
Brochure
Site Visit
Video
Meeting
MN
(Morris et al.,
1996)
Newspaper
Direct Mail
TV
Neighbors
Ext Service
Radio
Meeting
Local cable
    One clear implication is that watershed education efforts must utilize a mix of outreach techniques if they are going
to get the  message across to enough residents to make a difference in a watershed. Most existing watershed education
programs,  however,  cannot afford to use the  more sophisticated wholesale outreach techniques that are most effective
                                                     231

-------
I

1
+*
o
£
o
o
i
                                                                                     • Education Programs
                                                                                     D Residents
                                           Outreach Method
Figure 1. Outreach Methods Preferred By Residents Compared to Those Used by Education Programs.

at reaching the public  with their watershed message. This gap is evident in Figure  1, which compares the outreach
methods  actually used  by local watershed education  programs with the outreach methods that residents prefer, based
on responses from the Chesapeake  Bay survey (Swann, 1999).

Crafting  Better Watershed  Education Programs

    The first step in  crafting better watershed education programs is to  compile some baseline information on local
awareness, behaviors and media preferences.  Some of the  key  questions watershed  managers should consider  are:

      Is the typical individual aware of water quality issues in the watershed they live  in?

    .  Is the individual or  household behavior directly linked to water quality problems ?

    *•  Is the behavior widely prevalent in  the watershed  population ?

    >  Do specific alternative(s)  to the behavior exist that  might  reduce  pollution?

    >  What is the most clear and  direct message about these alternatives?

    >  What outreach methods are most effective in getting the message  out ?

    »•  How  much  individual behavior change can  be expected  from these outreach  techniques?

    The best way to  elicit this information  is to conduct a market survey within the watershed. These market surveys are
useful for two purposes: to gauge  the level of watershed awareness and  interest within the general population,  and to
determine if there is a segment of the population where education  efforts should  be focused to achieve the best  returns
in behavioral changes for the  money spent.
                                                     232

-------
    Perhaps the most critical step in crafting an education program is to select the right outreach techniques to send the
watershed message. Several communities have recently undertaken  before and after surveys to measure how well the
public responds to their watershed  education programs. From this research, two outreach techniques have shown some
promise in actually  changing behavior -  media campaigns  and intensive training. Media campaigns typically use a mix
of radio, TV, direct mail, and signs  to broadcast a general watershed message to a large audience. Intensive training use
workshops, consultation and guidebooks to send a much more complex message about watershed behavior to a smaller
and more interested audience.  Intensive  training often  requires  a time commitment of several  hours from residents.

    Both media campaigns and  intensive training can produce up to a20% improvement in selected watershed behaviors
among their respective target  populations (Tables 3 and 4).  Both  outreach techniques are probably needed in  most
watersheds, as each  complements the other.  For example, media campaigns cost just a few  cents per watershed resident
reached, while intensive training  can cost  a few dollars for each resident that is  actually influenced. Media campaigns are
generally  better at  increasing watershed awareness,  and sending messages about negative watershed  behaviors.
Intensive training, on the other  hand, is superior at changing  individual practices in the home,  lawn and garden.
Table 3. Effectiveness of Media Campaigns in Influencing Watershed Behaviors: Four Surveys
Location and Nature of Targeted
Campaign
San Francisco Radio, TV and Buses
(BHI, 1997)
ILos Angeles Radio and Newspapers
(Pellegrin Research Group, 1998)
Oregon Radio, TV (Advanced
(Marketing Research, 1997)
(Oakland County, Ml
IDirect Mail (Public Sector Consultants,
1994)
Effectiveness of Campaign
Awareness increased 1 O-l 5%
Homeowners who reduced lawn chemicals shifted from 2 to 5%
Best recall: motor oil and litter (over 40%)
Worst recall: fertilizer and dog droppings (<10 %)
Drop in car washing, oil changing, radiator draining of about 5 to 7%
Greater self-reporting of polluting behaviors: dropping cigarette butts, littering, watering and letting water run
on street, hosinq off driveways into the street (1 0% or more)
1 9% reported a change in "behaviors"-changes included being more careful about what goes down drain,
increasing recycling and composting, using more nature-friendly products etc.
44% of mail respondents recalled lawn care campaign
50% desired more information on lawn care and water quality
1 0% change in some lawn care practices as a result of campaign
(grass recycling, fertilizer use, hand weeding). No change in other
lawn care practices as a result of campaign
Tat
Location and Nature of Training
Campaign
Maryland Direct Homeowner
(Smith, 1996)
Florida Master Gardener
(Knox, 1995)
Virginia Master Gardener
(Aveni, 1998)
le 4. Effectiveness of Intensive Training in Changing Watershed Behaviors
Effectiveness of Intensive Training
10% shift from self to commercial car washing.
No change in fertilizer timing or rates.
Better claims of product disposal.
No significant change in fertilization frequency after program.
Some changes in lower rates, labels, slow release (8 to 1 5%).
Maior changes in reduced pesticide use (1 0 to 40%).
30 to 50% increase in soil testing, fertilizer timing and aeration.
10% increase in grass clippings and 10% decrease in fertilizer rate.
                                                     233

-------
    Both techniques work  best when  they present a  simple and direct watershed message, are repeated frequently,  utilize
multiple media  and are directly connected to local water resources that are most important  in the community.

    Other important  suggested considerations for effectively  marketing  a  watershed message are to:

    Develop stronger connections among the yard,  the street, the storm sewer, and the stream. Outreach techniques
should continually stress the  link between a particular watershed behavior and the undesirable water quality it helps to
create (i.e.,  fish kills, beach  closure, algae  blooms).  Several excellent  visual  ads that effectively  portray this link are
profiled in our  watershed  outreach award winners.

    Form  regional media campaigns.  Since most  communities operate on  small  budgets, they should consider pooling
their resources  to develop regional media campaigns that can use the outreach techniques that are proven to reach and
influence residents.  In  particular, regional campaigns allow communities  to hire the professionals needed to create and
deliver a strong message through the media. Also, the campaign approach allows a community to employ a combination
of media, such  as radio, television, and print, to reach a wider segment of the population.  It is important to keep in mind
that since no single  outreach technique will  be recalled by more than 30% of the  population at large, several different
outreach techniques  will be needed  in an effective media campaign.

    Use television wise/y. Television is the most  influential medium for influencing the public, but  careful choices  need
to be made on the form of television that is used. Our surveys found that community cable  access channels  are  much
less  effective than commercial or  public television  channels. Program managers should  consider using cable network
channels targeted forspecificaudiences, and develop thematicshows that  capture  interest of the home, garden  and lawn
crowd  (i.e., shows along the lines of "This Old Watershed"). Well-produced  public service announcements  on commercial
television are also a  sensible investment.

    Understand the demographics of your watershed. The middle-aged  male should usually  be  the prime target for
watershed education, as he is prone to engage  in more potentially polluting watershed behaviors  than other sectors of
the population.  Indeed, the most important audience for the watershed message includes men in the 35 to 55 year age
group with higher incomes and education  levels. Specialized outreach  techniques can appeal to  this  group, such as radio
ads on weekend sports events.

    Another target group worth reaching  includes what Pellegrin (1998)  terms the "rubbish rebels"- 18 to 25 year olds
who tend to have low  watershed  awareness, engage  in potentially polluting behaviors,  and are often employed in lawn
care  and other  service  industries. This age group is hard to reach using conventional techniques, but may  respond to ads
on alternative radio,  concerts, and other events that  celebrate  the watershed.

    As communities become  more diverse, watershed managers should  carefully track the unique demographics of their
watersheds.  For example, if many residents speak English as a second language, outreach  materials should be  produced
in other languages. Similarly, watershed managers should consider more direct channels to  send  watershed messages
to reach particular groups,  such as  church  leaders, African American newspapers, and  Spanish-speaking  television
channels.

    Watershed  educators  should also be careful about using the  traditional environmental education model  that uses
schools to educate children who in turn educate  their parents.  While this  model was instrumental in achieving greater
rates  of recycling, it may not be as effective in changing watershed behaviors.  While  it  is  important to educate  the next
generation of fertilizers, dog walkers, septic cleaners,  and  car washers, we  need to directly influence the  "boomer'
generation  now.

    Keep  the watershedmessagesimpleandfunny. Watershed education should not be  preachy complex,  or depressing.
Indeed, the  most effective  outreach techniques combine a simple  and direct message  with a dash of humor.

    Make information packets small, slick,  and durable.  Watershed educators should  avoid the ponderous and boring
watershed handbook that looks great to a bureaucrat but ends  up lining the bottom  of a bird  cage. One solution is to
create small, colorful and  durable packets that contain the  key essentials about watershed behaviors, with contact

                                                      234

-------
information to get better advice. These packets can be stuck on the refrigerator, the kitchen drawer or the workbench for
handy  reference when the  impulse for better watershed behavior strikes.

    Educate private sector allies. A wide number  of private sector companies may potentially stand to benefit from
changes in watershed behavior. Better watershed behavior can drum up more sales for some companies,  such as septic
tank cleaners,  commercial car washes, and quick oil change franchises - although they may need  some help in crafting
their watershed  marketing  pitch.

    Clearly, the potential exists for lawn care  companies and landscaping services to shift their customers toward more
watershed-friendly practices. Nationally, lawn care companies are used by up to 50%  of consumers,  depending on
household  income and  lot  size. Lawn  care companies  can  exercise considerable  authority over which practices are
applied to the lawns they tend, as long as they still produce a sharp looking  lawn. For example,  94% of lawn care
companies reported that they had authority to  change  practices, and that about 60% of their  customers were "somewhat
receptive to new ideas" according to a Florida study (Israel et al,  1995). De Young (1997) also  found  that suburban
Michigan residents expressed a high  level  of  trust in  their lawn care company.

    Indeed, a  small, but growing proportion of lawn  care  companies feel  that environmental advertising makes good
business sense and can  increase sales (Israel et al, 1995). Clearly, intensive training and certification will be needed to
ensure that watershed-friendly  ads reflect good  practice and not just slick salesmanship.  It  needs  to be  acknowledged
that lawn care companies strongly committed  to practices that reduce  fertilizer and pesticide  inputs need  to  be strongly
endorsed by local government. Right now, it is  not likely that such companies would be selected by the average
consumer,  as  consumers primarily rely on  direct mail, word of mouth,  and cost when choosing a lawn  care  company
(Swann, 1999  and AMR,  1997). For example, in  the Chesapeake Bay survey, only two percent of residents indicated that
they had chosen  a lawn care  company primarily on the  basis that  it was "environmentally friendly" (Swann, 1999).

    Lawn  and  garden  centers are another natural target for  watershed education. Study after study  indicates  that product
labels  and store  attendants are the  primary and almost exclusive source of lawn care information for the average
consumer. At first glance, national retail chains should be strongly opposed  to better watershed behavior,  since it would
sharply cut into lawn and garden product sales and the lucrative profits they produce  (even  at the expense  of the
community and environmentally friendly image they often market). The key strategy  is to substitute watershed-friendly
products for ones that are  not, and to offer training for the store attendants at the point of sale  on how to use such
products.

Summary

    For the watershed manager faced with new regulatory requirements under Phase II of the NPDES program, the
creation of an  effective watershed education program should be a high priority. Not only is public education  a mandated
component of  an  NPDES permit,  but in urbanized  areas it may the most cost-effective tool  available to  achieve water
quality goals.  For smaller communities with  scant budget and staff resources, it is imperative that these education
programs be productive in terms of changing behaviors  and  raising awareness of individual  actions on  local water quality.

    Perhaps the most  important factor in creating an effective watershed  education program is selecting the right outreach
methods.  Market  surveys will  often answer questions regarding  the  level of environmental awareness of watershed
residents, what forms of informational outreach attract their attention,  and resident willingness to change pollutant
producing  behaviors. This information  allows the watershed  manager to tailor outreach methods to specific target groups
where behavior change is most  likely.  These surveys will also  establish the demographics of the residents  and determine
whether multilingual outreach  is required.

    Watershed managers should also consider innovative approaches to sending out their pollution  reduction messages.
Pooling resources with other communities  to  create regional media campaigns and the  use of outreach opportunities
through  private sector education  are just two ways that program managers  can reach broader audiences without  spending
large amounts of money.
                                                     235

-------
    Continued development of productive outreach methods  and innovative techniques  is necessary to relay the basic
premise of watershed education - that we live in a watershed and how to properly live within it - in the most economical
and effective  manner.

References

Advanced Marketing Research (AMR). 1997. Stormwatertracking study. City of Eugene,  Oregon. Unpublished marketing
survey.

Assing, J. 1994. Survey of public attitudes. February and  July,  1994. Russian Hill Associates. Alameda County Urban
Runoff Clean  Water Program. San Francisco CA. 84 pp.

Aveni,  M. 1998. Water-wise  gardener  program:  summary report.  Unpublished  Data. Virginia  Cooperative Extension.
Prince  William County,  VA.

Big Honking Ideas,  Inc (BHI).  1997. Final report: Spring regional advertising campaign. Prepared for Bay Area Stormwater
Management  Agencies  Association. Oakland,  CA.

De Young, R.  1997. Healthy lawn and garden survey:  data  analysis report. Rouge River  National Wet Weather
Demonstration  Project.  Oakland County, Ml. 40 pp.

Elgin DDB.  1996. Public awareness study: summary report.  The Water Quality Consortium. Seattle,  WA. 24 pp.

Israel,  G., S. Pinheiro  and G. Knox. 1995. Environmental landscape management  - assessing practices  among
commercial  groups. University of Florida. Cooperative Extension Service. Bulletin 307.  Monticello, FL. 18 pp.

Knox, G., A. Fugate and G. Israel. 1995. Environmental  landscape management-use of practices by  Floridaconsumers.
University of Florida Cooperative Extension Service. Bulletin  307. Monticello,  FL. 26 pp.

Morris, W. and D. Traxler. 1996. Dakota County subwatersheds: residential survey on  lawn care and water quality.  Dakota
County,  Minnesota, Decision  Resources, Ltd.

National Environmental Education Training Foundation  (NEETF).  1999. National Report Card  on Environmental
Knowledge, Attitudes and Behaviors: Seventh Annual Roper Survey  of Adult Americans. National Environmental
Education  Training Foundation, Washington, D.C.

National Service Research (NSR).  1998.  Pesticide usage and impact awareness study: executive summary.  City of Forth
Worth Water Department. Fort Worth Texas. 44 pp.

Pellegrin Research Group. 1998. Stormwater/urban runoff public education program: interim evaluation, resident
population. Los  Angeles County Department of Public Works. 28 pp.

Public Sector Consultants, Inc (PSC).  1994. A strategy for public involvement. Rouge River  National Wet Weather
Demonstration  Project.  56 pp.

Simpson, J. 1994. Milwaukee survey used to design  pollution prevention  program.  Technical Note 37. Watershed
Protection Techniques.  1(3): 133-1 34.

Smith, J. 1996. Public survey used to estimate pollutant  loads in Maryland. Technical Note 73. Watershed Protection
Techniques. 2(2): 361-363.

Survey  Research  Center  (SRC).  1994. The  Chesapeake Bay attitudes survey. Communications subcommittee.
Chesapeake Bay Program. U.S. Environmental Protection Agency. 42  pp.
                                                    236

-------
Swann, C. 1999. A survey of residential nutrient behaviors in the Chesapeake Bay. Widener-Burrows,  Inc.  Chesapeake
Research  Consortium. Center for Watershed Protection. Ellicott City, MD. 112 pp.
                                                     237

-------
                    Lawn Care and Water Quality:  Finding  the Balance
                                               Jerry  Spetzman
                                     Minnesota  Department of Agriculture
                                           St. Paul, MN 55107-2094
                                   email: Jerome.Spetzman©State.MN.US
Abstract

    When land is converted from natural  areas to developed urban areas, pavement and rooftops replace grass and trees.
Water flows over driveways, streets and parking lots taking with  it particles and debris in its path  and depositing them,
via storm sewers, into  nearby  lakes,  creeks and  rivers. This non-point source pollution can contain sediment, debris,
fertilizers, pesticides, leaves, grass clippings, motor oil, or pet wastes. Small amounts of these materials entering a lake
or river are not generally considered harmful. But when small amounts are multiplied by thousands or tens of thousands
they can cause serious water quality  problems.

    Since 1993,  the  Minnesota Department of Agriculture,  in cooperation with  several other organizations, has been
gathering information on homeowner  use of pesticides and fertilizers in  the  Twin Cities metropolitan  area. This
information includes the amounts of lawn care  products used by  homeowners, where they are purchased, how they are
applied,  and whether or not they have an effect on nearby lakes, creeks or rivers.

    Residents  living in  two watersheds  were selected to participate in  a focused study. Lake Harriet represents urban
watersheds and Lake Alimagnet represents suburban watersheds. Based on survey  results and water quality monitoring,
education materials were developed to  promote public awareness of lawn practices and their potential to  affect water
quality. These  educational materials incorporate the concept that "everyone lives in a watershed" and that everyone has
the potential to affect water quality, whether or not they actually  live on a lake shore.

What are the Lake Harriet  and  Lake  Alimagnet  Watershed Awareness Projects?

    The projects  have two purposes: (1)  to inform  urban and suburban homeowners about living in a watershed, and (2)
to help them learn how  their lawn care habits can  affect the quality of Twin Cities water. The project's goal is to improve
water quality by  reducing  the quantity of pesticides and nutrients through responsible use  of those  materials.

How has this goal been  achieved?

    Project members have:

    .  Surveyed the  current  lawn  care habits of homeowners and measured  the effects of  those habits by  monitoring
      pesticide and  nutrient runoff into  Lake Harriet

    .  Informed homeowners about how their lawn care habits affect Twin Cities waters

    .  Asked homeowners how the projects have  affected their  lawn care practices

    .  Monitored  runoff  into Lake  Harriet to quantify  changes  brought about  by homeowner actions

    .  Drawn on  detailed  Lake Harriet experience to design  urban watershed education  materials  for use throughout
      Minneapolis and  Minnesota-these materials were tested  in the  Lake Alimagnet watershed

                                                    238

-------
What is a watershed?

    It is an area of land over which rain and melted snow flows to lakes, rivers,  and wetlands. There are 46 major
watersheds and  255 sub-watersheds in the Minneapolis-St.  Paul  metropolitan area.

    The Lake  Harriet watershed  is a 1,139-acre area in a well-established residential neighborhood with almost 6,000
homeowners.  The Lake Harriet  study area is a 148-acre portion of the watershed. About 40% of the study  area is
covered with  hard  surfaces, such as pavement and rooftops. About 700  homeowners live in the study area,  most  in
detached,  single  family  houses built in the early  1900s. The water quality in Lake Harriet is very  good for a Twin Cities
lake.

    A survey of 105 Lake Harriet watershed  residents most familiar with lawn care done on their own properties, showed
that:

    .  They are  highly  educated  (college or post-graduate degrees  predominate)

    .  Their average age is 47

    .  They have middle- to upper-level incomes

    .  They care for their lawns by mowing  regularly and using fertilizers and herbicides

    .  A few use professional lawn care services

    The Lake  Alimagnet watershed  has approximately 3600  suburban households. Of this number, a large proportion
is made up of townhouse residents. The Lake Alimagnet watershed was developed in  the 1960s.  Homeowners are well-
educated and  have middle- to upper-middle incomes.  The water quality in Lake Alimagnet is considered  poor, algae
blooms are common, and the predominant  fish  species is  bullhead.

Why are we studying the  Lake Harriet and Lake Alimagnet watersheds?

    Because they provide  good examples  (respectively) of how urban and  suburban development affect our water
resources  and because  they are  sources of year-round recreation for many Twin Cities residents.  These residents are
vitally interested  in keeping the water clean.

What are we  finding out?

    Though Lake Harriet had some of the highest quality water in the Twin Cities, that quality has also declined over time.
Lake Harriet has poor quality water  and area residents want to reverse the trends.

    These projects and others have  monitored storm water, rainfall, and lake water  to determine  the levels of non-point
source pollutants in Lake Harriet.  Specifically, this project monitored two types  of  pollutants: pesticides, which can affect
water quality,  and phosphorus, which can increase lake  algae growth  and reduce  water quality.

Pesticide monitoring

    The Lake  Harriet project monitored  storm  runoff, rainfall, and lake water. The water quality monitoring consisted of
a permanent automatic  sampling station installed in a storm drain outlet which carried watershed runoff into Lake  Harriet;
samples were taken during storm events, from which mass loading of pesticides and nutrients were calculated. In addition,
a rainfall monitoring station  sampled rain events. Finally  lake monitoring  samples were collected during the growing
season. Several  hundred samples were analyzed for more than 30  pesticides.

    The education  program  included  several methods  of education,  including  homeowner  meetings, direct mail, flyers,
billboards,  utility bill inserts,  local  newspaper articles and visits by Master Gardeners.  Because monitoring preceded the


                                                     239

-------
education, and  also followed  ^reductions in pesticides can be measured.  There was a  decrease in average  pesticide
loads between the earlier and  later monitoring periods. Therefore, the annual storm sewer runoff load of pesticides to  Lake
Harriet was reduced during the Lake Harriet project.  The largest decreases  came from the four compounds listed on this
table:
Lawn Herbicides
MCPA
Dicamba
2,4-D
MCPP
Percent Decrease
(1992-1 995)
86%
59%
58%
56%
    The  most prevalent pesticides found during monitoring  were herbicides (weedkillers). The  eight herbicides  listed
on the following table accounted for  95% of all pesticide detections.
Lawn Herbicides
MCPA
Dicamba
2,4-D
MCPP
Agricultural Herbicides
Alachlor
Atrazine
Cyanazine
Metolachlor
    .  Storm water runoff monitoring  summary:
      .  Lawn  herbicides were found in 80% of the  storm  runoff events sampled between April and October.
      .  Agricultural  herbicides were detected in 35% of the storm events sampled
    .  Rainfall  monitoring summary:
      .  The  agricultural herbicides  listed  above (the  only  herbicides  found in  rainfall  samples)  were  atmospherically
        deposited by wind  and rainfall onto  the watershed  and the accompanying water bodies.
      •  Lawn  herbicides were not detected  in rainfall  samples.
    .  Lake  monitoring summary:
      .  The three most commonly detected compounds in  lake water were MCPP, Atrazine, and 2,4-D.  They were also
        the  most frequently detected  compounds  in stormwater  entering  Lake Harriet.
    .  Phosphorus monitoring:
    Analyses revealed the phosphorus in runoff peaks twice a year, in the spring and in the fall.
      .  In the spring,  melting snow  carries phosphorus attached  to tiny particles of grit, sand, and organic matter as it
        enters the storm sewers.
                                                      240

-------
      . In the fall, phosphorus in leaves, grass clippings, and other organic debris enters the storm sewers. Studies
        conducted  by other researchers established that  only a small  percentage of phosphorus in  runoff results from
        appropriate  use  of fertilizer on lawns.
    .  Water  Quality  Education
    Lake Harriet and Lake Alimagnet project participants have concluded that educating homeowners living in the
watershed is one of the best ways of reducing pollution  in the  lake. Billboards,  brochures, and water bill inserts have
carried messages based on  the following two concepts:
    .  A  healthy lawn and landscape promotes  healthy waters. Home landscaping  with regionally adapted, healthy
      plants can help absorb and filter rainfall, irrigation,  and runoff from  melted snow.
    .  Keep your lawn and  landscape healthy as follows:
      .  Apply pesticides  and nutrients according to recommendations
      .  Aim roof downspouts onto lawns and gardens to filter and absorb runoff
      .  Keep grass clippings and leaves off streets, sidewalks, and driveways
      .  Leave grass clippings on the lawn  or compost them
      .  Use fallen  leaves as winter or summer mulch, compost  them,  or shred them and leave them on the lawn
      .  Keep lawn care products on  the lawn and  always follow label instructions
      .  Clean  up and reuse granular lawn care  products that fall on streets, sidewalks, and driveways
Project Evaluation
    .  Based on feedback from homeowners living in the Lake Harriet and Lake Alimagnet watersheds, we have
      concluded  the following:
      .  Most homeowners in the Lake Harriet watershed apply significantly  less lawn fertilizer than the University  of
        Minnesota's  recommended  guidelines.
      .  Most homeowners compost grass clippings or leave them  on  their lawns.
      .  Homeowners would rather spot-treat weeds than  apply herbicide to their entire yard or use non-chemical weed
        control  methods.
      .  Top soil in the Lake Harriet watershed is significantly  deeper than that found in the Lake Alimagnet watershed.
        In many cases when suburban areas are  developed,  top soil is removed,  but not replaced. This results in
        decreased  plant vigor, and an increase in the need to fertilize, water,  and maintain turf.
      .  Most  homeowners feel the educational initiative has increased their understanding of how lawn care habits affect
        water quality. Neighborhood  newspapers and direct mail, the most common source  of lawn care and water
        quality information, have the  greatest impact.
      .  Messages that are quick to read  and easy to understand are the  most effective in changing lawn care  habits.
      .  These messages are best delivered over an extended period.  Homeowners have  been reached  through fliers,
        newspaper  articles, brochures, direct mail,  handouts, and  personal contacts.
                                                     241

-------
    Feedback also shows that homeowners still need  to hear the following messages:
    . You can preserve water quality and have a healthy lawn by applying lawn  care products in appropriate amounts,
     at the right times, and during suitable (or appropriate) weather conditions.
    . By keeping leaves out of storm sewers, you  can help reduce the amount of phosphorus carried  to the lake in runoff
     water.
    . Fall is the best season to apply turf fertilizers  and lawn care  products that control broadleaf weeds. The primary
     growth of turf grasses is early fall  until late spring.
    . Erosion, leaves, grass clippings,  yard waste, pet  waste, and rainfall all contain pollutants that  can end up in  lake
     water.
Lake  Alimagnet  Project  Results
    During the  project  period, the water quality in Lake Alimagnet  improved significantly.
    . Project  cooperators  achieved  improvements in total  phosphorus  and chlorophyll-a.
    . The best  secci disk  reading ever on Lake Alimagnet was  recorded.
    • The Citizens  Assisted Monitoring Program improved the ranking of Lake Alimagnet from a "D" to a "C."
    These results may  be  a result of the project, a curly leaf pond weed  cutting,  or may be credited to El Nino.
    For more information on the Lake Harriet  or  Lake Alimagnet Watershed Awareness Projects  and samples of
homeowner education materials,  please  contact the  Minnesota  Department of Agriculture  at  (651) 297-7269.
                                                     242

-------
                              San Francisco  Bay  Area's Pesticide
                                   Toxicity Reduction Strategy
                                               Geoff Brosseau
                       Bay Area Stormwater Management Agencies Association  (BASMAA)
                                             Oakland,  California
Introduction

    Water quality research conducted by San Francisco Bay Area stormwater programs and wastewatertreatment plants
over the last several years has identified widespread toxicity in  local creeks  and wastewater treatment plant effluent
(California Regional Water Quality Control Board, 1997; San Francisco Bay Area Pollution  Prevention Group, 1998). The
toxicity problem was ultimately traced to diazinon and  chlorpyrifos-commonly  used  organophosphate pesticides available
in hundreds of consumer products  (Alameda Countywide Clean Water Program, 1997). Study results indicated that
pesticide use according to label instructions could  not be ruled out as  a cause of wastewater and stormwater toxicity
(Regional Water Quality Control Plant-Palo Alto, 1996). In May 1999, San Francisco Bay and 35 Bay Area urban creeks
were  listed by the U.S. Environmental Protection Agency (USEPA) as impaired by diazinon (USEPA, 1999).

Impact of 303(d) Listing on Local Governments

    In its action, USEPA listed 53 waterbodies  in California as impaired due to diazinon in urban runoff and 7 waterbodies
as impaired due to chlorpyrifos in urban runoff. By definition under the Clean Water Act, this action means that there is
a water quality  problem, regardless of the problem definitions  under the Federal Insecticide, Fungicide and  Rodenticide
Act (FIFRA)  (i.e., "unreasonable  adverse effect") or the Food Quality Protection Act (FQPA). The listing action put over
100 municipalities in the San  Francisco Bay Area  and Central Valley at immediate regulatory, legal,  and financial risk.

    .  Regulatory risk -The State Water Resources Control Board and USEPA can  take enforcement action against, and
      fine, these municipalities for violating their NPDES  (National Pollutant Discharge  Elimination System) stormwater
      permits.

    .  Legal  risk - Citizen  and environmental  groups  can  sue municipalities  for the same reasons.

    .  Financial  risk - These municipalities must now spend  local  public tax dollars proactively addressing this problem,
      and potentially  reacting  to fines and  lawsuits.

Municipalities'  Response

    To comply with their  NPDES stormwater permits, municipalities must meet two broad goals:

    1. Effectively  prohibit  non-stormwater discharges  into storm  sewers.

    2. Reduce the discharge of pollutants to the  maximum extent  practicable (MEP).

    To meet these goals and to address the 303(d) listing, there are a  number of actions Bay Area stormwater programs
have  taken or plan to take that  may  reduce pesticide-related toxicity in  surface waters. These actions are packaged  in
a Pesticide Toxicity  Reduction Strategy (BASMAA, 2000). The Strategy is  a  multi-faceted effort including:

    .   Education/outreach  including:


                                                     243

-------
      . limiting or prohibiting pesticide use by municipal staff and contractors and/or requiring use of best management
        practices  (BMPs) such  as Integrated  Pest Management (IPM)

      . providing adequate and convenient options for disposal  of unused  pesticides and pesticide containers through
        household hazardous waste collection  programs

      .  educating residents about pesticide-related toxicity and proper  use and  disposal through  distribution  of
        educational materials, and development and implementation of media  and advertising  campaigns

      .  educating residents about alternative methods and products through such programs as demonstration gardens
        and point-of-purchase campaigns in hardware stores and nurseries

      .  educating businesses about proper use and disposal, as well as alternative methods and products  for use around
        their own  properties and  facilities

      .  educating pest control operators and working with them  to develop BMPs  protective of surface  waters

      .  Regulatory- Identifying opportunities  to reduce toxicity and advocating state  and federal agencies to  seize these
        opportunities through regulation and  re-registration

      .  Monitoring -  Investigating  the extent and causes of toxicity, and assessing impacts on beneficial uses

The  IPM Store Partnership

    One exemplary part of the Pesticide Toxicity Reduction  Strategy worth describing  in  more detail is the IPM (Integrated
Pest  Management) Store Partnership. In 1997, the Central Contra Costa Sanitary  District (CCCSD), a wastewater
treatment  plant located  in Martinez, California, jointly developed and successfully piloted the IPM Store Partnership with
the Regional Water Quality Control Plant (Palo Alto, California) in four locally  owned garden centers and hardware stores.
In 1998, the  Bay Area  Stormwater Management Agencies  Association (BASMAA) and  the  San Francisco Bay Area
Pollution Prevention Group (BAPPG) joined together to fund the expansion of the IPM Store Partnership to more stores
in the San Francisco Bay Area.  By spring 1999, 116 stores in  eight Bay Area counties  were participating in the
Partnership.

The  Partners

    BASMAA  is a consortium of seven San Francisco Bay Area municipal stormwater programs. These programs
represent more than 90 agencies, including  79 cities  and 6 counties, and the bulk of the watershed immediately
surrounding San Francisco  Bay. BASMAA  agencies agree to  a  memorandum  of understanding  and each  year collect
dues, prorated by population,  from their  members for a "baseline"  program that  provides for staff and  finances
projects-like the IPM Store  Partnership-that are endorsed by all member agencies.

    The BAPPG  is a voluntary association of 39 wastewater treatment plants working together to prevent  water pollution
in the San Francisco Bay.  These agencies represent all of the publicly owned municipal wastewater agencies that
discharge  into San Francisco Bay in the  nine Bay Area counties, and almost  all of the watershed immediately  surrounding
San Francisco Bay. BAPPG's decision-making is done by consensus. Each year,  a work plan-with  an  associated
budget-is developed.  The budget is allocated among the 39 plants based on the average amount of treated  wastewater
discharged each  day. Contributions are voluntary, although all agencies do  contribute. These contributions are used  to
fund projects like  the IPM Store Partnership.

Integrated Pest  Management

    There are  many  definitions of  Integrated Pest Management. The definition used to guide the  IPM Store  Partnership
was the following:
                                                     244

-------
    Integrated  Pest Management is an approach that uses regular monitoring and cultural, biological, and  physical
    methods to keep pests at acceptable  levels. Only  less toxic chemicals are used  and  only as needed.

    IPM was used  as the basis for the program because it: (1) focuses on effective alternatives to traditional chemical
pesticides; (2) does not substitute another pesticide  that may become tomorrow's problem, and (3) does not preclude the
use of chemicals in all situations.

    Although promotion of IPM was the basis for the program, the term itself is somewhat problematic. The terms "IPM"
and "Integrated Pest Management" were used in speaking with experts and agencies  familiar with the jargon, but these
terms were avoided in communications  with the non-initiated (e.g., general public). In addition, the term IPM was not very
representative of the situation in  the store between  the  customer and the store employee. Most customers go to a store
for help when pests have already reached unacceptable levels,  so store employees  must  start  with controlling a pest
problem,  rather than preventing it, which is the first step in IPM. Despite this challenge, customers were exposed to the
full range  of IPM methods through the fact sheets and display materials, as  well as less-toxic products.

Goals

The goals of the  IPM Store  Partnership are to:

    .  Educate the public about the value of IPM approaches to pest control and safe use and disposal of pesticides, when
      used

    .  Deliver IPM-related  messages without negative  messages about any  products

    .  Develop partnerships with  retailers so that they can help spread the word about water quality problems related to
      residential  pesticide use

    .  Provide consistent  messages

    • Capitalize on economies of scale

    .  Prepare the stage  for regional program expansion into chain stores

Program Elemen ts

    The IPM Store Partnership is an education program  for employees and customers of locally owned garden centers
and  hardware stores. The project elements  include:

    .  development  and production of eight fact sheets  on less-toxic pest management strategies for the public  (Naturally
      Managing Pests, Controlling Ants, Controlling Aphids in Your Garden, Keeping  Cockroaches Out of Your House,
      Keeping Fleas Off Your Pets and Out of Your Yard, Living with Spiders,  Tips for a Healthy Beautiful Lawn, and Safe
      Use and Disposal of Pesticides)

    .  development of an  extensive list of less-toxic methods and products  preferable to diazinon and chlorpyrifos for
      various applications

    .  training sessions for store employees focusing on principles of Integrated  Pest Management and  successful
      application  strategies for products on  the less-toxic list

    .  design and production of  a program  logo and in-store promotion  materials including  "end  cap" displays, posters,
      shelf-talkers,  shelf signs,  and vinyl  banners

    .  program  evaluation  by a  San  Francisco State University-affiliated survey research  and data analysis  firm
                                                      245

-------
Results  and Discussion

    Although the final evaluation will not be complete until the end of the 1999 in-store promotion season (late 1999), the
sponsoring agencies expect it to be successful  based on preliminary findings and  on the success of the  pilot project.
Feedback from store owners and employees that  participated in the pilot  was uniformly positive. The sales data  in the
pilot IPM Store Partnership showed variations from store to store.  One store found that sales of all but one diazinon and
chlorpyrifos product dropped. At the same time, less-toxic product  sales experienced an overall 17% increase and  profits
were not affected.

     It is the hope of the participating agencies that the final evaluation and report on the 1999 Partnership will be  useful
as a model and primer for other agencies and jurisdictions  concerned about pesticide-related toxicity in surface waters
and  interested  in  building  educational partnerships with local  businesses.  While Bay Area  water pollution  prevention
agencies have  been coordinating their public education efforts since the early 1990s, the  IPM  Store Partnership  is the
first  point-of-purchase program implemented regionally. All of the general  benefits of  inter-agency coordination  (support
for smaller agencies, cost savings,  options  for pooled advertising and media relations) are  magnified in such a large
undertaking.

     Based  on the  partnership's success,  all  of the agencies  that participated in the  1999 Partnership allocated funds for
continuation of the  program in  1999-2000. BASMAA and the  BAPPG again contributed funds  to regional coordination.
Brainstorming sessions were held in  late summer 1999 to determine how to  improve the program, and minor modifications
were made for 2000.

    The IPM Store Partnership is one example of BASMAA and the BAPPG's commitment to use public resources
efficiently. Given that philosophy, materials developed  by the IPM Store Partnership are available to agencies interested
in implementing a similar program.

Other Aspects  of the  Strategy

     Despite the success  of the IPM  Store  Partnership and  many of the other educational aspects of the  Pesticide Toxicity
Reduction Strategy, it is  clear to Bay Area water pollution prevention agencies that their efforts alone will not be enough
to solve the problem. Study results indicate that less than 1%  of applied diazinon runs off, yet it takes  less than a fluid
ounce of active ingredient flushed  into  stormwater runoff to cause toxicity  in  urban creeks (Regional Water Quality Control
Plant-Palo  Alto, 1996). Educational  programs run by Bay Area water pollution prevention agencies are some of the most
developed in  the country and they have  won numerous awards for their quality  and effectiveness.  Nevertheless,  even
the  best education programs are  not  100%  effective.  It is clear that education alone  will not solve this problem.

    San  Francisco  Bay Area stormwater programs are  and will  continue to  address the problem of pesticide-related
toxicity in surface waters by way of meeting the MEP  requirement in their  NPDES permits. These agencies have gone
so far as to develop the Pesticide Toxicity Reduction  Strategy described  above  that includes three elements-education,
regulatory,  and monitoring. The  authority  and ability of local governments to implement the strategy varies with each
element. The most  cost-effective and appropriate aspect for local governments to  implement is education. For the
regulatory and  monitoring elements, local  governments can,  and have, identified  the issues and opportunities to reduce
pesticide-related toxicity,  but they have limited  ability or authority to actually  implement  corrective actions.

Regulatory

    Regulation of pesticides including their registration for use  in the Unites States is the responsibility of USEPA.
California's Department of Pesticide Regulation  (DPR)  has responsibility for regulating  the  sale  and use of pesticides in
California. California DPR,  with few exceptions, registers pesticides only  after they have been registered  by USEPA.
California DPR  can not register pesticides which have been denied registration by  USEPA. At the local government  level,
the California Food and Agriculture Code grants  some authority to  county agricultural  commissioners for local  enforcement
of pesticide regulations, record keeping, and  outreach to applicators. However, with the exception of county agricultural
                                                      246

-------
commissioners,  local governmentsare prohibited from regulating the registration, sale, transportation,  or use of "economic
poisons."  This  regulatory structure means that the ability and authority of local governments is limited to:

    .  identifying opportunities to reduce toxicity,  such as eliminating  potentially problematic uses,  and advocating  that
      state and federal agencies seize  these opportunities through regulation  and re-registration

    . in the case of wastewater and stormwater agencies, regulating the discharge of pesticides to the sewer or storm
      drain to ensure  local agencies' compliance with state and federal laws (e.g., Clean Water Act)

Monitoring

    Local governments have some ability, authority, and responsibility to use monitoring to address the problem of
pesticide-related toxicity of surface waters.  To-date, San Francisco Bay Area municipalities have  used monitoring to:

    .  identify and define the problem (Alameda County Urban Runoff Clean Water Program, 1995; Regional Water
      Quality Control Plant-Palo Alto, 1996; California Regional Water  Quality Control Board, 1997; San Francisco  Bay
      Area Pollution  Prevention  Group,  1998)

    .  characterize sources (Alameda  Countywide Clean Water Program,  1997)

    .  recommend corrective actions (Alameda County  Flood  Control and  Water Conservation  District,  1997)

    A review of monitoring  data from around the country shows that municipalities in the San Francisco Bay Area and
California  Central Valley are not alone in their identification of this environmental problem.

    .  Orange County, California  (Lee,  et al., 1999) - Multi-year studies of stormwater runoff in San Diego Creek as it
      enters Upper Newport Bay have shown  that the problem is not restricted to Northern California. Runoff from each
      stormwater event has been shown to be toxic, and about half of the  observed toxicity is due to diazinon and
      chlorpyrifos used in  urban  areas for structural termite  and ant control, and lawn and garden  pest control.

    .  NAWQA  (USGS, 1998) -  Results from the United States Geological Survey's (USGS) National Water Quality
      Assessment Program from 1992 through 1996 show that the problem is in fact  a national one. Over 300 samples
      have been  taken from eleven urban streams scattered across the country, from Florida to Connecticut to Oregon,
      as part of the Pesticides National  Synthesis Project. In  a recent report on the first  cycle  of  the program, USGS
      concluded that "urban and  suburban  areas  are substantial  sources of pesticides to streams" and  that "most urban
      areas have similar pesticides in streams...and many urban areas  may benefit from similar strategies for reduction."

    .  Publicly-Owned Treatment Works survey (USEPA,  1989) - Results from a survey done 10 years  ago by USEPA
      show that pesticide-related toxicity is a wastewater problem as well as a stormwater problem. USEPA's
      Environmental Research  Laboratory in Duluth, working through the National Effluent Toxicity  Assessment Center
      (NETAC), reported  on the  occurrence of diazinon in 28 POTW effluents. Diazinon was found in  sixteen (62%) of
      the effluents, and levels were greater than or equal to 250 ng/L for nine (32%) of the effluents. NETAC  concluded
      in part "The frequency with which we  have  observed diazinon in the past, in this survey, and  continue to find it in
      effluents  is indicative of  a widespread problem."

    Clearly this is a national problem caused  by  products that are registered at the national level and  sold across  the
country.

    The pesticide registration process provides  a built-in mechanism to use monitoring  and science to address this
national problem.  During the registration  process,  USEPA must review and summarize the  findings of studies  conducted
on each pesticide. During this step, USEPA may  request that "registrants" (e.g., pesticide manufacturers) submit specific
studies for review. Based  on its review, USEPA can confirm, deny, or change the  pesticide's registration including
approved  uses, sites of application, formulations, and label directions.
                                                     247

-------
    Local  governments are willing to use monitoring and science to further investigate local impacts and sources,  and
to host case studies, if USEPA will  provide financial and other support, with the goal of conducting  representative case
studies whose results can  be  extrapolated  across the country. But given the  established  mechanism in the pesticide
registration process, it would be inappropriate and ineffective for local governments to do more. USEPA must exercise
its federal authorities and use monitoring  and science information to make  more informed, up-to-date  registration
decisions.

Conclusion

    Rather than  being a tool in and of itself, the Pesticide Toxicity Reduction Strategy is really a toolbox.  It includes a
number of effective tools for reducing pesticide-related toxicity of surface waters-an increasingly important part of urban
water resource  management and  protection. Every job has its tool and in the right pair of hands, the job can be easy  and
cost-effective to complete.  The wrong tool  or the wrong hands  can make the job difficult, if not impossible to finish. It
is the responsibility of government  agencies to be clear and disciplined about which tool and which pair of hands go with
which job when  fixing environmental problems.  The extent to which they implement that  concept  will determine  how
successful  the work of environmental protection will be.

References

Alameda County Flood Control  and  Water Conservation District, 1997.  Strategy to Reduce  Diazinon Levels in Creeks
in the San Francisco Bay  Area. Prepared by J.  Scanlin and S. Gosselin.

Alameda  County Urban Runoff Clean Water Program, 1995. Identification  and Control  of Toxicity in Stormwater
Discharges to  Urban Creeks. Prepared by  S.R.  Hansen and Associates.

Alameda Countywide Clean Water Program, 1997. Characterization of the Presence and  Sources of Diazinon  in the
Castro Valley Creek  Watershed. Prepared  by  J.  Scanlin and A. Feng.

Bay Area  Stormwater Mgt. Association,  2000. Strategy for Reducing Organophosphate Pesticide-Related Toxicity in  San
Francisco  Bay Area  Urban Creeks.

California  Regional Water  Quality  Control Board, 1997. Diazinon in Surface Waters in the San  Francisco  Bay  Area:
Occurrence and Potential Impact. Prepared by T. Mumley (RWQCB) and R. Katznelson (Woodward  Clyde Consultants).

Lee, G. Fred, et al.,  1999.  Evaluation of the Water Quality Significance of OP  Pesticide Toxicity in Tributaries of Upper
Newport  Bay, Orange County, CA.  Environmental Toxicology and  Risk Assessment:  Recent Achievements in
Environmental  Fate and Transport: Ninth Volume, ASTM STP  1381, American  Society for Testing and Materials, West
Conshohocken,   PA.

Regional Water Quality Control Plant-Palo Alto,  1996. Diazinon in Urban Areas. Prepared  by A. Cooper.

San  Francisco Bay Area  Pollution Prevention  Group, 1998.  Diazinon & Chlorpyrifos Quantitative Identification for  San
Francisco  Bay Area Wastewater Treatment Plants. Prepared  by T. Chew, K. Easton, and A. Laponis (Central Contra
Costa Sanitary  District).

USEPA, 1989. Results of Diazinon Levels in POTW Effluents in  the United States. National Effluent Toxicity Assessment
Center (NETAC) Technical Report  14-89.  Environmental  Research Laboratory-Duluth.  Prepared by  T. Norberg-King,
M. Lukasewcyz,  and  J. Jensen.

USEPA, 1999. 1998  California 303(d) List and TMDL Priority Schedule.

USGS, 1998. Pesticides in  Surface and Ground  Water of the  United States: Summary of Results  of the National  Water
Quality Assessment  Program (NAWQA) (http://water.wr.usgs.gov/pnsp/allsumm).
                                                     248

-------
                     Administering the  NPDES  Industrial Storm Water
                                  Program at  the Municipal  Level
                                              Michael J.  Pronold
                                       Bureau of Environmental Services
                                                City of Portland
                                               Portland, Oregon
Abstract

    As part of the EPA  Phase 1 stormwater requirements, certain classes of industries are required to obtain Industrial
Storm Water  permits. The EPA, or a  state agency that  has been delegated  by EPA, administers these permits. The
Phase 1  regulations  also require that municipalities develop  a program to monitor and control  pollutants in storm water
runoff from industrial facilities.  These are potentially non-coordinated requirements and  can result in  redundant efforts
and  a  less than efficient program.  In  addition, EPA and/or  state agencies may not have the resources to adequately
administrate and enforce the permitting  program while leaving  the municipality liable for the discharges from the municipal
separated storm sewer  system (MS4).

    The City  of Portland, Oregon (City), met the  requirement in  its municipal storm water permit to control industrial
stormwater sources of pollution by developing a Memorandum of Agreement (MOA) with the Oregon  Department of
Environmental Quality (DEQ), (which is the delegated  authority) to  administer the permit  program. The MOA provided
the City with the mechanism to administer the industrial stormwater permits for those facilities that discharge to the City's
MS4. The City pursued this approach since  it was responsible for the discharge from the MS4 and wanted to ensure that
it had adequate oversight of these discharges. By coordinating this effort with other ongoing industrial water quality
programs, the City could provide a more cost-effective  program, considering the regulatory costs as well as cost to the
industry.  City Code was developed to support this approach.

    When the City took over the administration of the permits  in 1994, over 50% of the facilities with a permit had not met
the requirements  for the development  of a storm  water  pollution  control plan, the main requirement  of the permit.  In
addition,  nearly 60% of the  permitted  facilities had not performed the required stormwater sampling. Of the samples
taken, approximately 30% violated standards in the permit.  It was also  evident that  not all facilities required to obtain a
permit  had done so. Efforts  since 1994, have shown that only 25-30%  of the facilities required to obtain a permit had
applied.  A benefit of the local administration of the program is the detection of illicit discharges to the MS4.
Approximately  15%  of all industrial inspections have  identified illicit discharges.

    The City  has also identified certain classes of industries and  activities that can be significant sources of pollutants
to the MS4. This has helped  streamline the program efforts and redirect resources to  where the greatest cost benefit will
be  realized.

Introduction

    Stormwater 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.  These regulations established  NPDES permit application requirements for storm  water discharges

                                                     249

-------
associated with large- and medium-size MS4s. The regulations also established 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.

    A requirement of the City's application process was "A description of a program to monitor and control pollutants in
storm water  discharges to municipal systems from municipal landfills,  hazardous waste treatment,  disposal, and  recovery
facilities, industrial facilities that are subject to Section 313 of Title III of the Super-fund Amendments and Reauthorization
Act of 1986  (SARA), and industrial facilities that the municipal permit applicant determines are contributing a substantial
pollutant loading to the municipal storm sewer system." (40 CFR 122.26(d)(2)(iv)(C). This creates the potential for
redundant efforts and a less than efficient program.

    The stormwater regulations envision  that NPDES permitting authorities and  municipal operators  will cooperate to
develop programs to monitor and control pollutants in storm water discharges to MS4 from certain industrial facilities.
The NPDES permits for industrial facilities establish  requirements such as controls,  practices, and monitoring  of
stormwater discharges, as well as provide a basis for enforcement actions. An integral part of the requirement is the
adequacy  of legal authority.  This will allow the municipality to implement its program, which should include inspections,
review of stormwater  pollution control plans,  monitoring,  and implementation of control measures.

    The municipality  is  ultimately  responsible for discharges from its MS4. To meet the requirement in its municipal
stormwater permit, and to provide the oversight  necessary to protect itself from liability, the City developed  new  legal
authority and entered into an MOA with the authorized NPDES state authority  (DEQ), to administer the  permits for those
discharges to the MS4.

Program  Elements

Legal  Authority

    The City did not have adequate legal authority to oversee discharges to the MS4.  In response to this,  the City
developed code  in February 1994. Some of the  major provisions of the code are:

    . Authority  of the Director of Environmental  Services to  Adopt Rules

    . General  Discharge Prohibitions

    . Discharge  Limitations

    . Reporting Requirements

    . Storm Water Pollution Control Plan (SWPCP)

    . Storm Water Discharge Permits

    . Inspection  and  Sampling

    . Enforcement

    Key elements of the code include the requirement for permit holders to submit their SWPCP  and  monitoring results
to the City, the authority  for the Director to adopt administrative rules, make  inspections, and undertake enforcements.

Memorandum of Agreement

    The City entered into a MOA with the DEQ in March 1994. The MOA delineates the responsibilities for the
implementation of the  program between the two agencies. The MOA also prioritizes the implementation of the  program
to address those facilities that are of most concern first.  Key elements of the City's responsibilities include:


                                                      250

-------
    . Development of an inspection and  monitoring program

    . Informing DEQ  of any new or existing facilities that  require  a permit

    . Enforcement of City  Code

    Key  elements of DEQ's responsibilities include:

    . Issuance of NPDES  Industrial Storm Water permits  upon referral or approval  by the City

    . Denial of permit applications  for process wastewater discharge  into the  MS4

    . Enforcement where the City  lacks authority

Inspections and Monitoring

    The  section  responsible for the implementation  of the  program is  an autonomous work group. It is housed in the
Source Control Division,  which includes the pretreatment  program. In November 1994, two  inspectors  were  hired  to
implement the program under an  existing  supervisor. DEQ, which  had been  issuing permits since  September  1992,
provided  a list of facilities with stormwater permits. A letter was sent to the  permit holders requesting that they submit
their SWPCP and all monitoring results. The letter  referenced the MOA  and included  code  citing the City's authority.
Inspections were  prioritized based on problematic outfalls as determined from information gathered in the  Part  1 and 2
application   process.

    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.  Technical assistance is provided  and information  on Best Management Practices given in the form of verbal
suggestions  and reprints. Facilities are also evaluated for the presence of illicit discharges. 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.

    Industries are also inspected  if they are identified as potentially  needing a permit. These facilities are identified
through a systematic search using storm water outfall basins prioritized based on problematic outfalls. The basins are
delineated for drainage, the industrial facilities  identified within the basin using our database,  and facilities selected by
SIC Code. Inspections are also performed  in response to referrals, drive-bys, complaints, and  responses to an industrial
survey performed in support of the pretreatment program.  Prior to  an inspection, building records, existing files from the
pretreatment program,  and plumbing records are reviewed.

    Stormwater 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. This sampling  does not relieve the facility of its
stormwater  sampling  responsibilities.  Files are developed on the facilities and maintained separately from the
pretreatment files. An electronic database has been  developed and is used by both pretreatment and  storm water staff.

Enforcement

    Enforcement capability was developed  in City Code.  Where the City does not have enforcement authority, it seeks
voluntary compliance and  refers enforcement to the permitting  authority  when  necessary.

Funding

    The  program  is entirely funded through a surcharge on the storm water fee  for industrial and commercial accounts.
The storm water fee  is currently based on impervious area.  This  surcharge also funds portions of other  programs that

                                                     251

-------
have work related activities because of industrial  and commercial storm water discharges. The  current staffing level is
one supervisor and three technicians. Program costs amount to approximately $280,000 per year.  This is primarily
composed of salaries and benefits, but also includes approximately $25,000 for sample analyses The budget also
contains funds for the development of BMPs and educational materials.

Findings

Legal Authority

    It is essential that legal authority be developed in order to be able to implement and support a municipal program.
However, what is contained in the enabling legal  authority (code) can vary drastically.  It is important that the municipal
permit  holder review the NPDES  Industrial Storm Water permit to determine its adequacy in meeting the municipal permit
requirements. Most industrial storm water permits are general permits and they may not adequately  address  issues for
which the  municipal  permit holder is  responsible.

    For instance, if the municipality is responsible for meeting TMDLs for a  particular water  body, the industrial permit
may not even require that the facility  monitor for these pollutants in its  discharge. Provisions should  be placed in code
that allow the municipality to  require the  facility to conduct this monitoring. Another example  would be the requirement
to submit SWPCP and monitoring results to the municipality if this is not included in the permit. Nothing in federal
regulations prohibits the municipality from requiring additional  controls beyond the permit requirements.  A review of the
industrial stormwater permits  can help identify elements that  should  be  included.

    Another provision  that should be considered is the ability  of the municipality to require a facility  to obtain a  permit.
Currently, federal regulations base the requirement for a permit on SIC Code and exposure. There is a caveat that allows
the permitting authority to require a facility to obtain a permit regardless of its  SIC Code if that facility  is impacting water
quality.  However, this could  require that  the municipality  undertake  sampling and additional work to prove an  impact.
This reduces the efficiency of the  program  in terms of resources and uniformity. It may even be necessary to include
provisions in the code that allow the municipality to develop its own permit. Such a tactic is time consuming,  however,
and could  create confusion for the regulated community.

Memorandum of Agreement (MOA)

    The MOA should be developed  to clearly outline the responsibilities between the  permitting authority and the
municipality. Language should be  broad  enough  not  to constrict  how the municipality implements the inspection and
monitoring  program. This allows the municipality to alter the program as information is obtained from  inspections without
having to alter the MOA. Probably  the most important element of the MOA is the delineation of enforcement. Since the
municipality does not  have authority to enforce permit conditions, language should  specify that the municipality will enforce
applicable requirements of the Code and  seekvoluntary compliance where it has no independent enforcement authority.
If compliance is  not obtained using these methods, enforcement would be referred to the permitting authority.

Inspections

    The City has placed the responsibilities for implementation of the program within the Industrial  Source Control
Division. The section  also houses the pretreatment program.  It was  felt that  the responsibilities needed to be separate
because of the large number  of facilities that are to be addressed. The  City  has over 24,000 commercial and  industrial
facilities. Of these, nearly 3,000 have  the SIC Code that  potentially  places them in the permitting program. In addition,
a  Stormwater Work Group is responsible for addressing  other discharges to the MS4, such as pumped groundwater,
boiler blow-down, water supply line flushing,  washwater,  and  others.

    For the City's situation, this arrangement has worked very well. The Work Group is able to develop expertise in the
area while having access to existing information from the pretreatment program. Other municipalities have adopted this
approach  while others have  incorporated the responsibility into the  pretreatment program or other existing  programs


                                                      252

-------
including fire and  safety inspections.  The municipality needs to consider several items when determining  who  will  be
responsible  for implementing  the  program.

    . The number and type of industries

    .  Existing oversight of the industries (pretreatment, hazardous materials,.  . .)

    .  Existing  programs within  the  municipality

    If the municipality decides to place the responsibility in a Work Group that is not dedicated for this purpose, it needs
to ensure that adequate resources exist to implement the program and meet the conditions of the MOA. The stormwater
program may  not be the priority of the assigned work section and if resources become inadequate, this work may be
viewed  as low  priority and  may not be addressed at the level that makes it effective.

    The City has developed  several "partnerships" to expand the inspection program. Informational flyers and a poster
were  developed for county sanitarians to use  when they  inspect restaurants. A 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 issues of consequence would be addressed by storm water staff.

Permits

    The DEQ has been  issuing permits since September 1991. When the City took over administration  of the permits
in the fall of 1994, 63 facilities that discharged to the  MS4 had  permits.  Since that time, an additional 65  facilities have
been identified  through inspections of non-permitted facilities. Non-permitted facilities are inspected based on SIC Code
and prioritized by outfall basins that have been identified as  problematic. This approach was  necessary due  to the large
number of industries within the City that have the SIC Code included 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 to the issuance of a  permit as the City is a  mixture of combined sewers, sumps, and separated storm
sewers. Staff  members  spend  a considerable amount of time determining where stormwater drainage  discharges. A
municipality  may be able to utilize this approach  if the industrial  base  is smaller. Federal guidance states that a system-
wide approach to  establishing priorities for inspections should  be developed.

    Based on inspections of non-permitted industries  to date (approximately 15% require  a  permit), and  the  remaining
facilities that require inspections,  it is estimated that  an  additional 50-100 facilities will  be permitted. Based  on these
numbers, only 25-30% of facilities  requiring a permit had  applied when the City took over administration of the program.
However, a  large percentage of the facilities not requiring permits still had issues that needed to  be addressed or were
given  BMPs that they were requested to  implement.

SWPCP

    The original general permit developed by  DEQ did not require that the permit holder submit the SWPCP. When the
City took over administration of the permits, the plans were to be submitted using provisions of the  City Code. Over 50%
of the facilities (33 of 63)  had  not developed a plan within the 180 days allowed in the permit, and of these,  14 (22%) had
not even developed a plan. It is imperative that the municipality includes provisions in the code to obtain  these plans if
the provision does not exist in the permit. The  requirement to submit the plan allows the  City to  track its development
and review the plan prior to an inspection. Currently, only 5% of facilities have not met the requirement to develop the
plan in  the  required time period.

    Unfortunately, there  is no requirement in the  permit that the  plans need to be approved. As long  as they contain the
necessary elements required  in the permit they  would be in compliance. This has proven problematic in the quality of
some  plans. It also restricts the City's ability to  require that the facility implement certain pollution  control  activities. This
emphasizes the need to include these provisions in  the legal  authority. The City has taken the approach of strongly


                                                      253

-------
suggesting that certain activities be implemented and incorporated into the plan.  Once it is in the plan, it becomes part
of the permit and provides a mechanism to require the facility to implement these  measures. The City is then able to take
the position of providing  assistance in  evaluating  compliance with the permit. By noting these deficiencies and  seeking
voluntary compliance, the City believes it is assisting the industry in  meeting the conditions of their permit and benefiting
the environment through the implementation of the SWPCP.

Monitoring

    When the City  took  over administration of the  program in  1994,  monitoring  results  were requested from permitted
facilities. Nearly 60%  (36) of the facilities had not performed the required monitoring for the previous year. Of these, 22
had  not taken the required two samples, while the remaining 14 did not perform  the complete analyses. Of the  samples
taken, 30% violated standards in the  permit. Within the first  year, the City was able to  raise compliance  on sample
collection to over 80% and reduce violations of standards from 30% to 23%. Currently, over 90% of the facilities are in
compliance  with sample  collection. It is more difficult to compare compliance with standards because a new  permit was
issued in  1998, that includes benchmarks for  metals that were not  in the original permit.

    Monitoring results have limitations because they  are grab samples  taken from a discharge  that is short-term in nature
and  highly variable. However, they can be used as  a tool to measure effectiveness of BMPs and to identify  sources of
pollutants.  Based on  sample results, the City  has identified several classes of industries  that pose significant pollution
concerns. These  are,  in order:

    . Automotive recyclers  (SIC Code 5015) - metals, oil  and grease;

    . Recycling industry (5093) -  metals, oil and grease;

    . Transportation facilities (various 4000) - metals, oil and grease, TSS;

    . Heavy manufacturing (33--,  34-) - metals;

    . Food industry (20--) - TSS,  BOD, oil and grease.

    Other SIC Code groups either represent a  lower threat as a whole or are  not present in the MS4. The  City is now
using this  information to reprioritize their efforts in identifying industries that require a permit. While the  City is still
pursuing efforts based on outfalls,  they are also developing a  parallel effort to  inspect all the facilities in these classes.
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. These classes are not included in the
federal regulations,  but any municipal  program should  evaluate these facilities.

Enforcement

    Enforcement capabilities have  been developed in code for those discharges to the MS4. However,  the City does not
have enforcement capability on  permit  provisions. The  City must seek voluntary  compliance and refer those  matters to
the permitting authority  for which they don't have enforcement capability. This  has worked to date, but requires
coordination between  the City and  DEQ.  When seeking voluntary compliance, the City uses the threat of referral to the
permitting authority  or third party lawsuits to obtain  compliance. To make this effective, the permitting  authority must be
ready to follow up with enforcement upon the  municipality  referral.

Funding

    As with a number of environmental programs,  especially regarding storm water, it is  very  difficult to  measure  the
cost/benefit until the program has been in place  for a period  of time. Costs have been  identified, and certain benefits have
been realized. Compliance with permit conditions, for both industry and the City, have been,  for the most  part, met.
However,  has this resulted in  a benefit to the  environment? City data have shown  that industrial land use areas have


                                                      254

-------
significantly higher concentrations of pollutants than other land uses. Whereas the industrial land area in the MS4 is only
4%, it accounts for 11% of TSS, 15% of oil and grease,  and 24% of metals. It would reason that a program aimed at the
highest concentration of pollutants would produce a good  return on the investment, Another benefit of the program has
been the identification and removal of non-stormwater discharges. Approximately 15%  of the inspections  have identified
non-stormwater discharges, primarily washwater, that were of concern.

Conclusions

    The development of a program to monitor and control pollutants from industrial facilities is not one of the six BMPs
that Phase II 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.

    If a municipality decides to develop and  implement a program, it is recommended that it utilizes the accomplishments
of Phase 1 applicants. Phase 1 applicants can provide inspection forms, BMPs, MOAs, code language,  and other
necessary components to develop the program.  They can also share results of their work to help prioritize the efforts of
the municipality and  help decide how to incorporate the work into existing programs.  A municipality  may also become
a co-applicant 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 it at  least 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 it identifies 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 permit. Whatever the municipality chooses, it needs to understand that it is ultimately responsible for
discharges from its  MS4.
                                                     255

-------
          Lessons Learned from  Three  Watershed-Sensitive Development
                     Demonstration Projects in  the Great Lakes Basin
                                          Sarah Bennett Nerenberg
                                           The  Conservation  Fund
                                             Great Lakes Office
                                               Chicago,  Illinois
Introduction to The Conservation  Fund

    The Conservation  Fund (TCF) is a  national,  non-profit conservation organization that purchases and protects land
- more  than  1.6 million acres  since 1985. TCF also  assists  local  communities, private landowners and government
agencies with programs that balance conservation with economicdevelopment. TCF works with communities to improve
water quality,  build sustainable economic opportunities, and develop leadership skills, activities that put it at the forefront
of conservation across America.

    TCF has  been active in the Great Lakes Basin since it opened a regional office in 1995. The initial focus of its work
was the Great Lakes Watershed Initiative. This basin-wide effort was  designed to raise the local visibility of the nonpoint
source water  pollution  issue. The  Initiative adapted  many  of the innovative solutions showcased in the National Forum
on Nonpoint  Source Pollution. TCF worked with many  local  partners to launch a network of community-based  projects
addressing nonpoint source water pollution  in urban and rapidly urbanizing areas in  eight states and Canada. The
Initiative was  conducted in partnership with the Council of Great  Lakes Governors with major funding from the Great
Lakes Protection Fund  and Kraft  Foods.

    TCF expanded several projects as an outgrowth of the Initiative including the watershed-sensitive development work
outlined in this paper and a sustainable development effort in Michigan. In Michigan, TCF facilitates a broad, community-
based sustainable development effort in the  Saginaw  Bay watershed. The goal  of the initiative, which engages local
businesses, community  groups,  and government agencies, is to better link the environmental and economic well being
of Saginaw Bay communities in order to sustain  and  improve  the region's overall quality of life.  This year, the project
received the  National Award for Sustainability  from the President's  Council  for Sustainable Development and  Renew
America.

Introduction to Conservation Development Project

    Currently, TCF is targeting one of the remaining threats to  natural  resource quality,  enhancement, and preservation
in  urbanizing  areas -  conventionally designed subdivisions. In partnership with  local  developers,  community  groups,
and government agencies, TCF is working in the Great Lakes Basin on the Conservation Development project. This
project is designed to  demonstrate the environmental and economic benefits of watershed-sensitive design  through a
series of model developments.  In  particular,  we  are working to demonstrate the benefits of watershed-sensitive site-
planning and  best management practices that reduce impervious cover and conserve open space. The current model
projects are being developed in Huron,  OH,  Germantown, Wl, and  Niles,  Ml. The George Gund  Foundation and the
Great Lakes Protection Fund have provided major funding for  this project.

    We define watershed-sensitive development to include: open  space design, significant reduction in impervious
coverage,  natural stormwater conveyance and storage to the greatest extent possible, and  appropriate construction
mitigation measures. Watershed-sensitive design can be used  to build the same number of houses and still preserve
a significant  portion of the subdivision's original landscape.  These open spaces serve important community and
environmental functions. Agricultural land can  be farmed,  residents can enjoy recreational and aesthetic benefits, and
important natural areas and systems can be preserved.  Alternative designs also reduce the amount of impervious cover.

                                                    256

-------
Techniques including narrower streets, porous surface parking areas, stream buffers,  and open channelsforstormwater
conveyance  minimize runoff from  new development  and its negative impacts on water resources.
    When evaluating potential conservation development  projects, The  Fund considered the following criteria:
    .  Local community  must be interested and open to new techniques, including  flexibility on zoning and subdivision
      code issues;
    .  Property already  slated  for  development  and conventional development would  have significant negative impacts
      on the  site itself  or adjacent natural resources;
    .  Project partners represent one of the dominant development paradigms in  the Great Lakes (i.e., professional
      developer  building homes in farm fields, lay developer seeking to  hold and protect  family or other special  lands,
      government agency  seeking to encourage  sound practices);  and
    .  Project  site is suitable for demonstrating  broad array of site design techniques  and best management practices
      (BMPs).
    Through TCP's work in the Great Lakes, we  gathered many lessons-learned that may be applied to  other regional
and national efforts, This paper will review many of these lessons with the hope that other communities and
organizations  will be  able to benefit from our experiences. The paper is organized into the four sections listed below:
    I. Overall Lessons Learned
        1. Not "One  Size Fits All"
        2.  Measurable   Criteria for  Watershed-Sensitive  Development
        3. Adequate Oversight and Inspection
        4. Incentive System Needed
        5. Relationship  to Other  Smart  Growth Movements
    II. Lessons  Learned about the  Development Process
        1. Pace of Development  Often  Incompatible with Innovative Site Design
        2. A  Greater Initial Investment  in the Baseline Information  is Necessary
        3.  Initial  Cost  of  Watershed-Sensitive Developments
        4. Deed Restrictions
        5. Need Additional Lay Developer Education
        6. Aesthetics Do Not Equal Ecology
    III.  Lessons  Learned  about  Engineering/Site  Design
        1.  Educate the Engineers
        2. Lot Size Often  Dictated by Septic Issues
        3. Need Hard  Science
                                                     257

-------
    IV.  Lessons Learned about Working  with  Communities

        1.  Community Initiative

        2.  Local Official  Knowledge Varied

        3. Strong Local Partner is Key

        4. Final Lessons Learned

Overall Lessons  Learned

Not "One Size Fits AH"

    The approach used to create a watershed sensitive development must  be tailored to the individual, organization, or
developer creating it. The assistance  needed by a private landowner that is seeking to preserve portions of family lands,
for example, is  quite different from that needed by a professional developer. We found that the models  we developed
need to take the different skills and goals of the project's initiator into account very early in the process.   For example:

    Lay developers (i.e., the  individual landowners), not surprisingly, need help with the business aspects  of the  project,
and are more inclined  to make frequent changes to the preliminary site plan and architectural style of the development.
These changes  often reflect something the developer has "just learned" or "just considered." These new ideas can add
value to the project, but they also  require the technical assistants (e.g., landscape architects and engineers) to be more
patient,  more flexible,  and firmer than they  might be with  professional developers.

    Professional developers demand immediate turnaround on requests for assistance, and are looking  for "the facts"
on what is  required to make a development watershed-sensitive. They can be somewhat impatient with the notion that
there are not a  fixed and specified set  of best management practices and site design practices that, if employed, will
"always"  result  in an  "environmentally  friendly" development.

Measurable  Criteria for Watershed-Sensitive  Development

    As we  began to design  the model projects,  it became apparent that there were no  specific criteria available to
measure the benefits of the watershed-sensitive design. A tool was needed to encourage developers to  fashion
environmentally friendly site designs, to help communities add flexibility to their local ordinances, and to provide a
standard that can be  understood  by both homebuyers and existing community  residents.

    In response to this, TCF developed the  Conservation Development Evaluation System (CeDES) as a rating  system
to evaluate a conservation development over the  development's lifetime with emphasis on  water quality and landscape
impacts. The purpose of CeDES is to encourage  developers to think about environmental concerns earlier in the
planning process and to provide consumers and communities with a means  of assessing the impact of better site design
practices. It was developed with input from over thirty  national professionals skilled  in planning and  evaluating
conservation developments. It may be viewed at http://www.conservationfund.org/conservation/sustain/gloindex.html.

Adequate Oversight and Inspection

    One of the  biggest challenges  is  ensuring that the contractors are building in an environmentally responsible manner.
Even if the developer  is committed to minimizing the impact of the development on the environment, if contractors are
not  educated and committed it may not happen. This is  a challenge for many local agencies  and municipalities who have
limited staff for  constant inspections.  Even if the communities have ordinances that require  construction erosion  controls
etc., without constant inspection many  contractors do not follow the requirements. The nonpoint source  pollution from
construction, especially the sediment loadings, can negate any benefits from the alternative site design. One
recommendation is for the community to require that an environmental  inspector be  hired  specifically for the site. The
inspector may be from a  consulting firm or from a local Soil and Water Conservation District.

                                                    258

-------
Incentive  System Needed

    We expect that this process will proceed  much more quickly in communities that have recognized the threats
conventional developments pose and have  begun developing strategies to address them. In order to expand
conservation development practices to a broader constituency, state or county agencies may need to develop incentive
programs that prompt local developers to  undertake these  projects. With each community, we encouraged the creation
of incentives for watershed-sensitive development. These  included density  bonuses for the developer through  credits
for land preservation and minimization of impervious coverage. We also investigated the use of the Clean Water Act
State Revolving Fund (SRF). Among other uses, these funds are  used to  reduce nonpoint source pollution and could
encourage watershed-sensitive development.  The State of Ohio has successfully used  the SRF for this  purpose and
we hope to pilot the same use of the SRF loans in other states. Incentives such as Ohio's loan program, coupled with
the  higher  financial  returns these  developments are expected to generate, are making watershed  sensitive  developments
more the norm in the Great Lakes Basin.

Relationship to Other Smart Growth  Movements

    There are many "Smart Growth" movements currently  being debating and  promoted throughout  the country.
Watershed-sensitive development is just one  part of the  equation.  At times, we were challenged to show how this fits
into overall community sustainability efforts. The work that the National  Site Planning Roundtable completed to develop
"Better Site Design" principles has been  invaluable in demonstrating how these different movement can work together
(Center for Watershed Protection, 1998).  We often say that watershed-sensitive development is one option for Smart
Growth but that a community needs to find the  correct planning principles to  work for their residents and  issues. Those
in the Traditional Neighborhood Design (TND)  movement  challenge putting sidewalks only on one side  of the street,
which we  recommend  for reducing impervious coverage.  We also suggest that if sidewalks are on both sides of the
street at least  one  should  be made of pervious materials. There  also  are environmental groups that  challenge us for
encouraging  greenfield development instead of  infill development.  Again, watershed-sensitive development is  only one
option of many and if the market  is going to demand suburban fringe growth,  at least we can work with  the communities
and developers to ensure that it  is done with maximum possible protection and enhancement of the natural resources.

Lessons Learned about the Development Process

Pace of Development Often Incompatible with Innovative  Site Design

    The pace of development and the  pace of government decision-making often are absolutely incompatible.
Developers with outstanding loans on land need to move quickly to ensure a development is economically viable.
Government agencies, on the other hand, are very concerned about the  impacts of development, but move  very
cautiously,  especially when they are undertaking something new. The result is that it is easier for both government and
developer  to create conventional, environmentally  harmful  developments than  to do something better.

    On the demonstration projects, TCF took  special care at the outset  of the process to communicate  the timelines of
each  participant to the  other.  In this way, we hoped to keep the parties  from throwing up their hands and  giving up. For
the region, however, we explored  possibilities  to get communities to adopt "fast track" approvals  for watershed-sensitive
communities. The first need is to  show municipal authorities that these developments  deliver tangible benefits, then
we  can help them develop mechanisms such as a streamlined review process and updated subdivision  and
zoning ordinances that encourage their creation.

A Greater Initial Investment in the Baseline  Information is Necessary

    Before planning a  watershed-sensitive development,  fairly detailed baseline  information including topography, soils,
and wetlands delineations is  needed. Although  developers hope that they  get  enough  of the baseline site information
before beginning design work, inevitably the risk/benefit  of doing extensive  baseline work (e.g., soil borings)  may
preclude the developer from  getting all  of the necessary baseline data. The common  practice  is to use  "engineering


                                                   259

-------
judgment" based on existing data and extrapolation to the rest of the site. Unfortunately, especially when the drainage
plan is an integral  part of the initial site design, relying on "engineering judgment" simply is not sufficient. For example,
on the Ohio site we relied  on the existing soils information to design the swale system. After presenting a  preliminary
site plan, the developer discovered through additional research and sampling that  the available soil information  did not
accurately represent the  existing soil conditions and the drainage plan  had to  be  reconstructed.

Initial Cost  of Watershed-Sensitive Developments

    Planning and developing a watershed-sensitive development takes time and costs money. Both lay and professional
developers  often underestimate these  initial costs. Professional developers often leave site planning to their  engineers.
The  engineers typically obtain  a wetlands delineation  and examine soil  and topography  maps, but do not evaluate the
site from a watershed or ecological perspective. Although  lay developers may be more familiar with the special features
of their properties than professional developers, both need help to catalog all the features and to  understand the site's
role  in the surrounding landscape.  Quite  reasonably,  professional  developers  often are unwilling to undertake these
expenses until they have a sense of the  project's scale and niche in the  market. We believe, and existing watershed-
sensitive  developments indicate, that the costs of evaluating a property from an  ecological and a watershed perspective
will be recovered when the development is sold out.

Deed Restrictions

    The deed restrictions (i.e.,  covenants, conditions, and restrictions) necessary to ensure that  the  development will
continue  as a watershed-sensitive development in  perpetuity are a lot more extensive than typical deed restrictions.
Early in the process,  sample  restrictions for various  developments should  be  presented to the developer and to the
community so  that they understand the consequences  of  using some of the watershed  sensitive techniques. The
developer will  gain  an appreciation  for the long-term commitment necessary for a successful development  and  local
officials may be put at  ease when they recognize that  major additional  responsibilities (e.g., swale maintenance) rest with
the homeowners association and not  the  local government.

Weed Additional Lay  Developer Education

    Private landowners need to  be assisted  and educated through the  process. Although these  initiators often  have  a
deeper environmental  commitment than professional developers,  they often do not understand what kinds of activities
on their properties will have  negative watershed impacts.  For example,  on  one of our projects, the lay  developer
suggested that a pond  be built  each time an area of low-lying  ground is found to be  wet most of the year. Once informed
about the relationship of these  areas to more prominent wetlands  on the site, the developer agreed to treat these areas
more appropriately (i.e., preserving and enhancing the existing wetlands). The professional developers understand
stormwater and wetlands issues better because they operate in the regulatory arena. The lay developers may need to
be educated about the significance of these  issues and other  issues that are common knowledge to professional
developers.

Aesthetics  Do  Not Equal Ecology

    Another aspect of landowner education is the principle that aesthetics do not equal ecology. Just because  a
development  preserves or creates attractive  green  spaces does  not  necessarily indicate that it is not  harmful to the
surrounding  watershed. Accordingly,  the  criteria we  developed  for  watershed-sensitive  development  (see  discussion
above  under Measurable Criteria) incorporate appropriate  baseline evaluation of the site to  insure key resources are
protected, and  a thorough analysis of the stormwater impacts after development.
                                                     260

-------
Lessons Learned  about Engineering/Site  Design

Educate the  Engineers

    If any of the county, township, or city engineers are not comfortable with the techniques being  used, they can turn
down the project at any point in the review process. In all three of our projects, the "old-timer-" engineers were extremely
conservative and feared change more than any other local officials. We found that the developers' engineers need
constant oversight  and education to design the sites using the watershed-sensitive techniques. Unfortunately,  without
a broad effort  to educate engineers, they will have to be educated one community or county at a time. Once these
techniques become more commonplace, we assume that such a great  initial effort will not be necessary.

Lof Size Often Dictated by Septic  Issues

    Wastewaterissuesoften control the  form, location,  and economicfeasibilityof a new  residential subdivision.  In many
parts of the Great Lakes region, heavy clay soils strictly limit the  functioning of conventional septic  systems.  For this
reason, lot size is frequently dictated by septic issues as much as by local zoning. Although there are some  alternate
systems (e.g.,  constructed wetlands and community systems) being piloted and used in the region,  local  health officials
are very cautious about permitting them. This caution arises both  from  concerns about their technical functioning and
about long term maintenance issues.  Communities already feel  burdened by the need to monitor individual septic
systems. They are skeptical about a  homeowner association's ability to reliably maintain a community  treatment system.

    Wastewater treatment issues should be considered  up front in evaluating the feasibility of clustering homes on a
particular site.  If a public sewer does not serve the site, clustering  probably will not work as well (i.e., the individual lot
sizes will not be able to be reduced as much). There is the  possibility of placing the leach fields in the common  property
to increase the overall  open space  percentage.

Weed Hard  Science

    Although there is a  great deal of  national literature detailing watershed-sensitive development techniques,  there  is
not a lot of research documenting the extent of the water quality benefits they provide in the field. The Center for
Watershed Protection (CWP) recognizes that there is  a  lack of water quality monitoring data  that evaluates the
techniques  in varied site conditions and  is working to develop and encourage more  studies. Through consultation with
the CWP, the Northeastern Illinois Planning Commission, and the Wisconsin Department of Natural Resources (WDNR),
we found that funding for long-term monitoring of these  techniques  is scarce.  Without this data, many of these
techniques may be challenged successfully by skeptical  local officials.  With the assistance of Old Woman Creek
Estuarine  Research Reserve, one of our local partners, we  are monitoring the water quality at the Ohio site.  We hope
that they will be able to continue the monitoring after our grants are over. We also are working with the WDNR to secure
funding for long-term  water quality monitoring at the Germantown site.  It is our hope that this information will  continue
to back up  many of the claims of watershed-sensitive  development and that funding will continue to support these efforts.

Lessons Learned about Working with  Communities

Community Initiative

    Without community buy-in and interest in these concepts, even the most enlightened developer is not going to be
able to get a project approved. When we first started this project, we thought that the developers were going to be the
"hard sell." In two out of three of the communities, it actually has been the communities that needed more education.
In the Huron project, the developer was sold on many of the alternative site design techniques until he kept getting
negative feedback from the township board.  This site was chosen because of the commitment of the developer and the
obvious benefits to the surrounding water resources. What was  not realized was how much resistance there would be
in the political arena. At this  project, we  had several informal meetings with local officials prior to presenting a conceptual
plan, but because the process was developer-initiated, they continued to be resistant throughout the process.


                                                     261

-------
Local Official Knowledge  Varied

    Municipal, county, and state officials with similar regulatory responsibilities often have very different views about the
appropriateness of new techniques.  Although there are no  hard and fast rules about who is likely to be more
progressive, disagreements are common and a primary cause of frustration  among  developers.

    As development is now regulated, it is more expensive and  time-consuming for a lay or professional developer to
create a watershed-sensitive development.  The only way for a developer to address this situation is to inform local
regulators and planning officials  about the  project early on,  and to involve them in the process Unfortunately, this
involvement will probably not speed up the process for the individual developer, but after a few  such projects are
launched, we believe  the barriers  for these kinds  of developments will be lowered.

    Getting everyone with a  regulatory or permitting role on a  project involved at the very beginning is  absolutely vital.
If a project that includes techniques that have not been implemented in the region  before gets too far along before all
the regulators and municipal officials are brought in, the "stranger to the deal" can feel left out and derail the project.
Much of this  problem  will be  allayed once a few watershed-sensitive developments are built, but until then,  developers
and regulators pushing for these practices need to make special efforts to get everyone to the table early.  Of course,
this process increases the costs of doing the development initially, but it can keep it from falling apart  after significant
site planning  and related costs are  incurred.

Strong Local Partner is Key

    Throughout this project, TCP acted as a facilitator between  the communities and developers and as a representative
of the silent third party, the environment.  We believe that as each community  begins to look at this type of development,
this third party is key  to the success of a project. Although there are many merits to approaching communities as a
national organization, without a primary  local  partner who is well-versed in the trials and tribulations of the development
process (or willing to learn them), it is difficult to proceed. A preferable arrangement would be a local organization,  such
as a land trust,  leading the effort with support from a regional or national organization or technical assistance center.
A local organization will have a greater vested interest in  and  knowledge of the local  environment, will  know the  local
officials and political and personal histories, and will be able to track and monitor the day-to-day activities surrounding
the development. In the long term, local land trusts may become a key player in this area. They understand land
conservation and watershed issues, frequently have close ties with both local  landowners and local government officials,
and have some comprehension of the  development  industry.

Final Lessons Learned

    Several realities of the development process that have  little to do with the challenges of watershed-sensitive
development are important to mention for groups and communities considering this type  of project. One  is that the
personalities and reputations of the developers can make or break a project.  On our project in Ohio, the  developer
apparently had  a "history" with several  of the plan commissioners. Our partners in the community think that the plan
commission and the engineers were being  unduly unfair during the review process. Also, one of the developers  in
Wisconsin has a reputation for "low-end" development.  Because of this reputation, the Village is  afraid that the
developers will do their typical development in their town.

    Another reality of the development process  is that the Village Planner of Germantown estimated that 60%  of
submitted development plans are reviewed by the plan commission and less than 50% of zoning requests are approved.
All of our projects include a zoning request because the current  local ordinances do not include a provision for
watershed-sensitive development.

Conclusion

    At all three of our model sites there are already signs of new developments being proposed with many of the
watershed-sensitive techniques.  In Wisconsin, the developer was  approached by  neighboring communities to design

                                                     262

-------
similar subdivisions. In Michigan, several local officials have  stated interest in adding  language  in their new ordinances
that would encourage this type of development. In Ohio,  our  local agency partner, Old Woman Creek National  Estuarine
Reserve, was approached  by a developer who has been watching the process and is interested  in using  some of the
techniques at an adjacent site. While the review processes for all  three projects have not been  as easy as anticipated,
it is expected that the next round of developments will have an easier time because of the trailblazing work done before
them.

References

Arendt,  Randall. 1997.  Conservation Subdivision Design. Natural  Lands Trust, Media,  PA.

Center  for Watershed  Protection  (CWP).  1998.  Better Site Design: A Handbook for Changing  Development Rules in
Your  Community. Prepared for the Site Planning  Roundtable, CWP,  Ellicott City,  Maryland.

Center for Watershed Protection (CWP). 1998. Consensus Agreementon Model Development Principles to Protect Our
Streams, Lakes, and Wetlands. Prepared by the Site Planning Roundtable, CWP, Ellicott City, MD.
                                                     263

-------
                 Continuous Deflection Separation (CDS)  for  Sediment
                               Control in  Brevard  County, Florida
                            Justin  Strynchuk, John Royal,  and Gordon  England, P.E.
                                   Brevard County  Surface  Water Improvement
                                               Viera, FL  32940
Abstract

    In July 1997, Brevard County's Stormwater Utility Program  installed  a new type of trash and sedimentation control
device called a Continuous Deflection Separation (CDS) unit. This was  the first American  installation to  use the CDS
technology, which was developed in  Australia. After installation, autosamplers were placed  upstream and downstream
of the CDS unit and  a year's duration of sampling data collected. Monitoring has shown that the CDS unit has provided
an average 52%  removal  efficiency for total  suspended solids  and  31% removal  efficiency  for phosphorus.

Introduction

    Stormwater sedimentation is  a primary source  of pollution to Florida's Indian  River Lagoon. Suspended solids and
turbidity reduce sunlight  penetration in the lagoon which negatively impacts seagrass growth. Where  land is  available,
detention ponds  effectively reduce most of the suspended solids from Stormwater flows. When land is  not  available,
alternative,  less effective, treatment methods must  be  used.

    The CDS technology was initially  developed in Australia to provide an effective method for trash and solids removal
from Stormwater  flows. The screening action  within  the unit provides for 100% removal of trash  and particles down to
4700  microns. In addition,  the unique circular design  creates centrifugal action within the  round  concrete box which
propels suspended solids to the center of the box  and down into the storage chamber.

Methods

    The location chosen  for the CDS  unit installation  was along a ditch at  the north end  of  Brentwood Drive, south of Port
St. John. The drainage  basin for this location was  24.87  hectares (61.45  acres)  in area. This basin has Type A soils
along a sand ridge.  The  land uses  are 24.87  hectares (6.7 acres) of roadway (US Highway 1), 8.04 hectares (19.87
acres) of industrial park, 9.47  hectares (23.39  acres) of vacant land, and 4.65 hectares (11.49  acres) of commercial
property.  The industrial area has a permitted Stormwater system. A  significant land feature is a 2.02 hectares (5 acre)
dirt parking lot, 152.4 meters (500 feet) upstream of the site around a local restaurant. This parking lot has  a steep slope
and is composed of fine white base material.  There  is evidence of heavy  silt buildup in the inlets and pipes downstream
of this parking lot, along US 1.

    There is an earthen  ditch running eastward 76.2 meters (250 feet) upstream from the project location.  At the project
site,  there is  an existing 122 centimeter (48 inch) RCP driveway culvert in the ditch which discharges to a concrete
channel running 152.4 meters (500 feet) eastward to  the Indian River.  The time of concentration to the site is 63 minutes,
with a 1 0-year flow of 1,557.2 L/sec  (55 cfs) and mean annual flow of 1,177.9 Us (38.2 cfs). In Brevard County, the 10-
year storm  is 20.1 centimeters (7.9 inches) of rainfall and the mean annual storm is  13.97  centimeters (5.5 inches) of
rainfall. There is no base  flow at this location.
                                                     264

-------
    A diversion weir 68.58 centimeters (27  inches) tall is  placed in front of the 122 centimeter (48 inch) culvert so as to
divert flows  over 254.8 L/sec (9 cfs) around the unit.  In  18 months of observations, the water level has  risen over the
weir one time.

    A 76.2 centimeter (30 inch) concrete pipe was constructed adjacent to the existing  122 centimeter (48 inch) pipe  in
order to divert flows to the CDS  unit. The  76.2 centimeter (30 inch) pipe enters the CDS  unit tangentially to start the
circular flow within the unit.

    The CDS unit consists of three circular, concrete chambers stacked on top of each other. The top chamber, where
the water  enters the unit,  has a 1.524 meter (5 feet) inner diameter and is  188 centimeters (74 inches) tall. The middle
chamber has a 2.44 meter (8  feet) inner diameter and is 127.54  centimeters (51 inches) tall. In the middle chamber is
a  1.524 meter (5 foot) diameter stainless steel  screen matching  the walls of the top chamber. The screen has 4700
micron holes to filter larger materials. The bottom chamber has a  1.22 meter (4 foot) inner diameter by a  1.22 meter (4
foot) tall sediment sump.

    Water enters  the unit  in a  clockwise rotation. When the water passes through the screen, it flows counter-clockwise
between the screen and outerwall  until it reaches a 76.2  centimeter (30 inch) concrete pipe. This exit pipe is tangentially
placed for smooth exit flows. The elevation  of the exit pipe rises 96.52  centimeters (38 inches) from the lower chamber
to the outflow channel downstream of the 122 centimeter  (48 inch) culvert. This rise in elevation keeps the normal water
level in the  unit near the top of the second chamber at all times.  There is no base flow at this location.

    The top  of the unit is flush  with the surrounding ground and has a 0.91 meter (3 foot) square, lockable, stainless steel
access cover.  This feature allows  for easy access with a vacuum truck for cleaning purposes.

    The CDS unit was installed on July  17, 1997. Installation  took two days with the precast structures.  A large crane
was required to lift the chambers into place. A  4.57 meter (15 foot) deep hole was excavated for the structure.

    In conjunction with the CDS unit installation, County  forces cleaned  the ditch upstream of the unit. Two days later,
a significant rainfall event occurred and 2,294 kilograms (6,600 pounds)  of sediment from the upstream ditch wastrapped
in the unit. After that storm, the ditch was reworked and sod was laid. The  sod greatly  reduced the volume of sediment
washing into the  unit.

    Cleanouts were also performed on November 17,1997, removing 626.84 kilograms  (1,382 pounds) of  sediment  and
2.88 meters (34  cubic feet) of trash and debris,  and again on  May 6, 1998, with 998 kilograms (2,200 pounds)  of
sediment.  The solids removed  from the unit are taken to the Brevard County landfill for disposal.  The volume of water
stored in the unit is greater than the vacuum truck capacity, so decanting is performed on nearby sandy  soils to avoid
a second  trip to the  landfill for disposal.

Evaluation of the CDS Unit During Storm  Events

    The intent of the sampling was to evaluate  the effectiveness of the CDS  unit in removing pollutants  from a storm
event prior to discharging stormwater into the Indian  River Lagoon.

    Five storm samples  were collected at the CDS unit between April 1998 and March 1999. The storm events occurred
after dry periods ranging between 7 and 75 days. Protocol for this program dictated that if the sample collection devices
(autosamplers) were triggered  at intervals of less than three days between storms, the samples were  to  be discarded.
This situation did  not occur during the year, and near-drought conditions  were observed in  the sample area throughout
most  of the year-long monitoring  program.

    Rainfall  was measured at the  sampling  site  by a tipping bucket rain  gauge, and additional rainfall data  obtained from
the Orlando Utilities Commission (OUC) power generating plant 5.6 km (3.5 miles) to the north of the  CDS installation.

    Review  of the rainfall data collected indicates that the majority of the water passing through this BMP was from
precipitation falling on the upland, 18.72 hectare (46.25 acre) watershed. The variation noted in both coverage  and

                                                      265

-------
amount of rainfall  helps illustrate the  localized nature of the  storms  occurring along the  Lagoon coastline.  During  this
sampling period, water flowing off the drainage basin contributed much more flow through the CDS unit than would have
been expected based on the rainfall recorded at the sample  site.

    Samples were collected through the use of automated storm water samplers;  one at the  inlet and  another at the
outlet pipe of the CDS unit. All samples, associated blanks, and duplicates were collected in accordance  with our state-
certified Comprehensive  Quality  Assurance Plan.

    The stainless steel intake strainers for the samples were mounted on the reinforced concrete pipe, slightly off center
bottom, and both angled away from the  flow. This was to prevent the strainers from becoming silted over by sediments
and  allow collection of representative water  samples. Flow rates during the storm  events were measured  initially utilizing
water level meters (ISCO bubbler type) in  conjunction with a 90-degree V-notch weir, but eventually replaced with a
Doppler area-velocity flow meter which  provides a more accurate flow assessment.  Initially, two bubbler meters were
installed with both  bubbler tubes mounted on the upstream weir.  However, this led to difficulties in estimating just when
to trigger  (time delay) the  downstream sampler in order  to collect samples from the same "plug" of water.

    During the first three sample events, water levels recorded were correlated to flow,  and the samples  were manually
composited to give a flow-weighted composite sample from each sampler. Both inlet and outlet sample sets were
composited  identically, in  accordance with  the EPA NPDES  Stormwater Sampling  Guidance Document (July 1992).
Discreet samples were collected for the fourth and fifth  events.

    It was intended  that the  third sample event would  include a mass  balance  calculation. The CDS unit  sump  was
thoroughly cleaned utilizing a VAC-truck to ensure that the material collected  was a result  of the one storm to be
evaluated. Inlet and outlet stormwater composite samples were  again collected, with the  addition of a sediment (Table
1) and water  column sample from the sump. Sediment  depths were measured at  five  locations; four from the corners
of the lid opening and one in the  center. Based on a depth of 13.21 centimeters, a sump diameter of  1.22 meters (4 feet)
and  an estimated 1,410.6 kg/m3 (88 Ib/ft3), (based on previous sediment weight evaluation), approximately 217.3
kilograms (479.2  pounds) of sediment was collected  in the  unit from storm three. Based on  the concentrations
measured, 126.07 grams (4.443 ounces) BOD 5, 33.587 grams (1.184 ounces) of metals, and 122.81 grams (4.33
ounces) of Total Kjeldahl  Nitrogen  (TKN) were removed.

Table 1. Sediment Chemical Analysis For Storm #1
Parameter
Arsenic
Barium
Benzo(b)fluoranthene
BODS
Cadmium
Chromium
Copper
Iron
Lead
Nickel
Silver
Total Kjeldahl Nitrogen
Total Phosphorus
Zinc
Sediment
Grab
0.096
3.4
260
650
0.03
1.1
1.2
220
2
0.4
0.16
450
79
14
Grab Duplicate
0.11
2.9
ND
510
0.033
1.1
0.95
260
2.2
0.36
0.059
680
230
14
Average
Value
0.103
3.15
250
580
0.0315
1.1
1.075
240
2.1
0.38
0.1095
565
154.5
14
Detection
Limit
0.069
0.14
240
2.7
0.014
0.027
.0.027
0.55
0.041
0.069
0.014
37
9.2
0.27
Units
Mg/Kg
Mg/Kg
Ug/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Mg/Kg
Notes:
Equipment Blank Water Yielded ND for all listed analytes.
The benzo(b)fluoranthene mean value was calculated with the
RDL as the lower value for the duplicate.
Only parameters with values above detection limit are listed. Many others were tested below detection limits.
                                                      266

-------
    For this third sample event, the upstream, or intake flowmeter bubble tube was mounted on the 90-degree V-notch
inlet weir, as it was for previous sample events. The  downstream bubbler, however, was moved  and attached to the
downstream discharge  pipe. This  change was necessary to  account for the lag time between when the first sampler
received flow at the beginning of the storm,  the time required to fill the sump with 8,115 liters (2,144 gallons), and
discharge to flow  past the second sampler several minutes later. The problem encountered with this revised setup was
that the upstream V-notch weir used to determine the flow was overtopped,  allowing flow around and over it, preventing
an accurate  flow measurement. This led to  disparity in the estimation of actual flow through the unit. Due to the
questionable  flow measurements, it was not possible to calculate the mass balance.

    For the fourth sample event, an ISCO Doppler area-velocity flow meter was mounted in the bottom of the outfall pipe
of the CDS unit. Upon registering a water level rise of one  inch, this unit  triggered both upstream and downstream
autosamplers. The autosamplers were synchronized, collecting a bottle set  in each ISCO at the same  time. With this
methodology  and  placement, overtopping the weir, flow bypassing, and pressurization were no longer potential sources
of error.  Since the samplers now triggered only when the sump was full,  it was also somewhat  easier to accept the
premise of "what went  in, must have come out."

    Appropriate trigger points were selected in  orderto allow sufficient water depth forthe velocity meter probe to operate
properly. We found  that the Doppler area-velocity flow  meter probes appear to function erratically when covered  by less
than one inch of water,  and believe that measurements taken when the water was at this depth are  suspect.  Two-bottle
sample  sets were collected at  sampler initiation, and at 1 0-minute intervals during the storm. During previous  sample
excursions, samples were manually composited. Due to  a  high  suspended solids content, (heavy particles including
sand) that rapidly settled in the  sample container, it was  questioned whether the composite  samples  were truly
representative of the solids collected.  Therefore, discrete two-bottle sets collected  every  10 minutes were sent to the
laboratory without  being  composited.

    For the fifth sample event, two-bottle  sample sets were again collected at sampler initiation, and at lo-minute
intervals during  the  storm. As with the previous sample event, sample sets were not composited but sent for analysis
as six individual, two-bottle sets. The  sample  bottles for bottle sets six  were not collected due to insufficient water to
cover intake strainers, as the storm was not of adequate duration to produce the last 1  0-minute sample. Because of
numerous problems encountered in the  previous storm event samplings, along with refinements in sampler setup and
flow  measurement, the  fifth storm  sample event is considered the most  accurate to determine what pollutant reduction
is provided by the CDS unit for that storm.   The individual  two-bottle sets showed the variation  in pollutant loadings
throughout the storm event and the  corresponding removal under the varying loads. Unfortunately, this  was the  lowest
flow  storm encountered, which may account for higher than  normal removal efficiency. Maximum flow  was estimated
to be only 136 liters/sec (2.16 gpm). The average pollutant reduction between inlet and outlet samples for  this event was:
BODS 53%, COD 52.6%, TP 36%, TSS 56%, and Turbidity 74.8%.

    Sample results are presented in  Tables  2 through 4 for the five sample  events. Storm event 2 showed  a 23%
reduction in turbidity, but no reduction in the other parameters.   Storm  4 showed an  increase in most parameter
concentrations between inlet and outlet that could not be attributed to resuspension due to a full sump,  since the sump
had  been  cleaned prior to the third event. Data for these two  storms are therefore suspect. For events 1, 3,  and  5, the
average removal efficiencies for those parameters  that showed a reduction were: TSS 52%, Turbidity 46.9%, BOD
34.2%, COD 35%, and TP 30.6%

    After each sample  event, field observations were  made  of the appearance  of  the sample jars, each containing a
water sample that had been collected at progressive ten-minute intervals  throughout the storm flow.  Outlet samples
typically appeared to be less turbid than the corresponding inlet samples, and also  had less sediment on their bottoms.
An observation was also made of the water surface inside the CDS unit proper. There  was typically a thick layer of
floating grass and  other vegetation, an oil sheen,  glass and plastic bottles, plastic sheets and bits, seeds and nuts, sticks,
and  a surprising amount  of Styrofoam cups and particles.
                                                     267

-------
Table 2. Storm #1-#3 Test Results - Composite Samples

STORM 1


CDS Inlet
CDS
Outlet
Change
Percent
Reduction
pH Total
Turbidity
s u Suspended NTU
Solids
mg/l
7.6 220
7.4 110

0.2 100
3% 50%

Maximum flow rate = 5.488 literslsec (87 GPM,
Storm Duration
Rainfall @ OUC

STORM 2


CDS Inlet
CDS Outlet
Change
Percent
Reduction
= 67 minutes
0.254 cm (0. 1 inch), @ SITE
pH Total
s u Suspended
Solids
mg/l
8.4 350
8.2 350
0.2 0
2% 0%

Maximum flow rate = 8.39 liters/sec (133 GPM,
Storm Duration
Rainfall @ OUC

STORM 3

CDS Inlet
CDS Outlet
Change
Percent
Reduction
= 68 minutes
1.778cm (0.7 inch),.® SITED.
pH Total
s u Suspended
Solids(mg/1 )
7.6 300
7.6 150
0 150
0% 50%



180
100

80
44%

0. 19 cfs)

not recorded
Turbidity
NTU


440
340
100
23%

O.Scfs)

0762 cm (0.03 inch)
Turbidity
NTU

110
86
24
21 .8%

BOD5-Day COD
mg/l mg/l


28 150
23 110

5 40
18% 27%




BOD5-Day COD
mg/l mg/l


8.2 20
8.2 20
0 0
0% 0%




BOD5-Day COD
mg/l mg/l

12 71
8.2 53
3.8 18
31 .7 % 25.4

Total
Phosphorous
mg/l

1.4
1

0.4
29%




Total
Phosphorous
mg/l

0.86
0.86
0
0%




Total
Phosphorous
mg/l
1.3
0.95
.35
27%

Maximum flow rate = 149.75 liters/sec (2374 GPM, 5.29cfs)
Storm Duration
Rainfall @ OUC
= 113 minutes
4.064cm (1.6 inch), ©SITE 1

.27 cm (0.5 inch)




                                                             268

-------
Table 3. Storm #4 Test Results - Discrete Samples
Set 1
@ initiation
Inlet 1
Outlet 1
Change
Percent
Reduction/Gain
Inlet 2
Outlet 2
Change
Percent +/-
Inlet 3
Outlet 3
Change
Percent
Reduction/Gain
Inlet 4
Outlet 4
Change
Percent
Reduction/Gain
Inlet 5
Outlet 5
Change
Percent
Reduction/Gain
Inlet 6
Outlet 6
Change
Percent
Reduction/Gain
Maximum flow rate =
Storm Duration = 55
BOD5-Day
(mg/1)
2.1
5.4
+3.3
+61%

6.6
7
+0.4
+6%
6.7
6.7
0
0%

6.3
NT
Na
Na

5.6
6.4
+0.8
+13%

6
6.3
+0.3
+5%

60.30 liters/sec
minutes
Rainfall @ OUC 2.794 cm (1 .1 inch),
COD
(mg/1)
2
2
0
0*%

15
18
-3
+17%
25
24
-1
-4%

45
NT
Na
Na
t
33
30
-3
-9%

39
33
-6
-15%

(956 GPM, 2.13

PH
(SU)
8
7.8
-0.2
-3%

8.3
8.4
+0.1
+1%
8.2
8.3
+0.1
+1%

8.1
NT
Na
Na

8
8.2
+0.2
+2%

7.9
8.2
+0.3
+4%

cfs)

@ SITE 0.006 cm (0.002
Total
Phosphorous
(mg/1)
0.32
0.19
-0.13
41%

1.2
0.94
-0.26
-22%
1.2
1.5
+0.3
+20%

1.6
NT
Na
Na

1.6
1.6
0
0%

1.6
1.5
-0.1
-6%



inch)
Total Suspended
Solids
(mg/1)
690
320
-370
-54%

1400
1600
+200
+13%
830
550
-280
-34%

330
NT
Na
Na

290
170
-120
41%

220
270
+50
+19%




Turbidity
(NTU)
99
120
+21
+18%

1800
1000
-800
44%
530
430
-100
-19%

200
NT
Na
Na

300
260
40
-13%

120
230
+110
+48%




                                                           269

-------
Table 4. CDS Storm #5 Test Results - Discrete Samples

Inlet 1
Outlet 1
Change
Percent
Reduction/Gain
Inlet 2
Outlet 2
Change
Percent
Inlet 3
Outlet 3
Change
Percent
Reduction/Gain
Inlet 4
Outlet 4
Change
Percent
Reduction/Gain
Inlet 5
Outlet 5
Change
Percent
Reduction/Gain
Average Percent
Change
BOD5-
Day
(mg/1)
4.6
4.0
-0.6
13%

10
3.8
-6.2
62%
13
4.7
-8.3
64%

9.9
3.9
-6
61%

9.6
3.4
-6.2
65%

53%
COD
(mg/1)
68
18
-50
74%

51
23
-28
55%
55
33
-22
40%

53
29
-24
45%

53
27
-26
49%

52.6%
PH
(SU)
7.8
7.91
+.1
1 %

7.81
7.9
+.1
1%
8.2
7.6
-0.6
7%

9.2
7.7
-1.5
16%

9.4
7.6
-1.8
19%

-%
Total
Phosphorous
(mg/1)
0.23
0.18
-0.05
22%

0.25
0.18
-0.07
28%
0.3
0.18
-0.12
40%

0.35
0.18
-0.17
49%

0.29
0.17
-0.12
41%

36%
Total Suspended
Solids
(mg/1)
49
11
-38
78%

59
19
-40
68%
23
21
-2
9%

39
15
-24
62%

35
13
-22
63%

56%
Turbidity
(NTU)
16
4.3
-11.7
73%

38
6.9
-31.1
82%
23
12
-11
48%

61
7.2
-53.8
88%

56
9.4
-46.6
83%

74.8%
Maximum flow rate 0.136 liters/sec (2.16 GPM, 0.005 cfs)
Storm Duration =50 minutes
Rainfall @ OUC 1.016 cm (0.4 inch), @ SITE, 0.5842 cm (0.23 inch)
Conclusions

    While none of the sample events were a perfect combination of a good flow and everything working right, the data
collected,  and our observations, certainly indicate that the  CDS unit is  operating as intended and removing significant
quantities of debris and  suspended materials prior to discharge to surface waters. It was quite impressive to prevent this
trash and sediment from washing out into the lagoon  during a  normal rain.

    The phosphorus removals observed for the CDS Unit, as with any BMP of this type, will not have a high degree of
accuracy,  due to  leaching of nutrients from grass, leaves,  and other organic debris.  If there are no base flows, these
leached nutrients  will be washed out with runoff and  skew sample readings. A  much  more comprehensive analysis is
available in the library of the web site www.stormwater-resources.com.

Future Evaluations

    More  data are necessary to further evaluate this  BMP. Due to the inherent  inaccuracies in water quality sampling,
additional determination  of the efficiency of this type of BMP could be made by conducting a  mass loading and sediment
evaluation. Much of the sediment collected in this type of BMP is invisible to current testing techniques since it is
comprised of large particles that roll along the bottom of the pipe. Yet, known quantities of sediment are being collected.
                                                      270

-------
 A previous study of baffle boxes resulted in  the same  conclusion.  Future  sediment analysis from the CDS  unit could
be compared  to  the baffle box data previously collected.  Brevard County will be  conducting a  sediment evaluation at
three baffle box sites over the  next  12  months that will provide additional comparison. As time permits, Brevard  County
will also  collect additional sediment  data from he CDS unit.
                                                       271

-------
       Use of Automated Technologies in Watershed  Management Planning
                           Lake County Stormwater Management Commission (SMC)
                                               Libertyville,  IL
Introduction

    The Lake County Stormwater Management Commission's (SMC) is working with many agencies to develop
comprehensive watershed plans. These  watershed  plans  involve data collection and  collation, problem analysis,
alternative solutions  identification and  action plan  development. The watershed assessment includes  hydrologic and
hydraulic modeling; floodplain and floodway mapping; and water resource assessment. As part of a watershed
management plan, one of the end results is to update floodplain maps and to map depressional storage areas.
Other end products of this effort include location maps of water resources, including wetlands and regional detention
sites,  with identification  of those needing preservation, enhancement or  restoration. With this  information,  projects
can be prioritized and cost estimates determined in order  to assist local  governments  in implementing  the action
plans.

    Lake County, Illinois is located in the northeastern corner of Illinois and is one of the fastest  growing counties in the
country. The county has 61,000 acres of wetlands (12) and  400 miles of streams
and rivers throughout its  480 square miles.  The combination of growth and the
need to protect natural resources is driving the  Lake County Stormwater
Management Commission's  (SMC)  comprehensive  watershed planning efforts.
Plans are currently being developed for  urbanizing  watersheds between 2 and
50 square miles in area.

    With limited  personnel  and  funding, SMC is utilizing  in-house computer
capabilities and staff technical expertise to save time and money  as we increase
our ability to model and display watersheds. The Squaw Creek Plan  is an
example of how SMC is currently utilizing automated  technology for watershed
planning purposes. The  Squaw  Creek watershed  is 25.5  square miles and is
75% undeveloped (includesagriculture, vacant and open space).  The watershed
is  17.3  percent wetlands. TheNorthern  Illinois Planning Commission forecasts
a 155%  population change between  1990 and  2020.  The Squaw Creek
Watershed is located  in the western portion of the  county and drains into Fox
Lake,  on the Fox River.
Map 1: Lake County, Illinois, Sub-Watershed
    SMC is  integrating Geographic Information System (CIS)  (2) technology with  Computer Aided Design (CAD) and
the Army Corps of Engineer's HEC-1 (10) and HEC-RAS (11) models to create an "automated" watershed closely
resembling the existing Squaw Creek watershed characteristics. SMC used a variety of vendor software packages that
include Environmental Systems  Research Institute, Inc.'s (ESRI) ArcView (1) and its Spatial Analyst and Hydrologic
Extensions,  and Bentley's MicroStation.

Data Collection

    It is very important to  determine the methodologies  for  collecting, calculating,  and analyzing data early in the
automation  process. Methodologies were  determined for mapping floodplains,  inventorying  and analyzing  water
resources (8), and estimating runoff water quality. The floodplain mapping variables included time of concentration,
precipitation  runoff, stream storage, stream routing, sub-basin boundaries, and water surface elevations. Thesevariables
                                                   272

-------
had to be determined before final data could be formatted and  collated. We also had to determine how data could be
documented in the  report early in the study process,

    Since considerable map  data was available digitally, it was economical to  perform many  tasks on the computer rather
than on hard  copy. The  Northern Illinois  Planning Commission  (NIPC)  provided the digital  land  use map. Lake  County
Map Services  provided digital copies of the Soil Conservation Service (SCS)  hydrologic soil groups (HSGs) map, hydric
soil map,  United  States  Geological Society  (USGS)  orthophotos, Lake County Wetland Inventory  boundary map, and
Lake  County  parcel boundaries.  In addition to this digital data, SMC contracted to obtain 2-foot topographic contours,
detailed orthophotos,  stream cross-sections, and field-surveyed hydraulically significant  structures. Bridge  and  culvert
information and stream cross-sections were  also delivered digitally from  Illinois Department of  Natural Resource's  (IDNR)
land survey crew using  Global Positioning System  (GPS) and conventional  surveying. Photogrammetry  and cross-
section control points  were collected in the field  utilizing a GPS with accuracy of 1:50,000 horizontal and +/- 0.03 feet
vertical (5). Each USGS digital orthophoto map covers one quarter of a quadrangle and  used 45 MB of computer
storage. The topographic maps weredelivered in CIS  and CAD formats. Contracted  data were delivered by square mile.
This created a reasonable size data file, including:

Two foot  topographic contours and  breaklines                1.2 to 3 MB  per square mile.
Orthophotos                                                 35 MB per square  mile
Digital Elevation  Model                                      1 MB per square mile


    The cost for  the two foot contours overlaid on an orthophoto varied between  $2200 and $3300 per square mile.
Additional record  drawings of hydraulically  significant structures, such  as road crossings  and detention basin outlets,
were  collected from county  and state  highway departments  and local communities. The  townships  and communities
seldom had detailed information, so field  investigations were undertaken, where necessary, using topographic mapping
to establish a  reference  elevation.

    The  water resources inventory included  a stream assessment,  wetland inventory, and a wetland  restoration
assessment. The  stream  assessment data were collected in the  field along with short community interviews. The stream
inventory used an existing methodology created by NIPC (8). SMC created a methodology to identify potential wetland
restoration  locations.

    Surface runoff water quality was estimated using typical  measured  pollutant loading data for several general land
uses. NIPC had  an existing  procedure that assigned non-point runoff pollutant loads  to general NIPC land uses. The
typical pollutant loadings  were entered  into a worksheet  so this procedure could be  automated.

Creating Hydrologic and Hydraulic Data

    Several hydrologic and  hydraulic parameters  and other data were used to analyze the surface water  runoff and
generate floodplain  boundaries. These  included delineating sub-basin boundaries; determining  a runoff curve number,
time of concentration  and Clark's  coefficient of runoff for each  sub-basin; calculating  reservoir data; formatting HEC-1
model; and creating HEC-RAS model  geometry.

Sub-basin  Delineations

    The sub-basin  boundaries were  produced  automatically  using the  following steps. First,  a  Digital Elevation Model
(DEM) was produced from photogrammetry by a consultant. A DEM is a list of equally spaced  data points with a defined
easting, northing,  and elevation. A spacing of the DEM  points  of 10 and 30  feet  was evaluated.  The lo-foot spacing
would slightly increase the accuracy of the automated  sub-basin  boundary's but  it used  ten times  the disk storage as the
30-foot spacing. Therefore, a 30-foot grid DEM was used to determine  the sub-basin boundaries due to storage space
limitations.  Second, the DEM was  loaded into  ArcView and  converted to a DEM grid using ArcView's Spatial Analyst.
Third, the flow paths and the preliminary  sub-basin boundaries were created using ArcView's Spatial Analyst and
Hydrologic Extension  along  with the DEM  grid,  which delineated 180  preliminary  sub-basin boundaries in  2.5  hours.

                                                      273

-------
Fourth, these preliminary boundaries were edited with the digital contour map in the background to better model storage
areas and  road  crossings. This editing entailed splitting  basin  boundaries and joining basins together to produce more
accurate boundary  lines.  Editing was performed on portions  of approximately 50% of the preliminary  sub-basins that
were  automatically created  and ultimately reduced the number of sub-basins. Edited  boundaries  were checked against
hard copy maps  and a field investigation of storm sewers and field tiles. A check of maps and field investigation  identified
three  boundaries that needed additional modifications including the addition of one sub-basin. Finally, ArcView was used
to automatically  calculate each sub-basin's area and a sub-basin identification  was assigned to each of the 140 sub-
basin areas.

Runoff Curve Number

    SMC created a methodology to  estimate precipitation runoff. This required converting SMC defined land use
categories to Soil Conservation Service (SCS) runoff curve numbers (RCNs) (9) using ArcView and Excel (6). RCNs
were  calculated  using the following sequence. First, the 1990 NIPC land use  polygons were converted to SMC  land use
categories based on land cover using 1996 orthophotos as a backdrop. Land cover was divided  into six categories: 1)
impervious,  2) graded grass, 3) natural grass, 4) graded forest,  5) natural forest, and 6)  agriculture. Typically graded
grass and graded forest land cover categories  have increased  runoff compared  with their natural conditions as  soils are
compacted and depressions are removed during  grading. A SMC land use was created for the calibration year of 1996
and for the  model year of 2000. Second, concurrently with  the land use conversion, the digitally  mapped  soil  numbers
were  converted to HSGs using CIS queries. Third, the HSG map was intersected with the  SMC land use categories to
automatically create a land cover map.  Fourth, the  land use categories table and a land cover conversion table  were
joined so there was one RCN for each of the four HSGs.

Runoff Data

    HEC-1  requires specific input data to generate  runoff volumes for each sub-basin. The minimum input parameters
for each  sub-basin were identification,  area, the time of concentration (Tc), Clarks Coefficient of Runoff(R), and weighted
RCN. Sub-basin area was  delineated as previously described.

    The  weighted RCN was determined in two steps. First, intersecting the finalized sub-basin boundaries with the
RCNs boundaries using ArcView. This splits the RCN polygons with the sub-basin boundaries. This calculation took
just twenty  minutes. Then this table  of RCN attributes for each sub-basin was exported from ArcView into Excel where
the weighted curve  number  for each sub-basin was calculated in  one day.

    In addition, each sub-basin requires a length and slope  of travel to generate the Tc and R. To determine the length
and slope,  a line with two points  were needed, one upstream  and one downstream. The  line  represented the direction
of runoff from the farthest ridge to the outlet of  the sub-basin. GeoAnalytics, Inc., a  consultant,  created a program to
automatically generate a distance  point 10% and 85% from the sub-basin outlet along this  digitized line  in 30 seconds.
The point locations along the line were determined by the methodology used to estimate Tc and  R. These points were
queried  individually with the DEM  grid  and checked against the topographic map to determine their elevation, which was
entered into a table, ArcView calculated all sub-basin line  lengths in less than a minute.  The  stream line and its two
elevation points were associated  with  the sub-basin identification throughout this process.  Next,  the sub-basin
boundaries, the associated line, and two points were joined into one table and exported as a database file. This
database file was imported to an Excel worksheet where the slope, Tc,  and R were  calculated for each sub-basin.

Reservoirs

    To model  reservoir routing, the  reservoir volumes were determined using ArcView and the 2-foot digital contours.
The reservoirs consisted of a series of polylines in ArcView after conversion of the CAD contour map. The polylines were
modified  so they were completely  connected and then converted to a polygon in ArcView. This documented the location
of every  reservoir that was  modeled explicitly,  as not all reservoirs could be modeled within the scope of our project.
Second,  the elevation for each contour was entered into a table. The topographic contractor  now performs steps one
and two. After all the elevation polygons were created, ArcView calculated the area of the polygons with  one command.

                                                      274

-------
Finally, the elevation and area tables were opened in an Excel worksheet to calculate the elevation versus storage
relationship.  This  worksheet was referenced  by the HEC-1  formatted worksheet described  in the next section.  Stage
versus discharge  relationship was determined for each reservoir when data was available using  HEC-RAS  or HY-8.

Hydrology  Model Development

    All of this data was combined into one Excel workbook to generate the input needed for HEC-1. The sub-basin data
entry included: identification, area, weighted RCN, Tc, and R. Most sub-basins  also needed  reservoir or stream routing
data. An Excel worksheet was edited with  HEC-1 formatted  column widths so  the data could be  saved into a file that
the HEC-1 FORTRAN program can accept. Sub-basin data were entered  automatically by referencing other worksheets
in the same workbook. Once the first sub-basin referenced the  other worksheets properly, the first formatted sub-basin
data were copied to create another set of HEC-1 data for the next sub-basin. After the sub-basin identification was
entered for this new HEC-1 input data set, the remaining data were automatically retrieved in the worksheet and correctly
formatted, to avoid  data  translation errors.

Hydraulic  Data

    The stream cross-section data were initially generated in MicroStation. Each section was digitized as a series of
connected line segments that were exported to a comma-delimited file of easting, northing, and elevation  which was then
imported into HEC-RAS's "Import/Export Files for Geospatial Data."  The culvert and bridge data had to be coded in
separately.  The channel  stationing  was determined  automatically using Intergraph  In-Roads.  This procedure not only
provided data formatted to be exported directly in HEC-RAS,  but also created a 3D  map of the channel cross-sections
and stream centerline to document the model  spatially using MicroStation and ArcView. The cross-section segments
had to be manually identified for use in the  automated floodplain mapping.

Floodplain  Development

    Stream cross-sections and hydraulic structures were modeled using HEC-RAS to determine the water surface
elevation along the stream. ArcView's Spatial  Analyst Extension or Arclnfo  could be used to delineate the floodplain from
the HEC-RAS output. Final  maps were generated in ArcView.

    The HEC-RAS generated water surface profiles were exported  by HEC-RAS's "Import/Export Files for Geospatial
Data." GeoAnalytics Incorporated, Madison, Wisconsin, imported this data into an Arclnfo project that uses a 1  0-foot
DEM grid. Arclnfo needs a line and an elevation for each cross-section to map the floodplain. The cross-section line
and its identification were created in MicroStation, exported as  comma delimited points, and then referenced into ArcView
to create the cross-section  line.  The line with its cross-section identification was associated with the water surface
elevation. The grid was then "flooded" between  the two cross-sections with  a linear slope between  the appropriate  water
surface elevations. This creates a grid of the flooded area. For each flood profile that was to be mapped, a separate grid
of the flooded area must be completed.

    Reservoirs, such as lakes,  ponds, and depressions, that have their Base Flood Elevations determined using HEC-1,
were  mapped automatically. The storage areas  had  the water surface elevation defined using HEC-1 then the grid was
"flooded" for all points at  or  below that elevation.

    The flooded grids were then converted to  polygons and "smoothed" in Arclnfo for use in ArcView. Last the polygons
were  reviewed against the digital two-foot contours and adjusted as needed before final map production. Every reservoir
outlet had to be  manually mapped  between its outlet and the downstream reservoir or stream floodplain.

Water Resource Assessments

    A water resources inventory was  completed that included a stream assessment, a wetland inventory, and potential
wetland restoration site identification. All of the stream assessment data were collected in the field  along with short
community interviews and entered  into  a database.  Several  key stream  characteristics were  mapped  using CIS. The

                                                     275

-------
stream inventory data were queried for specific stream conditions and key characteristics were mapped such as degree
of bank erosion or sediment accumulation.

    A county wetland Advanced Identification (ADID) study was completed in 1992 prior to the assessment. One of the
criteria reviewed for each wetland was its storage potential, which was related to the area of the wetland. Querying the
spatial data for specified wetland areas  and creating a  new set of data easily identified these wetlands. The wetland
restoration  and mitigation bank site identification methodology was developed by SMC  in 1999. Several data sets were
queried to  identify  the former wetland sites that have the greatest number of characteristics necessary to  make them
restorable and usable as a wetland bank. A less stringent set of criteria was used to define all  former wetland sites with
restoration  potential.  The potential wetland restoration sites included all Advanced Identified (ADID) wetlands.  Potential
wetland banking sites excluded ADID wetlands and  restorable sites less than 20  acres.

    Surface water quality 'hot spots" were estimated using  non-point  pollutant  loading rates  for several general  land uses
via NIPC methodology.  Twelve pollutants were  evaluated. The pollutant  "hotspots" analysis  employed land  use,
impervious  surface area, annual runoff coefficients,  and  storm sewer conditions as surrogates to  determine the annual
pollutant loading by sub-watershed. The  pollutant loading database was attached to the land use  map database. It was
then mapped in ArcView resulting in 12  maps, one for each pollutant.

    The watershed advisory committee and NIPC identified which level  of pollutant loadings should be labeled  as
detrimental.  The pollutant load data were then grouped,  using  natural breaks in the data set,  as  low, medium or high.
These were mapped and queried to determine where water quality enhancement projects would be most beneficial and
highest priority.

Summary

    The Lake County Stormwater Management Commission has invested  a  significant amount of time and funding in
developing  the hardware, software,  and database necessary to  perform floodplain analysis. By making this commitment
and  establishing the  methodologies for  manipulating data and analyzing  watershed  parameters, we have created  a
powerful analysis tool. Mapping  accuracy, display flexibility and a wide range of CIS  analysis ability has been  created
through this process. This technology coupled with  other resource assessment efforts  has  created a strong foundation
for future watershed planning in this watershed. The technology  is transferable and will  be used throughout Lake County
as our agency resources allows.

References

Environmental Systems Research Institute, Incorporated  (ESRI) ArcView, 1995, "Using  ArcView CIS 3.0a" and  "Spatial
Analyst"

Environmental Systems Research Institute, Incorporated (ESRI) ArcView, 1996, "Using the ArcView CIS Spatial Analyst"

Lake  County Stormwater  Management  Commission, 1994, Watershed Development  Ordinance

Lake  County Stormwater Management Commission,  1998,  General Conditions of Topographic  Mapping

Lake  County Stormwater Management Commission, 1997, Attachment A,  Field Survey and Data Collection

Lake  County Stormwater Management Commission  and GeoAnalytics, Inc., Madison, Wl  53704, 1998,  Lake County,
Illinois Stormwater  Management Commission, CIS Data Handling Procedures for  Input into HEC-1

Lake  County Stormwater Management  Commission, Land Use to Land Cover Conversion Methodology

Northern Illinois Planning Commission, 1994, "Methods and Procedures for the Stream Condition  Form," from Appendix
A of Sequoit Creek Watershed  Management Project.
                                                     276

-------
Soil Conservation Service, Washington, DC, 1985, National Engineering Handbook Section 4 - Hydrology, PB86780494

U. S.  Army Corps of Engineers, 1990 "HEC-1  Flood  Hydrograph  Package User's Manual Version: 4.0"

U. S.  Corps of Engineers,  1997 "HEC-RAS 2.0 River Analysis System"

U. S.  Environmental  Protection Agency, Region  5,  Lake  County Stormwater Management Commission, and Northeastern
Illinois  Planning Commission, November, 1992, "Wetland  Advanced  Identification (ADID)  Study, Lake County, Illinois"
                                                    277

-------
              Sediment and Runoff  Control  on Construction Sites  Using
                     Four Application  Methods  of  Polyacrylamide  Mix
                                               A.  Roa-Espinosa
                                            Urban   Conservationist
                                  Dane  County Land  Conservation Department
                                          Assistant Visiting  Professor
                                        Biological  Systems  Engineering
                                        University of  Wisconsin-Madison

                                           G.D. Bubenzer, Professor
                                        Biological  Systems  Engineering
                                        University of  Wisconsin-Madison

                                    E. S. Miyashita, Former  Project Assistant
                                        Biological  Systems  Engineering
                                       University of  Wisconsin-Madison4
Abstract
    Fifteen small bare plots (1  meter x 1 meter) on a 10% slope were analyzed for runoff and sediment yield on a
construction site. A rainfall simulator applied 6.32 centimeters of rainfall per hour to each plot after a polyacrylamide mix
(PAM-mix CFM 2000*) treatment was applied. The following treatments: No PAM-mix applied to dry soil (control), PAM-
mix in  solution applied  to dry soil, dry PAM-mix application to dry soil, PAM-mix in solution with  mulch/seeding applied
to dry soil, and PAM-mix  in solution applied to moist soil.  Each treatment was repeated on three  plots. When a solution
of PAM-mix with mulch/seeding  was applied to dry soil and compared with the control (no PAM-mix  application to dry
soil), we  found an average reduction of  93% in sediment yield. An average reduction  of 77% in sediment yield was the
worst performing PAM treatment, and occurred when PAM-mix in solution was  applied to moist soil. The application of
dry PAM-mix  to dry soil reduced sediment by 83% and decreased runoff  by 16% when compared to the control.  Our
results  show that regardless of the application  method, PAM-mix was effective in  reducing  sediment yield in the test plots.
The  ease of application,  low maintenance, and relatively  low  cost associated with PAM make it a practical solution to
the costly methods  being implemented  today

Keywords:

Soil  binders,  soil erosion, polyacrylamides,  flocculation, infiltration, water  retention.

*Use of a product name  is for the convenience of the reader and does not imply endorsement by the authors, Dane
County Land  Conservation, or the University of Wisconsin.

Introduction

    One  effect of  rainfall is the initiation  of the erosion process where individual raindrops  fall and impact the  soil  surface.
Soil detachment and  particle transport by raindrop splash can  lead to serious soil deterioration. Once soil is eroded and
transported by surface runoff to  lakes, rivers,  and streams, a degradation of the aquatic  habitat occurs. Sediment is the
largest pollutant, by volume, in the State of Wisconsin (WDNR, 1994). In  order to maintain a healthy watershed, it is
critical  to  control erosion  and sediment  yield.

    Maintaining soil  structure and  aggregate stability helps control  erosion by increasing  infiltration and maintaining less
erodible-sized aggregates. Stable soil structures also help  maintain a healthy environment.  The use of polyacrylamides

                                                     278

-------
and polyacrilamide mixes (PAMs and RAM-mix) is a new tool to help maintain soil aggregate stability and reduce erosion
caused by surface runoff. Such materials can be applied at a rate of 20 to 30 Ibs per acre on construction sites, to
stabilize such  sites against erosion until  they  can be permanently protected through  vegetation.

    Water-soluble polymers and water polymer mixes do not create aggregates when applied to soil. However, they can
stabilize existing aggregates when the aggregates are saturated with a  solution of water soluble polymer mix.  Increasing
the aggregate  stability with polymers reduces the effect of raindrop impact on the soil, thereby reducing erosion. Polymer
application  to the soil may also retard surface  sealing, reduce particle  soil detachment, reduce sediment in suspension,
and compensate for low residue.

Objectives

    The objectives of this study are to determine the optimum application methods and  the effectiveness of the PAM-mix
under moist and dry soil conditions. The different application methods were  applied to a construction site in  Middleton,
Wl. Data were collected to determine the most effective method of application and the  effectiveness of the PAM-mix on
construction sites.

Literature Review

    The use of polymers as soil conditioners has been studied for decades. The most conclusive studies, done by Lentz
et al.  (1992),  determined that negatively charged PAM is an excellent soil erosion deterrent for furrow irrigated fields.
It was found to be a cost-effective and  safe technology. Sojka and  Lentz (1994) found that PAM, when applied in
irrigation waters at rates greater than 0.7 kg/ha, reduced furrow erosion by an average of 80 to  90% and  increased
infiltration on Portneuf silt loam by an average of 15%. Trout et al. (1995) reported  a 30 to 110% increase in cumulative
infiltration. Roa et al. (1996) found that soils treated with PAM had  infiltration  volumes more than  double that of untreated
soils over a two-hour period.  The  infiltration  volumes for  the untreated soils averaged 231 ml/38.5 square centimeters
while those for treated soils averaged 490 ml/38.5 square centimeters, or 98%  of the  volume of water to be infiltrated.
Roa et al.  (1996) also found  that the high infiltration rate of the treated sample was  associated with low concentration
of sediment in the effluent or infiltrated water.

    Nadler et  al. (1994)  found that  PAM mobility  in sandy loam,  as well as clay loam soils, was limited to the  top 25 cm
10 months after application.  Clays  were attached to anionic polymers  more easily when  salts were present in  solution.
With anionic polymers,  flocculation was  easier  and  more complete.  When polysaccharides are present with anionic
polymers in solution, fixation was also easier and more complete.  Khamraev et al. (1983) reported that clay fixation is
best achieved for PAMs with 30% anionic charges. The cementation provided by the clay flocculation stabilizes the
aggregate at the surface. Roa et al. (1997)  found that using  polysaccharides, a calcium source with  anionic polymers
orpolysaccharides with calcium nitrate  and anionic PAM, increased the  infiltration rate in saturated cores 5 times greater
than with no soil treatment.

    PAM use  for erosion control provides a  potent environmental benefit by halting furrow erosion  by about  half a  ton
of soil per ounce of PAM used. PAMs remove  most sediment, phosphorus, and  pesticides from return flows, and greatly
reduce return flow  BOD (Sojka and Lentz,  1996). The consequences of  reducing sediment and nutrient  loading of
construction areas can ultimately be expected to reduce the frequency and intensity  of algae blooms  and reduce turbidity
and sedimentation of stream channels.

    Lentz et al (1992) in Kimberly, Idaho, reported that when applied at  10 ppm, PAM provided a 94% reduction in runoff-
sediment in three years  of testing.  When used properly, PAM  has no  measurable toxicity to humans, plants,  or aquatic
organisms.  Molash  et al. (1997)  state that the Polyacrylamide  Allocation Standard for Reduction of Soil Loss is
necessary because other best management  practices (BMPs) are available  and have  varying degrees of effectiveness.

    Sojka and Lentz (1996)  summarized several advantages of PAMs over other erosion  control BMPs: (1)  PAM can
be applied  using irrigation equipment and  can  be effective for controlling erosion over large areas,  as demonstrated in
eastern Washington  and Idaho; (2) PAM is  very effective on fine silt/clay soils; (3) preliminary research conducted in

                                                      279

-------
Kansas and  California has indicated that  RAM is effective at  abating wind erosion; (4) RAM enhances precipitation of
fine  silts and  clay particles, providing water quantity  benefits; (5)  RAM increases  soil infiltration capacity that reduces
runoff volumes; and  (6) high benefit to cost ratio.

    The types of RAM used for erosion control should have an approximate molecular weight of 12-1 5 Mg/mole, with
an 8-35% negative charge density, and contain no greater than 0.05% Acrylamide monomer (Sojka and Lentz  1996).

    A recent study done by King et al. (1996) focused on  comparing the  uses of polyacrylamides and  straw  mulch  on
dry bean yields. It was shown that the sediment loss was  reduced for both  straw mulch  and the RAM  treatment.

     In three  years of studies in construction  sites using RAM for controlling soil loss,  RAM has provided a 60-97%
reduction  in runoff-sediment (Roa  et al. 1997).

Method

    Five treatments were applied to soil  test plots: (1) No RAM-mix application to dry  soil [control],  (2) RAM-mix in
solution applied to dry soil, (3) Dry RAM-mix application to  dry soil, (4) RAM-mix in solution with mulch/seeding applied
to dry soil, and (5) RAM-mix in solution applied to moist soil. Three replications of each treatment were  performed using
a randomized  block design on  1 m x 1 m non-vegetated plots in the Middleton Hills Development,  Middleton, Wl. The soil
was  a Dodge silt loam.  The average slope of the test site  was 10%.

    Plot preparations included  large boulder, cobble, and  excess debris removal. The surface was raked priorto testing.
Soil  moisture prior to testing was about 9%.

    The RAM-mix is a high molecular weight  anionic granular polymer. The  RAM-mix (2.25 g of RAM-mix added to 5
liters of water) was applied at a rate of 22.5 kg/ha, to the appropriate plots using a garden sprinkler. For the dry RAM-mix
application, 2.25 grams of the  RAM-mix was applied using a sifter.  For the RAM-mix applied to moist soil treatment, the
soil was pre-moistened by a 6.4 cm rainfall six hours before testing.

    The sprinkler infiltrometer (Bubenzer and Patterson,  1982 ) was used to collect data for this study. A rainfall
simulator was used that  produces 6.4 cm per hour. Actual rainfall depths were recorded using eight rain gauges for each
replication.  Runoff from each plot was collected into a tank where the depth of the water was recorded at approximately
2-minute intervals during each test. The average trial  time was 40-50 minutes or until  the runoff collection tankwasfilled.

    Runoff samples  were extracted at approximately 1 0-minute intervals  by diverting  runoff into a collection  container
during each  replication to  determine sediment yield. A representative sample was also taken at the end of each
replication from the tank.  The samples were dried at 11 0°C for 24 hours and weighed to determine an average sediment
load  for each trial.

Results and  Discussion

    Mean sediment yield, infiltration, and runoff depth  for the three  replications and  the controls are presented  in Tables
1, 2, 3 and 4. For Replication  1, the RAM-mix solution was prepared the evening before  field testing. It was noted that
the viscosity of the solution  decreased  throughout the day.  This change may have been due to UV  light, reaction with
the mix, and/or oxidative and photolytic interaction.  Thereafter,  the  solution was prepared  immediately before the  rainfall
simulation. After analyzing the  results, a lower viscosity of the  RAM-mix solution  was determined to be  less effective in
controlling sediment  yield. This difference  is presented in Table 4.  Future recommendationsforcommercial  applications
may  need to take into account the time of preparation of the  solution and handling before application.

    During the first replication of testing, the largest sediment reduction occurred when RAM-mix in solution was applied
to moist soil. The control yielded 184.4 grams per square meter and the RAM-mix in solution applied to moist soil yielded
36.4  grams per square meter resulting in a reduction  of 80% in sediment  yield  (Table 1). The sediment yield  reduction
for the  dry RAM-mix application to dry soil  and RAM-mix in solution with  mulch/seedling applied to  dry soil were


                                                      280

-------
Table 1.  Summary of rainfall, infiltration runoff, and sediment yield for Replication #1 .
       Treatment

Control
Dry PAM-mix/Dry Soil
Solution  PAM-mix/Dry
Soil
Solution PAM-mix/Moist
Soil
Solution PAM-
mix/Mulch/Dry  Soil
    Rainfall
     (cm)
     5.64
     5.79
     5.64

     5.72

     5.72
    Infiltration
      (cm)
      1.70
      1.91
      1.60

      0.05

      1.57
  Runoff
   (cm)
   4.01
   3.89
   4.11

   5.66

   4.14
             Sediment
    184.4
    68.3
    103.7

    36.4

    67.3
   Soil Loss    Runoff Rainfall
 % of Control    % of rainfall
    100%         71%
    37%          67%
    56%          73%
     20%

     36%
                 99%

                 72%
approximately 64%. The sediment yield for RAM-mix in solution  applied to dry soil was reduced by 44% when compared
to the control.

     In  Replications #2 and #3,  the lowest sediment  yield  occurred for the treatment of RAM-mix in solution with
mulch/seeding applied to dry soil. A sediment reduction of 97% and 89% occurred,  respectively. A sediment reduction
for the  treatment of RAM-mix in solution applied to dry soil was 87% and 57% respectively (Table  2 and 3).
Table 2. Summary of rainfall, infiltration runoff, and sediment yield for Replication # 2.
       Treatment

        Control
Dry PAM-mix/Dry Soil
Solution  PAM-mix/Dry
Soil
Solution PAM-mix/Moist
Soil
Solution PAM-
mix/Mulch/Dry Soil
   Rainfall
    (cm)
    4.57
    5.72
    4.72

    4.14

    4.55
    Infiltration
      (cm)
      0.51
      1.57
      0.61

      0.13

      0.38
  Runoff
   (cm)
   4.06
   4.14
   4.11

   4.01

   4.17
                                                                 Sediment
   377.67
   178.36
   48.77

   242.4

    12.04
 Soil Loss
% of Control
   100%
   47%
   13%

   64%

   3%
              Runoff Rainfall
                % of rainfall
                  88%
                  73%
                  87%

                  97%

                  92%
Table 3. Summary of rainfall, infiltration runoff, and sediment yield for Replication # 3.
Treatment

Control
Dry PAM-mix/Dry Soil
Solution   PAM-mix/Dry
Soil
Solution  PAM-mix/Moist
Soil
Solution PAM-
mix/Mulch/Dry  Soil
Rainfall
(cm)
     5.05
     5.38
     4.50

     4.42

     4.39
Infiltration
(cm)
      1.12
      1.96
      0.61

      0.28

      0.38
Runoff
(cm)
   3.94
   3.43
   3.89

   4.14

   4.01
Sediment
(gm)
   231.34
    43.29
    98.59

    47.65

    26.58
Soil Loss       Runoff Rainfall
% of Control    % of rainfall
    100%           78%
     19%           64%
    43%           86%
    21%

    11%
                 94%

                 92%
    When Replication #1 is excluded from the results, the average sediment reduction for RAM-mix in solution with
mulch/seeding applied to dry soil increased from 87% to 94% (Table 4). The sediment reduction for RAM-mix in  solution
applied to dry soil was 76%. For dry  RAM-mix applied to dry soil,  the sediment  reduction was 17%, and the sediment
reduction of RAM-mix in solution  applied to  moist soil was 77%.

Conclusion

    Our results show that,  regardless of the application method, RAM-mix was effective in  reducing sediment  yield in
the test  plots. The most effective  method of soil treatment throughout this study in  reducing  sediment yield  is RAM-mix
in solution  with  mulch/seeding  applied to dry soil. The  ease of application, low maintenance,  and relatively  low  cost
                                                          281

-------
4.01
3.81
4.04
4.60
4.11
264.51
96.65
83.71
108.82
35.32
100%
37%
32%
41%
13%
304.51
110.83
73.68
145.03
19.31
100%
36%
24%
48%
6%
79%
66%
81%
97%
84%
associated with RAM-mix makes it a practical solution to costly existing methods being implemented. The evidence from
the field application in this study reflects that RAM-mix is a tool to reduce soil loss on bare soil until vegetation cover is
established.

Table  4. Average summary of rainfall, infiltration  runoff, and sediment  yield for Replications 1, 2, and 3 and Replications 2 and 3, excluding
Replication 1.

       Treatment           Runoff     Sediment      Soil Loss %    Sediment (Gm)     Soil Loss %     % of Rainfall
                          (cm)        (gm)        Replication         Excluding     Replication 2 And 3
                                                1, 2, and 3      Replication 1
Control
Dry PAM-mix/Dry Soil
Solution PAM-mix/Dry Soil
Solution PAM-mix/Moist Soil
Solution PAM-
mix/Mulch/Dry Soil


    The  primary factor that  must be considered in future  studies is the time of polymer solution  preparation  and
application. It was noted that the optimal application procedure is to prepare the solution immediately prior to application.
This procedure is necessary in order to limit the amount of degradation and maximize the performance of the RAM-mix.

References

Bubenzer,  G.D.,  and Patterson, A.E., Intake Rate: Sprinkler  Infiltrometer,  Method of Soil Analysis, Part 1, Physical and
Mineralogical Method, Second Edition, Chapter 33, pp. 845870. (Agronomy Monograph Series #9,  1982)

Khamraev, S.S., Artykbaeva,  K., Nizamova, F. and Akhmedova, V.  (1983): Leaching of salts from takyr soils conditioned
by  amide-containing polyelectrolytes. Uzb. Khim. Zh.  4,16-20 (Russian). Chem. Abst. 99, 138730 (1983)

King,  B.A., B. Izadi, MS.  Ashraf, R.H. Brooks,  and W.H. Neibling.  1996. On-farm comparison of polyacrylamide and
straw mulch on dry bean yields,  irrigation performance and erosion.  University of Idaho Publication No.  101-96, pp.5559

Lentz, RD., I. Shainberg,  R.E.  Sojka, D.L. Carter 1992.  Preventing  irrigation  furrow erosion with small applications of
polymers. Soil Sci. Soc. Am. J. 56:1926-1932.

Lentz, R.D and. Sojka, R.E.  1996.  Five year  Research Summary  Using PAM  in Furrow Irrigation.  University of Idaho
Publication No. 101-96, pp.  1 I-20.

Nadler, A.,M. Magaritz, and L. Leib. 1994. PAM application techniquesand mobility in  soil.  Soil Science  158(4): 249-254.

Roa,  A., 1996. Screening of Polymers to Determine Their Potential Use in Erosion Control on Construction Sites,
University of Idaho Publication  No. 101-96,  pp.77-83.

Roa, A.,  Bubenzer, G.D, and Miyashita, E.,  1997.  Determination  of PAM Use  Potential in Erosion Control. Proceedings
of the  first European Conference of Water and the Environment.  Innovation Issues in Irrigation and Drainage. Edited by
Pereira et  al.  Institute  of Agronomy, Technical University  of Lisbon,  Portugal.

Molash, E. 1997.  WSDOT Stormwater Management Plan Page  148 V 5.3 3.25148

Trout. T.J., Sojka, R E. and R.D. Lentz.  1995. Polyacrilamide effect on furrow erosion and infiltration. Trans. ASAE. 38(3):
761-765.

Sojka, R. E. and Lentz, R.D., 1994. Polyacrylamide (PAM): A New Weapon in the Fight against Irrigation Induced
Erosion,  USDA-ARS Soil  and Water Management  Research Unit  Station Note #01-94.
                                                      282

-------
Sojka, R.E. and Lentz, R.D., 1996. A RAM Primer: A Brief History of RAM and RAM-Related Issues, University of Idaho
Publication No. 101-96, pp. 1 I-20.

Stephens,  S.H., Final Report on the Safety Assessment of polyacrylamides, J. Am. Coll. Toxicol., 10:193-202.

Wisconsin  Department of Natural Resources  (WDNR), 1994. The Wisconsin Quality Assessment Report to Congress,
PUBL-WR 254-94-REV, Wisconsin Department  of Natural Resources, Madison,  Wl.
                                                   283

-------
                   Construction  Site Planning and Management  Tools
                                   for Water Quality  Protection

                                              Thomas  Mumley
                                California  Regional Water  Quality Control Board
                                San Francisco Bay Region,  Oakland,  California


      California is seemingly a developer's paradise.  Population is on the  rise, the  economy is good, and there is little
or no rain to interfere with construction for nearly eight months of the year. To top off these benefits, the California
Regional Water Quality Control Board, San Francisco Bay Region (Regional Board) has a comprehensive Construction
Site Planning and  Management Program (Program).   It is based on the integration of a strong regulatory and
enforcement posture, an outreach and education  strategy, and technical  assistance. The  keys  to the success of the
program  are the balance of actions among these elements and  implementation tools for actions within  them.

Background

      The Regional  Board is the state agency in California  responsible  for protection of water quality and enforcement
of water  pollution  control  regulations,  including National Pollutant  Discharge  Elimination  System  (NPDES) permits. The
California Water Code  provides the  Regional Board with strong enforcement authority. This  authority  ranges from  a
notice to comply,  to a  notice  of violation, to  enforcement orders, to monetary penalties.  Penalties  can  be as high as
$10,000 per day of  violation or $10 for each gallon  of waste  discharged.  The Regional Board may also suspend part
of a penalty in exchange for  an environmentally beneficial project.

      In the San Francisco Bay Region, the Regional  Board is responsible for enforcement of  a  general  NPDES permit
for stormwater discharges from construction sites of five acres or greater. The general permit requires implementation
of an effective  Stormwater Pollution  Prevention Plan  (SWPPP)  that includes best management practices (BMPs) for
erosion  and sediment prevention  and control and management  of equipment, materials, and  wastes.  The  Regional
Board is  also responsible for enforcement of NPDES permits for municipal stormwater discharges that have been issued
to all municipalities (regardless of population) in the urban areas of the region. These permits include requirements to
control discharges from construction  sites (regardless of size). There is  an inherent overlap of Regional Board and
municipality  authority over construction of five acres  or greater.  The Regional Board's  Program recognizes and takes
advantage of this overlap of authority.

Inspections

      The Regional Board initiated an aggressive construction  site inspection and enforcement effort in 1997. This
resulted in  discovery of significant water  quality problems associated with sediment discharges  caused  by minimal or
token erosion and sedimentation control  actions. Some of the most common observations were:

      .  No permit.

      .  SWPPP not developed, not implemented, or deficient, especially in terms  of timing.

      .  Mass grading  allowed to continue throughout winter months until rain and muddy conditions make further work
        impossible.

      .  Mass grading continues past the time when grasses will grow and  germinate; first rains simply carry seed/mulch
        away with eroded soil.

      • No erosion  control measures;  reliance solely  on sediment basins.

      .  Sediment basins are  frequently  undersized,  improperly  designed, and not  maintained.

                                                    284

-------
      .  Site  not monitored to assess BMP  effectiveness.

      .  SWPPP not updated to reflect changes in site conditions.

      .  Hillsides stabilized with  hydroseed,  but  no  mulch (resulting in rains carrying  seed material away with  eroded
        soils).

      .  Control measures driven by "tokenism"  with control measures intended to demonstrate good intentions rather
        than  real  effectiveness.

      .  Willingness to risk fines in  order to  maximize work during winter (rainy season) months.

      .  Local agencies, specifically planners  and engineers with plan-approval authority,  often unaware of "best"
        management  practices.

      .  Sites approved by local authorities for  mass grading  during rainy  season.

      .  Local authorities  review  and approve erosion control  plans but do  not inspect sites.

Enforcement Actions

      Several types of enforcement actions evolved from these findings.  The first  consisted of the  development and
issuance of a "Notice to Comply" (Figure 1). Often (25 - 35 % of the time) operators at a site are unaware  of their
requirements or appropriate BMPs.   The  Notice to Comply is essentially a "fix-it" ticket that results in no  further
enforcement  action if corrective  action is  implemented.  Regional Board inspectors are authorized  to issue Notices to
Comply in the field, and  use of this simple  enforcement tool has proven to be an effective mechanism to gain timely
corrective  action at  construction sites.

Other enforcement tools are used in circumstances where the severity of the problem warrants  more intensive
enforcement  action.  These include,  in terms of  progressive severity:  a Notice  of  Violation, a Cleanup  and Abatement
Order,  and a Cease and  Desist  Order. Violations of any of these actions typically  lead to more aggressive enforcement
action.  The  most  aggressive enforcement action is imposition of administrative civil  liability (monetary  penalties).

      During  the 1997/98 rainy season the Regional Board imposed over $1 million in penalties, ranging from $10,000
to $230,000.  A major consideration  in determining penalty amounts is ensuring that it does not pay to  pollute. Due  to
the economic and  time pressures associated with many development projects, minor penalties may simply constitute
a cost of doing business. The  Regional  Board  has clearly stated its intolerance to this circumstance  and intends to
severely penalize  repeat  offenders.  Clearly,  such penalties  not  only get the attention of the  violator, but the building
industry as a whole. Substantive penalties have also provided  opportunities to fund  environmental education projects
in lieu  of direct payment  of penalties.  The Regional Board has favorably accepted development of technical assistance
tools as appropriate mitigation  projects.

Education and Outreach

     The Regional Board recognizes that regulation without education is ineffective. Consequently, its program includes
extensive  outreach efforts. These  include:

     .  Mass mailing to  construction projects of more than five acres and projects permitted for  winter grading
        summarizing  requirements  and  findings on inadequate  performance

     .  Meeting with development community and local agencies prior to the rainy season (August through September)
        to better communicate  concerns  and requirements  and to establish a  dialogue

     .  Providing detailed guidance and training for both developers  and municipalities on their responsibilities and  on
        effective control  approaches


                                                      285

-------
                               San Francisco Bay Regional Water Quality Control Board
                        1515 Clay Street, Suite 1400, Oakland, CA 94612 /Phone (510) 622-2300 - FAX (510) 622-2460
You are hereby notified thaL
Order No.	
a  NPDES Permit No. (if applicable).
a  California Water Code Section  _
• I Other	
              NOTICE TO COMPLY

             	(hereafter Discharger) has violated provisions of:
Federal, State, and Local Agency Contacts:
I. FACILITY INFORMATION
Inspection Date: 	
Discharger Contact:	
Site Name & Location:	
Headquarters/Owner Name & Address :
Time:
                Title:-
 Prior Notification: a Yes   a  &a    a  Unknown
	  Phone: (	)	
                                          County:.
II. NON-COMPLIANCE  INFORMATION

Nature of Violation :
                  Recommendation to Correct:
                         Time to Comply (Not to exceed
                         30 days)
III. SIGNATURE SECTION

I  acknowledge receipt of this Notice (must be owner, operator, or duly designated representative of facility):
RECIPIENT NAME (print):	      TITLE: 	
SIGNATURE:  	      DATE:	
STAFF NAME:	
SIGNATURE:	
IV. CERTIFICATION OF COMPLIANCE

Sign and return by mail or fax within 5 working days of achieving compliance

I  certify under penalty of perjury that the above violation(s) have been
corrected.
I  am aware that there are significant penalties for submitting false information.
                                     PHONE:
                                     DATE:	
Recipient Signature:
Print Name:	
                                                  Date:
                                               FOR REG. BD. USE ONLY

                                           Receipt Date:    Acceptable:
                                              /   /        a  sn
                                           Reviewed by:    D Yes
                                                          Recommendation:
                                           Date:
          Title:
Figure 1. Notice to Comply
    The objectives of these outreach efforts are:
                                                        286

-------
    . Commitment from the construction industry to include erosion  control  in  their  planning, scheduling, and (most
      importantly)  project  implementation
    . Commitment from municipalities to play a  greater role in SWPPP review  and  implementation, including training
      inspectors so that builders, municipal staff,  and Regional  Board staff are all on the same page - thereby allowing
      for consistent  regulation of construction activities  by applying  a uniform standard.
    In response, the building industry and  municipalities have  collaborated with the Regional Board on the production
of training  workshops  on  construction site planning  and  management for  both building industry and municipal staffs.
The workshops provide a  review of  regulations  and responsibilities including:
    . State  responsibilities
    .  Permits for work in  or near streams
    . Local  agency  responsibilities
    . Plan  approval  authority and  requirements
    . On-Site responsibilities (plans, permits,inspections)
    . Inspector responsibilities
    . Enforcement
    . Field inspection  coordination  (i.e.,  state  agency/municipality)
    The workshops also include training  on BMPs for  erosion  and sediment control  (principles, tools, corrective
measures,  inspections,  monitoring, reporting),  non-stormwater discharge  prevention and management, and  a  field  trip
to an active  construction site where vendors demonstrate both proper and improper installation  practices.
    Production of the workshops has  been funded in part through  mitigations associated with administrative civil liability
fines. Similarly, penalty mitigation funds  have been  used to develop  education tools  including:
    . An 18 minute training video entitled "Hold on  to Your Dirt: Preventing Erosion from Construction Projects" which
      provides information  on BMPs  for grading  projects  and for  stabilizing disturbed  land
    . An 18 minute training video entitled "Keep it Clean:  Preventing  Pollution from Construction Projects," which
      provides information  on BMPs to prevent water pollution from non-stormwater discharges from activities such as
      painting, stucco, concrete  washout facilities and saw cutting
    . A booklet of "Guidelines for Preparing a Storm Water Pollution  Prevention Plan"
Erosion  and Sediment Control Field Manual
      The  centerpiece of the Regional Board's Program is an Erosion and Sediment Control Field Manual (also
developed with  penalty mitigation funds). The Field Manual was produced in  response to a common complaint  by "field"
personnel that there  is a need for information on cost-effective and proven BMPs, and that existing references  were too
technical and difficult to read. The Field Manual contains concise descriptions of BMPs for erosion and sediment control
and  site management (Table 1). Overviews  of regulatory requirements  and inspection and monitoring responsibilities
are also provided.
                                                      287

-------
Table 1. Erosion and Sediment Control Field Manual BMPs
Erosion and Sediment Control Field Manual BMPs
Erosion and Sediment Control Practices
. Scheduling
. Preservation of Existing Vegetation
. Slope Grading
. Temporary Seeding and Mulching
. Permanent Seeding and Mulching
. Hydromulching - Hydroseeding
. Dust Control
. Erosion Control Blankets and Geotextiles
. Fiber Rolls
. Temporary Stream Crossing
. Stabilized Construction Entrance
. Entrance/Exit Tire Wash
. Outlet Protection - Energy Dissipation
• Check Dams
. Silt Fencing
. Temporary Straw Bale Dike
. Sand/Gravel Bag Barrier or Rock Filter
. Storm Drain Inlet Protection
. Catch Basin Inlet Filters
. Sediment Basin
. Sediment Traps
. Dewatering: sediments/toxic pollutants
. Secondary Filtration
General Site and Materials Management Practices
. Water Conservation Practices
. Solid and Demolition Waste Management
. Hazardous Waste Management
. Spill Prevention and Control
. Vehicle and Equipment Service
. Material Delivery, Handling, and Storage
. Paints and Liquid Materials
. Handling and Disposal of Concrete and Cement
. Pavement Construction Management
. Contaminated Soil and Water Management
. Sanitary/Septic Waste Management
. Landscaping Management
      BMPs are described  in a user-friendly format that features full-color graphics, including do and don't  illustrations
(Figure 2). Each BMP  description includes its purpose, application, limitations, practices,  inspection, and maintenance.
There is  a section  on Corrective Measures that  discusses what can go wrong and common  installation problems. This
latter section is essentially a troubleshooting guide  that contains a table of common problems  and corresponding
corrective measures. Overviews of regulatory requirements and inspection and monitoring responsibilities are also
provided. The Field Manuals waterproof 9" x 9" binder and coated pages make it ideal for use in the field. As such, it
provides  the  essential  connection between the  enforcement,  outreach, and  technical assistance  components of the
Regional  Board's  Program.

Overlap of State and  Municipal Authorities

    The  Regional  Board's  Program provides a clear demonstration of  how the Storm Water Phase II  Program's
construction requirements may be implemented. The Phase II rule  allows states to recognize  compliance with municipal
program  construction requirements as  equivalent  to  compliance  with a state-issued  NPDES  permit for construction,
if it can be demonstrated that the municipal program requirements are equivalent. In such situations, a construction site
deemed in compliance with  a municipality's requirements would be deemed in compliance with the state-issued NPDES
permit. The key is demonstration that the municipal  program qualifies as equivalent.

    In the San  Francisco Bay area, as previously noted, the Regional Board has issued  NPDES permits for municipal
stormwater discharges that  include  requirements to control  discharges from construction sites.  In essence,  there is an
overlap of Regional Board and municipal authority where municipalities are  in compliance with their permit requirements.
                                                     288

-------
                 SHORT CIRCUIT       -^|^
                      DO  ?
                 BAFFLES  WILL  INCREASE  DETENTION  TIME
Figure 2. Sediment Basin Design.

    Unfortunately, what may seem equivalent on paper may not be equivalent in practice. The case in point is that the
Regional Board's inspection program  noted above identified many construction sites out of compliance with their
construction NPDES  permits. Consequently, these same construction sites would be deemed in  non-compliance with
municipal  requirements,  In addition, the  same  inspection findings can be applied to the municipality. Since the
municipality's NPDES permit requires it to  control discharges from construction  sites, construction site non-compliance
means the municipality is not in compliance with  its  NPDES permit. In these circumstances, the Regional Board may
(and has) taken enforcement action against both  the construction site and the  municipality.

    To date, the primary enforcement tool used for the municipalities has been Notice to Comply. The Notice to Comply
requires  the municipality to report on the failure of its construction control program and to  implement timely corrective
actions.  Most municipalities have been very responsive to this "wake-up-call," and have  made improvements to
demonstrate the desired  "equivalency." The net result is a negative turning  into a  positive. The Regional Board's
Program, with its balance between enforcement  and education, has provided  a cfe facto mechanism for recognizing
municipal program equivalency allowed  by the Phase II  rule. By its design and  implementation, the program essentially
                                                  289

-------
requires municipalities to demonstrate such equivalency. Consequently, the Regional Board will significantly reduce  or
eliminate its inspections in  municipalities where Regional  Board inspections find construction sites in  compliance, thus
providing incentive and  reward  to both the building industry and municipalities.

Lessons Learned

    Lessons learned in  the development and implementation of the Regional  Board's  Program are summarized in the
following points:

    . The only effective means of controlling erosion is  erosion prevention,  which requires careful planning and
      adherence to seasonal time-lines. Sediment capture should be used  only as a  secondary  or back-up plan.

    . Regulation without education is  ineffective. Often, noncompliance is due to  lack of awareness  of the regulatory
      requirements  and cost-effective, proven BMPs.

    .  Education without enforcement  is impotent.  Despite good intentions, the  building  industry is constantly trying to
      maximize its  investment dollars, and environmentally sound BMPs are often  superseded by time  pressures to
      complete a project.

    .  Enforcement actions must be  severe enough that they cannot be accepted as a cost of doing  business.

    .  The balance between regulation and  education is  dependent on  readily available technical assistance and
      implementation  tools.

    . Outreach and technical assistance needs to be directed to the right audiences. Workshop agendas and
      attendance were initially misdirected toward planners  and  local decision makers. Key attendees are municipal staff
     who actually  review SWPPP  plans  and perform on-site inspections and  building  industry staff who are onsite.
      Evaluations revealed attendees wanted more technical information on installation and less time spent  on municipal
      general  plan/environmental  plug. Audiences  are especially  responsive  to  builders discussing their experiences
      in implementing BMPs.

    .  Both the building industry and municipalities have historically  short shrifted training. Workshop attendees
      expressed relief that practicable  training is finally available - especially information on vendors, cost comparisons,
      and practical  BMPs. The  building industry and  municipalities now realize costs of training are minimal relative to
     the benefit.

Conclusions

    The bottom line is that environmental  regulators, municipalities, and the  building industry have different  priorities that
must be reconciled. Regulators seek no adverse impacts to waters. Municipalities seek economic growth. Builders want
unfettered development. In the case  of construction-related erosion,  the means to each end is the same...effective
erosion and sediment control.  A little more work on the  part of each party involved has proven that their different
priorities are attainable  and  even harmonious.

    Since the Regional Board made  enforcement a top priority and began a collaborative effort with the building  industry
and municipalities to  provide cost-effective outreach and training, construction site compliance with NPDES permit
requirements has  risen from 20% three years  ago to  greater than 90% today.  Municipal compliance has risen similarly.
                                                     290

-------
             Regulating Sedimentation  and Erosion  Control  into Streams:
                                   What Really Works and Why
                                                Seth R. Reice
                                            Department of Biology
                                   University of North Carolina at Chapel Hill
                     Chapel Hill, NC 27599-3280 (address all correspondence to this author)

                                                JoAnn Carmin
                                          Department of Public Policy
                               Virginia Polytechnic  Institute and  State  University
                                               Blacksburg,  VA
Abstract

    The  overall objective of this project was to determine the  effectiveness of different environmental  policies,
regulations, and incentives in reducing the ecological risks and consequences of sedimentation  to streams.  We were
trying to  learn which sets of  regulations, enforcement strategies, and landscapes result in effective protection  of stream
communities from degradation, resulting from erosion and  sedimentation from  construction sites.  By connecting erosion
control efforts to environmental  impacts,  our aim was to  create  more effective  management strategies that  ultimately
provide  environmentally  sustainable social and economic  development in  our watersheds.

    We chose  four replicate construction sites in each of three regulatory jurisdictions that varied in stringency of
regulations and enforcement activities. At  each site,  we  conducted instream assessments  of water  quality and
biomonitoring  of macroinvettebrates and fishes to determine the success of the regulators in protecting stream
ecosystem  health. We combined these results with evaluations of the regulatory environment to link the  policies and
management styles  of the regulators  to the  effectiveness  of protection of the streams. While all construction sites did
some damage to the steams, we found that enforcement style and frequency of inspections  were far more important than
the nature of the regulations  in preventing sediment pollution of  streams.

    Keywords:  Development, enforcement,  rivers, sedimentation, streams,  regulations,  regulatory effectiveness.

Introduction

A  critical problem in American rivers and streams is sedimentation. Sedimentation degrades water quality, alters  habitat
for fish and macroinvettebrates, limits  ecosystem functions  and services, and reduces the aesthetic and economic value
of rivers  and  streams. Many  regulations  and policy incentives  have been devised to control sediment pollution  of our
rivers and streams. Yet there has  rarely been an attempt to  reconnect the  policies with the ecology of the rivers. That
was  the goal  of this research. This work integrates  the regulatory environment, sediment ordinances,  and policies with
resultant ecological impacts of sedimentation on rivers and streams. The question the ressearch sought to answer was
"What combinations of  policies,  regulations and on-site interactions between regulators and developers  really work to
enhance  stream biota and stream ecosystem health?"

    Research goals were accomplished by comparing similar streams in different regulatory jurisdictions (a comparative
watershed approach).  We  tested the effectiveness of different intensities of sediment control regulations and
enforcement. We used the streams to tell us what  matters ecologically.  The selected political jurisdictions differed  in
the stringency  of their erosion and sediment control requirements and the nature and intensity of enforcement  of the
regulations. We chose 17 construction sites  along streams in three different jurisdictions. We interviewed the  regulators


                                                    291

-------
and  developers at each site  and we studied the regulations and the attitudes of the  regulators and developers. At each
site,  we sampled  the  streams being impacted. Some  projects are still unfinished. Therefore, we will present only the
results  of the 'before  construction' and 'during construction'  samplings.

    We asked "Which erosion and sediment control regulations really work and why?" We have analyzed the erosion
and  sedimentation control regulations and  compared them among the respective jurisdictions.  Then we surveyed  the
attitudes and  enforcement activities  at  all levels within each jurisdiction. This paper  will briefly  outline our findings and
focus on what can be done to minimize sedimentation  into streams from construction sites.

Methods

S/te Selection

    We selected three regulatory jurisdictions so they would vary across a range of  two critical variables: (1) stringency
of regulations  (how strict and how rigorous the  rules are) and  (2) stringency of enforcement  (i.e.  frequency  of inspections,
severity of punishment of  violations). A  summary of some of the salient characteristics of the three  regulatory jurisdictions
is given in Table  1. Construction sites were selected  from the array  of applications for grading  permits filed with the
erosion and sediment control offices in  each jurisdiction.  The biggest constraint in locating study sites was  the availability
of construction sites on streams with  riffle  zones,  One  jurisdiction  (District IV)  extends eastward  into the coastal plain
as does Eastern Wake County. Therefore,  many  otherwise  promising sites, which  had  sandy  bottomed, slow flowing
streams, were eliminated  from  our study. To be selected for this study, the construction sites had to have  certain critical
characteristics. For example, streams had to be within 100m of the site. There  also had be a significant slope from the
construction site down to the stream,  so that  if erosion occurred it would impact the stream. These factors made site
selection extremely difficult. In this paper, we will discuss  only the impact of large construction  sites (>100 acres
disturbed). We have located and sampled ten large sites.

Table 1. Selected Characteristics of Erosion and Sediment Control Jurisdictions Used for this Project
Minimum Disturbed
Area Requiring
Erosion Plan Staff
Orange County
Wake County
District 4'
(16 Counties)
# Field Total Area
(Miles') Projects Ratio
0.5 Acres
1 .0 Acres
1 .0 Acres

# Active Site/Staff
3
4
4


400
858
8,116


-100
-400
-1000


33.3
100
250

* District 4, of the NC Division of Land Quality oversees all construction projects in all 16 counties without a Local Erosion and Sediment Control
Program. It covers all governmental construction in the District 4 area, including Orange and Wake Counties. So, a single stream can have adjacent
construction sites along the banks, one supervised by District 4 and the other by the Local Program.


Stream  Sampling  Procedures  and  Variables Sampled

    We monitored at least three  replicate sites per jurisdiction for the large construction sites. We sampled before,
during, and after construction. We cannot control the timing of the construction  projects, and since sampling must follow
a rain of >1/2" in  24 hours (i.e.,  a rain with  the potential to produce erosion and sedimentation), our sampling  was
dependent on the weather and the contractors. This  means that the time  between the  before, during, and after sampling
is highly variable. Since upstream  and downstream controls were sampled  on the same day as the "at the site" samples,
this did not cause  a significant analytical  problem.

    We sampled three sites on each  stream, including >1 00m upstream, at the site, and >1 00m downstream. We took
two replicate Surber  samples for  macrobenthos, identified to species whenever  keys permitted, including chironomids.
Chironomids are essential because they often constitute >90% of the individuals sampled, especially  in the impacted
reaches. The number of samples  is small since our  objective was not to  analyze any one stream in detail, but to  treat

                                                        292

-------
streams as replicates. In the overall analyses, multiple samples per site are pseudoreplicates. The mean of the replicate
Surber samples was used  in the analyses. We electroshocked for fish along one  50m  reach  of riffles and pools.  We
collected basic water chemistry data. Water quality parameters  included D.O., turbidity,  conductivity, Total N, NH,+, NO3"
1, Total  PO,-', SRP, pH, and temperature. We also studied leaf litter decomposition rates. Five g  leafpacks of Comus
florida (dogwood) leaves were incubated for two  weeks in situ at all three sites in the "during construction" period to
assess the critical ecosystem process of litter decomposition.

Environmental  Policy Analysis

    Surveys and semi-structured interviews  were used to investigate both the regulatory  agencies and developers. The
surveys focused on  the capacityof the agency, the external commitmentthat the agency receives, as well as the internal
commitment toward the environment, and the control measures that are  used.   The surveys and interviews  achieved
a 100% response and participation  rate. Although it has been harder to get their cooperation, we have nearly completed
data collection from developers.  The  survey data is being  augmented with documentary data  from  the sediment and
erosion  control  offices in  each of the counties.

    The evaluation of implementation  focused on  (1) the extent to which developers comply with sediment and erosion
control regulations and (2)  the way that regulatory  and  organizational factors interact  to  shape compliance behaviors.
The examination of outcomes combines social science and biological data  to examine associations among regulatory
styles, agency  activities,  and stringency of policy  enforcement. We further analyzed  how variations in  sediment and
erosion  control  enforcement are related to the ecological outcomes (including biological, chemical, and  physical factors)
in the impacted  streams.

Hypotheses

    Hypothesis 1. Greatest  degradation will  be evidenced at the construction sites, compared to upstream controls, with
moderate  to  complete recovery  downstream.

    Hypothesis 2. Tighter enforcement of erosion  and sediment control laws will result in less  damage to streams.

    Hypothesis 3. Stronger erosion and sediment control regulations will result  in less damage to streams.

Results

    Nearly all biotic and environmental variables measured  tell the same story. Figure 1 shows the  changes in  the EPT
Index for the during construction  sampling. That is the species  richness of the Ephemeroptera (the  mayflies), Plecoptera
(the stoneflies), and Trichoptera (the caddisflies). The tally of EPT taxa (i.e., EPT Richness or the EPT Index) is a well-
established and universally accepted measure of stream health. These groups of aquatic  insects are particularly
sensitive to  (and highly intolerant of)  high  temperature, low oxygen, toxic substances, a wide range of pollutants, and
burial by sedimentation. An abundance of EPT species  and individuals and  high EPT diversity are  clear indicators of
good stream health. Reductions in EPT  values demonstrate degradation of  stream conditions.

    EPT richness follows a pattern. The differences between jurisdictions are clear. The greatest decline in EPT values
from upstream to at-the-site occurs in District IV. The EPT Index in Orange County changes little at any site.  Wake
County  actually  shows some enhancement of the  EPT richness  as you go from upstream to at-the-site. We  sampled
many other variables  but the results parallel the EPT richness.

    A short summary  of the enforcement  activities and attitudes of the regulators in the various jurisdictions is  found in
Table 2. These  data show  that these agencies differ in these  aspects. Orange  County  had the strictest enforcement,
penalizing nearly 25% of all construction projects,  while Wake penalized -22% and District IV penalized  only ~  4.5% of
the projects they inspected. Orange County is most likely to use stop-work orders to halt construction due to
sedimentation violations, while District IV relies on fines. District  IV is perceived as being so understaffed that it is  unable
to make sufficient inspections. Consequently, some contractors  do  not feel obliged to follow their approved plans.  Some


                                                     293

-------
contractors  agree to a  plan and then cut costs by not following the sediment controls. This laxity is detectable from the
stream data (see Figure 1).
                                            Total  EPT Richness
         a
         x
         (8
        z
         0
                                                                                  Error bars = 1 std. error
                District IV Orange   Wake
                      Upstream
   District IV Orange   Wake

         At The Site
          District IV Orange   Wake

               Downstream
Figure 1. Total EPT Richness.

Table 2.  Regulatory Environment
Agency/Variables
Enforcement Action
Penalties enacted past year
Stringency of Penalties
Orange County
Very Strict
24
High
NC District IV
Average
44
Medium
Wake County
Strict
88
Medium
Attitudes of Regulators

Perception of official commitment
Percent of developers that regulators
Supportive
    8
Very  Supportive
      40
Indifferent
   10
    Orange County and Wake County  regulators  generally think that developers will try to avoid complying with erosion
and  sediment control  regulations. As the regulators' workload increases, their task becomes  more difficult.  This  may
result  in  regulators adopting a more forgiving  attitude toward developers and  less  vigorous enforcement  of the
regulations.  District IV regulators think that fully  40% of developers are  trying to comply with  the regulations.

Discussion

    There is a clear link between the attitudes and enforcement activities of the regulators of erosion and sediment
control ordinances and environmental outcomes in the streams near construction sites. If the regulations are completely
effective, all sites should  be similar to the upstream controls when the construction  is completed and the site has been
                                                        294

-------
stabilized  (i.e., revegetated). In our analysis,  the degradation is  clearly detectable  in the benthic community data (see
Figure 1). Benthic communities at the site are dramatically negatively impacted in District IV, unchanged in Orange
County, and  actually  enhanced  in Wake  County.  The effect is sometimes reduced downstream but the degradation
persists downstream in District IV.

    Wake County and District IV  have identical regulations, while Orange County's regulations are more stringent.
Comparison between  the two  jurisdictions with the same rules  but different inspection and enforcement intensities will
help us tease apart these factors. The stream data suggest that the laws, as written, are not particularly important. Wake
County has the best environmental results  while District IV has the worst stream degradation. Our analysis suggests that
differences in laws  and regulations have limited impact on the degree of degradation  of stream biota.

    The key factors seem to  be the attitudes and enforcement  behavior of the regulatory  agencies.  The frequency of
on-site inspections is  particularly important. In Orange County, every construction  site  is inspected  every week.  If it is
a problem site, the inspectors may  visit daily.  In Wake County, the inspections are closer to every other week. In District
IV, the goal is to visit every site once in the entire duration of the project. They also seek to respond  to any citizen
complaints within one week. In Orange County a  complaint generates an  inspection within one day.  Another  critical
factor is topography.  A very steep, erodible slope can undermine the best attempts at enforcement of erosion and
sediment  control regulations.

    Our analysis suggests  that differences in the  nature and frequency of enforcement and  inspections does matter.
Developers tell us that a rigid, command and control approach to enforcement is less  palatable to them  than a flexible
problem-solving  cooperative  approach. If the developers perceive that the  regulators are really tying to help them  keep
sediment on site and out of the streams, they do a better job. Flexibility enters in as follows. If the sedimentation
inspectors have  enough time to  analyze a sedimentation  problem in detail,  their suggestions will be better. Very often,
the inspectors need the authority to implement solutions which are not exactly "by the  book." When inspectors propose
innovative solutions, which can really solve the problem, this encourages the developers  to  be  more cooperative.  More
frequent inspections and a cooperative, flexible  approach by regulators does ameliorate  the  stream damage among
similar streams  in different jurisdictions.

    On the other hand, if the developers know that the regulators will in fact shut them down (with a stop-work order or
a court injunction), it is easier  for the regulators to get developers' attention.  Fines are notoriously  ineffective penalties
in  North  Carolina. Presently the maximum fine is $500 per day. When  developers are  pouring millions of dollars into
a project, this amount of fine is trivial. As one said, "It's just a cost of doing  business."  In essence,  the effectiveness of
erosion and sediment control depends more  on enforcement than on  how  the regulations  are written. Even  with weak
laws, the  success of Wake County's  Erosion and Sedimentation Control Program plainly depends on their on-site
enforcement  actions.

Recommendations

    . Provide sufficient inspectors to visit  each construction  site at  least weekly.

    . Give inspectors the authority and knowledge to implement innovative solutions to erosion problems on a  site-
      specific basis.

    .  Empower the inspectors to issue severe penalties  (stop-work orders)  in the case of sedimentation violations.

    .  Raise the  maximum level of fines to a meaningful  amount (we  suggest $10,000 per day).

    .  Educate the development  community to the  damage that  sedimentation does to stream communities.
                                                      295

-------
                     Effectiveness in Erosion and Sediment Control:

                                  New Initiatives in  Indianapolis


                                               James Hayes
                                               Marcia  Mathieu
                              Marion County Soil  and Water Conservation District
                                         Indianapolis, Indiana  46237
                                        marcia-matthieu @  iaswcd.org

                                               Greg Lindsey
                                 Center for Urban Policy and  the Environment
                         School of Public and  Environmental Affairs, Indiana  University
                                         Indianapolis, Indiana  46204
                                            glindsey@iupui.edu

    Since the late 1960s, when the  severity of pollution from sediment from construction sites was first documented,
many states and municipalities have worked to develop effective  programs  for erosion and sediment control. These state
and local programs were augmented in 1987, when  Congress required  in the Clean Water Act that operators of all
construction  sites over five acres prepare  erosion  and sediment control plans and  obtain National  Pollution Discharge
Elimination Permits (NPDES). At that time, some states, such as Maryland and  North Carolina, already had well-
supported, comprehensive approaches  that were developed largely in  response  to state law. Other states,  including
Indiana,  have relatively  new programs that were adopted only after the federal mandate. In general, these newer
programs are not as comprehensive, and  managers are still working to develop  systematic and effective methods for
implementation.

    This  paper describes a new initiative in Indianapolis, Indiana, to increase the effectiveness of erosion and  sediment
control programs.  The  paper  describes a general framework for evaluating erosion and sediment control programs.
Next,  it describes an intergovernmental, "S.W.A.T." team approach to inspection  that was  used in  Indianapolis in the
summer  of  1998. The paper summarizes the results of the  inspections and concludes with a discussion of the
implications for  managers of erosion and  sediment control programs.

Effectiveness  in Erosion and Sediment Control  Programs

    Managers and analysts in Maryland and North Carolina have used  a general  framework for evaluating erosion and
sediment control programs (Clevenger, n.d.; Departments of Civil  Engineering and City and Regional Planning, 1990).
The framework comprises five criteria, each of which must be satisfied for sediment pollution to  be controlled effectively:

    .  Complete  coverage

    .  Competent plans

    •  Careful installation

    .  Continual maintenance

    .  Consistent enforcement

    Overall effectiveness requires that the coverage  rate (the  proportion of construction sites with controls) approach 100%.
Operators of development sites must know of regulatory requirements and  make efforts to comply. Second, erosion  and
sediment control plans must be competent.  Best management practices (BMPs) incorporated into plans by engineers or
technicians must be able, if constructed properly, to control erosion and sedimentation.   Third, BMPs must be installed
completely and  correctly. Improper installation may result in failure and off-site sedimentation.  Fourth, BMPs must be

                                                    296

-------
maintained  for the duration of the construction  process.  Finally, consistently effective  approaches to enforcement must be
developed to ensure compliance with substantive criteria.  If any one of these criteria is not met, the objectives of erosion and
sediment control may not be achieved.

    When new programs are developed, these criteria can be considered sequentially. That is, when building a new program,
managers must first make sure that developers and builders are aware of regulatory requirements.  Next, they must work to
ensure that developer's engineers are preparing good plans. If developers are aware of requirements and are submitting good
plans, attention  can turn to installation and maintenance. Use of enforcement tools always is a last resort.

An  Initiative in Indianapolis

    Although  the City of Indianapolis has a sediment control  ordinance that  predates federal requirements, erosion and
sediment control programs in  Indiana have largely been developed  in response to a  state regulation [Title 327-IAC 15-5
(Rule 5)], that was adopted in 1992 to comply with EPA regulations.  Since the adoption of Rule  5, managers generally
have seen improvements in efforts  to comply.  Most  developers and builders are now aware of requirements, and
coverage is approaching 100%.  With  respect to plan review, the Division of Permits  in the  Department of Capital Asset
Management (DCAM) is  responsible  for plan review pursuant to the city's ordinance, while,  under a memorandum of
understanding  with the Indiana Department of  Environmental Management (IDEM) and the Department of Natural
Resources (IDNR), the Marion County Soil and Water Conservation District (District)  is responsible for plan  review
pursuant to Rule 5. MCSWCD reviewers estimate that the quality of  plans is improving, but that as many  as 60 to 70%
of all plans still  must be  returned and revised  before approval. Most plans are approved on the second  iteration.

    Although the review process now assures that competent plans are being prepared, installation  often remains
inadequate, BMPs often are  not maintained, and resources for inspection and enforcement are limited. IDNR has only
seven inspectors in the Division of Soil Conservation  for all 92 counties and 550 municipalities. IDNR inspectors
generally work  individually within  regions, inspecting sites sequentially and in  response to complaints. District personnel
lack enforcement authority and mainly visit sites  in response to complaints.  In Indianapolis, sediment control has been
a low priority with DCAM, which has  no  inspectors trained in or assigned exclusively to enforcement of sediment control
requirements.

    Managers  have struggled to find ways to overcome resource limitations and to  increase  the effectiveness  of
implementation. In 1998, IDNR and District staff conceived  of a "S.WAT." team approach to inspection.   In this
approach, all IDNR inspectors and District staff together focused their efforts on all open construction sites  in the county.
The objective was to visit all sites in a brief time period, thereby increasing the visibility  of the program. Managers
believed that intensive scrutiny of the county,  if only for a brief time,  would result in greater efforts at compliance. One
of the assumptions on which this approach was based was  that there are  both formal  and informal networks among
developers and  builders and  that this approach  would stimulate discussion about compliance issues.

    In Indianapolis, IDNR and District personnel  completed a county-wide survey  of construction sites on June 23 and
24,1998 (Hayes and Matthieu 1998). DCAM staff was  invited to participate. IDNR, District, and DCAM staff visited more
than 300 construction sites. Of these sites, 177 were active and were evaluated  for compliance  with Rule  5.
Construction had not yet begun at 23 of the sites, construction had  been completed at 61 sites, and the remainder were
not evaluated  because they were inaccessible or because construction  was just beginning. This summary is  restricted
to the sites under active construction. The results provide a good picture of the current status of implementation and the
general level of effectiveness of erosion and sediment control requirements in  Indianapolis.

    Inspectors  evaluated sites for compliance in nine categories  using  a standardized  checklist  developed by IDNR.
Sites  also were checked for obvious evidence of off-site sedimentation. The nine categories were:  (1) proper installation
of erosion and  sediment control measures; (2) perimeter erosion  control measures; (3) erosion and sediment  control
measures on individual building  sites; (4) protection of  storm-sewer inlets; (5) stabilization of disturbed areas, (6) proper
stabilization of drainage channels; (7) stabilization of drainage outlets; (8) maintenance of existing erosion  and  sediment
control measures; and (9) tracking or accumulation of  sediment on roadways.  These criteria generally can be grouped
within the installation and  maintenance stages of the evaluation framework outlined  above,  although most involve

                                                     297

-------
aspects of both installation and  maintenance. The  first seven criteria  primarily concern installation of BMPs; only two,
maintenance  and tracking,  primarily concern maintenance.

    Inspectors rated each  applicable criterion at each site on a  scale of Satisfactory,  Marginal,  Unsatisfactory  or not
applicable (NA). Items in compliance with Rule 5 were rated S, items that were in danger of becoming out of compliance
were rated M and items in violation of Rule 5 were  rated U.  Because all criteria were  not applicable at all of the sites,
the number of sites evaluated for with  respect to each criterion varies.

Disturbing Results from  Disturbed Sites

    The results  of the inspections are summarized in Figure 1 (Hayes  and Matthieu 1998). Overall the results show that
installation is inadequate  and that  maintenance  is worse.  Improvements in  implementation clearly are  needed.
Discussion of each of the  nine items reviewed follows.
                        Percent Satisfactory   D  Percent  Marginal   D Percent Unsatisfactory
Figure 1. Rule 5 Compliance Summary-Percent of Applicable Sites.

Installation of Erosion Control Measures

    Erosion control depends upon installation of appropriate control practices in given situations. Examples
of these practices include silt-fence  perimeter controls, sewer inlet and outlet protection devices, and the use
of stone or mulch to stabilize slopes. Proper installation of these devices and practices helps reduce the risk
of failure that may result in erosion and off-site sedimentation.   Erosion and sediment control measures were
installed correctly at only 32% of the active sites. Installation was marginal at 24% of the sites and had been
done incorrectly at 44% of the sites. Proper installation was marked not applicable in cases where no erosion
control  practices were in use.

Perimeter Erosion Control Measures

    Perimeter erosion control measures are designed to keep sediment from leaving a site directly at its perimeter
through  sheet or gully erosion.  Perimeter erosion control devices/practices such  as silt fence  or buffer strips  should be
installed before  land  disturbance  begins. The most effective  and  cost-efficient  perimeter control practice is to leave
existing  vegetation in  place, especially along waterways. Perimeter  measures were installed and in compliance  at 48%
of the sites. Marginal conditions were found at 21% of the sites, and 31%  of the sites were found to  be out of
compliance.  Perimeter erosion  control  measures were not applicable at level or  inward-sloping sites.
                                                    298

-------
Individual Building Sites

    Under Rule 5,  erosion control is the  responsibility of the  site operator, either the  developer or builder, throughout
construction. At some sites, after the infrastructure has been installed and the lots have been sold to individual builders,
the developer no longer has direct  control over erosion  and sediment control practices on those lots. Builders and
contractors may or may not  install and maintain erosion  control practices.  Erosion control on  individual building sites
is a serious problem in Marion County.   Erosion and  sediment control measures on  individual  building sites at
developments were found to be adequate at only 9% of the active construction sites. Measures were in marginal
condition at 21% of the sites, and 70% of the sites were found to  be out of compliance. Most of these sites lacked proper
construction entrances, storm-sewer protection, and  perimeter protection. This category  was not applicable for sites that
had not yet begun  construction of homes.

Storm-sewer Inlet Protection

    Sediment entering storm-sewer inlets  significantly reduces the capacity of retention/detention basins and drainage
channels to store and convey stormwater away  from flood prone areas effectively. If sediment is  not removed  prior to
site  closure, the specified volume and  dimensions  of retention/detention  basins  that were approved  by the City  can
change. Inlet protection measures  are especially important when sediment is tracked  into or  allowed  to accumulate in
roadways where it is conveyed directly to sewer inlets. Using measures such as seeding and silt fence  adjacent to inlets
will prevent sediment from clogging  inlet  protection  devices and accumulating  in the  streets. Storm-sewer inlets were
adequately protected from sediment at just 14% of the construction sites. Sewer inlets were  marginally  protected  at 17%
of the  sites, and inlet protection  measures were  inadequate and not in compliance at 69%  of the sites.  Inlet protection
was not applicable  to sites that  had not completed  sewer installation.

Stabilization of Disturbed Areas

    Stabilization of disturbed areas on construction sites may  be the single most important practice  for reducing  erosion
and off-site sedimentation. The best practice for achieving  stabilization  is to leave vegetation in place wherever possible.
If soil must be disturbed, stabilization is  relatively easily accomplished through temporary seeding  or application of
erosion control  blanket. Rule 5 requires that disturbed areas that will  be  inactive be temporarily seeded. Stabilization
by seeding results  in higher  perceived value by potential buyers,  offering developers a financial incentive  to vegetate
land as soon as possible once the infrastructure  is in place.  Of active sites that were visited, 30% were in compliance
with Rule  5 with respect to stabilization of disturbed areas, while 32% of the sites were marginal and  38%  were not in
compliance. This category of compliance was not applicable to sites that were being actively  cleared  or  nearly completed
at the time of the visit.

Drainage Channel Stabilization

    Ditches and swales designed to convey storm water  away  from development to natural drainage ways or  storm-
sewers are subject  to severe erosion  and  deterioration if not adequately protected. Erosion and damage  to conveyance
channels  results  in off-site sedimentation of waterways. This can be avoided by stabilizing the soil in conveyance
channels immediately with permanent seeding of grasses, or with stone,  mulch,  or straw cover. Conveyance channel
stabilization was satisfactory at 44% of the sites,  Approximately 23% of the sites had marginally  protected channels  and
33% of the sites had  channels in unsatisfactory conditions.  Conveyance channel stabilization was not applicable at sites
that did not have  or require channels or at those that did  not yet have them constructed.

Outlet  Stabilization

    Storm-sewer and drainage channel  outlets from a site need to be properly stabilized to prevent erosion and
sedimentation of the banks and  waters they empty  into. Outlet stabilization is best accomplished by protecting the  soil
around the outlet with stone riprap, geotextile  fabric,  or with  well-established vegetation. Outlet stabilization was
satisfactory at 73%  of active sites. Outlets were in marginal  conditions at 18% of the sites and unsatisfactory  at 9%. This

                                                      299

-------
category was not applicable at sites that did  not have outlets on the site or where infrastructure development was not
yet completed.

Maintenance of Erosion Control Practices

    Erosion control practices that have been  installed  properly must be maintained to be effective. In most cases, lack
of maintenance results in the same effects as not employing erosion control practices at all.  Examples of maintenance
of erosion control practices  include removing  accumulated  sediment  from behind silt fence and reinforcing inlet protection
after storms. Failure of erosion control  practices allows sediment to leave construction sites via storm-sewers, drainage
channels, roadways and sheet and gully erosion, An often-overlooked aspect of maintenance  is removal of devices after
work is completed. Maintenance of erosion control practices was satisfactory at only 18% of the sites. Maintenance was
marginal at 27% of them, and there was little or no evidence of maintenance at 55% of the sites. Maintenance of erosion
control measures was not applicable at sites that did  not employ erosion or sediment control practices.

Sediment Tracking and Accumulation  in Roadways

    Soil  and sediment in streets and  roads are readily washed into sewers and drainage channels and can be a
significant source  of pollution.  In addition, the sediment can  be a traffic hazard with the potential  for costly litigation
against the  local governments or  developers. Sediment  accumulated in  roads  is also unsightly and  may  discourage
potential home  buyers. Tracking and accumulation of soil  in roads was kept to an acceptable level at 24% of the sites.
Approximately 28% of the sites exhibited marginal compliance with the rule for keeping roads clear  of sediment. Sites
that were out of compliance with the rule made up 48% of this category. Large  industrial sites where equipment was
usually kept on site and residential sites that  did not yet require extensive coming and going  of vehicles were rated not
applicable for sediment tracking.

Off-site Sedimentation

    Sediment is the most abundant pollutant, by  volume, in  Indiana waters.  Residential and commercial  development
sites are potential  sources  of  high volume, sudden discharges of  sediment  that can cause  problems for land owners
down-stream of development.  Besides  the drainage and  flooding problems  caused by off-site sedimentation, sediment
can obstruct and widen  streams and erode stream banks. Sedimentation  of the state's streams and  rivers also causes
habitat damage for many aquatic species. There were obvious signs of off-site sedimentation at 21%  of the active sites.
This figure is believed to be  low, however, due to the large number of sites surveyed in a very short time. Only the most
obvious cases  were checked  as displaying  off-site  sedimentation.

Observations and  Implications: Priority-problem Solving

    A number of observations  that have important implications for managers of erosion  and sediment control programs
can be drawn from this inspection initiative.  First, it is useful to consider the initiative in the  more general framework for
effectiveness in erosion control. Indiana regulations for erosion and sediment control first were adopted in  1992.  Faced
with implementation of a new regulation with few  resources, IDEM, IDNR,  and District staff first  devoted efforts to
education and ensuring  complete coverage and competent planning. In  late 1997 and  early 1998,  program managers
determined that the plan  review process was fairly well established and  that additional effort needed to be devoted to
installation and  maintenance of BMPs.  Because resource shortages preclude  regular, periodic inspection, IDNR officials
developed  a S.W.A.T. team approach.  Teams  of state, district,  and available municipal officials  focused inspection
efforts, visiting and inspecting as many sites as possible in  a short time.

    In  Indianapolis, the results show that implementation generally is  poor. Installation of erosion and sediment controls
was unsatisfactory  on 44% of all sites,  and satisfactory on less than one-third. With the exception of outlet stabilization
practices,  which had  been  installed properly  at nearly three-fourths of the sites,  no practice was installed properly on
more than half of the sites. Perimeter controls, a basic practice,  were installed  satisfactorily on fewer than half of the sites
and they were  unsatisfactory at almost one-third.  Stabilization was satisfactory at less than one-third of the sites, inlet

                                                      300

-------
protection had been installed properly at less than 15% of the sites, and controls on individuals lots had been installed
properly  at just 9% of the sites where they were needed.  It is clear that installation is deficient  and that additional effort
is needed  to ensure  that practices identified on plans are installed properly.

    The  inspections show that maintenance  of erosion  and sediment controls is even worse than installation.  Inspectors
determined that  maintenance of controls was unsatisfactory on  55% of the sites and satisfactory on only 18%.  Mud is
being tracked on streets  and washed into sewers and drainage channels on  almost half the construction sites. Additional
field work to ensure proper maintenance of BMPs is a critical need.

    Although these results were disturbing,  they were  not  unexpected. Program officials  knew  that implementation was
inadequate and  devised  the S.W.A.T. team approach  to  provide a quick, comprehensive assessment of the status of
implementation.  Since the  inspections, program managers have used the results as  part of overall efforts to increase
understanding of requirements for erosion  and sediment  control and to build commitment to  the  programs. City staff
agreed to mail copies of inspection reports to all developers, and the district provided  a  summary of results to all city-
county  councilors.

    The  results provide information that program  managers can  use to establish priorities for problem  solving and
education.  For example,  installation of perimeter controls appears better than efforts to stabilize  disturbed areas on site.
Future inspections and  educational efforts therefore  can focus on the importance of stabilization.  Similarly, since it
appears  that site operators are doing a fairly good job  at stabilization of outlets, this requirement can be de-emphasized,
and additional  effort can be devoted to solving problems like installation of controls on individual lots  that are not
controlled by practices on  the larger development site. More generally, as  more people understand the different steps
in the process  of erosion  and  sediment control, implementation should become more effective.

    The  survey did not focus on discovering reasons behind compliance or  non-compliance, but several inferences can
be drawn from these  data. First, the data and experience  indicate that some developers  are unaware of their obligation
to control erosion and sedimentation and leave  the permitting and erosion control planning to engineers and contractors.
This can result  in a lack of commitment to  implementation. Second, some developers, engineers and  contractors clearly
do  not yet understand the purpose and  importance  of implementing erosion and sediment control  practices.  Education
is needed  to increase their understanding  and commitment. Third, some operators know the  requirements of Rule 5,
but do not take  them seriously, ignoring the Erosion  and  Sediment Control  Plan. For these  individuals,  enforcement
action may be required.  In addition, a general problem that was observed has to do with sequence of construction. All
too often, land  disturbance is beginning before erosion and sediment control  measures  are installed.  More emphasis
must be  placed on installation of practices prior to earth disturbance,  and site  operators must learn to follow the sequence
described on plans.

    Given that resource shortages  are likely to continue, problems in implementation are likely to continue and regulatory
programs are likely to remain less effective than they could be. Steps that  may be taken to increase effectiveness
include making  sure that the regulated community participates in on-site, pre-construction meetings that underscore the
scope and  importance of controls: increasing the visibility of IDNR and District staff and the frequency  of their  site visits;
educating developers, engineers and  contractors about erosion  and sediment control  practices and how to  install and
maintain  them; and emphasizing the need for erosion  and  sediment control  throughout the entire development process.


    The  S.W.A.T. team  approach clearly does not solve the problems of a relatively new,  understaffed erosion and
sediment control  program.  But the approach is  an  effective  way to  obtain  a  significant amount  of information  in a short
time,  raise  the visibility of  erosion  and sediment control programs,  and help establish priorities for problem solving.

References

Clevenger,  B. Miscellaneous unpublished materials.  Maryland Department  of  the Environment,  Sediment  and Stormwater
Administration, Baltimore, MD n.d.


                                                       301

-------
Department of Civil Engineering and the Department of City and  Regional Planning. Evaluation of the North Carolina erosion
and sedimentation control program. North Carolina State University, Raleigh, NC, 1990.

Hayes, Jim and Marcia Matthieu. "S.W.A.T. Summary." Marion County Soil and Water Conservation District, IndianapolisJN,
1998,
                                                      302

-------
                           Using Constructed Wetlands  to Reduce
                         Nonpoint Source  Pollution  in Urban Areas

                                                Jon Harbor
                                Department of Earth and Atmospheric Sciences
                                             Purdue  University
                                          West Lafayette, IN 47907

                                             Susan Tatalovich
                                           FMSM Engineers, Inc.
                                          Columbus,  Ohio  43229

                                         Ron Turco and Zac Reicher
                                          Department  of Agronomy
                                             Purdue  University
                                          West Lafayette, IN 47907

                                        Anne Spacie and Vickie Poole
                                Department of Forestry and  Natural Resources
                                             Purdue  University
                                          West Lafayette, IN 47907
Abstract

    Potential pollutants carried in stormwater runoff from urban surfaces are a major component of Nonpoint Source (NPS)
pollution. NPS pollution is a leading cause of reduced water quality in US rivers and lakes, and there are major efforts
underway to find innovative approaches to reducing NPS pollution from a wide range of sources,  In urban  areas, where
much of the land has existing structures, a major challenge is to find ways to retrofit built sites to reduce NPS pollution
associated with stormwater runoff. One component of this may be more widespread use of constructed wetlands that have
value not only in terms of water quality improvement,  but also in terms of urban ecology, aesthetics and education.

    We have begun a long-term monitoring program of the performance of constructed wetlands in two settings: 1) On
a commercial site where surface runoff is dominated by stormwater flow from parking lots and store roofs,  and, 2) On a
golf course that receives considerable surface flow from adjacent  commercial,  residential and highway areas. Monitoring
includes both continuous measurements of flow, temperature,  conductivity, pH and dissolved oxygen, and automated
sample collection during storm events for more complete chemical analyses. Initial results suggest that the commercial
site constructed wetland acts  as an efficient trap during the spring and summer for suspended sediment and some
dissolved matter. During the fall and winter dormant season trap efficiencies are much lower, and in some cases negative.
 The golf course site constructed wetlands also function as efficient traps during the summer, and plant growth in these
wetlands has been helped considerably by the  regular water supply provided by golf course irrigation.  Both wetland
systems  also provide value in terms  of improved aesthetics, their use by local educators, their diverse ecological
assemblages, and the public relations benefits associated with  visible efforts at improved environmental management.

    Replacing  portions of existing  parking lots with carefully designed constructed wetlands,  and adding constructed
wetlands to urban recreational sites  (such as golf  courses and parks) should be viewed  as  one of several elements of an
integrated approach for effective retrofitting urban areas to reduce NPS pollution from stormwater runoff.
                                                    303

-------
Introduction

    One of the major challenges facing urbanized areas is to find ways to improve environmental management in ways
that do not involve major, costly impacts on existing infrastructure.  Increasing recognition of the environmental impacts
of built areas on parameters such as runoff amount and quality has increased regulatory and public pressure to develop
and implement effective management practices.  However, many of the  best management approaches that can be
integrated into the design of new developments cannot be implemented in existing built areas without prohibitive costs.
Thus there is considerable  interest in best management practices that can  be used to retrofit urban areas for improved
environmental performance.

    Wetlands have the  ability to store large amounts of water, reducing flooding of surrounding  areas and  in some  cases
recharging groundwater (Mitsch  and Gosselink, 1993). In addition, wetlands are capable of improving runoff quality in
many situations (Perry  and  Vanderklein, 1996) because they trap both solid  and dissolved pollutants. Wetlands also can
have considerable aesthetic benefits, and provide habitat for  a wide range of plants  and animals. Constructed wetlands
are wetlands specifically designed and built for hydrologic and water quality management, as opposed to either natural
wetlands or created wetlands. Created wetlands are designed and built to  replace  lost wetlands or to compensate for
destruction of natural wetlands. Using constructed wetlands for water treatment attempts to take advantage of the benefits
of wetlands without compromising natural wetland areas.

    In urban areas there are unique challenges to be faced in proposing and designing constructed wetlands, Existing
built areas rarely include extensive undeveloped space that can  be converted to constructed wetlands. However, there
are several opportunities that arise in many areas, including: 1) Making space by reducing the size of an existing parking
lot; 2) Adding a constructed wetland to a redevelopment or urban renewal project; 3) Adding a constructed wetland to a
park or green space;  4) Adding constructed wetlands to existing recreational facilities such as golf courses. A second  major
challenge in proposing  constructed wetlands in built urban  areas  is to  maintain adequate hydrology  for  long-term wetland
survival. The extensive impervious surfaces of built areas generate large amounts of runoff during storm events, but this
water is usually routed  quickly away from the built area to  prevent flooding.  Because there is little opportunity  for rainfall
to infiltrate into the soil  in urban areas (because most soil is covered by impervious surfaces), shallow groundwater flow
is reduced. This means that wetlands in urban areas will receive far  less between-storm water recharge from shallow
groundwater than would be  expected for a  similar non-urban setting.  In essence, wetlands in urban areas  will  experience
a "flood and drought" hydrologic regime, which is poorly suited to  an ecosystem that is based on extensive periods of wet
conditions. One way around this problem is to look for locations where  water is applied regularly to adjacent areas, in
particular where extensive  irrigation is used. Golf courses and  lawns and  gardens of major corporate complexes  are
potential sites where  between storm irrigation might provide excess runoff and soil water drainage to adjacent constructed
wetlands.

    Given the potential use  of constructed wetlands to improve water quality in built areas, it is important to evaluate how
well wetlands function as pollutant traps in such settings. Such studies can  be used to drive design improvements, and
to evaluate the cost-effectiveness of using constructed wetland for NPS pollution control. Although there has been less
work done in the area  of stormwater constructed wetlands,  in comparison to wetlands used as part of a wastewater
treatment system (e.g.  Hicks and Stober, 1989), limited results so far suggest that wetlands can be effective in treating
stormwaterfor nonpoint source (NPS) pollution (Mitsch and Gosselink, 1993; Witthar, 1993; Livingston, 1989). Few data
sets are  available  because  of poorfollow-up of  constructed wetland performancethrough  appropriate monitoring  programs
(Perry and Vanderklein, 1996). However, available studies to  date and theoretical  reasoning suggest that NPS pollution
control  is enhanced by maximizing the distance between the wetland's inlet and outlet,  including deep and shallow sections
in the wetland, selecting vegetation on the basis of climate and water quality and supply conditions, maximizing the ratio
of treatment  area to base flow, and minimizing the slope along  which the water travels (Horner, etal., 1994;  Witthar, 1993).
The idea  in such a  design is to model the constructed wetland after a natural wetland, which not only has the ability to slow
down the flow of water  (as does a detention or retention pond), but also can  remove pollutants from  the runoff water. The
most important factor in the design and maintenance of constructed wetlands is hydrology (Mitsch and Gosselink, 1993).
Without the  proper water inflow and outflow, the newly created wetland  will fail and  be unable to accomplish  its task of
stormwater treatment.
                                                     304

-------
Aims and Objectives

    We are monitoring the performance of urban constructed wetlands in two settings, a constructed wetland  incorporated
into site development for a commercial facility and a series of constructed wetlands built into a recently renovated golf
course that receives runoff from an adjacent urban area. Both sites are in West  Lafayette, Indiana.  The goal  of long-term
monitoring is to provide insight into seasonal and longer-term variations in trap efficiency, both as the basis for improved
scientific understanding of constructed wetland processes and controls, and to form the basis for future improvements in
design.

Study Areas

    The commercial constructed wetland site  occupies approximately 0.51  ha, with a water surface area of 0.26 ha and
volume of 1300 m3. This wetland is intended to treat the "first flush" of runoff, and so was designed to accommodate the
volume of water  corresponding to first half-inch of precipitation on the store's impervious surfaces (the parking lot  and  the
rooftop). The mean depth of the constructed wetland is 0.5 m but this includes two deeper pools with a maximum depth
of 1.8 m  (Tatalovich, 1998). Conventional wisdom (which may  not be correct) states that 90% of the annual pollutant load
is transported in  the runoff produced by the first 1.3 cm of precipitation  (known as the first flush), and this has  been shown
to be true for the transport of most pollutants over impervious surfaces (Chang,  1994). At this commercial site, runoff that
exceeds the first-flush equivalent is routed to  a separate basin.

    One  motivating force behind use of a constructed wetland  on this site was concern over potential impacts on a natural
wetland (Celery  Marsh) adjacent to the property. In addition to the constructed wetland, this site includes: elimination of
a  proposed auto  care center,  abstinence  from  chemical ice-clearing  methods, and construction of additional  ponds to treat
stormwater runoff that could  potentially include harmful pollutants.  The constructed wetland  receives runoff primarily from
the 4.1 ha commercial parking lot, as well as minor additional input from an adjacent store, local access roads, and US
Highway 52.

    The golf course created wetlands are part  of Purdue's  new Kampen Golf Course and are positioned to  intercept both
runoff from much of the golf course and  the adjacent urban  area. The developed area includes two residential highways,
a section  of state highway, the parking lot of a motel, a gas station, and 200 residences. The water flowing  through the
Kampen  Course eventually enters Celery Marsh, but prior to reconstruction this water flowed directly through drainage
tiles and  overland transport to the marsh, with  no treatment. The golf course  constructed wetlands  serve several purposes:
providing  a water hazard and aesthetic component of the  course, and enhancing environmental  quality that  can  also be
used in environmental education. Runoff from the urban area travels through three constructed wetlands prior to leaving
the course. One particularly  notable aspect of these constructed wetlands is that they have flourished even during long
dry summer periods. Frequent watering of the greens and fairways,  common on most courses, has the added advantage
that it provides runoff and tile drainage to the wetlands throughout the summer.

Methodology

    To determine the  effectiveness of each constructed wetland in trapping potential pollutants, water samplers were
installed  at the inlet and outlet of the commercial constructed wetland (Figure 1), and at six locations in the golf course
constructed wetland complex to track the progress of water as it enters the course, moves through the wetland system,
and exits to the Celery Marsh. The samplers are equipped with ISCO® Submerged Probes that measure water  levels,  used
in conjunction either with  a  weir or pipe of known geometry. The sampler uses these levels and the corresponding
geometry of the sampling sites to calculate the flow into and out of the wetland.  Each  sampler also  has a YSI® 600 Multi-
Parameter Water Quality Monitor that measures dissolved oxygen, conductivity, temperature,  and pH. The samplers record
flow and water quality parameters every five minutes and are programmed to take water samples during  storm  events.
Storm sampling  is triggered  in most cases by a  change in water level, and at two locations, by rainfall  intensity as
measured with an automatic tipping bucket rain gauge. The trigger points were determined empirically, so that inlet and
outlet samplers begin to sample at approximately the same time. The sampling programs for each  sampler are split into
two sections. The interval of time between samples in  part A of each routine  is closer  together than those in the
corresponding part  B routines, so that sampling occurs more often during  the "first flush." After that, the  second stage
                                                      305

-------
Figure 1. Sampling equipment at a constructed wetland. The laptop computer is downloading monitoring data from the sampler, and in the foreground
is a set of 24 sample bottles for storm sampling.

of each routine samples at larger intervals to guarantee samples at times coinciding with the downward slope of the
hydrograph.

    Overall, the design of the experiment is to track flow and water quality into and out of the constructed wetlands
continuously, both during storms and between storms, for a multi-year period. This allows for determinations of storm,
seasonal and multi-year trends in constructed wetland  trap efficiency. Trap efficiency can be defined in a  number of ways,
depending on the likely application of the results. In this work we are interested in concentration  trap efficiency (percentage
change in potential pollutant  concentration between the inlet and outlet,  both maximum and average values) and load trap
efficiency (percentage  change in potential pollutant  load between the inlet and  outlet for given points in a storm, for storm
totals,  and seasonally  and  annually). Selected samples from each  precipitation event  are  analyzed by a Purdue University
laboratory for total suspended solids (TSS), hardness, total Kjeldahl nitrogen (TKN), and total phosphorus (TP). These
parameters are the same as  those measured for seven other local sites as part of a larger analysis of water quality in rural
and urban settings. In addition to the analyses performed at the Purdue laboratory, more complete chemical scans are
performed  once per season on selected samples by Heritage Environmental Services in Indianapolis, Indiana. The
selection of tests is based on the pollutants that might reasonably occur at each site. The reason for this more complete
scan is to determine whether any potential pollutants not routinely measured at the Purdue laboratory show up at unusually
high levels. Any parameters that were not detected  in the Heritage samples could potentially be excluded from future
testing, but those parameters considered to be problems would need to be monitored on a consistent basis in the future.

Results and  Discussion

    To illustrate possible types of analyses and some major trends in the performance data, without reviewing the entire
data sets available, this discussion includes three examples  from the two sites. These include a complete storm record
at the commercial site, between-storm sampling at the commercial site, and first-flush storm sampling at the golf course
site.

Sample Storm at the  Commercial Site

    A 0.97 cm-storm occurred on 26 October 1997, with a double peak in intensity (Figure 2).  As expected, the wetland
acts to damp  peak flows, so  discharge values at the outlet slightly lag those at the inlet and are lesser in magnitude.  Water
temperature in the constructed wetland inlet is high and uniform (no diurnal variations) prior to the storm (Figure  3),
                                                      306

-------
                             0.35 r—
                             0.30-
                             0.25'
                        Total
                        Hourly
                        Rainfall
                        (cm)
                                                                          Average
                                                                          Hourly
                                                                          Discharge
                                                                          (103 cmr/s)
                             0.10"
                             0.05 ''
                             0.00
                                 0   5   10  15  20  25  30   35  40   45   50  55  60  65   70

                                           Time afttr Midnight, 25 October 1997 (hours)



Figure 2.  Rainfall and inlet and outlet runoff records for a storm event at the commercial site constructed wetland.

                         14.0 T	
                         12.0
                      I
5'
<
                         10.0
                          80 -
                          60
                         4.0
                          20
                          0.0
                                     DO (Inlet)
                                     DO (Outfel)       \
                                     Temperature (Inlet) •
                                  	Temperature (Outlet);
16.0


140


12.0 £


10.0\


8.0  ft
    "Z
     »
6.0  £
     &
4.0   u
    <

2.0


0.0
                             0   5   10   15  20   2i  JO  35  40   4$   50  55 60  65  70

                                       Time after Midnight, 25 October 1997 (hours)
Figure 3. Dissolved oxygen and temperature records for the storm shown in Figure 1.

drops 6°C during the storm, and slowly climbs back2°C in the 20 hours after precipitation stops. The outlet temperature
shows a 4°C  diurnal cycle  prior to the storm, and a lower amplitude cycle  after the storm. At the same time, the dissolved
oxygen (DO)  values climb  during the storm  (Figure 3).  Inlet DO values vary within the 0 to 4 mg/L range before the storm,
jump up to 9 to 13 mg/L during the storm, and fall during the 24 hours following the storm event.  The outlet DO varies from
1 to 8 mg/L prior to the storm, is very stable between 8 and 9 mg/L during the storm, and has strong variations from 3 to
9 mg/L post-storm.  High DO values during the storm are due to the increased mixing of the water, which causes oxygen
to be introduced to the wetland, as well as the  addition of "new" water that is  higher in oxygen to the stagnant water.
                                                          307

-------
50

45

40

35




25

20

15

10

 5

 0
                                                                                Inlet
                                                                                 I Outlet
                              50     100     150     200     250     300     350     400    450

                                          Time after Start of Sampling (minutes)
                                  30                    200                  400

                                         Time after Start of Sampling (minutes)
Figure 4. Total suspended sediment load for the storm shown in Figure 1 (upper), and calculated trap efficiency (TE) if only grab sampling had been
used (lower). The actual load TE for the storm was 84%.
    Initially, hardness data from the 26 October 1997 storm showed much higher values for the outlet than the inlet. This
discrepancy relates to the movement of water through the wetland.  Soon after the start of a storm, water begins to flow
over the weir at the  inlet and inlet  sampling begins. This new runoff from the site has a low hardness, reflecting  the
naturally low hardness of rainwater. At the outlet, high hardness values show that the water initially being sampled is not
new inlet water being displaced from the wetland; rather it is water that was stored in the concrete outlet box prior to the
storm. Hardness values at the outlet fall throughout the storm, showing that hardness is lower in the wetland than in the
                                                        308

-------
concrete outlet box. Because hard water occurs when concentrations of Ca2+ and Mg2* are high (Zumdahl,  1989), the
concrete surroundings themselves can add to the hardness of the water. Ninety-five percent of all concrete used is made
from Portland cement,  which  is prepared  using finely  ground  limestone (Mindess and Young,  1981). The cement  is made
into concrete using admixtures such as calcium chloride. Also, the interaction of the surface water with groundwater, which
in this geographic location is very hard due to infiltrating rainwater that dissolves the calcite-rich till from limestone in the
region, can add to the hardness of the water in the wetland (Davis and Cornwell, 1991).

    The hardness data suggest that the outlet record probably does not include water that enters and exits the wetland
during the same storm. Other studies (e.g. Bhaduri, et al., 1997) have shown through chemical load distributions that the
inflow and outflow from a basin are actually two different water masses, except in extreme storm events. The only way
that the same water could appear at the inlet and outlet during the same storm would be if the inflow sheeted over the
water in the  basin, arriving at the outlet without significantly mixing with the water stored in the wetland prior to the storm;
or if the storm  produced enough water to completely displace the volume of water in the wetland.

    Total suspended solids (TSS) concentration data and flow values for the storm are used to calculate TSS loading
values, which depict the effect of the basin in reducing the overall sediment load. Total loads depict the actual physical
amount of sediment entering the wetland and are important for planning activities such as dredging. TSS load values for
the  inlet are  larger than the outlet (Figure  4).  The inlet values  start high, dip down,  and then increase again. This indicates
that the initial runoff has "first flush" (high load) characteristics, and then the load input rate decreases. A second, lower
peak later in the storm could  be the result of the later pulse of higher rainfall intensity (Figure 2). The values  of TSS at the
outlet remain fairly uniform throughout the storm. The initial  value presumably represents between-storm ambient TSS
loading  in standing water in the wetland. During the storm, the increase in flow creates more turbulence, which can stir
up some of the bed sediment, slightly increasing the  TSS concentration and, therefore, the load. More importantly, though,
the outlet values are lower than those for the inlet; thus there is a net decrease in  TSS loading from the inlet to the outlet
for this particular storm. In one  sense, this traditional approach is a valid efficiency measure because the water going out
is compared  to that going  into the wetland, but in another sense it is a skewed picture because the new inlet water is being
compared with "old" outlet water that arrived in the basin during a previous storm (Bhaduri, et al., 1997).

      Multiple  storm sequence sampling will provide  a better view of overall trap efficiency (TE) than a single storm, just
as a complete storm record  is better than a grab sample. Standard grab  samples do not always lead to accurate trap
efficiency calculations (Figure 4).  If one sample were taken each from the inlet and the outlet at exactly  same time, the
data could show a very high trap efficiency (30 minutes), no trap efficiency (200 minutes), or a fairly high trap efficiency
(400 minutes).  The overall load  TE for this analysis was 84%.  This is one of the  reasons that this particular study samples
several times after the start of a storm - to bridge the gap between standard grab samples and actual events within the
wetland. Continuous monitoring provides a more complete record of the constructed wetland's activity, more accurately
depicting the trap efficiency of the wetland. From conductivity  data, during the monitoring period, 137  kg of dissolved load
entered the  basin, and 59 kg left the basin, for  a total dissolved solids (TDS) load TE over the storm of  57%. Further
analysis of many  storms can be  used to determine  an overall trap efficiency over longer periods of time. This type of
analysis could  be used  to determine the  effects  of different storm intensities, seasonal variations,  and  increased
urbanization in the area.

Detailed Chemical Scan at the Commercial Site: Between Storm Conditions

    Samples for a detailed  chemical scan  were taken on 17  December 1997 using the sampler's  grab sampling
mechanism. At this time, there had not  been a  precipitation event in a  couple of weeks, so these samples  represent
between-storm conditions in  the wetland. Although these samples were tested for many possible pollutants, only a few
were detected (Table 1).

    Parameters which show reductions between the inlet and  the outlet were chloride, sulfate, ammonia nitrogen, calcium,
magnesium,  sodium, silicon, strontium,  and total dissolved solids (TDS). For instance,  chloride levels fell from 210 to 160
mg/L, calcium levels fell from 95 to 54mg/L, and strontium levels fell from 0.16mg/Lto below the detection level of 0.10
                                                     309

-------
Table 1.  Detailed chemical scan of the commercial site constructed wetland. All values are mg/L.

         Parameter                              Inlet                           Outlet                 Detection Limit


Chloride                                      210                             160                        2.5
Sulfate                                       49                               37                         1.3
Nitrogen,  Nitrate-Nitrite                          0.12                             0.11                        0.01
Nitrogen, Ammonia                             0.21                             0.14                        0.12
Chemical Oxygen Demand                      18                               28                         10
Aluminum                                    BDL                             1.3                        0.10
Calcium                                      95                               54                         0.10
Iron                                         0.33                             1.7                        0.70
Potassium                                    2.5                              2.7                        0.10
Magnesium                                   28                               15                         0.10
Manganese                                   0.32                             0.24                        0.10
Sodium                                      110                             85                         0.10
Silicon                                       5.1                               3.0                        0.10
Strontium                                    0.16                             BDL                        0.10
Total Organic Carbon                           BDL                             4.2                        7.0
Total Phosphorus                              BDL                             0.10                        0.03
Dissolved Solids                               720                             490                        10
Total Suspended Solids                         4                                13                         1

Notes: Italicized parameters are those which have an outlet value > inlet value.
BDL = below detection limit
mg/L. Also, TDS levels fell from 720 to 490 mg/L. When compared to the values calculated using conductivity data from
the sampler, these values are slightly  higher than the values calculated for the 26 October 1997 storm event. The
maximum TDS values calculated for the inlet and the outlet were, respectively, 568 and 365 mg/L, with average values
around 337 and 263 mg/L. The higher between-storm values could be a result of the ability of sediments to dissolve in
the wetland waters. Reductions in values between the inlet and the outlet indicate removal of certain pollutants within the
wetland and also suggest that at the beginning  of a storm, the  outlet values will be lower than those  of water near the inlet.
Because of this,  the best  TE should  be at the start of a storm,  which is  shown in the 26 October 1997 storm  chemical data.

    Not all of the detectable parameters were lower at the outlet than at the inlet. The ones that were actually larger at
the outlet than at the  inlet were:  chemical oxygen demand  (COD), aluminum, iron, potassium, total organic  carbon (TOC),
total phosphorus (TP), and total suspended solids (TSS). The increase in TSS is interesting, and may be the cause of
increases in  adsorbed pollutants. This could  be attributable  to the lack of growth of plants in the middle of December.
Plants slow flow within the wetland, allowing sediments in the water to settle, and  plants have the ability to take into their
roots pollutants carried by the sediments (Pond, 1995). Because of this, as the plants die, they  may release  the sediments
and pollutants trapped earlier in the year,  as well as releasing  products of the decay of the organic matter. Aluminum, iron,
potassium, and  phosphorus could have been attached to  these sediments, especially the finer particles. Findings such
as these agree with previous studies that noted a distinct reduction in the performance of stormwater wetlands in winter
months (Oberts,  1994;  Ferlow,  1993). Not only does plant death have  an effect, but also the formation of ice on the  water
surface can scour the margins and resuspend the sediments and the pollutants that they carry (Oberts, 1994).

Detailed Chemical Scan at the Golf Course Site

    First flush samples were collected for detailed chemical analysis during the first  pulse  of runoff from  a storm in
November 1998 and a second storm in June 1999 (Table 2). In November 1998, 14 water quality parameters declined
in terms of a comparison between the urban input (Site 1) and the golf course  output (Site 6). Four water quality
parameters improved between the urban  input and the water exiting the course  during the same storm.  This suggests that
the constructed wetlands were not working well soon  after initial  construction,  during the late  fall.  However, key
parameters such as ammonia and nitrate-nitrite nitrogen and pesticide  levels were  either  decreased as  the watercirculated
through the golf course wetlands or were not detectable at either sampling site.

                                                       310

-------
     A distinctly different pattern of results is apparent in the June 1999 sampling (Table 2). Fifteen water quality parameters
 improved between the urban input and the water exiting the course, and only 4 parameters declined. This suggests that
 the golf course's created wetland system is functioning well to improve the water quality in the late spring when wetland
 plants have become established.  Two parameters of particular interest for a golf course are nitrate-nitrite N and ammonia-
 N, which were undetectable in water exiting the course, but at 2.1  and 31 ppm, respectively, in water flowing onto the
 course.

 Table  2.    Detailed  chemical scan of the  golf course  site  constructed wetland,  selected parameters.    All values are mg/L.

                                           November 1998                                  June 1999
Parameter
Simazine
Atrazine
Oil and Grease
Chloride
Sulfate
Nitrogen nitrate-nitrite
Ammonia nitrogen
Chem. 0, Demand
Mercury
Total Organic Carbon
Phosphorus
Dissolved Solids
Suspended Solids
Silver
Aluminum
Arsenic
Calcium
Cadmium
Chromium
Copper
Iron
Potassium
Magnesium
Manganese
Molybdenum
Sodium
Nickel
Lead
Selenium
Silicon
Tin
Titanium
Zinc
Detection
limit
0.10
0.10
5
2.5
2.5
0.01
0.12
10
0.0002

0.03
10
1
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.10
Site 1
Urban
runoff
BDL
BDL
BDL
8.6
11
1.1
0.23
40
BDL
8.2
0.19
91
17
BDL
0.31
BDL
29
BDL
BDL
BDL
0.51
2.3
7.1
BDL
BDL
4.5
BDL
BDL
BDL
2
BDL
BDL
BDL
Site 6
Created
wetland outlet
BDL
BDL
BDL
22
55
0.06
BDL
37
BDL
10
0.17
270
290
BDL
5.8
BDL
61
BDL
BDL
BDL
4.7
7.8
24
0.21
BDL
6.8
BDL
BDL
BDL
14
BDL
0.14
BDL
increase/
decrease
BDL
BDL
BDL
+156%
+400%
-95%
-52%'
-8%
BDL
+22%
-11%
+197%
+1606%
BDL
+1771%
BDL
+110%
BDL
BDL
BDL
+822%
+239%
+238%
+133%
BDL
+51%
BDL
BDL
BDL
+600%
BDL
+56%*
BDL
Site 1
Urban
runoff
BDL
0.1
BDL
32
18
2.1
31
480
BDL
240
0.32
240
8
BDL
1.8
BDL
40
BDL
BDL
BDL
1.6
2.2
9.9
0.28
BDL
6.5
BDL
BDL
BDL
2.0
BDL
BDL
0.38
Site 6
Created
wetland outlet
BDL
BDL
BDL
20
31
BDL
BDL
25
BDL
1.6
0.08
220
2
BDL
0.16
BDL
34
BDL
BDL
BDL
0.26
0.37
28
BDL
BDL
8.7
BDL
BDL
BDL
4.8
BDL
BDL
BDL
increase/
decrease
BDL
-91 %'
BDL
-38%
+72%
-100%"
-1 00%*
-95%
BDL
-99%
-75%
-8%
-75%
BDL
-91%
BDL
-15%
BDL
BDL
BDL
-84%
-83%
+183%
-64%
BDL
+34%
BDL
BDL
BDL
+140%
BDL
BDL
-74%'
BDL = Below Detection Limit
.  where contaminant was BDL, the detection limit was used for % increase/decrease calculations
    No unusually high levels of any  of a wide array of potential pollutants,  including  pesticides and metals, were detected
at the  golf course sampling  sites. However, atrazine was detected in water  exiting the neighborhood and entering the golf
course (Site 1). Surprisingly, even from the urban runoff there was no measurable oil and grease. It is reassuring to note
that heavy metals of concern, such as mercury and lead, are below detection limits in all samples.

Conclusions

    Constructed  wetlands  can potentially  be used to improve NPS pollution management in urban areas. Although  finding
space  for constructed wetlands can be a challenge  in developed areas, these management tools can be incorporated into
the design of new or renovated commercial and industrial facilities. In some cases, they can be added  to recreational

                                                       311

-------
facilities such as parks and golf courses.  In each of these cases, good initial design and attention to continued water
supply for long-term wetland survival is critical.

    The constructed wetland monitoring program in West Lafayette, Indiana, includes both commercial and golf course
constructed wetlands. Selected results presented  here illustrate the  complexity of developing a program to evaluate
performance of such wetland systems.  Traditional grab sampling can  provide misleading results compared to continuous
sampling, and it is clear that apparent trap efficiency varies both within storms as well as between seasons. The type of
complete picture of constructed wetland performance that is needed to improve design and enhance understanding of
chemical  and biological processes  in constructed wetlands can  be  approached by  continuous  monitoring  through several
years. Initial data suggest that the  constructed wetlands studied here are generally performing well to reduce loads and
concentrations of a range of urban NPS pollutants, particularly during spring and summer storm events after wetland
vegetation has become established. However there is also a strong indication that trap efficiencies are much lower, and
in some cases negative, during winter months. The implications of this depend on the context provided by the receiving
area.

    Constructed wetlands also provide important benefits beyond water quality control. They  provide aesthetic diversity
in urban settings, they  represent islands of habitat types that are generally absent or underrepresented in  older developed
areas,  and they provide important local educational resources in urban areas. Overall, constructed wetlands should be
considered as a  potential element of urban retrofit projects, if there are situations where water supply is available to
maintain wetland hydrology.

Acknowledgments

    The Purdue  constructed wetland monitoring program would  not be possible without generous support from the
Showalter Trust,  the United States Golf Association, Pete Dye,  Inc., and Heritage Environmental. Equally important for
the  project has been the willing  cooperation of Jim Scott, superintendent of the Birck Boilermaker Golf Complex, Wal*Mart
Inc., and numerous undergraduate and graduate field assistants.

References

Bhaduri, B. L, Harbor, J. M., and Maurice, P. (1997) Chemical load fractionation  and trap efficiency of a construction site
storm water management basin. In Environmental &  Engineering Geoscience. 3(2): 235-249.

Chang, G. (1994) First flush of stormwater  pollutants investigated  in  Texas. In Watershed Protection Techniques. Thomas
R. Schueler, Editor. Herndon, VA: Center for Watershed Protection.  l(2): 88-89. From Chang, G.,  J.  Parrish,  and  C.
Souer. (1990) The first flush of runoff and its effect on control structure design. Environmental Resource Management
Division of Environmental and  Conservation Services. Austin, TX.

Davis,  M.  L. and Cornwell,  D. A. (1991) Introduction to Environmental Engineering. New York: McGraw-Hill, Inc.

Ferlow, D.  L. (1993) Stormwater runoff retention and renovation: a back lot function or integral part of the landscape?
In Constructed Wetlands for Water Quality Improvement. Gerald A. Moshiri, Editor. Boca Raton, FL: Lewis Publishers,
Inc. [CRC Press, Inc.] pp. 373-379.

Hicks,  D. B. and Stober, Q. J.  (1989) Monitoring of constructed wetlands for wastewater. In Constructed Wetlands for
Wastewater Treatment: Municipal, Industrial, and Agricultural.  Donald A.  Hammer, Editor.  Chelsea,  Ml:  Lewis
Publishers, Inc. pp. 447-455.

Horner, R. R., Skupien, J. J., Livingston, E. H., and Shaver, H.  E.  (1994) Fundamentals of Urban Runoff Management:
Technical and Institutional Issues. Washington D.C.: Terrene Institute, United States Environmental Protection  Agency

Livingston, E. H. (1989) Use of wetlands for urban stormwater treatment. In  Constructed Wetlands  for Wastewater
Treatment: Municipal,  Industrial, and Agricultural. Donald A. Hammer, Editor. Chelsea,  Ml: Lewis Publishers, Inc. pp.
253-262.
                                                     312

-------
Mindess, S. and Young, J. F.  (1981) Concrete. Englewood Cliffs, NJ: Prentice-Hall, Inc.

Mitsch, W. J. and  Gosselink, J. G. (1993) Wetlands. New York: Van Nostrand Reinhold.

Oberts,  G.  (1994) Performance of stormwater ponds and wetlands in winter. In Watershed Protection Techniques.
Thomas R. Schueler, Editor. Herndon, VA: Center for Watershed Protection. l(2): 64-68.

Perry, J. and Vanderklein, E. (1996) Water Quality: Management of a Natural Resource.  John  Lemmons, Consulting
Editor. Cambridge, MA:  Blackwell Science.

Pond, R. (1995) Pollutant dynamics within stormwater wetlands: plant uptake. In Watershed Protection Techniques.
Thomas R. Schueler, Editor. Herndon, VA: Center for Watershed Protection. l(4): 21 O-21 3.

Tatalovich, S.K. (1998). Chemical  and  Sediment Trap Efficiency of a Stormwater Detention Biofiltration Pond: A Study
of the Wal*Mart Constructed Wetland.  Master's Thesis, Civil Engineering, Purdue University.

Witthar,  S. R. (1993) Wetland water treatment systems. In Constructed Wetlands for Water Quality Improvement. Gerald
A. Moshiri,  Editor.  Boca Raton, FL:  Lewis Publishers, Inc.  [CRC  Press, Inc.]. pp. 147-155.

Zumdahl, S. S.  (1989) Chemistry. Lexington, MA:  D.C. Heath and Company.
                                                   313

-------
             Advanced Identification (ADID) Techniques  Used to Protect
           Wetlands and Aquatic Resources in  a Rapidly Growing County
                                             Dennis W. Dreher
                                  Northeastern Illinois Planning Commission
                                           Chicago,  Illinois 60606
Abstract
    McHenry County, Illinois, approximately 40 miles northwest of Chicago, is one of the fastest growing counties in the
state.   It also is home to many valued wetland  and  stream communities that are threatened by the impacts of new
development. Because of this, county government officials, with funding from the U.S. Environmental Protection Agency,
sought the assistance of resource experts from various local, regional, state, and federal agencies. Their task was to
assess the quality of the aquatic resource and to  develop strategies for improved wetland protection. The project initially
involved the development of an up-to-date countywide inventory of wetlands, lakes, and streams. This inventory showed
that over 11% of the county was covered with wetlands and waterbodies. These aquatic resources then were evaluated
and rated based on the habitat, water quality, and stormwater storage functions that they provided. While only a small
fraction of the total number of wetlands — about 11% — of the county's wetlands were designated high quality, these
wetlands represented over 60% of the total wetland  acreage.  Inventory and assessment data were transferred to a
customized CD-ROM mapping tool to provide ready access to project information by resource managers, planners, and
local governments. The project team also developed a protection strategy for aquatic resources that was tied to the
results of the  Advanced  Identification  (ADID) study. In  particular, it  identified four critical protection  components:
improved education, regulations and best management practices, acquisition, and restoration. Though the ADID study
is only recently completed, there are strong indications that this protection strategy is being taken seriously by officials
in the county.

Background

    McHenry County, Illinois lies approximately 40 miles northwest of Chicago along the Wisconsin border.  Reflecting
a history  of glacial activity, McHenry County possesses an abundance of wetland types in a variety of physical settings.
Predominant wetland types include palustrine, lacustrine, and riverine systems. Palustrine wetlands are found in a wide
variety of geographic settingsand terrains in the county and include marshes, bogs, fens, wet prairies, forested wetlands,
and ponds. Lacustrine wetlands are less common. They are found mostly in eastern portions of the county and  are
exemplified by the wetlands of the Fox River - Chain 0' Lakes. High quality rivers and streams, and associated riverine
wetlands, are relatively common. In fact, McHenry County has some of the highest-quality stream ecosystems in Illinois,
as exemplified  by  the Kishwaukee River and its  tributaries.

    While predominantly  rural, McHenry County is  one of the fastest urbanizing  counties in the state. From 1990 to 1998,
its population grew by 31.5% to 240,945. Population is forecast to grow to nearly 362,000 by 2020, an increase of an
additional 50%. This rapid population growth has raised  concerns over possible adverse  impacts  on the  county's
wetlands, lakes, and streams.

    Historically, wetland and stream protection measures in McHenry  County have included federal regulations, local
government ordinances, and acquisitions by government agencies, primarily the McHenry County Conservation District
and the  Illinois Department of Natural Resources.   However, with  the recent rapid  pace of urban development,
unacceptable loss and/or degradation of wetlands and aquatic resources have been observed. Concerned over the
possible  environmental  effects of rapid growth,  the county board invited the U.S. Environmental Protection Agency
Region 5 (EPA) to perform an "advanced identification" (ADID) study of its wetlands and  aquatic resources.
                                                   314

-------
    An  ADID  project can have  several objectives. One objective is  to shorten permit processing, while providing
increased predictability to the Corps of Engineers  regulations under Section 404 of the Clean Water Act. ADID also
provides information that can be used by state and local governments to aid in zoning, permitting,  or land acquisition
decisions. Another objective of ADID is to provide information to agencies, landowners, and private citizens interested
in restoration or acquisition of aquatic sites.

Approach

    The ADID study was initiated  in 1995, under the coordination of  the Northeastern Illinois Planning Commission
(NIPC). The study was a cooperative effort among federal, state, and  local agencies to inventory, evaluate, and map
high-quality wetland and aquatic resources in the county.  From the federal perspective, the primary purpose of this ADID
study was to designate wetlands  or other waters of the United States that are unsuitable for discharge of dredged or fill
material. From the local perspective, the  purposes of ADID were to  improve the  overall  protection  mechanism for
wetlands via  improved  local regulation, improved  predictability in  the permitting  process, identification of potential
mitigation/restoration sites, and  identification of potential sites for acquisition.

    The scope of work for the ADID project included the following tasks:

    .  form a Technical Advisory Committee and a  Planning  and Policy Committee;

    .  develop ADID objectives for McHenry County and  a strategy for protection and management;

    .  identify and map existing wetlands and  aquatic resources;

    .  develop  an evaluation methodology for  identified functions of wetlands and aquatic resources;

    .  apply evaluation  methodology utilizing Geographic Information System (GIS) technology and field inspection;

    .  map ADID sites for public  review;

    .  develop a CD-ROM tool that contains both the project data and a customized GIS interface for display, query, and
      mapping; and

    .  produce a final report and  brochure and  conduct a workshop for local governments, landowners, and consultants.

    The Technical Advisory Committee and the Planning and Policy Committee were formed soon after initiation of the
project. There were two general  purposes for these  committees: 1) provide technical  and policy assistance to NIPC and
EPA,  and 2) provide a forum for educating local interest groups regarding the value of wetlands and aquatic resources
in the county.

    The principal  role of the Technical Advisory Committee was  to advise project staff on scientific issues, particularly
the  development  of evaluation methodologies for wetlands, lakes, and streams. Technical committee members provided
expertise in wetland biology, soil  science, hydrology, engineering, water quality, and computerized mapping. Technical
committee members contributed substantial time evaluating wetlands, both  in the office and  the field. The technical
committee consisted of the following invited agencies and  organizations:

    .  U.S.  Environmental  Protection Agency,  Region 5

    .  U.S.  Army Corps of Engineers, Chicago District

    .  U.S. Department  of the Interior, Fish and Wildlife Service,  Chicago-Metro Wetlands Office

    .  USDA, Natural Resources Conservation Service

    .  Illinois Department of  Natural Resources

                                                     315

-------
    . McHenry County Department of Planning  and Development

    . McHenry County  Conservation District

    . McHenry County Soil and Water Conservation District

    . Fox Waterway Agency

    . Northeastern Illinois Planning Commission

    The principal  role  of the Planning  and Policy Committee was to advise  project staff on policy, particularly the
determination of wetland functions important to  McHenry County. The policy committee also provided advice on the
development of a  wetland protection and management strategy. The  policy committee included all of the members of
the technical committee as well as members of the following organizations:

    . Homebuilders Association of Greater Chicago

    . Illinois Audubon Society, McHenry  County Chapter

    . Illinois Environmental  Protection  Agency

    . Land Foundation  of McHenry County

    . McHenry County  Board

    . McHenry County Defenders (a local environmental group)

    . McHenry County Farm Bureau

    . McHenry County  Municipal Association

    . McHenry County  Realtors Association

    • McHenry County  Stormwater Committee

    . Openlands Project (a regional open space advocate)

Developing an Aquatic Resource Inventory

    Detailed inventories of wetlands, lakes, and streams were developed early in the project. Two principal existing
inventories were considered  for identifying and mapping wetlands: the National Wetland Inventory (NWI) developed  by
the US. Fish and  Wildlife Service with the assistance  of the Illinois  Department  of Conservation  (1986) in the early
1980s, and an inventory by the Natural Resources Conservation Service (NRCS) that was being completed in McHenry
County just as the ADID project began. While neither inventory was adequate alone (the NWI was becoming dated and
the NRCS inventory, while more recent, and only a data set, focused  principally on agricultural areas), in combination
they served as a good starting point. In finalizing the inventory, numerous revisionsand improvements were made based
on reviews of aerials photos, field checks, and the knowledge of local experts.

    The resultant  inventory identified 2,535 wetlands, including lakes, covering 37,846 acres. The  inventory identified
an additional 1,250 farmedwetlands covering 3,839 acres. In total, there were 3,785 wetlands in all categories covering
41,685 acres, or nearly  11% of the county.

    Lakes were identified as  a subset of wetlands. Specifically, lakes were distinguished based on a criterion of 20 acres
or more of open water.  Fifteen such lakes were  identified (excluding gravel pits).
                                                   316

-------

Type
Wetland
rarmed Wetland
Lake
River
Total

Number
2,518
1,250
15
2
3,785 |

Acreage
33,003
3,839
3,584
1,259
41,685
Percent of
County
Area
8.4
1 .0
0.9
0.3
10.6
    Streams were identified and mapped based on an inventory developed by the EPA. EPA's Stream Reach File,
Version 3 (1 :1 00,000 scale), with minor revisions, provided an accurate and complete inventory of county streams. The
inventory included over 570  miles of streams ranging in size from small, unnamed headwaters to large rivers like the Fox
and  Kishwaukee.

Evaluation  of the Functions  and Quality of Aquatic Resources

    As the first step in developing a wetland evaluation methodology, members of the policy committee were asked to
identify wetland and aquatic resource functions that were important to McHenry County. After considerable discussion,
the committee  recommended that the  following functions  be considered and  evaluated: biological/habitat functions, water
quality mitigation functions, stormwater storage functions, and groundwater functions.  These functions then were
evaluated and refined by the technical committee. Ultimately, it was concluded that groundwater functions of wetlands,
while having important water supply implications, could not be evaluated because of insufficient data.

    The project team and advisors then  proceeded to develop evaluation criteria and methodologies for the following
general categories: biological/habitat functions  and  water quality/stormwater storage functions. The development of a
methodology for  identifying  high-functional-quality wetlands  in  McHenry County relied  both on  existing wetland evaluation
methodologies and the technical expertise of members of the technical advisory committee. The resultant methodology
builds on a  methodology used in nearby Lake County,  Illinois (Dreher, et al., 1992) as well as other documented
methodologies, particularly the Wetland Evaluation Technique  (WET) manual (Adamus et  al., 1987), the Oregon Method
(Roth et al.,  1993), and the Minnesota manual (U.S. Army Corps of Engineers, 1988).

    The methodology was designed to accomplish two objectives: 1) identify the functions that individual wetlands were
performing, and 2) identify wetlands of such high quality that they merit special consideration for protection strategies.
The  evaluation of the identified functions for individual wetlands can be very complex and some of the referenced
methodologies describe fairly elaborate approaches to perform thorough evaluations. However, because of the large
number of wetlands to be considered  in  McHenry County, it was necessary to adopt a  simpler evaluation procedure.
The resultant methodology isfullydocumented in the final project report, "Advanced Identification (ADID) Study, McHenry
County, Illinois" (NIPC et al., 1998).  An overview of the evaluation criteria follows.

    Biological  functions  include wildlife  habitat, floristic  diversity,  stream aquatic  habitat,  and  lake aquatic habitat.
Wetlands were considered high qualitytor this function if they met one of several criteria. These criteria included:

    .  the presence of threatened or endangered plant or animal species;

    .  designation in the  Illinois or McHenry County Natural Areas Inventory (NAI);

    .  field evaluation as a grade A, B, or C wetland community following NAI methods;
                                                     317

-------
    . streams with Index of Biotic Integrity (IBI) scores of 41 or greater;

    .  streams with high quality physical habitat;  and

    .  healthy lake ecosystems with rich/diverse fish and  plant communities.

    Wafer quality/stormwater storage functions include shoreline and streambank stabilization, sediment and toxicant
retention, nutrient removal and  transformation, and stormwater storage and hydrologic  stabilization. In  order to be
designated  high functional value  for water quality/stormwater functions, wetlands were required to meet three of the four
following criteria:

    .  presence of stabilizing vegetation  adjacent to an open waterbody  or perennial stream;

    .  surface  area larger than five acres and  having characteristics indicating the propensity for sediment/toxicant
      retention;

    . surface area larger than five acres, upstream of a lake or impoundment, and having characteristics indicating the
      likelihood of nutrient removal/transformation; or

    .  surface  area larger than  five acres, at least  50%  outside the floodplain, and  having characteristics  indicating
      significant stormwater retention.

    Alternatively,  wetlands could  be designated high funcf/ona/va/ueforwaterqualityfunctions if they provided individual
water quality functions adjacent to or upstream of wetlands, lakes, or streams that provide high quality habitat.

    Individual wetlands and waterbodies were evaluated  using a three-step procedure of CIS screening; aerial photo,
map or desk-top  evaluation; and field evaluation  (as needed). Based on this evaluation,  it was determined that  154
wetlands totaling  17,489 acres, or about 42%  of the county's  entire wetland area, met the criteria for  high-quality habitats.
Most of the high-quality wetlands tended to be large parcels, averaging 114 acres in size in comparison to the average
wetland size of 11 acres countywide. An additional 274 wetlands totaling 8,292 acres  (average size of 30 acres)  met
the criteria for high value for stormwater and water quality functions.  Thus, while a relatively small number of wetlands
(about 11%) were designated high quality or high functional value, these wetlands represent  over 60%  of the total
wetland acreage.
Classification
High Quality
Habitat
High Functional
Value
High Quality Lake
Number
154
274
7
Percent of
all Wetlands
4.0
7.2
0.002
Acreage
17,489
8,292
1.346
Percent of
County Area
4.5
2.1
0.3
Percent of all
Wetland Area
42.0
19.9
3.2
    Of the 15 inventoried lakes, seven were determined to be high quality. A total of 572 miles of stream were evaluated
and 170 miles (or nearly 30%) were designated high quality. Interestingly, high-quality stream segments were found on
18  different named streams and rivers scattered throughout the county.

Using ADID for Protection and Restoration

    The ultimate measure of success for a project like the McHenry County ADID study is  how it contributes to the
protection and restoration of aquatic resources. With this in mind, the project scope included a work element to develop
                                                     318

-------
a strategy for protection and management of aquatic resources. With the assistance of the advisory committees, the
project team developed a four-part strategy involving:

    . improved education of local government officials,  landowners,  and the  public;

    . effective regulations and best management practices;

    . expanded acquisition  of aquatic sites and buffers; and

    . restoration of degraded sites.

    This strategy, which is described in detail in the project report  (NIPC et al., 1998), is summarized below. Also
described are some recent protection and management activities, although it is still too early to judge the long-term
success of the project.

    Improved Education: Educational initiatives are critical to improve awareness of wetlands and aquatic resources
among local citizens, land owners, and elected officials.  Improved awareness can enhance local support for protection,
acquisition, and  restoration programs.

    . A 12-page brochure, McHenry County's Wetlands, Lakes, and Streams, was developed to educate the public and
     local officials about the value of wetlands and  aquatic resources  in their communities. The brochure also
     discussed  the results of the ADID study and identified additional sources of information  and  agencies that can
     provide help.  Over 1000 copies of brochure have  been distributed by participating  agencies, such as the county
     soil and water conservation district.

    .  Maps and  information for all ADID sites were made available on a "user-friendly"CD-ROM. The CD-ROM includes
     simplified  mapping software developed from a  sophisticated CIS tool. The  software  enables querying and
     screening of various wetland characteristics at different geographic scales throughout the county.  It also enables
     printing out detailed information on individual wetlands. Over 100 copies of the CD-ROM  have been provided to
     local officials, consultants,  and landowners in the  county.

    . The  message  of wetland,  lake, and stream protection  also is  being  carried to local officials and the public by
     county-based environmental groups and consortiums called "ecosystem partnerships" that have been established
     for the two main river watersheds in the county (the Fox and the Kishwaukee). ADID will be a useful tool in aiding
     the efforts of these organizations.

    Effective Regulations: Effective regulations are needed to minimize the effects of  new development  on aquatic
resources. Specifically, improved regulations are needed to fill in the gaps in existing federal, state, and local regulatory
programs. It  was the conclusion of both the ADID team and the McHenry County Comprehensive  Stormwater
Managemenf Plan  (McHenry  County, 1996) that improved regulations are needed to address  concerns  such  as buffers
and setbacks, depressional storage volumes, pretreatment of stormwater runoff, and effective environmental mitigation
for  unavoidable disturbances.

    . Current federal regulations  authorized under Section 404 of the Clean  Water Act require a permit for the discharge
     of dredged or fill material into wetlands or other waters of the United States. Federal guidelines also authorize the
     EPA and the Corps of Engineers to identify in advance of specific permit requests, aquatic sites that will be
     considered as areas generally unsuitable for disposal of dredged or fill material. The Chicago District of the Corps
     has indicated  that it will  apply this discretionary authority to high-quality habitat and high- functional value sites in
     McHenry County. The  Corps also generally will require  an  individual permit  (which  allows  public input) for
     proposed modifications  of  ADID sites.

    . Stream and wetland regulations, based on a model ordinance developed by NIPC, also have been adopted by  a
     number of local governments in the county.  These regulations are intended to complement the federal regulations
     by discouraging development in buffers and setbacks adjacent to wetlands, lakes, and streams and requiring pre-

                                                     319

-------
      treatment  of  stormwater discharges.  The City  of Woodstock, the county seat,  recently applied  its  wetland
      protection  regulations in a  residential development review that resulted in  an innovative conservation design
      around a large wetland.  Not only will the wetland be avoided, but the site design calls for clustering of homes and
      buffers adjacent to wetland areas. Also, drainage swales and natural landscaping will be incorporated on upland
      portions of the site to reduce hydrologic and water quality  impacts of the development.

    .  ADID team members have worked  closely with staff and consultants to the McHenry  County Stormwater
      Committee in the development of a countywide ordinance for new development. It has been recommended that
      the  countywide ordinance include provisions for stream and  wetland  protection that complement,  but do not
      duplicate, federal regulations. While the ordinance adoption process has been challenged by financial constraints
      and political changes in the  county, it appears likely that significant stream and wetland protections will be added
      to existing county and municipal regulations.

    Acquisition: Acquisition of important wetlands and stream corridors is one of the best ways to assure their long-term
protection. In fact, recent experience indicates that these areas are becoming high  priorities for public land  acquisition.
Information developed in the ADID study, particularly the identification of high-quality habitats and high-functional-quality
wetlands,  will  be valuable to land acquisition agencies, including park districts,  the McHenry  County Conservation
District, the Illinois Department of Natural Resources, and local land trusts, in assessing acquisition priorities. In a recent
example, the Plan Commission of Nunda Township in east-central McHenry County is developing a comprehensive land
use plan that will utilize ADID maps to identify areas to be preserved as open space.

    Restoration: Restoration of degraded wetlands, lakes, and stream corridors, and ongoing management of higher
quality sites, are critical challenges for land management agencies. Management is  needed to counteract the effects
of disturbances such as site fragmentation, elimination of fire, invasive species, and hydrologic alterations. Notably, the
McHenry County Conservation District has  been a regional leader in restoring degraded streams and wetlands. The
ADID data base will  be very useful in identifying appropriate sites to continue this restoration. The availability of GIS data
bases and mapping, particularly in conjunction with other digital data such as soils maps and data on seeps and springs,
will  greatly facilitate this objective.

Lessons Learned

    ADID was a valuable experience in McHenry County that generally met its identified objectives. In considering ADID
studies in  other areas, there are several important lessons one can learn from the  McHenry County experience.

    1) Engage localgovernment sponsors and keep them informed throughout the project. The McHenry County ADID
       began after the county board passed a resolution soliciting EPA's assistance.  County staff and elected officials
      were invited  to participate  on  advisory committees. When support appeared  to waver at critical points in the
       process (e.g., staff changes and budget  difficulties at the county), the project team reached out by convening
      special  meetings reminding county officials of the benefits of the project with respect  to adopted  county objectives.

    2) Conduct an open study process involving  both traditional supporters of stream and weflandprofecfion efforts and
      potential adversaries. Groups ranging  from  environmental  organizations  to developers and  the agricultural
      community were invited to participate on advisory  committees where issues  and approaches were openly
      discussed. When a public meeting was held to present project results, over 200 individuals attended. The vast
       majority of those expressing opinions indicated support for ADID objectives and procedures, even though some
       had concerns over  the ramifications of federal wetland regulations.

    3) Utilize the expertise and local know/edge of federal, state, and local resource agencies. While EPA contracted
      with NIPC to coordinate the project, staff from numerous  resource agencies contributed invaluable expertise in
      hydrology, soils, aquatic ecology, and botany.  They also contributed  countless hours in  evaluating field sites.
      Scheduling such assistance from multiple agencies  resulted in some time  delays. However, without these
      voluntary  contributions,  the project could  not have been completed.
                                                     320

-------
    4)  Define  wet/and and aquatic resource functions from a multi-objective perspective. While there is a tendency
       sometimes to focus on just the habitat and recreational values of wetlands, lakes, and streams, it is important to
       consider a broader range of benefits to maximize local buy-in to the process. The McHenry County ADID
       specifically considered stormwater and water quality functions that were identified as being important in local
       plans, such as the McHenry County Comprehensive Stormwater Management Plan.

    5)  Distribute end-products in user-friend/y formats.  While ADID was a highly technical and complex project, efforts
       were made to  provide products that were readily understandable by local governments, land owners, consultants,
       and the public. The product receiving the most interest was the  CD-ROM  containing ADID data, as well as a
       user-friendly CIS-based interface for querying  and  mapping. The CD-ROM promises to be much more useful
       than conventional paper maps.

    6)  Engage the local press in covering the project. Limited attempts were made to inform the local  press during the
       course  of the study. While there was some resultant news coverage in local newspapers, particularly around the
       time of the public meeting, this coverage was not particularly effective in informing the public about the benefits
       of wetlands and the importance of the ADID study.  Focused efforts, such as targeted press releases, probably
       would  have improved the frequency and quality of coverage.

Conclusions

    The ADID study provides valuable information  to advance the protection and  restoration of wetlands and aquatic
resources in McHenry County.  It can aid residents and organizations desiring  to protect high-quality resources or restore
sites that have been degraded. It can inform landowners and developers about an appropriate course of action when
they are considering disturbances in  or adjacent to high-quality sites.

    While the final ADID products have been available  for only a short time, it is apparent that they will greatly facilitate
ongoing efforts to educate county residents and  officials, protect streams and wetlands from the effects of new
development, preserve sensitive stream corridors and  wetlands as public land, and restore degraded  sites. While the
ultimate success of county stream and wetland protection initiatives will depend on the will of landowners and local
government officials,  no one will be able to blame wetland loss on inadequate information.

References

Adamus, P., E. Clairain, R. Smith, and R. Young. 1987.  Wetland Evaluation Technique (WET). Department of the Army,
Waterways Experiment Station, Corps of Engineers, Vicksburg, Mississippi.

Dreher, D., S. Elston,  and  C.  Schaal.  1992. Advanced Identification (ADID) Study, Lake County, Illinois. For U.S.
Environmental Protection Agency, Region 5.

McHenry County  Stormwater Committee.  1996.  McHenry County Comprehensive  Stormwater Management Plan.
McHenry County,  Illinois.

Northeastern Illinois Planning Commission, U.S. Fish and Wildlife Service - Chicago-Metro Wetlands Office, and U.S.
Environmental Protection Agency,  Region 5. 1998. Advanced Identification (ADID) Study, McHenry County,  Illinois.

Roth, E., R. Olsen, P. Snow, R. Sumner. 1993. Oregon  Freshwater Wetland Assessment Methodology. Oregon Division
of State Lands.

U.S. Army Corps of Engineers, St. Paul District. 1988. The Minnesota Wetland Evaluation Methodology for the North
Central United States.

U.S. Department of the Interior, Fish and Wildlife Service, in cooperation with the  Illinois Department of Conservation.
1986.  National Wetlands Inventory.
                                                    321

-------
                  Local  Government  Involvement  in Mitigation  Banking
                                                Lisa T. Morales
                                               Wetlands Division
                                     U.S. Environmental Protection Agency
                                               Washington, D.C.
Abstract

    Mitigation banking is a valuable tool that can be used by local officials to achieve wetlands restoration and other local
goals. Mitigation banks can be established by local governments to provide compensation for wetland losses that result
from development projects. There are different strategies that local governments can use to establish a mitigation bank,
depending upon their goals and objectives. The success of a mitigation bank is dependent upon several factors, ranging
from bank location to the availability of funding. The Environmental Protection Agency conducted a survey of local
jurisdictions  to  identify the different strategies that  were utilized  for  effective development and  implementation  of
mitigation banking. The findings of the survey are presented in case-studies that characterize the approaches that were
used by local governments to achieve their mitigation banking goals and  objectives.

Introduction

    Mitigation  banking is  defined in the  Federal  Mitigation Banking Guidelines' as "wetland  restoration,  creation,  or
enhancement for the purpose of compensating for unavoidable wetland  losses in  advance of authorized impacts to
similar resources." Under  Section 404 of the Clean Water Act, applicants  for permits must first avoid and minimize all
impacts to wetlands and  other waters of the United  States. If there are still impacts, then applicants must provide
compensatory mitigation through the restoration, creation, and the enhancement of similar type of aquatic resources.
This "sequencing process" under the Section 404(b)(l) guidelines is a  central  premise of the Section 404 regulatory
program, and mitigation banking can play a role in providing compensatory mitigation for unavoidable wetlands losses.

    As a general  matter, on-site and in-kind  mitigation is preferred under the Section 404 Program.  However, in those
circumstances where it is  determined, on a case-by-case basis, not to be  practicable, then off-site, in-kind mitigation is
acceptable.'   Off-site mitigation can be  accomplished using a federally-approved mitigation  bank. Since the use  of
mitigation  banking to  offset permitted wetlands losses began in  earnest in the early 1990's, local governments have been
involved in developing banks to restore and replace lost wetlands  functions and values within their jurisdictions. By
simplifying the process for compensating for unavoidable wetlands losses, appropriate use of the banking concept can
improve  both permitting efficiency and environmental  protection.
1 Federal Guidance for the Establishment, Use and Operation of Mitigation Banks; Notice. Federal Register, Volume 60, No. 228, pages 58605-58614,
November 28, 1995.

2 Memorandum of Agreement between the EPA and the Department of the Army Concerning the Determination of Mitigation Under the Clean Water
Act Section 404(b)(l) Guidelines (February 6, 1990)

                                                     322

-------
Mitigation  Banking Objectives

    •  Local needs are best met when there are clear objectives and goals for establishing a mitigation bank. Examples
      of objectives for mitigation banking identified by local governments include:

    •  Use as part of a comprehensive watershed plan that addresses urban development and the need for preservation
      and restoration of wetlands

    •  Provide an incentive for economic growth by streamlining the process for providing compensation  of unavoidable
      wetlands impacts

    .  Provide for restoration of degraded wetlands that otherwise might not be improved because of insufficient funding

    .  Use as part of a multi-objective strategy to  manage stormwater, flooding, water quality, etc.

    •  Compensate for wetland losses from future local agency projects

Types of Mitigation Banks

    Once the objectives for the proposed bank have been established by the local agency, the type of bank to meet the
identified objectives must be determined.  Local governments can establish a mitigation bank for their individual use or
for credit sale as a commercial venture. The bank can be established solely by a public agency or as a joint venture with
a private entity (e.g., entrepreneurial business). Mitigation bank types need not be mutually exclusive; for example a
commercial mitigation bank may be established  through public/private partnerships and still be  part of a watershed
management plan.

    Commercialbanks: A commercial mitigation bank is one in which  the credits are sold to  a  party other than the bank
sponsor (banker). The banker sells  credits within the established service area to permittees who have approval from
the  U.S.  Army Corps of Engineers (Corps); the agency  responsible  for issuing wetlands development  permits; to
compensate for wetland impacts through  a mitigation bank.

    Sing/e-user: A single-user mitigation bank is established by a local agency to compensate only for wetland losses
associated with activities conducted by the agency.

Factors Contributing  to Mitigation Bank  Success

Conditions for successful bank establishment

    Local governments which have successfully  established a mitigation bank have identified several  conditions that
need to be considered in order to successfully meet environmental and/or economic objectives.

    •  There  must be a demand  for  compensatory mitigation  within  the local jurisdiction.  Demand results from
      development pressure in a rapidly growing  area where impacts to wetlands are expected to occur. The value of
      development in regions with rapid growth increases the willingness of the public agency or developer to  pay for
     wetland mitigation. Potential  bank sponsors should conduct a "needs" analysis to determine the demand for a
      mitigation bank in a given area. The analysis will show the extent of potential wetland impacts in the region and
     whether mitigation banking is a viable compensation option. Once demand for a bank is decided  upon, the size
      of the bank can be determined.
                                                   323

-------
      •  There must be a sufficient supply of available sites for restoration, creation, enhancement, or preservation.3 The
        availability of appropriate sites will vary by geographic region. For example in Florida, large tracts of degraded
        wetlands have been impacted by previous land-use practices and by invasion of exotic plant species. Such large
        areas are conducive to restoration and enhancement. In some areas, there are mitigation banks where property
        is in high demand, resulting in high land acquisition costs.  In other circumstances, finding an appropriate site
        may be difficult due to the lack of large wetland expanses. Land that is in public ownership can help lessen land
        acquisition cost, but may not have favorable physical attributes that would  allow for a mitigation bank to be self-
        sustaining over time.

    • Finally, regulatory coordination is important. The local agency  needs to  provide a prospectus to the Mitigation
      Banking Review Team (MBRT) established by the local Corps district.4 The prospectus will serve as the mitigation
      banking  instrument that identifies the objectives and administration of the proposed  bank. As pat-t of this
      coordination process, the local  agency should identify the proposed bank site, the geographic service area,
      wetland types  suitable for compensation  at the site, the debiting  and  crediting system, performance standards,
      monitoring  plan, contingency and remedial actions, and provisions for long-term management and maintenance.


        4 Proposing  a mitigation bank in the context of a regional plan  that integrates the bank into  a  comprehensive
          wetland or watershed management strategy may improve the likelihood of acceptance of the prospectus by
          the MBRT. A watershed management plan can provide greater certainty about the nature  and  extent of future
          wetland impacts and identify  the  most appropriate,  environmentally  beneficial  options for offsetting the
          anticipated impacts. In this way, the MBRT has a level of assurance that the bank is  a part of a broad goal
          to maintain or gain wetland functions in a given area or watershed.

Site selection criteria

    Local agencies need to carefully consider the sites that are  identified  for potential mitigation banks. Site selection
is one of the most important criteria affecting the successful establishment of a mitigation bank. The most significant site
selection criteria  is the potential that the site can be successfully restored or enhanced in a manner that will provide
appropriate  compensation  for anticipated  unavoidable  wetlands losses.  Selection  of such ecologically important or
desirable sites can further  regulatory acceptance.

    As identified in the federal  guidance on mitigation banking, mitigation sites should be self-sustaining overtime. Sites
with naturally occurring hydrology are preferable to sites that require  complex  hydraulic engineering  features that are
costly to develop, operate,  and maintain. Therefore,  sites that can be restored  without complex improvements should
be the first option when establishing a mitigation bank.

    Ideally, site selection would be undertaken in accordance with a watershed management plan under which existing
wetlands have been surveyed to determine which  sites are  the highest priority for protection,  the most suited for
restoration or enhancement, and the most likely to be impacted by development. Through watershed planning, wetland
functions that are lacking in a region can also be used to guide site selection. Additionally, planning at this level can help
determine compensation requirements, because anticipated wetland impacts can also be identified.  Two federal
programs, the Special Area Management  Plan (SAMP) and the  Advance Identification (ADID) Program, can provide
guidance and  technical assistance to  local sponsoring  agencies that meet  certain criteria and are interested in
establishing  a watershed scale planning  effort.
 The Federal Mitigation Banking Guidance states that the use of preservation in a mitigation bank may be authorized by the federal agencies when
it is demonstrated that the preserved areas contribute to the functions of the restored, created or enhanced aquatic resource.

* The Mitigation Banking Review Team is established as consistent with the Federal Mitigation Banking Guidance.

                                                      324

-------
    Mitigation bank site selection  can also be targeted  to address specific  environmental  objectives. For example,
wetlands in several counties in Florida were overrun by exotic vegetation. Mitigation banking provided the  funds to
improve the wetland habitat through exotic species eradication and revegetation with native  plants.

    Site selection may also be influenced by a regional demand for mitigation opportunities and the availability of suitable
restoration sites. Land acquisition is often cost-prohibitive for local agencies: land ownership, zoning restrictions,  and
allowable uses also are important factors in determining  a mitigation bank site. Important functional characteristics of
a potential restoration site include the presence of hydrology, hydric vegetation,  and/or hydric soils; size of the site;
historic conditions; and the degraded  condition of existing wetland (e.g.,  exotics, fill, compaction, trash). Experts in
restoration should be consulted regarding  specific wetland requirements and regional wetland attributes prior to  site
selection.

Funding

    The most common funding option used by  local governments  is the advance-credit sale option addressed in the
Federal Mitigation Banking Guidance.With this funding mechanism, advance sale of a percentage of the total credits is
allowed after certain criteria have been met by the bank sponsor, such as conservation easement, land acquisition,
and/or design plan approval. By selling a limited number  of credits in advance, the banker can collect sufficient funds
to begin conducting wetland improvements, then sell more credits as they are certified  by the regulatory agency. The
advance sale of credits, however, involves a degree of risk; for example,  problems can arise prior to  mitigation bank
implementation or completion, leaving the bank sponsor liable for any credits sold.

    Local governments have also  identified various other strategies for funding the establishment of their mitigation
banks. Some of the  options include:

    . Completing the project in phases so that initial credit sales will provide funding for the  remainder  of the project

    . "Borrowing" money from other local agency funds, then paying it back using money from credit sales

    . Using available federal or state grant money for wetland improvement as seed money to establish the banks;5

    . Partnering with a private company to share  the costs

    • Using combinations of the above mechanisms

Bank Administration and Operation

    Several issues related to bank administration and operation pose a challenge to local governments. The geographic
service area, credit ratios, credit valuation,  monitoring plans, and long-term maintenance provisions are all issues  that
must be addressed by the local sponsor and approved by the signatory agencies. Information on these  issues can be
found in  the Federal Mitigation Banking Guidance mentioned earlier,  as well  as in any  existing state  guidance. A
partnership  with  a private entity that  has experience in mitigation  banking  is one alternative  for  addressing bank
administration and operation challenges.

    The geographic service area for the mitigation bank should be established and presented in the mitigation banking
planning document. The service area is the area  in which credits from  a mitigation bank can be  used  to compensate
for unavoidable wetland losses. This area can be a watershed or a political boundary, such as a county or municipality.

    The number of credits produced by a mitigation  bank are generally determined by two factors, (1)  the number of
acres restored, created, enhanced, or, in exceptional circumstances, preserved, and (2) a quantified evaluation of the
5 EPA has also identified the State Revolving Fund (SRF) as another potential source for funding the establishment of mitigation banks, subject to
approval under each State's SRF regulations.

                                                     325

-------
functional value of the wetlands in the bank. As a general matter, the greater the wetland improvement, the more credits
generated.   For  example, more credits may  be  associated with  restoration  and fewer  credits  associated with
enhancement. The second factor involves a functional assessment of the improved or created wetland, and a method
for converting the functional units to credits.

    To determine the appropriate compensation requirements for wetland losses, a mitigation ratio must be developed.
This ratio can be based on lost wetland  function and area, the rarity of  the affected wetland, and the wetland  replacement
kind and function. The ratio will determine the number of credits required from the bank to replace the impacted aquatic
resource. Generally, the  mitigation ratio is greater than or equal  to one; i.e., at least one credit (acre of restored,
enhanced, or created wetlands) required to replace one acre of lost wetland resource.

    As a general matter, the bank sponsor monitors the bank for a period of time (e.g., five years) to determine whether
the bank is functioning at the level required by a previously determined set of performance standards established in the
banking instrument. After the performance standards are met, the  local agency may choose to transfer the mitigation
bank to another entity (e.g., land trust) for the long-term maintenance and monitoring. An agency can also use volunteer
labor to offset costs of monitoring and  maintenance and to improve public awareness and citizen stewardship for the
project. Among the local bank projects reviewed by EPA in the survey, local sponsors were  usually responsible/liable
for the performance of the mitigation site from the beginning, or after an established period of time if the project was a
joint-venture.

    Financial assurances for long-term maintenance and contingency plans for the bank most often  take the form of
additional fees added  onto the cost of a credit, with  a fixed amount from the sale of each  credit put into an escrow
account. The local agency then draws from  the account for  necessary  maintenance expenditures.  In some cases,
completed banks become part of a park system or are turned over to a public agency for long-term management.

Mitigation Bank Approaches with Case-Study  Examples

    The following case-studies are examples of the different types of mitigation banks that can be established by a local
agency, and highlight specific issues that are of concern to local governments. The banks discussed in the case-studies
are not mutually exclusive, in that a given bank may fit into more than one "bank type" category.

Mitigation Banking in the Context of a Watershed or Wet/and Management Plan

    Local governments can use a watershed or wetland management plan  as a means  of addressing economic
development and environmental concerns.  An important aspect  of such a plan involves compiling an inventory of
existing wetlands in order to determine both coverage area and functions  provided. This information can then be used
to protect important resources,  establish areas suitable for development, and determine the  best sites for restoration.
Mitigation banking is, therefore, a tool that can be integrated into a watershed plan to help meet resource management
objectives.

Advantages  of mitigation banking in the context of a watershed management plan are as follows:

    • Sites to be restored and critical sites to  be preserved are already identified in the plan

    . Market demand  for compensatory mitigation can be identified

    • The extent and nature of potential impacts have been determined

    . Likelihood of restoration  project success is improved

    • Uncertainty for regulatory  agencies is lowered

    . Multiple objectives can be attained
                                                    326

-------
    • Environmental benefits are  maximize

 West Eugene Mitigation Banking Program

    Overview: The Eugene Wetland Mitigation Bank Program (Bank) was established in 1995. The program is operated
by the City of Eugene, Oregon, under a separate fund within the City's financial structure. The goal of the Bank is to
provide a mechanism to fund wetland mitigation projects, carry out the West Eugene Wetlands Plan (Plan) adopted in
1992, and meet other community needs. This  program  is being conducted in cooperation with the City's wetland  partners
(the U.S. Bureau of Land Management [BLM] and the  Nature Conservancy [TNC]).6 The objectives that the  City seeks
to accomplish with the mitigation banking program include:

    • Creating credits in advance  of wetland losses

    • Meeting the mitigation needs of the community  of Eugene

    • Achieving  community  objectives, such as increased flood storage capacity, enhanced water quality, improved
      wildlife habitat, and  establishment of education  and recreation  opportunities

    . Targeting  areas with  high prospects for restoration success such as historic  wetlands,  disturbed agricultural
      wetlands, and  areas adjacent to waterways

    . Communicating the  banking program's value to the community

    • Operating the bank as part of a national model  wetland program in cooperation with the wetland partners

    .  Establishing a permitting process familiar to businesses, environmental interests, and regulatory agencies

    Funding for  identifying wetland areas for the  bank was provided through  EPA's Advanced Identification (ADID)
program. The West Eugene Wetlands Plan integrates wetlands protection and  community development  needs by
identifying areas best suited  for wetland preservation and areas with development  potential. Thirteen hundred  acres of
wetlands were identified within the plan area, with more than 1,000 acres designated  for protection  or restoration and
approximately 300 acres of lower value wetland designated as suitable for future fill and development. The  Plan calls
for the establishment of a mitigation banking program to compensate for unavoidable losses  of  wetlands  through
restoration  and enhancement in conjunction with  protection of important wetland resources.

    All mitigation bank projects will  be located within the Long Tom River watershed. Seven projects were initially
constructed for a total of 56 acres. Three additional projects totaling 60 acres were in  planning at the time of the EPA
survey.

    In this  program, mitigation bank credits can be established in three ways

    1.  The City or its partners may undertake wetland mitigation work,  then seek certification of credits by the  Corps and
       Oregon Division of State Lands (DSL). The City provides documentation on a site prior to the mitigation work and
       after completing improvements to demonstrate an increase in  wetland values.

    2. The  City  may create more credits than are needed to  compensate for permitted wetland  losses as part of
       concurrent or advance mitigation work. If the Corps and DSL certify the excess  credits, they will  be added to  the
       Mitigation Banking  Program's ledger for sale.

    3. Uncertified credits may be sold in order to fund initial mitigation work, which are later certified by the Corps and
       DSL when the initial hydrological and vegetative work is completed.
6 City of Eugene, Public Works Department. 1997. West Eugene Wetland Mitigation Bank, Annual Report 1996.

                                                    327

-------
    The City Stormwater Fund provided money for acquisition of site properties. Additional capital for the projects came
from the sale  of uncertified  credits. No uncertified credit can be sold unless the Corps and DSL have approved a
mitigation plan.

    As the bank sponsor, the City is responsible for monitoring  and maintaining the bank sites. Monitoring is required
for a minimum of five years and the sites will be inspected by the Corps and DSL on an annual basis. Monitoring goals
are determined on  a site-by-site basis. Maintenance will be conducted to ensure that the wetland functions and values
are fully  established and functioning. If the site  is not meeting  the  performance criteria, corrective measures will be
taken. Funding for  these tasks is included in the  credit price. At  least an additional 20% will be added to the estimated
credit cost and set aside to be used to  monitor, maintain and, where necessary, conduct remedial measures.

Single-user Mitigation Bank

    Local governments can establish a single-user bank. In this case, the local government initiates the bank, creates
the credits, and is the principal credit user. Local sponsors with  long-range project plans that involve potential impacts
to wetland areas may establish a mitigation  bank  or banking program in anticipation of the need for compensatory
mitigation. For example, a public works department with road expansion plans may know in advance that there will be
unavoidable impacts and provide the funds necessary to initiate a bank in anticipation of the project.

    An advantage  of a single-user bank is that long-term planning provides advance knowledge of potential wetland
losses associated with future projects. This reduces the uncertainty surrounding the demand for the bank and eventual
funding. Funding may be available to initiate the bank based on the predicted future use.

Snohomish County Airport Mitigation Banking Program

Overview. The Snohomish County Airport (SCA) needed to address the multiple objectives of their 20-year Master Plan,
the aviation needs of the airport, the economic growth of the region, and environmental  protection. The Master Plan
identified three development/construction projects that will occur  over five years, each with anticipated wetland impacts.
The objectives that the SCA seeks to accomplish through the banking program include:

    .  Creating a mitigation alternative for projected airport development

    . Replacing or augmenting wetland functions in the watershed

    •  Replacing habitats that will potentially be lost, including open-water habitat that cannot be created on-site

    •  Creating recreational and educational opportunities

    The SCA analyzed the three watersheds  containing the project areas to determine missing wetland functions and
values.  Potentially impacted wetlands were then characterized  according to acreage, wetland  category, and function.
Based on these analyses, two mitigation banks were designed to replace the wetland functions that would be lost by the
airport projects. The SCA banking program creates a "reserve"  of mitigation that accommodates Master Plan impacts
for approximately seven years.7

    Mitigation banking was an effective alternative for the SCA since the Federal Aviation Administration (FAA) prohibits
the creation of bird attracting  habitat within  10,000  feet of a  runway. In  this  circumstance,  on-site mitigation was
determined not to be a viable alternative as it would attract waterfowl and migratory birds. The SCA created or enhanced
a total of 32 acres  and preserved an additional 23 acres.

    Following  a format developed by the Washington State Department of Transportation, the SCA Memorandum of
Agreement (MOA) was established.  The MOA  describes the  procedures  for selecting,  managing, monitoring, and
  Snohomish County Airport. 1997. Snohomish County Airport MOA/IM Executive Summary.

                                                     328

-------
protecting mitigation banks. Additionally,  an Implementation Manual was created  by the  SCA that presents agency
recommendations for approval of the two specific bank sites. A technical oversight committee reviewed and advised
changes to the MOA, and provides input for credit evaluation and ratio development.

    Construction is  nearly complete at  both sites. One site has a nonprofit organization associated with it called the
Friends ofNarbeck Wetland Sanctuary thai is working to promote citizen stewardship, protection of wildlife and wildlife
habitat, native plant propagation,  and environmental research.

    Primary funding for the mitigation banks came from airport revenues. In addition, some funds were provided from
the FAA, since the program will be used to mitigate for projects required by FAA safety regulations.

    The SCA will be responsible for one of the sites in perpetuity. The other site will be donated to the City of Everett,
Washington, Parks Department after all  of the credits have been used. Monitoring will be conducted for five years in the
emergent wetland and ten years in the forested wetland. Both sites have been placed in a conservation easement and
a contingency plan will be executed if performance criteria are not met.

Public/Private Joint Venture

    Public/private joint ventures can facilitate mitigation-bankestablishment by providing investment capital and technical
expertise. Mitigation  banking  is  a complex  process  involving aspects  of design, engineering,  regulation, project
management, biology, ecology, marketing,  and long-term management.   Because of this, a mitigation  banking joint
venture may  be established to share the responsibilities and costs. A  local agency role may be bank sponsor, bank
creator, land owner, long-term manager, or all four. The private entity may conduct only the technical work, but can also
assume any of the aforementioned roles. The EPA survey found that the local agencies partnered in joint ventures were
most  likely to provide the  land for the mitigation bank and the  private entity would provide funds and/or technical
expertise.

City of Pembroke Pines and Florida Wetlandsbank, Inc.

    Overview: In 1992, the City of Pembroke Pines in Broward County, Florida, entered into a partnership agreement
with Florida Wetlandsbank, Incorporated to restore a heavily degraded site that was already overrun by exotic species
and becoming further deteriorated by  all-terrain vehicle  use and illegal  trash disposal. The  City did  not have the
resources necessary to restore the site or to establish a  mitigation bank. Concurrently, development pressures in the
area were  creating  a demand  for an  effective mitigation alternative.'  Florida Wetlandsbank agreed to design and
construct a wetland  system by eradicating the exotics and replacing them with a mixture of 10 typical Everglade habitats
including cypress stands, emergent marshes,  tree islands, and sawgrass prairie. Florida Wetlandsbank also provided
the management for the bank, selling the first credits in 1994. In 1997, restoration was near completion, with 396 credits
sold. The objectives for the Florida Wetlandsbank (the name of the mitigation bank) are to:

    .  Provide a mitigation alternative for the rapidly urbanizing City of Pembroke Pines

    .  Provide mitigation at a large site  with  high  potential for ecological success

    .  Restore native wetland ecosystems in an area that was degraded and infested with exotic species

    The mitigation bank provided  a number of economic and environmental benefits, including the following.
8 Paul Wattles, Assistant City Manager, City of Pembroke Pines, Florida. Personal communication, March 23, 1998.

                                                     329

-------
    • The  project provided  revenue for the City through the collection of franchise  fees  per acre (paid by Florida
      Wetlandsbank, Incorporated  after credits were sold).

    • Developers used the credits  as quickly as they became available.

    . The City plans to open part of the area as a park and offer public access and passive recreation opportunities.

    . To date, the site is ecologically successful, with little reinfestation of exotic plants.

    . Florida Wetlandsbank, Incorporated profited from the sale of credits.

    The construction is completed and  all credits sold for the original 350-acre site. An additional 100 acres has been
added to the site with construction nearly completed and credit sales pending.  The service area for the mitigation bank
was kept at the county level rather than  the basin level, which incorporates three  counties.  The County determined that
mitigation for wetland losses should remain within the county.9 The number of credits available in the mitigation bank
were determined  using a modified  Wetland Rapid Assessment Procedure (WRAP).

    Florida Wetlandsbank, Incorporated provided the  capital for initiating the bank. Pre-project sale of credits allowed
restoration work to begin. A trust fund was established to provide for the long-term maintenance of the site, with $1,000
per credit contributed to the fund.

    Site monitoring will be conducted  by Florida Wetlandsbank,  Incorporated from  the time that construction is completed
for five years. Quarterly monitoring reports are required  by the SFWMD. These reports contain  information regarding
the banks performance criteria. A performance bond was placed with the SFWMD for  each stage of construction. The
bond will be used in the event that any of the performance criteria are  not met.  Once  the performance criteria for the
site are met, the money will be released to Florida Wetlandsbank.

Commercial Mitigation Banking

    A local government  can establish a commercial mitigation bank  that sells  credits to anyone who  meets the
requirements to use the bank for compensatory mitigation. The  bank discussed here was established on public land,
solely by public agencies, with credits for sale to the public. Optimally, a public commercial mitigation bank is  part of a
larger watershed  plan.

    A major advantage of public commercial banks is that they are in public ownership, which can better provide long-
term  management, and the funds  generated from the bank can be  used to further improve public resources. The
disadvantage is that the up-front capital investment can be difficult for public agencies with limited availability of funds

DuPage County, Illinois

    Overview DuPage County is a highly  urbanized area located west of Chicago. Land development has negatively
impacted the natural drainage system of the  area by eliminating naturally  occurring storage,  reducing stormwater
infiltration, and increasing the velocity and quantity of runoff. In response, the DuPage County Stormwater Management
Plan (Management Plan) was developed in 1989, to reduce the potential for recurrent and increasing flood damage and
to  reduce further  environmental degradation  associated  with development. The ecological assessment conducted during
development  of the Management Plan found that wetlands represented a significant portion of the natural watershed
9 Desmond Duke, Project Administrator, Florida Wetlandsbank. Personal communication, June 19, 1998.

                                                     330

-------
storage in the county, and are essential for adequate stormwater retention, conveyance, sediment control, and water
quality  enhancement."

    Individual watershed plans, within the larger DuPage County management plan, were developed to identify wetlands
for protection, enhancement, and restoration.  The County attempts to  establish mitigation banks  in the four main
watersheds within the County. The service area for each bank will be the watershed in which the bank is located. The
County also  established  an  ordinance for stormwater management that requires developers whose development
proposals will affect  the function  and values of wetlands to consider mitigation banking as an alternative to compensatory
on-site mitigation. The objectives of the DuPage County commercial mitigation banking program are  to:

    • Manage and  mitigate the effects of urbanization on stormwater drainage

    .  Enhance  the quality,  quantity,  and availability of surface  and groundwater resources and  prevent further
      degradation

    . Preserve and enhance existing  wetlands, aquatic, and riparian environments

    • Encourage restoration of degraded areas

    Any investment in a  bank must be at least equal to the cost of planning, acquiring lands, constructing, and operating
and maintaining  mitigated wetlands of equivalent or greater functional value than those lost  to development.

    The land  for the bank sites was in public ownership. Funding for establishing the bank and initial project work came
from an advanced credit sale. One-third of the credits were allowed for sale prior to work beginning at the site. One site,
Winfield Creek, was in the process of selling credits for project initiation when public opposition to the site halted the
project. The County is now in the process of finding another site but, in the interim, is liable for the credits sold.

    The County  has five banks that have been approved by the agencies and are in various stages of implementation.
The oldest project, Cricket Creek, is in year two  and, to date,  is exhibiting successful hydrologic and vegetation
conditions.

Conclusion

    The use of mitigation banking by local governments can be an  effective tool to restore and protect their community's
valuable wetlands resources. The case-studies presented provide local governments with different mitigation banking
strategies that could be used to address  the needs of their community. As demonstrated in the case-studies,  the
establishment of sound objectives and goals by the local agency will help determine the type of mitigation bank that will
best meet the local  needs.  In addition, there are several conditions that local governments must evaluate and consider
for successful bank implementation,  including (1)  a demand for compensatory mitigation,  (2) a sufficient supply of
suitable sites for the bank, and (3) opportunities for working in partnership with the federal and state regulatory and
resource agencies.

    Mitigation banks that are established by local governments can address more than just the need for compensatory
mitigation. Wetlands mitigation  banks can achieve additional community needs  by  increasing local  flood storage
capacity, improving  wildlife habitat, and providing educational and recreational benefits while restoring and enhancing
important wetland resources.
10 DuPage County Stormwater Management Committee. 1989. DuPage  County Stormwater Management Plan.  DuPage County, Illinois.

                                                     331

-------
            Massachusetts Stormwater Management  Policy/Regulations:
                      Development,  Implementation, and  Refinement
                                             Bethany Eisenberg
                                     Vanasse Hangen Brustlin, Incorporated
                                          Watertown,  Massachusetts
Abstract

    In  March  of 1996,  the Massachusetts Department  of Environmental Protection, in conjunction with the  Massachusetts
Office of Coastal Zone  Management, released the Draft Version of the State's Stormwater Management Policy. The Policy
includes nine specific Stormwater Performance Standards, which are to be met to achieve compliance. The Policy presented
in two volumes: Volume One, the Stormwater Policy Handbook, which contains a description of the policy, its implementation,
and descriptions of the nine individual Stormwater management standards, and Volume Two, the Stormwater Technical
Handbook, which  contains more detailed information on Best Management Practices (BMPs), for Stormwater management
(i.e.,  detention basins, swales, etc.).

   The policy is the result of three years of work by a Stormwater Advisory Committee that included representatives from
regulatory offices (EPA, Department of Fisheries and  Wildlife, Natural Resources Conservation  Service, etc.),  engineers and
developers from the private sector, the highway department, and representatives of local conservation commissions. The Policy
is to  be implemented as an amendment to the existing Massachusetts Wetlands Protection Act, which is administered at the
local  level  by  local Conservation Commissions.

   The nine performance standards are the key components of the policy. General descriptions of the standards include the
following:

    1. No new Stormwater conveyances  may discharge untreated Stormwater directly to, or cause erosion in wetlands or
       waters of the Commonwealth.

   2.  Post-development peak discharge  rates may  not exceed pre-development rates.

   3.  Maximize recharge to groundwater: post-development must be similar to pre-development conditions.

   4.  Remove 80% of average annual load-post development of Total Suspended Solids (TSS).

   5. Use specific BMPs for discharges  from areas  with higher potential pollutant loads; untreated infiltration prohibited.

   6. Use specific BMPs for discharges to critical areas.

   7.  Redevelopment projects should not  meet performance  standards to  the maximum  extent practicable and  at a minimum,
       improve existing conditions.

   8.  Erosion and sedimentation controls are required  during  construction.

   9. Operation and Maintenance Plan for  Stormwater Management required.

   The Policy was introduced in March of 1996 for testing its effectiveness. Two Phases of training were provided over a two-
year  period across the state. The first phase focused on introducing the Policy and the second phase focused on  detailed
Engineering Companies, and local  DPWs and planning  departments.


                                                    332

-------
Introduction

    In March of 1997, the Massachusetts Department of Environmental  Protection (MADEP), in conjunction with the
Massachusetts Office of Coastal Zone Management (CZM), released a draft  version of the state's Stormwater Management
Policy (herein referred to  as the "Policy").  Prior to the  development of this policy, the control of peak stormwater discharges from
development sites to prevent flooding and erosion problems was a fairly standard requirement across the country; and was
well-implemented in Massachusetts. However, there were no state-level requirements for stormwater quality treatment,
maintaining groundwater recharge processes,  or maintaining stormwater treatment systems.  The Policy was developed to
provide standard minimum requirements  for stormwater management that could be consistently implemented on development
projects.

    The Stormwater Management Policy is currently an amendment to the Wetlands Protection  Act which is only applicable
when a project proposes work within the boundary, or buffer zone, of a Wetland Resource Area. Hence, the Policy is not
applicable to all developmental projects.

    While the regulatory implementation  of the  Policy is through local Conservation Commissions, under the State's Wetlands
Protection Act, the performance standards and design guidelines that define the Policy were developed  for use by  a larger
audience.  Development teams (typically the proponent and their engineers, architects, and planners) and the various reviewing
agencies (Conservation Commissions, Planning Boards, DPWs, etc.) were expected to be  users of the Policy. The goal was
to provide guidance to ensure  that  negative impacts from stormwater runoff generated as a  result of urban and suburban
development would be  minimized  without placing  an unjustifiable economic burden on developers for new projects, or
preventing redevelopment of existing sites.

    The Policy includes nine specific Stormwater Performance Standards for which compliance  must be achieved on
development projects.  Included  in these standards  is a requirement to provide  groundwater recharge, requirements for ensuring
proper stormwater treatment prior to  discharge to waters of the state, and provisions for waiving certain requirements if deemed
infeasible for redevelopment projects. The  Policy has been distributed  as an Interim Draft to allow for  refinements prior to its
final promulgation as state regulations. In orderto fully gage its effectiveness, however, the Interim Policy has been distributed
and implemented as  if the regulations were final.

    The Policy is  presented in two  handbooks: Volumes  One and Two. Volume One, the Stormwater  Policy Handbook,
contains a description of the policy,  its implementation,  and detailed definitions of the nine individual stormwater management
standards. Volume Two, the Stormwater Technical Handbook, contains detailed information on Best  Management Practices
(BMPs), with guidelines for the design of standard  stormwater management structures (such as detention basins and water
quality swales).

Development

    The Policy is a result of three years of work  by a State Stormwater Advisory Committee. In addition to leaders from MADEP
and CZM, this Committee included representatives from such regulatory offices as USEPA, Department of Fisheries and
Wildlife, and the Natural Resources Conservation Service. Also on the committee were engineers and developers from the
private sector (including the author) and  representatives from the  Massachusetts Highway Department and local Conservation
Commissions.

    The goal of the Committee was  to provide a cohesive set of performance standards that addressed key issues associated
with stormwater runoff control. The Policy was developed in such a manner that it provides adequate accompanying guidelines
and recommendations, to allow for consistent implementation, while still allowing for flexibility  in site-specific designs.  Given
the widely varying goals and perspectives of Committee members,  the Policy published in 1997 was a result of group
consensus  and  compromise.

    The Committee was divided into  two  sub-committees, the Policy  group,  and the Technical sub-committee. The  Policy group
was tasked with developing the process for legal implementation and the Technical Committee was responsible for developing
the specific technical requirements  of the  Stormwater  Management Policy.

                                                      333

-------
    Some of the performance standards introduced completely new  requirements for  development projects and  required
detailed evaluation and discussion during the development stages. Some of the key issues that generated substantial debate
include the following:

    • Recharge:~t\\Q loss of recharge to groundwater systems, which provide drinking water supplies and generate baseflow
      to streams and rivers, was a state-wide problem. A mechanism for requiring the maintenance of recharge after
      development was considered a high priority. The biggest issue relative to this requirement was: How much annual
      recharge should be required?

    . Water Quality Treatment Volume: The quality and quantity of  stormwater runoff from paved  and unpaved areas can
      vary greatly. It was determined by the Committee that runoff from impervious areas was the highest concern and should
      require treatment. The largest decision relative to this concern was: What volume of stormwater runoff should be treated
      for water quality? Should  it include runoff from both pervious and impervious areas?

    . Critical iinniffi.B  The Committee felt that certain sensitive  environmental areas should  have the maximum practicable
      protection. Under the Massachusetts Stormwater Management Policy, "Critical areas" are defined  as;  shellfish  growing
      areas, public swimming beaches, cold water fisheries,  recharge areas for public water supplies and designated
      Outstanding Resource Waters (ORWs). ORWs are further defined to include surface drinking water supplies,  certified
      vernal pools, and state designated Areas of Critical Environmental Concern (ACECs). The issue here was how to ensure
      that these areas are  provided adequate protection  and  how to define what adequate  protection is.

    • Exemptions: Some Committee members expressed concern that stormwater management requirements may be too
      costly for small  residential projects, or may be a deterrent for initiating redevelopment projects. The issue was what, if
      any, projects should be exempt from any or  all of the standards?

    • Operation/Maintenance: Maintenance of stormwater management practices is critical for their effectiveness. It is often
      difficult, if not impossible, to ensure that the operation and maintenance of BMPs will occur as necessary. The issue was:
      How can the necessary maintenance of BMPs  implemented on the project be ensured?

    A brief summary of the decisions on these  key issues is as follows:

    . Recharge: Annual recharge processes are  permanently changed by the introduction of impervious areas to a site.  In
      order  to minimize this impact, it was agreed that the existing  annual recharge should be determined and post-
      development annual recharge  should maintain this  to the  maximum extent practicable. A preliminary methodology  for
      determining existing  recharge on the site was provided  in the Draft Policy.

    • Water Quality Treatment  Volume:

      • For discharges to critical  environmental areas (defined in the Policy), the volume of stormwater runoff to be treated
       for water  quality control is  defined as 1  .0 inch of  runoff times the total  impervious area of the post-development project
        site.

      • For all other discharges, the volume to be treated is calculated as 0.5 inches of runoff times the total impervious area
        of the post-development project  site.

         These volumes represent  total runoff from the  smaller, more frequent  storms that occur annually, and the initial
      volumes of runoff from larger more infrequent storm events. The goal behind this decision was to fully treat the runoff
      from the majority of the storm events occurring annually,  without approaching  values where treatment system sizes
      would  result in increasing costs with decreasing additional benefit.

      • Critical Areas: It was decided, as stated above,  that  1  .0 inch  of runoff (as opposed to 0.5 inch) must be treated for
       water quality if the discharge is to a Critical Resource  Area. Also, specific approved BMPs are recommended for use
        in particular critical  areas. In addition, it was decided  that spill prevention/containment methods must be included  in
       the  Stormwater Management  System design.

                                                      334

-------
      .  ^mon,«n..H  Specific cases were provided  exemption from  the  Policy, including: single family house  projects,
        residential subdivisions with four or fewer lots that do  not discharge to critical areas, and emergency repairs to
        highway/roadways or their drainage systems. However, none of these projects are exempt from the standard requiring
        sedimentation and erosion control  requirements during  construction activities.

        While redevelopment projects are not exempt,  a redevelopment project may comply to the maximum extent possible
        if it can be proven that it is not practicable for the project to  achieve full compliance.

      .  Operation/Maintenance: All non-exempt projects require the development of a Stormwater Management System
        Operation and Maintenance Plan (O&M Plan). As defined in the Policy, the O&M Plan must contain the: names of the
        Stormwater Management System(s) owner and the person(s)  responsible for implementing the O&M Plan, a schedule
        for inspection and maintenance, and a description of maintenance activities to be performed. Recommendations for
        specific maintenance practices and schedules are  included in the Policy.

Stormwater  Performance  Standards

    The nine Stormwater Performance Standards most accurately describe the key components of the policy that came out
of the committee deliberations following resolution of the issues described above. The goal of the standards is to protect
groundwater, surface water, and  wetland resources from  the impacts  of Stormwater runoff generated as a result of development
and redevelopment projects. General  descriptions  of the standards are provided as  follows:

    1. No new Stormwater conveyances may discharge untreated Stormwater directly to, or cause erosion in wetlands or
      waters of the  Commonwealth.

    2.  Post-development peak discharge rates must not exceed predevelopment  conditions for the 2-year and 1 0-year storm
       events under  all conditions. The 100-year event must be analyzed to determine impacts and must be controlled if
       necessary.

    3.  Loss of annual recharge to groundwater should be minimized through the  use of  infiltration measures, to the maximum
       extent practicable. The recharge "requirement"  which is to mimic existing annual recharge on sites to the maximum
       extent practical, has not been  changed. However, a design  methodology for estimating existing annual recharge  at a
      site, and for designing recharge systems has been developed. The  methodology  uses soil classifications, soil gradation
      analyses and  specific Massachusetts regional rainfall data as data inputs.

    4. Stormwater management systems must be designed  to remove 80% of the average annual (post development) load
      of total suspended solids (TSS). It is presumed  that this standard is met when; (a) suitable nonstructural practices for
      source control and pollution prevention are implemented;  (b) Stormwater management BMPs are sized to capture the
       prescribed runoff volume; and (c) Stormwater management BMPs are maintained and designed as specified in Volume
      Two. The Policy provides estimates of the percent TSS removal for  individual BMPs when designed in accordance with
      the specified guidelines. Water quality treatment volume  is  0.5 inches of runoff from impervious areas (1  .0 inch if
      discharge is to critical environmental area).

    5. Stormwater discharges from areas that are defined as having "higher potential pollutant loads" (as defined in the Policy)
       require specific Stormwater BMPs. Infiltration of Stormwater from  these areas without pretreatment is prohibited.

    6. Specific BMPs must be used for discharges to critical areas and the water quality treatment volume is 1 .0 inch of runoff.

    7.  Redevelopment projects must  meet the performance standards to  the maximum extent practicable. It must be clearly
      stated why full compliance cannot be achieved and such projects must,  at a minimum,  improve existing conditions.

    8. Erosion and sedimentation controls  are required during construction.
                                                      335

-------
    9.  An Operation and Maintenance  Plan (O&M  Plan) for Stormwater  Management is  required.

       A detailed explanation of each of these Standards is available in  Volume One, the Stormwater Policy Handbook.

Implementation

    In  order to test  its effectiveness and identify potential problems, the Policy was introduced in  March of 1997, prior to the
promulgation of regulations.  Copies of Volumes One and Two were distributed to each Conservation Commission  office in the
state, and to other relevant regulatory agencies. Two phases of training were provided  across the state over a two-year period:
first to introduce the  Policy, and then to focus on detailed case studies and implementation issues. Training was made available
to regulatory agencies,  Conservation Commissions,  local  DPWs, planning departments, and engineering companies.  During
the training  sessions, the largest turnout was from Conservation Commission representatives and engineering/consulting firms,
with  minimal attendance by the other groups invited.

    At each of the training sessions, questions  from the audience on the  Policy were solicited and a list of most Frequently
Asked  Questions (FAQs) was  developed.  The FAQs  provided an excellent basis for outlining where  additional information
and/or  clarification was  needed. Based on  the number of  recurring questions,  the Committee decided to prepare  a survey to
solicit feedback from potential  policy users. The survey was comprised of 23  questions for characterizing the  respondent,
determining the usefulness and ease of implementation of the Policy, identifying particular problems in understanding or
implementing performance standards, and  determining what type of BMPs were currently being  implemented.

    The survey was not designed to be a statistically valid data set, but rather to gain a practical working knowledge of what
aspects of the Policy and/or its method of implementation needed to be refined. While this was  generally evident from the
FAQs,  the survey further substantiated the  specific areas  where additional work was  necessary. Some key findings from the
118  respondents to  the survey were as follows:

    .  The overall sense of the Stormwater Performance Standards was that the Stormwater Policy implementation was good
       (63%) and that they  consider the Stormwater Handbook to be a "useful resource for designing and reviewing systems
       (77%).

    .  In response to  the standards, respondents  were generally comfortable implementing peak discharge controls  and
       sedimentation and  erosion controls. This  was  not surprising since  these were existing requirements in most
       municipalities that needed to be achieved for most development projects.

    . Not surprisingly, new requirements for  groundwater discharge,  treatment of Stormwater runoff water, and  the
       preparation of Stormwater Management O&M Plans were the standards for which additional clarification and technical
       support were most requested.

Feedback

    The Draft Policy was issued prior to formal promulgation so that one to two years  of interim implementation could be used
for refinement. The FAQs, user survey, and ongoing feedback from the public  defined those areas where  refinements were
especially necessary, including the following:

    Standard No. 3  Recharge: The recharge requirement clearly  created  the most confusion, and required additional
technical support. The original brief description included in Volume One was  not sufficient for engineers or reviewers to prepare
a comprehensive program for achieving the annual recharge  requirement.

    As a result, the  technical sub-committee has focused  on providing a more detailed definition of the  recharge requirement
and appointed  a group to develop a design  methodology for achieving compliance with  this standard. The group has developed
a methodology using soil groups and soils analysis and  actual rainfall data to determine the existing  annual recharge on  project
sites. Methodology for designing a recharge system  to provide post-development recharge that best mimics pre-development
                                                       336

-------
conditions is being developed. This technical is currently in the final stages of development and is expected to be completed
and distributed  prior to December 1999.

    Standard  4 - 80% TSS Removal: The Total Suspended Solids (TSS) removal requirement was not developed with the
goal of removing only the TSS loads in stormwater runoff. Rather,  it was considered an indicator parameter, whereby if 80%
of the TSS is removed, a large portion of the additional pollutants carried in stormwater is also removed. This relationship does
not hold true, however, if the proponent chooses to use only mechanical methods where settling of fines, assimilation of
nutrients, or other biological processes that increase pollutant removal do not occur. For instance, a vegetated infiltrating swale,
or wet pond, that is designed to provide substantial stormwatervelocity reductions  may greatly increase fine sediment removal
and may also provide nutrient uptake  in addition to gross TSS removal. Astructure that removes solids only, and does not allow
for detention or contact with plants or  potential filtering areas does not comply with the goal/intent of the TSS removal standard.

    In addition,  users were having difficulty calculating TSS removals when  numerous BMPs were to be used in a series. The
specific percent removal rates  provided in Volume One of the Policy  are not additive  and, as such, must be calculated  as
percents of the  pollutant load that they receive.

    In response to these issues, further definition of the goal of the  TSS removal requirement and recommended practices  for
achieving the goal are being developed. A spreadsheet has been developed, which can be  easily filled out by hand, to assist
in calculating the percent  TSS removal for a project based on the BMPs implemented.

    Standard 9 - Operation  and Maintenance: This standard has consistently  raised the most concern relative to cost and
implementation.  Common questions  were as follows:

    .  Who  will pay for ongoing operation and  maintenance of stormwater BMPs if it  is not the town?

    .   How will the town fund the O&M  requirements if they assume responsibility?

    .  What long-term mechanism is there to ensure that maintenance will be completed?

    .  What is  the frequency of maintenance  required for specific BMPs?

    While  these questions are difficult to answer, for a specific site or on a statewide basis, in terms of required maintenance,
the Committee has responded by preparing operation and  maintenance checklists, which may be used by the operator/owner.
These checklists may be  submitted to the local Conservation Commissions  on an annual  basis, if deemed necessary.

    Specific  maintenance  practices and suggested frequencies have also been prepared and are currently being updated,  as
new information becomes available. It has also been noted that "one size does not fit all" in terms of required  maintenance.
For instance, an infrequently used residential guest parking area clearly does  not warrant the same frequency of street
sweeping  as a  commercial mall parking lot.

    The question of financing stormwater management system operation and maintenance activities is also a site- and location-
specific issue. The development of stormwater utilities and  management districts is ever  increasing and may be the option that
some communities choose. The promulgation of the Policy as regulations will require the project proponent and/or towns and
municipalities to develop a plan for ensuring O&M implementation. There are a variety of ways this may  be  achieved.
Resolution may  be somewhat facilitated in larger municipalities and/or USEPA-designated urban areas that must comply with
the  upcoming final USEPA NPDES  Phase II Stormwater Regulations.

Summary

    The development of the Massachusetts Stormwater Policy, which includes nine performance standards, was an integrated
effort between state and local regulators, policy makers, engineers, developers,  and the general public. The group effort and
the  implementation of the Policy as a Draft, to which refinements could and have been made, have contributed to the
usefulness of the standards. While the development and implementation of the standards are still in the early stages, the Policy
provides definitive goals  for achieving stormwater water quality and quantity control, and addresses annual recharge  on

                                                      337

-------
development sites. Until the development of these standards, there was no requirement to maximize groundwater recharge
on sites or to mandate the development of the stormwater O&M Plan (unless required under other regulatory programs). These
advances will provide an  overall benefit  to the natural resources of the Commonwealth of Massachusetts. The Stormwater
Management Policy Volumes I and II may  be obtained off the Internet or by request from:

      Tom  Maguire
      Massachusetts Department of Environmental Protection
      1  Winter Street
      Boston, MA 02108
      Phone (617) 292-5602
      email:  Tom.maguire@state.ma.us

References

Massachusetts Coastal Zone Management, Massachusetts Department of Environmental Protection, Stormwater  Management
Volume One: Stormwater Policy Handbook, March  1997.

Massachusetts Coastal Zone Management, Massachusetts Department of Environmental Protection, Stormwater  Management
Volume Two: Stormwater Technical Handbook, March  1997.

Massachusetts Coastal Zone  Management,  Massachusetts Department of Environmental Protection, The Stormwater
Evaluation Survey: Perspectives on the  Implementation of the Interim Massachusetts Stormwater Policy, June 1988.
                                                    338

-------
                          Implementation  of Michigan's  Voluntary
                      Stormwater  Permit-a Community  Perspective
                                         Kelly A. Cave, P.E. Director
                                      Watershed Management Division
                                  Wayne County Department of Environment
                                                Detroit, Ml

                                              Dale S.  Bryson
                                          Camp Dresser & McKee
                                                Detroit, Ml

                                     Kelly C. Kelly,  P.E., Project Engineer
                                           Canton Township, Ml

                                        Jack D. Bails, Vice-President
                                         Public  Sector Consultants
                                                Lansing, Ml
Introduction
    The Rouge River National Wet Weather Demonstration Project (Rouge Project) has made significant progress in
restoring beneficial uses to a large, urban waterway using a holistic, "bottom up" watershed approach. This project was
initiated in 1992, by the Wayne County (Michigan) Department of the Environment. The purpose of this document is to
present a summary of the activities and progress of the Rouge Project; discuss the watershed approach being utilized in
the Rouge Project, including the use of a general storm water permit; and summarize a community perspective on this
entire effort.

Rouge Project Background and Summary of Progress  to Date

    The Rouge River National Wet Weather Demonstration Project is a watershed-based effort, substantially sponsored
by the U.S. Environmental Protection Agency (USEPA), to manage wet weather pollution to the Rouge River, a tributary
to the Detroit River  in southeast Michigan (See Figure  1). The Rouge  River Watershed is  largely urbanized, spans
approximately 438 square  miles, and is home to over 1.5 million people in 48 communities in 3 counties. The Rouge River
has been designated  by the international Joint Commission as a significant source of pollution to the Great Lakes system.

    The early focus of the Rouge Project was the control of CSOs in the  older urban core portion of the downstream areas
of the watershed. As a finite number of point source CSO discharges could be identified, and responsibility for each
defined, the traditional regulatory approach of issuing National Pollutant Discharge Elimination System (NPDES) permits
mandating corrective  action worked relatively well. Additional  monitoring of the river showed,  however, that other sources
of pollution needed to be controlled before full restoration of the river would be achieved throughout the watershed. These
other sources of pollution  include storm water runoff, interflow from abandoned dumps, discharges from illicit connections,
discharges from failed on-site septic systems, and resuspension of contaminated sediment.

    The Rouge Project is  designed to  identify the most efficient and  cost-effective controls of wet weather pollution, while
assuring  maximum use of watershed resources. A great deal has been accomplished in these efforts. The following
summarizes some of those accomplishments, focusing on CSO controls first.  Approximately 50%  of the watershed is
served by separate sewer systems, with an additional 20%  served  by combined sewers (157 overflow points), and the
                                                   339

-------
Figure 1.  Rouge Subwatersheds & Location of Rouge Watershed in Michigan.

remaining area served by on-site sewage disposal systems. CSO controls are being implemented in phases. Under the
first phase, six communities have separated their sewers and eight communities have constructed or are constructing 10
retention treatment basins. Each of these  basins is sized  for different design storms and several employ innovative
technologies. A two-year evaluation study  of the CSO  control  basins began on June 1, 1997. The  results from the
evaluation study, coupled with efforts to control storm water and other pollution  sources in the watershed, will provide the
basis for the second phase CSO control program for the remaining CSO sources in the watershed. The information gained
from the evaluation of design storms and control technologies will  be useful to others nationwide to determine cost-effective
CSO controls to meet water quality standards.

    The Rouge Project is also evaluating innovative stormwatercontrol and watershed management  technologies. Twenty-
five different communities and agencies throughout the watershed are implementing over 100 pilot projects. Categories
of pilot  management  projects currently underway include  wetlands creation and restoration; structural storm water
practices, such  as grassed swales  and detention  ponds; erosion controls; stream bank stabilization; and habitat restoration,
to name a few.

    The Rouge Project also discovered that illicit connections and failing septic systems are major sources of pollution
problems in the Detroit urban area. Creative ways to remediate these sources of pollution  have been initiated.

    A suite of computer models has been developed by the  Rouge Project. These  models simulate the water quality and
quantity  response  of the Rouge River during  wet  weather events for existing and future conditions, and under various CSO
and stormwater runoff management alternatives. This effort has resulted in  a  very useful public communication tool on
water quality indices tied to actions needed to restore the Rouge  River. A comprehensive geographic information system
(GIS) and relational databases were designed and implemented to manage the wealth of data collected under the Project.
In addition, a special data exploration tool, DataView, was developed to support routine analyses of large time-series data
sets. DataView is  user-friendly and readily transferable to other locations. Related to DataView is the Rouge Information
Manager, also a user-friendly,  readily transferable tool  (an "electronic file  cabinet") for accessing multi-media information
about the Rouge  River restoration  effort.

    Effective, readily transferable tools have been developed, employed by the Project, and are being shared with other
cities and  state  agencies. Additional information  on the  Rouge  Project  can  be  obtained  from the  web  site at
h ttp:\www. rougeriver, com,

Evolution of the  Watershed Approach

    The Rouge River watershed has seven subwatersheds that range in size between 19 and 89 square  miles (See Figure
1). Older communities served by combined sewers dominate downstream portions of the Rouge River Watershed, while
                                                      340

-------
the headwater areas  are typically  open space, agricultural land, or low density  residential  developments that  are
undergoing rapid change due to growth pressures. Fully developed areas, typical of the middle portions of the Rouge
Watershed, have  separated  storm  sewers and limited opportunities to address stormwater problems with structural
solutions.

    Data gathered by the Rouge Project have shown numerous water quality and designated use problems, including high
bacteria levels and low dissolved oxygen levels during wet weather events in all areas of the watershed. Fish consumption
is prohibited in much of the watershed due to the threat to human health. Six of the seven subwatersheds have moderate
to severe degradation of wildlife habitats, with fish populations suffering severe impairment in three of the subwatersheds.
Aesthetic enjoyment is moderately to severely impaired throughout the watershed. Restrictions to small boat navigation
resulting from logjams,  garbage and  sedimentation  are  a moderate to severe  impairment in  virtually  all seven
subwatersheds.

    Based upon what was learned, the focus of the Rouge Project became more holistic to consider impacts  from all
sources of pollution and use impairments in receiving waters. The historic implementation of water quality management
programs in the United States at the federal and state levels has been to focus on  point sources,  which are the most
obvious sources of pollution to water bodies. This approach has worked well to control pollution from most point sources
but has also left a patchwork of regulated  and unregulated discharges of stormwater and nonpoint source pollution to
surface waters. This  patchwork is especially  evident in most urbanized areas where multiple local jurisdictions are located
in the same watershed.  More subtle sources of pollution, such as stormwater, have emerged as the next priority for
attention. The challenge for the Rouge Project became to develop innovative,  coordinated solutions to achieve water
quality objectives that may be: (1) be more cost-effective, (2) implemented in a more timely fashion, and (3) better able
meet local needs.

    It has become clear that water  resources management must have the support of the general public in order to be
effective and to become self-sustaining. A locally driven watershed approach to pollution management as a means to
achieve management goals is an exciting concept discussed by many, but for which there is limited practical experience.
This is particularly true in urban situations where there are multiple sources of impairment to a water body and  stiff
competition for limited local  resources to  address the pollution sources. The  Rouge Project  has  provided a unique
opportunity for a watershed-wide approach to restoring and protecting an  urban river system by using a cooperative, locally
based approach to pollution control.

The Michigan  NPDES General  Permit for Municipal Stormwater Discharges

    As concern expanded to sources of pollution in the upper portion of the watershed above the CSO discharges, and
water quality improvement focused more on watershed-wide approaches, the lack of a defined regulatory framework to
address stormwater pollution and diffuse nonpoint source pollution became a  major obstacle to further progress in
improving water quality and restoring beneficial uses to  the Rouge River. Beginning in 1995, the Michigan Department of
Environmental Quality (MDEQ), the Rouge Project, and the communities in the  Rouge Watershed jointly developed an
innovative, watershed-based  NPDES general permit ("General Permit") for municipal stormwater discharges. The permit
was issued on July  31,  1997 (MDEQ,  1997). This collaborative process was  outlined in a report entitiled "Adapting
Regulatory Frameworks  to Accommodating Watershed Approaches to Storm Water Management" (Fredericks, et al.,
1997).

    The MDEQ General Permit and USEPA's draft Phase  II stormwater regulations (U.S.  Environmental Protection
Agency, 1998) due to be promulgated in September 1999, were developed during the same time frame. Wayne County,
on behalf of the Rouge Project communities, was selected to serve on USEPA's Urban Wet Weather Flows  Federal
Advisory Committee, which  (among other activities) assisted USEPA with the development of the Draft Phase II
Stormwater Regulations. Participation on the federal advisory group on watershed approaches to  stormwater management
was invaluable to the development of the Michigan General Permit, and provides a high likelihood that the General Permit
will be acceptable for implementing the forthcoming federal Phase II Stormwater Regulations in Michigan.

    The General Permit incorporates the following elements:


                                                    341

-------
    General:

    - Permit coverage is voluntary until the final EPA Phase II Stormwater Regulations are promulgated

    - Public agencies who own, operate, or control stormwater systems are provided the opportunity for coverage

    - Watershed size is established by the potential permitees during the  application process

    - Application and permit process have limited required actions; the focus is to establish desired outcomes.

    Requires permittee to develop the following:

    - Illicit Discharge Elimination Plan (IDEP) that has the  goal of eliminating raw sewage discharges and includes
      addressing failing septic systems and improper connections of sanitary sewers to storm drains and open waterways.

    -  Public  Education P/an (PEP) designed to inform residents and businesses about what actions they should take to
      protect the river.

    -  In cooperation with others, a  Watershed Management P/an to resolve water quality concerns which includes: short
      and long-term goals for the watershed, delineation of actions needed to achieve the goals, estimated benefits and
      costs of management options, an opportunity for all stakeholders to participate in the process.

    - Storm Water Pollution Prevention Initiative, which includes evaluation and implementation  of pollution prevention
      and good housekeeping practices and the evaluation and implementation of BMPs to minimize impacts of new
      development and redevelopment.

    -  Monitoring and Reporting P/an, which includes a  schedule for revisions to the Watershed  Management Plan.

    The IDEP and PEP are submitted to MDEQ at the time of application and the implementation of these plans begins
when  the  MDEQ issues a certificate of coverage to a community/agency. Within six months after the issuance of a
certificate  of coverage to a community/agency, the General  Permit requires the submission of a public involvement plan
for approval by the state. This plan identifies  the approaches that will be used within the hydrologic area to involve
stakeholders  in  the development of a watershed plan that must be completed within 18 months after the certificate of
coverage is issued. Once a consensus, long-term management plan has been completed, each agency and community
within the watershed must prepare  and submit for state approval its own pollution prevention initiative that identifies actions
and schedules to address the pollution concerns identified in the consensus watershed plan. The watershed stormwater
management plans  developed by the communities and other public agencies do not require state approval; however,  the
individual pollution prevention initiatives emanating from the watershed planning process do require  state approval, as the
activities specified in the initiatives become permit requirements upon approval.

Rouge Community/Agencies Approach to Application and Permit Requirements

    A total of 43 communities  and agencies who own,  operate, or control stormwater systems in  the Rouge River
Watershed have obtained coverage under Michigan's new  watershed-based General Permit for  municipal  stormwater
discharges. The result is that over 95% of the watershed is  covered under this new program. The communities and
agencies requested that, for purposes of the General Permit,  the large Rouge watershed be  subdivided into the seven
subwatersheds  shown  in Figure  1.  Long-term management plans will,  therefore,  be  developed for each of these
subwatersheds, with  coordination  of the plans provided by the MDEQ  and the Rouge Project staff. The document
"Implementing a Model Watershed Approach  Through a State  General Storm Water NPDES Permit" (Cave, et al., 1998)
outlines key issues discussed and decisions  reached by  the communities  as they developed  their General Permit
applications during 1998. The following section presents additional information regarding the application  and permit
requirements recently approved for the communities and agencies in the Rouge Watershed under the Michigan  General
Permit.
                                                    342

-------
    The Michigan General Permit for municipal stormwater discharges is quite flexible  and allows those seeking coverage
under the permit to  use a wide variety of approaches to meet the public education, illicit  connection/illegal discharge, and
public involvement requirements. This flexible framework has allowed communities to experiment with various approaches
that recognize local constraints and seize upon  unique opportunities to meet the desired  outcomes. While the basic
requirements for what must be in the watershed plan are more detailed, the permitees  within a watershed are allowed
considerable freedom in deciding  upon their own priorities, remedial actions,  and  schedules. Pollution prevention initiatives
that are expected to be proposed by the communities will likely involve commitments to  continue  or expand  current
activities,  like soil erosion and sedimentation control; implementation of new activities to address priority issues such as
failing septic systems; and implementation of regional projects to reduce the frequency and velocity of storm flows in the
river.

    Table 1 and Table 2 outline the variety of public education and illicit discharge detection and elimination approaches
identified  by the communities and public agencies in the Rouge Lower 1  Subwatershed  (Figure 1).

    Across the watershed, communities actively sought ways to cooperatively address illicit connection/illegal discharge
investigations and public education projects. In one subwatershed group, a community with experience in the production
of videos agreed to  make a river  stewardship video that all other communities within the subwatershed  could use on cable
television, or through the distribution of cassette copies to local  libraries and/or schools. In the same watershed, another
community offered the use of a consultant to solicit bids for freestanding  public information display boards and to develop
stormwater information materials for the boards that could be used by all communities  at public gatherings and inside
public facilities. One community obtained the support of the local college to house and provide administrative support for
a well-established non-profit organization, the Friends of the  Rouge organization, whose public information  activities were
subsequently funded by several subwatershed groups to implement portions of their public education plans.

Advantages of  Watershed Approach

    The following section presents some of the lessons  learned as the communities and agencies in the Rouge Watershed
are beginning to implement the watershed-based, Michigan General Permit for  municipal stormwater discharges.

    Holistic Solutions/Local Ownership.  There are distinct advantages in managing stormwater on a watershed basis.  From
the work already completed on the Rouge Project, it is clear that an integrated approach is needed to address all sources
of water pollution and excessive flows in this urbanized watershed. By requiring those agencies and communities with
responsibility for stormwater to work together at the subwatershed level to establish goals and objectives, local agencies
and the state regulatory agency are forced to view solutions holistically. To achieve the desired level of river restoration,
there must be integrated action  plans that address not only stormwater but failing  on-site sewage disposal systems
(OSDS), CSOs, sanitary sewer overflows (SSOs), and significant nonpoint sources of pollution.

    Ideally, a watershed-based regulatory  framework  should  encompass all  dischargers so that  pollution sources  can be
addressed holistically. Practically, it must be recognized that prior NPDES permit programs at the state and federal level
are already in place for municipal and industrial point source waste treatment discharges, and for many  industrial and
commercial stormwater discharges. While the Michigan watershed-based stormwater General Permit  covers only public
agencies that own, operate, or  control  stormwater conveyance systems not currently under a Phase  I  Stormwater Permit,
the required  watershed management plan does provide aframeworkfor integrating activities under  other permit programs.
In addition, the  General Permit gives communities and agencies  the flexibility and encouragement to  incorporate nonpoint
source controls and  pollution  prevention  activities  as part of the required watershed  management plan. For example, many
                                                      343

-------
Table 1. Illicit Discharge Elimination Activities to be Implemented under Michigan Stormwater General Permit in Communities in Lower 1 Subwatershed,
Rouge River Watershed

Legal Basis
Locating Problem Areas
Finding the
Source
Minimizing Seepage
from Septic Fields and
Sanitary Sewers
SP
1
8
Activity
Existing ordinances for control of illicit connections and/or OSS have
been determined sufficient or community/agency will evaluate existing
ordinances
Will adopt additional ordinances for control of illicit connections and/or
OSDS if determined necessary
Review existing monitoring data to prioritize investigation areas
Plan developed w/County to locate sources of suspicious discharges
previously identified
Develop, modify, implement and/or continue to use complaint system
Procedure to coordinate complaint response/follow up
Develop and/or use GIS for tracking complaints and/or activities
Train field employees for identification & reporting of illicit discharges
Mapping of storm system, jurisdictions and/or locations of outfalls
Systematic dry weather screening of outfalls or manholes
Investigate possibility of systematic screening program
Screen drainage from facilities under jurisdiction
Dye testing when additions made to existing facilities
Establish priority of suspicious outfalls and/or initiate follow up visits
for further analysis of flow
Investigate to find sources of suspicious discharges using upstream
manholes or dve testing or televising
Identify and/or map areas served by OSDS
Determine feasibility of sewer extension/mitigation
Proposals for future sanitary sewer construction will consider existing
OSDS
Develop agreement/cooperate with county for implementing an OSDS
evaluation program
OSDS evaluation prior to sale of property
Continue sanitary sewer maintenance program
Reporting to MDEQ on investigations, violations found & corrective
actions taken
Investigate Funding Mechanism for Stormwater Related Tasks
Canton Township
II
X
X
X
X
X
X
X
X
X
X


X
X
X



X
X
X
X

Plymouth 1
Township 1
X
X
X
X
X


X
X
X



X
X



X

X
X

Van Buren |
Township
X
X
X
X
X
X

X
X
X



X
X
X

X
X

X
X
X
Wayne County























Salem Township 1
X
X
X

X
X
X
X


X



X




X

X

Superior
Township ||
X



X
X

X



X




X



X
X

Ypsilanti
Township |
X
X
X
X
X
X
X
X
X




X
X



X
X
X
X

                                                                344

-------
 Table 2. Public Education Activities to be Implemented  Under Michigan Stormwater General Permit by Communities in Lower 1 Subwatershed  Rouqe
 River Watershed
Activity
Cable TV ads
Clean Sweep program
Coordinate with Master Composters Program
Co-sponsor annual River Day
Co-sponsor educational workshops
Co-Sponsor Healthy Lawn & Garden Workshop
Cosponsor informational outreach workshops
Co-sponsor River Stewards program
Co-sponsor River Watch program
Co-sponsor Rouge River Day
Co-sponsor Rouge Education Project
Co-sponsor Rouge Friendly Neighborhood Program
Co-sponsor Rouge Friendly Business Program
Display maps of community, watersheds & boundaries
Distribute miscellaneous brochures and/or feet sheets
Distribute Rouge Recreational Guide
Distribute Rouge Repair Kit to homeowners
Distribute septic system maintenance packet to homes with OSDS
Distribute storm water general information package to new residents
Heighten visibility & promote school water/resource monitoring
Periodically provide Rouge Information Kiosk system in public buildings
Presentations
Provide articles, Information in community newsletter
Provide fliers/messages in water bills or tax notices
Provide water quality educational information on Website
Public service announcements
Publicize garden hotline
Publicize illicit discharge hotline
Storm drain marking
Tributary signage
Utilize "Our Actions" display at various community events
I
Canton Township
X


X






X



X
X





X
X





X

X
Plymouth 1
Township |

X



X



X
X
X
X


















Van Buren I
Township Q
X



X









X
X





X
X





X

X
Salem Township II




X



X

X


X
X

X
X
X
X


X



X
X



Superior |
Township |
X


X



X
X

X
X


X





X
X
X
X
X
X
X
X
X
X
X
Ypsilanti 1
Township
X

X


X








X




X

X
X



X
X
X

X
communities have initiated  pilot projects to evaluate how stormwater best management practices (BMPs) will control
stormwaterflow and prevent pollution. In some cases, these pilot projects have  permanently changed the way communities
and/or government agencies manage  stormwater. These  management practices will be  included,  as part of a watershed
management plan, and credit will be given to the entities that are performing those functions.
                                                       345

-------
    Many of the subwatershed units selected in the Rouge River Watershed involve communities that have combined
sewers, separated sewers, and OSDS. Some individual  communities have  all  three within their  corporate limits.  Once the
communities began to work together at the subwatershed level to establish goals to achieve water quality standards
necessary to restore the river, each found that they had a significant role in the process and that the control of flow in the
rapidly developing headwater areas was as significant as CSO problems in the lower portion of the watershed. Evaluating
the sources of water quality problems and/or the threats to existing uses of the river at the subwatershed level by local
agencies is leading to a better understanding of local constraints, opportunities for innovative solutions, ownership of the
long-term river restoration effort, and interagency cooperation.

    Overcoming Institutional/Regulatory Barriers. Local agencies and  communities in  urbanized areas  have a long history
of cooperative efforts to address the delivery of common  public services. Recent trends in Michigan, and elsewhere in the
country, to reduce the size and cost of government and limit local taxing power have accelerated efforts at the local level
to integrate or share the cost of a broad range of government services.  Local agencies are increasingly seeking ways with
their neighboring jurisdictions to reduce the cost of police and fire protection,  solid waste disposal, libraries,  recreational
facilities, infrastructure  maintenance and repairs,  public transit, water supplies,  and sewage disposal. Unfortunately, except
in a few isolated instances where a single authority has been created to oversee all aspects of water management, the
legal responsibility for stormwater is widely dispersed among local communities, county health and drain agencies, road
agencies, private developers, and autonomous school districts and  public colleges. The creation of a new level of
government in the form of a water  management authority with broad powers has been resoundingly rejected in the Rouge
River watershed by local agencies and is likely to receive the same reception in  many other urban areas of the country.

    State  and  federal water quality regulatory programs have  traditionally focused  on large point sources where
responsibility for obtaining  and complying with  specific permit limits  is easy to  establish. The regulatory framework to
control water pollution has generally  discouraged  rather  than encouraged  cooperative  solutions among communities and
has relied upon command and control to achieve results. The complexities involved in addressing wet weather pollution
problems  in urban areas, and the widely dispersed accountability for  managing  stormwater, demands a new regulatory
framework that will encourage cooperation among the  locally responsible public agencies to design integrated, holistic
solutions. The watershed  approach  to stormwater regulation developed in Michigan offers an opportunity to overcome the
institutional  and  regulatory impediments that  have discouraged cooperative  local  approaches  to restoring urban
watersheds.

    Institutional arrangements and financing options  necessary to implement  the General Permit  and  subwatershed
management  plans are among  the  many elements the local communities in the Rouge Watershed are addressing in their
working groups. As part of the subwatershed planning process, communities and agencies are also identifying issues
which cross subwatershed  boundaries. Rouge Project  staff and the MDEQ currently provide  coordination of individual
subwatershed efforts and assist subwatersheds in developing a comprehensive strategy for addressing watershed-wide
issues. The subwatershed communities are also identifying those activities, such as public education and water quality
monitoring, that may be most cost-effectively performed throughout the entire watershed by a single entity.

    Increased Local Accountability and Political Support. Building a watershed restoration project from the bottom-up by
helping local communities and agencies identify problems, set their own priorities for restoration,  select their own remedial
measures and  propose their own schedules requires a sharing of power among federal, state,  and  local agencies not
usually found in water pollution control programs. The General Storm Water  Permit program in Michigan is  voluntary at
this time and  it  has allowed state  regulators the ability to  provide  flexibility that might not otherwise be available. It has also
increased  the accountability  of local agencies who no longer have the  freedom to blame  federal and state officials for the
impositions of requirements, but now  are responsible for convincing local elected  officials that the  programs proposed are
in the best long-term interest of the local residents.

    Opportunities for Cost Eff iciencies/lnnovation.  As discussed earlier, the Rouge River communities  that have obtained
coverage under Michigan's General  Storm  Water Permit  and are working in subwatershed groups have  already developed
more  cost effective and efficient methods to meet public education  requirements through cooperatively  developed projects.
Similar joint  programs  are underway to train local community and agency staff in  illicit discharge elimination activities and
in sharing staff and equipment to conduct river and enclosed storm drain surveys. The three county health agencies are


                                                      346

-------
developing common  approaches to permitting and inspecting OSDS.  The county  road agencies are working with the state
highway agency to address the design, construction, and maintenance of transportation drainage systems.

    The county agencies in the three counties responsible for designated storm drains are working together and with local
communities toward  implementing  common standards for stormwater  management at new developments. County and local
officials  have worked together  to establish  protocols for rapidly developing independent GIS to assure that databases can
be integrated to assist in watershed-wide water quality/quantity management. The economic and political cost for each
community or county agency to develop these approaches has been an impediment in the  past.  The watershed approach
has enabled these cooperative programs to be established. It  is anticipated that the pollution prevention initiatives required
following completion of the watershed management plans would also involve joint projects.

    Establishing a broad range of cooperative programs  to address stormwater  problems across jurisdictional boundaries
is, in of itself, innovative. However, with the development of comprehensive watershed  plans, new practical approaches
to implementing total  maximum daily load (TMDL)  requirements of the Clean Water Act and  effectively using water pollution
trading options created at the state level become  possible. The Rouge River National Wet Weather Demonstration  Project
is funding  a pilot program to verify that the watershed management framework under the Michigan Stormwater General
Permit can be used to meet the TMDL requirements, ahead of state schedules (and at potentially lower cost), and the
objectives  of the Clean Water Action Plan program.  In addition, the pilot program will demonstrate how the General Permit
watershed framework can be  used as a basis for the proposed statewide water quality trading program currently under
development.

    The top-down, command-and-control approach often requires repeated threats or  legal action by state and federal
regulators to ensure compliance with requirements due to lack of political will at the local level. Locally generated
watershed plans containing specific action schedules that have been  adopted by elected boards, commissions, and
councils are  less likely to be abandoned or require  enforcement actions to  assure compliance. Peer pressure  and  citizen
support at the local level  should be sufficient incentive in most  instances for each  local agency to complete their
responsibilities on schedule. Where legal enforcement action is required, the state and  federal agencies are more likely
to find support among other local agencies who have met their obligations as outlined in the joint subwatershed plan.

Conclusions and Recommendations from a Community Perspective

    Local  communities  in  southeast  Michigan  and the state regulatory  agency are  attempting, for the first time, a
consensus, cooperative  approach to stormwater management and regulation under the NPDES program. The Michigan
General Permit is a watershed-based, general stormwater permit issued under the National Pollutant  Discharge Elimination
System. The permit requires  permitees to immediately initiate activities, such  as illicit discharge elimination, and  to
participate in watershed  management planning for a self-determined hydrologic unit. The watershed management plan
forms the basis for implementing watershed goals and objectives, including improved water quality and pollution control.
This new regulatory program implements the watershed  approach endorsed by USEPA and others and should facilitate
watershed-based integration of control  programs for different pollution sources, such as  stormwater CSOs which may  be
present within  a large, urban watershed. In addition, it is  believed that the new watershed-based stormwater program in
Michigan will achieve the objectives of the TMDL program, the Clean  Water Action Plan Program,  and water quality trading
programs under development across the country.  From the  perspective of local  communities and agencies, this approach
provides optimum flexibility to  solve the most pressing problems in  their subwatersheds by empowering them to identify
problems,  choose from  alternative solutions, establish  priorities and  schedules, and develop  common strategies with
neighbors. Communities and  others involved in this new program  are also addressing issues such as coordination of
subwatershed  efforts within larger watersheds.

    The Rouge Project (and others) have shown that by holistically addressing all sources of pollution, a cost-effective
action plan can be implemented to address impairments and restore river uses. Storm water issues cannot be corrected
in a vacuum, but must be integrated into an overall solution that addresses the physical, chemical, and biological stressors
in a waterway.  Stormwater adversely affects all three and,  therefore,  must be woven into the  fabric of the overall watershed
management plan and watershed control program. Without this integration,  stormwater control will become another "add
on" program that misses an opportunity to encourage an integrated program that addresses all sources of ecosystem


                                                     347

-------
stress in a cost-effective, prioritized manner. The approach being followed in the Rouge River Watershed should prevent
misplaced efforts and, most importantly, result in a restored  Rouge River on a much faster timetable.

    A key objective of the Rouge River National Wet Weather Demonstration Project has been to demonstrate alternative
methods to a "command and control," top down regulatory approach for water quality protection and improvement. The
alternative methods sought by the Rouge Project leverage "bottom-up" approaches that put place-based needs in the
forefront and use local initiatives to make progress on water quality restoration. This approach has led to a number of
institutional changes in the watershed that will  help sustain the watershed management efforts into the future. From the
perspective of the communities involved, the cooperative, iterative approach being followed appears to be working and
is a welcome change from traditional "command and control" relationships with regulatory agencies.

    The Rouge Project approach demonstrates that watersheds can be "managed." When water quality objectives can
only be reached through control of CSO, stormwater, and nonpoint sources, watershed management must involve the
active participation of local units of government. From a community perspective, this local involvement is critical to the
overall success of the Project and to stream restoration. Also, from a community perspective, undertaking a watershed
effort is not a simple matter. Watershed planning and  implementation takes a large commitment of time and  effort.

    The communities involved in  the Rouge  Project have a sense of overwhelming success  with the watershed restoration
efforts to date. Water quality and ecosystem health are improving, and the demonstration techniques have resulted not
only  in concrete and steel structures, but in  real  institutional changes that integrate the work of stormwater and watershed
improvement into the basic institutions of government. Most importantly, the public is able to  utilize new canoeing areas
and other river-based amenities, which are now possible due  the noticeable  improvement in water quality, aesthetics, and
other attributes of the river.  It  is hoped that this effort, and the  work of the  Rouge River  National Wet  Weather
Demonstration Project,  will continue to  identify and  quantify the benefits of cooperative, watershed-based efforts to  protect
and restore our nations water resources.

Acknowledgments

    This paper represents a summary of select elements from the ongoing efforts of many individuals and organizations
involved in the restoration of the Rouge River in southeast Michigan. The authors also gratefully acknowledge the
assistance of Ms. Sandra Kiser during the preparation of this manuscript.

    The Rouge  River National Wet Weather Demonstration Project is funded, in  part, by the United States Environmental
Protection Agency Grant #X995743-01  through 04 and #0995743-01. The views expressed  by individual authors are their
own and do not necessarily reflect those of USEPA. Mention  of trade names, products, or services does not convey, and
should not be interpreted as conveying, official EPA approval, endorsement, or recommendation.

References

Cave, K. and J. Bails (1998). "Implementing a Model Watershed  Approach Through a State General Storm Water NPDES
Permit". Proceedings of WEFTEC 98.

Fredericks, R., K. Cave, and J. Bails (1997). "Adapting Regulatory  Frameworks to Accommodating Watershed  Approaches
to Storm Water Management." Proceedings of WEFTEC 97.

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.

U.S.  Environmental  Protection Agency  (January  9,1998). "Draft Phase II Federal Storm Water Regulations."40 CFR Part
122 & 123.
                                                    348

-------
                   California's Model  Urban  Runoff Program  (MURP):
                     Urban  Runoff Programs for Small Municipalities

                                     Jennifer Hays, Assistant Engineer
                                         Public Works  Department
                                        City of Monterey, California

                                   Cy Oggins, Coastal Program Analyst III
                                       California  Coastal Commission
                                           Central Coast District
                                          Santa Cruz, California
Background
    Monterey Bay is a crown jewel of the California Coastline and has received special protection under the National
Marine Sanctuaries Act since September 1992, and the California Coastal Act since 1976. The 5,300 square mile Monterey
Bay National Marine Sanctuary includes a number of small coastal communities, and ranges from the City and County of
San Francisco on the north to Cambria on the south. The cities of Monterey and Santa Cruz have long recognized that
protection of the unique marine resources within the Sanctuary is  critical to the economic vitality and quality of life of their
communities. Monterey Bay, with  its world renowned  Monterey Bay Aquarium, rich bird and marine resources, recreational
opportunities that include the Santa Cruz Boardwalk, and commanding vistas has become a major tourist attraction.

    The Cities of Monterey and Santa Cruz developed and implemented a Model Urban Runoff Program (MURP) in a
cooperative team effort with the Monterey Bay National Marine Sanctuary, the California Coastal Commission, California
Regional Water Quality Control Board-Central Coast Region, and the Association of Monterey Bay Area Governments
(AMBAG) funded by an EPA 319(h) grant. The MURP was developed  to address and support a number of environmental
regulations and initiatives that applied to the Sanctuary and adjacent coastal areas including: the Sanctuary's Water Quality
Protection Program, requirements of the Coastal Zone Act Reauthorization Amendment (CZARA), Coastal Commission
plans and policies, the Regional  Board's watershed  management initiative and basin plans, the State Water Resources
Control Board's Ocean Plan  and Nonpoint Source Pollution Control Program, and EPA's proposed draft Storm Water
Phase II Rule (Storm Water Phase II).

    The concept of the MURP originated in the Sanctuary's Water Quality Protection Program for Monterey Bay National
Marine Sanctuary-Action Plan Implementing Solutions to Urban Runoff and a State's Urban Runoff Technical Advisory
Committee Report, developed to address Section 6217 of CZARA.

    One of the most important drivers in the development and implementation of the  MURP was an initiative of municipal
leaders to address the value that the community places on protecting the local creeks, streams, and wetlands and the
Sanctuary's marine biological resources.

Development of  Model Urban Runoff Program

   A key objective of the MURP was to produce a document that would assist other communities in the development of
their own urban runoff programs by providing an off-the-shelf "how-to" guidebook building on the experience  gained
through the  development and implementation of Phase I  Storm Water  Management Programs. A  second objective of the
project was the development of Urban Runoff Management Programs forthe cities of  Monterey and Santa Cruz that would
address the community's values and  achieve early  implementation  of and compliance  with the various  regulatory programs
and initiatives.

    Representatives from Phase I municipalities who were responsible for the development of their programs and had
gained experience in implementation were consulted throughout the  project. They participated in workshops when the
MURP was presented to Monterey Bay  municipalities and provided extensive information and examples of material used


                                                  349

-------
in their communities, including what worked and what didn't. Woodward-Clyde Consultants, with its extensive experience
in working with Phase I municipalities, was selected to assist the project team in developing and producing the MURP.

    The cities used a conceptual framework for developing urban runoff programs appropriate for their individual needs
following the guidance recommended by the MURP. It included four major steps:

    • Assessment - Institutional assessment and assessment of environmental resources and sources of pollutants of
      concern

    • Development - Program management, institutional arrangements and coordination,  legal authority, and fiscal
      resources

    • Implementation - BMPs and model programs for each of the  six control measures proposed in Storm Water Phase
      II and for control of commercial and industrial activities

    . Evaluation - Progress reporting and evaluation, water quality monitoring, and program update.

    Periodic meetings were held during the two-year grant period to share, review, and comment on individual city work
products, review progress, prepare and validate the  MURP, and discuss early  implementation actions. The  Sanctuary was
instrumental in the early implementation effort by providing public  information and outreach  support and  developing public
education materials.  The Coastal Commission played an active  role in formulating  strategies for addressing  Sanctuary-wide
water quality and land use concerns  and  providing support in the development  of GIS-based land use maps. Project status
reports were presented to the Monterey  Bay Regional Stormwater Management Task Force, AMBAG, city councils, and
Monterey Bay area public works officers.

The Model  Urban Runoff Program

    The MURP contains these four major steps for the development  and implementation of an  urban runoff program, and
a corresponding appendix containing additional information,  examples,  models,  references, and contacts for further
information.

    The Assessment Phase of the MURP describes the importance of information gathering and research to provide an
early survey of the municipalities, current  policies, programs, legal  authorities, and  fiscal resources applied to control urban
runoff. A similar institutional assessment of existing  regional  storm water, watershed,  and other water quality control
programs is recommended to avoid  duplication  and  to identify potential conflicts, opportunities for coordination, and areas
not previously addressed. This phase also provides guidance  and methods for (1) describing  a community's water
resources and activities that can be impacted by polluted runoff, (2) mapping  the storm drainage system, (3) developing
a relationship  of pollutant sources/activities to pollutants of concern, and  (4)  developing a working map to assist in targeting
problem  areas or pollutant sources.  Coordination with  and  building upon  existing efforts,  including joining  Phase I
programs, is encouraged.

    The Development Phase of  the MURP  describes the (1) selection of the program management structure, (2) identifies
individual and departmental responsibilities for management of individual program elements (public education,  control
measures, or BMPs) and (3) coordination with other internal and external programs and agencies. The legal authority to
ensure implementation of BMPs and achieve compliance with the MEP standard of the Clean Water Act was developed
through use of a model ordinance. The appendices  include examples  of language for the amendment of Local Coastal and
General Plans  as  required by the State of California. Revisions to the California Environmental Quality Act checklist were
recommended to  provide planners an additional tool to evaluate impacts of runoff from both new development and re-
development. This Phase includes an estimate of staffing resources  to implement each element of an urban runoff
program. It also describes the use of assessment districts, storm water utility fees, and other sources of funding program
implementation.

    The Implementation Phase  of the MURP describes eight program elements including six required by EPA's Phase II
draft regulations as minimum requirements MURP  program elements  include (1) public involvement/participation, (2) a

                                                    350

-------
public education and outreach program, (3) an illicit connection and discharge detection and elimination program, (4) a
municipal operations control program, (5)  a  construction  site  discharge control  program, (6) a new
development/redevelopment control program, (7) an optional control program for commercial facilities,  and (8) an optional
control program for industrial  facilities. The  MURP recommends objectives,  BMPs, implementation activities, and  methods
for  program evaluation and documentation.  This includes measurable goals for each of the eight program elements. The
appendices contain numerous examples of public participation and education; BMPs for residential areas, food service
operations, municipal operations, construction sites, vehicle service facilities, and shopping centers; sample SWPPPs for
corporation yards and construction sites; and reporting forms.

    The Evaluation Phase describes methods for (1) determining whether water quality  is improving and whether the
efforts and resources are directed at the right source and pollutants of concern; (2) reporting progress using the BMP
measurable goals, and (3) the developing and implementing of water quality monitoring programs and volunteer monitoring
programs.  This phase provides and stresses the need  for procedures for modifying and updating the urban runoff program
using the evaluation tools.

    Each section of the MURP contains an extensive list of references to assist municipalities in obtaining additional
detailed supporting information on how these programs were developed.

Implementation of the Model Urban Runoff Program

    Municipalities in California's major metropolitan  areas were encouraged, and  in some cases required, to file for
NPDES' permit coverage on an area-wide basis. Numerous smaller municipalities are already regulated by  Phase  I
requirements. There are now approximately 260  municipalities, with a combined population of 29 million, regulated  by
Phase I NPDES permits in California.

    In California, 76 incorporated places and counties are proposed to be  automatically designated  and 38 areas outside
urbanized areas that could  be  potentially  designated  under Storm Water Phase II. The Monterey  Bay Area has 13
incorporated  places and counties that would be automatically designated. The MURP will be of significant benefit to a
number of smaller California  municipalities, and particularly in the area covered by this project.

    Undertaken  as part of a 319(h) Grant, this  project was required  to conduct  an  outreach effort to ensure early
implementation of urban runoff programs. Two workshops were held in April of 1998 for planning, public works, building,
parks, public information/education, and general municipal operations staff in the Monterey Bay Area. These workshops,
attended by approximately 120  individuals, covered an  introduction to urban runoff pollution, and regulatory requirements.
They  featured presentations from individuals experienced in the development  and  implementation  of Phase I storm water
management programs and included four break-out workshops covering MURP development and implementation.

    The agencies participating in the development of the MURP have undertaken a  number of actions, described in the
following paragraphs, to implement the project recommendations.

City of Monterey

    During the development  of the MURP, the City of Monterey mapped watersheds, major storm drains, key streams,
creeks and channels. They also identified and mapped automotive servicing facilities, restaurants, several industrial  sites,
and pest and weed management activities as potential sources of runoff pollution. Fifteen potential sources or activities
that could  contribute primary pollutants of concern were identified. The City has adopted a water quality ordinance and
established a monthly storm  water utility fee, currently $3.43, to implement its urban runoff management plan.

    In cooperation with the Sanctuary and Coastal Watershed Council, the City has  also initiated a citizens participation
program to label  storm drain  inlets and perform volunteer monitoring (Urban Watch Stormdrain Monitoring Program). This
volunteer effort led to the  development of a Restaurant Outreach program  to  educate employees and eliminate pollution.
The City has commenced implementation  of BMPs for new and existing  sources, conducted water quality monitoring,
distributed public education  material, and  is currently working to implement its construction site pollution  prevention
program.

                                                     351

-------
    In cooperation with the Sanctuary, the City has obtained a grant to install and test storm water treatment devices at
the Monterey Harbor and Fisherman's Wharf parking lots to determine the effectiveness of removing oil, sediment, and
trash.

 City of Santa Cruz

    The City of Santa Cruz developed a  computer-generated map of watersheds and land  uses to identify potential
 pollutant sources. The City adopted a storm water ordinance in April 1998, regulating all water entering the storm drain
system, prohibiting illicit discharges and connections, and requiring implementation of BMPs published by the City. The
City has drafted BMPs for vehicle service facilities,  retail shopping areas, residential areas, and  food service  facilities. The
City Industrial Waste Inspectors will conduct initial inspections of 100 vehicle service facilities  in 1999 to determine any
actions which must be taken to comply with the ordinance, with second inspections scheduled to formally determine
compliance. The City hosted an  outreach presentation of the program and the proposed BMPs for the business  community
during its Pollution Prevention Week.

 City of Watsonville

    The City of Watsonville began implementing a storm water program in 1991, through its industrial facilities  Source
Control Program, and completed a bilingual storm drain stenciling program in 1992. Subsequent to the development of
the MURP, the City has completed a review of existing programs and policies, developed a new storm water  ordinance,
started an illicit connection program that has sampled 50% of the City's storm drain outfalls,  and established a public
education program in cooperation with the Sanctuary. The City  plans to implement a municipal,  industrial, and commercial
source control program, a targeted educational outreach program, and a construction and new development program.

Monterey Bay National Marine Sanctuary

    The Sanctuary's Water Quality Protection Program addresses a  number of water quality issues in addition to urban
runoff and targets nonpoint sources of pollution. The Sanctuary's program supports the  cities' urban runoff programs by
developing and distributing educational materials on urban pollution and co-sponsoring teacher training workshops with
the Monterey Bay Aquarium. It also collaborates with the City of Monterey on volunteer monitoring programs  and public
education. The Sanctuary has prepared a  plan for addressing polluted runoff from agricultural lands and has received
commitments from the California Farm Bureau and six regional Farm Bureaus to form a coalition to address  this issue.
The Coalition will  focus on educating its members on polluted runoff, establishing  landowner committees and  pilot projects
in several watersheds, and strengthening farm management practices by developing grower self monitoring and serving
as a liaison with the Sanctuary  and the Regional Board.

    The Sanctuary and the City of Monterey have a cost-sharing agreement, which funds a Sanctuary employee at half
time in return for the development of a City public education  program. This agreement is going into its third year, and has
resulted  in the development of public education brochures, posters, exhibits, BMP  pamphlets,  and the Restaurant Outreach
Program. Current work is focused on the development of a Public Service Announcement, a construction site education
program for  developers and inspectors, and signage for Monterey's Harbor.

State of California - Coastal Commission and Regional Water Quality Control Board

    Implementation of the MURP has been identified as a  priority in the California Nonpoint Source  Pollution Control
Program's first 5-Year Implementation Plan, which the State Water Resources Control Board and California Coastal
Commission developed pursuant to the Clean Water Act and Coastal Zone Act Reauthorization Amendments of 1990
(CZARA).  Key actions that the State will undertake  include the distribution  of copies of the MURP Guidebook to California
cities and the providing of technical support and  training to cities developing Urban Runoff Management Plans using the
MURP.
                                                    352

-------
    The Coastal Commission intends to review the experiences of Monterey Bay Area cities in implementing the MURP
and coordinating with the Central Coast Regional Water Quality Control Board to revise the MURP Guidebook as needed.
The document will be made available to other small coastal cities by printed copies, CD-ROM, or Internet web site.

Summary and Conclusions

    The Model Urban Runoff Program project, which is funded through a 319(h) Grant has provided small communities
in the Monterey Bay Area an excellent opportunity to develop their own urban runoff programs and to develop and validate
an off-the-shelf" how-to" guidebook on development of urban runoff programs.  The MURP will  potentially benefit over 100
communities in California that will be required to develop urban runoff programs implementing the six minimum control
measures contained in EPA's draft Storm  Water Phase II  Rule and the requirements of Section 6217  of the Coastal  Zone
Act Reauthorization Act.

References

California Coastal Act of 1976. (P.R.C.§§ 30000 et seq.).

California Coastal Commission. 2nd Edition 1996. Procedural Guidance Manual: Addressing Polluted Runoff in the
California Coastal Zone.

California Storm Water Quality Task Force.  March 1993.  California Storm Water Best Management Practice Handbooks.

California Regional Water Quality Control Board-Central Coast Region. Water Quality Control Plan,

Monterey Bay National Marine Sanctuary. February 1996. Action Plan I: Implementing Solutions to Urban Runoff. Water
Quality Protection Program for Monterey Bay National Marine Sanctuary.

Monterey Bay National Marine Sanctuary. 1998. Annual  Report for the Monterey Bay National Marine Sanctuary.

City of Monterey, City  of Santa Cruz,  California Coastal Commission, Monterey Bay National  Marine  Sanctuary,
Association  of Monterey  Bay Area Governments, Woodward-Clyde, Central Coast Regional Water  Quality Control Board.
July 1998. Model  Urban  Runoff Program, A How-To Guide for Developing  Urban Runoff Programs for  Small Municipalities.

State Water Resources Control  Board and California Coastal Commission. July 1999. Draft California's Nonpoint Source
Pollution Control  Program. Volumes I and II.

U.S. Department of Commerce, NOAA, U.S. Environmental Protection Agency-Office of Water. January 1993. Coastal
Nonpoint Pollution Control Program-Program Development and Approval Guidance.

U.S. Environmental Protection  Agency.  January  9, 1998. National Pollutant Discharge Elimination System-Proposed
Regulations for Revision of the  Water Pollution Control Program Addressing Storm Water Discharges: Proposed Rule

U.S. Environmental Protection  Agency.  January  1993.  Guidance Specifying Management Measures  for Sources of
Nonpoint Pollution in Coastal Waters.
                                                    353

-------
                    New Stormwater Treatment BMPs: Determining

                     Acceptability to Local  Implementing Agencies


                                             Gary R. Minton
                                      Resource Planning Associates
                                           Seattle, Washington

                                              Paul  Bucich
                                       Pierce County Public Works
                                          Tacoma,  Washington

                                             Mark Blosser
                                       City of Olympia Public Works
                                          Olympia,  Washington


                                                 Bill Leif
                               Snohomish County Surface Water Management
                                           Everett, Washington

                                              Jim Lenhatt
                                       Stormwater Management, Inc
                                            Portland,  Oregon

                                            Joseph  Simmler
                                           Entrance Engineers
                                          Bellevue,  Washington

                                              Steven  True
                                              Vottechnics
                                        Federal  Way, Washington
       Over 200 local governments in the Puget Sound watershed of western Washington require new developments
to install Stormwater treatment systems. The retrofitting of existing developments is also often required. With the guidance
of the State of Washington', each local jurisdiction developed a list of approved treatment technologies, mostly public
domain technologies such as wet ponds and swales, in the mid-1 990's. Having an approved list raises the issue of how
to add new unapproved technologies, in particular the manufactured technologies such as swirl concentrators and drain
inlet inserts. A protocol is  needed by which local jurisdictions can determine acceptability of "unapproved" treatment
technologies.  This paper presents a protocol* recently developed to assist local jurisdictions in  the  Puget Sound
watershed.

Protocol Structure

    The protocol has four parts:

    1.  Performance criteria to  compare  currently  unapproved treatment  technologies with currently approved  treatment
       technologies.

    2.  Types of data  to be used in the evaluation and methods of sample collection.
                                                  354

-------
    3. Factors that are to be considered in the evaluation other than performance.

    4. Content of the report provided to the local jurisdiction by the proponent.

Method to Compare  Performance

    There are many problems inherent in the development of a protocol, not the least of which is that there is no agreed-
upon  definition as to what constitutes acceptable performance against which to judge equivalency. An informal "standard"
that is much discussed and that has been adopted by some jurisdictions outside the Puget Sound region is 80% Total
Suspended Solids (TSS) removal over all storms less than a specified event.

    A one-number standard of 80% may be  inconsistent with two complementary observations about the performance of
stormwater treatment systems. The first observation3 is that removal efficiency of sand filters is directly related to influent
concentration: the higher the  influent concentration  the  greater the efficiency. This relationship may hold  for other
treatment technologies. The second observation' in Schueler (1996) proposes that there is a lower limit to the effluent
concentration.

    Further, a reasonably  strong and direct relationship exists between runoff rate and TSS concentration5. This is
particularly germane  to our region with its mild storm intensities. A comparison  of data from our region to areas with higher
average rainfall intensities indicates that we tend to have lower TSS concentrations in untreated stormwater.  It is therefore
possible that over an aggregate of storms, we cannot achieve 80% TSS removal in the Puget Sound region.

    A more appropriate method to compare  approved and unapproved technologies may be to relate efficiency to the
influent concentration. This approach allows the pooling of data from sites with different pollutant concentrations.

    The protocol presumes that if we are satisfied with the technologies that  are currently approved, we should approve
an unapproved technology with  similar performance. Therefore, the starting point is to identify the performance  of currently
approved technologies: swales, wet vaults, wet ponds, constructed wetlands, and sand filters. The results are presented
in Figures 1 through 3. The data points in the figures are of flow-weighted composite samples from individual storms. Only
data from studies conducted in the maritime climate of the Pacific Northwest are used.

    Although this is  a large region, from northern California to southwest British Columbia, comprehensive studies have
been  conducted  only in western Washington.  Data considered acceptable are from two wet ponds, three grass swales,
and two sand filters. The protocol' provides detail on the studies that were reviewed. The reasons for using only Pacific
Northwest data differ with the technology. With wet ponds there is concern about the possible effect of differences in
regional climates on effluent quality.  For sand filters, possibly because of differing sand  chemistry,  filters  used elsewhere
may be able to remove dissolved phosphorus6and zinc', a capability our sands do  not have.  Swales studies conducted
elsewhere do not provide the information needed to judge whether they were sized to criteria similar to that used in our
region.

    Figure 1 for TSS is used in all comparisons, and is the  first performance requirement that must be met. If the  receiving
water is nutrient sensitive, Figure 2 is also used. If the water body is of regional  significance because it supports salmon,
a central issue in our region, Figure 3 is also used. Zinc was selected to represent all toxics primarily because it is the only
toxic with influent concentrations that are commonly high enough to allow for the evaluation of efficiency.

    Each figure has a "Line of Comparative Performance," the origins of which are discussed later.  Each line is drawn
so that most of the data points fall above and to the left. This is called the region of acceptable performance.  The data
points of the unapproved technology under consideration  would be plotted using the same format.  If most of the data
points fall above  and to the left of the "Line", it can be concluded that the candidate technology is reasonably equivalent.
What  constitutes "most" is up to the local jurisdiction.   The protocol does not specify a hard rule but offers these
suggestions as to the percentage of data points falling above the "Lines": TSS, 90%; phosphorus, 80%; and zinc, 90%.
Note that low efficiencies generally occur at low influent concentrations.
                                                     355

-------
                                           FIGURE 1  TSS Evaluation
                                              PNW data - Individual storms
             -80%
                                              80      100     120     140
                                           Influent concentration (mg/L)
                                    FIGURE  2 Phosphorus Evaluation
                                           PNW data - Individual storms
                                                                                  Swale


                                                                                  Wet pond


                                                                                  Sand filter
          -80%
                                150    200    250    300    350
                                   Influent concentration (ug/L)
    How were the "lines"  derived in Figures 1 through  3? Judgment was used to select a point of "irreducible"
concentration, and to draw the line from this point to the upper right.  Regarding "irreducible"  concentrations:  for TSS Tom
Schueler has proposed4 20 to 40 mg/L, depending on the treatment BMP. A value of 15 mg/L was selected because the
data in Figure 1 suggest that 15 mg/L is attainable.  For TP  (Figure 2), Schueler4 proposed a concentration of 140 to 330
ug/L, depending on the treatment system. A lower value of 90 ug/L was selected for the  same reasons  as with TSS.
Schueler4 drew no conclusions with regard to zinc. However, his analysis4 suggests "irreducible" concentrations of 36 ug/L
for ponds and wetlands, so 35 ug/L was selected. It should be noted in Figures 1 through 3 that most of the incidents of
low or negative efficiency occurred at low influent concentrations.
                                                      356

-------
                                      FIGURE  3 Zinc Evaluation
                                          PNW data - Individual storms
      100%
                  50      100     150     200     250     300     350
                                      Influent concentration (ug/L)
                                                                        400
                                                                                450
                                                                                        500
    Attempts were made to identify a  line of best fit from which confidence limits could  be derived under varying
assumptions such as excluding all data below a particular influent concentration and/or removal efficiency. However, the
relationships were so poor as to make confidence limits meaningless. It was concluded that basing the lines on some sort
of statistical  construct would give an air of rigor that  is unwarranted at this time.  Therefore, the lines were drawn using best
professional judgment. It is expected that the Lines of Comparative  Performance will change with time as additional data
are collected.

    Figures  1 through 3 can  be used for inlet inserts if these pollutants are being considered. If,  however, an insert is being
considered for the removal of petroleum hydrocarbons, it must be compared to the effectiveness of oil/water separators.
A graph comparable to Figures 1 through 3 is  not provided for oil/water separators because of the  lack of data. The
criterion commonly used in the Pacific Northwest is that the concentration of individual samples shall not exceed  15 mg/L.
This protocol could be  used to generate data to compare to this criterion.

Data and Data  Collection

    It is the responsibility of the proponent of the unapproved technology, either the manufacturer or the development
engineer,  to obtain the required data. The protocol identifies the minimum requirements. Local jurisdictions are free to
request more data.  The  protocol  specifies the requirements for three sources of data: field studies with real storms, field
studies with artificial storms, and laboratory tests. The protocol calls for discussion of advantages and disadvantages of
each method. It is  left to the local  jurisdiction to decide the weight to place on each source of data. However, the protocol
recommends that the local jurisdiction not rely solely on laboratory tests, particularly when considering the removal of
dissolved  pollutants or oil/grease and related  products.  It also recommends a size distribution  of sediment for the
laboratory tests.

    The protocol is very specific with regard to the types and amounts of data that are to be collected. This aspect of the
protocol is summarized in Table 1.
Table 1.  Data Requirements
                                                      357

-------
Item	Requirement
Number of test sites               One to three sites: medium density residential, retail commercial, and non-retail commercial.
Number of sampled storms         A minimum of 10 per site, total of 30 if one site
Storm depth                     50% to the design storm depth (1.35' to 1.70")
Runoff duration                   50% to 200% of the mean annual storm duration (7.5 to 30 hours)
Average  rainfall intensity           50% to 200% of the mean annual intensity (0.02 to 0.08'Vhour)
Type of samples                  Flow-weight composite
Minimum aliquots per storm        10
Storm volume sampled             Samples are to be taken over not less than 75% of the total volume of each storm
Parameters                      TSS, pH, total zinc, copper, and cadmium, TP, BAP, and TKN. include dissolved metals and phosphorus if it
                               is claimed that the technology can remove dissolved constituents.  With catch basin inserts where the sole
                               objective is the removal of petroleum  hydrocarbons, measure oil/grease, TPH, TSS, and pH.
Additional                        At the end of the test period, the sediment accumulated in the treatment system shall be removed, quantified,
                               and anlyzed. The sediment shall be evaluated for total dry weight, moisture content, particle size distribution,
                               organic content, and zinc. Use ASTM wet and dry sieve test procedures to analyze the particle size distribution.
	Also determine the amount of floatables, i.e.  litter, and petroleum products.	

    The protocol  states that for a data point to be used  in  the analysis of efficiency, the influent concentration of the
parameter  shall either be at least ten times its detection  limit, or  be greater than the "irreducible" concentration, whichever
is greater.

     Efficiency is to  be calculated three ways.

    Method  1: Removal in each sampled storm calculated as:

               100 (average influent - average  effluent)/average influent

    Method #1 is  required  because it provides the data  points to plot figures like Figures  1 through 3.

    Method 2: Aggregate removal of the sampled storms as:

               100(A-B)/A

    Where:    A =   (influent cone. Storm 1 )(flow Storm l)+(influent  cone, of Storm 2)(flow of Storm 2)

                     + . . .  (influent cone, of Storm N)(flow of Storm N)

               B =   (effluent cone, of Storm 1 )(flow of Storm 1 )+(effluent cone, of Storm 2)(flow of Storm 2)

                     + . . .  (effluent of Storm N)(flow of Storm N)

    Method #2 is specified because it provides an overall efficiency of removal over the period of the research.   If the
amount of sediment that has accumulated in the bottom of the treatment facility has been determined from the separate
lab test (See Section A.4), another calculation can be done to  check the above estimate of efficiency.   This second
calculation  is done as follows: subtract  B  from A, and then compare this difference to the amount  of sediment  determined
from the separate laboratory test described  in Section A.4. These calculations  can also  be done for zinc  and  phosphorus.

    Method #3 Efficiency based  on geometric mean:

             100(A-B)/A

    Where: A = geometric mean of all influent samples

            B  = geometric mean  of  all effluent samples

    Method #3 is specified because it is the most correct method of calculating efficiency,  although it has  been used
sparingly to-date. All influent and  effluent data from multiple sites can be pooled.


                                                          358

-------
Consideration  of Factors  Other than Performance

    The protocol suggests that the local jurisdiction consider additional factors in making a decision should the technology
pass the first requirement of acceptable  performance. The protocol recommends the following factors, although the local
jurisdiction is free to include other factors.

    . Site characteristics

    • Constructability

    • Reliability

    • Robustness

    . Receiving water sensitivity

    . Groundwater risk

    . Operation and  maintenance

    • Habitat creation

    . Thermal effect

    . Aesthetics

    .  Recreational  use

    • Community acceptance

    It is left to the reviewer to place a weight on the relative significance of each factor and to develop a scoring system.
For example, the factors  could be categorized and weighted as: "critical/necessary, ""important,"and "desirable." A relative
score, say 1  to 10,  could  be identified for each factor,  and multiplied by the  corresponding weight of each of the categories.

Content of  the Applicant's  Report

    The proponent of the technology is responsible for producing a report for the  local government conducting the
evaluation. The protocol provides a very detailed list of items that are to be included in the report.

    Explanation of the technology, such as:

    • How it works, how it removes pollutants

    . Where it is currently being used

    . Available models

    • Treatment and hydraulic capacities of each model

    • Documentation of the treatment and  hydraulic capacities

    • Sediment storage  capacities of each model up to the point of recommended maintenance

    . Floatables storage capacities up to the point of recommended maintenance

    • Sizing methodology

                                                     359

-------
• Materials used in the construction of the product

• Recommended maintenance procedures and frequencies

Documentation of the field studies:

• A description of the test site including: total acreage, total impervious acreage, a description  of landscaping if
  relevant, the acreage draining to the device if it differs from total acreage. A description of the  drainage system
  including the size of the sumps, and whether the sumps were cleaned prior to or during the test period

. A description of the model used

• Complete drainage calculations showing the calculations to size the treatment device

. All raw data including laboratory reports. All data are to be reported including rejected data with an explanation for
  the rejection

. Statement from  the analytical laboratory certifying that the appropriate procedures were followed in the preservation
  and analysis of the samples

• Calculation of efficiency of each storm by comparing influent and effluent concentrations

• Calculation of the efficiency for all storms  by comparing the total aggregate inflow loading of all storms to the total
  aggregate outflow loading for all storms

• A graphic  of data  points showing influent  concentration versus efficiency  for each storm sampled for TSS, zinc, and
  phosphorus. Plot all data, including rejected data, with an explanation for the rejection

• Start and end times of the precipitation and runoff periods of each sampled storm.

. Start and end times of the sampling  period of each sampled storm

. Antecedent conditions during the 72 hours prior to each sampled storm

. Total rainfall depth of each sampled storm

• Total runoff volume of each sampled storm

. Runoff volume that occurred during the period of sampling of each sampled storm

. Total rainfall  during the period of all of the sampling, i.e. from the first storm sampled through the last storm sampled

• Total runoff that occurred during the  period of all of the sampling, i.e. from the first storm sampled through the last
  storm sampled

.  If artificial  storms are used, identify the method and application rates of water and translate those  rates  into
  corresponding rainfall intensities

• Statement of certification  signed by the proponent indicating that the protocol was followed

Additional considerations with inserts for drain inlets:

. Data showing the effect of accumulated litter and leaves on performance, flow capacities

• The point in the  maintenance cycle that the field tests were run:  i.e. were the units tested "fresh," without prior field
  exposure or were they in the field for some time


                                                  360

-------
    Documentation of laboratory studies:

    •  Description of the composition of the test water

    .  The size distribution of TSS in the influent and effluent

    .  The test flow rates

    .  The performance at each flow rate

    •  Mass balance of influent, effluent, and collected sediment

Conclusions

    The protocol  offers a reasonable and defensible  approach that provides rationale for the consideration of technologies
that are not currently approved for use in new developments as well as for public projects. The protocol is most suitable
for the maritime climate of the Pacific Northwest. It is anticipated that the protocol will change over time, particularly as the
data base for approved technologies becomes more extensive and as we learn from its use. More performance data are
needed for public domain technologies located in our region, in particular wet vaults and constructed wetlands for which
there  are currently no data.

References

1.   Washington Department of Ecology, 1992, Stormwater Management Manual for the Puget Sound Watershed.

2.   Minton, G., M. Blosser, P. Bucich, B. Leif, J. Lenhart, J. Simmler, and S. True, in press, Protocol for the Acceptance
    of New Stormwater Treatment Technologies in the Puget Sound Watershed, APWA Washington Chapter.

3.   Bell, W., et al., 1995, "Assessment of the Pollutant Removal Efficiencies of Delaware Sand  Filters", Alexandria, Va.

4.   Schueler, T., 1996, Irreducible Pollutant Concentrations Discharged From Urban BMPs, Tech Note 75, Watershed
    Protection Techniques, 2, 2, 369.

5.   Roundtree, L, 1995, The Relationship Between Rainfall  Intensity and Urban Stormwater Constituent Levels: A
    Comparison  of Puget Sound to Other United States Climates  Using Pollutographs, Master's Thesis, University of
    Washington,  Seattle, Washington.

6.   City of Austin, 1996, Evaluation  of Nonpoint Source Controls, 319 Grant Project.

7.   Wellborn, C.T., and J.E. Veenhule,  1987, Effects of Runoff Controls on the Quantity and Quality of Urban Runoff at
    Two Locations in Austin,  Texas, U.S. Geological Survey, Water - Resources Investigations Report 87-4004.
                                                    361

-------
                                          By Any Measure.. .
                           Thomas R. Adams, P.E., Vaikko P. Allen and Andrea Perley
                                               Vortechnics, Inc.
                                                Portland, Maine
    The introduction of a diverse array of stormwater quality management tools in the last few years has created problems
for the growing number of individuals and organizations who would  like to compare the performance of these tools.
Comparison  is complicated by differences in treatment capacities,  targeted pollutants, and treatment approaches.  Several
methods of evaluation have emerged in response to the need for verification of theoretical performance predictions: yet
none of these "yardsticks" are appropriate in all situations and results from each are often not readily comparable to results
from other measures.

    Complicating the matter further is the confusion regarding what is  being compared.  In some cases, a technology will
be compared with  another technology. In other cases, the technology  is compared to a performance standard. (Analogy:
My maple syrup may be better than your maple syrup, but does that make it Grade A maple syrup?)

    The confusion stemming from this is greater than meets the eye. For example, many specifiying engineers and
hydrologists want to meet a performance standard of 80% TSS removal on an annual average  basis. They go to guidance
manuals and product manufacturers seeking something that will meet the standard. As the selection  process develops,
they grapple with cost, maintenance, the availability of land needed, and many other issues under the heading of "cost-
effectiveness" for their clients. But by the time the selection has been made, it has become more a question of who has
the better maple syrup rather than whether or not the selected product meets the standard. That is partly because 80%
TSS removal on an average  annual basis is virtually an impossible standard to achieve.

    It is not the purpose of this paper to propose the adoption of one standard over another.

    It is the purpose of this paper to review the merits of various measures of performance and, more importantly, to
stress that any monitoring program that attempts to measure performance of a stormwater quality management system
should begin with a consideration of how the observations will be reported.

   Any monitoring program  will consist of:

    .  Sampling

    •  Testing

   •  Reporting

    Often, researchers start with the selection of samplers and procede with questions of deployment and maintenance
of the samplers. Only when samples have been collected are the questions of testing, and eventually reporting, given much
thought. We propose that the reporting aspect, even though it is "last" chronologically, should dictate how the testing and
sampling are done. For  example, if the report is considered most informative when its focus is the  mass of pollutants
removed, as is often the case, there is no need for samplers. To get those results usually requires only the very simple
task of periodically  measuring accumulations of precipitated  sediments or floatable petroleum  products (or other pollutants)
and doing simple  arithmetic calculations to determine the mass removed.  We know of projects that went to the
extraordinary effort that it takes to procure, set up, operate, and maintain automatic samplers to obtain influent and effluent
concentrations, only to "back into" an estimate of the mass of pollutants removed by a convoluted set of "volume-times-
concentration" calculations. This is a classic case of doing something the hard way, not to mention the very expensive

                                                     362

-------
way. So the first rule of thumb for any monitoring project is to decide first how to report the results. Then design the
sampling and testing around the information that is to be reported.

    In this paper, we will present the following four "measures" of performance from which to choose in making those
decisions. We consider these measures to be the current "status quo of the art."

    •  Mass of pollutants removed

    •  Event mean concentration (EMC)

    .  Lines of Comparative Performance (Minton, et. al.)

    .  Plotting  efficiency versus operating  rate

    First, we would  like to  discuss in some detail some of the broader issues involved  in monitoring stormwater treatment
systems and measuring and reporting on their performance.

Setting

    The setting is nearly always in the field  or in a laboratory. We feel that the many benefits of testing in a laboratory are
generally underestimated by technical professionals  and non-professionals alike.  What we refer to as the setting tends
to be  pre-determined well  before a study gets underway. Since the setting tends to influence the important decision of
what  performance measure is best, this cursory overview is provided for perspective.

Field testing

    The drawback with  field testing is that it cannot be replicated very well. Every site is different. Every storm is different.
There are  "wet years" and "dry years." There are seasonal variations that can produce easily treated heavy sediment  loads
in winter and spring; hard-to-treat loadings, such as pollen and grass clippings,  in summer; and moderately treatable
loadings (leaves etc.) in the fall.  It is poor science to compare the results of a field test of any treatment system to the
results of a test of the same system at a different time and  place.

    Still, field testing has tremendous appeal because the stormwater and the sediments are "real."

    .  Field testing of individual facilities is usually adopted to evaluate the facility's performance  in comparison to a
      performance standard (such  as 80% TSS removal) to  see if it "measures up." This setting is the simplest  and  most
      common, and is adaptable to any of the measures. The treatment system can be set up to treat all runoff or it can
      bypass flows that exceed the treatment capacity.

    •  "Side-by-side" field testing  of several different facilities is increasingly popular, at least in concept. But, to our
      knowledge, this approach has not yet been successfully implemented. A key element is  the design of the "flow-
      splitter" that takes all of the runoff and "splits" it into an equivalent discharge to each treatment system. The easy
      part of the design of the flow-splitter is achieving equal flow rates of wafer to each of the facilities (and  even this
      "easy part" is not always all that easy). The hard part is getting equal  discharge of pollutants to each of the
      facilities.

    Whether testing  an  individual system or multiple systems, researchers have the potential to learn something very  basic
and very important from field testing that has, to our knowledge, eluded researchers to date. That is the determination of
an appropriate threshold for bypassing peak flows. Consider a site that is estimated by conventional runoff modeling to
discharge stormwater runoff at a rate of 3 CFS in a 1 O-year storm. Consider further two proposals for treatment. One
claims efficiency of 90% and a treatment capacity of 1 CFS. The other claims 80% and a treatment capacity of 3  CFS.
Which treatment option should be selected? If the prevalent standard of 80% is in place, the "safe"  choice would seem
to be  treating all runoff  from a lO-year event with 80% efficiency. But what if price is a factor? Maybe the system with 1
CFS capacity costs  less. Even at the same cost, would 90% efficient be preferable to  80%? And with that consideration


                                                     363

-------
comes the $64,000 question. Will the system that claims 90% efficiency with a capacity of just 1 CFS even meet the
standard of 80% overall? How much of the pollutant load will bypass the system altogether?

    Some proponents of small-flow/high-efficiency technologies have stated that 90% of all storms are less than 1 inch
of total rainfall and, therefore, treating 90% of all rain  at 90% efficiency will yield a net annual  removal of 81%. This
argument is fundamentally flawed. It assumes that the 90% of rain from small storms carries 90%  of the pollutants. This
is simply not the case. The rate of mobilization of virtually all pollutants depends on rainfall intensity, not depth of rain.
Therefore, it  is  important to treat high-intensity flows resulting from  the infrequent event,  which tend to carry  a
disproportionately high pollutant load.

    If 1 inch  of rain falls in 24 hours, virtually any system that  is  reasonably proportioned, designed and, of course,
maintained for the treatment of stormwater will do a good job. Efficiencies of TSS removal should be in the 90% range if
the runoff is fairly dirty with silty-to-fine sandy sediments.

    It is questionable, however, as to whether or not all of the runoff would be dirty if the rain that produced the runoff
totaled 1  inch and fell over a 24-hour period. Intuitively, the "last flush" of such a storm would be very clean. But even the
first flush may be very  clean in comparison to what it would be if  1  inch of rain fell in 1 hour. This highly variable "dirtiness"
gives rise to another interesting question when trying to measure efficiency. That is the question of how to account for the
inevitable reduction in treatment efficiency when the water to be treated is clean in the first place.  No treatment system
can remove what is not there.  So it has been  argued that some  accounting should be made for the  fact that there  is some
lower limit to the physical treatment that can be provided. Minton's "Lines Of Comparative Performance" (see figure 2.)
take this important consideration into account and are discussed later.

The  "Double Whammy" of the "2-Month Storm"

    Infrequent, high-intensity storms are important to the effective treatment of stormwater for two reasons:

    •  Over time, the higher intensity of less-frequent rainfalls, and the resulting higher stormwater runoff velocity, is what
      transports most of the sediment off of streets.

    •  The treatment facility is overloaded by the high flow of water that is transporting the sediment at the same time that
      most of the sediment is being transported to it.

    Schueler and Shepp  (1993) performed  monthly observations documenting a random pattern of  accumulation and loss
of sediment in a study of 17 different oil/grit separators in Maryland. Overall, the losses of sediment "outnumbered" the
accumulations.  In  other words, the observed systems  lost previously accumulated sediments once every two months.
We have inferred from their work that the "2-month storm" is a reasonable benchmark for stormwater treatment.  To be
"measurably"  better than the poorly reputed  conventional oil/grit separator, a system or a  facility should be  able to
demonstrate,  at a minimum, that it can continue to function in the 2-month storm. If a system is found to lose sediment
in a 2-month storm, it  should not be considered any better than conventional technology. Similarly, if a system needs a
bypass to protect it from washing out in 2-month storms, it should be considered only marginally  better than conventional
oil/grit separators. Bearing in mind that high flows  transport much of the total sediment, treatment systems should  be able
to handle more than the 2-month storm without bypass. Otherwise, much of the total sediment  load may be discharged
to the receiving waters that the system is supposed to protect.

    Clearly the statements of the preceeding paragraph are more of a hypothesis than a statement of fact.  One way to
validate or invalidate the hypothesis is described  in the  following section on side-by-side testing in the field.

Side-By-Side  Testing

    Testing stormwater facilities "side-by-side" has recently become a very  popular idea. The premise is that a well-run
comparison of systems treating "the same stormwater and the same pollutants at the same rates  of flow" will go a long
way to reduce the tremendous "scatter" in the data that has been obtained to-date by testing individual systems at different
sites.  If two systems are evaluated at different sites, even if the study is carried out by the same  researchers using the

                                                     364

-------
same protocol, the results will probably not be comparable. Every site is different, and from the point of view of stormwater
treatment, differences that appear slight can actually be significant. We have observed a dozen systems installed on a
single site  (a large shopping mall parking  lot) which were specified by the same engineer and installed by the same
contractor at more or less the same time and, of course, subjected to the same weather. They exhibited decidedly different
results when we measured sediment accumulations in the systems. The sediment depths ranged from a light dusting to
accumulations of over two feet in less than a year.

    So it is important in  "side-by-side" testing that there be just one flow stream to the two (or more)  systems being tested
and that the flows be split, so that each  system gets exactly the same rate of flow and the same pollutant concentration
at all  times.

    The main benefit of "side-by-side" testing is that it can provide an answer to the question, posed earlier, of whether
it is preferable to have, the arrangement should be as shown in Figure 1.
                        Influent Sampling
                        Point 1
                                                       80% TSS Removal Efficiency
                                                       Treatment Capacity of 6.0 cfs
Effluent Sampling
Point 1
                                                             High Flow
                                                             Bypass 1
                                                      90% TSS Removal Efficiency
                                                      Treatment Capacity of 1.0 cfs
          Flow
          Splitter
   Effluent Sampling
   Point 2
                         Influent Sampling
                         Point 2
                                                             High Flow
                                                             Bypass 2
 Figure 1. Recommended Arrangement For Side-by-Side Field Testing.

     By sampling at points 1 and 2, the overall efficiency of the treatment system and bypass can be assessed objectively.
 Also, the question of "Which is the better system?" is answered. There are two shortcomings:

     . Lack of repeatability. If one system gets 80% efficiency overall and the other gets 70% overall during one year of
      testing, there is no assurance whatsoever that either number will be repeated the next year. The test  results should
       be regarded as indicative of performance. They are certainly  not an assurance of performance over time. Such is
      the inconsistency, or "noisiness," of stormwater treatment data. A study like this should be conducted over a period
       of no less than two years. If the second year's results  are  reasonably close (in terms of statistical correlation) to the
                                                       365

-------
      first, it can be considered complete.  If not, it would be tempting to average all results. We feel that it would be poor
      science to do so, however. With the seasonal variability of stormwater pollutant loadings, one year's results will
      produce a single data point. A second year's results will produce a second data point.  Many people seem to regard
      each storm result as a single data  point, but as long as standards continue to be based on "average annual removal
      efficiencies," that is simply not the case.

    .  Variability with other sites. We have already  mentioned the differences from one site to the next. The basic premise
      of side-by-side testing is to determine relative performance of two or more systems (i.e., which is best). As long as
      such a study is limited to this premise, the variability from one site to the next will not be a problem.  But we know
      from experience that any "study," even the most cursory, tends to be overly generalized. We  can  only caution
      against doing so.

Laboratory Testing

    Testing stormwatertreatment systems in a laboratory setting offers some very significant advantages over field  testing.

    •  It is repeatable and demonstrable.

    •  It is more productive in the sense  that decades of rainfall can be simulated in a matter of days.

    •  It is more economical in terms of labor, sampling equipment, and flow-metering equipment costs.

    Laboratory testing achieves these benefits by controlling operating rates, particle sizes, and pollutant loading. When
influent concentrations are very low, removal efficiency will be low; but for concentrations that are generally recognized
as representative for stormwater, all concentrations tend to produce comparable removal efficiencies.

    In the lab, a set of tests can be run using one particle size (at representative concentrations) at operating rates from
zero to the system's capacity. At the conclusion  of these tests, a curve can be drawn plotting efficiency versus operating
rate on the y-axis and x-axis, respectively. Such a curve typically slopes  downward to the right, reflecting reduced
efficiency and higher operating rates. Any point along a constructed curve  should be reasonably reproducible when using
the same influent sediment load.

    Subsequently,  a whole family of other curves can be constructed using different particles. Also, to  more  closely
simulate "typical" sediment, a graded sediment sample can be developed and tested in the same way.

    Laboratory testing should not be considered the "last word" in documentation of a system's performance, but can be
considered a "benchmark" which is very useful in comparing systems operating at flow rates up to their capacity. Some
field testing, where it is feasible, should  supplement the work in the lab and,  as previously discussed, side-by-side field
testing is the only way to determine the  impacts of bypasses  on different systems.

The  Four Most Common Measures of Performance

    1. Mass of Pollutants Removed

    This is easily the simplest approach  to stormwatertreatment measurements in the field. By measuring the depths of
sediment accumulations in the facility on  a  periodic basis, it  becomes a simple arithmetic exercise to  calculate the  volume
and mass of sediments removed  by the system.

    Additional information is made available by this measurement.  It may be recalled, from our earlier discussion of the
2-month storm,  that Schueler and  Shepp used measurements  of sediment  accumulations to  document the poor
performance of conventional oil/grit separators.

    Researchers should consider using  the same approach for the  newer technologies that have come along since their
important work was published. The approach can be made even more informative by correlating observations to such


                                                    366

-------
things as activities in the drainage area (e.g., winter sanding, sweeping, a spill, etc.) or meteorological events such as
observed rainfall intensities or precipitation depths.

    2. Event Mean Concentration (EMC)

    These are sometimes referred to as "flow-weighted" or "flow-based composite" samples. They are nearly always
obtained using automatic samplers, a flow-meter and a flow totalizer that arithmetically converts the flow rate measured
by the flow-meter to flow volume over time and keeps track of the volume.

    The sampler receives a signal that causes it to take a sub-sample when a programmed volume of flow is measured.
For example, one sub-sample might be taken every 200 cubic feet of flow through the system. Over the course of the
storm, all sub-samples would  be combined into one large sample container from which the concentration will be  obtained
that represents the flow- weighted average for the entire storm.

    Without a flow meter, the  samplers could be set up to take a "time-based composite" sample; i.e., to sample every 30
minutes. Flow-weighted samples  are much more representative, as a simple example will show. Consider a volume of
1,000 gallons with a uniform  concentration of 300 mg/l flowing at a uniform rate past the sampling point in 15 minutes,
followed by half as much volume (500 gallons) with a 100 mg/l concentration flowing  by in the next 15 minutes. The correct
representation of the concentration would be calculated as:

                                        (I.000x300) + (500*100)
                                              1,000 + 500

    Flow-weighted sampling  will  more accurately reflect this. For example, if the sampler were programmed to pull a
sample  every 500 gallons, then  2 samples  at the  higher concentration would  be taken and just one at the lower
concentration. The average concentration would be calculated as:
    Time-based  sampling would, if the programmed time interval were 15 minutes, take one sample with a concentration
of 300 and another with a concentration of 100, and the average would be calculated as:

                                               300 + 100 =200
                                                   2

    Automatic samplers that can take flow-based composites have become a very valuable tool for sanitary engineers
measuring concentrations of pollutants in wastewater. We believe that they have been too quickly applied to stormwater
monitoring without regard for some of the inherent differences.  Waste streams have "highs and lows" of both flow rate
and concentration, but they are not nearly as wide as the variability of stormwater, which can change from flow rates of
zero to a deluge in a matter of minutes and concentrations that can also exhibit a minimum of zero. These "spikes" can
cause very brief periods of negative efficiency if a system is prone to washing out (as stormwater systems were shown
by Schueler and Shepp to do regularly). If a wash-out occurs, it is an important phenomenon to note, but the briefly
elevated concentration in the effluent will be "composited" with the rest of the (presumably lower) effluent samples. This
will  reduce the "event-mean-concentration," but will not reveal that a washout has occurred. Noting washouts, and the
flow rate that caused them, is a very important aspect of a stormwater monitoring program; but they are not likely to ever
be revealed  by EMC data.

    The second drawback of EMC data is that when influent concentrations drop to very low levels that cannot be further
reduced by physical treatment,  the efficiency, as measured by EMC's, will be reduced. This tends to obscure the  fact that
higher efficiencies can be achieved when they need to be achieved; i.e., when influent concentrations are higher.

    3. Minton's  "Lines of Comparative Performance"


                                                    367

-------
    It is widely acknowledged that there is a lower limit to the capabilities of physical treatment systems for stormwater.
This means that it is  very  unlikely  that  effluent concentrations would ever  be zero.  It also  means that  very  low
concentrations would not be significantly reduced.

    Minton et. al. has proposed the following mathematical expression to describe this lower limit:

                                             Influent - LowerLimit
                                                   Influent

If the lower limit is 20 mg/l, then a plot of this expression  is that shown in figure 2.
                Performance Standard = (Influent Concentrati9n - Lower Limit) / Influent Concentration
           c
           
-------
                        Pollutant removal efficiency vs. operating rate for various particle sizes
                                                                                                250 Micron
                                                                                                150 Micron
                                                                                                100 Micron'
                                                                                                75 iMirisnp
                                                                                                63 Micron
                                 20           40           60   >       80

                                      Percent of maximum opeating rate
100
Figure 3. Removal Efficiency versus Operating Rate.

    Field data is less likely to fit the relatively tight curves that can be generated in the lab. At the same operating rate,
you may have vastly different influent concentrations,  particle gradations, organic  content, etc.,  depending on  such factors
as the time of the storm, anticedent dry period, and time of year.  Removal efficiency is a function of all of these factors
combined.

    We  feel that this performance measurement technique and presentation  is the most informative. Its repeatability under
controlled conditions makes it ideal for comparing one system to another.  Certainly, if one system's performance curve
on 100-micron particles,  for example, is higher at all flow rates than another,  it could reasonably be judged to be the higher
efficiency system. If the curves are similar at low rates of operation, but either system drops down to zero efficiency at
some higher flow rate, that flow rate should, of course, be considered the peak capacity for that system. This approach
cannot show compliance with any standard for a stated percentage of TSS removal on an annual average basis.

Conclusion

    To our knowledge,  these  four measures represent all of the techniques  that have been used to measure the
effectiveness of various stormwater treatment systems.

    Measuring sediment  accumulations in the field provides  a good  deal of  useful information on mass removals and the
ability to retain (or fail to retain) previously captured pollutants during  periods of high flow. This approach costs very little
to implement.

    Event-mean-concentrations are the most widely accepted measure, but  may  not report all  efficiencies and will  almost
certainly allow any failures to go undetected. This approach requires the use of automatic samplers at considerable cost,
in terms of both time and money.

    Minton's Lines of Comparative Efficiency are more fair to the treatment system because they account for the inability
of any system to remove pollutants that are not present (or  present in very  low concentrations).  If EMC data is collected
to plot against the lines, then there are the same drawbacks of cost and automatic samplers allowing failures to go
undetected. Both of those  drawbacks can be overcome, but only with a very dedicated effort to take samples manually.
Taking manual samples throughout the duration of all storms  is very time-consuming  and unpleasant work. For that reason,
it is almost never done.
                                                     369

-------
    Plotting efficiency versus operating rate, whether in the field or in the laboratory, is arguably the most informative
approach. In the field, automatic samplers are used (with individual samples in individual bottles and not composited), so
there are those costs to consider. In the lab, samplers are needed but the construction  of a model treatment facility, and
the pumps and tanks to handle the required flow rates and volumes of water, will more than offset that cost saving.

    Since none of these measures provides an ironclad confirmation that the widely  prevalent  standard of 80% TSS
removal is being met, we submit that a different standard should be adopted by stormwater management jurisdictions.

References

Minton, Gary,  M. Blosser, P. Bucich, B. Leif, J Lenhart, J Simmler and S. True, in press, Protocol for the Acceptance of
New Stormwater Treatment Technologies in the Puget Sound Watershed, APWA Washington Chapter.

Schueler, Tom and  David  Shepp, The Quality of Trapped Sediments and  Pool Water Within  Oil Grit Separators  in
Suburban Maryland, Metropolitan Council of Governments, 1993 (revised).
                                                    370

-------
                      Stormceptor  Hydrology and Non-point Source
                                  Pollution  Removal Estimates
                                                 G. Bryant
                                            Stormceptor Canada
                                        Etobicoke, ON, L3T 3K2, CAN

                                                 R.  Grant
                                New England Pipe Wauregan, CT, 06387, USA

                                               D.  Weatherbe
                                      Donald G. Weatherbe Associates
                                      Mississauga, ON, L4X 1  H7, CAN

                                                 V. Berg
                                          Stormceptor  Corporation
                                         Rockville, MD, 20852, USA
Abstract
    A model has been developed to estimate total suspended solids (TSS) removal using Stormceptor, an oil/sediment
separator. The  model was based  on  a  commonly  used, continuous simulation  model United  States Environmental
Protection Agency Stormwater Management Model (USEPA SWMM) for hydrological processes.  The suspended solids
loading was estimated using build-up and  wash-off equations.  The solids were assumed to be distributed into five particle
sizes for settling calculations. Simulations were conducted using various assumptions of loading and settling velocities to
determine the sensitivity of the  model to assumptions. Simulations were  also conducted for a diverse range of geographic
areas to determine the sensitivity of the  TSS removal rates to  regional hydrology. The model was sensitive to the selection
of settling velocities and pollutant loading. The model was less sensitive to changes in hydrology, although significant
changes in hydrology did impact TSS removal estimates.

Keywords: Stormwater; suspended solids; model; hydrology; Stormceptor, separator

Introduction

    The Stormceptor  is a mechanical water quality separator designed to remove oil and  sediment from Stormwater. A key
feature of the design is an internal high flow by-pass to prevent scouring and  re-suspension  of previously trapped
pollutants. Since the separator is based on treating "the everyday storm," the effectiveness of the separator is dependent
on the distribution of pollution  in Stormwater and the frequency and magnitude of Stormwater flows throughout the year.

    In 1995,  sizing guidelines were derived for the Stormceptor based on field monitoring  of sludge  accumulation over time
in Toronto, Ontario, Canada. The accumulation data were used to derive estimates of annual total suspended solids (TSS)
removal. Two key assumptions were made in the 1995 analysis to estimate TSS removal: (1) a TSS loading rate of 185
mg/l based on theUSEPA's Nationwide Urban Runoff Program (NURP) median (U.S. Environmental Protection Agency,
1983),  and  (2) a sludge water content of 75% water.  Actual  Toronto rainfall data,  combined with the NURP TSS
concentration, provided estimates of annual TSS loading. Figure  1 shows the performance relationship derived from the
Toronto monitoring, which  forms the basis for the existing sizing guidelines.

    Toronto rainfall time-series data (5  minute timestep) were input to a continuous hydrologic simulation model (SWMM
Version 4.3) to determine  the percentage of annual  runoff treated based  on the sizing  criteria shown in Figure  1. The
                                                   371

-------
                                                10.0
                                                           15.0
                                                                      20.0
                                                                                 25.0
                                       Total Stormceptor Storage (m3/ha)
Figure 1. TSS Removal versus Stormceptor Total Storage (Toronto, 1995).

analysis of Toronto rainfall indicated that 80% - 90% of the annual runoff would be treated if the Stormceptor were sized
according to these 1995 guidelines. This study was initiated to address concerns about the applicability of the Toronto-
based sizing criteria to meteorological conditions in other regions.

Methodology

    A computer simulation model was developed based on USEPA's SWMM Version 4.3. Solids build-up, wash-off, and
settling calculations were added to the hydrology code to estimate suspended solids capture by the Stormceptor.

    The model accommodates the use of either  an EMC (event  mean concentration) or  build-up/wash-off calculations to
estimate suspended solids loads. The build-up/wash-off model is more theoretically and physically correct.  The EMC
method has been shown to provide  reasonable estimates of total solids  loads (Charbeneau and Barrett, 1998) alone, if
the distribution of the load is not important.

    The distribution of pollutant load is important for measures that incorporate  a  high-flow by-pass (commonly known as
"first flush" measures). Accordingly, preference  is given to the build-up/wash-off calculations to correctly distribute the
pollutant load with flow, recognizing the need to  optimize the sizing of small-site stormwater quality measures.

    In the model, solids build-up and wash-off are both approximated using an exponential distribution. The distribution
of solids build-up  is a function of antecedent dry days according to Equation  1  (Sartor and Boyd, 1972).
               P, =Pi + (PA-Pi)(1-e-kt)

    Where:     P, = solids accumulation up to day t (kg)
               P = maximum solids build-up (2.4 kg/ha)
               A = drainage area (ha)
               Pj = initial solids load on the surface (not washed off from the previous storm) (kg)
               k = exponential build-up factor (0.4) (days-1)
               t  = antecedent dry days
(1)
                                                      372

-------
    The maximum solids build-up  (P) load  was adjusted to provide similar long-term solids loading rates (124 mg/l) when
compared to the EMC method. An exponential build-up factor (k) of 0.4 was used based on previous literature (SWMM
4.3 Users Manual). A k-value of 0.4 translates into 90% of the maximum solids build-up occurring after 5.66 days. Once
the pollutant  build-up  reaches the 2.4  kg/ha  limit,  additional  build-up  is not allowed  (assumed to be wind  re-
suspended/driven off the surface). Wash-off is estimated using  Equation 2.

               P. = Pie-"                                                                           (2)

    Where1      ^ = so''c's remaining on the surface at day t (kg)
                Pi = initial solids  load (from equation 1) (kg)
                k = exponential decay factor (0.2) (mm-1)
               V = volume of accumulated runoff from the surface (mm)

    The exponential decay factor (k) of 0.2  was based on a review of previous literature that indicates k-values range from
0.03 to 0.55 (Alley, 1981; Charbeneau and Barrett, 1998).

    Charbeneau and Barrett (1998) found  that the simple wash-off  model  adequately described observed  solids  wash-off
in Austin, Texas.  Other researchers have cited that the wash-off Equation (2) is reasonable  for fine material  but may not
be  reasonable for larger solids that require a high rainfall intensity for mobilization  (Metcalf and Eddy,  1991; Ball  and
Abustan, 1995). The SWMM  model treats wash-off as a function of the runoff rate to account for mobilization. This
correction is applied indiscriminately to the entire solids load and does not account for the variation in wash-off rate with
particle size. If an "availability" factor is applied to all particle sizes uniformly, the model will underestimate the wash-off
of solids with increasing runoff volume if the majority of particles are fine in size. The approach taken in this study was to
use an availability factor for particles 400 [Jm in size or larger.  Smaller particles follow the simple wash-off estimates given
by  Equation  2. The larger particles (> 400 |Jm) require  greater runoff intensities to induce wash-off according to the
availability factor  provided in Equation 3.

               A = 0.057 +0.04(r)11
                                                                                                    (3>
    Where-     A = availabilitV factor
                r  =  runoff rate (mm/h)

    Equation 3 is based on research by Novotny and  Chesters (1997). The  runoff rate is used instead of rainfall intensity,
recognizing that the wash-off will  lag the rainfall based on the time of concentration.  The availability factor varies each
timestep and is only applied to the runoff volume for that timestep, as dictated in Equation 4.  The availably factor has an
upper limit of 1.

               V = Vi + A(Vt)                                                                        (4)

    Where:     V = accumulated runoff volume used in Equation 2 (mm)
               V, = accumulated runoff volume prior to current timestep (mm)
               A = availability factor (equals 1 for particles smaller than 400 (Jm)
               V, = runoff volume for current timestep (mm)

    The correction in Equation 4 effectively re-defines the accumulated runoff volume to be the runoff volume sufficient
to mobilize the particles. This  methodology requires more  accounting in the model but provides a more physically correct
wash-off  model.

    The separator was treated as a completely stirred tank reactor (CSTR). Alterations to the concentration of solids in
the separator will vary according to Equation  5 (Tchobanoglous and Schroeder, 1987).

               CV = QC( - QC, - rcV                                                                (5)
                                                      373

-------
    Where     C' = the change in concentration of solids in the tank with time (kg/m3s)
               Q = flow rate through the tank (m3/s)
               C| = solids concentration in the influent to the tank (kg/m3)
               C, = solids concentration in the tank (kg/m3)
               V = tank volume (m3)
               rc = reduction in solids in the tank (kg/m3s)

    For gravity settling devices rccan be estimated using Equation 6.

               rc = VSC/D                                                                         (6)

    Where     rc = reduction in solids in the tank (kg/m3s)
               Vs = settling velocity of solids  (m/s)
               D = depth of tank (m)
               C = concentration of solids in the tank (kg/m3)

    Substituting Equation 6 into Equation 5, solving the first-order differential equation and integrating provides the general
form of the non-steady state solution (Equation 7) for the concentration in the tank at time t.

               C = QC/(V(VS/D + Q/V))( 1 -e'(Vs/D +Q/V") + C,e-(Vs/D + W)t                                   (7)

    Where     C = concentration in the tank at time t (kg/m3)
               Ci = concentration in the flow influent to the tank (kg/m3)
               C, = concentration in the tank at the beginning  of the timestep (kg/m3)
               Q = flow rate through the tank (m3/s)
               V = volume of water in the tank (m3)
               Vs = suspended solids settling velocity (m/s)
               D = tank depth
               t = time

    Equation 7 was used to estimate the suspended solids concentration in the tank, and in the discharge from the tank
each timestep. Equation 7 assumes the suspended solids are completely mixed within the tank volume.

    During periods without flow (inter-event  periods) the solids are not assumed completely mixed at the beginning of each
timestep and the depth of suspended solids in the separator decreases each timestep until all of the solids are removed
or there are subsequent flows into the separator. The concentration of solids in the tank during periods without flow was
calculated using Equation 8.

               C = C,(1-Vst/D)                                                                     (8)

    Where:     C = solids concentration in the tank (kg/m3)
               C, = initial solids concentration in the tank at the beginning of the timestep (kg/m3)
               Vs = settling velocity (m/s)
               t = timestep (s)
               D = depth of solids  in  the separator (m)

    The depth of solids (D) in the separator in Equation 8 decreases each timestep based on the settling velocity until all
of the solids are removed  or there are  subsequent inflows to the tank.

    The model can be used with either hourly or 15  minute rainfall data.  Fifteen minute data are preferred, recognizing that
the Stormceptor is only applicable for small drainage areas. Small drainage areas have short times of concentration and
require data with  a suitable timestep. Internally, the model performs calculations with a 5 minute timestep.
                                                      374

-------
    The choice of particle size distribution and settling velocities is a key part of the modeling exercise. Different settling
velocities can be applied to the same particle size distribution, based on the specific gravity of the particles or to account
for the effect of non-ideal settling or flocculation on settling. In this study, a typical stormwater particle size distribution
(USEPA, 1983) was used for analysis (Table 1). The distribution given in Table 1 is commonly accepted by most regulatory
agencies in North America.

    The model  allows the user to alter the percentages of each size based on site-specific conditions, if required.  In most
areas, it is anticipated that the particle size distribution will not vary significantly since it is primarily related to vehicle wear
and atmospheric deposition. There may be certain instances, however, where the native soils contribute loading and the
default distribution needs to be altered. The default percentages were used in our study.

                               Table 1. USEPA Default Particle Size Distribution
                                Particle Size (pm)
                                      20
                                      60
                                      130
                                      400
                                     4000
                                                    % by Mass
                                                        20
                                                        20
                                                        20
                                                        20
                                                        20
    Settling velocities were then assessed for each of the particle sizes provided in Table  1. Settling velocities were either
calculated or  based on empirical literature (USEPA, 1983).  The calculation of settling velocities for small particles  follows
Stokes' law (Equation 9) since the Reynolds number (Equation 10) is less than 0.3.
    Where
                Vs = g(p,-pjd2/18u

                Vs = settling velocity for particle diameter d (m/s)
                g = gravity (m/s2)
                ps = density of particles (kg/m3)
                pw = density of water (kg/m3)
                d =  particle diameter (m)
                u = viscosity of water (kg/ms)
                                                                                                      (9)
                N
                 R = VsdPw/u
(10)
    Where
                NR = Reynolds number
                Vs = settling velocity for particle diameter d (m/s)
                pw = density of water (kg/m3)
                d = particle diameter (m)
                u = viscosity of water (kg/ms)

    If the Reynolds number is greater than 0.3, drag on the particles reduces the settling velocity.  An iterative solution was
used (solving for the Reynolds number, drag coefficient, and settling velocity until changes in the settling velocity were
insignificant) for particle sizes with the Reynolds numbers. The drag coefficient is given by Equation 11, and the settling
velocity is calculated by Equation 12.
                CD = 24/N, +3/(NH0.5) + 0.34

    Where      CD = drag coefficient
                NR = Reynolds number

                Vs=(4g(ps-Pw)d/(3CDpw))05
                                                                                                      (11)
                                                                                                      (12)
                                                       375

-------
    Where      Vs = settling velocity for particle diameter d (m/s)
                g = gravity (m/s2)
                ps = density of particles (kg/m3)
                pw = density of water (kg/m3)
                d = particle diameter (m)
                CD = drag coefficient

    Table 2 provides a comparison of the settling velocities used in this study.

 Table 2. Discrete Particle Size Settling Velocities (mm/s)

 Particle Size (|jm)      S.G. = 1.3 calculated    S.G. = 1.8 calculated    S.G. = 2.65 calculated   USEPA (1983) empirical
 20                  0.07                0.17                0.36                 0.00254
 60                  0.59                1.57                3.23                 0.02540
 130                 2.50                5.70                11.20                0.12700
 400                 16.00               37.00               65.00                0.59267
 4000                180.00               300.00              450.00               5.50330
  S.G. = Specific Gravity


    The settling velocities that are based on the empirical USEPA data are 65 to 150 times smaller than the settling
velocities based on a specific gravity of 2.65. A specific gravity of 2.65 is commonly associated with sand-size particles
whereas the fines in stormwater are commonly associated with a lower specific gravity. The use of a higher specific gravity
may be justified, however, if the values are considered representative of the settling velocities of fines in a flocculated or
coagulated state. Research  indicates that there is  a high potential for coagulation amongst particles (Ball and Abustan,
1995) which will increase settling velocities and TSS removal rates. Furthermore, historical settling velocity calculations
have been based on discrete particle methodologies  (vertical settling column tests)  that do not account for potential
coagulation. Coagulation would effectively offset the settling velocity columns in Table 2 (i.e., discrete settling velocity for
60 )Jm represents coagulated 20 )Jm particle size).

    Numerous field tests on the Stormceptor (Labatiuk, 1996; Henry et al., 1999;  Bryant, 1995) have indicated a high
percentage of  fines in the Stormceptor. This empirical  evidence lends credence to the coagulated settling theory, indicating
that the USEPA discrete particle settling velocities may underestimate actual  TSS removal rates. Settling velocities based
on a specific gravity of 1.8 were chosen in this study as the default or benchmark selection. The solids loading was
segmented into the particle size distribution and the concentration of solids in each particle size was tracked individually
during the settling calculations.

Meteorological  Data

    Rainfall data from the City of Toronto (5 minute timestep, 0.25 mm  resolution, 10 years record, 1987-1996) was
agglomerated into 15 minute data for use with the model. Fifteen minute data were obtained for the entire U.S.A. from
Earthlnfo on CD ROM.  Stations were selected based on location, period  of record, data resolution and completeness
within the period of record. Data were also obtained  from CSR  Humes for various stations throughout Australia. The rainfall
data were converted into National Climatic Data Center (NCDC)  format for input to SWMM.

    Fifteen minute data were utilized, recognizing the small time of concentration that  would typically be encountered in
most Stormceptor applications. Simulations were also conducted using hourly data to determine the sensitivity of the
results to the precipitation timestep. Numerous  hourly stations were available on the Earthlnfo CD for this purpose. The
model uses a  5 minute timestep at all times regardless of the rainfall timestep.

Modeling Parameters

    SWMM models catchments  and conveyance systems  based on input rain, temperature, wind speed, and evaporation
data. Only rain data were used in these analyses. The default SWMM daily evaporation values (2.5 mm/day) were used.
Evaporation data will not be important in this analysis since the catchment area  is small  (c 10 ha) and  has minimal


                                                      376

-------
depression storage.  The Morton equation was chosen for infiltration. The method of infiltration chosen is unimportant  due
to the small amount of pervious area (1%). Table 3 provides a list of the parameters used in the SWMM model.


  Table  3. SWMM Area  Parameters

   Area - ha (ac)                               variable
   Imperviousness                              99%
   Width - m (ft)                                variable
   Slope                                     2%
   Impervious Depression Storage - mm (in.)          4.7 (0.19)
   Pervious Depression Storage - mm (in.)            0.5 (0.02)
   Impervious Mannings  n                        0.015
   Pervious Mannings n                          0.25
   Maximum Infiltration Rate - mm/h (in/hr)            62.5 (2.46)
   Minimum Infiltration Rate - mm/h (in/hr)            10 (0.39)
   Decay Rate of Infiltration (s"')                    0.00055

    The width of catchment  was assumed to be equal to twice the square root of the area.

Results

EMC versus  Build-up/Wash-off

    The suspended solids removal results based  on the build-up/wash-off model were compared to those based on an
EMC of 124 mg/l (USEPA, 1983) to demonstrate the sensitivity of the model to the different solids loading approaches.
The use of an EMC assumes an equal concentration of suspended solids in  all of the stormwater that is conveyed to the
Stormceptor.

    Figure 2 shows a comparison of results using an event mean concentration loading and build-up/wash-off loading,
given the  default particle size distribution and settling velocities based on a  specific gravity of 1.8.

    The results in Figure 2 show that the TSS removal rates using the EMC  approach are lower by 14% when compared
to the build-up/wash-off method even though the total loads are similar. This is expected due to the by-pass nature of the
Stormceptor. The estimated  TSS removals for the existing (1995) sizing guidelines, which are based on an early field study, are lower
than both the EMC and build-up/wash-off estimates for low values (50% TSS removal) of separator storage/drainage area and  are higher than the other
estimates for larger values of separator storage/drainage area (80% TSS removal).

    The range  of TSS removal values based on computer modeling is smaller than the empirical  TSS removal rates.
Doubling the size of unit for the same area results in an increase of 30% for TSS removal, based on the current sizing
guidelines, whereas the increase  in performance based on the modeling is less dramatic (a 5% to 10% increase in TSS
performance). This finding indicates that the modeling results will  be less sensitive to changes in the model size for  any
given drainage area.

Selection of Settling Velocities

A comparison was made regarding the choice of settling velocities using Toronto rainfall data and the build-up/wash-off
TSS  generation methodology.  Figure 3  provides the results of this analysis.  The TSS removal estimates using the USEPA
settling  velocities are  an average of 20% lower than the original TSS removal estimates, 29% lower than the estimates
using the SG=1.3 velocities, and 39%  lower than the estimates using the SG=2.65 velocities. These results indicate  that
the TSS removal performance results  are very sensitive to the selection of settling velocities.
                                                      377

-------
Figure 2. TSS Removal vs. Loading Methodologies
on0/
g1 80% -
5f 70% -
o 60% -
» 50% -
co 40% -
CO
^ 30% -
| 20% -
< 10% -
no/.
Ti * -^
•" — — y ^^*^ "

• EMC
• Build-up/Wash-off
—a— Existing Sizing

6.00 8.00 10.00 12.00 14.00 16.00
Storm ceptor Storage (m 3/ha)
Figure 2. TSS Removal vs. Loading Methodologies.

Annual Flo w Treatment

    Numerous regulatory agencies design stormwater quality measures using a "design" event. The design event used
generally ranges from the 25 mm storm, or annual storm, to the 25-year storm. The modified SWMM  program  was used
to calculate the percentage of annual runoff that would be treated (not by-passed) with different by-pass flow rates. This
analysis was conducted using the Toronto rainfall for a drainage area of 2.25  ha. Figure 4 shows that the volume of runoff
that is treated prior to by-pass quickly becomes asymptotic with increasing treatment flow rate. A device that treats 30 Us
prior to by-pass would treat approximately 80% of the annual runoff. A device that treats 70 Us (over 2x higher flow rate)
only treats 10% more runoff (90%). Although the relationship between conveyance (% of annual runoff treated) and TSS
removal is non-linear, Figure 4 shows that high-rate treatment devices are not required for small drainage areas.

    The relationship provided in Figure 4 will vary with local meteorological conditions and is inherently accounted for in
the TSS removal modeling.

Regional  TSS  Removal  Performance Analysis

    The model was used to compare results from different areas in North America and Australia to determine the effect
of regional hydrology on TSS removal performance. All analyses were conducted using 15  minute rainfall  data and  based
on the TSS build-up and washoff model and settling velocities for a specific  gravity of 1.8.

    Table 4 shows the results for various sized Stormceptors with a 2 ha drainage area. The locations of stations listed
in Table 4 were selected to cover a wide geographic area, provide rainfall on a 15 minute timestep with a  0.25 mm
resolution, and provide results representative of large nearby cities. Most data from city airports are recorded hourly, and
therefore were not included in the comparison. The  results in Table 4 are plotted in order of decreasing performance
expectations in  Figure 5.

    Of the 16 stations analyzed, 12 stations provided TSS removal estimates within ±5% of the Toronto values.

    Although the majority of stations provided similarTSS removal estimates,  there were  areas with significant differences.
The performance estimates were lowest for the  southeastern United States. This area  is well know for its intense seasonal
rainfall distribution. Figure 5 indicates that the TSS  removal rates may vary up to 20% under different hydrological
conditions on the same land use/site conditions. The use of local or regional rainfall data is therefore appropriate for  design
purposes.
                                                     378

-------
90% -i
80% -
_ 70% -
ss
:r 60% -
re
| 50% -
1 46%' •
V) 30%
(/}
*~ 20%
10% -
0% 4
Figure 3. TSS Removal Performance vs. Settling Velocities
* **-r
iff- Z"-
D ^^.^
«*r*i
n ' <
V V
g^* ^ , "a"
>U A
10

0 USEPA
• 1 DSG=1.3
x:3G=1.SG=1.8
A SG=2.6S=2.6

6 8 10 12 14 16
Stormceptor Storage (m3/ha)
Figure 3. TSS Removal Performance vs. Settling Velocities.
                               Figure 4. Annual Runoff Treatment
                                        (Toronto - 2.25 ha)
                                20
40            60

Flow Rate (Us)
80
10C
Figure 4. Annual  Runoff Treatment.
                                                379

-------
      Table 4. Regional Comparison of TSS Removal Performance (2ha)
            State/  Province
            Colorado
            Alberta
            British Columbia
            California
            Massachusetts
            Ontario
            New South Wales
            New York
            North Carolina
            Queensland
            Minnesota
            California
            Maryland
            Missouri
            Florida
            Texas
                                   Location
Fort Collins
Calgary Forest
Vancouver
Davis
East Brimfield Lake
Toronto
Sydney
Rhinebeck
Cataloochee
Brisbane
Le Sueur
Orange County
College Park
Miller
St. Lucie New Lock
Houston Addicks
                  Stormceptor Model (CDNIUSA)
300/     750J      1500/     3000/    5000/    6000/7200
450     900      1800      3600     6000
  49%      63%       65%      71%      76%        79%
  48%      63%       65%      71%      76%        79%
  48%      65%       66%      71%      76%        78%
  44%      61%       63%      69%      74%        77%
  43%      59%       61%      67%      73%        75%
  43%      58%       60%      66%      72%        75%
  42%      57%       59%      66%      72%        76%
  41%      57%       59%      65%      71%        74%
  41%      56%       58%      64%      71%        74%
  41%      55%       57%      64%      71%        74%
  41%      56%       57%      84%      70%        74%
  39%      57%       59%      65%      71%        74%
  37%      53%       54%      61%      67%        70%
  34%      50%       51%      59%      65%        69%
  30%      43%       44%      52%      59%        64%
  27%      41%       42%      49%      57%        61%
                            Figure 5. Regional Comparison of TSS Removal Performance
                                                      (2ha)
                              3001450    750190KB   1500/1800  300013600  5000/6000  6000/7200
                                             Stormceptor Model  (CDNIUSA)
Figure 5. Regional Comparison of TSS Removal Performance

Rainfall Timestep

    An analysis was conducted to determine the sensitivity of the model to changes in rainfall resolution. Results based
on hourly rainfall data (0.25 mm resolution) were compared to those based on 15 minute rainfall data, to determine the
impact of using the hourly data. Hourly data are more readily available than 15 minute data and most large cities have
airports that collect rainfall on an hourly basis.
    The model reads the hourly data as rainfall that falls during the first fifteen minute timestep of each hour. This will
produce  higher intensities  since  the  rain  is not distributed correctly over the  entire  hour. The greater intensity  is
compensated for, however, by the completeness of the hourly records which translates  into  a greater number of small
rainfall values.
    Four areas were  analyzed (Rockville, Maryland;  Boston, Massachusetts; Miami, Florida;  and Houston, Texas). The
results of this analysis  (Figure 6) indicate that the use of hourly data does  not significantly alter the TSS removal estimates
                                                       380

-------
for units that are designed to remove over 40% of the annual TSS load. Greater discrepancies can be expected at large
ratios of drainage area to separator storage.
 90% -
 80%
 70% -
60% -I
50% -i
40%
30% -
 20% -
 10% -
  o%
                        Figure 6. TSS Removal versus Rainfall Timestep
                                      (by Stormceptor  Model)
                                                                                     -•-MA (15 min)

                                                                                     4  MA (60  min)
                                                                                       A  MD(15 min)
                                                                                     —A— MD (60 min)
                                                                                     —•— FL(15min)
                                                                                     -e-FL  (60 min)
                                                                                     —•— TX (15 min)
                                                                                     -e-TX  (60 min)
                                             3         4
                                          Stotmceptor Model
Figure 6. TSS Removal vs. Rainfall Timestep (by Stormceptor Model).

Conclusions

    The TSS removal results were sensitive to the selection of settling velocities for the specified particle distribution.
Differences in TSS removal of up to 40% were obtained, depending on the settling velocities that were evaluated.

    Results were also affected by the TSS loading method. The  use of an EMC underestimated  TSS removal performance
by approximately 15%, when compared to using the build-up and wash-off equations. This difference is expected since
the EMC method increases the load that is by-passed and  provides higher loads during higher treated flow rates when the
detention time, and hence settling effectiveness of the unit, is reduced.

    The model indicates that high percentages of the annual runoff can be treated with low-flow treatment devices such
as the Stormceptor. The model also predicts that the TSS removal performance is less sensitive to the size of separator
than observed from previous field studies.

    Regional hydrology affected the TSS removal estimates  provided by the model. Although differences of up to 20%
were observed, significant hydrological differences between the sites were needed to obtain this variance. Most of the
rainfall station locations tested provided TSS removal estimates similar to those of Toronto,  where the original sizing
guidelines were developed.

    Testing the model with different rainfall timesteps (15 minute versus hourly) indicated that hourly rainfall records can
provide an adequate estimation of performance if the rainfall  is collected  at adequate resolution (0.25 mm increments).

    The modeling indicated that significant TSS removal rates can  be achieved  using small  infrastructure control measures
if the drainage area is limited. The results lend credence to the positive field monitoring results obtained to-date for the
Stormceptor,  and to the concept of small storm  hydrology being the predominant  parameter for urban  stormwater quality
design.
                                                     381

-------
References

Alley, W., Estimation of Impervious-Area Washoff Parameters, Water Resources Res., 17, 1161, 1981

Ball, J., and Abustan, I, An Investigation of Particle Size Distribution durina Storm Events from an Urban Catchment,
University of New South Wales, 1995

Bryant, G., Misa, F., Weatherbe, D., and Snodgrass, W., Field Monitoring of Stormceptor Performance, 1995

Charbeneau, R., and Barrett, M., Evaluation of methods for estimating stormwater pollutant loads, Water Environment
Research, Volume  70, Number 7, 1998

Henry, D., Liang, W., and Ristic, S., Comparison of Year-Round Performance for Two Types of Oil and Grit Separators.
Draft  paper,  1999

Labatiuk, C.,  Nataly, V., and Bhardwaj, V., Field  Evaluation  of a Pollution Abatement Device for Stormwater Quality
Improvement, CSCE Environmental Engineering Conference, Edmonton, 1997

Novtony, V., Unit Pollutant Loads. Water Environment & Technology, 1992

Novtony, V., and Olem, H., Water Quality:  Prevention, Identification,  & Management of Diffuse Pollution. Environmental
Enaineerina Series, Wiley & Sons, 1997

Sartor, J. , and Boyd, G., Water Pollution Aspects of Street Surface Contaminants. EPA-R2-72-081, U.S> Environmental
Protection Agency,  Washington, D.C.,  1972

Tchobanoglous, G., and Schroeder, E., Water Quality, University of California at Davis, 1987

Tchobanoglous, G., and Burton,  L.,  Wastewater Enoineerina.  Treatment. Disposal & Reuse. Water Resources  &
Environmental Enaineerina. McGraw-Hill, 1991

U.S. Environmental Protection  Agency, Final Report of the Nationwide Urban Runoff Program, Water Planning Division,
Washington, D.C.,1983

U.S. Environmental  Protection Agency,  Storm Water  Management Model, Version 4.3. -User's -Manuali Washington.  D.C.,
1988
                                                   382

-------
                                 NPDES Phase II Cost Estimates
                                             Andrew J. Reese, P.E.
                                 Ogden Environmental and Energy Services, Inc.
                                              Nashville, TN, 37211
 Introduction

    The United States Environmental Protection Agency (EPA) has published final rules expanding  the existing  stormwater
 NPDES permitting program to smaller cities and other urban areas throughout the United States.  Due both  to external
 pressures and directives from the current and past administrations,  EPA is conscious of attempting to make the current
 stormwater  NPDES program "cost-effective." For example:

    "EPA believes this rule will cost significantly less than the existing 1995 rule that is currently in place, and will result
    in significant monetized financial, recreational and health benefits, as well as benefits that EPA has been unable
    to monetize, including reduced scouring  and erosion  of streambeds, improved  aesthetic quality  of waters, reduced
    eutrophication of aquatic systems, benefit to wildlife and endangered and threatened species, tourism benefits,
    biodiversity benefits and reduced siting costs of reservoirs." 1

    ".. . the  Agency recognizes the continuing  imperative to  assure that environmental regulations accomplish statutory
    objectives in the least burdensome and most cost-effective fashion. As explained further in this preamble, the form
    and substance of NPDES permits to address the sources designated in today's  proposal would provide greater
    flexibility for the newly  covered sources than the existing "standard" NPDES permit."'

    While the "benefit" side  of the  proposed  regulations exists in  the realm of gross estimates, the "cost" side is also filled
with unknowns. What will the mandated and negotiated stormwater program cost a local community? Are there ways to
 reduce costs? What should a local community be doing now to prepare for this regulatory program? This paper seeks
to address these related questions.

    The final regulations were published on December 8,  1999 and the changes  from the draft regulations are only minor3.
 But it is still  not possible to say what the regulations  will cost everyone in toto. This is so because:

    .  there  is great flexibility inherent in the regulations to create a stormwater quality program tailored to meet an
      individual community's needs and situation;

    •  each permit writer has preferences and "hot buttons" that will  color what any particular program will look like; and

    •  each community setting is different in terms of climate, topography, pollutants of concern, and current condition of
      local  waters.
1  Federal Register, January 9, 1998 p. 1536
2  ibid. p. 1550
3  Federal Register, December 8, 1999 pp. 68722-68851
                                                     383

-------
Basic Approach to Permitting

    Under proposed § 123.35(g), an  NPDES  permitting authority issues a general permit to authorize stormwater
discharges from regulated small municipal separate storm sewer systems.  The NPDES permitting authority will also
provide a  menu of regionally appropriate and field-tested Best Management Practices (BMPs) that the permitting authority
determines to be "cost-effective."The regulated small municipal  separate storm sewersystems could choose to select from
this menu or select other BMPs that they feel are appropriate.

    Under Phase II each  regulated community will need to develop a set of BMPs under each of six specific program
minimums. These BMPs can be any combination of programs, structures and other controls that, in the agreed opinion
of the  permit writer and the regulated community, meet the standard of reducing pollution discharge to waters of the state
to the Maximum Extent Practicable (MEP).  In this process, permittees and permit writers would evaluate the proposed
stormwater management  controls to determine whether reduction of pollutants to the MEP could be achieved with the
identified BMPs. EPA envisions that this evaluative process would consider such factors as condition of receiving waters,
specific local concerns, and other aspects included in a comprehensive watershed plan.

    Under the proposed  approach,  implementation of BMPs  consistent with  stormwater management program
requirements at § 122.34  and  permit provisions at § 122.33 would constitute  compliance with the standard of "reducing
pollutants to the maximum extent practicable."  That is, "if you do what you say you will do,  you  are by  definition in
compliance." It is  important to note that states implementing their own NPDES programs may develop more stringent
requirements than  those proposed in the Federal Register.  In fact, we  anticipate that many states will require more specific
and  rigorous  requirements under special  circumstances relating to the condition  of the receiving water within, and
downstream from, the community. For example, if a certain stream is required to  have a Total Maximum Daily Load
(TMDL) or  similar  study performed on it (for example, a watershed assessment for the purposes of wastewater treatment
plan permitting or expansion),  the NPDES stormwater Phase II permit conditions may reflect the allocation of pollutants
to that community.

    The steps  for  a community are: (1) review the conditions of the general permit,  (2) develop and submit a Notice of
Intent (NOI) to comply with the general NPDES permit through description of a BMP-based  program under each  of the six
minimum controls or program areas (see below), (3) negotiate this proposed  program with the permit writer, (4) receive
approval  of the submittal,  and (5) begin implementation of the conditions and programs described in the NOI including
record keeping and submittal of appropriate reports describing attainment of "measurable goals" for each BMP  as
described in the NOI.

Current NPDES Phase II  Program Cost  Estimates

    There is naturally much speculation on the actual program elements and costs for a particular stormwater program
developed under Phase II.  There  have been  several  attempts at  estimating Phase  II program costs based on current costs
of "similar" programs.

    In  the draft regulations, EPA had provided estimates of the  probable cost implications of the NPDES Phase  II Permit.
These  estimates were based on summary information from the  permit applications from 21 Phase I cities. Very high and
very low figures were thrown out  by EPA in developing these estimates. Figure 1 shows the summary table developed by
EPA.

    The range depicted in Figure 1 is from $1.39 to $7.83 per person per year for the first permit five-year period, and
$1.28 to $5.63 for other permit cycles.   For a city of 50,000 that is a very wide range  of $69,500 to $391,500 annually for
the first permit cycle. This is clearly not helpful in attempting to estimate a specific community's costs.

    There  is question about the vagueness in  the regulatory language, and the high degree of potential flexibility inherent
in briefly described program elements. For example, for the first of the minimum controls the regulatory language states:
                                                    384

-------
                 P?j «Ui«l Ml-
                                          frsF **! j

„ ,1, r ,,.,,., 1... .,,,,, ,,,,'.„..,,,,- „,
.......
il for lir-,1
...
„.,
•$:•!££-
,H ^tH.,,,,,1 (
A Is
-.,

Iti.S
•3a^a^^™..-
i l'L"J."'" ''.'«'»'.'," >.;«'B<.«" T,'i,,,u,,
                                                                   DAPITA COSTS FOR i



Flra






2nd







Manvuie

( Pei'fTNt Oycl«
f>ubilc eoocftiioo
H-ubtta Involvement
IHiCil Diacruirge D«,E
OooM Oil* SW nunolf Oonirol
Po-wl ConotrucDon SW MH>
PF>/<;IM nf tu*i*ik;ip»l Ops
and 3rd «*«milt CyClB*

Public lnvotv»m«4il
IllkJ.I O>»<^tn.>u» »AC
CooM SltA SW Ruoofl Com ot
r>o«t C&mtlrut.aon SW M»«
PP-/OM of Municipal Os>«
inuitlctpa.ll-
pecied to
(p«r««nt)








30
1OO

ab
4
CT
Lo~.™«
ciipftn coats


SOO^

Otwt
0.0
1 O
o o

o o

0.0*

1,O9
o.r>
«^«K.«
^a^m ^.^fU


so.a-4

a.e*

1 OO
2.OO

034

a.t?

1 O»
*.OB
Figure 1. EPA Cost Estimates for Phase II NPDES Compliance.

    1. Public education and outreach on storm water impacts4.. You must implement a public education program to
       distribute educational materials to the community or conduct equivalent outreach activities about the impacts of
       storm water discharges on water bodies and the steps that can be taken to reduce storm water pollution.

       (You may use stormwater educational materials provided by your State, Tribe, EPA, or, subject to the approval of
       the  local government,  environmental or other public interest or trade organizations. The materials or outreach
       programs should inform individuals and households about the steps they can take, such as  ensuring proper
       septicsystem maintenance,  limiting the use and runoff of garden chemicals, becoming involved in local stream
       restoration activities that are coordinated by youth service and conservation corps and other citizen groups, and
       participating in storm drain stenciling, to reduce storm water pollution. In addition, some of the materials or outreach
       programs should be directed toward targeted groups of commercial, industrial, and  institutional entities likely to
       have significant storm  water impacts.  For example, information to restaurants on the impact of grease clogging
       storm drains and to garages on the impact of oil discharges. You are encouraged to tailor your outreach program
       to address the viewpoints and concerns of all communities, particularly minority and disadvantaged communities,
       as well as children.)

    The "regulatory" wording in parentheses is not mandatory  but suggested. There is wide room for interpretation of the
intensity and detail necessary to accomplish this minimum control.  The devil is always  in the details, and there will always
be great variability in what two different programs intend to do to  accomplish the same general goals.

    NAFSMA (1999a, 1999b)  published a survey on potential Phase II program costs responded to by 121 cities and
counties nationally. Ten communities responded with programs that had  three  or  more suggested elements in the first
minimum control: Public Education  and Outreach. The annual per capita costs for these ten ranged from $0.04 to $1 .17
- again a wide range.

    Of those responding, only one community stated that it  had program activity in each of the six  minimum control
measure areas  and it spent $15.11  per capita annually, well above the EPA estimate (the city has a population of about
25,000). Of the  121 respondents only 26 had  programs in at least three (most had only three)  of the six mandatory
minimum control areas, and these can be considered far from complete. Figure 2 shows the distribution of costs for these
26 programs. The vertical axis is the annual per capita cost for these elements.  The median was $1.44 and the average
was $4.07. The low value was $0.04 and the high was $26.00.
4Federal Register, January 9,1998, p. 1639.
                                                     385

-------
s m . 00
524.00 •
$22.00
s m . 00
$18410
$16.OO
914.00
$12.00
s lo.OO
saoo
$aoo
$4.00
$200
$- -1






n


^-"\
_ *. . — -/*" \ Pt



*- n v> t*. o> ^ t£ iPt 5^"
1

|





K
i •



fi
Figure 2. NAFSMA Study Program Costs.

    We can speculate that if many of these communities had a fully developed Phase II program, the average costs could
more than double, since each community would be adding both new program areas and upgrading  their existing programs
they had to make them comply with the details of the Phase II permit writers requirements.

    In the final regulations, USEPAtook a different approach to making estimates of the costs of compliance, using both
the NAFSMA information  and past experience with Phase I  (EPA, 1999). EPA estimated annual costs for the municipal
programs based on a fixed cost component and a variable cost component. The fixed cost component included costs for
the municipal application, record  keeping, and  reporting activities.  On average,  EPA estimated  annual costs of $1,525 per
municipality. Variable costs include the costs associated with annual operations for the six minimum measures and are
calculated at a rate of $8.93 annually  per household (assuming 2.62 persons per household). The the cost estimating
equation is:

                                 Annual cost = $1,525 + population/2.62*$8.93

    Finally, rule of thumb estimates based on the author's experience working in over 100 communities  indicate that
comprehensive stormwater programs that  include  advanced stormwater quality programs cost  between $7.00 and $20.00
per capita per year- above the EPA estimates. The quality portion is normally between 20 and 30% of the total average
program  cost.

Estimating Costs from Anticipated Programs5

    The methods used above do not provide details of the components of the stormwater programs resulting  in the costs,
and thus are not very helpful in assisting other communities in their thinking about the regulations. An effort was made
to develop cost estimate  ranges based on a direct interpretation of the stormwater regulations as applied to  example
communities at each end  of the spectrum, in terms of size and intensity of water quality program. This has an advantage
in that it deals directly with the stormwater regulatory requirements and illustrates specific program components so that
we can control and define all details. The following sub-sections will develop two hypothetical permit applications for the
six minimum controls.

The Two Permittees

    Permittee one ("Smallville") is a community of 10,000 that is adjacent to a larger city that has obtained a Phase I permit
or that can assist Smallville in many of  its permit responsibilities. It is a small bedroom community interested in compliance
with minimum disruption and cost. It does not really have an  engineering or planning component of its city staff, but relies
on a city administrator and hired consultants.
5 Based on a presentation made by Andy Reese of Ogden Environmental at the APWA seminar, "Designing and Implementing an Effective Storm
Water Management Program, Denver, 1998.
                                                     386

-------
    Permittee two ("Midtown") is a larger and more self-contained community with a population of 50,000 located within
an urbanizing county whose total population  makes it a designated "urbanized area." It is aggressively annexing growth
areas, and has a thriving economy. It has a City  Engineer/Public Works director, road maintenance staff, and other
municipal capabilities and resources. It also has a growing stormwater quantity program  and wishes to bring quality
together with quantity in a comprehensive and integrated approach. It wants to take advantage of its GIS database and
capability.

The Programs

    We can assume that contained within, or subsequent to obtaining the general permit, the permit writer will publish a
list of regionally appropriate BMPs  to be used  in permit applications. The  general  permit will have narrative effluent
limitations which  describe goals or  narrative standards for each of the minimum controls.  Each  permittee must then
develop basic program objectives and measurable standards (not included here) under the goals  provided  by USEPA for
each  of the six minimum controls.  These measurable standards can be stated in  terms  of actions taken or results
achieved. It is best to state  them in terms of things that can be controlled and which do not have uncontrollable and
unpredictable results.

    It is also smart to schedule the programs (the schedule is not demonstrated here)  in terms of phases,  pilot programs,
demonstration projects, trials, etc., with an evaluation process at some point in the permit. It should then be written into
the NOI that this program will be modified, expanded, curtailed  or even abandoned if it is not effective.

    Smallville sought to obey only the letter of the law, but did not see many ways to  proceed. It had no real stormwater
program, no  known water quality problems, and few current responsibilities. This community sought to take advantage
of "big brother" next door in joint programs or education,  and to adopt more regionally uniform development regulations
enforced locally. Smallville sought to fund any program needs through budget changes and through economies gained
by taking advantage of regional programs, free information, and expanding duties of existing staff.

    Midtown sought to meet the program minimums in a more proactive way focusing  on  perceived needs within the
community. They took advantage of the strength of existing local programs,  a strong economy,  a strong environmental
awareness, and outside assistance where available in the form of copied resources and shared efforts. Midtown  expanded
its current program using  EPA suggestions to build  a more comprehensive and meaningful program in several key areas6.
Because they did not have the ability to try to work regionally (the adjacent county had no resources for developing a
stormwater program, but would cooperate as necessary) it needed to build the program alone  and to  work  extra-territorially
as appropriate. Midtown looked at each program  to  insure the existence of: adequate  legal authority, competent  technical
approach, dedicated financial resources and appropriate administrative procedures and staffing.

    Because  program  funding became an issue,  Midtown  sought to establish  a stormwater user fee  system (often  called
a stormwater utility) to provide stable, adequate and equitable funds. The costs and steps of the utility development are
not included here.

Program  Objectives

    Table 1  develops the basic objectives of each of the programs in each of the six minimum areas. In real life these
objectives would be developed through a series of discussions with staff and, perhaps,  a citizen's group, and through early
coordination with  the permit writer.

    Table 2, which is attached  as an Appendix, gives basic cost-estimate information for the two programs.  The costs are
approximate  and would vary depending on how all costs  are accounted for, availability of staff, etc. The intent is to give
ballpark estimates and not to quibble over details. In these estimates all personnel time is costed at $50/hr regardless of
6 NAFSMA has taken an earlier version of the Midtown values, refined them, and developed a minimal and advanced program concept out of this
information. That information can be obtained from NAFSMA by calling 202-218-4122.

                                                      387

-------
the source of the  labor (in-house or contracted).  This  corresponds to a fully burdened salary  rate plus allocated overhead
costs for a mid-level technical person.

Table 1. Basic Program Objectives
                     Smallville
                                                                                                      Midtown
                                           Public Education and Outreach on Storm Water Impacts
    Acquire and mail existing public domain informational brochures
    Encourage  and facilitate newspaper articles
    Educate the few industrial and commercial stakeholders individually
                         1.  Acquire and mail  existing and specifically pertinent public domain
                            informational brochures to the general public
                         2.  Develop  a  stratified database of stakeholder groups  and develop
                            and execute targeted  education programs
                         3.  Develop and implement elementary school education programswith
                            preexisting  curriculum
                         4.  Develop and advertise complaint hotline as  a pollution hotline
                         5.  Develop press information and briefings with  the objective of having
                            a quarterly  news  article
                         6.  Develop and make available a slide show and speakers bureau
                                                      Public Involvement/Participation
    Develop and implement a citizens advisory group appointed by the mayor
   Encourage citizen participation in the  neighboring  city's  programs for
    used oil, household hazardous waste, adopt-a-stream, etc. through news
    articles in  local  neighborhood newspaper
                         1.  Develop  and implement a stratified and diverse citizens  advisory
                            group/task  force
                         2.  Develop a citizen monitoring and/or adopt-a-stream program - may
                            be partially federally funded
                         3.  Develop a student storm drain stenciling program and student dry
                            weather screening  program (see illicit connections program)
                         4.  Encourage  the  development of watershed groups for each major
                            watershed within the jurisdiction (see BMP control)
  Illicit Discharge  Detection and Elimination
    Develop a stormwater major outfall map on USGS  base map
    Modify slightly and adopt a generic ordinance available from the state or
    other organization.
                         1.  Develop a major stormwater system map and inventory on existing
                            GIS topo.  Base mapping
                         2. Cross-reference map with existing databases  on NPDES permit
                            holders (available from the state) and SARA  Title  III database to
                            identify likely source of dry weather pollution
                         3. Develop an illicit  connections  and illegal dumping  ordinance
                            including hotspot program
                         4.  Perform initial dry weather screening in several key parts of the city
                            by student volunteers
                         5. Develop inspection and enforcement capabilities and  resources,
                            and develop a  detection program using city staff and a database of
                            potential specific locations
                         6.  Advertise hotline and write news articles (see  public education)
                         7. Advertise  existing  private used  oil disposal sites  (see public
                            education)
                         8.  Educate all public employees to recognize and report problems (see
                            pollution prevention)
                         9. Develop automotive  industry sponsorship of  spill prevention,
                            materials  management, and  inspection  and education programs
                            (see public education for part  of this)
Construction Site Storm Water Runoff Control
1.  Modify the adjacent city's  sediment and  erosion  control  ordinance  to
    meet the regulatory minimums
2.  Modify  plans  review and  inspection  procedures  to include  program
    minimums
3.  Train  city secretary  to collect phone complaints and take  appropriate
    action on erosion complaints
4.   Advertise the  complaint line as part of the  public education program.
                        1.  Modify existing  sediment and erosion  control  ordinance to
                            include all the requirements of the  regulations
                        2.  Add a BMP section and clear design steps to  the drainage
                            manual
                        3.  Conduct training and familiarization  program for developers,
                            contractors and engineers, as well  as  in-house training for
                            inspectors
                        4.  Insure hotline has a formal and defined ability to receive and
                            properly process  erosion complaints
                        5.  Upgrade the  erosion  control  inspection and  enforcement
                            program
                                                                   388

-------
                         Post-Construction Storm Water Management in New Development and Redevelopment
1. Modify  and  adopt  the adjacent city's  stormwater  ordinance  regarding
   stormwater quantity and  quality  requirements to  require similar controls
   and  requirements.  Add a maintenance  requirement for BMPs and
   detention designs
2. Transform the inspection process to be  able to inspect and enforce the
   new ordinance
3. Communicate the  new requirements
                                1.   Investigate and seek to institute zoning and policy changes to
                                    encourage density restrictions, transferable  development rights,
                                    easier use of PUDs, limitation of impervious areas, conservation
                                    easements, mandatory floodplain dedication, etc.
                                2.   Develop design guidance  for the  use  of structural  and non-
                                    structural  BMPs
                                3.   Develop and conduct an ongoing training program in the proper
                                    use of BMPs
                                4.   Develop several  BMP  pilot projects to demonstrate  and gain
                                    experience in BMP use
                                5.   Overhaul and develop a comprehensive storm water  ordinance
                                    for both water quantity and quality which includes mandatory use
                                    of BMPs and a maintenance requirement
                                6.   Establish  inspection program for private BMPs
                                7.   Develop a monitoring  program  for  local surface waters and to
                                    monitor their long term changes
                                8.   Develop master plans for areas facing  new development and
                                    establish  and enact  policy for regional  BMP design and
                                    maintenance
                                9.   Develop ways to  improve  extra-territorial planning and zoning
                                    input
                                10.  Identify key environmentally sensitive areas and take steps to
                                    protect such areas through ordinance, overlay districts, etc.
                                11.  Seek to  establish  local watershed organizations and
                                    neighborhood adopt-a-stream programs to assist in compliance
                                    and build  public support
Pollution Prevention/Good Housekeeping for Municipal Operations
1. Review all current municipal  procedures and document ways to reduce
   pollution
2. Make changes and document
3. Obtain and distribute materials on ways to  reduce pollution as available
   and appropriate.
                                    Conduct  an outside  review of all  applicable procedures and
                                    criteria and make  recommendations  for change,  implement
                                    changes
                                    Obtain  available  information  and conduct sensitivity and
                                    familiarization training for all applicable city employees
                                    Seek to control  floatables partially  through  adopt-a-stream
                                    program  (see public participation)
                                    Review existing flood control  projects to insure advantage is
                                    taken of pollution reduction opportunities in design and operation
   Hours are given in  most cases. Italicized numbers are one-time costs that are experienced some time in the first permit
period, assumed to fill the year in which they initiate. For ongoing programs,  the  program initiates beginning in the next
year. The annual costs are the anticipated costs thereafter. I have assumed that  all programs initiate in year one for the
total five-year cost estimate. Obviously if a  program initiates in  a later year there will be  savings in annual  costs not
incurred until the program initiates. The five-year total is four times the annual cost plus the initial cost — making a total
of five years. Some programs are five-year programs only, ending after the first  cycle.

   A schedule of tasks and of manpower requirements is not developed in this paper. The costs are given as initial costs
and  as ongoing costs (clear from the context of the table). Because not all program elements will be developed  and in-
place for the whole permit term, there will be a ramp-up process. Also,  most of the program  elements will continue to
change and  evolve  over time, and  program costs will  also change (up  or  down)  in  subsequent permit periods.
Extraordinary volunteer efforts have not been assumed (e.g. writing news articles, manning a  hotline,  etc.).

   It is important to realize that some  per capita costs go down for large cities because  they have a large fixed component.
For  example, it may cost the same to develop a one-page brochure whether  the  city  has 20,000 or 200,000 people in it.
Expenses are based on medium levels of effort wherever appropriate. Detailed expenses (e.g. long distance phone costs)
have not been  estimated.
                                                           389

-------
   Measurable goals have also not been provided in this  handout. But for each BMP measure or program it will  be
necessary to develop some measurable standard by which to judge success. The standard may be based on internal
activities where it cannot easily be based on external results.  For example, sending out brochures three times per year
can be measured. But, the effectiveness of those brochures can only be measured through phone surveys of public
knowledge before and after the brochure was sent, or based  on statistics on increased public participation in whatever
program the brochure was about. Neither measure is easy and reliable.  And, should  a certain percent "effectiveness
increase" be stated as the measurable goal, if it is not achieved the city would, technically, be out of compliance. Better
to make  the goal controllable, especially in the first permit cycles when little is known on the effectiveness of certain
(especially non-structural) BMP measures.

   In  no case have the costs of structural BMPs been estimated or included. Cost estimates are available in several
references including the Center for Watershed Protection (1997) and Northern Virginia  Planning District Commission
(1994). The economic benefits of structural BMPs are discussed in EPA (1995).

   Monitoring costs are developed for Midtown based on both  receiving stream monitoring and some pilot BMP program
monitoring; they are non-existent for Smallville. EPA estimates that about 50% of permittees may incur monitoring costs
in subsequent permit cycles.  It is also assumed that there are no TMDL or other types  of watershed  assessment actions
going on in the watershed which may radically modify the permit conditions, and that there are no regional or state-wide
programs which could simply be adopted by reference for portions of the NPDES minimum requirements.

Summary  Results

   The summary results of the analysis are presented in Table 3, in terms of cost per capita, for each of the programs in
a manner comparable to the EPA estimates.

   The range of results is similar  to that experienced by EPA in making its original estimates of the cost of the Phase II
program.  The details of this program development can assist a local  community  in fashioning its own stormwater program
in response to the regulations.
Table 3. Summary Results
Minimum Control
                                                                             Annual Per-Capita Cost
                                                                  Small
                                                                                                     Midtown

1
2
3
4
5
6


- Public Ed.
- Public Inv.
- Illicit Connections
- Construction
- Post Const.
- Housekeeping
Totals
First 5-year Permit Period
0.39
0.21
0.24
0.20
0.14
0.15
1.33

1.24
0.62
1.77
0.96
5.78
0.59
10.96
                                          Subsequent 5-year Permit Periods
1 - Public Ed.
2 - Public Inv.
3 - Illicit Connections
4 - Construction
5 - Post Const
6 - Housekeeoina
 Totals	
0.36
0.24
0.10
0.18
0.13
0.10
1.40
0.51
1.16
1.10
1.26
0.20
5.63
                                                     390

-------
The Phase II Action Plan

    Given the great range in costs for the Phase II program it makes sense to get a jump start on planning for it. Many of
the requirements or potential inter-local arrangements that could be developed take  time to  implement,  more time than is
available if the community waits until the general permit  has been finalized and the NOI is due. There are steps that a local
government should take now to prepare itself for the regulations and to position itself to meet compliance in the most cost-
effective manner. These steps can be performed as part of a Phase II action plan:

 1. Assess your status

    Ask yourself if you are "in," "potentially in," or "out." Find out who else is in your category.

2.  Get to know the permit writers

    Find out what the permit writers are thinking about the permits, what the general permit will look like, when you will
know more, how they will evaluate those potentially in, what other actions are going on in  the state that may impact the
permit, etc. Find out their ideas about what is important in the permit, what their special interests are, do they strongly
support the permit, etc. Plan to establish an ongoing dialog.

3.  Assess your surface waters

    Find out if there are any ongoing actions which might designate surface waters in your jurisdiction as not meeting water
quality  standards. See if there are any planned watershed assessments or TMDL  requirements coming in the future.

4.  Assess your  own program

    How much of your own  stormwater program looks  like the regulations, even with  some  minor modifications. Can you
get a jump on the requirements through transformation of your current programs?

5.  Check out your neighbors

    Are there some other programs nearby that might result in savings to you? Can  you simply be covered under another
program?  Can parts of the requirements be waived because they are already being done by someone else? Can you
plan to be part of a  regional permit?  Can you split the permit requirements with  an adjacent  entity and perform them
together at savings to both of you?

6.  Get a team together

    Once  you have answered some of these questions, it is time to pull the action team together. This may include only
your own staff,  a multi-disciplinary staff within your own jurisdiction, or a multi-jurisdictional or regional team.  Get together
to brainstorm and come up with a proposal to the permit writer which has mutual benefits. Remember, permit writers are
being encouraged to think regionally and on a watershed basis.

7.  Develop an  action plan

    Once  you have a team,  it is time to have a plan.  Begin to formulate what you will need to do to apply for the permit
and to  carry it out. What might your program minimums look like? Are there some things you can  do  now, over several
years, that you cannot afford  to do in any one year, or that will take too long to get going if you wait until the permit is upon
you? Can  you begin  the  program transformation process now? What about data collection and mapping? Are there other
uses for any data you will collect which will create synergy?
                                                    391

-------
8. Get started

    Some things are best started early. But do not jump the gun by committing resources in areas that are not yet
anticipated to be firm. Ask the permit writer for his or her opinion.

References

Center for Watershed Protection, 1997, The Economics of Stormwater BMPs in the Mid-Atlantic Region, August.

The National Association of Flood and Stormwater Management Agencies (NAFSMA), 1999a, Phase II Survey Raw Data
Report, 1299 Pennsylvania Ave. NW,  Washington DC, 20004.

The National  Association of Flood and Stormwater Management Agencies (NAFSMA), 1999b, Survey of Stormwater Phase
II Communities, 1299 Pennsylvania Ave. NW, Washington DC, 20004.

Northern  Virginia  Planning District Commission,  1994, Urban Retrofit  Techniques: Applicability, Costs,  and Cost-
effectiveness, November.

United States  Environmental Protection Agency,  1995, Economic Benefits of Runoff Controls, EPA 841-S-95-002,
September.

United States Environmental Protection Agency, 1999, Report to Congress  on  the Phase II Stormwater Regulations, EPA
833-R99-001, October.
                                                   392

-------
                                                                           Appendix
Table 2. Hypothetical Program Detail and Cost Summary

                                    Smallville
        Program Element

Public Education and Outreach on Stonnwater Impacts
                                                     Midtown
cost
            Program Element
                                                                                  cost
Acquisition of available mailers and information from private
institutions and other governmental entities - 20 hrs
Keep up with available literature - 20 hrs/yr
$1 ,000
$1 000/yr
Acquisition of available mailers and information from private
institutions and other governmental entities - 40 hrs
Keep up with available literature - 50 hrs/yr
$2000
$2500/yr
Coordination with neighborhood or shoppers newspaper to run
articles on pollution sources - 4 hrs
Develop 2 articles per year - 24 hrs/yr
Coordination with the few individual potential sources of pollution
about the program and their needs - 10 hrs
Series of three mailings - stuffers in utility bill
One mailing per year afterward
Responding to information requests - 1/2 hr/wk




$200 Stratified mailing database development for key stakeholder
groups - commercial, automotive, minority, etc. - 100 hrs
$1 ,200/yr Maintenance of database - 1 hr/wk
$500 Obtaining or developing educational materials for the specific
outreach and stakeholders' programs, printing - 30 hr
Updating materials - 1 00 hrs/yr. Mailing 5,00 brochures per year
$3,600 Developing outreach and educational programs - 200 hrs
$1,050/yr Executing programs - updating, mailing, training, presentations -
200 hrs/yr
$1,300/yr Develop elementary and middle school education programs - preexisting
Material/curriculum - free materials - 100 hrs
Ongoing program maintenance - refresher training, 5 schools - 100 hrs/yr
Advertising of hotline - radio spots developed in-house and on public
And other radio service spots and Newspaper ad, 3 times per year -
1 40 hrs - donated spots
Develop white paper and press package - initial, brief - 32 hrs
Develop quarterly press package/briefing - brief press - 24 hrs per +
expenses
Development of a short, scripted stormwater pollution slide show,
Presentation and speakers bureau & initial presentation - 60 hrs
Give presentations - 48 hrs/yr + expenses
General informational brochure development and mailing - once/year
- 60 hrs/yr - 25,000 inserts @ 0.50 per
$5,000
$2,600/yr
$4,000
$7,500/yr
$10,000
$12,000/yr
$5,000
$5,000/yr
$12,000/yr
$1 ,600
$5,000/yr
$3,000
$2,600/yr
$15,500/yr
Continued
                                                                               393

-------
Table 2.  Cont.
                                      Smallville
                                                                                                                              Midtown
Program Element

Total


Total Cost
Cost (ongoing 5
cost

Initial Cost
Annual Cost
(first 5 years)
-year period)


$5,350
$4,550
$23,500
$22,750
Program Element
Responding to information requests - 2 hrs/wk
Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5-year periods)
cost
$5,200/yr
$30,600
$69,900
$310,200
$349,500
Public Involvement/Participation
Development and implementation of a citizen advisory committee
appointed by the mayor - 2 initial meetings - 14 hrs
Quarterly meetings - 32 hrs/yr


Advertisement of the larger city's stream cleanup program in local
shopper newspapers - news articles, and coordination with them in
all such programs - 1 6 hrs/yr



$700
$1,600/yr
$800/yr

NOTE: italics are initial cost - for first year only
Total
Total Cost
Cost (ongoing 5
Initial Cost
Annual Cost
(first 5 years)
-year period)
$700
$2,400
$10,300
$12,000
Development and implementation of a citizen advisory committee
appointed by the council - 5 initial meetings - 70 hrs + expenses
Bimonthly meetings - 60 hrs/yr
Initial coordination of monitoring program and/or adopt-a-stream
- 60 hrs - equipment purchase
Ongoing^ coordination and equipment, database maintenance
- 100 hrs/yr + expenses
Student storm drain stenciling program development and
implementation - 80 hrs
Annual cost
Watershed group encouragement - presentations, advertising -
50 hrs + expenses
Ongoing coordination, education - 4 groups - 20 hrs. per
Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5-year periods)
$3,700
$3,000/yr
$40,000
$15,500/yr
$6,500
$3,000
$2,800
$4,000/yr
$53,000
$25,500
$155,000
$127,500
Illicit Discharge Detection and Elimination
Collect and plot field information on
5 hrs - contract
system locations
and sizes -
$7,000
Develop system map, perform inventory of major structures - 60
hrs + contract
$150,000
                                                                                         Update map - 60 hrs
$3,000/yr
Adopt ordinance - 20 hrs
Enforcement of ordinance - 20 hrs/yr
$1 ,000
$1 ,000/yr
Database development and GIS programming and mapping -
200 hrs + expenses of $3k
Database maintenance - 100 hrs
$13,000
$5,000/yr
                                                                                                                                                              Continued
                                                                                   394

-------
Table 2.  Cont
                                      Smallville
                                                                                                                            Midtown
Program Element





Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5 -year period)
cost





$8000
$1 ,000
$12,000
$5,000
Program Element
Ordinance development with public participation - contract
Initial dry-weather screen in parts of city - student volunteers - 240 hrs
One staff member 1 day/week for inspection and enforcement of
Illicit connection program - + expenses
Development of automotive or other specialty programs - 1 00 hrs +
1 k exp.
Annual implementation of inspection and education - 1 day/wk
Initial Cost
Annual Cost
Total Cost (first 5 years) $433,000
Total Cost (ongoing 5-year periods)
cost
$20,000
$12,000
$28,000/yr
$6,000
$22,000/yr
$201,000
$58,000
$290,000
Construction Site Stormwater Runoff Control
Modify and pass new erosion control ordinance - 40 hrs
Enforcement ordinance in inspection process - 50 hrs/yr
Modify development procedures - 4 hrs
Train secretary to handle calls - 8 hrs
Handle erosion calls - 1 0 hrs/yr


Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5 -year period)
Post-Construction
Modify and get ordinance passed • 40 hrs
Enforce/explain new ordinance provisions - 1/2 hr/wk


$2,000
$2,500/yr
$200
$800
$500/yr


$2,600
$3,000
$14,600
$15,000
Stormwater Management in
$2,000
$1 ,300/yr


Modify existing ordinance - public participation - 60 hrs
Add BMP section to design manual - 140 hrs + printing cost
Conduct training sessions for staff and local development related
persons - 80 hrs
Ongoing biannual training - 32 hrs/yr
Develop hotline procedure for complaints reception - 10 hrs
Hotline @ 150 hrs/yr + expenses
Upgrade erosion control program for more sites and more activities -
one person two days/wk + expenses
Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5-year periods) $275,500
New Development and Redevelopment
Work on major policy changes in land use regulations - contract +
200 hours
Develop design guidance for BMPs - contract
Training program for BMP use - debvelopment - 24 hrs + contract
Annual training - 60 hrs/yr
$3,000
$12,000
$4,000
$1 ,600/yr
$500
$8,500/yr
$45,000/yr
$19,500
$55,100
$239,900

$100,000
$25,000
$3,000
$3,000/yr
                                                                                                                                                            Continued
                                                                                  395

-------
Table 2.  Cont.
                                           Smallville
                                                                                                                               Midtown










Program Element







Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5 -year period)
Pollution
Review of all current procedures - modification of procedures - 40 hrs
Obtain
Annual


T
0
T
A
L
T
0
T
A
L
and distribute educational materials - 10 hrs
cost of changed procedures - SWAG


Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5 -year period)
Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5 -year period)
cost







$2,000
$1 ,300
$7,200
Program Element
BMP Pilot projects - federal funding assistance - 5-year program •
contract 5 yrs
Comprehensive stormwater ordinance with public participation -
contract
BMP inspection and enforcement program - one person one day/wk
+ expenses
Data collection program - SWAG
Master planning for new areas for both quality and quantity - 2 mile
Planning zone around -cSyeipr program -40 mi2
Costs of administration of regional BMP program - SWAG
Sensitive area identification program, ordinances and policy enactment-
5-year program - 100 hrs incl. Mapping 5 yrs
Initial Cost
Annual Cost
Total Cost (first 5 years)
Total Cost (ongoing 5-year periods)
Master planning
cost
$200,000
$40,000
$25,800/yr
$30,000/yr
$800,00
5 yrs
$4,000/yr
$25,000
$393,000
$62,800
$644,200
$314,000
Prevention/Good Housekeeping fo Municipal Operations
$2,000
$500
$1,000/yr


$2,500
$1 ,000
$6,500
$5,000
$21 ,500
$13,250
$74,150
$66,250
Review and modification of all applicable procedures and criteria
contract
Site inspections and corrections - 5-year program - $5k/yr
Training for city employees on new procedures - 40 hrs + 10 hrs @
75 persons + expenses
Review flood control projects for retrofit opportunities - contract
Annual cost of changed procedures - SWAG
Initial Cost (without master planning)
Annual Cost
Total Cost (first 5 years without master planning)
Total Cost (ongoing 5-year periods)
Master planning
Initial Cost (without master planning)
Annual Cost
Total Cost (first 5 years without master planning)
Total Cost (ongoing 5-year periods)
Master planning
$25,000
$25,000
5 yrs
$42,000
$15,000
$10,000/yr
$107,000
$10,000
$147,000
$50,000
$800,000
$804,100
$281 ,300
$1 ,929,300
$1 ,406,500
$800,000
                                                                                   396

-------
                   The Stormwater Utility Concept in the  Next  Decade
                                       (Forget the Millenium)
                                           Hector J. Cyre, President
                                       Water Resource Associates, inc.
                                                Kirkland,  WA
Abstract

    In the mid-1 970's, the first stormwater utilities were viewed as novel innovations in a few western states. Today, just
25 years later,  more than four hundred cities,  counties, and special districts throughout  the United States  have established
such utilities. The pace  is accelerating, and the stormwater utility concept has moved from a novelty to a well-accepted
management and funding approach. What will we see in the next decade?

    The stormwater utility has been adapted to fit diverse stormwater management problems and needs across the United
States. Program content, priorities,  institutional and organizational structures,  and rate  methodologies have  been tailored
to fit local needs and municipal authority and practices that vary widely.  Courts in several states, and  even federal courts,
have been engaged in resolving key issues, including but not limited to the legality of utility service  fees and the use of
other funding mechanisms.

    Major changes in the concept are still emerging today. Stormwater quality has  become a concern equal to flood
control in many communities.  The National Pollutant Discharge  Elimination System  Phase II stormwater  permits have
spurred a new  round of interest in the stormwater utility concept among smaller communities. This is creating a demand
for basic utility  concepts  suitable for small cities and towns,  which will need to  be  less costly  and simpler to implement and
maintain. Concurrently, more large  cities,  urban and  urbanizing counties,  regional service agencies  such as metropolitan
sewer districts, and consolidated governments are investigating the utility approach. They  will require more complex
institutional and funding solutions.

    Stormwater management itself is also changing  rapidly. Interest and involvement in stormwater management have
broadened. As combined sewer overflow programs,  total  maximum daily load (TMDL) negotiations, stormwater quality
mandates, coastal zone management measures,  and  safe  drinking water supply issues converge,  more wastewater and
even water  supply  utilities  are engaging in  stormwater management.  Regional  resource management programs,
watershed-based  master planning,  multi-purpose cooperative efforts involving  urban forestry  and  riparian  corridor
protection, and use of state revolving loan funds for stormwater quality projects are becoming more common.

    Local  programs are  quickly evolving as well. They have become more comprehensive in scope, more  costly,  and more
demanding of technical and  administrative skills while  the  pool  of resources  has grown  relatively  slowly. Local
governments are accepting responsibility for more components of the stormwater drainage systems or, in some cases,
being forced to take on such responsibilities. Open streams, historic remnants or agricultural ditches  and levees, and
detention facilities are being included among the system components actively improved,  operated, and maintained by local
stormwater management agencies, A preventive orientation that minimizes  problems is  replacing  reactive measures.
Technology, such as geographical information systems and hydrologic and hydraulic  modeling, is more widely available
and more  productive in support of stormwater management, even in smaller communities. Public involvement in  decisions,
policies, and even the operation of systems is increasing.

    This paper examines these  and other emerging trends that characterize where  stormwater utilities are heading in the
next decade.
                                                    397

-------
Pressures Moving us from  Draining the Swamp to Stormwater Management

    Historically,  local drainage flooding, erosion, and water pollution due to stormwater runoff have not been high  priorities
for  municipal governments. Unless homes,  businesses, valuable agricultural land, or public properties  have been
devastated by flooding or other "drainage" problems, competing priorities have generally garnered more public  concern
and thus more support from elected  officials. As a  result, stormwater management operations, regulatory measures,  and
capital investment were historically ignored or, at best,  received inadequate attention and erratic funding.  Stormwater
management has been a  "stepchild" among municipal programs.

    Symptoms of this past disregard are evident in many cities and counties.

    .  Improvements to stormwater systems in many communities have been limited to site-specific facilities installed by
      subdivision and commercial developers.
    . Design practices have traditionally emphasized collecting and discharging runoff from each property as quickly as
      possible, without regard for downstream consequences.
    . Public maintenance of stormwater systems has typically been reactive, and usually limited to road rights-of-way
      where uncontrolled stormwater might impact traffic  safety,  degrade the integrity of road surfaces,  or threaten
      valuable adjacent properties.
    .  Maintenance of stormwater systems located outside  of road  corridors has commonly been left to private  property
      owners, who are rarely capable of or willing to properly improve, clean, and repair such facilities.
    .  Municipal governments have usually improved and maintained individual structures or reaches instead of entire
      drainage systems, creating a patchwork of pieces  having widely varying  capacity  and reliability.
    •  Failures of substandard components  frequently impair the performance of  otherwise adequate parts of the  systems
      and damage properties near them.

    As described by one  municipal public works official, this stepchild is also the "sleeping giant" of unmet municipal
infrastructure needs. Long-term stormwater remedial repair costs potentially exceed street and bridge repair needs in
many  older cities. Learning the high cost of correcting stormwater management  deficiencies through master planning  may
have frightened as many  local jurisdictions into inaction  as it has spurred others. Perhaps the classic example is the
stormwater master plan for Key West, Florida,  which  (in the early 1990's)  identified $78 million in capital needs for  that four
square mile island community of less than 30,000  people.

    Several factors are now changing local governments' traditional orientation to stormwater management.

    .  Citizens' service expectations are higher than in  the past.   In many cities and  counties the  number of citizen
      complaints about stormwater problems exceeds those about potholes in  roads.
    .  Crumbling inlets and silt clogged ditches along roadsides spawn complaints even  though they are on  public property.

    .  Individual citizens or neighborhood associations  no longer  tolerate minor problems like localized flooding  and
      channel erosion in backyards.

    Environmental  awareness in general  is  greater than in the past,  and much more attention is being focused on
stormwater impacts on receiving water quality in recent years.

    .  Stormwater management is now recognized as being part of an effective water resource  protection strategy.
    .  Local concerns about acute threats of water pollution from spills and surreptitious dumping of toxic materials into
      stormwater systems are becoming more common.
    .  Phase II  of the National  Pollutant  Discharge Elimination System (NPDES) stormwater permitting  program is
      extending  the program to  smaller communities and those  larger urban cities that escaped Phase  I due to combined
      sewer service area  exemptions.
    .  Programs  proposed  by local governments in NPDES Phase II permit applications will cost many thousands of dollars
      per year in cities, towns, and  urban counties.


                                                    398

-------
    An encompassing, umbrella perspective of water resource management is emerging.

    . Solutions to combined sewer overflow (CSO) problems will have to balance optimization of wastewater transport
      and treatment facilities against stormwater quantity and quality concerns.
    • Several coastal states have instituted restrictive limitations on stormwater runoff to protect fragile estuaries and
      offshore waters from stormwater impacts.
    . Drinking water supply watershed protection measures have imposed stormwater runoff regulations on developers
      independent of local stormwater management control practices.
    . The point is becoming clear. Drinking water is water. Wastewater is water. Stormwater is water. Ground water
      is water. It is all WATER!

    In the face of  these pressures, the inadequacies of traditional stormwater management practices and funding are more
widely recognized. More comprehensive and cohesive programs that address both stormwater quantity and quality are
emerging. Clearly, however, the diversity of our  communities and their problems and priorities means that  no single
solution is appropriate for every county, city, town, and village. Nor can a single funding method or stormwater utility rate
structure fit every  situation. Stormwater service fee methodologies can  be designed to meet the specific needs of each
community and provide equitable, adequate, and stable funding. The  key is to tailor the funding  to a clear program
strategy.

"Stormwater Utility" can have Many Meanings

    The fact that the simple term "stormwater utility" obscures the various meanings it may encompass,  results in many
misunderstandings. The term may imply  a funding and accounting method, an organizational approach,  a management
concept, or a combination of all these. In reality a "utility"  provides an umbrella under which the financial, organizational,
and management  approaches  of each local stormwater  program  can be orchestrated to achieve practical and efficient
solutions.  Responsibilities may be  consolidated  and  focused.  Substantial  new funding may be generated. New
technology,  different management concepts,  and  upgraded support  systems may be adopted. A comprehensive,
preventive program may be instituted.

Changes  in the  Approaching  Decade

    The spectrum  of  the stormwater utility concept will broaden more in the next 10 years than  it has in the 25 years since
the first  utilities were established. The definition of "conventional" will change. Smaller towns and even villages will need
to employ simpler variations of the concept. Larger cities, urban counties, consolidated governments, and coordinated
regional approaches will demand  more complex institutional, organizational,  and funding solutions.  The following are a
few of the changes that may occur.

NPDES Phase II Will Impact the Stormwater Utility Concept

    The findings  of  a  survey   of Phase II cities conducted by  the National  Association  of Stormwater and  Flood
Management Agencies (NAFSMA) and published  in July, 1999 indicate that 17% of all the respondent communities did
not know how they would obtain funding to  meet the stormwater regulations. Nearly half indicated  they were not currently
spending money on any of the  stormwater program elements  mandated  by the regulations.  Nearly three-quarters  did not
have a public information or education program as the regulations mandate. The 54% of respondents that currently fund
programs or  activities that fit the Phase II  regulations on average spend upwards of $4,000 per  square  mile, or about $2.79
per capita. The implication  seems clear that the NPDES Phase II program poses  demands on local governments that may
cause them to look to the stormwater utility concept to meet their stormwater quality program funding requirements.

    The typical community that found the stormwater utility concept attractive in the past was a mid-size to larger city
undergoing rapid development. Analyses of their stormwater management needs and programs typically revealed initial
costs of service ranging from $25,000/sq. mile to $50,000/sq. mile annually. Costs per capita were typically $10 to $30
annually, with smaller cities trending toward the higher end of the range.  In this context, the NAFSMA survey data is not

                                                   399

-------
alarming. It suggests that the Phase II program costs will likely be in the neighborhood of five (5) to fifteen (15) percent
of the typical cost of stormwater management at the outset of utility-based programs.

    Given its cost implications, NPDES Phase II makes the stormwater utility concept attractive to a broader variety of
cities, counties, towns, and villages. Many communities that do not suffer flooding or other drainage problems will find the
revenue potential and flexibility of a utility  service fee attractive in the face of NPDES permit requirements. This will result
in utility approaches that are outside the current spectrum of our experience. The needs in individual small communities
may be  less diverse  than in large cities and  urban counties, but the  range will  be  cumulatively broader among the
communities involved in NPDES Phase II than in those that have implemented utilities previously.

    New institutional arrangements  and relationships will  have to be  devised. The  "utility" concept will take on new forms.
Use of interlocal agreements among several local jurisdictions  will  increase, with  responsibilities  in some cases
concentrated in one entity capable of providing the range of services required or, conversely, allocated among several
participants.

    The  limits of existing authorizing legislation in  some states will be tested. Many  states will need  to adopt new
legislation and amendments,  giving  local  governments greater flexibility in dealing with their water resource management
responsibilities. Courts in the various states, and perhaps even federal courts, will be challenged to arrive at some sense
of continuity  among the  institutional and financial solutions characterized as "stormwater utilities." Whether the court
decisions will enable rather than hinder local governments' efforts to comply with NPDES mandates is a key question.

    Organizationally  and financially  independent stormwater utilities have been common  to date.  In the  next decade more
stormwater utilities will be integrated with other water resource programs: organizationally, through formalized working
relationships,  or through financial arrangements. Other resource management  agencies  and programs,  including but not
limited to health departments and growth management  authorities, will demand a seat at the stormwater table.

    As NPDES permitting is applied to smaller urban  areas of less than 100,000 people, more of the regional wastewater
and water utilities already serving those communities  will assume stormwater management responsibilities.  In some cases
their involvement will be limited to water quality aspects. In others they  will address both quantity and quality. Funding
of stormwater management costs will simply be assumed by some of these existing utility agencies without changes in
their rate methodologies. Others will establish independent  stormwater cost centers  and  rate components  to track
spending and  allocate costs.   Some will even modify  existing wastewater  and/or water rate methodologies to better reflect
the impact of stormwater control on costs of service.

    As more  small  cities  and  counties seek to establish  utilities, stormwaterfunding strategies and rate methodologies will
need to minimize implementation costs,  yet  be  more  flexible to accommodate  stormwater quality management costs  and
unique local needs. The urge to use a "cookbook" solution will cause some to adopt approaches that are poorly suited
to their circumstances. The desire for more precision in service fee rate algorithms will lead to  methodologies that give
an illusion of greater refinement without actually achieving it.

    The mandated involvement of smaller  jurisdictions  and more rural communities in stormwater management will spawn
"paper utilities" established solely to generate added revenue.  Most of these will be initiated without the foundation of a
solid program strategy. Accountability will  become a key issue in some of these communities within a few years.  Political
challenges based on accountability issues will cause some of these storm water utilities to be melded into other local
agencies'or programs or even dissolved entirely before  they have geared up to address their program priorities.

    Despite NPDES storm water permit  mandates,  locally  perceived needs  will still predominate in setting priorities.
Flooding will remain  a  more important local issue than storm water quality.  NPDES  mandates will  influence actual spending
priorities only slightly. Few communities will  need to institute a utility service fee just to support their NPDES Phase II
programs, but many will justify it (at least partially) on that basis because it is easy to blame unfunded federal mandates
for new local taxes,  assessments, and service fees.
                                                     400

-------
    The technological resources and expertise required will change from the traditional engineering emphasis to a multi-
disciplinary mix. More natural science and social science skills will be needed. Operational practices will change as new
technology and information management systems enable innovative approaches and result in greater efficiency. Greater
use will be made of outsourcing because of limited personnel resources and the high cost of specialized equipment.
 The Stormwater Utility Concept will Impact NPDES Phase II
    Local  approaches to stormwater management will influence the content of Phase II permits and attainment of NPDES
objectives. Stormwater utilities offer both financial capability and flexibility. Except in rare instances, stormwater utilities
will not be established strictly to address stormwater quality and NPDES permit requirements. Rather, they will have a
broader stormwater management perspective.  For many communities this will mean that water quality management
activities to comply with their NPDES permit will be tacked onto other stormwater efforts. NPDES Phase II permitting will,
within a few years,  adjust to accommodate this  reality in terms  of permit mandates,  technical and scientific standards, and
reporting requirements.

    Related  issues  ranging  from combined sewer overflow strategies to drinking water protection will  be melded with Phase
II permit requirements because they have to be. Local stormwater quality management cannot independently meet the
entire range of regulatory expectations operating strictly by reference to NPDES Phase II. Conflicts and primacy battles
will identify inconsistencies and  gaps between the issues and programs, and  will ultimately filter down to changes in
NPDES Phase II program priorities and the permit requirements imposed on local  governments. The unknown is whether
this result  in responsibility shifting toward bigger agencies with more resources and  a broader perspective or toward  local
entities that have the ability to identify and activate locally acceptable solutions.

    Watershed-based regulatory programs will overtake jurisdictional-based regulatory programs  like NPDES Phase II.
The utility  approach will broaden to encompass watersheds through agreements among counties and cities simply because
utility funding has the proven capacity to generate sufficient funding in politically acceptable ways. The transition has
already begun in some areas. Where TMDLs affecting discharges of all sorts into receiving waters  are an issue, they will
supercede the six minimum practices identified for NPDES Phase II, making them essentially meaningless. Scientifically
based, public  health driven measures to  protect drinking water supplies, estuaries, lakes, fisheries,  and recreational
beaches will overwhelm the programmatic approach represented by NPDES Phase II.

 You May Need a Program to Identify All  the Players

    Stormwater utilities were first established because no one wanted responsibility for stormwater management. Those
involved were concerned only about the impact of stormwater on their "real" jobs. The utility approach provided a way to
focus  responsibility  and  obtain dedicated, if not  always adequate, funding for stormwater  management. If there had  been
another option that was working, the stormwater utility concept probably would never have emerged.

    A key issue in the next decade will be whether stormwater utilities will be major protagonists or bit players among all
those  now crowding onto the stage.  More established and  better-funded  water and wastewater utilities now recognize that
stormwater influences their operations directly and, in some cases, dramatically. For example, TMDL-based wastewater
discharge limitations may severely curtail development in some areas. Will local wastewater utility administrators  (and
local elected officials) allow independent stormwater  management utilities to address stormwater quality when  economic
vitality is at risk?

    Other interests are becoming involved in stormwater management. Water supply  utilities face the requirements of
federal and state legislation regulating sources  of supply and treatment. Coastal zone management has recognized that
many priority uses of the shorelines and near-shore areas are dependent on good water quality.  Growth management
is  an  emerging concept,  and concurrency of infrastructure  improvements with development  approvals highlights the  issue
of deficient stormwater systems in many communities. Protection of endangered and threatened species, urban forestry,
and riparian corridor protection all have a relationship with stormwater management.
                                                    401

-------
 The Walls Will ComeTumbling Down (or at least they better)

    The proliferation of federal, state, and local water resource (and related) regulations in recent years has created an
environment in which  dispersed responsibility for water in various forms and for various purposes is rapidly becoming
unworkable. The  institutional barriers that have been created over the  past hundred  years or so to focus attention, energy,
and responsibility no longer fit the public needs. As watershed-scale studies, planning efforts, and the concept of TMDLs
clearly illustrate,  water resources are inextricably bound together regardless of their temporary form, use, and character.

    The next decade will see accelerating consolidation of water resource management responsibilities at the local level
of government. This is contrary to the control interests of some individuals and entities, and will not happen silently or
easily. Will cities, counties,  and special districts relinquish a little (or a lot) of their control over water resources through
interlocal agreements? Will they accept a regional entity for water supply, wastewatertreatment, stormwater management,
or even water quality? What will be the effect on stormwater utilities?

    What are  the organizational implications of the coming  changes in storm water management? Realistically, local
governments change slowly. Public Works  and Street Departments  have  historically been  the lead organizations of storm
water programs, but they rarely have had much involvement in water  quality issues. If storm  water quality begins to
influence  local priorities, it is more likely that water and wastewater  utilities will  assume storn  water management
responsibilities from  Public Works and Street Departments than the reverse. Public Works agencies will  have to  upgrade
their engineering  and scientific capability or risk losing their storm water management role to water and wastewater utilities
that are typically well-established, well-funded, and well-understood  by the public.

 The Ability to Innovate Will Exceed the Need

    Most of the early stormwater utilities programs were rather narrowly focused, and the  funding mechanisms supporting
them were relatively simple. In recent years, however, there has been  a shift toward more sophisticated and complex
approaches to all aspects of stormwater management-from master planning to rate methodology design. Much of the
credit goes to the explosive  growth in information processing capability associated with the computer revolution of the past
20 years. It is not clear, however, that  much  of the added capability to  innovate is  necessary to meet stormwater
management needs.  This is  not to suggest that opportunities to improve should be ignored simply  because  they are  based
on increasing capability to do so. The following examples demonstrate how the ability to innovate in stormwater through
technology can run amok, and suggest how it should be managed to the benefit of people and the environment.
    There is no substitute for understanding what is really important. One Southeastern United States city invested over
$1 million dollars assembling a highly detailed location  inventory of its stormwater systems on a relatively sophisticated
data processing  platform. Unfortunately, the need  for the inventory was not premised on a clear program strategy, nor
was  adequate funding available  or established  concurrently to  support  capital improvements or maintenance
enhancements that could be facilitated by the inventory. The local elected officials finally tired of the seemingly mindless
spending on the  inventory and refused to discuss program improvements. Today, nearly 10 years later, the inventory has
not been maintained and is  out of date, and few improvements have been made  in  the stormwater management program.

    What is technically possible does not always make  common sense, and what makes sense is not always technically
possible. A Northeastern city recognized that the stormwater component of its wastewater service fee rate methodology
(one that was based on water meter size and internalized within its water/wastewater rates) was not reasonably allocating
the cost of stormwater services and facilities across the community. Change to a  more rational  approach was desired,
so a thorough  assessment of the range of options was undertaken. A broadly representative advisory committee aided
in the selection process. A  relatively simple stormwater rate  concept was selected that segregates stormwater funding
from wastewater and water service.  It will allocate a portion of the cost of stormwater service on the basis of gross area
and a portion on the basis of impervious area.  Once the impact of the change  on certain rate payers was recognized,
however, the advisory committee decided that phasing in the new rate over three  years was a better idea than making the
change in one step.  While the technical support requirements  of  the phased approach are not especially  demanding, the
public information and  education challenge  is enormous.  Not only must the new rate methodology explained to the public;
it and the phase-in concept must be explained every year for three years.

                                                     402

-------
Errors by a  Few Will Make Life Miserable for Many

    As the number of stormwater utilities grows, there is a natural tendency among municipal managers to assume that
the process  and results have become standardized, and the experiences of another community can simply  be  transferred.
In an effort to save money, some cities  and counties  have established utilities without sufficient foundation and have even
adopted service fee ordinances without the benefit of a cost of service analysis or a rate study. Such misjudgments have
led to some  monumental errors that have the potential to erode if not destroy the viability of the utility concept in a region,
a state, or even nationally.

    One city recently established a stormwater utility and adopted rates based on internal analyses that did not define a
program, project the cost of service, or estimate the rate base available to generate revenue. As a result, the initial service
fee billing was  for nearly three times as much total revenue as the administration  had  indicated it hoped to  raise for
stormwater management.  Furthermore, sufficient public information and education had not been conducted prior to the
initial billing, so the public did not understand  the purpose of the billing. A lawsuit was filed, and a same judgement on
behalf of the plaintiff has resulted in the servie fees being rescinded and revenues returned with interest.

Expectations Will Advance Faster than Programs

    One common experience of the cities and counties that have established stormwater utilities is that public expectations
for the program have exceeded the  utility's ability to perform. This means that creating accurate expectations before a
utility is established must be a high priority.  One cause for unfulfilled expectations is that stormwater utility revenue
streams are  usually insufficient to address all the accumulated problems in  a  relatively short time.  Initial stormwater utility
service fees have typically been less than $3  month for single-family residences.

    Perhaps more significant,  however, is the fact that most stormwater utilities inherit programs and systems that are not
only deficient, but also do not offer an adequate foundation for a good, more comprehensive, program.  Utilities often  must
invest one to three years creating the  foundation for the program before real results begin to emerge in the  form of capital
improvements,  remedial repairs, upgraded maintenance, and more effective regulations.  Ratepayers tend to have little
patience, however, when they are writing checks regularly to a stormwater utility.

    In the context of NPDES Phase  II permits, public expectations  of improvements in water quality need to reflect the
complexity of water quality issues and the limited ability of local government to quickly alter conditions in receiving waters
through informational and regulatory programs. Attempting to sell a utility to a community as a response to federal water
quality mandates has been unsuccessful in several  communities. The  public recognizes  that stormwater quality, while
important, is still a minor part of the total cost of stormwater management.  Unless a  comprehensive quantity and quality
control program strategy is apparent, it is difficult to generate support for a stormwater utility.
                                                     403

-------
                     Are Green  Lots Worth More Than  Brown Lots?

   An Economic Incentive for Erosion Control on  Residential Developments'


                                       Martha Herzog and Jon Harbor
                                Department of Earth and Atmospheric Sciences
                                             Purdue University
                                       West Lafayette, Indiana 47937

                                              Keith McClintock
                                 Geauga Soil and Water Conservation District
                                            Burton, Ohio 44021

                                                John Law
          Indiana Department of Natural Resources, and Saint Joseph Soil and Water Conservation District
                                         South Bend.  Indiana 46614

                                               Kara Bennett
                                          Departments of Statistics
                                             Purdue University
                                       West Lafayette, Indiana 47937
Abstract

    Construction sites are major contributors to nonpoint source (NPS) pollution. However, a lack of personnel to enforce
erosion control regulations and limited voluntary compliance means that few developers apply effective erosion control.
New approaches are needed to increase erosion control on construction sites if this source of NPS pollution is to be
significantly reduced. We have tested whether an economic advantage exists for developers who use vegetative cover
for erosion control, independent of advantages gained in addressing environmental or regulatory concerns. Improving
residential lot appearance from muddy brown to green grass may increase the appeal of the lot to buyers. A market survey
shows that homebuyers and Realtors perceive vegetated lots to be worth  more than unvegetated lots, and this increased
value exceeds the  cost of seeding.  Thus, developers can now be encouraged to invest in vegetative cover because of the
potentially high return on the investment.

Introduction

Sediment Pollution and Construction  Sites

    Nonpoint source (NPS) pollution, produced from diffuse sources such as runoff from agricultural land, construction
sites, and urban surfaces, is now the leading cause of surface waterqualitydegradation  in the United States  (Novotny and
Chesters, 1989; Federal Register, 1990). In developing areas, construction sites  are a major source of NPS  pollution
because soil erosion rates are increased dramatically when land is exposed and disturbed by excavation and vehicular
movement (Harbor et al., 1995; Goudie, 1994; Goldman et al., 1986, Fennessey and Jarrett, 1994). In fact, some of the
greatest soil  erosion rates ever reported are associated with construction activities (Crawford and Lenat, 1989); erosion
rates on construction sites are typically 2-40,000 times greater than rates  under preconstruction conditions (Wolman and
Schick, 1967; Harbor, in press).  Sediment contributed to streams by construction  sites can  exceed that previously
1 This paper is reprinted from the Journal of Soil and Water Conservation, Spring Issue, 2000. We thank the Soil and Water Conservation Society for
their permission to reprint this article.

                                                   404

-------
deposited over many decades under pre-development land uses, radically altering stream geomorphology and ecology
(e.g.,  Wolman and Schick, 1967). The larger-than-normal  sediment deposition in waterways frequently exceeds the natural
capacity of the receiving water system to assimilate and equilibrate to the sediment influx (Paterson et al., 1993), causing
rapid  channel  changes and increased probability of flooding, erosion,  and sedimentation problems (Goldman et  al., 1986).

    In addition to sediment, construction sites generate other pollutants such as  pesticides,  nitrogen, and phosphorus from
fertilizers, petroleum products such  as oil  and gas from machinery, soil  stabilizers, construction  chemicals,  and washings
from concrete  or bituminous  mixing  and flushing operations (Koehn and Rispoli,  1982; Lemly, 1982). In  some cases these
pollutants are in particulate form or are adsorbed by soil  particles and are transported with the suspended sediment in
runoff from construction sites (Paterson et al.,  1993; Bhaduri et al., 1997).

    Although  construction sites generate a wide range of potential  pollutants, sediment overshadows all the other
construction site pollutants in total ecological and economic impact on receiving waters (Lemly, 1982). It was  estimated
that 15 million tons of sediment were released from urban  construction sites to surface waters in or near heavily populated
areas in 1975 (Lemly, 1982). The North Carolina Department of Natural  Resources and Community Development has
stated, that "sediment and its effects on stream environments" is the "most widespread water quality problem in North
Carolina" (Crawford and  Lenat, 1989). Because construction activities  predominantly occur near existing population
centers,  the waters that are  most seriously degraded are generally those that are most frequently used (Lemly, 1982).

Economic  Consequences of Sedimentation

    In addition to environmental impacts,  enhanced  delivery of sediment to off-site areas from construction activities has
significant economic effects (Table 1). These  economic impacts  result from lakes and streams becoming turbid  and filling
with silt,  destruction  of commercial aquatic  species, the need for additional treatment of turbid waterfor industrial  use, filling
of harbors and navigation channels,  loss of storage capacity of reservoirs, damage to  drainage ditches,  increased
frequency of flooding, loss of aesthetic value in the environment, and loss of game habitat (Lemly, 1982; Wolman and
Schick, 1967;  Koehn and Rispoli, 1982). The economic burden of  mitigating these environmental impacts is almost always
placed upon the taxpayer, rather than on the  operator of the construction site that is producing high sediment yields
(Harbor,  in press). By  not paying to  prevent the  off-site transport of sediment through the use of erosion  control measures,
the developer allows sediment from  the construction site to reach waterways where the economic and environmental costs
of any impacts are paid by downstream landowners and the community as a whole, and not the developer.

    Overall, annual  expenditures for in-stream and off-stream  impacts due to sedimentation  in the United States  exceeds
$11.6 billion (Table 1).  In-stream effects include  impacts while  sediment is in a waterway (stream, river, lake, or  reservoir).
Off-stream effects can occur  before  or after sediment reaches  a waterway,  either in floodwater or in water withdrawn  from
waterways to be used for industries, municipalities, or agriculture  (Clark, 1985; Clark et al.,  1985; Paterson et al., 1993).
Although  agricultural areas are  far more extensive than construction sites, the  mass of sediment  per unit volume  of runoff
from urban and construction areas  is 5 to 20 times greater than  that of runoff from agricultural lands (Fennessey and
Jarrett, 1994).  In addition, construction sites are usually located in  developing  or  developed areas,  where potential impacts
on infrastructure and other water uses are more severe than in rural areas. Estimates of agriculture's contribution to off-
site effects range from 1/3 to 2/3 of the total (Clark, 1985;  Clark et al., 1985; Colacicco  et al., 1989, Pimentel, et al. 1995).
Thus,  urban off-site environmental impacts  are probably on  the order of $3.9 to $7.8 billion  per year (1/3 to 2/3 of the  total
off-site effects), and are often borne by off-site landowners and communities. One of the main goals of erosion and
sediment control regulations is to avoid these  costs. The problem, however, is that developers have to pay  to reduce
erosion yet do not see any immediate return on this investment. Because there is little economic incentive for developers
to control erosion, regulatory and educational  approaches have  been developed to  improve construction site erosion
control, and requests to impose impact fees on developers have  increased (Trotti, 1997).
                                                      405

-------
 Table 1. Off-site damage costs from soil erosion by water in the United States

                                                                                              Cost
	Type of Damage	(millions in 1997 dollars*)
In-stream damage
   Recreational (fishing, boating, swimming)                                                          3,886.0
   Water storage facilities (dredging, excavation, construction of sediment
     pools)                                                                                   1,340.7
   Navigation (accidents, dredging)                                                                 1,088.1
   Other in-stream uses (commercial fisheries)                                                        1,748.7
     Subtotal in-stream                                                                          8,063.5

Off-stream effects
   Flood damages (sediment damage to urban and agricultural areas)                                      1,496.1
   Water conveyance facilities (sediment removal of drainage ditches and                                   388.6
     irrigation canals)
   Water treatment facilities                                                                       194.3
   Other-off stream uses (municipal and industrial,  steam electric power                                    1,554.4
     plants, irrigation)
     Subtotal off-stream                                                                         3,633.4

Total water erosion costs	11,696.9*	
(Data based on: Clark et al., 1985)
'Conversion using Consumer Price  Index from 1980-1997.
* Assuming that effects are the same today as in 1980.

 Vegetation and erosion  control

    The significant ecological and economic impacts of sedimentation provide strong motivation for erosion control. Soil
erosion involves the detachment of soil particles by raindrop impact,  wind-blown particle impact, wetting and drying cycles,
freezing and thawing, and runoff, and the  transport of detached soil particles by rain splash, wind and runoff

(Ekwue, 1990; Goldman et al.,  1986). Climate, topography, vegetative cover, and soil characteristics are the principal
factors that control soil  erosion  potential. Climate and soil characteristics cannot be readily controlled on  a  site, and
topography is constrained by pre-existing conditions and the grading plan,  leaving surface cover as the most easily
modified variable that controls oil  erosion on a site. Increasing vegetative cover on barren areas such as construction  sites
is  an excellent way to impede soil erosion and decrease sedimentation (Fig. 1).

    "Vegetative cover is the  most effective form of erosion control...a properly revegetated soil will be protected  from
erosion indefinitely without any need for human attention" (Goldman et al., 1986, p. 6.23). Vegetation (especially close
to  the ground surface) protects the surface  from raindrop impact and reduces the velocity of water flowing  over  the surface
by  increasing surface roughness  and  disrupting overland flow (Clark et al.,  1985;  Rogers and Schumm,  1991;  Satterlund,
1972). The reduction of water velocity flowing over the  surface and the breaking up of soil by plant roots increases the
amount of infiltration, thereby reducing the amount of surface water flow (Clark et al., 1985). Vegetation also depletes
subsurface water  between rainfall events, which reduces the amount of runoff during storm events. In fact, vegetative
stabilization on construction sites has been shown to reduce soil  loss by 80% (Harbor et al., 1995) to 99%  as compared
to  bare soil (Koehn and  Rispoli,  1982).  The cost  of reducing soil erosion using vegetative cover depends on the materials
used, but typical temporary seeding on a one-third acre residential lot in the Midwestern US costs  from $250 to $325.

Regulations requiring  construction site  erosion control

    In the US, the biological and physical impacts of off-site sedimentation have prompted  local, state, and federal
regulations requiring erosion  and sediment control for construction sites.  The National Pollution Discharge  Elimination
System (NPDES)  is a national program that issues, monitors and enforces permits for stormwater discharges associated
                                                        406

-------
                                 800
                                 600

                         Sediment
                          Yield
                           (kg)  400
                                 200
180
                                             JO        20        30       40
                                                 Vegetation Cover (%)
                   50
Figure 1. Sediment yields for different vegetative cover densities at 30, 60, 120, and 180 minutes of simulated rainfall on a 10% slope (Rogers and
Schumm, 1991).

with  industrial activity such as construction underthe Clean Water Act (Federal Register, 1990). State and local  regulators,
under the NPDES program, require erosion and sediment control for construction sites with 5 acres or more of land
disturbance (Federal Register,  1990). Because vegetative  cover greatly reduces soil erosion, many federal and state
regulations, such as Rule 5 in  Indiana and the Model Regulations for Urban Soil Sediment Pollution Control in Ohio,
encourage the use of surface cover as an important element of erosion control on construction sites.

    In Indiana, for example, state regulations  mandate that sediment should be contained on the construction site and not,
for example, allowed to run onto public or private roadways. Rule 5 requires that if vegetative practices such as seeding
and mulching are used, they must be implemented within seven days of the "last land-disturbing activity" at the site and
that these actions are the responsibility of the person in charge of the construction activity, which usually is the developer
(Indiana Department of Natural Resources,  1992). Similarly, in Ohio, Model Regulations for Urban Soil Sediment Pollution
Control (1980) require that the  responsible party for the development stabilize denuded areas  with permanent ortemporary
soil stabilization within seven days for any denuded  area that has reached its final grade or is to remain dormant for more
than 45 days. The permanent vegetation is not "considered established until ground cover is achieved which., .provides
adequate coverand is  mature enough to control soil erosion satisfactorily and to  survive adverse weatherconditions"(0hio
Department of Natural Resources, 1980).

    Enforcement of erosion control regulations varies significantly among states. For example, in Indiana, the Indiana
Department of Environmental  Management (IDEM)  controls permitting and enforcement, but  local soil  and water
conservation districts (SWCD) review and evaluate erosion control plans. At typical staffing levels, SWCDs in developing
areas find it very hard  to keep  up with the large number of developments they  are responsible for. The  local SWCDs
inspect the construction sites to establish whetherthe developer is implementing the soil erosion control plan correctly and
to observe whether the possibility of or the actual transport of sediment off-site exists. The SWCD will provide the
developer  with written recommendations  describing which erosion control measures  need to  be improved,  maintained,  or
installed. The developer then has two  weeks to comply with the recommendations.  If the developer is not found in
compliance with the requirements  after recommendations have been made, the SWCD reports the site to the Urban
                                                      407

-------
Erosion Control Specialist from the Indiana Department of Natural Resources (IDNR), who has been receiving copies of
all written warnings to the developer. The Urban Erosion Control Specialist will then visit the site and determine whether
the site should be reported to the IDEM. Subsequently, the IDEM determines whether further action, such as levying a
fine against the developer, is warranted. This process can lead to delays of many months  between identification of a
problem and regulatory enforcement.

    In reality, it takes a great deal of coercion to get developers to promptly apply erosion control measures on their sites.
Developers find applying erosion control measures inconvenient, costly, and time consuming, and are fully aware of the
lack of regulatory  personnel to  enforce local, state, and federal mandated erosion control (Harbor et al., 1995). Therefore,
developers  often do not comply with  the regulations and let their sites remain  bare (Harbor, in  press;  Harbor et al., 1995).
When inspected, sites are often either lacking erosion  control measures or maintenance of existing control measures is
long overdue. The effort (if any) on the developer's part to maintain or implement the erosion control measures is often
inadequate  and is done to appease the local SWCD, rather than with the goal of achieving 'best management' of the site.
Aside from regulation, there is little incentive for a developer to use erosion and sediment control practices. In fact, a
developer who uses erosion control may  be at a cost disadvantage compared to other developers who do not, thereby
making construction less profitable (Harbor, in  press).

Origin  of this  study

    In a study evaluating the use of rapid seeding and  mulching to reduce NPS pollution from construction sites, one
developer commented  that he liked seeding because he thought that  it made his developments  more marketable (Harbor
et al., 1995). The developer soon began  to include extensive seeding on his other developments to  achieve the same
neat, green looking result. Even though the developer was interested in seeding because he thought it would give him
a competitive edge over other developers,  he was voluntarily using vegetative  erosion control (Harbor et al., 1995; Harbor,
in press). As similar anecdotal evidence accumulated, it seemed possible that a higher market value for a seeded site
might provide an  incentive  for  voluntary  erosion control. If an economic advantage can be established, then it may be
possible to  persuade developers to use erosion control on the basis of a profit  motive, where regulation and education have
proved ineffective. If widespread voluntary  application can  be achieved by this means of  increased  profitability, it will make
it easier to  obtain  compliance with erosion  control programs and reduce the burden on regulators.  Furthermore,  and  most
importantly, the NPS pollution  load from construction sites would be  reduced.

Methodology

    We hypothesize that green, grassed  lots are more  attractive to buyers and therefore may be valued more highly and
sell faster than bare, dirt lots. There are several ways to test this hypothesis, with the most thorough being a detailed
tracking of the sales prices and sales timing of a large number of randomly selected treated and untreated control lots on
residential constructions sites throughout the US.  In the absence of data to perform this type  of highly detailed  approach,
we undertook a  pilot  study using photos of treated  and untreated  lots in a market  survey questionnaire  aimed at
establishing whether lots with green vegetative cover are valued higher than barren ones by Realtors, developers, and
homebuyers. In the work reported here, however, we do not evaluate whether green, grassed lots sell faster than bare,
dirt lots.

    Randomly selected lots in  three residential housing developments in Ohio and Indiana were seeded  and mulched.
Photographs of these lots were taken prior to seeding and then when the grass  was approximately  one  inch high (Fig. 2).
Lots were photographed from three angles (front  left, front center,  and front right), and selected photos were used in  a lot
valuation survey.  The market survey was designed  as a  broad tool  to investigate  a  wide range of factors which
homebuyers,  Realtors,  and developers  find important when buying/selling a lot in a residential housing development. The
survey included open- and  closed-ended questions, and those surveyed were not told the actual purpose of the survey.
The survey  is reproduced in Herzog (1997). Included within the wide range of questions in the survey were specific
questions on the  importance of lot appearance, and  a lot  valuation question in which those surveyed  were asked to place
prices on lots shown in photographs. Those surveyed were told the lots were in the same neighborhood/subdivision, with
the streets  and curbs  installed and  had the samesewer/septic system, water system, and noise level.  They were then
                                                      408

-------
Figure 2. Examples of grassed and bare lot photographs used in the survey.

given 10 lot  photos and asked to  establish  prices for each lot,  having been told that the average  lot value  in the
development was $20,000.

     Most Realtors and developers were interviewed at their offices in St. Joseph County, Indiana and Geauga County,
Ohio. Potential homebuyers were interviewed either at a neutral location or at their place of work, and included residents
of St. Joseph County and West Lafayette, Indiana, as well as personnel at a chemical engineering facility in Buffalo, New
York which was relocating to  Philadelphia, Pennsylvania. The survey typically took 10-20 minutes depending on the
responsiveness of the individual.

    After completion of the surveys, comparative statistics were used on the lot valuation data to assess whether there
was  any significant difference between "brown" and  "green" lot values for  Realtors,  developers, and homebuyers. Analysis
of variance was initially used to be able to test for the existence  of significant interaction between the fixed variables
(respondent group and color), while taking into account variation that occurs in the random variables (eg., subjects). The
assumptions  needed to appropriately apply this method, such as normality of the  error terms, were found to be satisfied
(Montgomery, 1997).
                                                      409

-------
                                  Price Differences for Green and Brown Lots
                                         Results  From a Lot Valuation Study
                                  Developers     Realtors    Homebuyers    Everyone
Figure 3. Differences in average prices for green and brown lots between survey groups.

Table 2. Effect of lot treatment on price for three different survey groups
Group
Realtors
Homebuyers
Developers
Green Lot
Mean Price ($),
sample size (n)
20,71 1
(n=155)
20,250
(n=36)
20,469
(n=48)
Brown Lot
Mean Price {$),
sample size (n)
19,967
(n=154)
19,500
(n=36)
20,218
(n=48)
Price
Difference
($)
744
750
251
Test
Statistic
t=4.0085
t= -1 .7957
t= -0.9200
Significance
Level
(p-value)
0.0001
0.0788
0.3609
    An important element of the economic analysis of lot greening is the actual cost involved in applying seed.
This can vary widely depending on the method used to apply the seed, and the density of vegetation desired.  In
this work, we restrict the analysis to an amount and type of cover intended for erosion control, as opposed to grass
species and density intended for final lawn cover. For this study we used independent contractors to apply seed,
mulch, water and  fertilizer by hydroseeding. Other common approaches include use of a hand seeder, and
mulching with straw either  by  hand  or using a blower.  During dry seasons in some areas, watering  may be
necessary to produce successful germination and early growth. Thus there is a wide range of possible  costs of
lot greening. In this study  we use the actual cost  of hydroseeding for our study sites in Indiana and Ohio, $300 per
lot,  although we could  have  applied seed and  mulch by  hand for about $100 per lot. Readers may want  to contact
their local Soil and Water Conservation District to get estimates of typical costs for their areas.

Results

    A total of 478 lot valuations  (310  by Realtors, 96  by developers, and  72 by homebuyers) were made.
However, during the survey process, it became apparent that two of the photographed lots were  being ranked
either highest or lowest based on their specific  background (one with a fire hydrant and another with lush tree
vegetation behind the lot giving an  appearance of more  privacy than the  other lots). Lot  valuations based on these
two photos, one green lot and  one brown were eliminated prior to the statistical analysis.
                                                  410

-------
    Initial statistical evaluation of the entire data set focused on determining if the data fit a model in which price
was a function of the overall mean price,  effects related to the group, the individual surveyed, the individual lot and
the lot color plus interaction and random error terms. In this model we assume that the effect of the particular lot
and the particular individual are random variables that are independently and normally distributed with a  mean of
zero, and also that the error terms are independently and identically distributed as normal random variables with
mean of zero and variance O2. Analysis of variance followed by a normal probability plot of the residuals, and
plotting  of the residuals versus the predicted values, demonstrated that the error terms were normally distributed
with constant variance (Montgomery,  1997). This analysis also demonstrated that there appeared to  be significant
differences in the variations of prices between groups, which complicates  analysis of the data as a combined
group. Thus it was necessary to analyze each group separately, using at test to evaluate the overall effect of color
within each group.

Realtors

    The Realtors surveyed gave an average value of $20,711 on the green grassed lots and $19,969  on the brown
dirt lots. The distributions  of lot values for green and brown lots were statistically significantly different at a 99.99%
confidence level (Fig. 3, Table 2). As a simple difference between means, the perceived added value for green
lots was $742 per lot.

    Narrative questions on the survey  revealed additional qualitative insight into Realtors' perceptions of lot value,
and reasons for preferring green lots. One Realtor commented: "I don't like these  mud lots." Others said the grass
was more appealing and "easier on the eye," and that the lots look better because they are green.  Other  Realtors
did not see the importance of seeding and believed that grass should not enter into the decision because it will
be  destroyed in the house building process.  "Grass makes it look better but means nothing for what's coming."

    Overall Realtors perceived that  homebuyers would prefer the grassed  lots ("I think people like grass,") and
the green lots would sell first because the green grass will remind homebuyers of a yard and allow them to
visualize what a house and yard would look like on the lot. One Realtor stated that the grass/ground cover was
more appealing than dirt,  and that homebuyers "wouldn't like the bare ones very much." Another noted that buyers
would be more willing to walk a grassed lot in inclement weather lot because the grass would absorb the moisture
and that buyer would not walk a dirt lot  because it would become muddy and puddle. One Realtor stated: the
green lots look "lush and fertile;" some people cannot visualize dirt lots as possibly being lush and fertile. This
Realtor  also brought up the concern that a buyer may ask about the drainage if the lot is wet, and if the dirt lot is
dry, caked, and cracked,  the buyer will wonder if anything can grow on it.

Homebuyers

    The homebuyers surveyed  placed an average value of $20,250 for the green  lots and $19,500 for the brown
lots. The  distributions of lot values  for green and  brown lots were statistically significantly  different at a 92%
confidence level (Fig. 3, Table 2). As a simple difference between means, the perceived added value for green
lots was $750  per  lot. The added value of $750 (the greatest added value among the groups surveyed) is
particularly significant because homebuyers are the ones who actually pay for the lots.

    Homebuyers stated that grass gives a realistic impression of the future appearance of the lot and it is more
appealing; and that the  final  product is more  difficult to  visualize on dirt lots. In general, the homebuyers
acknowledged that the grass  looks good and has more appeal, while understanding  that the lots would be
disturbed  during construction.  One homebuyer said that the green look was nicer but that it "wouldn't effect my
decision to buy," because grass was  not a "big deal." Even though comments such as these were made, on
average homebuyers valued green lots $750 more  than brown lots.
                                                  411

-------
Developers

    The developers surveyed placed an average value of $20,469 on the green lots and $20,219 on the brown
lots. The distributions of lot values for green and brown lots were statistically significantly different at only a
64% confidence level (Fig.  3, Table 2). Typically this would be viewed as indicating no statistically significant
difference. As a simple difference between means, the perceived value of the green lots was $250 greater
than the brown  lots. Clearly, the small difference between the green and  brown lots data sets and the
comparatively low significance level indicate that developers perceive little or no difference based on lot color.

    During the survey, developers addressed the difference between the  grass and dirt lots and stated that it
should not be a factor in lot price. They pointed out that the green lots will become brown lots during
construction and that the homebuyer will put in a yard anyway. Other developers saw that ground cover was
more attractive  ("I like the green") and perceived that homebuyers would  like the grass. Also, some perceived
that the green look made a development more  marketable  compared to other developments; one developer
said he "greens  up" his developments to make them look more attractive. Another  stated that grass makes a
lot look like it has topsoil, and if there are soil concerns, the grass demonstrates that vegetation can be grown
and is holding soil. One developer remarked how ground cover may be important to homebuyers for  more
than just appearance. He stated that grass cover is more significant when there is rolling ground because if
there is unseeded soil on an adjacent lot, the soil may erode onto the grassed property to the dismay of the
homeowner.

The Economic Incentive

    Although green lots may be  priced higher  than brown  lots, this gross value is only significant if  the price
differential exceeds the cost of seeding. The difference in value between grassed and bare lots compared to the
cost of seeding  provides a measure of potential net economic benefit to the developer. In terms of a simple net
return on investment, seeding a lot provides potentially excellent return. Homebuyers valued grassed  lots $750
more than brown lots, and as it cost $300 to seed a lot in this study, the developer stands to profit by $450 per lot,
which is a 150% return on the initial investment. The ability to more than double an initial investment should be
an attractive and sensible advantage for the developer, if the  perceived value difference actually translates into
a sale price difference.

Present Limitations and Future Work

    This pilot study is an initial step in developing information  that can be  used to persuade more developers to
make widespread use of vegetative  cover, and other forms of environmental protection. The results of this pilot
study are  most  relevant in areas where climate conditions  allow for relatively easy  establishment of temporary
vegetative  cover, and are not applicable to  arid or semi arid areas.  In  addition, budget restrictions limited  the scale
of the study. Although we  collected 478 lot valuations from 62  respondents, a much larger study with respondents
from many areas of the United States would overcome a  potential  criticism that the current study only represents
conditions in  a small portion  of the  Midwest. Developers are also more  likely to notice results  based on data
collected within  their region, especially if these are  coupled  with regional demonstration projects. Thus, the next
logical step is to initiate a network of coordinated studies in regions experiencing rapid residential development.
This would provide for analysis on a national as well as a regional level, and for comparisons between regions.

    An additional limitation of the results presented here is that  they consider only perceived increase in lot values.
In actual sales transactions, buyers  may not actually behave  in the way they say they would on a survey. As a
linked project, it  would be  desirable to track actual sales histories (timing  and pricing) to provide a more complete
picture of the actual economic impact of lot greening. Future research should include analysis of a large number
of real-world transactions for which lot condition is known. This could be based on a  large-scale, long-term study
in which researchers intervene to change  lot conditions on selected  lots or developments. Alternatively, if some
                                                 412

-------
landowners are convinced by the results of this study, the experiment might occur naturally in the marketplace as
the findings of our work are disseminated.

    Further extension of the basic concept of examining the direct profitability of environmental protection is also
possible. We were recently contacted by a consultant who had heard of the lot valuation study, and wanted a
similar study performed on the increased value of lots next to ponds on developments. Although ponds are often
built for stormwater control, and  also aid  in reduction of nonpoint source pollution, they  can also  have considerable
aesthetic appeal. Thus, it would be potentially very useful to know what the return on investment on a pond is for
a new residential development, both in terms of the increased price of lots adjacent to the pond, as well as the
increase in  average price in lots for the development as a whole because of the improved appearance of the
development.

Conclusions

    Showing that erosion control may be profitable provides a new way to reach developers who have failed to
act on the logic that erosion control provides environmental protection and is required  to comply with local, state
or federal erosion control regulations. Evaluating the cost of environmental damage is not only very difficult, but
also of little  direct relevance to a developer who  does  not directly  pay the cost of the damage.  Land development
is  a business,  with profit as a leading motive, so  appealing to  increased profitability is one potentially  effective way
to  change behavior.

    The pilot study described here  indicates that vegetated lots are perceived to be more valuable and more
desirable by Realtors and homebuyers. Realtors perceived that vegetated lots are worth more than barren lots
(by $742). They also perceived that vegetated lots are worth more to homebuyers and that homebuyers would
be willing to pay more for grassed lots.  Homebuyers also perceived grassed lots to be  more valuable and  put the
largest premium  ($750) on the lots for  all those surveyed. The added lot value is only significant if the price
differential exceeds the cost of seeding  ($300), which was the case for Realtors and  homebuyers by $417 and
$450, respectively. Developers valued the vegetated lots higher than non-vegetated lots by an average of $250,
but the difference was  not statistically significant. Even if it were significant, this price differential is less than the
cost of seeding and indicates  that developers perceive that seeding  costs are greater than the benefits  of
vegetation. This perception of a net cost associated with greening  a lot is perhaps why the market has largely
failed to recognize this simple way  to increase a property's value.  Some developers did recognize the visual
appeal of the  grass and believed that a greened development would attract homebuyers more rapidly  than a
development that  appears unkempt.  However, thevaluation  study indicates  that  developers have not  aligned  their
perception of lot value  with that of homebuyers.

    An alternative way of interpreting the results is to consider the potential  return on the investment in the
vegetative cover. For  a $300 investment the developer can receive a return of $750, i.e. a 150%  return on
investment. Such  a rate of return is  difficult to achieve in most conventional investments. Finally, price differential
is  not the only economic benefit  of lot greening; if lots sell faster because of greening, profits will increase because
of lower financing costs for capital invested  in the development process.  Further research is  needed to  clearly
define the value  of this  potential economic impact  associated with lot greening. However, at this stage  it is
possible to  state that in addition to the environmental benefits, and regulatory requirements associated with using
vegetation for  erosion control, there may be significant marketing and thus economic  returns associated with lot
greening.

    Education  concerning the environmental benefits  of erosion control, and enforcement of regulations have not
produced widespread, effective use of vegetative  cover for  erosion control. Because  developers generally  do not
perceive much incentive to vegetate their developments aside from complying with often-ineffective regulations,
they typically do  not. Typically  a developer who is using erosion  control practices  believes  s/he  is  at  a  cost
disadvantage  compared to  other developers who are not, thereby making the developer following regulation
believe s/he will be less profitable.  Furthermore, the developer does not directly pay for the mitigation of the
environmental impacts caused by the sediment leaving the site; it is the burden of the taxpayer instead.  In this


                                                  413

-------
study, we have demonstrated that vegetating a development may be a profitable investment. Appealing to the
profit motive will hopefully provide a way to generate widespread use of vegetative cover that also provides erosion
prevention on construction sites.

References

Bhaduri, B., J. M. Harbor, P. Maurice. 1997. Chemical load fractionation and trap efficiency of a construction site
storm water management basin. Environmental and Engineering Geoscience. Ill: 235-249.

Clark, E. H. 1985. The off-site costs of soil erosion. Journal of Soil and Water Conservation 40: 19-22.

Clark, E. H., J. A.  Haverkamp, and W. Chapman. 1985. Eroding Soils: The off-farm impacts. The Conservation
Foundation, Washington, D.C. 252 pp.

Colacicco, D.,  T.  Osborn,  and K. Alt. 1989. Economic damage from soil erosion. Journal of Soil and Water
Conservation 44: 35-39.

Crawford, J. K. and D. R. Lenat. 1989. Effects of land use on the water quality and biota of three streams in the
Piedmont Province of North Carolina. US Geological Survey Water  Resources Investigations Report 89-4007.
Raleigh, N.C. 64 pp.

Ekwue, E. I. 1990.  Effect  of organic matter on splash detachment and  the process involved. Earth Surface
Processes and Landforms. 15: 175-1 81.

Federal Register.  1990. Fed. Reg. 55. November 16. 47990.

Fennessey, L.A. and A. R. Jarrett. 1994. The dirt in a hole: A review of sedimentation basins for urban areas and
construction sites. Journal of Soil and Water Conservation 49: 317-323.

Goudie, Andrew. 1994. Human Impact on the  Natural Environment. The MIT Press. Cambridge. 302 pp.

Goldman,  S. J.,  K. Jackson, and T. A. Bursztynsky. 1986. Erosion and Sediment Control Handbook. McGraw Hill
Publishing Company. New York. 454 pp.

Harbor, J. M., J. Snyder, and J. Storer. 1995.  Reducing nonpoint source pollution from construction sites using
rapid seeding and mulching. Physical Geography. 16: 371-388.

Harbor, J. M. In press. Engineering  geomorphology at the cutting  edge of land  disturbance:  Erosion and sediment
control on construction sites. Geomorphology.

Herzog, M., 1997.  Reducing erosion  problems from land  development:  An  economic  incentive for erosion control.
Unpublished MS Thesis, Department of Earth  and Atmospheric Sciences, Purdue University.

Indiana Department of Natural Resources.  1992. Indiana Handbook for Erosion Control in Developing Areas,
Indiana Department of Natural Resources. Indianapolis, Indiana.

Koehn, E. and J. A. Rispoli. 1982. Protecting the Environment During Construction. Journal of the Construction
Division, Proceeding ASCE. 108(CO2): 233-246.

Lemly,  D.  A. 1982. Erosion  control  at construction sites on red clay soils. Environmental Management.  6: 343-352.

Montgomery, , D.C., 1997. Design and analysis of experiments. 4th edn.  John Wiley & Sons, Inc., New York.
704 p.
                                                414

-------
Novotny, V. and G. Chesters. 1989. Delivery of sediment and pollutants from nonpoint sources: A water quality
perspective. Journal of Soil and Water Conservation 44: 568-576.

Ohio Department of Natural Resources. 1980. Model Regulations for Urban Soil Sediment Pollution Control.
Columbus,  Ohio.

Paterson, R. G., M. I. Luger, R. J. Burby, E. J. Kaiser, H. R. Malcolm,  and A. C. Beard. 1993. Costs and benefits
of urban erosion and sediment control: The North Carolina experience. Environmental  Management. 17:  167-1  78.

Pimentel, D., C. Harvy, P. Resosudarmo, K. Sinclair, D. Kurz, M. McNair, S. Crist, L. Shpritz, R. Saffouri, and R.
Blair. 1995. Environmental and economic costs of soil erosion and conservation benefits. Science. 267: 1117-
1122.

Rogers, R.D.  and Schumm, S.A.  1991.  The effect of sparse vegetative cover on erosion and sediment yield.
Journal of Hydrology 123: 19-24.

Satterlund, Donald R.  1972. Wildland Watershed  Management. The Ronald  Press Company.  New York. 370  pp.

Trotti, T. 1997. Developer fees don't hack it. Erosion Control. Nov/Dec 1997: 6.

Virginia Department of Conservation  and Recreation. 1992. Virginia  Erosion and Sediment Control Handbook.
Conservation  and Recreation,  Richmond, Virginia.

Wolman,  M. G. and A.  P. Schick. 1967. Effects of Construction on Fluvial Sediment, Urban and Suburban Areas
of Maryland. Water Resources Research 3: 451-464.
                                                415

-------