-------
SELECTION OF THE MODEL
After the models and their capabilities have been introduced in the preceding
sections, one question managers and urban planners may ask is which model
to select and for what purpose. The ideas of an universal model that could
do everything that was advocated in 1970's by some developers of the mathema-
tical models, may not be as popular and feasible today. The progress and
availability of small and relatively cheap micro- and minicomputers have re-
volutionized the model applications. A highly powerful (500K RAM memory) fast
microcomputer can now be purchased for $1500 to $2500. These computers are
comparable in their capacity and speed to medium size computers of the 1970's.
What this means is that a menu of smaller models may be more useful to a
practicing urban engineer and manager than one large model. Thus, one should
look at the model selection question from the point of view of what purpose
and tasks the model should serve. The model itself should be looked upon
as a tool to accomplish the task and objectives.
As stated in the preceding section, most of the hydrological models are
limited in their areal extent. A typical size of the watershed best suited
for detailed hydrologic modeling would be from a fraction to several km2.
For example the applications of ILLUDAS model should be limited to basin
areas of less than 13 km2. Thus, most of the applications have been limited
to small experimental watersheds and techniques on how to expand the modeling
to larger watersheds are still in evolution. Bearing in mind that commonly
only a portion of a typical urban or urbanizing watershed is responsible for
most of nonpoint pollution, overview or screening modeling techniques should
be made compatible with the hydrological detailed models. In a typical water-
shed nonpoint pollution abatement process the modeling effort should proceed
in the following phases:
(I) Overview screening models should identify the problem areas.
(II) Detailed continuous models (level III or V) could be used for obtaining
more accurate nonpoint pollution loadings figures and for screening
the various management practices and structural measures and their
water quality impact. Such modeling activities must consider both the
source strength and pollutant delivery.
(Ill) A detailed event oriented model (level IV) could be used to finalize
the design of some technically complex structural and nonstructural
measures selected and evaluated in Phase II. Such modeling may include
design of storage for multiple overflows, various treatment schemes,
and redesign of drainage systems.
A schematic diagram of the process is shown on Figure 3.
Calibration and verification with field measured data may be necessary for
Phase II modeling if accurate loadings of pollutants are desired. Calibra-
tion is less important if the effect of various management practices on the
loadings is studied.
This approach leads to a hierarchical modeling process whereby the knowledge
and extensive testing associated with some large and complex urban runoff-
pollution models can be extrapolated to the practical load management models
and approaches used for selecting of hazardous critical areas within the
K-VI-10
-------
Goals and
Objectives
Preliminary Data
Collection and Surveys
Delineation of Hazardous
and Problem Areas
Screening Model
Level I or II
Selection of Critical Areas for
Detailed Modeling, Selection of
Alternatives and Management
Practices
Data Collection for
Calibration
Detailed Modeling and
Evaluation of Alternatives
Final Detailed Design
Selection of a Detailed
Hydrologic Model -
Level III or V
Model Calibration
and Verification
Event Oriented
Model (Level IV)
Figure 3
Schematic of the Modeling Process
K-VI-ll
-------
watershed and for selection of appropriate management practices. A detailed
hydrologic model calibrated and verified by small plot field studies can be
used to estimate unit loads for a screening overview model as shown by
Novotny and Bannerman (1980).
This concept has been used by the Wisconsin Department of Natural Resources
to identify the problem areas within so called "priority watersheds" that
is watersheds with a serious nonpoint pollution problem.
The Model Enhanced Unit Load concept (MEUL) enables selection of the priority
watersheds and the problem areas within them. The units loadings were gen-
erated by a calibrated and verified hydrologic-nonpoint model. In this
study loadings were generated for 1 km2 (100 ha) area! units located on four
hydrologically uniform soil types with a uniform slope of 6%. A time series
summarizing several years of meteorological data was statistically analyzed
to provide long term average loadings under different land use conditions.
In a subsequent study that focused on urban areas only, a hydrological level
III model, calibrated and verified by 3 year monitoring program encompasing 12
small urban watersheds, was used to extrapolate the findings into more general
relationships (Novotny (1983, 1984), Novotny et al. (1985)). Figure 4 shows
such a relationship whereby pollutant loadings from hypothetical unit urban
areas (residential and commercial land use) are related to the percent imper-
viousness of the area soil type, and street sweeping practices.
USE OF THE MODELS FOR SELECTION OF CONTROL OF URBAN RUNOFF
There are a number of measures that can be used to control pollution by urban
runoff. These methods will be presented in detail in a subsequent paper.
From the modeling standpoint, urban runoff (quantity-quality) control measures
can be categorized as follows:
(1) On-site source control measures that include control of atmospheric de-
position, litter control programs, leaf and grass pick-up, street sweep-
ing. These measures affect the input and accumulation of pollutants
on urban impervious areas.
(2) Control of urban pervious areas, including mostly control of bare and
unprotected soils within the watershed, especially those due to con-
struction. This is modeled by adjustment of a few hydrological para-
meters (depression storage, runoff curve number, infiltration rate) and
by the cover factor, C, in the Universal Soil Loss Equation.
(3) Hydrologic modification of urban watersheds. Many measures in this
category of urban nonpoint pollution control are similar to those
used for control of flooding. But not all flood control practices can
control nonpoint pollution. Simple flow retardation, for example, in a
dry detention pond or surface ponding on impervious surfaces in areas
served by storm sewers may have no effect on reduction of pollution
loads. Hydrologic modification for modeling purposes can be described
as follows:
(a) practices that increase permeability and enhance infiltration, such
as the use of pervious pavements or vegetation infiltration strips;
(b) practices that will increase hydrologic storage, such as rooftop
storage on flat roofs, temporary surface ponding, and restriction
of stormwater inlets;
K-VI-12
-------
500
O 400
LJ
CO
300 -
CD
z 200
o
100 .
IMPERVIOUS /
UNSWEPW
20 40 60 .
% IMPERVIOUSNESS
80
100
Figure 4 - A schematic representation of sources of urban particulate
nonpoint pollution related to imperviousness, soil type and
street sweeping. The loadings are expressed for 1 km2 unit
watersheds with 6% slope.
K-VI-13
-------
(c) practices that will reduce directly connected impervious areas,
that include disconnnecting roof drains, letting surface runoff
overflow on adjacent pervious areas, use of dry wells, and use
of pervious pavements.
(4) Reduction of delivery of pollutants. After the particulate pollutants
leave the source areas, they can settle or be removed on route from the
sources to the receiving water body. A delivery ratio factor expresses,
in the models, the ratio of the amount of pollutants delivered to a
receiving water body (pollutant yield) to the amount of pollutant
generated at the source.
Urban areas with a good storm drainage have a delivery ratio close to
one, meaning that almost all pollution, after it leaves the source,
is eventually delivered to the receiving water body.
Residential areas with a natural (swale) drainage have delivery ratios
ranging from a few percent to about 50 percent. As a consequence, urban
areas with a natural drainage produce much less pollution than areas
with a good storm sewer system.
(5) Runoff detention-retention. The terms "detention" and "retention" have
been used synonymously in most literature interpretations and in common
usage. However, with recent advancements of the detention-retention
technology, detention refers to a stormwater storage facility that is
normally dry and is designed to temporarily hold stormwater during high-
peak runoff events. Such basins have only a very limited efficiency
in controlling urban runoff pollution. A retention facility always
contains a substantial volume of water to serve other purposes such as
recreation or aesthectics. The modeling of the effects of these facili-
ties to control urban runoff is significant. The modeling of the effects
of these facilities may vary in sophistication. The simple techniques
will adjust the erosion control practice factor P of the Universal Soil
Loss Equation using rule-of-thumb factor. Such factors have been
developed by Driscoll (1983) and they can be used in combination
with the Level II or III models.
A number of more complex models of retention facilities evolved in the
last ten years. A summary of the models was presented by Prasad (1984).
One of the most advanced retention facility models is SEDIMOT (Wilson
et al. (1983)). Most of the Level IV and V models mentioned herein
have a retention storage segment.
Example of the Model Appli cati on
The Level III Urban Runoff Model developed by the author (Novotny (1983)
and Novotny et al. (1985)) was used to study the effect of several low cost
management practices.
In the first phase of the research the model was calibrated and verified by
monitoring data of eight small urban experimental watersheds gather, by the
Wisconsin Department of Natural Resources (Bannerman et al. (1984)).
K-VI-14
-------
For each watershed the monitoring program provided the following information:
(a) atmospheric deposition rates (wet and dry),
(b) pollutant accumulation in the curb storage,
(c) runoff quantity and quality,
(d) traffic count and land use characteristics.
The calibration process preceded as shown on Figure 5:
(a) The generated street pollutant accumulation was matched against the
monitored data (Fig. 6).
(b) When adequate fit of street accumulation data was obtained the cali-
bration then preceded to matching runoff quantity for storm events
with monitored data.
(c) Runoff quality calibration was the last step (Fig. 7).
Generally good to satisfactory fit of measured and computed data was obtained
for most of the pertinent watersheds.
After the calibration precipitation data for 1981 and 1982 was used to generate
seasonal and annual (excluding winter) loadings, the output from the model
provided event runoff volume and quality for five constituents -- suspended
solids, volatile suspended solids, nitrogen, phosphorus, and lead. These
simulated loads and breakdown of sources of pollution are shown in Table 1.
This type of study provides the manager with information on magnitudes of
nonpoint pollution and on the source strength. For example, in the Milwaukee
study modeling indicated that atmospheric deposition accounted for less than
5% of the input, traffic became an important source of lead only in high
traffic watershed, and effect of street sweeping was not significant.
The effect of several management practices on water quality was then docu-
mented on a case of a "hypothetical" watershed that was set-up using the
surface, soil, and land use data similar to those established for the study
sites. The watershed consisted of the following land uses:
medium density residential zone 100 ha
low density residential zone 7.5 ha
open unused land (natural grass) 23.6 ha
commercial (retail) 12.7 ha
transportation (freeway) 14.5 ha
parks 19.50 ha
Figure 8 is a map of the hypothetical watershed. The soils were in silt
and silty clay loam textural classes. Slopes ranged from 0 to 6 percent.
The following alternatives were studied:
Alternative 1 - No action.
Alternative 2 - Effective street sweeping.
Alternative 3 - Street sweeping and grass strips along the roads.
Alternative 4 - Street sweeping and two retention ponds providing 50%
reduction of pollutants.
Alternative 5 - Street sweeping, two retention basins and grass strips
along the roads.
K-VI-15
-------
STREET POLLUTION
ACCUMULATION
MONITORED DATA
RUNOFF QUANTITY
RUNOFF QUALITY
Litter Deposition
Rate
Pervious Area - CN
Pollutants Cone.
in Litter
Litter Particulate
Distribution
Impervious Area - DS
Breakdown Factor
DUSTFALL
TRAFFIC
1
4 *
MEDIUM
i
LITTER
1
I
COARSE
BREAKDOWN
Figure 5 - Calibration Process Schematics
K-VI-16
-------
250
MEASURED COMPUTED
FINES A
MEDIUM X
COARSE O
100 140
DAY OF THE YEAR
Figure 6 - Calibration Results of Street Solids
Accumulation Process
1000-^
0>
S 100 -d
O
CO
Ol
CO
z>
Q
LU
OC
UJ
CO
CD
O
10
-i 1 1—i—i—i i i
100 1000
COMPUTED SUSPENDED SOLIDS, mg/l
Figure 7 - Calibration Results of Runoff Quality
K-VI-17
-------
Figure 8 - The Experimental Watershed
K-VI-18
-------
Table 1
Computed Sources and Outputs of Pollutants
From an Urban Watershed
Overall Mass Balance
Street Loads
Initial
Final
Daily Inputs
Dry Atm.
Litter
Street Dirt
Total
Event Inputs
Wet Atm.
Soil
Total
Outputs
Blown Out
Swept Out
In Runoff
Total
Susp. Solids
(ton)
9.225
6.641
8.052
93.614
271.149
372.814
1.496
14.577
16.073
328.965
.000
62.507
391.472
Organics
(kg)
897.443
529.347
797.155
24527.625
59919.688
85244.438
748.006
831.012
1579.018
73871.187
.000
13320.262
87191.437
Nitrogen
(kg)
14.761
8.715
12.883
149.813
362.217
524.913
299.202
346.410
645.613
465.394
.000
711.177
1176.571
Phosphorus
(kg)
21.139
12.645
8.052
257.288
631.968
897.308
3.740
20.118
23.858
801.693
.000
127.966
929.659
Lead
(kg)
13.838
24.615
8.052
140.444
960.308
1108.804
3.740
5.831
9.571
961.323
.000
146.275
1107.597
K-VI-19
-------
Figure 9 shows the unit loads of solids for the no action alternative (Alt. 1).
Such figures can aid managers and designers in their decision making to iden-
tify the problem areas and sources of pollution. In a typical mixed land use
watershed identification of hazardousland is necessary since not all lands and
land use segments require management. For example, established low density
residential areas with natural drainage have loadings comparable to undis-
turbed natural prairies and woodlands.
Figure 10 shows the cummulative probability plot of suspended solids concentra-
tions. The results indicated that grass strips along the roads (2 meter grass
strips between the impervious road surface and the drainage system - sewer
or swale) provided the best water quality benefit both in reducing the con-
centrations by 80% and runoff volumes by 50% for medium and by 25% for longer
storms, respectively.
Street sweeping was not an effective pollution control measure in this sensi-
tivity analysis, although it did reduce the pollutant outputs from highly
impervious segments.
The suggested optimal best management practices for this hypothetical case
study would include street sweeping in the commercial zone, two meter grass
strips along the roads and streets, and retention ponds to intercept flows
from open lands and residential areas located on high slopes and poor (silty
clay loam) soils.
In addition, grass clipping control, leaf pick-up during the fall season and
litter control ordinances should be implemented.
Conclusions
The urban runoff quantity-quality models are indispensable tools in the eval-
uation of sources of pollution from urban and urbanizing areas and the design
of the abatement measures. Such models can:
(1) Estimate loadings of pollutants from nonpoint sources under different
hydrological and meteorological conditions.
(2) Estimate source strengths of different contributing areas within urban
and urbanizing watersheds.
(3) Evaluate relative efficiency of various management practices and
structural measures.
4) Enable design of nonpoint control measures.
5) Aid in watershed planning.
Both screening and hydrological models should be used, however, not on the
same scale and with the same detail. Screening models are useful for identi-
fication of problem areas within larger watersheds. Hydrological models are
then subsequently used for detailed modeling of nonpoint pollution loadings
and evaluation of the impact of abatement.
K-VI-20
-------
UNIT LOAD
kq/ha year
250.300
500-IOOO
XOOO
Figure 9 - Units Loadings of Solids from the Experimental
Watershed for Alternative 1 (Ho Action)
K-VI-21
-------
1000
9oo-
o
V)
o
ui
o
z
llj
a.
CO
D
10-
10 20
40 60
i
80
90 95
PROBABILITY 4.
Figure 10 - Cumulative Probability Plot of Suspended Solids Concentrations
1n the Runoff for 5 Alternatives Measured by the Model
Alt. 1 - No actlpn,
Alt. 2 - Street sweeping only,
Alt. 3 - Street sweeping in coranericial zones plus grass strips along the
roads in residential zones,
Alt. 4 - Street sweeping plus two retention ponds,
Alt. 5 - Same as Alt. 3 plus two retention ponds.
K-VI-22
-------
REFERENCES:
American Public Works Association (1981). Urban Stormwater Management.
Spec. Report No. 49, Chicago, IL.
Anon. (1975). Urban Storm Water Runoff - STORM. The Hydro!ogic
Engineering Center, U.S. Army Corps of Engineers, Davis, CA.
Bannerman, R., et al. (1984). Evaluation of Urban Nonpoint Source Pollution
Management in Milwaukee County, Wisconsin, Vol. I. Urban Stormwater Charac-
teristics, Pollutant Sources and Management by Street Sweeping. A Report
for U.S. EPA, Region V., Chicago, IL.
Barfield, B.J., Kao, D.T.Y, and Toller, E.W. (1975). Analysis of the Sedi-
ment Filtering Action of Grasses Media. Res. Pap. No. 90, University of
Kentucky Water Resources Research Institute, Lexington, KY.
Chow, V.T. (1972). Hydrologic Modeling. Journal of the Boston Society of
Civil Engineers, 60, pp. 1-27.
Donigian, A.S., and Crawford N.H. (1976). Nonpoint Pollution Model from
the Land Surface. EPA 600/3-76/083. U.S. EPA, Washington, DC.
Driscoll, E.D. (1983). Performance of Detention Basins for Control of Urban
Runoff Quality. Proc. International Symposium on Urban Hydrology, Hydraulics,
and Sediment Control. University of Kentucky, Lexington, KY.
Haith, D.S., and Tubbs, L.J. (1981). Watershed Loading Functions for Nonpoint
Sources. Journal Env. Eng. Div., ASCE, 107, pp. 121-137.
Hall, M.J. (1984). Urban Hydrology. Elsevier Applied Science Publ., Ltd.,
Essex, England.
Hydrocomp, Inc. (1979). User's Manual for the Hydrologic Simulation Program/
Fortran. Mountain View, CA.
Huber, W.C., Henry, J.P., Nix, S.J., Dickinson, R.E., and Polinon, D.J. (1984).
Stormwater Management Model User's Manual - Version III, EPA-600/S2-82-84-109
a & b, U.S. EPA, Cincinnati, OH.
Johnson, M.J., et al. (1978). Management Information Base and Overview Model-
ing. Inter. Joint Commission, Windsor, Ontario, Canada.
McElroy, A.D., et al. (1976). Loading Functions for Assessment of Water
Pollution from Nonpoint Sources. EPA 600/2-76-151. U.S. EPA, Washington, DC.
Novotny, V. (1983). Model of Nonpoint Pollution of Eight Small Watersheds in
Milwaukee County. A Report by Novotny Associates to Wisconsin DNR, Milwaukee,
WI.
Novotny, V. (1984). State-of-the-Art of Hydrological and Water Quality Models
Used for Simulation of Nonpoint Pollution from Agriculture. Proc. International
Conference on Agriculture and Environment 1984. Venice, Italy, June 11-15.
K-VI-23
-------
Novotny, V. (1984). Efficiency of Low Cost Practices for Controlling Pollu-
tion by Urban Runoff. Proc. 3rd International Conference on Urban Storm
Drainage, G'dteborg, Sweden, June 4-8, pp. 1241-1250.
Novotny, V., and Banner-man, R. (1980). Model Enhanced Unit Loadings of
Pollutants from Nonpoint Sources. Proc. Hydraulic Transport Modeling
Symp., Publ. No. 4-80, ASAE, St. Joseph» MI.
Novotny, V., and Chesters, G. (1981). Handbook of Nonpoint Pollution:
Sources and Management. Van-Nostrand-Reinhold Publ., New Yrok, NY.
Novotny V., Sung, H.M., Bannerman, R., and Baum, K. (1985). Estimating
Nonpoint Pollution from Small Urban Experimental Watersheds. Paper to be
published by Journal of Water Pollution Control Federation.
Prasad, N.S. (1984). Modeling Urban Stormwater Retention Basins. M.S.C.
Thesis, Department of Civil Engineering, Marquette University, Milwaukee, WI.
Price, R.K. (1984). Development and Implementation of the Wallingford, Pro-
cedure. Proc. Third International Conference on Urban Storm Drainage.
GSteborg, Sweden, June 4-8, pp. 445-452.
Sonzogni, W.C. et al. (1980). WATERSHED: A Management Technique for Choosing
Among Point and Nonpoint Control Strategies. Proc. of the Seminar on Water
Quality Management Trade-offs. U.S. EPA, Chicago, IL.
Sutherland, R.C. (1980). An Overview of Stormwater Quality Modeling. Proc.
International Symposium on Urban Storm Runoff, University of Kentucky, Lexing-
ton, KY, July 29-31.
Terstriep, M.L., and Stall, J.B. (1969). Urban Runoff by Road Research
Laboratory Method. Journal of the Hydr. Div.-ASCE 95, pp. 1809-1834.
Terstriep, M.L., and Stall, J.B. (1974). The Illinois Urban Drainage Area
Simulator. Bulletin 58, Illinois State Water Survey, Urbana, IL.
U.S. Soil Conservation Service (1975). Urban Hydrology for Small Watersheds.
Technical Release No. 55.
Watkins, L.H. (1962). The Design of Urban Sewer Systems: Research into the
Relation Between the Rate of Rainfall and the Rate of Flow in Sewers. Road
Res. Tech. Paper No. 55 (HMSO, London)
Wilson, B.N., Barfield, B.J. and Moore, I.D. (1983). Design Manual for the
SEDIMOT II - Hydrology and Sedimentology Model. University of Kentucky,
Dept. of Agr. Engineering, Lexington, KY.
Wischmeier, W.H., and Smith, D.D. (1965). Predicting Rainfall-Erosion Losses
from Cropland East of the Rocky Mountains. U.S.D.A. Agricultural Handbook
No. 282, Washington, DC.
K-VI-24
-------
MANAGEMENT AND CONTROL OF POLLUTION BY URBAN RUNOFF
Richard Field, Chief
Storm and Combined Sewer Program
Office of Research and Development
U.S. Environmental Protection Agency
Edison, NJ 08837
Abstract
The paper reviews progress on urban stormwater management and pollution
control with emphasis on non- and low-structurally intensive techniques along
with the total system approach encompassing control-treatment.
Many of the United States Environmental Protection Agency's demonstra-
tion-evaluation projects are presented to exemplify: Land Management
Techniques, i.e., land use planning, best use of natural drainage, dual use of
retention and drainage facilities required for flood control designed
simultaneously or retrofitted for pollution control, porous pavement, surface
sanitation, and chemical use control; Collection Systems Control, i.e.,
catchbasin cleaning, flow regulators (including swirl and helical blend
devices), and the new concepts of elimination or reduction of unauthorized
cross-connections in-channeI/conduit storage and/or other forms of storage for
bleed back to existing treatment plants; In-Receiving Water Storage;
Treatment, i.e., physical/chemical, disinfection, and a treatment-control
planning and design guidebook; Sludge and Solids Residue from Treatment; and
Integrated Systems, i.e., storage/treatment dual-use wet-weather flow/dry-
weather flow facilities, and reuse of stormwater for non-potable purposes.
Recommendations for the future in the areas of: control based on
receiving water impacts, toxics characterization and their control, sewer
system cross connections, integrated stormwater management, and
institutional/sociological/economic conflicts are also presented.
Keywords: Storm sewers, Overflows—sewers, Drainage, Surface water runoff,
Runoff, Hydrology, Combined sewers, Water pollution, Water quality,
Wastewater, Sewage, Contaminants, Waste treatment, Storage tanks,
Disinfection, Mathematical models, Remote control, Hydraulics
K-VII-1
-------
INTRODUCTION
The mission of the Storm and Combined Sewer Program (SCSP) of the U.S.
Environmental Protection Agency (EPA) is to develop methods for controlling
pollution from urban stormwater discharges and combined sewer overflows
(CSO). This paper will emphasize research relevant to urban stormwater. This
program has been involved for 20 years in the development of a diverse
technology including CSO and stormwater control technology, instrumentation,
problem assessments, best management practices (BMP) development and
evaluation, stormwater management models, sludge handling and disposal
methods, infiltration inflow (I/I) control, erosion control, and many
others. Because the time allotted will not allow complete coverage, just the
more important products will be highlighted with emphasis on some recent
developments.
POLLUTION PROBLEM ASSESSMENT
Characterization
The quality and quantity characterization of urban stormwater discharges
is necessary for problem assessment, planning, and design. Summaries of
/I _0\
characterization from many research studies are available v '.
About 70 Ib/ac/yr of biochemical oxygen demand (BOD) in urban stormwater
runoff discharges contribute about 45% of the annual BOD load if secondary
treatment is provided for the dry-weather flow (DWF).
Since 1974, EPA supported an urban rainfall-runoff-quality data base
// 5\
v ' ' for two important data requirements: characterization of urban storm
runoff and calibration and verification of urban stormwater management models.
From an in-house project, preliminary screening of urban wet-weather
discharges from 24 samples from nine urban areas found that approximately one-
half of the 129 EPA's priority pollutants (Table 1). The heavy metals were
consistently found in all samples. Polynuclear aromatic hydrocarbons from
petroleum were the most frequently detected organics.
K-VII-2
-------
Table 1. Distribution of Priority Pollutants Frequently Detected
Pollutant
**
Frequency* ^
LiquidSediment Liquid (pg/1)Sediment (yg/kg)
Range
Phenols
Polynuclear Aromatic
Hydrocarbons (PAH's)
Fluoranthene
Pyrene
Anthracene
1,2-Benzanthracene
Benzo(a)pyrene
3,4-Benzofluoranthene +
11,12-Benzofluoranthene
Chrysene
Fluorene
Phenanthrene
5
4
3
3
3
2
2
2
2
2
2
0.3-0.9
0-98
0-76
1,200-4,000
1,200-3,000
375-7,000
1,100-4,100
600-25,000
2,800-4,600
1,500-4,100
210-300
375-2,000
Esters
Bis(2-ethylhexyl)phthalate
Di-n-butyl phthalate
Butylbenzyl phthalate
Heavy Metals
Arsenic
Beryllium
Cadmium
Chromium
Copper
Lead
Nickel
Silver
Zinc
Antimony
Selenium
Thallium
Mercury
6
6
—
7
7
7
7
7
7
7
7
7
7
7
7
7
3
3
2
3
3
3
3
3
3
3
3
3
3
3
3
3
0-160
0.5-3.4
—
2-15
1-2
2-7
9-36
11-49
30-400
9-40
3-8
90-330
<20
3-10
2-10
0.2-0.56
130-59,000
300-25,000
450-1,200
Sediment (mg/kg)
2.8-4.8
0.1-0.2
0.06-1
8.10-28
7-140
24-370
3-16
0.5-2.1
37-380
3-3.3
0.06-1.4
<0.2
2.1-2.4
*Reported when found four or more times out of seven liquid samples and two
or more out of three sediment samples.
**These are mostly grab samples and therefore cannot be accurately related to
pounds of pollutants discharged. Also, the mass emission of pollutants will
vary for each rainfall event.
K-VII-3
-------
A project in Syracuse, NY using the Ames Test to evaluate urban runoff
mutagenicity' ' ' found detectable responses for 22% of the samples. It is
significant that some mutagenic substances are present with a potential for
entering the food chain.
The discharge of pathogens from CSO and storm sewers is a major public
health concern. Indicators such as fecal coliform have long been known to be
present in stormwater discharges in densities sufficient to cause contraven-
tion of standards. A study in Baltimore, MD identified actual pathogens and
enteroviruses in storm sewer discharges' '. Cross-connections from sanitary
sewers were strongly implicated as the major cause. This problem is not
confined to Baltimore. The problem is likely to be widespread particularly
where sewers are older and cross-connections are present.
Receiving Water Impacts
Knowledge of the receiving water impacts resulting from urban wet-weather
discharges is a basis for determining the severity of problems and for
justifying control. Problem studies of receiving water impacts are described
in a proceedings from a national conference^ ' and in a journal paper^ '.
Under certain conditions, storm runoff can govern the quality of
receiving waters regardless of the level of DWF treatment provided. Based on
national annual mass balance determinations, urban wet-weather oxygen demand
loads are greater than the dry-weather Loads from the same areas^ ' '•
Hence, control of runoff pollution can be a. viable alternative for maintaining
receiving water quality standards. However, documented case studies of urban
runoff impacts are scarce.
The SCSP has had only partial success in finding urban storm flow
generated receiving water impacts employing the conventional dissolved oxygen
(DO) concentration criteria' ''. The problem appears to be in the
application of conventional dry-weather monitoring techniques to the unsteady-
state flow regimes caused by storms. Further studies should be undertaken to
resolve the anomaly between actual runoff loadings and observed receiving
water impacts. The SCSP has been more successful in sediment analysis than in
water column analysis for DO depletions. Direct evidence has been obtained
(from the Milwaukee River Project' ') of how a disturbed benthos depletes DO
from the overyling waters. Studies have also shown that storm and CSO
K-VII-4
-------
adversely affect sediment by toxics enrichment and resultant biological
upsets^* »"~'. Since particulate matter in untreated storm and CSO is
larger, heavier, and in significant quantities when compared to treated
sanitary effluent, more needs to be known about the fate and transport of
settleable and separable materials. Hydrodynamic solids separation and
sediment transport routines must be added to receiving water models to take
care of the neglected or presently omitted significant particulate and bed
flow fields.
A simplified continuous receiving water quality model has been developed
for preliminary planning and screening of areawide wastewater treatment
alternatives in terms of frequency of water quality violations' •'. Also,
a general methodology has been developed for evaluating the impact of CSOs on
receiving waters and for determining the abatement costs for achieving various
water quality goals' '. In addition, a methodology for defining criteria for
wet-weather quality standards has been developed' !""/. in recognition of an
important gap in the developed methodologies, the duration of water quality
standards vs. species survival was taken into consideration.
USER'S ASSISTANCE TOOLS
User's assistance tools include instrumentation, stormwater management
models, manuals of practice (MOP), methodologies, compendiums, and state-of-
the-art (SOTA) reports.
Instrumentation
Storm flow measurement is essential for process planning, design,
control, evaluation, and enforcement. Sampling devices do not provide
representative aliquots. Conventional flowmeters apply to steady-state flows
and not to the highly varying storm flows.
Flowmeters have been developed to overcome these adverse storm condi-
tions^ ' '. A prototype sampler for capturing representative solids in
storm flow has also been developed and a design manual is available^ '. This
gave manufacturers the incentive to perfect samplers by increasing intake
velocities and in other ways. SOTAs are available for flow measurement and
K-VII-5
-------
sampling'z'»z°). Because storm flow conditions are extremely adverse, the
manuals and instruments developed are useful for monitoring all types of flow.
Simulation Models
The SCSP has fostered the development of models for assessment, planning,
design, and control of urban stormwater pollution. The program's urban water
management analysis approach involves four levels of evaluation ranging from
simple to complex that can be worked together.
The various levels of the Storm Water Management Model (SWMM) are the
most significant model products in terms of past SCSP resources, user accep-
tance, and overall popularity. SWMM is one of the most widely used urban
models and its benefits for planning and design have been demonstrated. It
has been employed by consulting engineers to design sewers and to analyze
pollution control alternatives.
There are significant recent enhancements of SWMM. Probably the most
significant is Version IIl'"»J1' which includes: better documentation and
improved stability of the dynamic flow routing model, EXTRAN (Extended
Transport Block); documentation of continuous simulation and snowmelt
capabilities; new flexible physically-based storage and treatment routines
which can provide estimates of treatment (by settling) in storage basins; an
addition of the physically-based Green Ampt infiltration model; and flexible
water quality routines and other routines.
Documentation and user's manuals are also available for three continuous
stormwater planning models. One is a detailed model, the Quantity-Quality
(32 33}
Simulation model (QQS)V ' . The other two are macroscopic planning models,
ABMAC<34»35> and EPAMAC<36'37>.
Operational models which have been implemented in Detroit^ ',
Minneapolis' ', Seattle' ', and San Francisco produce control decisions
during storm events.
SOLUTION METHODOLOGY
A major emphasis of the SCSP was solution methodology, implemented
through development of SOTA reports, MOPs, and user manuals.
K-VII-6
-------
The SOTA texts, user's guide and the assessment of urban stormwater
technology are recommended documents^ ' '. Separate engineering manuals are
available for storm flowrate determination^ »', porous pavement
design^5~^8\ cost estimating^9"51 \ storm sewer design^52"5^, planning and
design guidance' ', and for conducting stormwater studies' '. Seminar
proceedings with themes of "modeling, design, operation, and costs" have been
published.
The SOTA document on particle size and settling velocity^ ' offers
significant information for solids treatability and their settlement in
receiving waters, important areas overlooked in planning and design. An
excellent film is being distributed by the General Services Administration
(GSA), National Audio Visual Center which covers the EPA CSO Research,
Development, and Demonstration Program, and in particular full-scale control
technologies' '.
A report entitled, "Urban Stormwater Management and Technology: Case
Histories,"' ' presents 12 case histories which represent the most promising
approaches to CSO and stormwater control. The case histories were developed
by evaluating operational facilties that have significant information for
future guidance.
Three illustrative methodologies for conducting stormwater facility
planning have been published' '.
MANAGEMENT ALTERNATIVES
The next major SCSP area is management alternatives. First the choice of
where to attack the problem; at the source by land management, in the
collection system, or off-line by storage. Pollutants can be removed by
treatment and by employing integrated systems combining control and treatment.
Land Management
Land management includes structural, semi-structural, and non-structural
measures for reducing urban and construction site stormwater runoff and
pollutants before they enter the downstream drainage system. Various concepts
have been fostered by the SCSP including:
K-VII-7
-------
. Land use planning;
. Natural drainage which will reduce drainage costs and pollution, and
enhance aesthetics, groundwater supplies, and flood protection; and
• Retention and drainage facilities, and other management techniques
required for flood and erosion control which can be simultaneously
designed for pollution control.
Retention on-site or upstream can provide for the multiple benefits of
aesthetics, recreation, recharge, irrigation, or other uses.
Porous Pavement
Porous pavements provide storage, enhancing soil infiltration that can be
used to reduce runoff and CSO. Porous asphalt-concrete pavements can be
underlain by a gravel base course with whatever storage capacity is desired
(Figure 1).
Results from a study in Rochester, New York, indicated that peak runoff
rates were reduced as much as 83%' ' where porous pavement was used. The
structural integrity of the porous pavement was not impaired by heavy load
vehicles. Clogging did result from sediment from adjacent land areas during
construction; however, it was relieved by cleaning (flushing). The construe-
tion cost of a porous pavement parking lot is about equal to that of a
conventional lot with stormwater inlets and subsurface piping.
A project in Austin, Texas developed design criteria for porous
pavements^ ' and compared porous asphalt pavement to six other conventional
and experimental pavements' »'.
Surface Sanitation
Maintaining and cleaning urban areas can have a significant impact on the
quantity of pollutants washed off by stormwater. Tests under real work
conditions in San Jose, CA showed that street cleaning can remove up to 50% of
the total solids and heavy metal yields in urban stormwater with once or twice
a day cleaning' '. Typical street cleaning programs of once or twice a month
proved ineffective. Organics and nutrients could not be effectively
controlled even with intensive cleaning.
In Bellevue, WA, street cleaning proved ineffective; however, a modified
regenerative air Tymco* street cleaner showed promise' ' '. Street cleaning
K-VII-8
-------
POROUS ASPHALT COURSE
FILTER COURSE
RESERVOIR COURSE
VOLUME DESIGNED FOR RUNOFF
DETENTION AND FROST PENETRATION-
EXISTING SOIL
MINIMAL COMPACTION TO RETAIN
€ POROSITY AND PERMEABILITY
Figure 1. Porous Asphalt Paving Typical Section
K-VII-9
-------
is no panacea for stormwater pollution control (and is site specific dependent
upon climatic conditions), but if integrated with other methods, could reduce
city-wide costs.
Chemical Use Control
Until assessment in 1971^ ', there had been limited research on highway
/£Q \
deicing effects. A search to define alternatives*- ' confirmed the need for
an economic impact analysis and for identifying a hydrophobic substance to
reduce ice adhesion. The economic analysis has been conducted^ ' and
hydrophobic substances have been identified and investigated' » '. Even
though their costs appear greater than salt, when considering an estimated $3
billion annual damage to the environment, the costs are acceptable.
The 1973 assessment' ' identified the problems of sloppy salt storage
practices and over application, resulting in manuals of practice' *' for
improvement in these areas. These manuals were recognized as highly
significant. The Federal Highway Administration reprinted them, and over
8,000 copies have been distributed.
SCSP work has prompted several states to enact legislation controlling
salt application and storage. Also the SCSP's expert testimony helped
Wisconsin to change its laws to contain controls for roadway deicing. And
lastly, Program work has greatly assisted the Federal Highway Administration
and the Army's Cold Climate Research Laboratory who are utilizing SCSP results
and doing further work with hydrophobic substances.
Collection System Controls
The next overall SCSP category, collection system controls, pertains to
management alternatives for stormwater interception and transport. These
include: improved maintenance and design of catchbasins, sewers, in-pipe and
in-channel storage, elimination of sanitary and industrial cross connections,
and remote flow monitoring and control.
Catchbasins
In a project conducted in Boston, MA, catchbasins were shown to be
potentially quite effective for solids reduction (60-97%)^75>76\ Removals of
associated pollutants such as chemical oxygen demand (COD) and BOD, were also
K-VII-10
-------
significant (10-56% and 54-88%, respectively). To maintain the effectiveness
of catchbasins for pollutant removal requires cleaning probably twice a year
depending upon conditions.
Regulators and Hydrobrakes
At present, there is a strong need to develop and have a reserve of
control hardware for urban runoff control and to effectively reduce the
associated high cost implications for conventional storage tanks, etc. The
dual functioning swirl flow regulator/solids concentrator has shown
outstanding potential for simultaneous quality and quantity control^ '. A
helical type regulator/separator has also been developed based on principles
similar to the swirl. It is felt that the swirl/helical type regulators,
previously applied only to CSO, can also be installed on separate storm drains
before discharge and the resultant concentrate flow can be stored in
relatively small tanks, since concentrate flow is only a few percent of the
total flow. Stored concentrate can later be directed to the sanitary sewer
for subsequent treatment during low-flow or dry-weather periods, or if.
capacity is available in the sanitary interceptor treatment system, the
concentrate may be diverted to it without storage' '.
These methods of stormwater control may be more economical than building
huge holding reservoirs for untreated runoff, and offer a feasible approach to
the treatment of separately sewered urban stormwater.
A completed demonstration project in Boston, MA evaluated the swirl and
helical bend for separate stormwater treatment. A complete swirl/helical
design textbook has been published' '.
Hydrobrakes™ have been demonstrated by the SCSP in Rochester, NY.'*'3) and
Cleveland, OH' > '. Hydrobrakes have no moving parts, require no external
energy source to operate and are not susceptible to blockages. They can be
used as upstream off-line attenuators, and in-line flow back-up devices.
Conceptually, they allow flow release without orifice movement or closure at a
constant predetermined rate regardless of upstream head.
New Concepts for Stormwater Control
Research efforts have shown that sanitary and industrial contamination of
separate storm sewers is a nationwide problem.
K-VII-11
-------
The SCSP has fostered a new concept: in response to this problem which
includes simple methods of checking for cross-connections. If found, they can
be eliminated or the storm sewer can be dealt with as though it were &
combined sewer whichever is more feasible. Other new concepts are using storm
sewers and storm drainage systems for in-channel storage applying CSO in-sewer
storage and routing technology and other storage facilities for ties into the
existing sewage treatment system. Thus making better use of existing
facilities for overall pollution control.
Storage
Because of the high volume and variability associated with stormwater,
storage is considered a necessary control alternative. It is the SCSP's best
documented abatement measure. Project results and theory indicate storage
must be considered at all times in system planning, because it allows for
maximum use of existing dry-weather and other treatment plant facilities and
results in the lowest cost in terms of pollutant removal.
Storage facilities may have auxiliary functions, such as sedimentation
treatment, flood protection, sewer relief, flow transmission, and dry-weather
flow equalization.
Storage concepts investigated include the conventional concrete holding
tanks and earthen basins, and the minimum land requirement concepts of:
tunnels, underground and underwater containers, underground "silos," natural
and mined under and above ground formations, and the use of abandoned
facilities and existing sewer lines^82'83^.
The in-receiving water flow balance method (Figure 2), is a recently
developed storage alternative' '. In-receiving water storage facilities
contain stormwater between floating plastic curtains. After cessation of the
overflow, pumps start automatically and the surrounding waterbody will enter
the compartments and push the stormwater back towards the first compartment
where it is pumped to the plant. Thus,, the waterbody is used as a flow
balance medium.
The storage method is low cost due to the employment of low cost
materials (plastic and wood), the time required to install the unit (several
days to months vs. months to a year for more structurally intensive
K-VII-12
-------
Figure 2. In-Receiving Water Flow — Balance Method
K-VII-13
-------
alternatives), and the absence of land requirements. Studies show that costs
could be about 5 to 15% of conventional concrete tank costs.
The facility which was tested at three locations in Sweden performed very
satisfactorily, and was able to take ice and wind loads without adverse
impact. It is desirable to demonstrate a. facility in a harsh urban estuarlne
or marine site, such as the SCSP project will be doing shortly in Fresh Creek
Basin in New York City, NY.
A storage/sedimentation planning/design manual' ' is being finalized
which includes the following:
Major Issues
Applicability,
Integration into existing treatment systems,
Costs,
Effectiveness as treatment/control device.
Secondary Issues
Citing impacts,
O&M requirements,
Sludge production and handling.
Treatment
Due to adverse and intense flow conditions and unpredictable shock
loading effects, it has been difficult to adapt existing treatment methods to
storm-generated overflows, especially the microorganism dependent biological
processes. Physical/chemical treatment techniques have shown more promise
than biological processes in overcoming storm shock loading effects. To
reduce capital investments, projects have been directed toward high-rate
operations approaching maximum loading.
Storm-flow treatment methods demonstrated by the Program include
physical, physical-chemical, wetlands, biological, and disinfection' '.
These processes, or combinations of these processes, can be adjuncts to the
existing sanitary plant or serve as remote satellite facilities at the
outfall.
Physical/Chemical Treatment
Physical processes/systems with or without chemicals, such as: fine mesh
screening, swirl degritting, fine mesh screening/high-rate filtration (HRP),
K-VII-14
-------
sedimentation, and fine mesh screening/dissolved air flotation (DAF), have
been successfully demonstrated. Physical processes have shown importance for
storm-flow treatment because they are adaptable to automated operation, rapid
startup and shutdown, high-rate operation, and resistance to shock loads. The
high-rate processes/systems (DAF, fine mesh screening, and HRF) are felt to be
ready for municipal installation.
The swirl has also been developed for grit removal. The small size, high
efficiency and absence of moving parts offer economical and operational
advantages over conventional degritting facilities. A full-scale demonstra-
tion of a (16 ft diameter/11 mgd design flowrate) swirl degritter has been
completed in Tamworth, Australia' *'• Removal efficiencies confirmed
laboratory results. Compared with a conventional grit chamber, construction
costs are halved, and operation and maintenance costs are considerably lower.
Disinfection
Because disinfectant and contact demands are great for storm flows' ',
research has centered on high-rate applications by static and mechanical
mixing, higher disinfectant concentrations' ', and more rapid oxidants,
i.e., chlorine dioxide^91'9^, ozone^90\ and ultraviolet (UV) light; and on-
site generation' »»'. Demonstrations in Rochester, NY' '; Syracuse,
. Eagt Chicag0) iN; and Philadelphia, PA^93\ indicate that adequate
reductions of fecal coliform can be obtained with contact times of 2 minutes
or less by induced mixing and dosing with chlorine and/or chlorine dioxide. A
pilot scale UV demonstration with a contact time of less than 10 seconds was
conducted in New York City, NY.
The hypochlorite batching facility is still being used in New Orleans, LA
to protect swimming beaches in Lake Ponchartrain^ ' . The SCSP supported the
development of a brine hypochlorite generator now being used in industry' ' .
Treatment/Control Design Guidebook
A compilation of the SCSP's best research efforts in CSO treatment /con-
(Qf.\
trol over the past 18 years has been published'' '.
Treatment Process Performance
Treatment process costs and performance in terms of design influx rate
(gpm/ft2) and BOD5 and SS removal efficiency is provided in Table 2. The
K-VII-15
-------
high-rate performance of the swirl, and the microstrainer, screening/HRF and
screening/DAF systems, is apparent when compared to sedimentation.
Table 2. Wet Weather Treatment Plant Performance Data
Device
Control Alternatives
Design
Loading Rate
(gpm/ft2)
Removal Efficiency (%)
BO DC
SS
Primary
Secondary
Swirl concentrator
Microstrainer
High-rate filtration
Dissolved air flotation
Sedimentation
Representative Performance
Contact stabilization
Physical-chemical
Representative Performance
60
20
24
2.5
0.5
25-60
40-60
60-80
50-60
25-40
40
75-88
85-95
85
50
70
90
80
55
60
90
95
95
Sludge-Solids
Another SCSP area is the sludge and solids associated with storm flow
treatment. Sludge handling and disposal must be considered an integral part
of treatment because it significantly affects the efficiency and cost of the
total waste treatment systems. A study quantified solids residue and sludge
that would result from the treatment of separate urban stormwater on a
nationwide basis(98»99).
INTEGRATED SYSTEMS
The most promising and common approach to urban storm flow management
involves the integration of control and treatment. Integrated systems are
divided into storage/treatment, dual-use wet-weather flow/dry-weather flow
facilities, and control/treatment/reuse.
Storage/Treatment
When there is storage, there is treatment by settling, pump-back/bleed-
back to municipal works, and sometimes disinfection. In any case, the break-
K-VII-16
-------
even economics of supplying storage must be evaluated when treatment is
considered. The SCSP has demonstrated all of these storage/treatment concepts
at full scale.
Dual-Use Wet Weather Flow/Dry Weather Flow Facilities
The concept of dual use is — maximum utilization of wet-weather facilities
during non-storm periods and maximum utilization of dry-weather facilities
during storm flows. The SCSP has demonstrated the full-scale dual-use of
high-rate trickling f ilters^100), contact stabilization(101~103),
and equalization basins^ . Various municipalities are employing dual-use
microscreening.
Control/Treatment/Reuse
"Control/Treatment/Reuse" is a "catch-all" for all integrated systems. A
prime consideration should be the various nonstructural and land-management
techniques. In Mt. Clemens, MI, a series of three "lakelets" have been
incorporated into a CSO treatment /park development' '. Treatment is being
provided so that these lakes are aesthetically pleasing and allow for
recreation and reuse for irrigation.
RECOMMENDATIONS FOR THE FUTURE
Receiving Water Impacts
Ties between receiving water quality and storm flow discharges must be
clearly established and delineated. Quantification of the impairment of
beneficial uses and water quality by such discharges is a major goal. Project
results indicate the potential for significant impact of wet-weather flows on
receiving waters. Control of runoff pollution can be a viable alternative for
maintaining receiving water quality standards. However, the problems found
seem to be site-specific in nature. Therefore, site-specific surveys are
required that must consider the effects of larger materials and floatables
near the outfalls. Based on results from these surveys, control may be
warranted.
K-VII-17
-------
Toxics Characterization and Control/Treatment
Results from a limited in-house effort and the Nationwide Urban Runoff
Program (NURF) indicate that urban stormwater runoff and CSO contains
significant quantities of priority pollutants. Additional investigation of
the significance of concentrations and quantities of toxic pollutants with
regard to their health effects oi: potential health effects and ecosystem
effects is required. A need exists to evaluate the removal capacity of
alternative treatment technologies for these toxics and to compare their
effectiveness with estimated removal needs to meet water quality goals.
Sewer System Cross Connections
Investigations have shown that sanitary and industrial contamination of
separate storm sewers is an extensive nationwide problem. In practice, a
significant number of separate stormwater drainage systems actually function
as combined sewer systems. Therefore, a nationwide effort on both federal and
local levels, to alleviate the pollution impacts from discharges of these
systems is required. It is better to classify such cross-connected drainage
systems as combined systems for pollution control priorities.
Integrated Stormwater Management
The most effective solution methodology for wet-weather pollution
problems must consider: (1) control of wet-weather pollution impacts in lieu
of continually and blindly upgrading existing municipal plants, (2) structural
vs. non-structural techniques, (3) integrating dry- and wet-weather flow
systems to make maximum use of the existing drainage during wet conditions and
maximum use of wet-weather control/treatment facilities during dry-weather,
and (4) the segment or blend on the percent pollutant control vs. cost curve
in which incremental costs exceed incremental control efficiencies subject to
load discharge or receiving water requirements.
Flood and erosion control technology must be integrated with pollution
control, so that the retention and drainage facilities required for flood and
erosion control can be simultaneously designed or retrofitted for pollution
control. If land management and non-structural techniques are maximized and
integrated costs for the extraction of pollutants from storm flows in the
potentially more costly downstream plants will be reduced.
K-VII-18
-------
Institutional and Socio/Economic Conflicts
Some of the most promising opportunities for cost-effective environmental
control are multipurpose in nature. However, there are institutional problems
that hinder their implementation. First, the autonomous Federal and local
agencies and professions involved in flood and erosion control, pollution
control, and land management and environmental planning must be integrated at
both the planning and operation levels. Multi-agency grant coverage must be
adequate to stimulate such an approach. For example, the EPA would have to
join with the Corps of Engineers, Soil Conservation Service, Department of
Transportation, and perhaps other Federal agencies as well as departments of
pollution control, sanitation, planning, and flood control at the local level.
Another problem is that construction grant (and other) incentives are
geared towards structurally intensive projects which may counter research
findings in the area of optimal solutions. Optimized wet-weather pollution
involves a city-wide approach including the integration of structural as well
as low-structural controls. The low-structural measures are more labor
intensive. Construction grant funding does not presently address this expense
and accordingly, municipalities are discouraged from using them.
CONCLUSIONS
In general, on a mass basis, toxic compounds, oxygen demanding, and
suspended and visual matter in urban stormwater are significant. Ignoring the
problem because it appears to be too costly or too difficult to solve, will
not make the problem go away. The integrated approach to wet-weather
pollution control is the only way which is going to be feasible, economical
and, therefore, acceptable. Potentially tremendous savings can be derived
from wet-weather pollution control research fostering integrated solutions.
As you can see, the SCSP has investigated a problem, proven its significance,
and developed a gamut of design and control techniques. Better advantage
needs to be taken of proven technology.
K-VII-19
-------
REFERENCES
1. Lager, J. A., and W. G. Smith, Metcalf & Eddy Engineers, Inc., Palo Alto,
CA - "Urban Stormwater Management and Technology: An Assessment," USEPA
Report No. EPA-670/2-74-040, NTIS No. PB 240 687.
2. Lager, J. A., et al., Metcalf & Eddy Engineers, Inc., Palo Alto, CA -
"Urban Stormwater Management and Technology: Update and User's Guide,"
USEPA Report EPA-600/8-77-014, NTIS No. PB 275 654.
3. Sullivan, R. H., et al., American Public Works Assoc., Chicago, IL -
"Nationwide Evaluation of Combined Sewer Overflows and Urban Stormwater
Discharges, Vol. Ill - Characterization of Discharges," USEPA Report No.
EPA-600/2-77-064c, NTIS No. PB 272 107.
4. Huber, W. C., et al., University of Florida, Gainesville, FL - "Urban
Rainfall-Runoff-Quality Data Base," USEPA Report No. EPA-600/2-81-238,
NTIS No. PB 82-221 094.
5. Huber, W. C., et al., University of Florida, Gainesville, FL - "Urban
Rainfall-Runoff-Quality Data Base," USEPA Project Summary No. EPA-600/S2-
81-238.
6. Spiegel, S. J., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY, and
Ott, R., Department of Drainage and Sanitation, Onondaga County, North
Syracuse, NY - "Evaluation of Urban Runoff and Combined Sewer Overflow
Mutagenicity," USEPA Report No. EPA-600/2-84-116, NTIS No. PB 84-211168.
7. Spiegel, S. J., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY, and
Ott, R., Department of Drainage and Sanitation, Onondaga County, North
Syracuse, NY - "Evaluation of Urban Runoff and Combined Sewer Overflow
Mutagenicity," USEPA Project Summary No. EPA-600/S2-84-116.
8. Olivieri, V. P., et al., The Johns Hopkins University, Baltimore, MD -
"Microorganisms in Urban Stormwater," USEPA Report No. EPA-600/2-77-087,
NTIS No. PB 272 245.
9. Proceedings of National Conference, Orlando, FL, November 26-28, 1979,
"Urban Stormwater and Combined Sewer Overflow Impact on Receiving Water
Bodies," USEPA Report No. EPA-600/9-80-056, NTIS No. PB 81-155 426.
10. Field, R., and Turkeltaub, R., USEPA, Edison, NJ - "Urban Runoff
Receiving Water Impacts: Program Overview," Journal of the Environmental
Engineering Division, ASCE, Vol. 107, No. EE1, Feb., 1981, pp. 83-100.
11. Sullivan, R. H., et al., American Public Works Assoc., Chicago, IL. -
"Nationwide Evaluation of Combined Sewer Overflows and Urban Stormwater
Discharges, Volume I - Executive Summary," USEPA Report No. EPA-600/2-77-
064a, NTIS No. PB 273 133.
K-VII-20
-------
12. Heaney, J. F., et al., University of Florida, Gainesville, FL -
"Nationwide Evaluation of Combined Sewer Overflows and Urban Stormwater
Discharges, Volume II - Cost Assessment and Impacts," USEPA Report No.
EPA-600/2-77-064b, NTIS No. PB 266 005.
13. Reefer, N., et al., The Sutron Corporation, Arlington, VA - "Dissolved
Oxygen Impact from Urban Storm Runoff," USEPA Report No. EPA-600/2-79-
156, No NTIS number.
14. Melnholz, T. L., et al., (Rexnord) Metropolitan Sewage District of County
of Milwaukee, WI - "Verification of the Water Quality Impacts of Combined
Sewer Overflows," USEPA Report No. EPA-600/2-79-155.
15. Pitt, R., Private Consultant, Blue Mounds, WI; Bozeman, M., Woodward-
Clyde Consultants, San Francisco, CA - "Sources of Urban Runoff Pollution
and Its Effects on an Urban Creek," USEPA Report No. EPA-600/2-82-090,
NTIS No. PB 83-111 021.
16. Pitt, R., Private Consultant, Blue Mounds, WI; Bozeman, M., Woodward-
Clyde Consultants, San Francisco, CA - "Sources of Urban Runoff Pollution
and Its Effects on an Urban Creek," USEPA Project Summary No. EPA-600/S2-
82-090.
17. Tomlinson, R. D., et al., Municipality of Metropolitan Seattle, Seattle,
WA - "Fate and Effects of Partlculates Discharges by Combined Sewers and
Storm Drains," USEPA Report No. EPA-600/2-80-111, NTIS No. PB 81-118 390.
18. Medina, M., Duke University, Durham, NC - "Level III: Receiving Water
Quality Modeling for Urban Stormwater Management," USEPA Report No. EPA-
600/2-79-100, NTIS No. PB 80-134 406.
19. Medina, M. A., et al., Duke University, Durham, NC - "River Quality Model
for Urban Stormwater Impacts," USEPA Journal No. EPA-600/J-81-234, NTIS
No. PB 81-221 087.
20. Moffa, P. E., et al., Stearns & Wheler, Civil and Sanitary Engineers,
Cazenovia, NY "Methodology for Evaluating the Impact and Abatement of
Combined Sewer Overflows: A Case Study of Onondaga Lake, New York,"
USEPA Report No. EPA-600/8-80-048, NTIS No. PB 81-141 913.
21. Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "Development of
Methods to Define Water Quality Effects of Urban Runoff," USEPA Project
Summary No. EPA-600/S2-83-125.
22. Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "Development of
Methods to Define Water Quality Effects of Urban Runoff," USEPA Report
No. EPA-600/2-83-125, NTIS No. PB 84-122 928.
23. Mancini, J. L., Mancini & DiToro Assoc., Valley, NB - "A Method for
Calculating Effects on Aquatic Organisms, of Time Varying Concentra-
tions," Water Research, Vol. 17, No. 10, pp. 1355-1362, 1983.
K-VII-21
-------
24. Foreman, K. M., Grumman Aerospace Corp., Bethpage, NY - "Field Testing of
Prototype Accoustic Flowmeter," USEPA Report No. EPA-600/2-79-084, NTIS
No. PB 80-121 544.
25. Anderson, R. J., and Bell, S. S., City of Milwaukee, WI - "Wastewater
Flow Measurement in Sewers Using Ultrasound," USEPA Report No. EPA-600/2-
76-243, NTIS No. PB 262 902.
26. Shelley, P. E., EG&G Washington Analytical Services Center, Inc.,
Rockville, MD - "Design and Testing of a Prototype Automatic Sewer
Sampling System," USEPA Report No. EPA-600/2-76-006, NTIS No. PB 252 613.
27. Shelley, P. E., and Kirkpatrick, G. A., EG&G Washington Analytical
Services Center, Inc., Rockville, MD - "Sewer Flow Measurement - A State-
of-the Art Assessment," USEPA Report No. EPA-600/2-75-027, NTIS No. PB
250 371.
28. Shelley, P. E., and Kirkpatrick, G. A., EG&G Washington Analytical
Services Center, Inc., Rockville, MD - "An Assessment of Automatic Sewer
Flow Samplers - 1975," USEPA Report No. EPA-600/2-75-065, NTIS No. PB 250
987.
29. Huber, W. C., et al., University of Florida, Gainesville, FL - "Storm
Water Management Model User's Manual, Version III," USEPA Report Nos.
EPA-600/2-84-109a, NTIS No. PB 84-198-423 and EPA-600/2-84-109b, NTIS No.
PR 84-198-431 (EPA-600/2-84-109b is the "Addendum EXTRAN").
30. Huber, W. C., et al., University of Florida, Gainesville, FL - "Storm
Water Management Model User's Manual, Version III," USEPA Project Summary
No. EPA-600/S2-84-109a & b.
31. Roesner, L. A., et al., Camp Dresser & McKee Inc., Annadale, VA - "Storm
Water Management Model User's Manual Version III, Addendum I - EXTRAN,"
USEPA Report No. Pending (Interim availability: University of Florida).
32. Geiger, W. F., and Dorsch, H. R., Dorsch Consult Ltd., Toronto, Ontario,
Canada - "Quantity-Quality Simulation (QQS): A Detailed Continuous
Planning Model for Urban Runoff Control - Volume 1 Model Description,
Testing, and Applications," USEPA Report No. EPA-600/2-80-011, NTIS No.
PB 80-190 507.
33. Geiger, W. F., and Dorsch, H. R., Dorsch Consult Ltd., Toronto, Ontario,
Canada - "Quantity-Quality Simulation (QQS): A Detailed Continuous
Planning Model for Urban Runoff Control - Volume 2 User's Manual," USEPA
Report No. EPA-600/2-80-016, NTIS No. PB 80-221 872.
34. Litwin, Y. J., et al., Ramlit Assoc., Berkeley, CA - "Areawide Stormwater
Pollution Analysis with the Macroscopic Planning (ABMAC) Model," USEPA
Project Summary No. EPA-600/S2-81-223.
K-VII-22
-------
35. Litwin, Y. J., et al., Ramlit Assoc., Berkeley, CA - "Areawide Stormwater
Pollution Analysis with the Macroscopic Planning (ABMAC) Model," USEPA
Report No. EPA-600/2-81-223, NTIS No. PB 82-107 947.
36. Smith, W. G., and Strickfaden, M. E., Metcalf & Eddy Engineers, Inc.,
Palo Alto, CA - "EPA Macroscopic Planning Model (EPAMAC) for Stormwater
and Combined Sewer Overflow Control: Application Guide and User's
Manual," USEPA Report No. EPA-600/2-83-086, NTIS No. PB 83-259 689.
37. Smith, W. G., and Strickfaden, M. E., Metcalf & Eddy Engineers, Inc.,
Palo Alto, CA - "EPA Macroscopic Planning Model (EPAMAC) for Stormwater
and Combined Sewer Overflow Control: Application Guide and User's
Manual," USEPA Project Summary No. EPA-600/S2-83-086.
38. Watt, T. R., et al., Detroit Metro Water Department, Detroit, MI -
"Sewerage System Monitoring and Remote Control," USEPA Report No. EPA-
670/2-75-020, NTIS No. PB 242 107.
39. Metropolitan Sewer Board, St. Paul, MN - "Dispatching Systems for Control
of Combined Sewer Losses," USEPA Report No. 11020FAQ03/71, NTIS No. PB
203 678.
40. Leiser, C. P., Municipality of Metropolitan Seattle, Seattle, WA -
"Computer Management of a Combined Sewer System," USEPA Report No. EPA-
670/2-74-022, NTIS No. PB 235 717.
41. Lager, J. A., et al., Metcalf & Eddy, Inc., Palo Alto, CA - "Urban
Stormwater Management and Technology: Update and User's Guide," USEPA
Report No. EPA-600/8-77-014, NTIS No. PB 275 654.
42. Lager, J. A., and Smith, W. G., Metcalf & Eddy Engineers, Inc., Palo
Alto, CA - "Urban Stormwater Management and Technology: An Assessment,"
USEPA Report No. EPA-670/2-74-040, NTIS No. PB 240 687.
43. Brater, E. P., and Sherrill, J. D., University of Michigan, Ann Arbor, MI
- "Rainfall-Runoff Relations on Urban and Rural Areas," USEPA Report No.
EPA-670/2-75-046, NTIS No. PB 242 830.
44. Yen, B. C., and Chow, V. T., University of Illinois, Urbana, IL - "Urban
Stormwater and Runoff Determination of Volumes and Flowrates," USEPA
Report No. EPA-600/2-76-116, NTIS No. PB 253 410.
45. Field, R., et al., USEPA, Storm and Combined Sewer Program, Edison, NJ -
"Porous Pavement: Research; Development; and Demonstration,"
Tranportation Engineering Journal of ASCE, Vol. 108, No. TE3, May 1982,
pp. 244-258.
46. Diniz, E., Espey, Huston & Assoc., Inc. Albuquerque, NM - "Porous
Pavement: Phase I - Design and Operational Criteria," USEPA Report No.
EPA-600/2-80-135, NTIS No. PB 81-104 796.
K-VII-23
-------
47. Goforth, G. F., Espey, Huston & Assoc., Inc., Austin, TX; Diniz, E. V.,
Resource Technology, Inc., Albuquerque, NM; and Rauhut, J. B., Austin, TX
- "Stormwater Hydrological Characteristics of Porous and Conventional
Paving Systems," USEPA Report No. EPA-600/2-83-106, NTIS No. PB 84-123
728.
48. Goforth, G. F., Espey, Huston & Assoc., Inc., Austin, TX; Diniz, E. V.,
Resource Technology, Inc., Albequerque, NM; and Rauhut, J. B., Austin, TX
- "Stormwater Hydrological Characteristics of Porous and Conventional
Paving Systems," USEPA Project Summary No. EPA-600/S2-83-106.
49. Benjes, H. H., Gulp, Wesner, Gulp Consulting Engineers, El Dorado, CA -
"Cost Estimating Manual-Combined Sewer Overflow Storage Treatment," USEPA
Report No. EPA-600/2-76-286, NTIS No. PB 266 359.
50. Benjes, H. H., Jr., Gulp, Wesner, Gulp Consulting Engineers, El Dorado,
CA - Field, R., Storm and Combined Sewer Program, Edison, NJ - "Estimate
Sewer Overflow Facility Costs," Water & Wastes Engineering, September
1978, pp. 56-62.
51. Heaney, J. F., et al., University of Florida, Gainesville, FL -
"Nationwide Evaluation of Combined Sewer Overflows and Urban Stormwater
Discharges," Volume II - Cost Assessment and Impacts," USEPA Report No.
EPA-600/2-77-064b, NTIS No. PB 266 005.
52. Sonnen, M., Water Research Engineers, Walnut Creek, CA - "Abatement of
Deposition and Scour in Sewers," USEPA Report No. EPA-600/2-77-212, NTIS
No. PB 276 585.
53. Yen, B. C., et al., University of Illinois, Urbana, IL - "Stormwater
Runoff on Urban Areas of Steep Slope," USEPA Report No. EPA-600/2-77-168,
NTIS No. PB 272 755.
54. Kaufman, H. L., and Lai, F. H., Clinton Bogert Assoc., Fort Lee, NJ -
"Conventional and Advanced Sewer Design Concepts for Dual Purpose Flood
and Pollution Control - A Preliminary Case Study, Elizabeth, NJ," USEPA
Report No. EPA-600/2-78-090, NTIS No. PB 285 663.
55. Field, R., and Weisman, D. A., USEPA Storm and Combined Sewer Program,
Edison, NJ - "A Planning and Design Guidebook for Combined Sewer Overflow
Control," USEPA Report No. EPA-600/2-82-084, NTIS No. PB 82-259 235.
56. Wuelschleger, R. E., et al., Envirex Environmental Sciences Div.,
Milwaukee, WI - "Methdology for the Study of Urban Storm Generated
Pollution and Control," USEPA Report No. EPA-600/2-76-145, NTIS No.'PB
258 243.
57. Dalrymple, R. J., et al., Beak Consultants for American Public Works
Assoc., Chicago, IL - "Physical and Settling Characteristics of
Particulates in Storm and Sanitary Wastewater," USEPA Report No. EPA-
670/2-75-011, NTIS No. PB 242 001.
K-VII-24
-------
58. Field, R., Storm and Combined Sewer Program, USEPA, Edison, NJ -
"Stormwater Pollution Control: A New Technology," 28 Minute Film
distributed by the General Services Administration (GSA), National Audio
Visual Center, Washington, B.C. 20409.
59. Lynard, W., et al., Metcalf & Eddy Engineers, Inc., Palo Alto, CA -
"Urban Stormwater Management and Technology: Case Histories," USEPA
Report No. EPA-600/8-80-035, NTIS No. PB 81-107 153.
60. Heaney, J. F., et al., University of Florida, Gainesville, FL -
"Stormwater Management Model: Level I—Preliminary Screening
Procedures," USEPA Report No. EPA-600/2-76-275, NTIS No. PB 259 916.
61. Heaney, J. F., et al., University of Florida, Gainesville, FL -
"Stormwater Management Model: Level I—Comparative Evaluation of
Storage/Treatment and Other Management Practices," USEPA Report No. EPA-
600/2-77-083, NTIS No. PB 265 671.
62. Moffa, P., and Karanik, J., Onondaga County, NY - "Methodology for
Evaluating the Impact and Abatement of Combined Sewer Overflows - A Case
Study of Onondaga Lake, NY," USEPA Report No. EPA-600/8-80-048, NTIS No.
PB 81-141 913.
63. Murphy, C. B., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY;
Quinn, T. J., and Steward, J. E., Monroe County Division of Pure Waters,
Rochester, NY - "Best Management Practices: Implementation," USEPA
Report No. EPA-905/9-81-002, NTIS No. PB 82-169 210.
64. Pitt, R. E., Woodward Clyde Consultants, San Francisco, CA - "Demonstra-
tion of Nonpoint Pollution Abatement Through Improved Street Cleaning
Practices," USEPA Report No. EPA-600/2-79-161, NTIS No. PB 80-108 988.
65. Pitt, R. E., Consulting Environmental Engineer, Blue Mounds, WI -
"Characterization, Sources, and Control of Urban Runoff by Street and
Sewerage Cleaning," USEPA Draft Report for Project No. CR-805929,
January, 1984.
66. Pitt, R. E., Consulting Environmental Engineer, Blue Mounds, WI -
"Characterization, Sources, and Control of Urban Runoff by Street and
Sewerage Cleaning," USEPA Draft Project Summary for Project No.
CR-805929.
67. Struzeski, E., Storm and Combined Sewer Overflow Section, USEPA, Edison,
NJ - "Environmental Impact of Highway Deicing," USEPA Report No.
11040GKK06/71.
68. Murray, D. M., and Eigerman, M. R., Abt Associates, Inc., Cambridge, MA -
"A Search: New Technology for Pavement Snow and Ice Control," USEPA
Report No. EPA-R2-72-125.
K-VII-25
-------
69. Murray, D. M., et al., Abt Associate's, Inc., Cambridge, MA - "An Economic
Analysis of the Environmental Impact: of Highway Deicing," USEPA Report
No. EPA-600/2-76-105.
70. Ahlborn, G. H., and Poehlmann, H. C,,, Jr., Ball Brothers Research Corp.,
Boulder, CO - "Development of a Hydrophobic Substance to Mitigate
Pavement Ice Adhesion," USEPA Report No. EPA-600/2-76-242.
71. Krukar, M., and Cook, J. C., Washington State University, Pullman, WA -
"Optimization and Testing of Highway Materials to Mitigate Ice Adhesion,"
USEPA Report No. EPA-600/2-78-035.
72. Field, R., et al., Storm and Combined Sewer Technology Branch, USEPA,
Edison, NJ - "Water Pollution and Associated Effects from Street
Salting," USEPA Report No. EPA-R2-73-257.
73. Richardson, D. L., et al., Arthur D. Little, Inc., Cambridge, MA -
"Manual for Deicing Chemicals: Storage and Handling," USEPA Report No.
EPA-600/2-74-033.
74. Richardson, D. L., et al., Arthur D. Little, Inc., Cambridge, MA -
"Manual for Deicing Chemicals: Application Practices," USEPA Report No.
EPA-670/2-74-045.
75. Freeman, P. A., Peter Freeman Assoc., Inc., Berlin, MD - "Evaluation of
Fluidic Combined Sewer Regulators Under Municipal Service Conditions,"
USEPA Report No. EPA-600/2-77-071, NTIS No. PB 272 834.
76. Field, R., USEPA, Storm and Combined Sewer Program, Edison, NJ - "The
Dual-Functioning Swirl Combined Sewer Overflow Regulator/Concentrator,"
USEPA Report No. EPA-670/2-73-059, NTIS No. PB 227 182.
77. Field, R., et al., USEPA, Storm and Combined Sewer Program, Edison, NJ -
"Treatability Determinations for a Prototype Swirl Combined Sewer
Overflow Regulator/Solids-Separator," Prog. Wat. Tech. Vol, 8, No. 6, pp.
81-91. Printed in Great Britain.
78. Field, R., and Masters, H. E., USEPA, Edison, NJ - "Swirl Device for
Regulating and Treating Combined Sewer Overflow," USEPA Report No. EPA-
625/2-77-012, EPA Technology Transfer Capsule Report, CERI (Cincinnati)
2012.
79. Sullivan, R. H., et al., American Public Works Assoc., Chicago, IL -
"Swirl and Helical Bend Pollution Control Devices: Design Manual," USEPA
Report No. EPA-600/8-82-013, NTIS No. PB 82-266 172.
80. Matthews, T. M., et al., Snell Environmental Group, Akron, OH -
"Hydrobrakes Regulated Storage System for Stormwater Management," USEPA
Report No. EPA-600/2-83-097, NTIS No. PB 84-110 378.
K-VII-26
-------
81. Matthews, T. M., et al., Snell Environmental Group, Akron, OH -
"Hydrobrakes Regulated Storage System for Stormwater Management," USEPA
Project Summary No. EPA-600/S2-83-097.
82. Field, R., and Struzeski, E. J., USEPA, Storm and Combined Sewer Program,
Edison, NJ - "Management and Control of Combined Sewer Overflows,"
Journal Water Pollution Control Federation, Washington, D.C., Vol. 44,
No. 7, pp. 1393-1415, July 1972.
83. Field, R., Storm and Combined Sewer Program, Edison, NJ; Lager, J. A.,
Metcalf and Eddy, Inc., Palo Alto, CA - "Urban Runoff Pollution Control -
State-of-the-Art," Journal of the Environmental Engineering Division,
American Society of Civil Engineers, Vol. 101, No. EE1, pp. 107-125,
February 1975.
84. Soderland, H., Kjessler & Mannerstrale AB, Swedish Council for Building
Research, Stockholm, Sweden - "Flow Balancing Method for Stormwater and
Combined Sewer Overflow," ISBN 91-540-3765-4: 017:1982, pp. 1-27, 1982.
85. Smith, W. G., et al., Metcalf & Eddy Engineers, Inc., Palo Alto, CA -
"Storage/Sedimentation Facilities for Control of Storm and Combined Sewer
Overflows Design Manual," USEPA Report No. Pending, (USEPA Contract No.
68-03-2877).
86. Field, R, and Weisman, D. A., USEPA, Storm and Combined Sewer Program,
Edison, NJ - "A Planning and Design Guidebook for Combined Sewer Overflow
Control and Treatment," USEPA Report No. EPA-600/2-82-084, NTIS No. PB
82-259 235.
87. Shelley, G. J., et al., George J. Shelley Consulting Engineers, Tamworth,
New South Wales, Australia - "Field Evaluation of a Swirl Degritter at
Tamworth N.S.W., Australia," USEPA Report No. EPA-600/2-81-063, NTIS No.
PB 81-187 247.
88. Shelley, G. J., et al., George J. Shelley Consulting Engineers, Tamworth,
New South Wales, Australia - "Field Evaluation of a Swirl Degritter at
Tamworth N.S.W., Australia," USEPA Project Summary No. EPA-600/S2-81-063.
89. Field, R., et al., USEPA, Storm and Combined Sewer Program, Edison, NJ -
"Proceedings of Workshop on Microorganisms in Urban Stormwater," USEPA
Report No. EPA-600/2-76-244, NTIS No. PB 263 030.
90. Glover, G. E., and Herbert, G. R., Crane Company, King of Prussia, PA -
"Microstraining and Disinfection of Combined Sewer Overflows - Phase II,"
USEPA Report No. EPA-R2-73-124, NTIS No. PB 219 879.
91. Drehwing, F. J., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY -
"Combined Sewer Overflow Abatement Program, Rochester, NY - Volume H
Pilot Plant Evaluations," USEPA Report No. EPA-600/2-79-03lb, NTIS No. PB
80-159 262.
K-VI1-27
-------
92. Moffa, P. E., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY -
"Bench-Scale High-Rate Disinfection of Combined Sewer Overflows with
Chlorine and Chlorine Dioxide," USEPA Report No. EPA-670/2-75-021, HTIS
No. PB 242 296.
93. Maher, M. B., Crane Company, King of Prussia, PA - "Microstraining and
Disinfection of Combined Sewer Overflows - Phase III," USEPA Report No,
EPA-670/2-74-049, NTIS No. PB 235 771.
94. Drehwing, F., et al., O'Brien & Gere Engineers, Inc., Syracuse, NY -
"Disinfection/Treatment of Combined Sewer Overflows, Syracuse, New
York," USEPA Report No. EPA-600/2-79-134, NTIS No. PB 80-113 459.
95. Leitzs F. B., et al., Ionics, Inc., Watertown, MA - "Hypochlorite
Generator for Treatment of Combined Sewer Overflows," USEPA Report No.
11023DAA03/72, NTIS No. PB 211 243.
96. Pontius, U. R., et al., Pavia Byrne Engineering Corp., New Orleans, LA ~
"Hypochlorination of Polluted Stormwater Pumpage at New Orleans," USEPA
Report No. EPA-670/2-73-067, NTIS No. PB 228 581.
97. Conrick, D. J., et al., Environmental Design & Planning, Inc., Allston
(Boston) MA - "Evaluation of a Treatment Lagoon for Combined Sewer
Overflow," USEPA Report No. EPA-600/2-81-196, NTIS No. PB 82-105 214,
98. Carr, D. J., et al., Marquette University, Milwaukee, WI - "Characteris-
tics and Treatability of Urban Runoff Residues," USEPA Report No. EPA-
600/2-82-094, NTIS No. PB 83-133 561.
99. Carr, D. J., et al., Marquette University, Milwaukee, WI - "Characteris-
tics and Treatability of Urban Runoff Residues," USEPA Project Summary
No. EPA-600/S2-82-094.
100. Homack, P., et al., E. T. Killlam Assoc., Inc., Millburn, NJ -
"Utilization of Trickling Filters for Dual Treatment of Dry and Wet
Weather Flows," USEPA Report No. EPA-670/2-73-071, NTIS No. PB 231 251.
101. Benedict, A. H., and Roelfs, V. L., Whitely-Jacobsen and Assoc.,
Portland, OR - "Joint Dry-Wet Weather Treatment of Municipal Wastewater
at Clatskanie, Oregon," USEPA Report No. EPA-600/2-81-061, NTIS No. PB
81-187 262.
102. Benedict, A. H., and Roelfs, V. L., Whitely-Jacobsen and Assoc.,
Portland, OR - "Joint Dry-Wet Weather Treatment of Municipal Wastewater
at Clatskanie, Oregon," USEPA Project Summary No. EPA-600/S2-81-061.
103. Agnew, R. W., et al., Envirex, Milwaukee, WI - "Biological Treatment of
Combined Sewer Overflow at Kenosha, WI," USEPA Report No. EPA-670/2-75-
019, NTIS No. PB 242 107.
K-VII-28
-------
104. Innerfield, H., and Forndran, A., New York City Deparatment of Water
Resources, New York, NY - "Dual Process High-Rate Filtration of Raw
Sanitary Sewage and Combined Sewer Overflows," USEPA Report No. EPA-
600/2-79-015, NTIS No. PB 296 626/AS.
105. Welborn, H. L., Y-T-0 & Assoc., Walnut Creek, CA - "Surge Facility for
Wet- and Dry-Weather Flow Control," USEPA Report No. EPA-670/2-74-075,
NTIS No. PB 238 905.
106. Mahida, V. U., and DeDecker, F. J., Spalding DeDecker Assoc., Madison
Heights, MI - "Multi-Purpose Combined Sewer Overflow Treatment Facility,
Mt. Clemens, MI," USEPA Report No. EPA-670/2-75-010, NTIS No. PB 242
914.
K-VII-29
-------
NONPOINT POLLUTION ABATEMENT PROGRAMS: THE MANAGER'S DILEMMA
Frederick W. Madison
Associate Professor, Department of Soil Science,
Wisconsin Geological and Natural History Survey,
University of Wisconsin-Madison
Madison, WI 53706
ABSTRACT
The passage of PL 92-500 in 1972 set governments at all levels—state, local
and federal—on a collision course. The legislation changed national
pollution abatement policy in that it directed the pollution control efforts
be focused on treating pollution at its source rather than in receiving
waters.
The challenge of dealing with non-point sources of pollution is that it
interfaces land-use and water quality problems. In the country,
historically, the federal government has dealt with water policy whereas
responsibility for land-use decisions has been delegated to local units of
government. As a result, there exists a tremendous institutional complexity
which must be overcome if nonpoint problems are to be dealt with
effectively. Implementation programs and mechanisms have evolved and
nonpoint programs can, and should be put in place.
Keywords: Nonpoint source pollution, water policy, land-use policy,
nonpoint source programs, PL-92-500, institutional arrangements,
abatement program implementation
K-VIII-l
-------
NonPoint Pollution Abatement Programs: the Manager's Dilemma
The passage of PL 92-500 in the early 1970's set governments at all
levels—state, local and federal—on a collision course. This landmark
legislation was the result of extensive deliberations by the Congress which
focused its attention on the continuing deterioration of the nation's
surface waters even after several years of fairly significant federal
investment in the upgrading of municipal sewage treatment plants.
The legislation, of course, represented, in the final analysis, a
significant change in policy for the nation in its struggle to control the
continuing degradation of its waters. Pollution control efforts were now
going to be focused on treating pollution at its source, rather than in
receiving waters. Permits were to be issued to control any pollutants
discharged from a point source and pollution arising from the wide array of
man's land-use activities was also to be treated.
The term nonpoint source pollution was coined to cover basically all
pollution arising from diffuse sources. It remains interesting that in the
entire text of PL 92-500, the term nonpoint source pollution was never
defined; it was basically taken to mean everything else.
Section 208 of that same legislation launched the nation on a massive
planning effort designed to devise strategies for the control of both point
and nonpoint sources of pollution. Unfortunately, the language in that
section was also apparently vague and it took several years and many fits
and starts before that program got underway.
The challenge of dealing with nonpoint sources of pollution is that it
interfaces land use and water quality problems; the inevitable collision
course that it puts units of government on is deeply rooted in the
institutional and legal history of this country. Since the early days of
the nation, the federal government has played a major role in water
resources development as well as in water quality and policy. In 1824, the
U.S. Supreme Court concluded that navigation and related water resource
programs were within the purview of the Congress under its authority to
regulate interstate commerce.
Over the years, the federal government has been directly involved in the
construction of dams, canals, flood control projects, irrigation projects,
and the like. A minimal responsibility for pollution control was placed in
the U.S. Public Health Service in 1912 although the first abatement
authority was probably given to the Corps of Engineers under authorities
granted in the Refuse Act of 1899. Pollution abatement efforts, of course,
have been strengthened enormously over the years with significant
legislative overhauls in 1948, 1965, and 1972.
The states, basically, have exercised authority over water resources
problems only in those areas clearly without federal jurisdiction. Since
the 1950s, they have been involved in implementing federal water quality
programs, all of which has been done with pretty strong federal policy
guidelines. States do have the responsibility for the regulation of
municipal and private use of water, water supply, groundwater, and
industrial waste. It is noteworthy that these regulatory responsibilities
K-VIII-2
-------
have been carried out primarily by the states themselves and that very
little responsibility in these areas has been delegated to local units of
government.
Policy governing land-use management has followed a very different course.
The idea that a "man's home is his castle" is deeply rooted in our national
tradition. Much of the impetus behind the westward movement of the pioneers
in this country was the desire to own a piece of land to which an individual
could do whatever he or she wanted, presumably without any interference from
anybody. Although the states, primarily through their police powers, have
the authority to regulate land use, this authority has traditionally been
passed on to local units of government. States will often define areas of
interest where land use activities affect more than local jurisdiction or
will set limits to local land use activities but above and beyond this, the
tendency has been to interfere as little as possible in land-use issues.
The result is a very diffuse approach to land management questions which is
in sharp contrast to the rather sharply focused, water-quality management
programs.
The federal role in land-use plicy is very limited. They do, of course,
administer public lands which does assume significance in some western
states where those holdings make up a large percentage of the land area.
Federal policies on soil and water conservation, however, are carried out at
the local level by unique state-local institutions which were created at
federal insistence, the Soil and Water Conservation Districts (SWCDs).
But PL 92-500 set the stage. It established the fact that nonpoint source
pollution was a significant contributor to the degradation of surface water
resources and that it was a problem that had to be dealt with. Dealing with
it meant that land-use policies had to be reviewed, amended, updated, and in
some instances, established in order to solve water quality problems.
PL 92-500 also stimulated a number of research projects to provide the
technical backdrop for institutional, legal, and programmatic changes that
the implementation of nonpoint source control programs would require. Many
of the efforts examined and attempted to quantify, loadings from land-use
activities and the impacts of those loadings on surface-water quality.
Funds for some of this work came from Section 208 although perhaps the
largest program was mounted by the Great Lakes National Program Office in
Region V of the U.S. Environmental Protection Agency (EPA) with funds
appropriated under Section 108 of PL 92-500. That program was linked
closely to a massive effort by both U.S. and Canadian researchers under the
auspices of the International Joint Commission.
Rural and urban nonpoint problems were identified and assessed. Best
management practices were evaluated. Technical solutions to nonpoint
problems were tested and a fairly good understanding of how to deal with the
majority of the diffuse pollution problems emerged. Of all these projects,
the Washington County Project in southeastern Wisconsin was unique because
it was designed to examine the legal and institutional complexities
associated with the implementation of nonpoint source control programs and
even had, as one of its specific program objectives, the development of a
sediment control ordinance. It was a bold and innovative effort funded by
the Great Lakes National Program Office.
K-VIII-3
-------
The project involved 11 cooperating agencies representing state, local,
federal and regional units of government with the main actors being the
Water Resources Center of the University of Wisconsin-Madison, the State
Board of Soil and Water Conservation Districts, the Wisconsin Department of
Natural Resources (DNR), and the Washington County Soil and Water
Conservation District Supervisors. Project staff included a smattering of
soil scientists, agricultural engineers, and water chemists but was
dominated by lawyers, planners, sociologists, and economists. The challenge
was fairly straight-forward, namely, to work through the legal-institutional
maze which had evolved as a result of the differing approaches taken to land
and water management in this country and to determine if the very
heterogeneous mix of institutions could be woven into some kind of pattern
that could deal effectively with the control of nonpoint source pollution
problems.
At the inception of the Washingotn County Project, EPA's interest was
twofold, i.e., they were committed to focusing responsibilities for
implementation on a local unit of government, preferably the Soil and Water
Conservation District, and they were committed to regulating sediment loss
in rural areas.
In Wisconsin at that time (the mid-1970s) the set-up of Soil and Water
Conservation Districts was somewhat unique in that the districts were
coterminous with county boundaries and the SWCD supervisors were, by law,
those members of the county board who served on the Agriculture and
Extension Education Committee. A project sponsored study of the County
Board in Washington County concluded that change—in this instance, perhaps,
even radical change, namely sediment control--could probably best be
accomplished at this government. The reasoning behind this was basically
that county board supervisors were somewhat isolated from their constituents
who generally have very little notion of what county government was all
about. As a result of this, county supervisors tended to be re-elected time
after time without much voter concern about what they were doing. The
thinking was, then, that perhaps major change could be accomplished at this
level of government if it were done carefully.
County government in Wisconsin is basically an arm of state government and
really has only those authorities that the state grants it. A legal review
suggested that much could be accomplished within the existing statutory
framework but that the role of the SWCD could be strengthened by making it a
standing committee of the County Board and by integrating its function with
other county land use functions. This led to an extensive revision of state
statutes which abolished Soil and Water Conservation Districts in Wisconsin
and replaced them with a committee of the County Board called the Land
Conservation Committee. Authorities of the old SWCDs were retained; new
responsibilities for planning for erosion control and animal waste
management have been added. Major funding for implementing these programs
may be forthcoming from the current session of the Wisconsin legislature.
A sediment control ordinance was developed which set a performance standard
for soil losses from an individual farm based on the Universal Soil Loss
Equation. A farmer not meeting the standard could adjust his management
practices in anyway he wished as long as the predicted soil loss rate met
requirements established by the county. The Washington county SWCD
K-VIII-4
-------
supervisors were prepared to take the proposal to a referendum vote in a
part of the county when a certain amount of waffling on the question of
sediment regulation occurred both in the state legislature in Madison and at
EPA headquarters in Washington. That was the end of that proposal; the
substance of the idea was adopted finally by the SWCD as a policy or
position paper.
Efforts to regulate sediment losses from subdivisions in urban and urbanized
areas proved to be more successful. In a simple sense, urban residents are
much more accustomed to regulation than their rural counterparts and are
required to obtain permits and to meet all kinds of requirements for almost
everything they do. One more requirement one way or another didn't seem to
make much difference. The major problem encountered involved convincing
local officials that, in fact, a problem existed.
During this period, the Wisconsin Department of Natural Resources launched
its Nonpoint Source Pollution Abatement Program. Five watersheds were
selected for the program initially involving multiple governmental
jurisdictions. Local government agencies with implementation authority--
called Designated Management Agencies or DMAs—were identified to run the
programs. One watershed, the Root River in southeastern Wisconsin, had 23
DMAs. The program established the pattern for nonpoint source
implementation, namely, that local units of government with appropriate
authorities would implement programs guided by firm state policy
directives. The institutional complexities, however, remain.
This past year the EPA convened a Nonpoint Source Task Force to develop a
National Nonpoint Source Policy to accelerate the implementation of nonpoint
source control programs. The proposal policy urges continued federal
leadership and coordination but goes on to emphasize that the states will
have the lead in developing and implementing nonpoint source management
strategies. Local governments are to be charged with developing nonpoint
source strategies in coordination with the states and ultimately, with
implementing those strategies.
All the pieces would seem to be in place, yet the question still remains:
Can nonpoint control programs be implemented? The linkages between the
water and land agencies have been forged although a great deal of
institutional complexity remains. Federal agencies, particularly those in
the U.S. Department of Agriculture, are falling prey to the budget cutter's
ax.
This year is the 50th Anniversary of the establishment of the U.S. Soil
Conservation Service. The agency was formed in response to the Dust Bowl
crisis of the 1930s and under the leadership of the inspirational Hugh
Hammond Bennett represented a bold and innovative effort to solve nagging
soil erosion problems.
Now, 50 years later, we are again faced with many of the same problems and
perhaps now it is time once again to look for innovative answers to the
problems plaguing nonpoint control programs.
Authority for passing money through to states for technical and cost-sharing
assistance has existed for some years now. Maybe the time is right for
trying that approach.
K-VIII-5
-------
Participation of landowners in rural nonpoint programs today is a problem
much as it was with PL-566 Watershed Projects 20 and 30 years ago. Is
regulation the answer? The jury seems to still be out on that question.
More and more states are looking to regulation and the idea of establishing
some realistic goal, like T by 2000, seems to have become very appealing.
The established goal is to be reached in steps and most proposals call for a
regulatory mechanism to kick in if interim goals are not met.
It goes without saying that as point sources of pollution are increasingly
controlled, nonpoint sources assume even greater significance in the overall
effort to control the continuing degradation of the nation's water. Many
problems have been solved, some still remain. The where-with-all to mount a
major program to control nonpoint source has been developed. The time has
come to put that knowledge into action.
SELECTED REFERENCES
Arts, J.L., and Church, W.L. (1982). Soil Erosion—the Next Crisis?
Wisconsin Law Review, Vol. 4, No. 5, pp. 535-626.
Best Management Practices for Nonpoint Source Pollution Control Seminar
(1976). EPA 905/9-76-005.
Final Report on the Federal/State/Local Nonpoint Source Task Force and
Recommended National Nonpoint Source Policy (1985). EPA,
Washington, D.C.
Madison, F.W., Arts, J.L., Berkowitz, S.J., Salmon, E.E. and Hagman, B.B.
(1979). The Washington County Project: Final Report. EPA
905/9-80-003.
Madison, F.W. (1981). Institutional and Legal Complexities of Nonpoint
Source Pollution Control. EPA 905/9-81-005.
Voluntary and Regulatory Approaches for Nonpoint Source Pollution Control
(1978). EPA 905/9-78-001.
K-VI1I-6
-------
INSTITUTIONAL ARRANGEMENTS FOR NONPOINT POLLUTION CONTROL PROGRAMS
Floyd E. Heft
Past-president
Soil Conservation Society of America
4319 Brookie Court
Columbus, OH 43214
U.S.A.
ABSTRACT
Institutional arrangements for nonpoint pollution control must be structured
to provide maximum local acceptance, implementation, and enforcement. The
more remote the institutional arrangements are from the local polluter and his
understanding and trust, the more suspect and difficult implementation
becomes. The ultimate in institutional arrangements is about 180 degrees from
point-source pollution abatement arrangements.
Strong educational and technical assistance to the polluter, coupled with
realistic subsidies for pollution abatement practices that are not
cost-effective for the polluter, are the key to acceptance and adoption.
Local institutional arrangements must prevail in these efforts. Enforcement
must be established with a low profile and used as a last resort.
State and federal institutional involvements must carry a low profile a's well,
with realistic assistance to the local institutions. State and federal
dictates will only slow implementation. Federal and state subsidies and
implementation funds can be administered most efficiently at the local level
and should be provided through grants.
The standard procedures for monitoring point-source discharges and
establishing tolerances becomes nearly impossible and indefensible for
nonpoint pollution, particularly from an economic point of view. This dilemma
merits the alternative solution of preventing nonpoint pollution as the most
acceptable means of control.
Local institutional arrangements must and will vary from state to state and
community to community according to the forms of institutional authorities and
responsibilities and the severity of nonpoint pollution problems. State and
federal efforts to stereotype local institutions and programs will usually
hinder local initiatives and implementation.
Coordination of local, state, and federal institutions for maximum effect and
efficiency is necessary for success and for gaining of the public's confidence
that the nonpoint pollution problem is being addressed in a realistic and
productive manner. Rearranging authorities, responsibilities, and support
programs of existing institutions can and must prevent the formation of
additional institutions at any level of government.
Institutional arrangements adopted within a state must provide the capability
for implementation if responsibilities are not carried out by the primary
institution of authority, thus assuring the public of positive action.
K-IX-l
-------
INSTITUTIONAL ARRANGEMENTS FOR NONPOINT POLLUTION CONTROL PROGRAMS
Distinct differences exist between point-source and nonpoint-source pollution.
These differences must be taken into account in developing acceptable
institutional arrangements for the control of nonpoint-source pollution.
Following are the major differences:
* Nonpoint-source pollutants are primarily diffuse in nature. Most are
associated closely with surface runoff, usually following land-disturbing
activities, in both rural and urban settings. In contrast, point-source
pollutants are primarily associated with industrial and municipal
discharges that emanate at a singular point.
* Point-source pollutants frequently are by-products of commercial activity
with little or no use or value. Nonpoint-source pollutants, on the other
hand, often are of economic value to the generating unit and, therefore,
represent an economic loss. Pollution abatement efforts must consider the
fact that the motivation to pollute or not to pollute involves these forces
of opposing value.
* Nonpoint pollutants from agricultural operations mainly include sediment,
fertilizers, pesticides, and animal wastes. All represent purchased or
natural elements required for the production of food and fiber. The
positive value of these elements seemingly would be sufficient motivation
for agriculture to attempt to retain such pollutants on the land.
Unfortunately, it is necessary to expose these potential pollutants, which
are necessary for agricultural production, to rainfall and wind during
land-disturbing activities.
* An individual's or firm's ability to recover point-source pollution
abatement costs is possible—the individual or firm generally controls a
product's supply and market price. Nonpoint polluters, particularly
agricultural operations, have little, if any, ability to recover pollution
abatement costs through supply control or price setting. Therefore,
effective pollution abatement costs cannot be recovered in normal
production and marketing activities.
* Monitoring and permitting point-source polluters is quite simple and
precise. These activities usually generate the factual data, records, and
analysis that are crucial to enforcement and prosecution if necessary.
Because of their diffuse nature, the identification of sources of nonpoint
pollution and the volumes of pollutants they contribute makes meaningful
monitoring, control, and enforcement nearly impossible.
* Issuing permits to point-source polluters, a relatively easy process,
usually is accomplished by establishing an acceptable level of pollutants
to be discharged or issuing an order to the polluter requiring compliance
with an established standard. The technical and financial burden for
compliance becomes a production cost to the point-source polluter; this
cost is programmed into the cost of producing a product and ultimately is
passed along to consumers. The same procedure could be used to control and
finance the elimination of municipal sewage pollution. However, 75 percent
K-IX-2
-------
of the technical and financial burden for constructing such facilities is
assumed by the federal government in the name of public benefits, which
alleviates a large portion of the municipality's direct cost. Increased
federal taxation is necessary to pay for this federal subsidy. The higher
tax becomes an added cost to the nonpolluter having no direct obligation.
Permitting nonpoint sources of pollutants has little, if any, support from the
public or governmental institutions because of the impossibility and
prohibitive costs of monitoring the discharges. Fluctuations in discharges,
both volume and concentration, by virtue of variations in both rainfall
intensity and soil characteristics create a nearly impossible condition for
reliable and defensible monitoring. Enforcement under a permit system would
be questionable at best.
A strategy for nonpoint pollution control
The differences between point-source and nonpoint-source pollution are the
basis for development and execution of two drastically different abatement
strategies.
Success in reducing point-source pollution through the sequence of permitting,
monitoring, prosecution, and penalties has been effective because of
governments' extensive authorities.
Success in reducing nonpoint pollution cannot make use of this same sequence
of tactical actions by governments. Success can only be attained through
education, demonstration, technical assistance, subsidization, and, as a last
resort, enforcement by an institution capable of assessing a penalty.
Technological deficiencies are substantial in nonpoint pollution abatement.
Identification of a pollutant's source, its concentration or volume, best
control techniques, economic impacts, and the practicality of solutions are
all matters that agricultural research has dealt with in only limited fashion.
Agriculture instead has pursued profit-motivated research and development
strategies generally acceptable within our free enterprise system. Little
effort has been devoted to determining the environmental effects of new
products or technologies.
Extensive investigation gradually produced a prevention strategy as the most
acceptable and defensible approach to nonpoint-source pollution control. As a
strategy, prevention represents a system of management practices designed to
reduce, alter, or eliminate landowner and land operator actions that generate
nonpoint pollutants. Other than sediment, the agricultural pollutants of
primary concern are phosphates, nitrates, pesticides, and animal wastes.
Phosphates and pesticides affix to soil particles upon application to the
land. Nitrates pollute water by becoming an ingredient of surface water
runoff or by percolating down through the soil into underground water
supplies; the degree of pollution depends upon rainfall intensities, nitrate
application rates, and timing of application. Animal wastes pollute water if
applied to the land in excessive quantities or through the uncontrolled
concentration or confinement of animal units without adequate animal waste
management facilities. The focus of the prevention strategy has been to keep
agricultural pollutants from entering streams and underground water supplies
K-IX-3
-------
through the reduction or elimination of soil movement and storm water runoff
from the land into streams and lakes.
This prevention strategy focuses directly on farmers and their actions to
produce food. Owners and developers of residential, industrial, and
transportation sites are examples of nonagricultural land-disturbing
activities causing erosion and sediment pollution of our waters. Nonpoint
pollution by sediment itself contributes little chemical pollution, although
it does produce extensive turbidity and damage to the biological environment.
The identification of major nonpoint pollutants and why they enter water
supplies puts the prevention strategy squarely in front of millions of
landowners and land operators. Neither owners or operators have an economic
motivation to pollute, nor do they understand to what extent they may or may
not be polluting.
Nonpoint Abatement Goal—A First Priority
Identification of nonpoint pollutants by volume and source provides the
foundation for developing institutional arrangements for their control.
Acceptable and attainable nonpoint pollution control goals dictate what
institutions should be involved and the extent of that involvement.
Institutional structure must also recognize three facts: that landowners and
land operators are independent, private entrepreneurs by choice; that most
dislike regulation; and that resistance to regulation magnifies as higher
levels of government become involved. Many landowners still harbor the
position that a fee simple title grants complete management authority,
regardless of the impact on the public's interest. With few exceptions,
however, most will accept change if they understand the need for pollution
control and if control efforts are both justified and fair.
For 10 years I was involved in the development and early implementation of
Ohio's nonpoint pollution abatement program. The merits of that state's
strategy deserve attention. Other states have structured similar strategies
with minor differences due to varying forms of local and state governments.
Four major determinations influenced Ohio's final institutional arrangement:
1. All means were used to avoid the creation of a new institutional structure
to address nonpoint pollution control.
2. It was determined which existing institutions might logically and
willingly expand their responsibilities and authorities to implement a
nonpoint pollution abatement program.
3. Consensus was secured among agricultural, governmental, and environmental
interests regarding their desired institutional arrangement.
4. Legislation was written to authorize the selected institutions to produce
the necessary standards, rules, and regulations and implementation
procedures for an effective nonpoint pollution abatement program.
K-IX-4
-------
Among the Important factors influencing the selection of a final institutional
structure and implementation strategy were the following:
1. Existing federal and state agricultural agencies were unwilling to
restructure their voluntary-service approach to include enforcement.
2. Controlling the largest volume of agricultural pollutants was attainable
through application of soil erosion control and other land management
practices recommended by existing local conservation institutions.
3. Good working relationships and a reasonably high level of trust and
respect existed between local conservation institutions and farmers.
4. Farmers seemed relatively comfortable with the identification of pollution
problems by local agencies and recommendations from those agencies for
abating the problems.
5. Pollution abatement practices, enforcement procedures, penalties for
noncompliance, and legal recourse for polluters were generally formulated
and agreed to prior to legislative action granting state and local
institutions the necessary authorities for implementation.
6. Institutional arrangements acceptable to the U.S. Environmental Protection
Agency and, presumably, the public for point-source pollution control
proved both inadequate and inappropriate for nonpoint-source pollution
control programs.
Local Responsibilities
Conservation districts in Ohio accepted the authority to investigate and
specify the necessary combination of land management practices required to
reduce soil loss to the state-adopted limits. Districts provide technical
assistance to landowners and land operators when necessary and recommend
cost-sharing where applicable. Should enforcement procedures become
necessary, a district's administrative board issues the first warning to a
landowner or land operator. The board also establishes a time frame for
compliance and offers the farmer any technical or financial assistance
available. If the landowner or land operator complies within the designated
time frame, the enforcement process ceases. If the landowner or land operator
fails to comply, he or she is declared to be in violation of the rule, and the
district board informs the individual that the designated state enforcement
agency, the Division of Soil and Water Conservation within the Ohio Department
of Natural Resources, is to be notified of the violation. The violator is
also informed that he or she has a designated period of time within which he
or she may appeal the local board's proposed action, if desired, prior to the
board's forwarding of the citation to the state agency.
Public complaints about pollution problems can be registered with districts or
the Division of Soil and Water Conservation. Complaints must be in writing,
and they must be signed by the complainant. Complaints registered with the
state agency are referred to the conservation district for investigation and
recommendations. A district representative discusses the complaint with the
accused violator. Technical and financial assistance are offered, if
K-IX-5
-------
appropriate, to facilitate installation of any needed management practices.
This contact usually results in a cordial solution to the problem. Both the
individual who filed the complaint and the accused are informed of the
district's findings and disposition of the complaint.
This strategy for controlling agricultural pollution is premised on education
and information, technical assistance, and financial subsidy, where
appropriate. Enforcement is a last resort. But to view the strategy as weak
and cumbersome is inappropriate because enforcement is administered, when
necessary, to attain compliance.
State Responsibilities
Every state has an agency with administrative responsibility for soil and
water conservation and conservation districts. In Ohio, this unit is the
Division of Soil and Water Conservation, which also has been assigned certain
responsibilities for nonpoint pollution abatement.
Specifically, the state legislature assigned the Division responsibility for
developing and implementing a program for abatement of agricultural and urban
sediment nonpoint pollution. The program consists of rules and regulations
for the control of pollution from soil erosion, pesticides, fertilizer, animal
wastes, and urban sediment generated by land-disturbing activities. Criteria
were also established for subsidizing the installation of sediment control
practices and animal waste control facilities that are not cost-effective to
the landowner but were required for public benefits.
The Division disburses these subsidies and administers the enforcement
procedure for conservation districts when recommended. It also grants the
accused landowner or land operator a hearing prior to its final determination
and issuance of a court order and penalty schedule if that becomes necessary.
Federal Responsibilities
The U.S. Environmental Protection Agency has responsibility for setting the
parameters within which a state formulates its nonpoint pollution abatement
plan for meeting requirements of the Clean Water Act. These parameters
provide the needed flexibility to states for implementing a nonpoint pollution
abatement program tailored to a state's particular problems and institutional
arrangements. These institutional arrangements at the state and local levels
are not and should not be a federal concern beyond the assurance that the
chosen institutional structure has the capability to get the pollution control
job done. Periodic progress reports to EPA are justified as a measure of a
state's progress in meeting the nation's clean water goals.
EPA must assume leadership responsibility for other federal institutions in
compliance with state-adopted program standards and local implementation
procedures.
Urban Sediment Control Strategy
Institutional arrangements and strategies for controlling sediment pollution
from land-disturbing activities in urban areas are basically the same as in
K-IX-6
-------
agricultural areas. The major responsibility for action must rest at the
local level. County and city governmental units should be given the authority
if they are not already authorized to enact and enforce sediment control
ordinances. Both governmental units control development and building permits
within their jurisdiction, and such permits should contain sediment and storm
water management plans. No special permit need be required. In fact,
additional permits are economically and politically frustrating to the permit
seeker.
Conservation districts may encourage adoption of such sediment control
ordinances and offer technical assistance, or even enter into working
agreements for review, approval, and implementation services, as the city or
county may desire.
Should cities and counties fail to accept these local responsibilities, state
initiatives should be authorized to assure progress and compliance with
state-adopted rules and regulations.
Enforcement should automatically fall to the legal arm of the enforcing unit
of government, be it local or state. Also, local units of government should
be given the authority and opportunity to address urban sediment pollution
problems before involving state and federal institutions, beyond encouragement
and guidance from those levels.
The state can play a vital role by developing and adopting urban sediment
control rules, regulations, and a model ordinance as guides for local
initiatives.
Logistical Support Strategy
Local institutions that assume the responsibilities for nonpoint-source
pollution abatement have many state and federal agencies available to assist
them in varying degrees and ways. It is not only logical but imperative that
these agencies be used to the maximum extent. The structuring of new
governmental units should be avoided.
The major areas of assistance needed and normally available include the
following:
Education and Information. Education of the public and landowners
specifically about the chemical and physical aspects of nonpoint pollution is
essential to acceptance of an abatement program. The Cooperative Extension
Service has a long-standing, highly respected reputation for public education
in agricultural technology. This education is founded on research by state
and federal agencies. Extension's participation in education should be
sought. But information, of course, can be disseminated by any public agency
as well as commercial interests and the public media. These educational and
informational forces are extremely important to landowner acceptance and
public support.
Research. Research must address the causes of and cures for nonpoint
pollution. The Agricultural Research Service and state agricultural research
K-IX-7
-------
institutions must produce practical and economical pollution control
solutions.
EPA has essentially been ineffective in its strategy of allocating research
funds for nonpoint pollution control by requiring answers in three years. No
defensible data or research finding can be produced in such a short time span
when incorporating such unpredictable variables as rainfall quantity and
intensity. A more realistic time frame would attract new research urgently
needed and produce reputable findings.
Technical Assistance. Technical solutions to preventing nonpoint pollution
are directly related to land and water management systems. Both are directly
affected by soil and water conservation practices. The Soil Conservation
Service, state conservation agencies, and conservation districts have
specialized in planning and applying these practices in cooperation with
landowners for about 50 years. These agencies can continue to provide the
technical assistance needed to deal with nonpoint pollution. These same
agencies must exercise caution in using the Universal Soil Loss Equation as a
pollution abatement equation. The equation must be used only as a measurement
of soil loss from a given site. Delivery of such sediment to a given stream
or body of water requires the development of a delivery formula or model.
Local institutions must provide technical and administrative expertise above
and beyond that received from state and federal institutions. Acceptable
implementation rates for pollution abatement systems dictate that local
institutional capabilities be strengthened.
Financial Assistance. Financial support is imperative for the successful
implementation of a nonpoint-source pollution abatement program. All levels
of government and the polluter must share the financial burden for pollution
control.
Federal funding to help solve nonpoint pollution problems must be commensurate
with the point-source pollution abatement funding relative to public benefits.
To date, spending for nonpoint pollution has been insignificant from a federal
point of view. Institutionally, federal funds for nonpoint pollution
abatement should represent a realistic portion of the federal funding of
pollution abatement efforts by the congressional public works committees to
the U.S. Environmental Protection Agency. Such funds should not be provided
through the congressional agricultural committees to agricultural agencies.
It is logical and imperative that such funds be viewed by the public as water
quality funds benefiting the public and not just agriculture as another
subsidy. Federal funds need to be allocated to state and local institutions
in the form of grants for implementation and subsidies. Experience dictates
that this procedure will not only be more productive but more efficient per
dollar spent. Any other system of federal financial support will be less
effective and produce financial waste.
State funding should supplement federal funding to assure that local
implementing institutions are adequately staffed and subsidy needs in a state
or area are sufficiently met. State funds can be most effective if they are
used by local institutions responsible for implementation.
K-IX-8
-------
Local funding should provide for basic facilities and support personnel.
Special needs and cooperative efforts can best be funded locally to assure
jurisdictional control and direction.
Subsidies to private landowners from any institutional or governmental level
should be restricted to those practices required of the landowner that are not
cost-effective to the individual but are required to achieve public benefits.
Any subsidized landowner practice must be accompanied by a mandatory
maintenance provision, and there must be a means for recovering public funds
if a practice is not maintained. This mandatory maintenance provision is now
used by many state and local institutions, but it has never captured the
support of federal agricultural institutions administering subsidy programs
for erosion control. The public has a right to require this protection of its
investment.
Other Assistance and Structural Issues. Other assistance, such as legal
support and private-sector support, must be built into institutional
arrangements on the basis of various state and local jurisdictions and
capabilities.
The most successful institutional arrangements for nonpoint pollution
abatement must be acceptable to the local implementing institutions first,
then to state and federal institutions, in that order. The nature and
complexity of the problem dictates these priorities of acceptance.
Institutions implementing nonpoint pollution programs must accept the fact
that nonpoint pollution cannot and will not be eliminated, but reduced by the
best of institutional arrangements. Nature by itself produces significant
nonpoint pollutants through natural processes.
The number of polluters essentially equals the number of landowners
nationwide. This requires a strong, persuasive effort prior to enforcement
action. It also requires a graduation of compliance, dealing with flagrant
violators first, which will generate rapid public support.
Progress toward nonpoint-source program goals and clean water objectives must
and will move at a somewhat slower pace than has been the case in point-source
programs.
A benefit secondary to the immediate purpose of pollution abatement, but
possibly an even greater benefit in the long run, is that pollution abatement
practices will equally benefit the protection of the food-producing base of
our nation. These co-benefits should generate public support politically and
financially if properly implemented through realistic institutional
arrangements.
K-IX-9
-------
WHO SHOULD PAY FOR
NONPOINT POLLUTION ABATEMENT?
Lawrence W. Libby
Professor, Department of Agricultural Economics
Michigan State University
East Lansing, Michigan
ABSTRACT
Nonpoint pollution is an expensive social problem, but abatement of nonpoint costs
society as well. Nonpoint policies involve implicit comparison of these two categories of
cost and their distribution among policy participants. This paper assembles economic
evidence of costs of nonpoint and cost of abatement, discusses distribution, and suggests
policy direction.
Emerging policy in nonpoint abatement must place greater emphasis on national efforts
rather than state or local. Because those who cause nonpoint are usually separated by
both time and space from those damaged by pollution, institutions that bring these
interests together are essential. Further, a "polluter pays" philosophy will become more
prominent in nonpoint policy. Society will insist that the polluter bear a greater
obligation for the impacts. Economic incentives that redistribute cost (a tax, control, or
cross compliance) or benefit (benefit share) will improve chances for design of
acceptable, workable policy.
Key Words: Nonpoint pollution policy, economics, institutions, distribution of cost.
K-X-1
-------
Who Should Pay for
Nonpoint Pollution Abatement
by
Lawrence W. Libby*
Introduction
Nonpoint pollution is an expensive social problem. It imposes enormous costs on
water users, a cost recently estimated to total about $6 billion a year (Clark, 1985a).
Sediment and pollutants ruin the ecosystem habitat for fish, waterfowl and other
organisms that are valued by people for various reasons. Clogged harbors and channels
raise the cost of water transport. Nutrient enriched lakes and streams are less
attractive for recreation or as part of a living environment; damaged ground and surface
water must be treated before use in municipal systems. The costs and consequences of
nonpoint have been discussed and documented throughout this conference. But
abatement of nonpoint costs society also. Nonpoint policies involve implicit comparison
of this cost of abatement with the cost of not abating and the distribution of the two
categories of cost. Nonpoint is the unfortunate yet predictable side effect of activities
that are generally worthwhile — food production, construction, and others. Abatement,
then, may mean less of that valued activity, or can be direct financial outlay to reduce
the water runoff.
This paper addresses the economics of abatement — both the efficiency and
distributional implications of those actions taken to reduce nonpoint damage.Primary
emphasis is on agricultural nonpoint pollution. Effficiency consideration basically
establish the technical parameters, the feasible set within an abatement strategy may be
selected. Distributional impacts influence choice by discribing who pays and who gains
from the options available. The most economically efficient abatement technique,
whatever it may be, has important distributional consequences. Thus it is just one among
many abatement options, evaluated by political actors in terms of what it gives them at
what price.
Purpose of this paper is to assemble the economic evidence being generated and
draw conclusions that may be useful for this conference and for any subsequent policy
recommendations that might be developed. A further purpose is to help set the stage for
more specific technical sessions to follow.
First, a series of assertions about "the nature of things" in the political economy of
nonpoint will establish context.
1. "Efficient" solution of the nonpoint problem implies the greatest possible
abatement for the money spent, achievement of a given level of abatement at least
possible cost and/or most importantly, comparing the cost of an additional increment of
abatement to the benefits from achieving that increment. The assumption is that when
Professor of Agricultural Economics, Michigan State University. Paper prepared
for the Symposium on Nonpoint Pollution Abatement, Marquette University, Milwaukee,
WI, April 23-25, 1985.
K-X-2
-------
an increment costs more than it gains, whoever gains or loses, society has acquired
enough reduction of nonpoint pollution. Economists love to repeat this basic rule, but it
has real meaning in the social process of allocating public effort among many valued
services, of which clean water is just one.
2. Benefit, cost and therefore efficiency in resource use are defined within the
structure of property rights that expand or constrain the rights of individual water
users. These rights are established and reinforced by public authority. Thus any solution
to pollution is valid or efficient only to the extent granted by publicly acknowledged and
protected property rights. There is nothing meta-physical about economic efficiency, in
abatement or any other production process. It is a fabrication, a result of transactions
among actors granted the priviledge of commerce, with publicly protected rights to
impose a price for a productive factor. Those rights could change, thus altering the
efficient solution.
3. As with most forms of pollution, nonpoint agricultural pollution involves impacts
that are separated in time and space from those causing the problem. Pollution is an
additional output of a production process generating a "good" of some kind. Property
rights protect the producer, limiting his liability for the cost of that undesireable output
in the production decisions that determine quantity and form of output. Thus, the
benefits of abating nonpoint are distributed differently from the costs of abatement.
While those benefiting from abatement might be willing the "bribe" those causing the
problem through some sort of compensation program, the costs of organizing the
transaction would be formidable.
4. Participants in the policy process form positions on options based on a
comparison of separable benefits to separable costs. There is no inclination to pay or
bear inconvenience if that action produces no obvious result of consequence to the
actor. Some people will bear personal cost to create benefits for others out of a sense of
community or altruism. But it is risky to construct nonpoint policy on the assumption
that that will occur.
5. In nonpoint policy as in other areas, good policy is acceptable policy. Policy
changes that may be technically correct but require too great a sacrifice of other valued
services are irrelevant. "Non-degredation by 1985" is and has been such an irrelevant
policy goal. Changes occur in incremental adjustments as competitors compare the
relevant consequences of options and bargain on that basis. Policy goals and objectives
may be stated and used as targets so long a participants do not expect to actually
achieve them. Goals and objectives evolve in policy bargaining. They must be flexible
enough to accomodate reality (Lindblom, 1979).
In nonpoint abatement, the efficiency questions involve comparing the
consequences of pollution with the consequences of abatement. Policy participants must
understand the cost of failing to reduce erosion and compare to the cost of cleaning up,
to arrive at the marginal benefit/marginal cost comparisons discussed above.
The Evidence on Cost; A Brief Review
Various empirical studies have sought to establish the magnitudes of the two types
of cost — pollution and abatement. There is further development of methodology to
improve the comprehensiveness and reliability of these estimates. Shortage of site
specific water quality data continues to be a problem in seeking economic estimates of
damage.
K-X-3
-------
Cost of Nonpoint. The cost of nonpoint pollution represents a reasonable starting
point for estimating benefits of abatement. Some non-user benefits might be omitted
from the cost figures, thus cost of pollution may understate the actual benefits from
abatement. But certain instream costs to the biological ecosystem are also omitted.
There is an implicit property rights issue imbedded in this cost/benefit template of
nonpoint pollution. Those hurt by pollution are basically trying to defend something they
already have ~ a clean water supply. The polluter, however, implicitly owns the right to
convert that clean water into the combination of dirty water and a product,both of which
are outputs of the production process in question. The payment that the downstream
user would be willing to accept to part with clean water would likely be greater than
what he would be willing to pay to acquire a new supply of clean water (Bishop and
Heberlein, 1980). Initial allocation of property rights, therefore, will influence benefit
estimates of nonpoint abatement. Magnitude of that difference would be a useful
empirical question to pursue.
The most definitive national estimates of nonpoint costs have been generated
recently by the Conservation Foundation(Clark, 1985b). Authors assemble the most
useful empirical literature on the subject, impose a few defensible assumptions, and
develop some estimates. While more specific case studies within particular watersheds
could improve the quality of the national figures, the CF estimates are a significant
contribution to policy development in the area. Pollutants include sediment, nutrients
and chemicals. Instream damages are defined to include impacts on the aquatic
environment, reduction in recreation benefits from fishing and other water sports,
reduced water storage and holding capacity of reservoirs and lakes, increased cost of
maintaining harbors and water ways, and loss of certain non-user values associated with
clean water. Their estimate of these instream damages is $4.3 billion per year, of which
agriculture causes $1.6 billion. Off-stream damages result from increased flooding
associated with suspended sediment, less efficient water conveyance systems (canals),
and more expensive water treatment systems for power generation. The total off-stream
estimate is $1.9 billion of which agriculture accounts for $660 million. These are
acknowledged to be "order of magnitude" estimates (Clark, 1985a, p. 22).
A recent case study of the Obion-Forked Deer River Basin in Tennessee estimated
off site damagaes from erosion totalling $74 million per year plus losses associated with
flooding of productive hardwood forests (USDA, 1980). Impact of declining water quality
on value of waterside property along St. Albans Bay on Lake Champlian was estimated
using an hedonic model. The water quality variable constituted approximately 20% of
property value or $4,500 per property difference attributed to quality of the adjacent
waterbody. Total damage experienced by owners of 430 single family residences was $2
million (Young and Teti, 1984). Other case studies are needed to get a more accurate fix
on economic burdens associated with nonpoint pollution.
Benefits of Abatement. Another perspective on the cost of pollution side of the
nonpoint policy question may be gained by estimating benefits of cleaning up the water.
Benefit is more than just damage avoided. It also includes the liklihood of continued
availability of clean water and the impact of that liklihood on willingness to invest in
waterside property or other services that require clean water.
The Economic Research Service of USDA is putting considerable effort into
estimating benefits of abatement in the various rural clean water projects (RCWP)
around the country. Work is still underway, but a few results are available. Improving
the quality of water in St. Albans Bay in the northeastern part of Lake Champlain in
Vermont would produce annual benefits estimated at $230,300 through a contingent
valuation technique and about twice that using a travel cost approach (Ribaudo, 1984).
K-X-4
-------
The travel cost model involved estimating the cost that a recreator has been willing to
bear to enjoy the bay for various recreation experiences and the amount of travel
expense that recreator would pay if the bay were cleaner. Number and costs of trips
from various origins in the region were estimated. Relationship between the willingness
to bear travel cost and distance from St. Albans, availability of alternative sites and
income levels was calculated for current water quality and for the hypothetical situation
of a cleaner bay. Mean difference in willingess to pay for recreation attributable to
improved water quality was $123 per recreator, for a total recreation benefit of
$530,700. The contingent valuation model involved interviews of prospective recreators
who were asked to rate several recreation opportunities involving St. Albans Bay and
then re-rate those options with the possibility of a cleaner bay. These indifference maps
were then converted to monetary benefits associated with improved water quality in the
bay.
Benefit estimates of several other clean water projects will be published soon.
Each is unique, with land use, soil, and water characteristics that differ significantly
from site to site. Generalization is impossible. Yet these point estimates can improve
chances for public agencies and various political participants involved in clean water
debates to draw conclusions that are economically sound. With reasonable estimates of
abatement costs and their distribution, one may draw a conclusion about the efficacy of
a particular abatement program.
Cost of Abatement. The greatest amount of published work in the economics of
agricultural nonpoint pollution deals with the costs that selected management techniques
could impose on the farmer. These studies generally assume initial distribution of
property rights that gives the farmer the right to decide for or against a reduction in
run-off. In a 1976 linear programming analysis, Kasal examined farm income
consequences of imposing limits on fertilizer use and soil loss as measures to reduce
nonpoint pollution. For the sample farms selected and alternative measures imposed,
nonpoint abatement reduced farm incomes by from 10 to 36 percent (Kasal, 1976). A
national LP model run by Wade and Heady made similar kinds of estimates of increased
production costs associated with nonpoint measures, within the overall requirement of
meeting projected food demands. The minimum sediment solution to the production cost
minimization model was 42% higher than the unrestricted solution. In both cases there
was considerable cropland adjustment within the model to retain sufficient production to
meet output requirements (Wade and Heady, 1977).
Alt, e£ al (1979) examined the farm level costs implicit in reducing sediment
delivery to a specific reservoir in central Iowa from the Iowa River watershed. A 10 ton
acre limit on gross erosion from farms in that watershed would increase production costs
17% while reducing sediment delivery by 91%. White and Partenheimer (1979) examined
income consequences of erosion reduction plans for a sample of Pennsylvania dairy
farms. Impacts varied among the specific farms studied, with two farms experiencing
increased revenues from the recommend soil conserving plan, four with income
reductions of less than 5% and six with reduction from 7% to 30%. A no-till option was
introduced with the result that all but one farm showed increased returns. The net
on-farm/off-farm economic consequences of erosion reduction were examined for a
particular river basin in northeast Texas. The general conclusion reached in this analysis
was that the positive on-farm impacts of soil conservation overshadowed whatever
revenue reduction might be associated with complete attention to reducing off-farm
damages. Thus, authors conclude, it is in the farmer's interest to conserve soil and no
regulation or subsidy is warrented assuming that farmers are economically rational
(Reneau and Taylor, 1979).
K-X-5
-------
Results of these and many other case studies of the economic costs of various
erosion control techniques have provided important analytical back-up for recommending
best management practices in specific state and local water quality plans. The whole
approach to nonpoint reduction has been local solutions to local problems, with "best
practices" a function of the physical, economic and institutional circumstances prevailing
in that area; A systematic approach for considering the various factors involved in
selecting best management practices for a particular political/economic/hydrologic
sub-region has been suggested by Bailey and Waddell (1979). The general policy goal has
been to establish an abatement program that avoids major cost impact on local farmers.
Procudures for considering the most important firm level costs have been recommended
by USDA and EPA (1975) for the development of those BMP's (1*). Inter-regional and
national consequences of these strategies have been estimated also. The more
aggregated the analysis or modeling becomes, the more imprecise and unreliable the
estimates because of obvious differences among areas.
A significant and widely acknowledged gap in the evidence on alternative nonpoint
strategies is the link between technique and water quality improvement
(Christensen, 1983). We know that nonpoint costs money; we know that abatement costs
money; we even know that the two costs are distributed differently. But we need much
better evidence on the aggregate water quality results of farm level practices that
reduce run off, so that the estimates of economic benefit of those actions may be
improved.
Reduced tillage alternatives show particular promise since they can, under some
circumstances, lead to increased net farm revenue while reducing run-off. Pope, et a[
(1983) concluded from their study of Iowa agriculture that conservation tillage systems
have less negative impact on farm income than is true for conventional systems. Study
of alternative tillage systems on Ohio farms found that no-till options that reduced
erosion to T or less also produced higher net revenue than the base solution of fall
plowing on soybeans (USDA, 1983). In his study of cash grain farms in the Jackson
Purchase Area of Kentucky, Kugler found that for nearly all farms there was a tillage
conversion that would increase net returns while reducing erosion (1984). Black, et al
(1984) concluded that since yield is not affected by tillage system for the sample of
Michigan farms studied in the Saginaw Bay watershed, the net revenue advantage for
conservation tillage comes from reduced machinery and labor costs. Determination as to
whether or not reduced tillage makes sense as a best management practice for nonpoint
abatement depends on the net effect of that practice on the environment. While
conservation tillage often reduces soil movement and runoff, it may also result in
increased chemical concentration in the run-off water that does leave the field. The
plant residue left in the field as part of the reduced tillage system tends to increase
insect problems and reduce the effectiveness of some herbicides (Baker, and Laflen,
1984). Any calculation of the cost of nonpoint abatement using reduced tillage or any
other erosion reducing technique must consider offsite damage as well as income impacts
for the farmer. Conservation tillage may reduce erosion at little direct economic cost to
the farmer but if it also increases chemical contamination of rivers and streams because
of increased use of pesticides, its viability as a best management practice is
questionable.
The real difficulty for an individual or government in reaching an efficiency
decision on nonpoint investment is that information comes in fragments. Cost and
benefit are not felt at the same time or place. At the farm level, an action that reduces
nonpoint may have advantages in net revenue. On the other hand, a practice that
reduces erosion and increases the farmer's revenue, may worsen the water quality
problem. As analysts we try to isolate factors, examine them one at a time for their
K-X-6
-------
positive and negative consequences. In fact, however, the farmer makes decisions within
a complex of incentives, risks, and options. It may be that an incentive necessary to
attract socially rational investment in nonpoint abatement should be artifically linked to
the action that would reduce runoff. That is the general idea behind cross-compliance in
soil conservation policy. As long as the positve incentive is greater than any income
burden from the resulting nonpoint technique, it can be a rational choice for the farmer
(Dinehart and Libby, 1981). So long as the public cost of the incentive, including
whatever unintended side impacts it may cause, in less than the incremental cost of
pollution it can be an improvement.
Success of alternative policies to reduce nonpoint pollution will depend primarily on
the distribution of both the cost of pollution and the cost of abatement. In policy we are
more concerned with who pays than with the overall size of the bill.
Who Pays?
Distributing the burden for reducing agricultural nonpoint pollution, assuming there
is general agreement that nonpoint should be reduced, is the central policy question in
this area. The prevailing theories are 1) those who benefit from abatement should pay
for it, and 2) those who cause the problem should pay for its solution. The former implies
that the right to permit run-off remains with the polluter; those who want less pollution
because they currently experience a pollution cost, must pay for abatement. The
problem, then, is to decide who benefits and levy a tax accordingly. In large measure
this is the current system. Farmers and other polluters retain the right to pollute;
abatement is voluntary; cost sharing and technical assistance are paid for by taxpayers.
State level cost sharing programs exist in many states, implying that benefits are
received by all state taxpayers. As of 1983, sixteen states have their own cost share
programs (Braden and Uchtmann, 1984). Federal cost sharing, through ASCS of the US
Department of Agriculture, implies that all US taxpayers would benefit from reduced
pollution and should therefore pay for it. In practice, state programs actually
supplement this federal support.
The latter approach to distributing the cost of pollution implies that the polluter
has full liability for his actions and that any cost of abatement must be included in the
cost of production for the commodity or service that creates to pollution. Thus far, the
only method for implementing this approach to distributing abatement cost has been
through regulation. County soil conservation districts in most states have the authority
to impose regulation against excessive erosion, though few have exercised that
authority. Five states have sediment control ordinances that include agriculture. Iowa
has the most aggressive and comprehensive regulations against excess erosion. Illinois
has a newly enacted law designed to "jawbone" farmers toward acceptable erosion levels
by the year 2000. Pennsylvania has an innovative permit system functioning through the
state's Clean Stream Law (Holmes, 1979, p. 63-93).
Other mandatory approaches being examined for soil conservation programs include
various cross compliance measures that would require soil conservation as pre-condition
for eligibililty for price supports and other income support programs. While these have
not really been designed as nonpoint abatement measures, they could help meet water
quality goals.
The matter of "who should pay" for nonpoint is obviously a matter of opinion.
Conclusions presented here are based not on perception of right and wrong (though some
element of that may creep into the discussion) but on a judgement of what is likely to
K-X-7
-------
happen. The "benefits received" approach is well received in the agricultural community
including, incidentally, most of the Land Grant system. The obvious problem is, however,
that it has not worked. Policy development for the next decade will shift more generally
toward the "polluter pays" approach reflecting concern for cost effectiveness in nonpoint
and other natural resource policy as well as a growing impatience among the pollution
control professionals about progress toward reducing nonpoint. The magnitude of the
estimated national cost of nonpoint pollution and agriculture's portion of that total has
the attention of the Soil Conservation Service. Estimated annual costs imposed by
agricultural nonpoint pollution are several orders of magnitude higher than annual cost of
soil productivity lost to erosion. SCS must give greater attention to off-site
consequences of erosion in these days of national budget stringency for agriculture
natural resources and for SCS in particular.
Policy Conclusions. The following conclusions on the general trends in policy and
cost distribution are based on a review of the economic evidence on nonpoint summarized
briefly above, review of policy literature concerned both with soil conservation and
water pollution, and a general "testing of the wind." Following these brief conclusion, is
a list of suggested innovations for policy in this area.
1. Abatement of nonpoint must be a national program, not just the sum of state
programs. Despite the fact that 208 water quality planning, the Model Implementation
Program and the Rural Clean Water Program emphasize state action, and the Reagan
administration prefers decentralization, water pollution is just as is national as national
defense. The implicit assumption that beneficiaries of abatement are only the first line
pollutees who directly feel the results of dirty water, and that they reside in the same
state as those causing the problem cannot be sustained. In fact, benefits of a clean
water supply go well beyond immediate users. They include various non-user benefits
from continious availability of water when and if it might be used. They include various
multiplier impacts on the overall economic environment of a region where the farms,
cities and countryside become attractive places to live and work. Clean water is part of
a broad sense of well being or security that will encourage investment and general
popular support for change.
A recent article in the 3ournal of Soil and Water Conservation was entitled "Saving
the Chesapeake: Maryland's Agricultural Education Program," (Magette, 1985). Even if
the title were changed to refer to "Maryland's Ironclad Regulation Against Erosion," it
would be wishful thinking. There is just no way that the problems of the Chesapeake Bay
can be handled separately by Pennsylvania and Maryland. Even when the pollution
problem is reasonably well confined as in Saginaw Bay, Michigan or Green Bay, Wisconsin
benefits of abatement extend well beyond. There has been important institutional
innovation in states, and that must be supported. The Wisconsin nonpoint program is an
exceptional example of a balanced and integrated program with all components from
problem identification to implementation (Konrad, et al, 1985). RCWP's have become
testing grounds for coordination among federal and state agencies. They have been
strongly endorsed by EPA and UDSA (Groszyk, 1979; Unger, 1979). But beyond these
intermediate purposes, reducing nonpoint to acceptable levels will require federal
action. It is unreasonable to expect an agricultural state to enact abatement measures
sufficient to produce adequate water quality benefits for downstream users in other
states. To a large extent we have allowed the traditions of state and local prerogative
on land use management to run the nonpoint program. We begin with the irrevocable
presumption that locals know best when it comes to land use. Perhaps that is true, but it
will not solve the nonpoint problem. If land use controls are the only solution to
nonpoint, and land control must be a local government function, then nonpoint will be
with us for a long time.
K-X-8
-------
A national program for nonpoint abatement means that the cost will be shared
among payers of national income tax. State programs may supplement but never replace
the national focus. Taxpayers in agricultural states will pay more than taxpayers
elsewhere since farmers in those states will likely have sufficient political clout toshift
the abatement cost and get additional cost sharing programs installed.
2. Farmers and others causing nonpoint pollution will pay an increasing portion of
the abatement cost. Most of this cost will be indirect, in the form of fewer property
rights to use land in ways that cause nonpoint pollution and through any reduced revenue
associated with less erosive farming systems that might reduce output. There is simply
no possible way that continued reliance on technical assistance, cost sharing and
voluntary action can satisfactorily cope with the nonpoint problem. Reduced tillage has
really been the only hope for this softer approach and recent evidence on the pollutant
concentration in run-off raises serious question about the viability of these techniques.
This should not suggest that the voluntary approach with or without the other elements is
inadequate in every case. There have been important successes. But in a national
nonpoint program, these measures are simply inadequate.
There is nothing earthshaking about this conclusion. Resource policy professionals
have known for somtime that only by changes in farmer behavior will agricultural
nonpoint be reduced, and current incentives do not favor the necessary changes. The
RCWP era in water quality policy has been one of support building and institutional
design within the prevailing balance of power and responsibility among levels of
government. It has been necessary. But there will be pronounced shift in philosophy
toward the "polluter pays" approach. Those who stand to gain from abatement, a broadly
defined community of water users, will be unwilling to leave all of the land use discretion
in the hands of those who cause the problem. Epp and Shortle suggest that given this
inevitable policy shift, the water quality agencies and research community should focus
on design of effective and economically efficient mandatory programs (Epp and Shortle,
1985, p.111). That is sound advice.
Distribution of burden within agriculture deserves policy attention as well.
Farmers on erosive land will find it particularly painful to meet mandatory limits. They
suffer the initial disadvantage of less productive land and then the higher cost of
compliance. The inevitable result will be further concentration of agricultural
production on the most responsive and easily protected lands of the country. Additional
people and areas will be displaced from agriculture. There must be special attention to
these poeple to facilitate transition and limit the hardship involved. But attempting to
mask the inherent disadvantage of farming these lands through special credit or disaster
payments would be an expensive and apparently inappropriate policy response. With
mandated erosion and run-off standards, the inherent advantage of some lands would
likely be reflected in land price. Technology may reduce the differences between erosive
and non-erosive lands, however.
Directions for Policy Innovation
The major needs in nonpoint involve, first, new organizational strategies for getting
those who cause the problem together with those who gain from its reduction. As noted,
costs of pollution and costs abatement are distributed differently. Only by creating a
political environment conducive to negotiations between these categories of participants
can lasting policy change be accomplished. The second major need is for financing
schemes that take advantage of the economic facts of pollution and abatement.
Economists have long been reviled for seeking social purpose through inherent selfishness
of the individual. That is an unfair assertion, of course. The more positive perspective is
K-X-9
-------
to facilitate expression of the true consequences of alternative actions so that private
choices will yield results that are socially responsible.
1. The first suggestion is not new at all but entails giving new life and new-
missions to an old institution ~ the river basin commission. There is logic in hydrologic
units! As water resource professionals have known for many years, the river basin unit
can overcome much of the transaction cost associated with bargaining between gainers
and losers from a pattern of water use. This capacity can be particularly important in
nonpoint abatement where the source of the problem is diffuse, benefits widely
distributed in small increments. Some of these inter-state linkages have been established
for the Chesapeake. Experience in the Deleware and Susquehanna River Basins has been
well documented (Libby, 1970). Bay area regional commissions with multi-state
memberships could be established or focused more clearly on nonpoint problems. Purpose
of the water shed, or more generally "problem-shed", is to establish the setting for
bargaining over the terms of nonpoint policy. There should be no claims of designing the
optimal water management program. Instead, interested parties could iterate toward
acceptable levels of pollution and abatement, and a distribution of burden that is
tolerable. With focus clearly on nonpoint, voting and membership rules would have to
encompass the essential parties involved. A different organization might be appropriate
for other resource problems in the basin.
Hydrologic units are awkward to adminster. They cross many jurisdictional
boundaries and lack political legitimacy. The administrative overhead can be
substantial. The problems are not much different from those of multidisciplinary
research is an acedmic institution. But there can be no lasting reduction of nonpoint
pollution without some mechanism of forcing confrontation among the interested
parties. Provision must be made for beneficiaries outside the basin and some basic
understanding of problem source and benefit of improvement. Water quality standards
would help establish targets. Even with mandatory controls on erosion, the mechanism
for bargaining would be valuable.
2. Within a decision or bargaining unit such as a river basin, a mechanism for
redistributing the benefits of abatement could improve chances for incremental
resolution of the nonpoint problem. Such an innovation has been analyzed by Park and
Shabman (1981) for the Occoquan River Basin in Virginia. Under such a scheme, those
who gain from abatement could spread the benefits sufficiently to generate support
necessary for a compensation scheme that will bribe the polluter not to pollute.
Alternatively, this system might produce the plurality to enact mandatory controls.
Areas receiving major measurable benefit from abatement could compensate areas that
would benefit less to help finance installation of BMP's with the basin commission acting
as banker. As long as net benefits to all parties benefits are positive, it is in their
interest to share to build support. The basin commission's role is to reduce transaction
cost associated with bargaining both among beneficiaries and between those who cause
and those hurt by nonpoint pollution.
The Chesapeake Bay situation would seem to lend itself to the type of innovation
suggested by Park and Shabman. Any such system would require improved information on
water quality benefits and impacts of alternative abatement techniques.
3. Another device for employing an economic incentive to encourage private
actions with a social purpose is installation of a tax in pollution. Economists like this
idea particularly well because it forces the producer to internalize social costs of the
production process. The tax also creates the incentive to reduce pollution to avoid the
tax. While there apparently is no real experience with a tax on nonpoint pollution, Seitz
K-X-10
-------
ejt al used a linear programming model of hypothetical farms representing western Illinois
agriculture to conclude that a soil loss tax would be more efficient than an erosion
restriction in reducing nonpoint (Seitz, et al, 1979, p. 375-376). Under a tax scheme, the
farmer would seek the least costly means for reducing run-off, unique to the land
conditions on that farm. The erosion standard or restriction, on the other hand, might
impose a more costly technique than is necessary.
The obvious difficulty with tax schemes is absense of specific information on
performance of practices, cost of pollution and other essential variables for a given
farm. It may be that trial and error would be a reasonable research design in this area of
building institutions. The theory of tax schemes is well developed, but implementation is
lacking. Selected pilot studies could be a reasonable use of scarce research funds.
4. Marketable or tradeable rights to pollute constitute a third type of innovation
for marshalling the forces of greed in the interest of society. The sediment and waste
assimilative capacity of a stream or lake might be allocated among farmers on the basis
of some readily determinable index such as frontage on water course or acreage adjusted
by an indicator of accessibility to water. Then the farmer would have the right to sell or
trade those rights when their value exceeds the cost of reducing run-off. Implementation
of such a scheme would require monitering to determine compliance for a particular
farm. There might also be a problem of localized pollution, where farmers decide to
acquire pollution rights and let the pollution occur. Those farming areas that can reduce
run-off most readily will have much cleaner water. The average , or overall assimilative
capacity, could hide extreme situations that impose local hardship. There would be
overhead cost involved for the basin commission or other managing unit in defining terms
of trade and upholding the result.
After thorough review of the economics and institutional literature related to
nonpoint pollution one could easily conclude that there is really very little that is new.
There are few revolutionary new ideas. In fact, there is a fair amount of repetition at
the conceptual level. What is lacking is experience, empirical observation of various
means of coping with disparate patterns of distribution of the cost of pollution and cost
of abatement. There are limits on acceptable means for redistributing those burdens, but
the limits are less severe than traditionally assumed. Acceptability is a function of
prevailing attitudes about who can or should shoulder more of the burden for meeting a
widely held social objective for clean water.
K-X-11
-------
References
1. Alt, K. F., 3. A. Miranowski and E. O. Heady. 1979. "Social Costs and
Effecitveness of Alternative Nonpoint Pollution Control Practices",
Best Management Practices for Agriculture and Silviculture, Ann Arbor,
Michigan: Ann Arbor Science Publishers, Chapter 23, p. 321-328.
2. Bailey, G. W. and T. E. Waddell. 1979. "Best Management Practices for
Agriculture and Silviculture: An Intergrated Overview", Best
Management Practices for Agriculture and Silviculture, Ann Arbor,
Michigan: Ann Arbor Science Publishers, p. 33-56.
3. Baker, 3. and 3. Laflen. 1983. "Water Quality Consequences of Conservation
Tillage". 3ournal of Soil and Water Conservation. 28(3); p. 186-193.
ft. Black, 3. R., H. Muhtar, 3. Posselius and 3. Schwab. 1984. "Results of an
Economic Comparison of Conventional and Conservation Tillage
Systems in the Southeastern Saginaw Bay Coastal Drainage Basin",
Proceedings for a System Approach to Conservation Tillage, East
Lansing, Michigan: Michigan State University Cooperative Extension
Service.
5. Bishop, R. C. and T. A. Heberlein. 1979. "Measuring the Values of Extra Market
Goods: Are Indirect Measures Biased?" American 3ournal of
Agricultural Economics, 61 (1979): 926-930.
6. Braden, 3. E. and D. L. Uchtmann. 1984. "Soil Conservation Programs Adrnidst
Faltering Environmental Commitments and the New Federalism",
Environmental Affairs Law Review, 10(3): 689-696.
7. Christensen, L. A. 1983. "Water Quality: A Multidisciplinary Perspective",
Water Resources Research; Problems and Potentials for Agriculture
and Rural Communities, Ankeny, Iowa: Soil Conservation Society of
America, p. 36-62.
8. Clark, E. H. 1985a. "The Off-Site Set Costs of Soil Erosion", 3ournal of Soil and
Water Conservation, 40:1, p. 19-22.
9. Clark, E. H., 3. A. Haverkamp and W. Chapman. 1985b. Eroding Soils; The Off-
Farm Impacts, Washington, DC: The Conservation Foundation.
10. Dinehart, S. and L. Libby. 1981. "Cross-Compliance: Will It Work? Who Pays?"
Economics, Ethics, Ecology; Roots of Productive Conservation,
Ankeny, Iowa: Soil Conservation Society of America, p. 407-415.
11. Epp, D. and 3. Shortle. 1985. "Commentary: Agricultural Nonpoint Pollution
Control, Voluntary or Mandatory?" 3ournal of Soil and Water
Conservation, 40:1, p. 111-114.
12. Groszyk, W. S. 1979. "Nonpoint Source Pollution Control Strategy", Best
Management Practices for Agriculture and Silviculture, Ann Arbor,
Michigan: Ann Arbor Science Publishers, p. 3-10.
K-X-12
-------
13. Holmes, B. H. 1979. Institutional Basis for Control of Nonpoint Source Pollution,
Environmental Protection Agency, Washington, DC, WH-554.
14. Kasal, 3. 1976. Trade-offs Between Farm Income and Selected Environmental
Indicators, A Case Study of Soil Loss, Fertilizer, and Land Use
Considerations, Washington, DC: Economic Research Service, US
Department of Agriculture, TB #1550.
15. Konrad, 3. G., G. S. Baumann and S. E. Bergquist. 1985. "Nonpoint Pollution
Contgrol: The Wisconsin Experience", 3ournal of Soil and Water
Conservation, 40(1) p. 55-61.
16. Kugler, D. E. 1984. "Variable Cost Sharing Level Program Implications for
Kentucky's 3ackson Purchase Area: An Economic and Policy Study of
Cash Grain and Production Considering Soil Depletion." East Lansing,
Michigan: Department of Agricultural Economics, Ph.D. Dissertation.
17. Libby, L. 1970. "The Political Economy of Water Management: Conceptual
Model and Decision Strategy for the Susquehanna River Basin", Ithaca,
New York: Cornell University Department of Agricultural Economics,
Ph.D. Dissertation.
18. Lindblom, C. 1979. "Still Muddling, Not Yet Through", Public Administration
Review, 39: 517-526.
19. Magette, W. L., R. A. Weismiller and K. C. Gugulis. 1985. "Saving the
Chesapeake: Maryland's Agricultural Education Program", 3ournal of
Soil and Water Conservation, 40(1) p. 79-81.
20. Park, W. and L. Shabman. 1981. Securing Support for Nonpoint Pollution Control
with Local Compensation, Blacksburg, Virginia: Virginia Water
Resources Center.
21. Pope, A., S. Bhide and E. Heady. 1983. "Economics of Conservation Tillage in
Iowa", 3ournal of Soil and Water Conservation, 38(4): 370-373.
22. Reneau, D. R. and C. R. Taylor. 1979. "An Economic Analysis of Erosion Control
Options in Texas", Best Management Practices for Agriculture and
Silviculture, Ann Arbor, Michigan: Ann Arbor Science Publishers, p.
393-418.
23. Ribaudo, M., C. E. Young and D. Epp. 1984. Recreation Benefits From An
Improvement in Water Quality at St. Albans Bay, Vermont, Washington,
DC: Economic Research Service, US Department of Agriculture.
24. Seitz, W. D., C. Osteen and M. C. Nelson. 1979. "Economic Impacts of Policies to
Control Erosion and Sedimentation in Illinois and Other Corn Belt
States", Best Management Practices for Agriculture and Silviculture,
Ann Arbor, Michigan: Ann Arbor Science Publishers, p. 373-382.
25. Unger, D. G. 1979. "Improving Water Quality in Agriculture and Silviculture",
Best Management Practices for Agriculture and Silviculture, Ann Arbor,
Michigan: Ann Arbor Science Publishers, p. 11-16.
K-X-13
-------
26. United States Department of Agriculture and Environmental Protection Agency.
1975. Control of Water Pollution from Cropland; A Manual for
Guideline Development, Washington, DC: Agricultural Research
Service, Report No. ARS-H-5-1.
27. United States Department of Agriculture. 1980. Obion-Forked Deer River Basin,
Tennessee -- A Summary Report. Nashville, Tennessee.
28. United States Department of Agriculure. 1983. Pickaway Soil and Water
Conservation District Resources Inventory.
29. Wade, 3. C. and E. O. Heady. 1977. "Controlling Nonpoint Sediment Sources with
Cropland Management: A National Economic Assessment", American
Journal of Agriculural Economics, 59(1) p. 13-24.
30. White, G. B. and E. J. Partenheimer. 1979. "The Economic Implications of
Erosion and Sedimentation Control Plans for Selected Pennsylvania
Dairy Farms", Best Management Practices for Agriculture and
Silviculture, Ann Arbor, Michigan: Ann Arbor Science Publishers, p.
3*1-358.
31. Young, C. E. and F. A. Teti. 198*. The Influence of Water Quality on the Value
of Recreational Properties Adjacent to St. Albans Bay, Vermont,
Washington, DC: Economic Research Service, United States
Department of Agriculture.
K-X-1.4
-------
Panel Reports
-------
SUMMARY
Panel I
Planning Processes for Nonpoint Pollution Abatement Programs
PANEL MEMBERS
Moderator: John Konrad, Chief, Nonpoint Pollution Section, Wisconsin
Department of Natural Resources, Madison, WI
Panelists:
Richard Homer, Research Assistant Professor, University of Washington,
Seattle, WA
Erhard Joeres, Professor and Chairman, Water Resources Management Program,
University of Wisconsin, Madison, WI
Walter Knisel, Agricultural Research Service, U.S. Dept. of Agriculture,
Tifton, GA
Gary Oberts, Environmental Planner, Metropolitan Council,
St. Paul, MN
Michael Smollen, North Carolina State University at Raleigh,
Raleigh, NC
Jy Wu, Associate Professor, University of North Carolina,
Charlotte, NC
Recorder: David Lee, Marquette University, Milwaukee, WI
INTRODUCTION
Panel discussion and audience participation for Panel I involved plan-
ning processes for Nonpoint Pollution Abatement programs. Although well
documented scientific data exist to support the fact that pollution other
than from "end-of-pipe" or point sources exists, defining what is con-
sidered nonpoint pollution and a plan to remedy its consequences is still
largely undefined. Identifying problems and goals, preparing a technical
and economic attack, and drawing together conclusions to determine the
effectiveness of the program, are essential to the overall success. The
engineering, scientific and planning communities are dealing with such
immense pollution problems taking place in broad, complex ecosystems,
that no one person or approach can completely solve them. This panel
has quickly concluded that general planning steps (see attached flow
sheet) are necessary to implement a regional nonpoint pollution abatement
program. However, one must be aware that program needs will vary greatly
from one region to the next and, therefore, a wide variation in applica-
tion and achievement in different programs can be expected.
FINDINGS
Many extremely important issues were raised concerning proper and effective
P-I-1
-------
program planning and implementation. The following issues, aspects and
conclusions were found to be the most significant:
1. Widely accepted means of defining nonpoint pollution problems are
missing. Water quality standards, per se, such as are employed in
point source work, do not exist and probably will not exist for some
time to come. Criteria based on receiving water, beneficial use
preservation, may represent the best strategy but more understand-
ing of the effects of nonpoint discharge is required to implement
it.
2. Substantial nonpoint source research findings are already available
to partially support the development of problem assessment methods,
although little effort has been made to do so. Further progress in
this area will require much additional study. Success in these forth-
coming studies will require programs in which great attention has
been given to definition of general and specific objectives, and
prioritization of program elements to achieve the maximum gain in
knowledge for the resources expected.
3. In the nonpoint source water pollution field, unlike many other
scientific and technical disciplines, there is little standardiza-
tion of investigation methods. Greater standardization should be
promoted to allow more interchange among investigators and reduce
at least some of the duplication of effort now occurring.
4. Dealing with superimposed point and nonpoint loadings to remedy the
cumulative, concurrent pollution burden in receiving waters must
be addressed.
5. Nonpoint pollution assessment via the use of mathematical models
is an indispensible tool, but is rendered useless without well or-
ganized sampling and data collection efforts. In addition to the
site monitoring that takes place preceding model development and cali-
bration, a continued long-term monitoring effort must be made to
verify the model and evaluate any control techniques that are im-
plemented. The time lag between project implementation and expec-
ted results is politically unattractive and can be a very sensitive
issue.
6. Effective, well-defined legislation regulations are imperative to
research project goals, but the effort is merely rote if proper
enforcement does not exist.
7. The effects of nonpoint pollution loadings to groundwater must be
assessed. This investigation must be concerned with trade-offs
between surface water improvement and increased groundwater con-
tamination resulting from strategies promoting infiltration of non-
point runoff. More research is needed to determine the extent of
chemical and biological pollutant reduction processes occurring with-
in the reduction processes occurring within the soil layer. If these
mechanisms are overlooked, erroneous conclusions may be reached
regarding hydraulic and pollutant loading rates that can be permit-
ted on land.
P-I-2
-------
8. The impact of nonpoint loadings from the atmosphere appears to be
an area in which little research exists and which poses a great
problem to nonpoint water pollution managers. The political and
institutional implications are many and could be extremely complex.
9. Our society works on an action/reaction basis. The professional
community must determine what action(s) must be undertaken in order
to make the public aware of the extent of nonpoint water pollution
and to stimulate the reaction that is needed, i.e., support for
study and abatement programs.
10. It appears at this point that the engineering and technological
hurdles are lower than those encountered by planners. Identifying
and researching nonpoint pollution problems in "critical areas" is
essential, but projects based in these areas are doomed to fail if
public participation and cooperation are not acquired.
CONCLUSIONS
It is the consensus of the panel that, although researchers are well aware
of nonpoint pollution problems and many of their implications, very little
planning and program implementation has taken place. Solving our point
source pollution problems has been difficult but the challenge, politi-
cally, technically, and economically, to improve the overall quality of
our environment has just begun. Public awareness and cooperation is essen-
tial and the overall challenge will stem more from sociological than tech-
nical feasibility problems. Essentially, the nature of the problem stems
from the fact that water moves. In its movement, it does not recognize
municipal, state or any other political boundaries. Truly, what happens
in Minneapolis affects those living hundreds of miles away in New Orleans.
What programs will be implemented to halt or reduce these problems, who
will pay and how much and what level of political authority will take
precedence in assuring continued environmental regulation, compliance
and enforcement,are all issues that were examined in our panel discussion
and must continue to be addressed.
P-I-3
-------
»! Problem Identification
Goal Definition
Target Selection
Plan Preparation
Data Collection
\
f
Problem Reevaluation
Management Practice
Evaluation
Institutional and Financial
Formulation
Goal Realization
Planning Process Steps For Abatement Of Nonpoint Pollution,
From "Fundamentals of Watershed Management",
Presented by Gary Oberts at the International
Symposium on Lake and Reservoir Management,
North American Lake Management Society, Oct.16-19,
1984, McAfee, N.J.
P-I-4
-------
IMPROVEMENT OF ENVIRONMENTAL IMPACT ASSESSMENT
OF NONPOINT SOURCES OF WATER POLLUTION
Richard R. Horner
Research Assistant Professor* Department of C1v1l Engineering
University of Washington
Seattle, Washington, U.S.A.
ABSTRACT
Abatement of nonpolnt source water pollution problems relies upon the
ability to assess the dimensions and effects of those problems. This
ability 1s not highly developed, and advancing 1t 1s hindered by difficult
problems. Nevertheless, progress can be made by systematically building on
a research foundation to develop Impact assessment strategies providing the
necessary basis for decision-making. This paper discusses examples of such
strategies from our research. Included are techniques to design cost-
effective monitoring programs to gather nonpolnt source data and to carry
out stormwater runoff sampling. Also discussed 1s a stepwlse guide to
assess quantitatively the aquatic Impacts of operating highways and to
allocate resources to their mitigation. These approaches offer means of
overcoming a number of the problems Impeding nonpolnt source Impact
assessment and have potential application 1n other settings.
P-I-A-1
-------
INTRODUCTION
The passage of federal and state water pollution control legislation 1n the
past 15 years has spurred widespread actions to Improve surface water
quality 1n the United States. For much of this period the majority of the
attention has been directed toward control of point sources of water pollu-
tion. Continuing water pollution problems 1n cases where some success has
been achieved 1n reducing point source contributions has suggested, how-
ever, that more distributed sources also are major factors 1n surface water
quality. These distributed contributors have been termed nonpolnt sources
and are associated with stormwater runoff from land surfaces put to various
uses.
A prerequisite of acting effectively to solve nonpolnt source water pollu-
tion problems 1s to be able to assess their dimensions and their Influence
on affected ecosystems. This process of environmental Impact assessment
has a fifteen-year history stemming from the passage of the National Envi-
ronmental Policy Act (NEPA). Despite this rather lengthy period, little
agreement has developed concerning how to conduct assessments 1n nonpolnt
source water pollution cases. The underlying legal and regulatory frame-
work provides little distinct guidance, and practitioners have not had
discussions 1n the breadth and depth needed to adopt and disseminate proven
techniques and to promote standards of quality. As a result, nonpolnt
source Impact assessments, much too frequently, cannot provide meaningful,
objective problem analyses that can support effective decision-making.. In
fairness, 1t should be noted that our field of Interest 1n this symposium
1s certainly not alone 1n this state.
Much of the applied environmental research performed since the adoption of
NEPA has had as an Implicit objective the enhancement of abilities to
conduct environmental Impact assessments. While substantial data bases and
many useful methods have resulted from these efforts, there have been few
attempts to apply the knowledge gained to developing comprehensive assess-
ment protocols. Even more rare has been Implementation of these research
results 1n the practices of organizations that must prepare or evaluate
environmental Impact statements.
It 1s the central premise of this paper that environmental Impact assess-
ment of nonpolnt sources of water pollution can be Improved by developing
systematic methods firmly rooted 1n research results. Further, the
research community should take the lead 1n formulating these methods and
should work closely with practitioners and regulators to Implement them.
At the University of Washington we have been addressing these Issues 1n our
nonpolnt source work. This paper will discuss the problems we have recog-
nized, the steps we have taken to solve them, and where we see the need for
more widespread efforts. We believe our general approaches have potential
application 1n other locations, to different nonpolnt problems, and 1n many
other situations.
PROBLEMS IN NONPOINT SOURCE WATER POLLUTION IMPACT ASSESSMENT
Developing means of assessing environmental Impact necessitates some degree
P-I-A-2
-------
of generalization* whereby knowledge gained 1n one Instance can be applied
with acceptable reliability at another time and, often, at a different
place. The diffuse nature of the nonpolnt source water pollution problem
creates numerous and significant difficulties 1n meeting this necessity 1n
the presence of the limitations on financial resources that always exist.
These difficulties arise at every step 1n the process, from data collec-
tion, through data analysis and Interpretation, to Implementation of re-
sults 1n usable form.
Probably, some progress could be made 1n systematizing nonpolnt assessment
by applying methods already developed and reconsidering existing data with
this objective 1n mind. Great progress, however, will require substantial
new data collection. Planning effective programs for this purpose 1s the
first obstacle. In any sizable nonpolnt source water pollution problem,
the potential subjects for monitoring are so extensive that full coverage
1s Impossible. In the past, there has been little consideration of how to
select scientifically among all the possible tasks and how to allocate the
available resources. Further, there 1s little standardization 1n the
nonpolnt source water pollution field of specific sampling and data
analysis techniques. Therefore, results frequently are not transferable
from one time and place to another.
Natural conditions encountered 1n nonpolnt source water pollution work
reinforce and add to these problems. The extreme variability documented
(Wan1el1sta, 1978; Homer and Mar, 1982) points out that nonpolnt water
pollutant concentrations and mass loadings are very site- and time-speci-
fic. Part of this reported disparity undoubtedly is due to method Incon-
sistency, but variations 1n topography, meteorology, climatology, soils,
geology, specific features of land use, and other factors seem to make
great dispersion 1n the results of nonpolnt monitoring an inherent charac-
teristic. The result is that storm runoff usually has a greater diversity
1n contaminants than Industrial and municipal sewage effluents, with less
predictability of their levels. In addition to the measurement of physical
and chemical constituents of runoff 1s the Issue of whether receiving water
biota should be observed la sJt.u or la vitro. All of these considerations
expand the 11st of operative variables beyond a measurable number.
Intertwined with all of the problems noted above are the issues of long-
term versus short-term Impacts of nonpolnt water pollution sources and the
cumulative effects of a number of activities 1n a watershed. As 1n other
endeavors in the environmental sciences, most of the existing understanding
is based on short-term observations of the acute effects on aquatic biota
of high levels of single contaminants. Low-level, longer-duration expo-
sures, synergistlc effects of multiple pollutants, responses of communities
of species, and how human actions Influence the accumulating burden on
receptors are all subjects that have received little study.
Assuming that these problems of data collection and analysis can be over-
come, putting results to use in impact assessment is stll 1 a formidable
problem. Not all occurrences will have significant Impacts, and analysts
should have means of distinguishing those that almost certainly will not
from those that may. Most of the analytical effort then could be directed
toward the latter. In this task, analysts need techniques that can produce
P-I-A-3
-------
an objective and fully justifiable assessment 1n a convenient fashion. The
limited resources available for Impact mitigation should be allocated to
actual problems Identified through such a process.
Although the difficulties with nonpoint assessment described are daunting*
some strategies are developing to make them manageable. The next section
will present our major contributions to this progress and the remaining
needs that we bel1eve exist.
APPROACHES TO IMPROVING NONPOINT SOURCE WATER POLLUTION IMPACT ASSESSMENT
Monitoring of Nonpoint Sources
We have just begun the development of a methodology to guide the design of
monitoring programs for assessing nonpoint source water pollution problems
throughout Washington State. This effort 1s an extension of work we have
completed for the electric power Industry (Mar et al.» 1985). We contem-
plate creating a conceptual framework to assist a user 1n Identifying the
potential causes of Impact and subjects for monitoring and 1n formulating
hypotheses to test 1n monitoring. A central feature of the framework will
be a procedure (based on eigenvalue mathematics) to rank monitoring sub-
jects and hypotheses with respect to designated criteria* so that the many
possible variables can be reduced to those most likely to yield the needed
Information. Another feature will be a computational algorithm to deter-
mine the optimum sampling program design (number of stations* frequency*
and replication) for a cost constraint or, alternatively, a constraint, on
statistical power. The product of this research 1s expected to focus
Washington State nonpoint source monitoring programs on definite objectives
and to Improve their cost-effectiveness. We believe similar approaches are
possible and advisable elsewhere.
In the absence of widely agreed upon specific nonpoint sampling procedures*
we have given considerable attention to this subject. Upgrading the
quality of the methods used and promotion of 1nterchangeab1l1ty of results
will not occur until a standard-setting body outlines and promulgates
standards that all will adopt. Increasing the standardization of experi-
mental practices 1s a major prerequisite to the Improvement of nonpoint
source Impact assessment.
The most common means of sampling stormwater runoff has been to collect a
number of samples at discrete points throughout the runoff period using
automatic equipment. This strategy produces a relatively complete record
but 1s costly 1n both sampling equipment and laboratory analysis. We
developed an Inexpensive technique to fractionate and composite runoff from
an entire storm, thus reducing both costs greatly (Clark et al.» 1981).
Our strategy 1s to perform enough discrete sampling through storms to
understand maximum concentrations, followed by storm composite sampling to
assemble a large whole-storm data base. The first data set allows assess-
ment of Intense, acute effects, which we also Investigate with laboratory
bloassays (Portele, 1982). The large latter set assists 1n resolving the
problems of temporal and spatial variability discussed earlier and permits
both deterministic modeling and statistical analysis, as described below.
P-I-A-4
-------
Protocols for Impact Assessment
We conducted a five-year study of storm runoff from operating highways 1n
Washington State that concluded with the development of a stepwlse guide
for assessment of aquatic Impacts (Mar et al.» 1982; Horner and Mar, 1982;
1983). This guide has several features that Illustrate our approach to
systematic nonpolnt Impact assessment based on research results. First,
Its organization 1s hierarchical, so that cases having differing potentials
for significantly Impacting aquatic ecosystems can be distinguished and
proportionately allocated resources for problem assessment and solution.
An Initial screening process, based on specific criteria, removes those
cases highly likely to create an Insignificant Impact from further consi-
deration. For those cases analyzed further, the basic premise of the
assessment 1s that the highway Impact on the receiving water can be
assessed most realistically 1n the context of the aggregate burden created
by all activities 1n the watershed. This approach attempts to deal with
the Issue of cumulative effects, although much more Insight 1s necessary
before this phenomenon can be represented adequately.
The large storm composite data base resulting from the research was
employed to develop a model of accumulated pollutant loadings delivered by
highway runoff to the receiving water over an extended period (Asplund et
al., 1982; Chu1 et a!., 1982). Variability among Individual storm events
was too great for deterministic modeling, but the data were applied suc-
cessfully to develop a probabilistic method for assessing short-term
effects (Little et al., 1983). This approach exploits the log-normal
distribution of the individual storm data and permits the impact analyst to
estimate the frequency with which a given pollutant concentration, such as
a water quality criterion, would be exceeded in a receiving water as a
result of highway runoff. A similar technique was developed by the U.S.
Environmental Protection Agency (1982) for assessing the effects of urban
runoff. It also was recommended by Loftis et al. (1983) in a general
review of statistical models that might be applied 1n water quality regula-
tion. Modeling of cumulative pollutant generation and probabilistic
evaluation of Individual events allows estimation of both long- and short-
term Impacts.
The guide offers an opportunity to forecast potential aquatic Impacts of
highway projects at an early stage of development and to allocate resources
for impact mitigation on the basis of need. This advance Improves the
cost-effectiveness of stormwater runoff management. We have worked with
the Washington State Department of Transportation to Implement the use of
the Impact assessment guide in Its practices.
SUMMARY AND CONCLUSIONS
Assessment of the impacts of nonpolnt sources of water pollution is rather
poorly developed, and formidable problems impede its advancement. We
believe these problems can be overcome by systematically applying research
results to formulate assessment strategies that will yield objective,
decision-making criteria. Such strategies are needed in the areas of both
data collection and its application to analyze potential impacts. In the
former category, we are developing an instrument to identify all of the
possible elements of Washington State nonpolnt monitoring programs and to
P-I-A-5
-------
select the priority ones. In previous work we devised a strategy and tools
to collect representative stormwater runoff data more cost-effectively than
with traditional techniques. Greater standardization of methods is a
definite need in the nonpoint source monitoring field. With regard to
application of research data to formulate impact assessment protocols* we
have developed a guide allowing Washington State highway designers to
identify and scope actual environmental problem areas and to apply abate-
ment resources accordingly. This guide reflects concern for cumulative and
both short- and long-term impacts to the extent supported by current know-
ledge. We recommend implementation of similar strategies in other loca-
tions and in other nonpoint source problem areas.
REFERENCES
Asplund, R.L., Ferguson, J.F., and Mar* B.W. (1982). Characterization of
Highway Runoff in Washington State. J. Environ. Engin. Div. ASCE, 10J?,
391-404.
Chui, T.-W., Mar, B.W., and Homer, R.R. (1982). A Pollutant Loading Model
for Highway Runoff. J. Environ. Engin. Div. ASCE, 108, 1193-1210.
Clark, D.L., Asplund, R.L., Ferguson, J.F., and Mar, B.W. (1981).
Composite Sampling of Highway Runoff. J. Environ. Engin. Div. ASCE, IfiJ,
1067-1081.
Horner, R.R., and Mar, B.W. (1982). Guide for Water Quality Impact
Assessment of Highway Operations and Maintenance, FHWA WA-RD-39.14.
Department of Civil Engineering, University of Washington, Seattle.
Horner, R.R., and Mar, B.W. (1983). Guide for Assessing Water Duality
Impacts of Highway Operations and Maintenance. Iran sportation..Research
Record, 9J£, 31-40.
Little, L.M., Horner, R.R., and Mar, B.W. (1983). Assessment of Pollutant
Loadings and Concentrations in Highway Stormwater Runoff, FHWA WA-RD-
39.12.1. Department of Civil Engineering, University of Washington,
Seattle.
Loftis, J.C., Ward, R.C., and Smillie, G.M. (1983). Statistical Models for
Water Quality Regulation. J. Water Pollut. Control Fed., 55, 1098-1104.
Mar, B.W., Horner, R.R., Ferguson, J.F., Spyridakis, D.E., and Welch, E.B.
(1982). Summary - Washington State Highway Runoff Water Quality Study,
1977-1982, FHWA WA-RD-39.16. Department of Civil Engineering, University
of Washington, Seattle.
Mar, B.W., Lettenmaier D.P., Horner, R.R., Richey, J.S., Palmer, R.N.,
Mil lard, S.P., and Mackenzie, M.C. (1985). Sampling Design for Aquatic
Ecological Monitoring, Volume 1: Summary Report. Report to Electric Power
Research Institute by Department of Civil Engineering, University of
Washington, Seattle.
Portele, G.J., Mar, B.W., Horner, R.R., and Welch, E.B. (1982). Effects of
Seattle Area Highway Stormwater Runoff on Aquatic Biota, FHWA WA-RD-39.11.
P-I-A-6
-------
Department of C1v1l Engineering, University of Washington, Seattle.
U.S. Environmental Protection Agency (1982). Preliminary Results of the
Nationwide Urban Runoff Program, Vol. 1. USEPA, Water Planning Division,
Washington, D.C.
Wan1el1sta, M.P. (1978). Stormwater.Management, Quantity and.Qua!1ty. Ann
Arbor Science Publishers, Inc., Ann Arbor, Michigan.
P-I-A-7
-------
QUANTIFICATION OF NONPOINT SOURCE POLLUTION
Jy S. Wu
Associate Professor
Department of Civil Engineering
University of North Carolina at Charlotte
Charlotte, North Carolina 28223
ABSTRACT
Storm runoff pollution can be characterized, in magnitude and in
concentration of pollutans, as intermittent and impulsive-type discharges
into receiving waters, causing shock loading problems to the ecosystem of
these water bodies. The classical approach, using critical low flow as the
design criterion for water quality management schemes, must include the
effect of storm runoff. This paper presents state-of-the-art methodologies
for quantifying storm runoff loads. Four categories of assessment methods
are presented: these include zero-order, rational, statistical and
simulation methods, for achieving different levels of prediction, i.e. (i)
average annual storm load, (ii) storm load per event and (iii) storm load
distribution within events.
P-I-B-l
-------
QUANTIFICATION OF NONPOINT SOURCE POLLUTION
Jy S. Wu, Ph.D., P.E.
INTRODUCTION
In recent years, pollution from nonpoint sources has become an
increasingly important consideration in water quality planning and
management. Recognizing that point sources are obvious in nature and
relatively easy to identify and characterize, the remaining problem is the
evaluation of storm-related or nonpoint source pollution.
naturally arises: what techniques of varying cost,
The question
accuracy, and
sophistication are available for achieving a successful assessment program ?
Characteristics of storm runoff from various sources have been
reported extensively in the literature (McElroy and Bell, 1974; Wu and
Ahlert, 1976 ; Novotny and Chester, 1981). In genreal, storm runoff
pollution can be characterized, both in magnitude and in concentration of
pollutants, as intermittent and impulse-type discharges into receiving water
bodies. The classical approach, using critical low flow as the criterion
for water quality management, must take the effect of storm runoff into
account. This paper presents state-of-the-art methodologies for defining
and quantifying storm runoff pollution.
LEVELS OF PREDICTION
Various levels of refinement can be employed to define storm runoff
pollution. They range from a relatively simple yearly loading to quite
detailed descriptions that introduce temporal variation into the analysis.
Three levels of detail can be defined (U.S. EPA, 1976; Wu and Ahlert, 1978a
and 1978b).
Level 1: Average Annual Storm Load
This level of detail defines the average annual storm-generated loads
as if they occur continuously, during both wet and dry periods. It is used
to assess the cumulative long-term effects, e.g. sediment depostion in.
reservoirs, and may be used to make comparisons with continuous municipal
and industrial point-source loadings and describe the relative magnitudes
from each source type.
Level 2; Storm Load per Event
This level of detail considers the actual temporal distribution of
storm event, including variability of storm loads from one event to the
other.
Level 3: Storm Load Distribution Within Events
This level of prediction describes the actual runoff loading rate, as
a function of time, within each event. It allows the evaluation of
transient and shock-loading impacts on receiving streams.
P-I-B-2
-------
In general, level 1 is an integral form of level 2 and, in turn, level
2 is an integral form of level 3. These classifications provide useful
guidelines for water quality planners to define the necessary levels of
stormwater impact anlaysis, and to select appropriate assessment techniques.
For example, in characterizing storm loads, an average annual loading rate
is useful as a measure of the relative impact of storm runoff. In the case
of sediment deposition, the cumulative effect of all storms is more
significant than the effect of a single storm or a specific period within a
storm. In the case of transient water quality impact analysis, e.g. oxygen
depletion of receiving waters during storm conditions, the long-term average
loading becomes much less important, because it does not provide enough
information to determine how often stream water quality standards are
violated.
ASSESSMENT TECHNIQUES
Numerous methods have been used to estimate the magnitudes of storm
runoff loads, for both urban and nonurban areas and for long-term as well as
short-term predictions. These methods can be categorized as follows.
Zero-Order Method
This is the most crude and least expensive method. Storm runoff loads
from an area are estimated from data in the literature. The method is
inaccurate in most cases and applicable only to long-term predictions. .This
is due to the questionable transferability of data from areas of different
climate, topography and management practices, etc. The term "zero-order"
indicates the relative flexibility of estimation from literature data.
Rational Method
The Rational equation, Q=CIA, for predicting peak runoff rate is known
by every water resoucres engineer. The rational type of equation takes the
form of a product of several independent variables. If storm runoff and
pollutant concentrations are independent, the mean runoff loading rate, w,
will equal to the product of a mean concentration, c, and the mean
discharge, q :
w = q . c (1)
Assuming "q" can be estimated from basic hydrology, "c" can then be
calculated from:
(a) published data from a matched watershed;
(b) the computed flow-weighted average concentration, or a
representative concentration, obtained over a period of grab
sampling (Whipple et al., 1976);
(c) an equation which relates "c" to land use, population density,
and street sweeping effectiveness (U.S. EPA, 1976). Street
sweeping has been a subject of debate for its effectiveness
in reducing nonpoint pollution (Novotny, 1983).
P-I-B-3
-------
Typical examples of the rational type method include the Univeral Soil
Loss Equation (Wischmeimer and Smith, 1965} and its modified form such as
the loading function approach (McElroy et al., 1976). The latter approach
can be used to predict sediment yield from a watershed, while other water
quality parameters are related as fractions of the sediment load.
Statistical Methods
Statistical methods employ techniques such as regression, correlation,
frequency analysis, etc., to predict storm runoff loads.
A simple multi-regression equation was employed to predict total
pollutant loads for single events (Wells et al., 1971). The variables
included in the regression equaiton are rainfall volume, storm duration,
time between storms, volume and duration of last rainfall. Heaney et al.,
(1976) developed correaltion equations for predicting annual average
loading rates for each pollutants, as a function of land use, precipitation
and population density. This method is proposed as a desk-top procedure for
estimating the quality of urban runoff from combined, storm, and unsewered
portions of urban areas.
In most cases, it is important to consider high-flow pollutant laoding
as a phenomenon separate from low-flow pollutant loading. Paterson (1977)
developed the idea of a "concentration matrix" for estimating in-stream
storm runoff loading. The matrix is a 4x4 square matrix having 16 blocks.
These blocks contain a specific pollutant concentrations representative, of
the 4 seasons of the year and , also, the 4 flow ranges of low, medium,
medium high and high. The representative concentration in each block is
obtained by averaging a large number of water quality data (from STORET, for
example) in that particular flow range at the corresponding season of the
year. For each day, the storm runoff loads are determined by multiplying
the daily flow data and the respective concentrations in the matrix. Daily
loads can be summed to derive the seasonal or annual loads.
Wauchope and Leonard (1980) employed regression analysis to estimate
the maximum edge-of-the-field concentration of pesticides from an individual
event. The regression equation relates pesticide concentration to
application rate, time of runoff since application and an availability
factor describing the pesticide-soil interaction. The accuracy of
prediction using regression equations is usually limited by the range of
numerical values of the variables originally used for deriving the
equations.
Simulation Methods
Simulation methods deal with deterministic rainfall-runoff mechanisms
that affect the quality of runoff within a watershed.
Sartor and Boyd (1972) defined dust-and-dirt accumulation on street
surfaces as a major source of pollution in urban runoff. Later developments
involve the use of their results as basic guidelines for the prediction of
P-I-B-4
-------
pollutant accumulation rates on street surfaces, prior to a storm event.
Accumulated dust-and-dirt is subject to a first-order washoff mechanism by
rainwater. Other pollutant concentrations are predicted, as fractions of
the total solid loads, as described in the original version of SWMM (Metcalf
& Eddy et.al., 1971). Sutherland and McCuen (1974) related the accumulation
rate of each quality constituent to land use, pavement type and condition,
traffic volume and length of time since rainfall or street sweeping. The
rate constant of the first-order accumulation equation was correlated with
pavement condition, for single family and muti-family residential areas, and
with traffic volume, for industrial and commercial areas.
The use of homogeneous land use approach in conjunction with the
concepts of pollutant accumulation and first-order washoff was demonstrated
by Wu and Ahlert (1978a). The first-order washoff model is calibrated
against the various types of homogeneous or single land-use subareas, to
obtain estimates of the two parameters, i.e. initial loading and washoff
rate constant. The calibrated model can then be applied to the mixed land-
use areas. In this case, the initial loading and the washoff constant are
evaluated as composite, weighted values representing the different
percentages of land uses for the whole area. Output will be estimates of
storm runoff loadings for single events and the time distribution of
pollutants within events. This approach minimizes the drawbacks of dust-
and-dirt concept and sediment-related loads; model parameters are estimated
independently for each quality constituent. However, it requires a
relatively intensive sampling effort to obtain a detailed time history of
runoff flows and quality, from a number of single land-use subareas and for
a number of storms.
Donigian and Crawford (1976) modeled sediment yield from land
surfacess including the processes of net daily accumulation of sediment as
dust-and-dirt, detachment of particles by raindrop impact on fine sediment
and soil, and transport of sediment by overland flow. Other quality
parameters are related as fractions of the sediment loads. Wu (1980)
modified this approach and assumed the BOD load consists of two parts, a
sediment-related part and a soluble part. The soluble BOD load is subject
to the first-order washoff by surface runoff. A stormwater assessment model
was developed to interface with the SCS TR-20 hydrology program.
A storm runoff model can also be constructed to perform continuous
simulation. The use of continuous simulation is to supplement the limited
capability of field sampling programs in assessing the watershed responses
to hydrologic conditions and management practices, over an extended period
of time. One of the difficulties encountered in continuous simulation is
the lack of adequate data-base for calibration and verification. In many
cases, the collected data contain missing records and thus the use of
statistical technique will be needed to generate these missing records (Wu,
1983). Nevertheness, the use of continuous simulation will provide results
that can be used to develop probabilistic pollutant loading to assess the
cost effectiveness of various storm water management alternatives at an
acceptable level of storm runoff load.
P-I-B-5
-------
CONCLUSION
Various levels of detail and analysis of storm runoff pollution have
been highlighted in this paper. However, as new problems will be emerged
the assessment methodology will have to be refined. One such example is the
uncontrolled discharge of landfill leachate that represents a new horizon
of nonpoint source pollution (e.g. see Wu and Hilger, 1984). Future
research on nonpoint source pollution should include the development of
assessment programs for those nonpoint sources other than the conventional
ones such as combined sewer overflow, agricultural or urban runoff etc.
REFERENCES
Donigian, A.S. and N.H. Crawford (1976). Modeling Nonpoint Pollution From
the Land Surface. EPA-600/3-76-083, U.S. Environmental Protection
Agency.
Heaney, J.P., W.C. Huber and S.J. Nix (1976). Stormwater Management Model
Level 1; Preliminary Screening Procedures. EPA-600/2-76-275, U.S.
Environmental Protection Agency.
McElroy, F.T.R. and J.M. Bell (1974). Stormwater Runoff Quality for Urban
and Semi-Urban/Rural Watersheds. PB-231, NTIS.
MeElroy, A.D. et al. (1976). Loading Functions for Assessment of Water
Pollution From Nonpoint Sources. EPA-600/2-76-151, U.S. Environmental
protection Agency.
Metcalf & Eddy, Inc., University of Florida and Water Resources Engineers,
Inc. (1971). Storm Water Management Model, Volume 3. User's Manual.
WPCRS 11024 DOC 09/71, U.S. Environmental Protection Agency.
Novotny, V. and G. Chesters (1981). Handbook of Nonpoint Pollution: Sources
and Management. Van Nostrand Reinhold Co.
Novotny, V. (1983). Effectiveness of Low Cost Practices for Urban Runoff
Pollution Control"!A minicourseinstructionmaterialpresentedat
International Symposium on Urban Hydrology, Hydraulics and Sediment
Control, University of Kentucky, Lexington.
Paterson, R.B. (1977) A Preliminary Investigation of the Phosphorus Laoding
Characteristics of Lake Carnegie, Princeton, New Jersey. B.S. Thesis,
Rutgers University, New Jersey.
Sartor, J.D. and G.B. Boyd (1972). Water Pollution Aspects of Street
Surface Contaminants. EPA-R2-72-081, U.S. Environmental Protection
Agency.
Sutherland, R. and R. McCuen (1975). A Mathematical Model for Estimating
Pollution Loadings in Runoff from Urban Streets. in Proceedings
International Conference on Mathematical Models for EnvTronmental
Problems, edited by C.A. Brebbia, John Wiley and Sons.
U.S. EPA (1976). Areawide Assessment Procedure Manual. EPA-600/9-76-014,
U.S. Environmental Protection Agency.
Wauchope, R.D. and R.A. Leonard (1980). Maximum Pesticide Concentrations in
Agricultural Runoff: a Semiempirical Prediction Formula, J. Environ.
QuaJL, 9(4), 665-672.
Wells, D.M., T.A. Austin and C.C. Cook (1971). Variation of Urban Runoff
with Duration and Intensity of Storms. PB-204 235, NTIS.
P-I-B-6
-------
Whipple, W. Jr., J.V. Hunter and S.L. Yu (1976). Characterization of Urban
Runoff: New Jersey. Water Resources Research Institute, Rutgers
University, New Jersey.
Wischmeier, W.H. and D.D. Smith (1965). Predicting Rainfall-Erosion Losses
from Cropland East of the Rocky Mountains.Agricultural Handbook 282,
U.S. Department of Agriculture.
Wu, J.S. and R.H. Ahlert (1976). State-of-the-Art Review: Nonpoint Source
Pollution. Technial Report WRE-76-3, College of Engineering, Rutgers
University, New Jersey.
Wu, J.S. and R.C. Ahlert (1987a). Assessment of Methods for Computing Storm
Runoff Loads. Water Resources Bulletin, 14(2), 429-439.
Wu, J.S. and R.C. Ahlert (1978b).Prediction and Analysis of Stormwater
Pollution. in International Symposium on Urban Storm Water
Management, University of Kentucky, Lexington, 183-188.
Wu, J.S. (1980). Development and Application of a Storm Water Assessment
Model. Doctoral dissertation, Rutgers University, New Jersey.
Wu, J.S. (1983). Data Management for Continuous Hydrologic Simulation, in
Proceedings of Stormwater and Water Quality Model, edited by T.O.
Barnwell, EPA-600/9-83-015, U.S. Environmental Protection Agency, 161-
176.
Wu. J.S. and H. Hilger (1984). Evaluation of EPA's Hazard Ranking System.
J. of Environ. Eng., ASCE, 110(4), 797-807.
P-I-B-7
-------
SUMMARY
Panel II
Legislative Means and Financing of
NPS Control Programs
PANEL MEMBERS
Moderator: James Arts, Dept. of Agriculture, Trade and Commerce Protec-
tion, Madison, WI
Assistant Moderator: Jim Peterson, Attorney, Milwaukee Metropolitan
Sewerage District
Panelists:
Kathleen Falk, Public Intervenor, Department of Justice,
Madison, WI
Lawrence Libby, Professor of Agricultural Economics, Michigan State
University, East Lansing, MI
Kathleen Segerson, Assistant Professor of Agricultural Economics, Uni-
versity of Wisconsin, Madison, WI
Ann Weinberg, U.S. EPA, Nonpoint Pollution Section, Washington, DC
Recorder: James D'Antuono, Southeastern Wisconsin Regional Planning
Commission, Waukesha, WI
INTRODUCTION
The charge of this panel was to investigate the means of legislative
treatment for nonpoint source pollution control, to suggest how to treat
this with legislation, and to suggest how to finance the control of non-
point source pollution. The discussions were wide ranging, expressing
a broad spectrum of views. The spectrum was so broad, in fact, that
distillation of the matters discussed into findings agreed to by all
is impossible. Accordingly, this report must necessarily reflect only
the ideas that received the most uniform approbation.
FINDINGS
General
1. The panel and the attendees uniformly reaffirmed that legislation
is the means to achieving nonpoint source pollution control.
2. There is a not-entirely-understood interrelationship between ground-
water protection and nonpoint source pollution control programs.
P-II-1
-------
3. Legislation seems most appropriate on a national or federal basis
to avoid regional disadvantages. EPA, however, eschews any national
legislative policy, in favor of state and local control of the problem.
4. Substantial social benefits from nonpoint source control measures
were perceived. Society would be positively benefited from the crea-
ation of incentives to control nonpoint source pollution.
5. Financial incentives for nonpoint source pollution control in volun-
tary programs are generally inadequate; some form of regulatory im-
perative is needed.
Legislative Means
6. Legislation requiring nonpoint source pollution control is the pre-
ferred means to generate a "community will" to achieve the desired
levels of nonpoint source pollution control.
7. Legislation should avoid water quality standards as the medium for
achieving nonpoint source pollution control because the benefits of
any control program are quite difficult to identify in strictly
water quality terms,and enforcement of a violated water quality
standard against a particular source of nonpoint pollution is prob-
lematic and ineffective.
8. Legislation should either identify best management practices or
authorize administrative determination of best management practices
that are not related to performance standards. The authorizing
legislation should also contain provisions requiring compliance.
9. Efforts should be made to attach nonpoint source pollution control
onto legislative efforts aimed at other environmental problems.
For example, research money could be garnered from toxic control
legislation.
1.0. Legislation should require affirmative efforts toward nonpoint
source pollution control as a condition of other environmental
grant programs.
Financing
11. Funding of nonpoint source pollution control is logically part of
the public support for a clean, safe and healthy environment. Accord-
ingly, nonpoint source control programs should be attached to other
environmental projects well received by the public for funding. For
example, erosion control was added to park development on a recent
Missouri Constitutional Amendment.
12. Use of sales tax money garnered from food purchases, at rates as
low as .1%, can generate needed research and implementation moneys.
13. "No-cost" control programs can be found by merely reviewing present
governmental practices. For example, in new residential areas,
abandonment of curb and gutter requirements can alleviate storm
P-I1-2
-------
water run-off.
A Contrary View
14. A thought-provoking paper was presented which utilized a perform-
ance standard predicated upon achieving designated water quality
standards, with a responsible group being required to pay taxes
when the standards are not achieved and being provided bonuses or
rewards when the standards are achieved. Although a majority of the
panel and attendees did not concur, the approach illustrates that
control of nonpoint source pollution may ultimately be found in
creative use of economic reward and penalty.
CONCLUSION
The panel concludes, federal regulation which establishes best manage-
ment practices is needed. Funding must come by "piggy-backing" nonpoint
source pollution control onto other programs and having funds earmarked
for nonpoint source pollution.
P-II-3
-------
LEGAL/REGULATORY APPROACHES
TO CONTROL OF NON-POINT SOURCES
OF GROUNDWATER CONTAMINATION
Robert F. Ehrhardt
Applied Management Sciences, Inc.
Silver Spring, Maryland
ABSTRACT
The U.S. Environmental Protection Agency Issued Us Groundwater Protection
Strategy 1n August 1984. The strategy outlined the agency's proposed approach
to addressing both point and non-point sources of groundwater contamination.
Non-point sources receiving attention In the Strategy Include leaking
petroleum and chemical storage tanks and chemical use, handling, and storage
(Including pesticides) over certain groundwater formations. The strategy
envisions restrictions or bans on such activities or substances particularly
over groundwater that 1s "ecologically vital" or an "Irreplaceable source of
drinking water."
Because state groundwater quality programs will be required to reflect the
provisions of EPA's Strategy to some, as yet undetermined extent, a survey was
conducted of existing legal/regulatory approaches to groundwater quality
management at the state level of government Including programs to control
potential non-point sources of groundwater contamination, primarily pesticide
and other chemical handling and storage. Few state programs were Identified
which regulate such sources to a degree commensurate with the EPA Strategy's
approach. Significant changes In state legal/regulatory approaches may be
required for states to conform to this component of EPA's Strategy.
P-II-A-l
-------
GROUNDHATER QUALITY MANAGEMENT CONTEXT
Groundwater quality management and protection has been described as one of the
top two priorities on the environmental agenda for the 1980's and possibly
beyond, the other top priority most often mentioned being add precipitation
control. The Importance of the Issue 1s Justified by a battery of now-
famlHar statistics: over 50 percent of the U.S. population depends on
groundwater as a drinking water supply, over 80 percent In many rural areas of
the country. Groundwater also supplies at)out.40 percent of the nation's
Irrigation requirements and approximately 25 percent of the requirements for
so-called "self-supplied" Industrial purposes.
In addition to these characteristics of groundwater use, characteristics of
"typical" groundwater contamination Instances also give evidence as to the
Importance of groundwater quality protection. Groundwater contamination has
often been discovered near heavily populated and not coincidentally
Industrialized areas. The cost of remedial actions to remove, replace, or
renovate contaminated groundwater can be measured In the millions of dollars
at a single site alone. The Congressional Office of Technology Assessment has
sounded a further ominous note In Us recent study of groundwater quality
management by observing that what we know today about groundwater
contamination problems nationally 1s at least as reflective of how we have
discovered or looked for problems and what we have been looking for as 1t 1s
representative of the major outlines of actual groundwater contamination
problems across the country. Almost all Interested parties agree that there
1s at present Insufficient groundwater quality monitoring 1n most locations to
truly characterize the nature and extent of groundwater quality contamination
problems accurately or comprehensively. As monitoring programs Increase and
more data becomes available 1t 1s more than likely that more contamination
problems may be found as well.
U.S. EPA'S GROUNDWATER PROTECTION STRATEGY
In this context of groundwater management needs and problems, the U.S. EPA
Issued Its Groundwater Protection Strategy 1n August 1984. In It the Agency
outlined Initiatives 1n four areas designed to Improve groundwater quality
protection efforts:
• Strengthen state groundwater agencies and Institutions
• Strengthen EPA's Internal organization related to groundwater
management
• Assess problems from unaddressed sources of contamination
Issue guidelines for EPA decisions affecting groundwater protection
and clean-up to promote consistency In decision making and regulatory
requirements.
The subject of this conference, non-point source pollution, was addressed more
than once by EPA's Groundwater Strategy. Under the category of unaddressed
sources of contamination, the Agency proposed to Increase efforts to protect
P-II-A-2
-------
groundwater from pesticide and nitrate contamination. EPA committed Itself to
evaluate more precisely the danger of groundwater contamination from pesticide
use, to use labelling restrictions to restrict pesticide use over certain
types of aquifers, to provide more health advisories and technical assistance
to states, and other types of studies and analyses to Increase understanding
of the problem and Us possible solutions.
The guidelines the Agency proposed to Increase the consistency of Us own
groundwater-related decisions are based on the concept of classifying
groundwater formations according to their present quality, current or best
use, and vulnerability to contamination. Depending on Us classification (the
EPA system has three classes) groundwaters would be protected to a greater or
lesser degree, but all groundwater In the same class would theoretically, at
least, be protected to the same degree. The strategy outlined, for example,
how hazardous waste facility siting restrictions would differ according to
what class of groundwater underlled a particular site. The degree of clean-up
required under Superfund could also vary according to what class of
groundwater was Involved.
Non-point sources were also addressed In this groundwater classification
portion of the strategy. The highest class of groundwater 1n EPA's
classification system, "Special Groundwaters", Include groundwaters that are
"Irreplaceable source(s) of drinking water" or "ecologically vital." The
Agency proposes to look Into developing additional restrictions on the use,
disposal, or storage of potentially threatening chemicals over this class of
groundwater. This could potentially Include pesticides, fertilizers, road
salts, storage tanks of various types, and other substances typically Included
on the roster of non-point sources of pollution.
While the groundwater classification guidelines were proposed to apply to
EPA's own decision-making they have Implications for states as well. States
with EPA-delegated authorities 1n groundwater-related program areas (such as
RCRA, Superfund, Safe Drinking Water, etc.) will be required to show that
their programs are no less strlgent than Federal requirements 1n those areas
which will result from Implementation of the Agency's groundwater strategy.
EPA has said 1t will keep regulatory requirements for states general and
performance-oriented to the extent possible In these areas.
The strategy, particularly In Us groundwater classification components, has
generated great Interest, particularly among some segments of the regulated
community, as to how states regulate groundwater quality now and how they
might be required to change their approach to conform to EPA's Strategy. The
General Accounting Office, the Office of Technology Assessment the
Environmental Law Institute, and the American Petroleum Institute have all
completed studies on state groundwater protection programs. EPA's Office of
Groundwater has a similar study scheduled for completion In Spring 1985.
Various task forces and commissions have been formed to address how best to
protect groundwater quality, and national groundwater protection legislation
1s an almost sure topic for congressional debate 1n 1985.
A groundwater quality management study recently completed by the author
compared existing state groundwater programs with the provisions of EPA's
Groundwater Strategy. The purpose of the study was to Identify the nature and
P-II-A-3
-------
extent of changes that might be required of states 1n order to bring their
programs 1n line with the Strategy. One of the topics of this study was the
non-point source-related provisions of the strategy mentioned earlier 1n this
paper. Framed as a research question, this portion of the study asked: Do
any states presently regulate pesticide, fertilizer, or other chemical use,
handling, storage, or disposal on an aquifer specific or groundwater-speclflc
basis as proposed In EPA's Strategy? Legislation, regulations, policy
statements and related program documents from all fifty states were reviewed
and environmental agencies In all states were contacted for answers to this
question. Some example approaches and general themes from the state-by-state
regional analyses are presented In the next section.
STATE LEGAL/REGULATORY PROGRAMS
With regard to non-point source regulation, most states have not gone beyond
the requirement of best management practices (BMPs) as part of their general
water quality programs. As has been pointed out by EPA Hselt recently, BMPs
designed primarily for the protection of surface water quality may actually
have resulted 1n Increased degradation of groundwater quality, e.g.,
percolation "catch basins" for stormwater runoff which may result 1n
concentrated pollutant streams reaching underlying aquifers.
Many, 1f not most state agency contacts consulted 1n the course of the study
reported the perceived political Impossibility of generating necessary levels
of support 1n their state legislatures for any regulation of non-point sources
of groundwater contamination beyond BMPs. Agency contacts offered this
assessment while at the same time reporting that their existing statutory
authorities probably provided a sufficient basis for developing and
Implementing such programs.
Some states with otherwise comprehensive and Inclusive programs to regulate
point sources of potential groundwater contamination, explicitly exclude
non-point sources from their regulatory requirements. Montana, as Just one
example, Is a state which employs a classification system similar to that
proposed by EPA's Strategy and requires permits for most point source
discharges to groundwater: a combination of management approaches employed by
well less than half of all the states. The state has a relatively strict
non-degradation policy for groundwater quality, yet expressly excludes from
degradation restrictions changes In groundwater quality due to non-point
sources "where all reasonable land, soil and water conservation practices have
been applied."
This Is reflective of the comments received from most state agency contacts,
that given the present state of knowledge with regard to groundwater
contamination dynamics and the expected high costs of thoroughly and
positively regulating diffuse non-point sources, recommended management
practices and guidelines are the only feasible management option for non-point
sources capable of being Implemented on a broad scale. Some states and
localities have begun to address at least In part the non-point concerns of
chemical storage use, and handling as outlined 1n the EPA's Groundwater
Strategy. Approximately fifteen states have enacted specific regulatory
programs to control leaks from above and below ground tanks for storing
petroleum products, chemical feedstocks, and other substances. Over half of
P-II-A-4
-------
the states with such programs are 1n the Northeast region of the country.
Host states have general authorities to respond to tank leaks and require
clean-up. State agency contacts report that these authorities generally do
not provide a sufficient basis for an effective, preventive, regulatory
program. Some existing tank leak prevention programs at the state and local
level consider the hydrogeologlcal formation over which a tank will be sited;
more often than not this 1s not an explicit consideration. Many agency
contacts 1n states without tank regulation programs have reported that they
are waiting to see what requirements will be forthcoming under EPA's new tank
authorities from the 1984 RCRA amendments before proceeding further on program
development of the state level.
Apart from storage tank programs, few states have taken steps to restrict
chemical use, handling, or storage, over certain groundwater formations which
could be susceptible to contamination from such sources. Some states' major
facility siting laws perhaps come the closest to embodying the types of
authorities and restrictions apparently envisioned for pesticide and other
chemical handling and storage (at least for Class I aquifers) by EPA's
Strategy. Even states which currently classify aquifers or groundwater
Include restrictions only on certain point sources, such as hazardous waste
landfills, and even these are allowed to locate over aquifers 1f certain
design and operating conditions are met. Seldom 1f ever 1n such state
programs are non-point sources singled out or Included for regulatory
attention.
New Jersey 1s an example of a state which has explldty addressed chemical
handling and storage as a possible source of groundwater contamination In Us
water quality regulations. Major chemical storage facilities (greater than
400,000 gallons capacity) must provide for secondary containment structures
lined with Impermeable materials wherever hazardous substances are handled.
Leaks must be able to be detected by Inventory controls, automatic leak
detection devices or groundwater monitoring wells. Regulations are applied to
above and below ground storage of petroleum and other hazardous substances,
drum storage areas, transmission pipelines and 1n-plant piping.
In Massachusetts, many local communities have enacted recharge area zoning
ordinances which regulate and restrict the handling, storage, use, etc. of
chemicals over particular aquifers and types of hydrogeologlc formations. The
state also has an aquifer/land acquisition program through which localities
are assisted to purchase and thereby remove or prevent all threatening land
uses from property overlying their local groundwater supplies.
In Virginia, a county government has created a special Natural Resource
Conservation Overlay zone 1n the recharge area of one of Its Important
drinking water aquifers. This special zoning legislation Is directly aimed at
restricting the use of agricultural chemicals 1n the recharge area to protect
groundwater quality
CONCLUSION
These are a few examples of Initiatives at the state and local level to
control one category of non-point sources with the potential to affect
groundwater quality: chemical use, handling and storage. These approaches
P-II-A-5
-------
are neither typical nor wide-spread among the states. The most prevalent
approach to non-point source pollution of groundwater continues to be
recommended practices and guidelines. Many states have requirements that
pesticides be registered for specific uses and that applicators be certified.
But few of these programs consider underlying groundwater explicitly as Is
proposed 1n EPA's Groundwater Protection Strategy.
The EPA Strategy's proposed Initiatives relating to pesticide and other
chemical handling, use, and storage has the least number of analogies among
state programs of any of the Initiatives proposed 1n the Strategy. The number
of local government programs of this type Is not surprising, since what 1s
being proposed In the Strategy Is a variety of restrictive zoning authority
usually associated with traditional local government powers. It appears
likely that any Initiation of even a limited regulatory program under this
component of EPA's Strategy will require significant changes 1n existing state
approaches to protection of groundwater quality from these and other non-point
sources of pollutants.
REFERENCES
Ground-Water Protection Strategy for the U.S. Environmental Protection Agency
(U.S. EPA, August 1984).
Groundwater: Strategies for State Action (Environmental Law Institute, 1984).
Protecting the Nation's Groundwater From Contamination (U.S. Congress, Office
of Technology Assessment, October 1984).
Pye, V.I., Groundwater Contamination 1n the United States (University of
Pennsylvania Press, 1983).
Trends In U.S. Groundwater Law, Policy, and Administration (Edison Electric
Institute, January 1984).
P-II-A-6
-------
POSITION PAPER
URBAN WISCONSIN SHOULD JOIN THE FIGHT
AGAINST NONPOINT POLLUTION
Kathleen M. Falk - Wisconsin Public Intervenor
The Wisconsin Public Intervenor, an Assistant Attorney
General in the Wisconsin Department of Justice, is expressly
charged by statute with the duty of intervening where necessary
to protect natural resources in the State of Wisconsin. The need
for protection of these natural resources from pollution through
nonpoint sources is now at a critical stage, yet there are still
no comprehensive programs to either prevent or clean-up nonpoint
pollution. Urban Wisconsin must do its part to ensure that the
future will be different from the past by regulating urban
nonpoint pollution. My intention is to summarize some of my
observations as to why there are still no comprehensive programs
to prevent or clean up nonpoint pollution and why legislation
must be enacted to change this. Legislators must recognize that
one of the most promising ways to prevent future nonpoint
pollution is through the control of urban sprawl.
In the early 1970's, a rash of fundamental changes occurred
in federal and state environmental laws, turning around the way
the country was polluting its resources. The immediate focus of
the new laws was on pollution that one observed spewing out. of a
smokestack into the air, or pouring out of an industrial pipe
into a river. Very little attention and effort was directed at
nonpoint pollution.
We are at the point, historically, where most agree that
tremendous advances have occurred in meeting many of the goals
and mandates contained in the 1970's laws to clean up our state
and nation from point source pollution. It is imperative that we
now turn our political and technical skills forward to solving
the problem we let lie underneath the rug for the past decade.
Instead of cities simply pointing their fingers at pollution
caused by their rural neighbors, and rural Wisconsin citizens
pointing their fingers at the city culprits of nonpoint
pollution, and instead of debating who causes more pollution and
whose pollution is the worse, it is time to join together to
remedy and prevent future nonpoint pollution. Of the many
sources of nonpoint pollution — agriculture, urbanization and
transportation, to name a few — one specific cause of nonpoint
pollution gets too little attention: urban sprawl.
In a broadly historical sense, urban sprawl is a relatively
new problem. After World War II, suburban development was
considered a goal, almost an ethic; paving over our countryside
was equated with "progress" and "growth." Numerous government
programs subsidized suburban sprawl - many still do.
Urban sprawl causes enormous environmental and economic
problems for Wisconsin. One of the major environmental harms is
P-II-B-1
-------
urban runoff, one kind of nonpoint pollution. A recent study of
the Menominee River Watershed reveals that approximately 75% of
the total pollutant loadings in the river comes from urban
nonpoint pollution. Urban runoff contains almost every type of
pollutant: suspended solids and toxic substances, particularly
heavy metals, bacteria, nutrients, asbestos, oil and grease and
other pollutants. One study in Washington, B.C., found that, in
water from streets, the concentration of suspended solids was 104
times higher than effluent from a secondary sewage treatment
plant, and the lead concentration was 1,015 times higher.
(President's Council on Environmental Quality, 1980.)
A recent study by Bauman, Domanik and Konrad, 1980,
identifies the two major types of nonpoint pollutants generated
by urban land use: 1) nutrients and sediments, and 2) toxic
materials*
Urban sprawl is a major nonpoint source of these two kinds
of pollutants because of the widespread land development,
increased traffic density and deficient wastewater treatment
facilities that are associated with it.
The magnitude of the nonpoint pollution generated by new
land development is reflected in the following statistic on the
Menominee River Watershed, taken from the Bauman, Domanik and
Konrad study. Developing urban areas represented only 2.6% of
the total area of the watershed, but contributed 37% and 48%,
respectively, of the suspended solids and total phosphorous.
Subdivision construction was cited as a major culprit in the
generation of nutrient and sediment pollutants.
To further underscore the excessive pollutant amounts
contributed by land under development, it is helpful to compare
it with industrial nonpoint pollution. In 1978, industry put an
estimated 5,100 kg/ha/yr of suspended solids into the Menominee
River Watershed. Land under development was the source of 43,700
kg/ha/yr. That means new urban development, such as that which
makes up urban sprawl, generates roughly eight times more
suspended solid pollutants than industry (Bauman, Domanik and
Konrad, 1980). The need for regulation of development is
apparent.
Another by-product of urban sprawl is an increase in
nonpoint pollution generated by transportation. Greater
commuting distances traveled by each car will influence the
amount of pollutants emitted into the air.
Automobiles are the primary source of airborne lead in the
environment (Carberry, 1980). As a toxic, lead is one of the two
major types of urban nonpoint source pollution. A recent
compilation of studies on the source of lead and its transport in
urban run-off indicates that emission of lead into the air by
automobiles leads to fall-out of the lead particule in rainwater
and soil. One study measured lead concentrations in rainwater in
P-II-B-2
-------
Pittsburgh, Pennsylvania, and found average concentrations of 41
ppb for urban samples. This was compared with 19 ppb for rural
samples. The maximum allowable lead concentration in drinking
water at the time of this study was set by the EPA at 50 ppb.
The conclusion drawn by J.B. Carberry, the compiler of the
studies, was that "rainwater analysis indicated that urban
regions were more highly contaminated from automotive exhausts
than rural regions." To increase the emission of airborne lead
as a by-product of urban sprawl is not only senseless, it is a
threat to health. In man, accumulation of lead is faulted with
hemoglobin interference and kidney function impairment. The
increase in traffic density caused by commuters from suburban
areas can only add to the nonpoint pollutants already threatening
the environment and human health.
Along with the sprawl of urban development into former rural
areas comes the need for wastewater treatment facilities. The
cost of wastewater treatment facilities borne by each household
and subsidized through grants is much greater for these new
facilities than for those in established urban areas. Regardless
of these prohibitive costs, facilities must be built so that
adequate treatment of waste can be achieved before it is
discharged into the environment.
An example of the cost discrepancy between the old and the
new facilities is the following statistic from a study of
wastewater treatment costs done by James A. Hanlon (1980).
Wastewater treatment costs in Greater Chicago, Illinois, were an
estimated $45.00 per household/year in 1980. The treatment costs
for the Village of Elwood, Illinois, with a population of 783
were an estimated $215 per household/year. Urban Wisconsin can
no longer afford to allow new sprawling developments to consume
private and public monetary resources in addition to compounding
the nonpoint pollution problem.
The state can prevent multiple ills by preventing urban
sprawl. Wisconsin should get to work by beginning to get its own
house in order. Each day state agencies make many, many
decisions which actually cause, not prevent, urban sprawl and
nonpoint pollution. These agencies should redirect their efforts
and prevent suburban development and control this major source of
nonpoint pollution. Urban areas should start reducing nonpoint
pollution by regulating urban sprawl. Legislation is long
overdue.
P-II-B-3
-------
References
Bauman, J., Domanik, A., and Konrad, J. (1980).
Nonpoint source pollution In Urban Areas. Seminar on
Water Quality Management Trade-Offs. U.S. EPA, 309-322.
Carberry, J.B. (1980). Water quality degradation due to nonpoint
pollution from urban sources. OWRT-B-018-DEL/14-34-0001-
8070, University of Delaware. 1-53.
Hanlon, James A. (1980. Costs for Wastewater Treatment,
Seminar on Water Quality Management Trade-Offs,
U.S. EPA, 81-86
llth Annual Report of the President's Council on Environmental
Quality (1980).
Southeastern Wisconsin Regional Planning Commission (1980).
A nonpoint source water pollution control plan for the Root
River Watershed. 19-24.
p-H-B-4
-------
ECONOMIC INCENTIVES FOR CONTROL
OF NONPOINT POLLUTION
Kathleen Segerson
Asst. Professor, Department of Agricultural Economics
University of Wisconsin
Madison, Wisconsin, USA
ABSTRACT
Attempts to control agricultural nonpoint pollution (NPP) have focused on the
strong encouragement of the use of "best management practices (BMPs)", with
cost sharing used as an inducement to participate. However, few farmers have
chosen to implement approved BMPs, and it appears that these voluntary
programs are unlikely to result in levels of participation that are sufficient to
combat NPP adequately. In addition, imposing mandatory practices to reduce
NPP is unlikely to result in the use of least cost abatement techniques since
the effectiveness of alternative techniques will vary across farms.
An alternative to the use of mandatory farming practices is the use of economic
incentives to induce pollution abatement. This paper suggests one possible
incentive mechanism that could be used when the suspected sources of the
pollutants can be identified (e.g., a group of farms bordering a waterway) but
the polluting activity (e.g., pesticide application or runoff) cannot be directly
observed and/or monitored. The general mechanism combines a system of
rewards for water quality above a given standard with a system of penalties
for sub-standard water. Both the advantages and disadvantages of this
incentive scheme are discussed.
Although the discussion is in the context of agricultural pollution, a similar
incentive mechanism could also be applied in cases of non-agricultural NPP of
either surface or ground water.
Keywords; Nonpoint pollution, pollution abatement, agricultural runoff,
economic incentives, pollution taxes.
P-II-C-1
-------
INTRODUCTION
Twelve years ago the U.S. Congress passed the Federal Water Pollution
Control Act Amendments (P.L. 92-500), which were intended, among other
things, to address and control nonpoint water pollution. Yet, in its assessment
of the success of existing pollution control programs, the General Accounting
Office (1982) identified NPP as one area in which little progress could be
claimed. Since agricultural runoff is one of the main sources of NPP, the
problem is particularly acute in the largely agricultural North Central Region.
Attempts to control agricultural NPP have focused on the strong encouragement
of the use of "best management practices (BMPs)," with federal or state
cost-sharing used as an inducement to participate. However, few farmers have
chosen to implement approved BMPs. For example, in Wisconsin, the
Department of Natural Resources (DNR) set a goal of 70% participation after
two years by landowners in priority management areas, while actual
participation rates have ranged only between 11% and 38% (DNR, 1984).
There are at least two possible reasons for low participation rates. The first
is that, even with cost - sharing, individual farmers might expect that the costs
to them of implementing BMPs are not justified on the basis of their expected
returns. For example, if the benefits of the use of BMPs result primarily from
reduced downstream sedimentation and pollution rather than from increased
productivity or reduced input costs, these benefits will not accrue directly to
the farmer and it is unlikely that farmers will be willing to bear the associated
costs voluntarily.
A second possible reason for low participation is that farmers tend to be very
independent and object to anything they consider to be an infringement on
their freedom to operate their farms as they choose. Since participation
involves a commitment to certain operating practices, it limits a farmer's
freedom to operate independently.
Given these disincentives to participation, it is not surprising that the U.S.
Environmental Protection Agency (EPA) should conclude that "In such cases,
sole reliance on voluntary programs is not likely to accomplish adequate
reductions in pollutant loads and, as a result, other approaches may be needed
(e.g., economic incentives or regulation)." (EPA, 1984, p. xv). The most
likely form of regulation to reduce NPP would be some form of mandatory use
of BMPs or other farming practices. This approach has two drawbacks.
First, it does not in general provide sufficient flexibility to control NPP in a
site and source-specific manner that guarantees the most efficient use of
available funds. For practical reasons, direct regulation or control is
generally applied across-the-board, without regard to differences between
polluters in terms of either costs of abatement or damages imposed. In the
case of NPP, since the effectiveness of alternative BMPs depends on conditions
that vary across farms (i.e., any individual BMP is not necessarily "best" for
all farms), mandating the use of these practices across-the-board is unlikely to
result in use of least-cost abatement techniques. In other words, resources
will be wasted unless site-specific controls can be imposed, since farmers
would be forced to use practices that are inappropriate for their specific
conditions.
P-II-C-2
-------
Direct regulation has a second potential drawback as well, namely the
difficulty of monitoring to ensure compliance with the regulations. Although
the use of some practices (such as terracing) can be easily monitored, others
(such as a modification of fertilizer or pesticide application techniques to
reduce runoff) would be difficult to monitor on individual farms. Since
farmers would often have little incentive to comply with regulations, voluntary
compliance cannot be assured. Thus, in the absence of substantial
enforcement efforts, the effectiveness of direct regulation may be further
reduced.
As suggested by EPA, an alternative to direct regulation is the use of
economic incentives to induce participation. This paper describes one possible
incentive mechanism that could be used under certain conditions to reduce
NPP. This mechanism combines a system of rewards for water quality above a
given standard with a system of penalties for sub-standard water. The
possible use of this incentive scheme and its advantages and disadvantages are
the subject of this paper.
IMPORTANT CHARACTERISTICS OF NPP
An important characteristic of NPP that makes the standard solutions that have
been successful in controlling point source problems unworkable for NPP is
that, although the likely polluters can often be identified, it is generally not
possible to identify a one-to-one relationship between the level of abatement or
discharge and the damages from pollutants in the water system. The reason
for this is twofold: (1) given any level of abatement, the resulting water
quality effects are uncertain due to the contributing effects of stochastic
variables, and (2) the level of abatement or discharge often cannot be directly
monitored by the overseeing authority without excessive costs.
More specifically, the pollutant loadings that result from any given operating
practice depend on a number of climatic and topographic conditions in a
manner that cannot be predicted with certainty. For example, as noted by the
EPA,
The timing (intermittent discharge caused by rain or snow),
concentration, and dilution of the pollution from nonpoint sources
constitute only part of the picture when one considers the nature of
associated water quality impacts; the transportation and ultimate fate
of the pollutant constitute the other part.
. . .The potential dilution of pollutants during high flow must be
considered against the velocity with which pollutants are dislodged
and transported. Thus, it is difficult to make generalizations about
the concentration of loads from nonpoint sources. (EPA, 1984,
p. 1-8)
The result is that associated with any given abatement practice or discharge
level at any given time is a range of possible loadings for each pollutant.
(More generally, there is a range of possible damages in terms of the impacts
P-II-C-3
-------
on human health and welfare that depend not only on loadings but also on
factors such as stream flow and exposure risks. Although conceptually the
analysis could be applied to this broader range of impacts, for simplicity we
focus here only on the range of possible loadings.) This range can be
represented by a probability density function (p.d.f.) that is conditional on
the abatement practice. The p.d.f. gives the probability that loadings of a
given magnitude will occur at the specified time, where the probability depends
on the abatement practices being used.
The objective of policies to reduce NPP is then to shift the distribution
represented by the p.d.f. to the left, as illustrated in Figure 1, to increase
the probability that actual loadings will fall below some tolerance level.
Probability
Tolerance Level
distribution with
abatement
distribution without
abatement
mean
loading with
abatement
mean
loading without
abatement
pollutant
loading
Figure 1: Distribution of Pollutant Loading With and
Without Abatement
If direct monitoring of all farm operations were economically feasible or
voluntary compliance with regulations were guaranteed, then the distribution
could be shifted through mandatory abatement practices. However, as noted
above, achieving this at least cost would in general require mandating
site-specific practices. In addition, monitoring certain farm activities such as
irrigation and pesticide or fertilizer use on a continual basis is generally
impractical, and voluntary compliance with restrictions on use in the absence
of monitoring is unlikely. Thus, the use of direct regulation to try to shift
the distribution seems inappropriate; the use of a mechanism that provides an
incentive (either positive or negative) for compliance is likely to be more
effective.
In addition to shifting the p.d.f. of pollutant loadings, incentives can also
result in risk-sharing. The existence of a distribution of possible loadings
implies that there are certain risks associated with any given level of
abatement. For example, if the actual level of abatement is set so that
expected marginal costs equal expected marginal benefits, there is the risk
that mitigating factors will be favorable and therefore that this level of
abatement will have been higher than that necessary to achieve a target level
of environmental quality, i.e., actual loadings will be even less than expected.
P-II-C-4
-------
Conversely, under adverse conditions this level of abatement could turn out to
be insufficient to meet the environmental quality goal. To date, U.S. pollution
control policy has in general been structured so that polluters bear none of
this risk associated with a given level of abatement and society bears the full
risk. Since abatement requirements do not depend on the realized
environmental conditions, i.e. the realized damages, the cost to the polluter is
the same regardless of those actual conditions. His payoff from pollution
abatement is nonstochastic. However, society's payoff in terms of reduced
damages is a random variable depending on the realized environmental
conditions. If mitigating factors are favorable, society will "win" the gamble
in the sense that the actual damages will be lower than expected. Likewise
adverse factors imply that society "loses" in that it suffers damages that are
greater than expected. If society is adverse to taking risks, it can increase
its expected well-being by shifting some of the risk to the polluters through
the use of policy instruments that result in risk-sharing. This could benefit
polluters as well if sharing risks induced victims to be content with more
lenient target levels.
The use of incentive mechanisms and risk sharing to induce desired behavior
has been studied by many previous authors, e.g. Stiglitz (1974), Holmstrom
(1979), Shavell (1979), and Mookherjer (1984), in the context of what has
come to be known in the economics literature as the "principal-agent" problem.
A principal-agent problem exists when the welfare of one party, called the
principal, depends directly on actions taken by another party, called the
agent. The principal's challenge is to devise a payment schedule for the agent
to induce the agent to take those actions that best serve the interests of the
principal. However, the principal is unable to control the agent's actions
directly and cannot even observe them. He can only observe his payoff that
results from the agent's actions, which depends not only on those actions but
also on stochastic conditions or events. The following question is then asked:
what form of incentive mechanism should the principal use to induce the agent
to take those actions that best serve the interests of the principal? The
question can also be asked in the context of a more general model where there
are many agents instead of just one.
If we interpret society as the principal and the farms generating NPP as the
agents, the principal-agent problem describes the challenge faced by society to
induce independent farmers to take steps that will contribute to improved
water quality and thereby serve the interests of society. Viewing NPP in this
way allows the insights that have been gained for solving general
principal-agent problems to be applied in solving problems of NPP.
AN INCENTIVE MECHANISM
Single Polluter Problem
In devising an appropriate incentive scheme, the first point to note is that
with regard to NPP we are interested in a reduction of loadings (or, more
generally, the damages that result from them) rather than a reduction in, for
example, soil erosion per se. This is emphasized by the EPA (1984), which
notes that lands with high erosion rates are not necessarily those causing
significant NPP problems, and vice versa. Thus, incentives designed solely to
P-II-C-5
-------
reduce soil loss (such as soil loss taxes) are inappropriate for control of NPP.
Instead, an incentive scheme designed to shift the distribution of loadings is
needed.
Consider first the case of a single suspected polluter, e.g. a single farm
whose land drains into a nearby stream. Let x be the level of actual loadings
of a given pollutant in the stream, and let x be a specified target or tolerance
level of loadings, which is set by authorities and could, for example, be
adjusted seasonally. Actual loadings x will depend upon both the abatement
actions taken by the polluter (e.g., the use of various BMPs) and random
variables reflecting unpredictable weather and stream conditions, as illustrated
in Figure 1.
A general incentive scheme designed to shift the distribution of actual loadings
would take the form of automatic, required payments T(x) that depend upon
the level of actual loadings as compared to the target level x and are given by
if x £ x
__ I
if x < x
where t and k are constants that can be set by the regulating authority to
ensure that the payment scheme provides the incentive necessary to induce the
polluter to undertake the level of abatement that is deemed socially desirable.
The optimal values of t and k will depend upon marginal abatement costs and
benefits as well as how increased abatement shifts the distribution of loadings.
(See Appendix A for details.)
The payment scheme is composed of two parts. The first, reflected by t, is a
tax/subsidy payment that depends upon the extent to which x differs from x.
If actual loadings exceed the target level, the polluter pays a tax proportional
to that excess, while actual loadings below the target level result in a subsidy
or credit to the suspected polluter. Note that actual loadings may differ from
target levels because of either the abatement actions of the polluter or the
influence of the random variables. Thus, the polluter may be liable for tax
payments that result from influences outside his control. Likewise, however,
his liability may be reduced (and he may even receive subsidies if x falls
below x) due to favorable environmental conditions even if he has taken no
action to control NPP. Thus, in choosing his level of abatement, he gambles
on what his actual tax liability will be and weighs the additional cost of
pollution abatement against the decrease in expected payments that results
from increased abatement.
The same type of incentive is provided by the second component of the
payment scheme, reflected in k, which is a fixed penalty imposed whenever
loadings exceed the target. The amount of the penalty is independent of the
extent to which the target is exceeded. Again, the polluters can weigh the_
cost of abatement against the decrease in the probability that x will exceed x,
i.e., that he will incur the penalty, that results from increased abatement.
Note that the effect of this penalty scheme is different from that of penalties
P-ll-C-6
-------
applied to actions (or inactions) that are directly under the control of the
polluter (e.g., penalties for point emissions in violation of standards). In the
stochastic case, additional abatement will in general always decrease the
expected penalty by decreasing the probability that x will exceed x, whereas
under penalties for emissions in excess of standards incentives exist to reduce
emissions to the standard level but not below.
Either component of the incentive mechanism can be used by itself to induce a
desired level of abatement, or they can be used in combination as given above.
(See Appendix A.) Since a comparison of these alternatives is more relevant
when there are many polluters, they are discussed below in that context.
Multiple Polluters Problem
In many cases of NPP, it is likely that several polluters, e.g., several farms,
will be suspected of contributing to the loadings of a given waterway. An
incentive scheme similar to the one introduced above can still be used, if t and
k are allowed to vary across polluters, i.e., if tax payments of polluter i are
given by
ft.(x-x) + k. if x £ x
T.(x) = 1 l l
I t.(x-x) if x < x
Again, t. and k. can be set to ensure optimal levels of abatement by each
source. (See Appendix A.) Polluters with high marginal abatement costs or
whose abatement is likely to have a relatively small effect on the distribution
of loadings will tend to have lower optimal payments than other polluters.
Note, however, that each polluter's liability depends on loadings from the
whole group, not just his individual contribution, since at any given time
individual contributions are not known or observable.
As in the case of a single polluter, possible forms of the incentive scheme
include (1) a pure tax/subsidy scheme where k. = 0 and t. > 0 for all i, (2) a
pure penalty scheme where t. = 0 and k. > 0 for all i, and1 (3) a combined
scheme where t. > 0 and k. > 0 for all i. Although the optimal abatement level
can be induced1 using any \>ne of the three alternatives, their implications in
terms of total polluter/government payments are clearly different. For
example, under the pure tax/subsidy scheme, combined subsidy payments to
all polluters when x < x could far exceed the benefits of the reduced loadings,
since each polluter would in some sense get credit for the entire reduction in
loadings. In addition, it provides no way to reward or compensate polluters
who abate "more than their share" and thereby create additional benefits for
all other polluters. A pure penalty scheme has the advantage of requiring no
government outlays for low loadings, but also suffers from the inability to
compensate "good" polluters. The combined scheme avoids some of these
problems. Under this scheme the t. values are not constrained (see Appendix
A) and thus can be chosen so that \he sum of subsidy payments when x < x
does not exceed the benefit of the reduced loadings. Although this still
requires government outlays when x < x, those outlays can theoretically be set
as low as desired (as long as the k. are adjusted to maintain proper
incentives). In addition to choosing the sum of the t. to avoid excessive
P-II-C-7
-------
outlays, the regulatory authority can also set the individual t.'s to reward
"good" polluters. Thus, the combined scheme allows the distributional effects
of the incentive mechanism to be adjusted to satisfy other, non-efficiency
objectives.
ADVANTAGES AND DISADVANTAGES
Regardless of which form of the incentive mechanism is chosen, the use of
such a mechanism has several advantages.
First, it involves a minimum amount of government interference in daily farm
operations, and farmers are free to choose the pollution abatement techniques
that are least cost for their farms. Since individual farmers are in a better
position to determine the abatement practices that will be most effective for
their land (and will have an incentive to do so), their freedom to choose the
techniques used provides the flexibility necessary to ensure that any given
level of abatement is achieved at the lowest possible cost.
Secondly, once in place, the incentive mechanism can be easily administered
since it does not require continual monitoring of farm practices or metering of
runoff or soil loss. It instead requires that the regulatory authority monitor
pollutant loadings regularly and calculate the necessary tax or subsidy
payment. Accounts can be cumulated over time with payments made
periodically. If, over the time period, tax liability exceeds subsidy payments,
then no government outlays would be necessary even under the pure tax/
subsidy or combined approaches. The subsidies would simply act as credits
against tax liability.
Thirdly, the cost-sharing mechanisms of the existing NPP programs can be
maintained to prevent placing excessive burdens on the agricultural sector,
and other considerations regarding an appropriate distribution of costs can be
accommodated, as long as the parameters of the payment scheme are adjusted
accordingly to maintain proper incentives. As mentioned above, under the
combined approach the t. values can be chosen to reflect distributional
considerations. In addition, federal or state cost-sharing to cover a portion of
investment or operating expenses is consistent with use of the incentive
scheme. The payment mechanism simply provides the incentive for
participating in cost-sharing programs that seems to be missing under the
current structure.
A fourth advantage of the incentive scheme is that it focuses on water quality
rather than erosion or runoff, which is more appropriate for controlling NPP.
In addition, to the extent that some of the fluctuations in pollutant loadings
can be anticipated, there would be an incentive for farmers to try to offset
peaks by, for example, avoiding heavy pesticide or fertilizer applications prior
to anticipated rain or wind storms.
Finally, these incentives allow the risks associated with given abatement levels
to be shared by victims and polluters alike, which differs from the typical
approach to environmental policy in the U.S. under which only victims are
forced to gamble on whether given abatement levels are sufficient. This could
benefit both polluters and victims if sharing risks induced victims to be
content with more lenient target levels.
P-II-C-8
-------
The disadvantages of this incentive scheme include the information
requirements that are necessary to set the levels of the t. and k. parameters
initially to provide the correct incentive. (In general, tliiis is a problem with
any regulatory device seeking to achieve socially optimal outcomes.) The
necessary information includes abatement cost estimates, estimates of damages
from pollutant loadings, and estimates of how each polluter's abatement affects
the distribution of those loadings, which would require the use of individual
watershed models.
A second possible disadvantage of the mechanism is its implications with regard
to distortionary taxation. It would have to be structured so that allowing the
t. and k. parameters to vary across sources would not be considered to be
d\stortionary taxation.
REFERENCES
Holmstrom, B. (1979). Moral Hazard and Observability. Bell Journal of
Economics, 10(1), 74-91.
Holmstrom, B. (1982). Moral Hazard in Teams. Bell Journal of Economics,
13(2). 324-40.
Mookherjee, D. (1984). Optimal Incentive Schemes with Many Agents.
Review of Economic Studies, 51(3), 433-446.
Shavell, S. (1979). Risk Sharing and Incentives in the Principal and Agent
Relationship. Bell Journal of Economics, 10(1), 55-73.
Stiglitz, J. (1974). Incentives and Risk Sharing in Sharecropping. Review
of Economic Studies, 4, 219-255.
U.S. Environmental Protection Agency. (1984). Report to Congress: Nonpoint
Source Pollution in the U.S., Office of Water Program Operations, Water
Planning Division, January.
U.S. General Accounting Office. (1982). Cleaning Up the Environment;
Progress Achieve!
Washington, D.C.
Progress Achieved but Major Unresolved Issues Remain, GAO/CED-82-72,
ihir
Wisconsin Department of Natural Resources. (1984). Wisconsin Water Quality;
Report to Congress, WDNR-SWQS-84-7, May.
ACKNOWLEDGEMENTS
The author would like to thank Michael Carter, Jean-Paul Chavas, and Daniel
Bromley for comments on an earlier draft, without implicating them for any
remaining errors.
P-II-C-9
-------
APPENDIX A
This appendix describes how the parameters of the incentive scheme described
in the text could be set to ensure proper abatement incentives. The scheme is
similar to one suggested by Holmstrom (1982) in the context of optimal
organizational structure.
Let the random variable x(a) be actual loadings of a given pollutant which
depends on the level of abatement of a single polluter, denoted a. Let C(a)
be the cost of abating to level a, and let F(x,a) be the probability that x(a)
is less than a target level x, given a. Finally, let B(x(o)-x(a)) be the
benefit of increasing abatement from a base zero level to a. Then society
seeks the level of abatement that maximizes E[B(x(o)-x(a))] - C(a), where E is
the expectation operator over the random variable x. The optimal level of
abatement, denoted a*, is implicitly defined by the first order condition
E[B'-dx/da] + dC/da = 0.
Given the payment scheme T(x) in the text, the expected_value of those
payments, E[T(x)], is given by t-E[x(a)] - tx + k (l-F(x.a)). The polluter
thus chooses the level of abatement that minimizes E[T(x)] + C(a). His
choice, a, is implicitly defined by t-E(dx/da] - k(3F/3a) + dC/da = 0. A
comparison of the optimality conditions for society and the polluter implies that
the polluter will bejjiduced to choose the level of abatement that is optimal for
society, i.e., a = a , if:
(1) k = 0 and t = -(dC/da)/E[dx/da] evaluated at a*,
(2) t = 0 and k = (dC/da)/(3F/3a) evaluated at a*, or
(3) t is arbitrary and k = (dC/da + tE[dx/da])/(3F/3a) evaluated at a*.
Note that under the pure tax/ subsidy case (case 1), the optimal tax rate t is
equal to marginal benefits B' if B' is a constant. Under a nonlinear benefit
function, t ^ E(B'). However, E(B') may be a sufficient local approximation to
the optimal t, or serve as a guide in setting t.
If there are multiple polluters, let a. be the abatement level _of polluter i, let
C.(a.) be i's abatement cost, and interpret a in x(a) and F(x,a) as the vector
(a"..,1. ..,a ) (where n is the number of polluters). Society then seeks the
vector a* = (a|,...,a*) that maximizes E[B(x(o) - x(a))] - [C1(a1) + ...+
C (a )]. Individual polluters choose a. to minimize E[T.(x)] + Cjta.) given a
set oi expectations about the actions or other polluters. An analysis similar to
that for the single polluter case indicates that, in equilibrium where all
expectations are realized, the possible optimal incentive schemes are given by:
(1) k. = 0 and t. = -(dC./da.)/E(3x/3a.] evaluated at a*,
(2) t. = 0 and kt = (dC./da. J/(3F/3a.) Waluated at a*, or
(3) t* is arbitrary and k =l (dC./da.1 + tjEtSx/SajD/OF/Sap evaluated at
a*.
P-II-C-10
-------
LEGISLATION, NOT LITIGATION, IS THE SOLUTION
TO NON POINT SOURCE POLLUTION
by
James H. Petersen, Senior Staff Attorney
Milwaukee Metropolitan Sewerage District
PREMISE
Heightened recognition by society of problems arising from
non-point source pollution have raised an increasingly large
voice towards controlling such discharges. Often times, however,
those adversely affected by non-point source pollution do not
have available the tools or knowledge of the problem sufficient
to obtain affirmative relief. Indeed, it is the premise of this
paper that traditional legal remedies cannot adequately be relied
upon to remedy non-point source pollution. Thus, it is important
that society's interests be protected by the legislature, the
source of our ordered form of government.
INTRODUCTION
Recourse to legislation is particularly apropos in dealing
with non-point source pollution if for no other reason than non-
point source pollution does not respect traditional boundaries.
From any source of pollution, the effects pass on through the
watershed. The resulting pollutant path, the pollution shed,
effects many jurisdictions, transcending municipal and county
lines and can, in the many cases, cross state lines providing a
federal cast to the problem. Many of the traditional remedies
provided through the courts are frustrated by this concept of
pollution shed, and where not frustrated, the remedies are
rendered so expensive that only the most well-heeled of litigants
can realistically hope to finance a favorable result.
A further reason for recourse to legislation is the
critical need to factor economic and societal values into the
calculus of environmental rights. This analysis must of
necessity require society's judgments of the future and the cost
society is willing to bear today to ensure a sound environment in
the future. As will be seen, this is particularly true when the
need for a solution of non-point source pollution points toward a
change in more or less traditonal uses of water resources to
remove and pass along non-point source pollutants.
P-II-D-l
-------
I.
TORT REMEDIES ARE INADEQUATE
The use of litigation as a vehicle to obtain relief from
damaging non-point source pollution is the most familiar vehicle
for people to look to for relief and yet at the same time it is
probably the least adequate way to obtain the relief necessary.
This inadequacy is a necessary component of the legal system from
which the relief must be generated. Relief would be based upon a
theory of tort violation; that is, an injury to a protected
personal right caused by another's acting in an unreasonable
fashion.
Two aspects of traditional tort litigation prove to be the
most frustrating in the pursuit of relief from non-point source
pollution: the need to establish a discharger's act was the
cause of damage and that damage in fact occurred. These
limitations in traditional tort litigation create significant
problems with the allocation of responsibility for it is the
plaintiffs burden in any litigation to identify a defendant whose
conduct was the "but for" or "substantial factor" cause of the
harm done to the plaintiff person injured.
With non-point source pollution, identifying a "but for"
cause of damage is not easy. Perhaps a few statistics will
illustrate the point. In the United States there is enough
animal waste produced annually as would equal the waste generated
by two billion people, nearly ten times the number of people
currently living in the united States. Each day, a head of
cattle produces six times the waste of the average human being.
Erosion control in and of itself is a major cause of non-point
source pollution but the costs of remedying it are astronomical
and not entirely calculated. The Environmental Protection Agency
has made various estimates in just three of those areas.
EPA has estimated that highway erosion control could cost
anywhere from a $130 million to $7 billion; EPA has estimated
that urban construction erosion control would cost anywhere from
$140 million per year to $1.4 billion per year. EPA has
estimated that sediment control in streams would cost
approximately $200 million per year to $3 billion per year.
These statistics illustrate that the problem of non-point source
pollution is multi-faceted and its effect on water is
pervasive. Indeed, with such diverse sources as animal waste,
erosion, and sedimentation causing non-point source pollution of
waters, it tends to make the allocation of responsibility in
traditional tort litigation almost impossible.
P-II-D-2
-------
A convenient starting point for review of the practical
effectiveness of the tort system to treat of non-point source
pollution is to review the source of the rights to be enforced.
The most obvious rights against non-point source pollution are
the rights of a riparian owner. A riparian owner is the owner of
property bordering upon a water body. The riparian owner has a
right to use of the water subject only to the use being
reasonable and to the return of the used water in as to close to
its original state as possible. A riparian owner obtains rights
simply by virtue of the ownership of land bordering the water
body. The owning of land bordering a water body brings with it a
common law right to use the water in common with all other
riparians.
An action by a riparian to protect his rights is limited to
bringing an action against another riparian user. If a riparian
owner seeks to obtain damages from a non-riparian, the riparian
is limited to suit alleging nuisance.
In a suit to protect its riparian rights, the riparian need
only show that another riparian has interferred with the suing
riparian's use of the water. It need not show negligence nor any
form of intentional invasion of a property interest. While this
simplifies the burden of proof on the part of the riparian, it
does not entirely answer the non-point source pollution problem
since the polluter is subject to damages for its pollution
only. Further, even a polluting riparian is allowed a reasonable
use of the water which may include some reasonable amount of
pollution caused by such use. This may be compounded by a
downstream owner in effect granting an easement to pollute to an
upstream riparian by acquiescing in a course of conduct which is
known, and notorious.
The other major source of rights is found in the law of
nuisance. A nuisance is an unreasonable interference with a
person's enjoyment of land. In the case of a riparian owner,
nuisance would be established if it could be shown that the
defendant did an act that polluted the water in a material
fashion and that the defendant's conduct was a proximate cause of
that unreasonable interference.
It is the concept of proximate cause, however, which again
poses the most substantial impediment in an action for
nuisance. Proximate cause is at best difficult to establish when
there are multiple discreet dischargers, making the tracing of
water quality at best unachievable. It was the effective
inability of government to trace a water quality violation from
point source dischargers that led to the Federal Water Pollution
Control Act Amendments of 1972. Tracing non-point source
pollution dischargers presents an even tougher cause and effect
problem.
P-II-D-3
-------
The point of the above discussion is to show that
traditional tort remedies for relief from non-point source
pollution are simply inadequate to address the problem. The
number of polluters are many, their pollution is diverse, and the
cost of ascertaining proof to the requisite degree of certitude
is almost prohibitive.
The law has attempted to find ways around these limitations
resulting from a "but for" causation standard of damage by a
defendant. Most of these attempts have developed in the field of
products liability litigation. This alone makes use of these
theories suspect when treating of non-point source pollution.
There are grave conceptual problems in superimposing these
doctrines upon non-point source pollution and it is accordingly
not a recommendation that this be done. Nonetheless, it is
important that we recognize what those theories are so that by
understanding them we can appreciate why they should be rejected.
The four prevalent theories are: 1) alternative liability,
2) market share liability, 3) enterprise liability, and 4)
concertive action liability. In the State of Wisconsin, only the
latter, concertive action liability, is a recognized theory for
allowing recovery in a products liability circumstance. A brief
description of each of the four and its limitations:
1. Alternative Liability; If more than one defendant could
have caused the liability, the defendants may all be
joined in the litigation and the defendants must prove
among themselves who is not liable. This theory of
liability may not be used when a court does not have
jurisdiction over all of the defendants. This poses a
major problem with multiple jurisdiction pollution
sheds. It also causes problems with large, undefined
groups of defendants as are likely in non-point source
pollution. Further, some of the defendants may be
governmental bodies which are protected by dollar limits
on their potential exposure.
2. Market Share Liability; This is based upon the theory
that when the entire market is divided among a fixed
group of defendants, all participants receive benefit
from the existence of the market and should share in
liabilities arising from marketing. All market
participants are not needed in the litigation as a
defendant, but any defendant in the litigation would be
responsible for only its pro rata share of the liability
in proportion to its share of the market. This is not
readily available as an alternative for non-point source
pollution because non-point source pollution is not a
"market" oriented event similar to manufacture or
distribution of a pharmaceutical product.
P-II-D-4
-------
3. Enterprise Liability; This applies where there is an
industry wide practice which created the unreasonable
risk of harm, which risk is not capable of attribution
to a single defendant. Like the alternative liability
theory above, the defendants are required to prove among
themselves who is not liable; unlike the alternative
theory but like the market share theory the enterprise
liability theory does not require that all defendants be
in the suit. What is required, however, is that there
must be a well defined industry and it requires an
awareness of the risk and a joint liability to reduce
the risk among all of the defendants. These
characteristics remove non-point source pollution from
the enterprise liability theory of recovery since there
is no well defined industry nor is there a joint
liability to reduce risk;
4. Concertive Action Liability; This theory of liability
requires that there be a common design or agreement to
act in a certain fashion. Under this theory, all
participants in the concerted action are responsible for
any harm that is created. The concerted action theory
does no require any written agreement, and can be
established by a showing of conduct estabhlishing action
in concert. The fault with this theory in relationship
to non-point source pollution is that the showing of
concertive action across the broad cross sections of
society that create non-point source pollution is
virtually impossible.
In a sense the harm caused by non-point source pollution is
a public harm, a public harm that cries for some form of
intervention by the government. It is this no longer exclusively
private harm which should attract the attention of the respective
legislatures to conceive of means to address the need for a
remedy to non-point source pollution.
The next problem, of course, is what form should that
legislation take.
THE MANY FACES OF LEGISLATIVE SOLUTIONS
Recourse to legislation as a solution to the problems of
non-point source pollution, while undoubtedly the preferrable
means, nevertheless creates the problem of what form the
legislation should take. It must borne in mind that the
legislative forum has a number of limitations. Legislation often
times requires political compromise and political compromise is
P-II-D-5
-------
not susceptible to analytical review. Further, three important
problems arise with legislation.
First, it must be borne in mind that fixing a problem is
always more expensive than creating a problem. It is human
nature that if something can be done cheaper but create a
problem, it will have been done cheaper nevertheless. Second, it
is hard to assess environmental problems in terms of unit
costs. Thus, unlike fixing costs of pollution based upon
production, the benefit of fixing an environmental problem or the
harm caused by an environmental problem cannot be reduced to
economically meaningful values that fluctuate in a predictable
fashion. Third, environmental problems themselves are subject to
change by merely redefining what the problem is. In the context
of non-point source pollution, if stream erosion is eliminated
from the definition non-point source pollution you have created
an entirely different perspective on the problem. By like token,
if urban non-point source pollution is eliminated from the
legislation, you have a different problem than if you were
dealing with all non-point source pollution.
Bearing in mind these limitations on the legislative
process, there are basically four ways to achieve a legislative
solution to environmental problems. These are: (1) change the
legal status of the problem; (2) regulate the problem; (3) tax
the problem; and (4) subsidize the solution.
Legislation that changes the legal status of an
environmental problem is easiest to proclaim, but its
ramifications are often the most pervasive and hence disruptive
manner of dealing with the pollution problem. The Federal Water
Pollution Control Act Amendments of 1972 are perhaps the best
example of legislation which changed the legal status of a
pollution problem. The Federal Water Pollution Control Act
Amendments were predicated upon a change from water quality
standards dictating what the effluent limitation should be to a
system whereby a legislatively determined effluent limitation
would be achieved. The legislation decreed a zero pollution
goal. Notwithstanding the unreality of attaining a zero
pollution goal, the idea behind it was essentially to remove the
nation's waterways from the order of public goods free and
available to all. In its stead, Congress created a prohibition
against the use of those waterways as discharge points for
pollutants.
The use of legislative change in legal status to address
non-source point pollution may not correct the problem,
however. Politicians are pragmatic people in many ways. It will
require more than just a definitional change in legal status to
work the full change in legal status that is necessary to control
P-II-D-6
-------
non-point source pollution. The Federal Water Pollution Control
Act Amendments illustrate this point. The Congress enacted a
complete package of changes, including a construction grant
program, a permit program, water quality standards and state
implementation plans program, and a creation of effluent
standards program. A change in legal status for non-point source
pollution would require a similar, all pervasive response. I
submit that such an all pervasive change in the status of non-
point source pollution, while possible, is not yet politically
expedient.
A second legislative form to deal with non-point source
pollution would be to regulate the amount and means of non-point
source pollution discharges. Regulation is favored in America
for a number of reasons. Americans have a great love of quick
solutions and are willing to abdicate their ability to attain
greater societal change to a legal solution for a social problem
rather than develope other practical ways of dealing with that
problem. Many social problem solvers are also lawyers, resulting
in a tendency on their part to create remedies with which they
are familiar. Further, regulations alleviate the need for people
who are relatively unsophisticated in economic and social theory
to engage in intellectually painful analysis of how their acts
economically effect society as a whole. Last, Americans favor
regulation of environmental problems because it permits them to
ease their consciences more easily. This is important for many
because environmentalism has taken on virtually a religious
fervor.
Regulation is, contemporaneous with its ease in being
understood, economically quite troublesome. The relative ease in
understanding regulation of non-point source pollution comes from
the use in most regulatory shemes of the prescribed standard:
one standard for all like discharges. It is this recourse to a
uniform standard that gives rise to the problems, however,
because this solution does not admit of any inefficiency in its
implementation and because the standards are more or less
rigid. Coupled with this is recognition that there is never a
situation of governmental regulation without a cost of its own,
if in no other way than salaries and the like for administration
of the regulation, the creation of a bureaucracy. Government
regulation creates distortions in economic efficiencies. This
inefficieny is caused by the tendency to regulations to require
"equally proportionate reductions" of pollutants from each source
regardless of variance in control costs.* Although it is easy to
*Stewart and Trier, Environmental Law and Policy, 556
(Bobbs-Merrill, 1978)
P-II-D-7
-------
prescribe a standard, it is not so easy to justify the
standard. Thus, determination of standards can be quite
difficult.
The third legislative avenue for dealing with non-point
source point pollution is to create a tax on discharges.
Essentially, a tax would take the form of an emission fee per
unit of discharge. While taxing a unit of discharge is a ready
shibboleth to the economic sloganeer it is not a concept that is
free of significant problems.
Costs of abatement may vary per unit taxed, which is
economically good. Taxes do provide an incentive for reduction
and taxes are more adaptable and more easily administered.
Nonetheless, the standards to be achieved tend to be quite
static. Changing the standard to reflect varying efficiencies is
not easily achieved.
The tax concept has many of the economic inefficiencies of
the regulation except the tax furnishes an economic incentive to
abate the pollution. Establishing a standard to be achieved is
at best guesswork aided by political biases. The unit of '
detection, the pricing mechanism, and monitoring are even more
problematic.
The last form of legislation that may be utilized to
control non-point source pollution is the creation of a
subsidy. Subsidies take essentially two forms. One is tax
credits or increased depreciation for capital expenditures
related to reduction in non-point source pollution discharges.
Tax incentives and increased depreciation for capital
expenditures create problems notwithstanding that the government
in essence underwrites the efforts being made to relieve non-
point source pollution because there is a cost to the operation
which had been making the discharge. Depending upon the
discharger, that cost may be serious enough to turn what is
otherwise a profitable function into a nonprofitable function,
primarily when these costs are not common across the competitive
sphere. For example, where the government for any region less
than the whole nation, such as the State of Wisconsin, requires
its farmers to engage in various non-point source pollution
elimination activities, including capital expenditures, no matter
what tax incentives are given and no matter what increased
depreciation is allowed, there will nonetheless be a capital
expense to the farm operation which farm operations in other
states or regions do not have. In a marketplace where dairy
products, grain products, cattle products are essentially
fungible, requiring but one group of competitors in that market,
Wisconsin farmers, to take on the additional costs puts an undue
hardship upon them and puts them at a distinct market
P-II-D-8
-------
disadvantage.
A second form of subsidy is the concept of buying up units
of non-point source pollution discharge. The legislature would
establish an acceptable level of discharge that must be met and
create a fund with which to eliminate discharge to that level.
For every unit of historic discharge which is not discharged
following the enactment of the subsidy, the discharger would
receive a bounty from the government. Those dischargers who have
relatively low costs in achieving discharge curtailment will then
rapidly move into that field and will curtail their discharge and
obtain the bounty money; those units with higher costs to curtail
discharging will not be able to partake of the bounty.
Nevertheless, this latter group would have the benefit of being
able to use more of the acceptable limit for their specific
discharges.
Both subsidiary ideas, however, tend to break down in the
real world because the transference from theory to practice
contains a cost for government administration which cannot be
readily or accurately ascertained. The cost for ascertaining the
historic level of non-point source pollution discharges is quite
illusive and not readily calculable to the requisite degree of
certitude.
The four forms of legislative response to the problems of
non-point source pollution each have their own problems. When
the public wants a solution to non-point source pollution, the
legislature must weigh which is more likely to meet the public's
demand. The outside influence of the citizen demand for action
can force a choice.
III. PROPOSED SOLUTION
It seems that the most ready way to deal with non-point
source pollution is to regulate the discharge of pollutants into
the water. Although it has severe disadvantages, the use of
regulation does have a simple, quick solution aspect to it which
tends to be favored in American society. Regulation itself could
establish preferred pollution avoidance practices which could be
designated and enforced either by a state agency or as private
rights.
Furthermore, once legislation takes up the subject matter
of non-source point pollution it may remove it from local
politics. Additionally, regulation need not address the whole
problem at one time. Regulation would permit the legislature to
address a major segment at a time. For example, it is not
inconceivable for the legislature to designate the Department of
Natural Resources or the Department Health, Social Services or
the Department of Industry, Labor and Human Relations as the
P-II-D-9
-------
authority to identify construction site pollution emission
discharge limitations. The legislature could then delegate to
the department the authority within another year to promulgate
regulations on sediment control and yet further regulations on
highway rion point source pollution discharges. In this fashion
the legislature would address the entire problem on a timetable.
The State of Florida has began such efforts. Its
Department of Environmental Regulation has enacted a number of
non-point source pollution elimination steps concerning highway
erosion control which have subsequently lessened the discharge of
non-point source pollution to affected water bodies. Such a role
could be achieved in Wisconsin and in other areas nationwide.
1791L
P-II-D-10
-------
SUMMARY
Panel III
Efficiency and Feasibility of Best Management Practices
For Nonpoint Source Control
PANEL MEMBERS
Moderator: W. Trevor Dickinson, Professor of Watershed Engineering,
University of Guelph, Ontario, Canada
Assistant Moderator: Robert Biebel, Chief Environmental Planner,
Southeastern Wisconsin Regional Planning Commission, Waukesha, WI
Panelists:
Bradley Crowder, Institute for Research on Land and Water Resources,
Pennsylvania State University, University Park, PA
John Harkin, Professor of Soil Sciences and Water Resources, University
of Wisconsin-Madison
Jackie Robbins, Department Head and Professor of Agricultural Engineering,
Louisiana Technical University, Ruston, LA
Thomas Schueler, Department of Environmental Programs, Metropolitan
Washington Council of Governments, Washington, DC
Stuart Walesh, Associate, Donohue and Associates, Waukesha, WI
Recorder: David Kendziorski, Southeastern Wisconsin Regional Planning
Commission, Waukesha, WI
INTRODUCTION
The panel discussed the effectiveness and feasibility of both urban and
rural BMPs, and addressed analytical techniques for evaluating the prac-
tices. The need to properly define nonpoint source pollution problems
and evaluation criteria was also noted. It was acknowledged that pro-
blems are multi-dimensional, and that implementation of remedies can be
greatly affected by social, cultural, religious, economic, and personal
preferences.
It was questioned whether too much emphasis in evaluating nonpoint source
control measures is placed on modeling, and whether more attention should
be given to monitoring the water quality effects of BMPs. Panel members
felt that modeling was most valuable for guidance in selecting individual,
or an optimal mix of, BMPs. There should be good feedback between moni-
toring and modeling, with monitoring data being used to calibrate and
verify models. The most beneficial use of models is in the insight they
provide, rather than in the absolute quantification of variables. Models
P-III-1
-------
also provide an economical means for assessing alternative conditions
which could not realistically be evaluated with monitoring data alone,
It was concluded that modeling and monitoring programs for nonpoint
source control are not substitutes; rather, they complement each other,
RURAL NONPOINT SOURCE CONTROL
Concerns
1. Voluntary NFS control programs may not be effective and achieve
successful implementation in areas with intensive agricultural
production, particularly livestock raising.
2. BMPs may be applied for water quality improvement purposes without
regard to cost-effectiveness and pollutant removal capability.
3. BMPs which effectively reduce soil loss by reducing surface runoff
may not sufficiently control nitrate leaching from excessive
fertilizer use of manure applications.
4. Manure storage facilities may simply encourage farmers to expand
livestock operations.
5. Cost-effective waste control measures for unconfined livestock
operations still need to be demonstrated.
Recommendations
1. Cost-effective soil conservation measures include contouring, vege-
tative cover, conservation tillage, and strip-cropping. Terraces
are effective, but not relative cost-effective. Waterways and
diversions cost less than terraces, but are also less effective.
2. Nitrate leaching into the groundwater can be controlled by reducing
heavy applications of manure and fertilizers, (especially in highly
permeable soils), and by increased education of farmers.
3. The most efficient water quality improvement programs would utilize
the most cost-effective BMPs in critical NPS problem areas.
4. Confined livestock waste discharges should be collected and treated
as a point source. Ultimate disposal of the animal waste should be
land application or use as a feed source. Lagoon treatment and
direct discharge to surface waters is unacceptable.
5. Unconfined livestock waste discharges should be treated with practices
consistent with optimal forage production on a long-term basis. Con-
gregations of animals in critical areas should be prevent. Grazing
areas should be maximized and rotated.
P-III-2
-------
URBAN NONPOINT POLLUTION CONTROL
Concerns
1. Proper selection of urban BMPs is difficult because of the wide range
in the ability of the BMPs to remove various pollulants.
2. There is a lack of available BMPs which can achieve a reasonably
high level of pollutant removal in fully developed urban areas.
3. Stormwater storage in wet basins, one of the few highly effective
urban BMPs, cannot be used in areas with high slopes, low perme-
ability, or limited space.
4. Because of the limited knowledge available concerning the design,
operation, and effectiveness of urban BMPs, mandatory specifications
of BMPs would be premature and would result in "action", but little
progress.
Recommendations
1. Effective urban BMPs include wet basins, porous pavement, infiltra-
tion systems, and artificial wetlands. Dry basins, grassed swales,
and street sweeping have not been found to be very effective.
2. Among stormwater storage measures, dry basins cost least; extended
dry basins cost about 10 more, wet basins 20 to 45 percent more.
Infiltration systems are cost-effective only in smaller residential
areas.
3. Urban nonpoint source control programs are best implemented under
the guidance of a comprehensive stormwater management plan which
addresses both water quantity and quality concerns.
4. Screening procedures should continue to be developed to assist local
developers in selecting those urban BMPs which are most appropriate
within a given range of site characteristics.
5. Stringent inspection and maintenance programs must be developed to
help ensure the continued effectiveness of urban BMPs especially
storage facilities.
CONCLUSIONS
Following review of the urban and rural BMPs available, some panel members
expressed concern about the inability to specify the relative cost-effec-
tiveness, and to quantify the pollutant removal effectivenss of individual
BMPs. The panel concluded, however, that many excellent studies and de-
monstration projects have been completed, and implementation of both
urban and rural practices should proceed. More will be learned as these
practices are implemented on a wide scale. The advancing state-of-the-
art have provided a strong direction toward which types of BMPs work best
P-III-3
-------
under certain conditions. The panel expressed confidence that the techr
niques are available, and the knowledge exists, to sufficiently control
both urban and rural nonpoint source pollution in most areas. New and
emerging technologies will likely "fine tune" the system, enhancing
water quality improvements achievable with the present range of BMPs.
Inability to guarantee that any BMP selected for implementation is the
wisest choice should not be used as an excuse for inaction. Immediate
prudent use of existing patently cost-effective BMPs should be encour-
aged because they are all steps in the right direction.
P-III-4
-------
EVALUATING BMPs IN PENNSYLVANIA'S CONESTOGA
HEADWATERS RURAL CLEAN WATER PROGRAM
Bradley M. Crowder and C. Edwin Young
Agricultural Economists, USDA-Economic Research Service
The Pennsylvania State University
University Park, Pennsylvania, USA
ABSTRACT
This paper discusses the relative cost effectiveness of best management
practices (BMPs) for controlling agricultural nonpoint pollution. The
discussion herein applies directly to the control of soil and nutrient losses
resulting from crop production, and is derived from the economic evaluation
of the Conestoga Headwaters Rural Clean Water Program (RCWP) project in
southeastern Pennsylvania. The purpose of the project is to provide
assistance for the installation and maintenance of BMPs to control nonpoint
pollution and improve water quality.
Due to the current status of BMP implementation in the project, monitoring
data are not available to evaluate the impacts of BMPs on water quality.
Physiochemical and economic modeling of BMP impacts was undertaken to
characterize the cost effectiveness and overall effectiveness of BMPs for
controlling agricultural nonpoint pollution. CREAMS, a field scale model for
Chemicals, jUmoff, and Erosion from Agricultural Management ^Systems, is a
continuous computer simulation model that was used to estimate field-edge
losses of soil, surface runoff losses of N and P, and NO^ leached out of the
root zone.
Contouring and stripcropping (where the crop rotation makes it feasible) were
the most cost-effective practices for controlling erosion and surface runoff
losses of nutrients. Conservation tillage practices were also cost-effective
alternatives. Terrace systems were highly-effective runoff control measures
but were generally the least cost effective. Other structural practices,
waterway systems and diversion systems, were found to be less effective than
terraces for soil loss control, but were more cost effective.
NO^ leaching, perhaps the project's most significant water quality problem,
may not be alleviated and may increase when soil conservation practices are
implemented. Sound nutrient management practices are necessary to control
subsurface and surface deliveries of nutrients to water supplies. Nutrient
inputs from animal wastes, commercial fertilizer, and crop residues must be
managed in a way that prevents excessive inputs and limits movement off the
field. Manure storage and application techniques should be consistent with
maximizing the economic value of nutrients while minimizing field losses.
Keywords: Agricultural nonpoint pollution, best management practices,
economic modeling, cost effectiveness, water quality, CREAMS.
P-III-A-1
-------
INTRODUCTION;
Increasingly intensive crop and animal production have characterized
southeastern Pennsylvania's Conestoga Headwaters Watershed and the remainder
of Lancaster County. This production intensity has been motivated by:
(1) urban pressures on agricultural land, especially near the city of
Lancaster; (2) Amish and Mennonite family farming traditions, which have
resulted in the division of farmland into small parcels among children; and
(3) proximity to large East Coast markets. The watershed is 1 of 5
experimental RCWP projects nationwide to be selected for comprehensive
monitoring and evaluation. There are 44,516 hectares of land, of which 26,305
is in farms.
The "average" farm in the watershed has 21 hectares of land, almost entirely
used to produce feeds for animal operations that usually exceed 5 animal units
(AU) per hectare of farmland (USDA, 1982). Over a 13-year period from 1970 to
1983, the number of milking cows increased by 67 percent, swine production
increased by 189 percent, and poultry production increased by 145 percent.
During the same period, the percentage of total field and forage crops that
was in corn grain and silage increased from 33 and 12 percent in 1970 to 36
and 24 percent in 1983, respectively (Pennsylvania Crop Reporting Service,
1970, 1983). Commercial N fertilizer purchases increased over the same period
to meet the perceived needs of the additional corn.
Manure is applied at an average annual rate of 90 MT per hectare in the
watershed. The average fertilization per hectare for corn, from animal manure
and commercial fertilizers, is: 482 kg N, 304 kg ?2Q5' and 3Ql* k8 K2°
1983). These high nutrient inputs and the intensity of agricultural
production have resulted in local water quality degradation.
WATER QUALITY PROBLEMS
Agricultural nonpoint pollution in the Conestoga Headwaters Watershed affects
the water supplies of 175,000 people. The water quality parameters of
greatest concern are: (1) high NO^ concentrations in groundwater, especially
in private wells; (2) fecal coliform (FC) bacteria in surface and groundwater
supplies; (3) triazine pesticides in both surface and ground waters; and (4)
sediment and turbidity in streams.
Most private wells exceed the Environmental Protection Agency (EPA) drinking
water standard of 10 ppm NO^-N, with many exceeding 30 ppm N03~N concentration
during the year. Abnormally high levels of N, P, pesticides, and FC bacteria
also have been found in streams and wells (USDA, 1983). Average annual soil
losses are estimated to be 19.7 MT per hectare in the watershed, with annual
losses from individual fields often exceeding 100 MT per hectare. EPA has
identified southeastern Pennsylvania agriculture as a major contributor to the
Chesapeake Bay of N (especially N03 during spring runoff events), toxic
substances, sediment, and to a lesser extent P (EPA, 1983).
It is the goal of the RCWP project to substantially reduce the delivery of
agricultural pollutants to receiving waters by implementing conservation and
P-III-A-2
-------
other BMPs on farms. The purpose of this paper is to discuss the effective-
ness and cost effectiveness of selected management practices for controlling
field losses of soil and nutrients and ultimately for improving water quality.
This discussion is directed toward an area characterized by erosion problems
and high manure inputs.
BMP EFFECTIVENESS FOR CONTROLLING POLLUTION
Farmers in the RCWP project area currently are implementing contracted BMPs.
Water quality is being monitored and data are available only for current
management practices. It could be years before newly installed RCWP practices
cause measurable changes in water quality. As a result, computer modeling was
chosen to evaluate selected RCWP practices.
Modeling soil and nutrient losses
In order to isolate the effects of BMPs, both individually and in combination,
the CREAMS model was used to obtain estimates of field-scale soil and nutrient
losses. CREAMS is a computer simulation model, constructed by USDA's
Agricultural Research Service to compare relative field losses of soil and
selected chemicals among different management practices (Knisel, 1980).
Computer modeling provides a method of estimating BMP effectiveness for
pollution control without incurring the financial and time costs of BMP
implementation and monitoring.
CREAMS estimates of field losses were used with economic optimization and
budgeting techniques to evaluate the cost effectiveness of: (1) field BMPs;
(2) manure storage and handling systems (and their effects on nutrient
conservation/availability and field losses); and (3) manure transport (in
order to reduce high rates of manure application). Surface runoff losses of
soil, N, and P, and subsurface losses of NOj leached below the root zone, were
estimated using CREAMS and published data (see USDA, 1984). Pesticide losses
in surface runoff also can be modeled with CREAMS, but are not addressed in
this paper. Deliveries of soil and nutrients to surface and ground waters can
be assumed to be some fraction of the CREAMS estimates of field losses.
Soil losses
Table 1 shows the field BMPs that were modeled, their approximate costs, and
the reductions in losses of soil and nutrients associated with the BMPs. The
baseline soil loss was 24.3 MT per hectare per year. This represented the
soil loss for conventional practices — tillage up and down the slope with a
moldboard plow followed by 2 passes with a harrow. A distinction is made
between nonstructural management practices and structural practices.
Permanent vegetation, hay or pasture, is a highly-effective soil conservation
practice. It also appears to be relatively cost effective, with cost
effectiveness defined here as the percent reduction in field loss per dollar
of expenditure. Permanent vegetation is applied with RCWP cost sharing only
on severely erosive land that generally is adjacent to waterways. The
opportunity cost of removing land from row crop production is prohibitive
P-III-A-3
-------
without 50-percent government cost sharing, and therefore the practice has
limited applicability.
Contour stripcropping involves performing tillage operations along the
contours of the fields, and alternating row crops with close-seeded hay or
small grain crops in strips (modeled to be about 37 meters in width using
CREAMS). The effectiveness of stripcropping is primarily due to the reduction
in average surface runoff caused by the strips of close-seeded crops, not
accounted for by modeling the strip of corn in Table 1. Soil loss was reduced
by 39 percent at no cost, making contour stripcropping a highly cost-effective
practice for erosion control.
Conservation tillage practices — reduced tillage (chisel plowing followed by
harrowing once) and no-till — were found to be both effective and cost
effective for erosion control. No RCWP cost sharing is provided to convert
tillage practices. This is consistent with findings by Crowder, et al. (1984)
that conservation tillage practices reduce ownership and operating costs,
resulting in greater profitability.
Winter cover and residue management are important components of any
conservation, tillage system. When combined with the zero-cost practices of
conservation tillage and contour stripcropping, the overall management system
provides significant erosion control for the cost of winter cover and residue
management, $0 to $50 per hectare annually (depending on the crop residue).
The more important structural practices being implemented with RCWP cost
sharing include: (1) pipe-outlet terraces; (2) diversion systems; and (3) sod
waterway systems. The first column of Table 1 shows the approximate annual
costs per hectare served for structural practices, amortized at 12 percent
interest over 10 years.
Terraces are highly effective for erosion control, but at relatively high cost.
If a 10-year useful life is assumed, the annual cost per hectare is $163.70.
With 75-percent cost-sharing rates, however, terrace systems can be made
attractive to farmers. Sod waterways and diversions may be more effective for
preventing soil loss to streams than for reducing erosion. Given these
circumstances, they are probably more effective for protecting water quality
than soil productivity. Because they are considerably less expensive than
terrace systems, sod waterways and diversions appear most cost effective than
terraces for protecting water quality.
The last entry in Table 1 illustrates the reductions in pollution resulting
from a typical combination of soil conservation practices. An 89 percent
reduction in soil loss is possible at an approximate annual cost of $225 per
hectare. The cheapest combination of management and structural practices can
be chosen that will meet targeted soil loss reductions on a given field. In
this way, soil losses can be reduced at minimum cost to both farmers and
taxpayers .
P-III-A-4
-------
Table 1. Annual costs and effectiveness of conservation practices for continuous corn silage, daily spread
45 metric tons manure per hectare per year, 5 percent slope, Duffield silt loam.
I
M
H
M
>
Ul
Conservation
practice
Conventional
practices
Permanent vegeta-
tive cover
Contour strip-
cropping
Winter cover and
residue manage-
ment
Chisel plowing/
reduced tillage
No-till
Terrace system
Diversion system
with 6-meter wide
sod filter strip
Sod-waterway system
Reduced tillage
TJ*Q 4~ 1 m at" a f\
CiSt/imaueu
annual cost
per hectare
N/C
$83*
N/C
$0-50
N/C**
N/C**
$163.70*
$30.09*
$20.35*
$225
Soil
__
95
39
14
44
68
72
43
64
89
— — — rGFCGnU PGQUCulOn
N in
surface runoff
87
30
11
34
50
56
42
49
78
1 n f* \ ^1 t\ 1 f\QQ&Q
in i iej.u losses
Total N
__
45
18
9
20
38
32
24
29
43
Total P
_»__
91
33
13
37
54
62
46
52
82
with terraces,
contouring,
stripcropping,
residue manage-
ment , and sod
waterway
^Permanent vegetation was amortized over 5 years at 12 percent interest, and is cost shared at rates up to
50 percent. Structural practices were amortized over 10 years at 12 percent interest. Cost-sharing rates
of 75 percent on structural practices reduce farmers' annual before-tax costs to 25 percent of those shown.
**Cost of winter cover and residue management must be included for continuous corn.
Source: USDA (1984).
-------
Nutrient losses
As part of the RCWP economic evaluation, *J approaches to farm nutrient
management and conservation were identified:
1. Use field conservation practices (BMPs) to control surface runoff
losses of N and P.
2. Reduce animal manure production and/or commercial fertilizer
applications .
3. Improve the allocation and application of manure and nutrients
throughout the farm.
4. Export manure from farms with excess nutrient problems.
Nutrient losses in runoff can be reduced by implementing conservation
practices which restrict surface runoff and soil loss. The last 3 columns of
Table 1 show the reductions in nutrient losses for the BMPs discussed earlier.
Annual losses for conventional practices were 58 kg N per hectare in surface
runoff (of which almost 90 percent was sediment associated), 97 kg total N per
hectare (39 kg of NO^-N per hectare was leached through the root zone), and
28 kg total P per hectare in surface runoff.
The effectiveness and cost effectiveness of soil conservation practices for
controlling surface runoff losses of nutrients parallel those for controlling
soil loss. This follows because most nutrient losses in surface runoff were
sediment-associated for the baseline conditions modeled. However, soil
conservation practices have little effect on controlling NC>3 losses in
percolate. In fact, such practices may actually increase the load of NO^ lost
by leaching because of the increased water percolation through the soil
profile, caused by reductions in surface runoff. From the perspective of N
mass, the gains in N control are significant although not proportionate to the
reductions of soil and P losses. However much, and perhaps most, of the
sediment-associated N will not become biologically active, so the efficiency
of erosion control methods for controlling biologically active N losses is
even less than projected in Table 1.
As for other water quality parameters discussed, permanent vegetative cover
and the combination of BMPs (last entry in Table 1) are the most effective for
controlling total N losses, followed by no-till, terraces, and sod waterways.
Diversions, reduced tillage, and contour stripcropping estimates indicate that
they can significantly reduce total N losses. Among individual practices
(ignoring permanent vegetative cover due to its limited applicability and
effects on farm forage production), contour stripcropping, no-till, and then
reduced tillage are the most cost-effective nonstructural practices for
controlling total N losses. In addition, nonstructural practices have the
advantage of keeping soil and nutrients in place on farm fields. For
structural practices, total N loss reduction per dollar of cost is greatest
for sod waterways, then diversions, and last terraces — the same cost-
effectiveness ranking as for soil and P losses.
Decision makers should be aware that groundwater and stream pollution problems
are not likely to be alleviated by the implementation of soil conservation
practices alone if NOg is a critical water quality parameter. Improvements in
P-III-A-6
-------
surface water quality from surface runoff control may be offset by increased
NOg loads in groundwater and base flow. When groundwater NO^ levels are a
concern, as they are in the RCWP project area, nutrient management practices
probably will be necessary to improve water quality substantially.
jteducj^_nut£ient^ Applications
The second way to reduce farm nutrient losses is to match manure and
fertilizer nutrients more closely to crop needs. Available nutrients in
manure and legumes should be credited toward the fertilizer needs established
by soil test recommendations.
Part of the over-application of N could be prevented if farmers had
inexpensive, accurate estimates of available soil N. Unfortunately, such
estimates are not readily available, and the problem is further compounded
because the N that becomes available from current and previous manure appli-
cations is not easily predicted. Due to these uncertainties, it makes
economic sense to apply excess N commercial fertilizer as insurance to
guarantee crop yields. Over-application of nutrients would be expected to
decrease as the degree of uncertainty is reduced.
An example of the effects of over-application of nutrients, due to excess
manure (which is responsible for much of the nutrient problem in the project
area), is shown in Table 2. A manure application of 45 MT per hectare results
in substantially less nutrient loss than the average watershed application of
90 MT per hectare. Fertilizer and manure management plans are being imple-
mented through the RCWP program. Better timing of manure applications to
enhance crop uptake and prevent field nutrient losses, and eliminating the
large commercial N fertilizer applications that exceed crop requirements, can
reduce the stream delivery of nutrients as well as other pollutants. Reducing
fertilizer purchases also reduces farmers' costs. Reducing the application of
manure nutrients is difficult when excess manure is produced on a farm.
Storage, handling, and application systems which encourage N volatilization
can be used, or the number of animal units on the farm can be reduced — not
an attractive option for farmers.
f animalmanur
Significant reductions in farm nutrient losses may be obtained by applying
manure evenly throughout farm fields, and by properly storing manure during
crop growth periods and the winter months. For example, variation in soil NOj
levels within a single field varied from 3.3 to 230 kg of N03~N per hectare.
This was attributed primarily to "a relatively poor job of spreading nutrients
from manure." It was predicted that most groundwater pollution results from
those sites on fields with extremely high soil NOg levels (USDA, 1984). It is
believed that significant nutrient conservation and pollution reductions are
possible when manure nutrients are spread evenly on fields. Such a claim,
however, is difficult to quantify.
Table 2 provides an illustration of the estimated nutrient losses for
different manure storage periods and application methods. Losses shown are
field losses only, and do not account for barnyard runoff losses (which
usually are significantly greater on farms with unimproved barnyards, and
on farms where manure is stacked and spread periodically — a practice common
P-III-A-7
-------
Table 2. Effects of storage period on nutrient losses (kg/ha) for corn silage
following corn silage, 5 percent slope, Duffield silt loam, 45 and
90 metric tons of manure annually per hectare.
Application
Nitrogen Losses
Percolate Surface Total
Phosphorus Losses
Total
Daily spread
on surface
41
13
1,5 MT/ha
Six-months
storage with
plowdown
Six-months
storage with
injection
Twelve-months
storage with
plowdown
Twelve-months
storage with
injection
45
33
20
12
13
13
15
57
58
35
Daily spread
on surface
Six-months
storage with
plowdown
Six-months
storage with
injection
Twelve-months
storage with
plowdown
Twelve-months
storage with
injection
66
82
25
21
yu ni/na -
91
103
9
8
86
62
72
24
24
28
110
86
100
Practices include reduced tillage, winter cover, contouring, stripcropping,
with terraces and sod waterways installed.
Source: USDA (1984).
P-III-A-8
-------
among farmers in the RCWP project area). Manure storage systems conserve
nutrients, especially N, which is undesirable when plant-available N from
animal manures alone exceeds county crop nutrient requirements by 35 percent.
The P and K available from manures exceed crop nutrient requirements by 170
and 50 percent, respectively (USDA, 1983).
Field nutrient losses were similar for daily spreading and 6-month storage
systems at an annual loading rate of 45 MT per hectare of manure (Table 2).
Percolate NC^-N and total N losses were substantially less for 12-month
storage, an effective strategy for reducing nutrient loss. Also, nutrient
conservation by manure storage can offset most, if not all, of the amortized
costs of a storage structure. The economic modeling results indicate that,
for a small 45-cow dairy with 24 hectares of tillable land (resulting in
slightly less than 45 MT of manure per hectare applied to corn land), a
6-month earthen basin liquid storage system can increase annual net returns to
the farmer by $750 (with no cost sharing for the structure) compared to a
daily spread system. A 6-month solid storage system was found to decrease
annual net returns by $700 compared to daily spread.
If, however, a farmer in the RCWP project area stores manure and applies it at
the average rate of 90 MT per hectare, total N losses are about the same or
substantially greater than the daily spread system (Table 2). If a farmer has
excess manure nutrients, a storage structure no longer pays for itself because
the marginal productivity of the conserved nutrients is no longer positive.
Some reduction in total nutrient losses is possible by storing manure, but at
substantially more cost than for farmers who do not have excess nutrient
problems.
A critical concern for farmers is the effect of on-farm nutrient-loss
abatement on net income. For the representative small dairy farm mentioned
above, the results indicated that it was possible to reduce field losses of
nutrients by storing manure, by uniform spreading of manure, and by growing
less erosive crops or rotations on erosive fields. Further low-cost
reductions in nutrient losses can be obtained by spreading manure evenly
within the field borders, by performing conservation tillage operations along
field contours, and other conservation management practices discussed before.
These more cost-effective practices should be pursued to reduce field losses
before those which significantly reduce income and cost the government large
amounts of cost-sharing funds. In addition, if manure storage is planned on
farms with high animal-to-land ratios, it should be designed not to conserve
but to release unwanted nutrients by chemically degrading them to non-noxious
or biologically unavailable compounds.
_0f£-f_ar_m_di_s£os_al_ of_manure_
The last approach to reducing farm nutrient losses involves transporting
excess manure to other locations, where its application will not be environ-
mentally harmful and where its nutrients can be utilized more fully. With the
volume of excess manure that exists in Lancaster County and surrounding areas,
it is not anticipated that profitable manure marketing (for farmers with
excess manure) is possible on a wide scale. The trend toward more intensive
dairy and animal production will exacerbate the problem, and probably will
require future manure hauling over longer distances than those required
currently for safe disposal.
P-III-A-9
-------
Off-site manure disposal is costly for a number of reasons. First, manure is
a bulky material to handle and transport. Second, ensuring safe application
to land, so as to minimize nutrient losses, limits the time period that manure
can be applied. Manure applications must be compatible with crop growth
cycles. Thus, manure must be stored for up to 10 months, to allow most of the
manure to be applied prior to spring planting of row crops. Last, sufficient
hauling and application equipment should be available to apply manure during
this limited time period. Such equipment will stand idle during the remainder
of the year., Therefore, manure hauling over long distances is not a low-cost
method for reducing nutrient losses.
An option for hauling manure up to 10 miles, from the representative small
dairy farm discussed earlier, was modeled using linear programming. It was
assumed that farmers would pay for hauling and give the manure free to farmers
receiving it in order to reduce manure applications on the originating farm.
The receiving farmer would pay for field application. Using the CREAMS
estimates of N loss, it was found that for a farmer using a daily spread
system, income was reduced by about 20 percent when manure was hauled
40 miles to reduce total N losses by 30 percent. It was found under the
assumptions employed, that if a farmer stored manure for 6 months in an
uncovered solid storage structure, income was reduced by only about 8 percent
while achieving the same 30 percent reduction in total N losses. Thus, manure
storage structures may be necessary to cost-effectively institute off-farm
manure disposal in the RCWP project area and elsewhere. Cost-sharing funds
provided by RCWP and other programs for storage structures may make manure
export economically viable for farmers with manageable manure problems. It is
expected that manure export will impose substantial income penalties on
farmers who have substantially more manure than can be used on their farms,
unless government subsidies are provided.
Uses of manure other than fertilizer, most notably as fuel for power
generation, usually require economies of scale at the farm level to be
profitable. Refeeding manure to livestock causes health problems when done at
significant levels. Ongoing research on these and other alternatives may
provide hope for profitable off-farm manure utilization in the future.
CONCLUSIONS
It is apparent that while reductions in soil and P losses from farm fields can
be achieved with soil conservation practices, the control of N losses (as well
as FC bacteria) is more complicated. Reducing surface runoff and its
constituents results in increased percolation of water through the soil. This
potentially increases leaching of N03, pesticides, FC bacteria, and some small
amount of P to groundwater. In short, solving one problem likely aggravates
the other.
A comprehensive plan to promote sound management of manure, fertilizer, and
pesticides is necessary to improve water quality in the RCWP project area.
The rigorous application of manure management and soil conservation practices
could impose severe financial hardship on farmers if government assistance for
such practices is not continued. Even with the traditional forms of
assistance provided through RCWP and other programs, farmer participation in
P-III-A-10
-------
conservation planning has been poor in the project area due to the religious
and cultural mores of the farm population.
A number of problems must be addressed to solve southeastern Pennsylvania's
agricultural pollution problems. First, the level of pollution reductions
necessary to meet water quality goals must be identified for critical water
quality parameters. Second, those watersheds which have pollution problems,
and the critical areas within the watersheds, should be identified. The most
cost-efficient combination of practices should then be identified and selected
to meet the pollution reductions. Third, implementation policies are required
that will achieve the necessary farmer participation to efficiently and
equitably obtain the desired pollution reductions. Finally, the trend toward
increasing the intensity of land use should be constrained. Future expansions
of animal operations should be conditional on the adoption of adequate
nutrient control technologies. Gains made through better farm management
could be negated if cropping intensity and animal intensity continue to
increase without these controls.
REFERENCES
Crowder, B.M., Epp, D.J., Pionke, H.B., Young, C.E., Beierlein, J.G., and
Partenheimer, E.J. (198M). The Effects on Farm Income of Constraining
Losses of Soil and Plant Nutrients: An Application of the CREAMS
Simulation Model. Bulletin 850, Pennsylvania Agricultural Experiment
Station, The Pennsylvania State University, University Park, Pa.
Knisel, W.G., Ed. (1980). CREAMS: A Field Scale Model for Chemicals, Runoff.
and Erosion from Agricultural Management Systems. Conservation Research
Report No. 26, U.S. Department of Agriculture-Agricultural Research
Service, Washington, D.C.
Pennsylvania Crop Reporting Service (1970, 1983). Pennsylvania Crop and
Livestock Annual Summary. Harrisburg, Pa.
U.S. Department of Agriculture, Agricultural Stabilization and Conservation
Service (1982). Conestoga Headwaters Rural Clean Water Program - Plan of
Work. Harrisburg, Pa.
U.S. Department of Agriculture, Agricultural Stabilization and Conservation
Service (1983, 1984). Conestoga Headwaters Rural Clean Water Program
Progress Report. Harrisburg, Pa.
U.S. Environmental Protection Agency (1983). Chesapeake Bay: A Framework for
Action, Appendices. Philadelphia, Pa.
P-III-A-11
-------
BEST MANAGEMENT PRACTICES
FOR ANIMAL PRODUCTION
Jackie W. D. Robbins
Professor and Head
Agricultural Engineering Department
Louisiana Tech University
Ruston, Louisiana, USA
ABSTRACT
The environmental effects of grazing animals and how to control these are
presented. In general, pollutant yields from pasture and range!and operations
are very low. They are not directly related to the number of animals or
amounts of wastes involved. Rather, they are intimately related to hydro-
geological and management factors and are best described as the results of
the erosion/sediment phenomenon. Control is related to following practices
that are consistent with optimal forage production on a long-term, sustaining
bases and regulating animal movement/congregational patterns to minimize
development of highly impacted areas.
Keywords: Animal wastes, Pasture, Water pollution, Agricultural wastes,
Agricultural land runoff, Nonpoint pollution, Rangeland,
Livestock, Animal husbandry
P-III-B-1
-------
INTRODUCTION
This report summarizes state-of-the-art knowledge related to the yield and
control of water pollutants from animal production systems. Using findings
from recently completed research projects, this report updates earlier work
by the author that emphasized the character of the problem (Robbins, 1979).
Herein, emphasis is on control techniques or best management practices (BMP's)
aimed at keeping wastes from reaching surface waters.
By way of orientation, the difference between confined and unconfined animal
production is that wastes generated in confined systems are subject to
handling/diverting for conventional control or treatment, while those in
unconfined systems cannot be handled and, thus, must be controlled through
the management/operation scheme. Correspondingly, the contrast between
confined and unconfined animal production systems is very similar to the
differentiation between point and nonpoint sources of pollution, where a non-
point source is one whose specific point of generation and exact point of entry
into the environment cannot be defined. All grazing systems—where livestock
have free access to pasture, range, woodland, or cropland and utilize the
associated forage/residue as a major feed source—are unconfined systems.
About 40% or 360 x 106 ha of the land area of the United States is used for
grazing livestock. Forages, which in the main are grazed, account for the
production of more than 50% of the nation's milk, nearly 80% of the total feed
for beef cattle, and 90% of the total feed for sheep. Less than 5% of swine
feed and almost no poultry feed comes from grazing.
While approximately 50% of all livestock wastes in the United States are
estimated to be produced in confinement, about 75% of cattle, 85% of sheep,
and 10% of hogs are in unconfined systems at any given time. Table 1 lists
the numbers of animal production systems and animals maintained in unconfined
production systems. The vast numbers of individual units involved in
unconfined livestock activities are reflected by the values given for numbers
of production units (about one million). Although specialization, labor
reduction, and improved efficiency have caused a trend toward confined
production systems, unconfined production is expected to continue to
predominate the beef and sheep industries.
IMPACT ON PASTURE AND RANGELAND
As grazing animals traverse pasture and rangeland, the stresses applied to
the soil beneath their hooves often exceed the strength of the soil.
Trampling may cause poorer water infiltration. Reductions in infiltration
rates due to grazing correlate with increases in runoff as indicated by
numerous studies. For example, White et al. (1982) found that where winter
feeding of cattle occurred, the runoff volume was three times that of normal.
Winter rotational grazing increased runoff by 75% and summer rotational
grazing only, increased runoff by 17%. The increased runoff volumes were
primarily associated with the dormat season, winter feeding and pasturing.
Trimble et al. (1951) measured average infiltration rates of 52 and 11 cm/h
on ungrazed and grazed pastures, respectively. In a New Mexico study of
grazing intensity-infiltration relationships, infiltration rates were
determined to be 10.5, 5.5, and 2.1 cm/h on undergrazed, overgrazed, and
depleted ranges, respectively (Flory 1936).
P-III-B-2
-------
TABLE 1. DISTRIBUTION OF LIVESTOCK (After Robbins, 1979)
Type Animal
No. of
Production
Units
No. of
Animals
No.
Confined
No. Un-
Confined
Percent
Unconfined
Cattle and Calves
Beef Cattle
Dairy Cattle
Sheep and Lambs
Goats and Kids
Hogs and Pigs
Horses and Ponies
490
220
50
10
260
Thousands
123 000
101 500
21 000
12 700
1 300
55 000
8 000
30 000
20 000
10 000
2 000
200
50 000
93 000
81 500
11 000
10 700
1 100
5 000
76
80
52
84
85
9
P-III-B-3
-------
After a comprehensive review of over a dozen studies, Moore et al. (1979) were
able to summarize that light and moderate intensities of grazing during
unsaturated soil conditions will not decrease infiltration rates. However, a
concentration of livestock is likely to produce significant decreases in
infiltration. Adverse impacts on sheet and rill erosion and water quality
constituents are an outcome of these changes. The phenomena of heavy
trampling damage to infiltration rates may be an insignificant problem on fine-
textured high clay content soils where undisturbed rates are less than 25 cm
per hour. It is well documented that periods of non-grazing, however built
into the grazing system, allow certain natural processes to progressively
improve soil water infiltration rates to predisturbance levels.
A decrease in infiltration and increase in runoff on grazed lands may lead to
more arid conditions than normal. For example, Hanson et al. (1970) found
that the reduction in available moisture on heavily grazed South Dakota
ranges averaged 2 cm a year, or about 8% of the annual precipitation available
for plant production. A decrease in moisture available for forage production
is usually accompanied by an increase in soil erosion.
Selective defoliation by grazing animals decreases the preferred forage species
and allows the proliferation of less desirable species. Close grazing of
desired plant species during critical stages of growth, such as before seed
maturation, hinders their competitive capacity. Reduction of the leaf area of
forages by close grazing reduces their photosynthetic abilities and may affect
plant succession (Stoddart et al., 1975). Heady (1975) attributes the most
prevalent cause of range!and retrogression to overgrazing and other faulty
management of livestock operations.
The general congregational and waste elimination behavior patterns of live-
stock follow ancestral instincts, but may be influenced by management
practices. Behavior patterns and preferences of livestock species tend to
result in certain areas showing intense environmental stresses. These
"problem areas" are usually only a small fraction of the total unconfined
production site and are caused by unregulated animal behavior that leads to
selective—and over—grazing (of sites as well as plant species) and to animal
congregation and waste elimination in or near water sources or other critical
areas. While only a small portion of the total production area, these problem
areas often contribute the major portion of pollutants and constitute the only
sources that when controlled results in a discernable improvement in runoff
quality.
BACKGROUND LEVELS OF POLLUTANTS
In assessing the environmental impact of unconfined animal production, the
concept of "background (natural) levels" of pollutants is important. From
a study of 12 agricultural watersheds, Robbins et al. (1971) found that dis-
tinguishing between pollutants from farm animal production units and natural
pollutants in receiving streams is difficult or impossible. More recent
studies designed to establish the contributions of pollutants from unconfined
animal production to surface waters by Moore et al. (1979), Doran et al.
(1982), White et al. (1982), Powell et al. (1982), and Saxton et al. (1982)
all reemphasize that control of pollutants from unconfined animal production
units may be to no avail unless other pollutant sources that naturally occur
in the watershed are controlled as well. When management is directed to
P-in-B-4
-------
optimizing forage production, pollutant yields from unconfined animal
production systems are not more than would occur under native conditions.
Only problem areas as noted earlier and areas used more to confine animals
rather than to graze them would seem to be subject to corrective actions.
EROSION/SEDIMENT
Sediment is both a pollutant and a carrier of pollutants. As a first
approximation, factors that govern erosion and sediment yields are the same
factors that control pollutant yields from unconfined animal production
systems. Numerous studies including Robbins et al. (1978), Smeins (1976),
Moore et al. (1979), Saxon et al. (1982), White el al. (1982), and Powell
et al. (1982) have pointed to the need for good soil and water conservation
practices to minimize the movement of pollutants from animal production units
into streams. Excellent, detailed and thorough reviews and discussions of
erosion, erosion models, and sediment yields as related to agricultural non-
point sources have been prepared by USEPA (1973), Stewart et al. (1975),
and Sweeten and Reddell (1978). Sediment yield to streams and lakes exceeds
2 t ha'lyr"! on the average. On-site erosion is estimated to be twice this
value for an average of more than 4 t ha~lyr~l or a total of 3.6 x 10^ t/yr
for the total U.S. land area. According to Froehlich (1976), the smallest
levels of sediment yield are from certain undisturbed forestlands in the
Rocky Mountains where sediment reaching high elevation streams may range up
to 0.36 t ha-lyr'l. Other examples of background sediment yields are 0.59,
0.90, and 1.10 t ha-1yr~^ from three neighboring watersheds in western Oregon.
Well stocked southern pine forests my yield 0.7-1.1 t ha"lyr~l. In south-
western forests with low precipitation, yet high intensity storms, sediment
yields average 1.1-1.4 t ha~lyr~l.
Background sediment yields may be compared with yields from cropland. Crop-
land has been credited with responsibility for 50% or 0.9 x 10^ t/yr of the
1.8 x 10^ t/yr total sediment delivered to U.S. streams and lakes. As noted
by Sweeten and Reddell (1978), 70% of the nation's cropland yields more than
6.7 t ha'iyr'l. Representative values of on-site erosion from several sources
as reported by USEPA (1973) are given in Table 2. Here, grassland includes
pasture and rangeland. While the erosion rate from grassland is 10 times
that from forestland, it is considerably less than the average rate of
4 t ha~lyr~l for all land; and it represents the average background or
natural rate for grassland ecosystems.
Sediment yields are very erratic. The largest annual sediment loads for a
given stream are often 20 times greater than the smallest sediment load and
generally can be correlated with those years of greatest runoff. Large
differences in sediment yield can exist on adjacent streams discharging at
the same rate. And, even in the same stream, suspended sediment concentration
can vary tenfold at a given discharge rate, depending on many factors
(Froehlich, 1976).
Most sediment from a watershed may come from a relatively few small areas
needing corrective attention. Stewart et al. (1975) listed conditions
indicative of high sediment yield potential that usually can be identified
by observation. Overgrazing and resultant loss of groundcover can increase
greatly the credibility of pasture and rangeland.
P-III-B-5
-------
TABLE 2. REPRESENTATIVE RATES OF ON-SITE EROSION FROM VARIOUS LAND USES
(USEPA 1973)
Rate, Relative Rate, Total, Relative Total
Land Use t/ha yr Forest = 1 t/yr Forest = 1
Forest 0.085 1 16.8 1
Grassland 0.85 10 185 11
Cropland 17 200 2840 168
Harvested Forest 42 500 187 11
Construction 170 2000 100 6
p-iil-B-6
-------
Results from four test sites that exemplify water quality effects resulting
from suspected problem sites and unimpacted sites are given in Table 3. These
results vividly emphasize the need to consider problem areas separately from
the remainder of the unconfined production system and to design corrective
actions to effectively control the impact of these. Yields of pollutants from
heavily stocked pastures are exemplified in Table 4, along with that from
other land uses. As indicated, heavily stocked pastureland contributes less
than cultivated fields but more than permanent grasslands. Thus, improved
pasture management may be called for to control the pollutants from
such heavily used areas. The pollutant levels from other than problem areas
and heavily stocked pastures are seldom discernible from background levels and,
on an area! basis, are of the same magnitude as yields to land from rainwater
and yields to surface waters from undisturbed lands.
Prediction of increases in pollutant yields to receiving waters due to
unconfined animal production is not possible with existing technology. As
measured in receiving streams, these increases usually are not only small in
value but extremely erratic as well. As a first approximation, models for
erosion/sediment yields may be adopted/used to predict pollutant yields from
unconfined livestock systems.
BACTERIAL COUNTS
Bacterial indicators are reputed to be the most sensitive index of water
quality impacts resulting from unconfined livestock. But the actual signi-
ficance or meaning of elevated bacterial counts in this case is not clear.
While the yield from pastured areas may be sometimes several fold that from
control areas (White et al., 1982), the opposite is also not uncommon (Saxton
et al., 1982). Their use in evaluating the effect of unconfined livestock
production on runoff is certainly difficult, if not impossible. A growing
consensus among researchers is that their use in evaluating nonpoint pollution
is inappropriate. These organisms are properly used to designate the
bacteriological safety of potable water. Their presence in water is accepted
by public health authorities as an indication of fecal pollution. Whenever
water is polluted with fecal material, it poses a potential health hazard if
used untreated for drinking or body contact recreational purposes. There is
a dearth of documented cases of health problems correlated with bacterial
indicator counts in water related to unconfined animal production systems,
but this may be due solely to the lack of intensive and systematic research in
this area.
CONTROLS
While certain specific questions such as desired level of control and best
methods to achieve water quality goals remain unresolved, the basic concepts,
methods, processes, and procedures needed to reduce pollutant yields from
unconfined animal production systems are available in principle. Since the
pollutants are nonpoint in origin, control methodology is effected through
changes in the production management system; i.e., by incorporating practices
involved in controlled grazing, regulated animal congregation patterns, sus-
tained forage production (establishment and protection of desired species),
intensive erosion control, and proper land use. This may, under extremely
adverse site conditions, call for complete exclusion of livestock.
P-III-B-7
-------
TABLE 3. WATER QUALITY BY LAND USE (SEWELL AND ALPHIN 1972)
Land Use
BOD5
DO
NOa-N , POa-P
ppm
Total Coliform,
cts/100 ml
Ungrazed
Woodland
Heavily-Grazed
Pasture
Farm Pond in
2.5
13.8
8.6
6.1
0.05
4.5
0.05
7.1
1500
330000
Pasture
Farm Pond in
Woodland
10.0
3.1
7.9
7.6
0.20
i
0.12 '
0.05
0.02
600
1500
TABLE 4. POLLUTANT
Pollutant
Total Residue
Suspended Solids
Total Phosphorus
Nitrate-Nitrogen
Total Kjeldahl Nitrogen
Chemical Oxygen Demand
YIELD BY LAND USE
Cultivated
334
286
0.30
0.37
0.91
48
(DORNBUSH ET AL.
Yield, kg ha" yr
Pasture
58.2
11.8
0.25
0.40
1.12
28
1974)
-1
Grassland
32.4
4
0.1
0.24
0.73
13
P-III-B-8
-------
Pasture and rangeland management practices leading to optimal forage pro-
duction on a long-term sustaining basis, such as that resulting from maintaining
proper stocking rates and forage production practices with erosion control, are
also those practices which minimize environmental impacts. On the other hand,
when exploitive (poorly managed) production is followed, unnecessary detri-
mental effects on surface water quality are likely to occur. These systems
may require major modification/adaptation to meet environmental goals as well
as to achieve other objectives which govern pasture and rangeland use.
As detailed by Robbins (1978), when developing or modifying management practices
for unconfined animal production systems, consideration should be given to the
following concepts:
1. Multidisciplinary and multiagency review teams should be used to evaluate
specific environmental problems resulting from unconfined animal production
and to recommend changes in management practices to correct the problems.
Their recommendations should be applied in concert with established
regulatory procedures.
2. Regulatory programs should account for local (site specific) conditions.
Regulated changes in management practices should be restricted to those
documented to have measurable water quality benefits for receiving waters.
Due to the low level of pollution associated with unconfined animal
production, regulatory programs that would in general discourage or
restrict livestock production on pasture and rangeland should be avoided.
3. Install and maintain an effective and complete program of soil erosion
control.
4. Follow stocking rates and controlled utilization of forages (e.g.,
rotation, deferred, and seasonal grazing) that reduce erosion and minimize
waste accumulation. Reduce stocking rates in problem areas and at
critical times or seasons. Stocking rates and grazing programs should
be tailored to the soil, vegetation, topography, hydrogeology, and micro-
climate of the particular site.
5. Avoid animal stocking rates and other practices that create holding areas
rather than grazing areas. Promote necessary animal congregation in
areas that are hydrologically remote from streams and other major drainage
channels. Periodically move bedground, shelter, salting, feeding, and/or
watering areas to control waste accumulation, soil compaction, and
erodible paths and areas.
6. Maintain to the extent ecologically feasible, highly productive forage
and dense ground cover on the land to decrease volume and rate of runoff,
to entrap and hold animal wastes, to utilize fertilizer nutrients, and to
prevent erosion. Increase herbaceous cover in proximity of stream banks,
downslope from animal congregational areas, and on other critical areas.
7. Where the number of animals per unit area or the characteristics of the
site present pollution problems, appropriate drastic management alternative/
practices may include:
p-lii-B-9
-------
(a) Restrict animal access to critical areas. Use fencing to prevent
livestock from entering highly credible areas and critical stream
or pond reaches and to prevent animals from wading in water.
Provide summer shade (trees or artificial shelters) and insecticides
to lessen the need for animals to enter water for relief from heat
and insects. Block erodible paths with physical barriers and
revegetate eroded paths. Move drinking facilities outside
critical areas.
(b) Increase rate of fecal degradation and incorporation. Modify feed
formulation and/or texture. Use tillage to break up, manipulate,
and incorporate wastes in particular problem areas.
(c) Use land forming and diversions to modify drainage patterns.
Total stream fencing and other drastic controls will usually be both unnecessary
and impractical except for a few problem areas within an operation.
CLOSING REMARKS
One major challenge still remaining is to demonstrate cost-effective routes
toward achievement of various levels of pollution control for unconfined animal
production systems. Whether resources should be used to reduce pollutant yields
from a given production system really requires an on-site evaluation of whether
the pollutants have a real adverse effect on the quality of the receiving water
and whether the yields can be significantly reduced. In view of the many
desirable characteristics of unconfined animal production compared with the
problems associated with alternate livestock production techniques, adequate
research, management, and educational resources should be allocated to assessing
the importance and significance of increased levels of pollutants (indicator
bacteria, plant nutrients, and oxygen demands) in surface waters (streams and
impoundments), with consideration given to fate of the pollutants and uses of
the waters. In cases where pollutants from livestock activity are not dis-
cernible from background levels, little or no change may be noted in stream
quality even when significant efforts and resources are expended on control.
When grazing is approached using well-planned management practices that include
pollution control as an integral part, rather than a laissez-faire approach,
the impact of unconfined animals will be well within acceptable limits and in
keeping with water quality goals.
ACKNOWLEDGEMENTS
This paper is based on work supported in part by funds from the U.S.
Environmental Protection Agency. The grant, No. R805559, was administered
through the Robert S. Kerr Environmental Research Laboratory, Office of
Research and Development, Ada, Oklahoma 74820.
P-III-B-10
-------
REFERENCES
Doran, J.W., Schepers, J.S., Swanson, N.P., and McCalla, T.M. (1982).
Chemical and bacteriological quality of runoff from a central plains pasture-
land. Unpublished report for IAG 78-D-X0088, U.S.EPA, Ada, OK.
Dornbush, J.N., Andersen, J.R., and Harms, L.L. (1974). Quantification of
pollutants in agricultural runoff. EPA-660/2-74-005. Superintendent of
Documents, U.S. Government Printing Office.
Flory, E.L. (1936). Comparison of the environment and some physiological
responses of prairie vegetation and cultivated range. Ecology 17:67-103.
Froehlich, H.A. (1976). Inorganic pollution from forests and rangelands.
Publication No. SEMIN-WR-021-76. Water Resources Research Institute, Oregon
State University.
Hanson, C.L., Heinemann, H.G., Kuhlman, A.R., and Neuberger, J.W. (1970).
Grazing effects on runoff and vegetation on western South Dakota rangeland.
J.Range Management 2^:418-420.
Heady, H.F. (1975). Rangeland Mangement. McGraw-Hill Book Company, New York.
Moore, E. (ed.) et al. (1979). Livestock grazing management and water quality
protection. EPA-910/9-79-67. National Technical Information Service,
Springfield, VA.
Powell, J., Crow, F.R., and Wagner, D.G. (1982). Rangeland watershed water
budget and grazing cattle waste nutrient cycling. Unpublished report for Grant
No. R-803735. U.S.EPA, Ada, OK.
Robbins, J.W.D., Howells, D.H., and Kriz, G.J. (1971). Role of animal wastes
in agricultural land runoff. EPA-13020DGX. Superintendent of Documents, U.S.
Government Printing Office.
Robbins, J.W.D. (1978). Environmental impact resulting from unconfined
animal production. EPA-600/2-78-046. National Technical Information Service,
Springfield, VA.
Robbins, J.W.D. (1979). Impact of unconfined livestock activities on water
quality. Transactions of the ASAE (22|6:1317-1323.
Saxton, K.E., Elliott, L.F., Papendick, R.I., and Jawson, M.D. (1982). Effect
of animal grazing on water quality of nonpoint runoff in the pacific northwest.
Unpublished report for IAG D6-0030 and 78-D-X0249. U.S.EPA, Ada, OK.
Sewell, J.I. and Alphin, J.M. (1972). Effect of agricultural land use on the
quality of surface runoff. Progress Report 82. Tennessee Farm and Home
Science, University of Tennessee.
Smeins, F.E. (1976). Influence of vegetation management on yield and quality
of surface runoff. Annual Report No. C-6310. Texas Water Resources Institute,
Texas A&M University.
Stewart, B.A. (ed.) et al. (1975). Control of water pollution from cropland:
volume I. EPA-600/2-75-026a or ARS-H-5-1. Superintendent of Documents, U.S.
Government Printing Office.
Stoddart, L.A., Smith, A.D., and Box, T.W. (1975). Range Management, 3rd
edition. McGraw-Hill Book Company, New York.
Sweeten, J.M. and Reddell, D.L. (1978). Nonpoint sources: state-of-the-art
overview. Transactions of the ASAE (21)3^:474-483.
Trimble, G.R., Jr. and Weitzman, S. TT951). Effect of soil and cover conditions
on soil-water relationships. Station Paper No. 39, U.S. Forest Service, North-
eastern Forest Experiment Station.
USEPA. (1973). Methods for identifying and evaluating the nature and extent of
nonpoint sources of pollutants. EPA-430/9-73-014. Superintendent of Documents,
U.S. Government Printing Office.
P-III-B-11
-------
White, R.K.,, VanKeuren, R.W., Owens, L.B., Edwards, W.M., and Miller, R.H.
(1982). Effects of livestock pasturing on nonpoint surface runoff. Unpublished
report for Grant No. R-804631. U.S.EPA, Ada, OK.
P-III-B-12
-------
COMPARATIVE POLLUTANT REMOVAL CAPABILITY. ECONOMICS
AND PHYSICAL SUITABILITY OF URBAN BEST MANAGEMENT PRACTICES
IN THE WASHINGTON B.C. METROPOLITAN AREA
Thomas Schueler, Robert Magill, Michael P. Sullivan and Cameron Wiegand
Department of Environmental Programs
Metropolitan Washington Council of Governments
1875 Eye St. Washington DC 20006
ABSTRACT
Recent field studies have assessed the comparative pollutant removal
capability of six kinds of urban stormwater best management practices
(BMPs) in the Washington metropolitan area. While some urban BMPs were
found to be very effective in removing pollutants entrained in urban runoff
(wet ponds, extended detention, porous pavement), the most frequently used
stormwater BMPs, dry ponds and grassed swales, exhibited little or no
pollutant removal capability. Local governments have recently sought to
incorporate water quality objectives into existing stormwater management
programs by encouraging BMPs with high pollutant removal capability. This
strategy has not led to widespread implementation because of uncertainties
about the physical, economic, and environmental constraints of these BMPs
at the site level.
A systematic approach for screening urban BMP options is presented, based
on a comparative review of the suitability of BMPs with respect to
pollutant removal, cost, and site/environmental suitability for nine
hypothetical development scenarios. A three level screening matrix is
developed which provides rapid guidance to site planners on the suitability
of seven urban BMP options. Major elements evaulated in the matrix include
site restrictions such as soils, slope and site area, space requirements,
pollutant removal, local and downstream habitat modification, landscaping
and aesthetic value, total cost, and operation and maintenance
requirements. The use of the screening matrix for developing effective,
low cost urban nonpoint source control strategies is also discussed.
INTRODUCTION
Stormwater management programs have been in effect in the Washington, D.C.
area since the early 1970s with the objective of controlling increases in
post-development peak runoff rates. To date, nearly 3,500 structures have
been constructed. At least 75% of all new regional development is now
serviced by some kind of stormwater control, with dry ponds being applied
in about two-thirds of all development situations (MWCOG,1984a).
P-III-C-1
-------
As more information has become available on the impact of urban runoff
loadings on important area receiving waters, Washington area local
governments have actively sought to incorporate nonpoint pollutant removal
into their existing stormwater management programs, chiefly through the use
of alternative urban "Best Management Practices" (BMPs). These
alternatives rely on either greater retention of stormwater (wet ponds,
extended detention ponds), or enhanced infiltration (porous pavement,
grassed swales, infiltration basins and trenches).
Despite strong local institutional and legislative commitments to improve
urban stormwater quality, less than 20 percent of all stormwater control
structures installed or designed during 1982 and 1983 had significant
pollutant removal capability. (MWCOG,1984a) At the site planning level, the
most frequently cited reason for the lack of widespread implementation has
been uncertainty about the cost, efficiency and suitability of alternative
BMP options.
In 1983, the Metropolitan Washington Council of Governments completed a
four year BMP monitoring and cost survey under the EPA Nationwide Urban
Runoff Program (NURP). The objective of the research was to provide a
comparative assessment of the relative suitability of selected urban BMPs
for nonpoint source control. Subsequent research has examined the degree
to which urban BMPs can be applied over the range of site conditions
encountered in the region. This paper summarizes the major findings of
these efforts, and presents a systematic screening procedure for evaluating
alternative BMP options at the development site level.
URBAN BMP ASSESSMENT RESEARCH
BMP Pollutant Removal Field Studies.
The pollutant removal capability of six different urban BMPs was monitored
over a 12 to 18 month period in suburban Maryland and Virginia. BMP
monitoring sites included dry ponds (1), extended detention dry ponds (1),
wet ponds (3), grassed swales (3), infiltration pits and trenches (2) and
porous pavement (1). Flow-activated automated water quality samplers were
installed at the inlets and outlets of 11 BMP sites, which drained small (3
to 90 acre), single land use watersheds. Flow weighted composite samples of
nutrients, sediments, COD and trace metals were collected for 20 to 50
storm events at each site. Pollutant removal efficiency was calculated in
terms of total mass outflow load/total mass inflow load for volume BMPs,
and outflow median loading rate/inflow median loading rate for infiltration
BMPs. More detailed discussion of monitoring and data analysis methods
employed in the study can be found in the final project report
(MWCOG,1983a)
BMP Cost Surveys
Construction costs for pond BMPs were evaluated in an extensive local cost
survey (MWCOG,1983b). Construction components and requirements were
identified for 31 representative SWM ponds in the Washington metropolitan
area (e.g. cubic yards of excavation, linear feet of pipe etc.). In-place
unit cost estimates for each pond component or construction requirement
were then independently solicited from 12 public and private construction
P-III-C-2
-------
and engineering firms. By applying the unit cost estimates to each of the
31 ponds, 372 separate estimates of construction cost were generated.
Subsequent statistical analysis of the dataset yielded a strong (^=0.84),
curvilinear relationship between pond storage volume and construction cost
(Table 1).
No systematic cost surveys of infiltration BMPs could be developed since
very few have been installed in the region. As an indirect approach, the
basic geometry of infiltration BMPs was inferred from design standards for
each site (MD DNR.1984). Construction requirements for each site were then
estimated based on site geometry, and construction costs were estimated by
multiplying construction requirements by unit cost data obtained in the
pond cost study (Table 1).
Estimates of average O&M costs for pond BMPs were obtained in a survey of
O&M budgets and component cost records of six public works agencies
(MWCOG,1983b). Typical O&M activities reported included inspection, grass
mowing, debris removal, bank stabilization, structural repair, nuisance
control and sediment removal. Analysis of O&M budgets suggested that
regular performance of these tasks would entail an annual cost equivalent
to 3 to 5% of pond BMP construction cost. O&M requirements for
infiltration BMPs are generally different in scope and frequency; the best
available methods for evaluating these costs are shown below.
Table 1. METHODS USED TO CALCULATE COSTS FOR SELECTED URBAN BMPI IN THE WASHINGTON DC METROPOLITAN AREA
BMP TYPE
CONSTRUCTION CONTINOfNCY 0 fc H COST
COST (C) COST (CC) (annual|
COMMENTS/NOTES
DRY POND
EXTENDED
DETENTION
DRY POND
GRASSED ROADSIDE
SWALES
82.51Vs*«0.51 C * 0.25
C • 0,05
82.5Ws"»0.51 C • 0.25
C • 0.05
82.51Vs«0.5l C « 0.25
C • 0.05
Equations derived in Washington DC Stonawater
Pond Cost Study (MWCOG,1983).
O&M Costs were computed assuming a conservative
value of 5%
Contingency costs are defined as costs Incurred In
planning, designing and overseeing construction of
BMP
Vs
volume of pond up to the crest of the
emergency spillway (in cubic feet)
INFILTRATION
BASIN
INFILTRATION
TRENCH
POROUS PAVEMENT
(PPfl)
Not Space
Limited
POROUS PAVEMENT
(PPI2)
Space
Limited
66.23Vs*"0.51
Construction
Requi rements
Determined by
Design Geometry
Increments I
Cost over
Conventional
Pavement
As above, +
cost of extra
reserve! r base
under asphalt
for runoff
storage
C • 0.25
C • 0.25
C * 0.35
Includes
extra cost
for tddl.
site surveys
soil testing
inspection
C • 0.35
C • 0.05
$286 • SA
of trench
* Buffer
SA in acres
SO. 003 «
SA of PPf 1
(commerc la 1
rate for
vacuum street
sweeping
it times/yr
SO. 003 • SA
of PPf 2 site
(SA In sq.ft)
Equation derived by factoring out the costs of
outlet works and structures from the pond cost
equation.
OlcM requirements assumed to be similar to ponds.
Construction requirements Include: excavation,
fill of coarse stone or gravel, filter cloth,
test wells, and seeding/mulching for buffer strip
OteM based on cost/maintained acre In (MWCOC,1983)
Incremental costs include: higher unit costs of
porous asphalt relative to conventional asphalt,
extra costs for filter cloth, and test wells.
SA for PP|1 * 60 to 70% of total site area
Depth of extra base course computed from State of
Maryland Design Specifications. Associated extra
cost for excavation and coarse stone fill derived
from unit cost data contained in pond cost study.
SA for PP4>2 • impervious site area/2
(13 to M3X of total site area)
Not Not Not Grassed roadside swales generally assumed to be
Calculated Calculated Calculated loss expensive than the curb and gutter convey-
ance systems they replace.
TOTAL BMP COST It defined (C + CC) + OHpv; where OMpv Is the present value of all future O&M payments, assuming a twenty
year project life and an 8{ discount rate. Land costs were assumed to be zero for all BMP*.
ANNUAL BMP PAYMENT was computed assuming a twenty year loan at 6% interest rate
P-HI-C-3
-------
Physical and Environmental Suitability Surveys
Physical or environmental conditions which might restrict the application
of a particular BMP were evaluated in a review of local and state BMP
design manuals and other relevant literature. Approximately thirty site
planners, engineers and public works officials involved in BMP construction
were also interviewed to obtain their perceptions and experiences about the
limitations and attributes associated with each BMP option.
Development of BMP Screening Model
To arrive at a comparative assessment of BMP options at the site level,
nine hypothetical development site scenarios were examined in detail.
These scenarios included single family residential (2 DU/Ac--25%
impervious), townhouse (8 DU/Ac--65% impervious), and commercial
development (85% impervious) situated on 1, 10, and 25 acre sites. The
TR-55 tabular method (SCS,1975) was used to determine the increase in
post-development peak discharge and runoff volume associated with the two
year design storm at each site. Curve Numbers (CN) of 70, 85 and 92 were
assumed for the single family residential, townhouse, and commercial
developments, respectively. All sites were assumed to have soils in
hydrological group "B", and slopes of 3-5%. The stormwater management
criteria for all sites was to maintain post-development peak discharges
(pond BMPs) or increases in upland runoff volumes (infiltration BMPs) to
pre-development levels for the two-year storm (defined as meadow,CN 58).
Required runoff storage volume was determined for each BMP by either, the
short-cut routing method (pond BMPs) or State of Maryland infiltration BMP
design standards (MD DNR, 1984).
The total cost of installing and maintaining each BMP on the nine
development sites was computed using the methods outlined in Table 1.
Physical constraints, such as space requirements, were evaluated assuming a
standard geometrical design for each site. Environmental factors were
assigned to each BMP site based on the consensus among site planners and
engineers surveyed.
RESULTS OF THE COMPARATIVE URBAN BMP ASSESSMENT
Comparative Pollutant Removal
Long term pollutant removal observed at the BMP sites monitored is
summarized in Table 2. As can be seen, the degree and kind of pollutant
removal achieved varied widely among the BMPs studied. For example, the
dry pond investigated was not effective in removing sediments, nutrients,
trace metals or oxygen-demanding materials. Urban stormwaters passed
through the dry pond too rapidly to achieve significant removal via
settling. During some storms, the dry pond actually became a net source of
pollutants, as pollutants were either scoured or resuspended from the pond
bed.
Extending detention time in dry ponds by 6 to 12 hours proved to be
effective in removing sediment, organic nutrients and trace metals. Removal
of biologically available nutrients and other soluble pollutants, however,
was negligible. Wet ponds were found to be the most effective pond BMP
P-III-C-4
-------
studied. Removal of sediment, trace metals and organic nutrients was
comparable to rates observed at the extended detention pond. In addition,
biological processes occurring within the pond effectively removed the
soluble nutrients most readily utilized in algal growth.
No statistically significant improvement in runoff quality could be
attributed to grassed swales when compared to traditional curb and gutter
conveyance systems. The poor performance of swales appeared to be due to
the characteristics of the sites monitored--moderate slopes (4 to 6%),
relatively permeable soils, and short grass. It is possible that better
removal might have been achieved by swales with check dams on sandy,
relatively flat sites. However, these conditions are not common in the
Washington metropolitan area.
Monitoring difficulties and site limitations prevented a valid statistical
assessment of pollutant removal by infiltration trenches. Trench BMPs did
appear to be capable of removing sediments and trace metals, but did not
show any capability for removing nutrients. It should be noted that the
trenches monitored had outlet structures, and thus did not function as
"true" infiltration BMPs (which infiltrate incoming runoff completely into
the soil profile). If complete infiltration is accomplished, as is
currently recommended, it is likely that infiltration BMPs would exhibit
much higher removal efficiencies. This, in fact, was the case at the
porous pavement site monitored. Although the site did have perforated
underdrains (and therefore could be monitored), significant infiltration of
runoff into the underlying soil was achieved. As a result, removal of-most
pollutants at the porous pavement site approached 90%.
Table 2. COMPARATIVE POLLUTANT REMOVAL OF URBAN BEST MANAGEMENT
PRACTICES IN THE WASHINGTON D.C. METROPOLITAN AREA NURP
FIELD MONITORING STUDY
DRY EXTENDED WET INFILTRATION POROUS GRASSED
POND DETENTION POND BASIN TRENCH PAVEMENT SWALES
DRY POND
PARAMETER
PERCENT REMOVAL OF POLLUTANTS
SEDIMENT
14
55
50
>95
NSD
TOTAL
PHOSPHORUS
ORTHO-
PHOSPHORUS
TOTAL
NITROGEN
NITRATE &
NITRITE
TKN
EXTRACTABLE
LEAD (Pb)
20
0
10
9
11
-5
EXTRACTABLE
ZINC (Zn) -10
COD 0
10
24
10
15
84
57
40
66
84
28
60
15
65
51
28
60
-8
48
37
60
>50
88
>70
>90
>95
>99
83
NSD
< 0
NSD
NSD
NSD
NSD
< 0
< 0
P-III-C-5
-------
Relative Cost of Urban BMPs
Figure 1 displays the computed total costs associated with seven urban BMP
options for several land use and site area combinations. In general,
conventional dry ponds represent the least cost BMP investment. Extended
detention ponds involved an additional cost of 5-10% at most sites, whereas
wet ponds and infiltration basins were 25 to 45% more expensive, depending
on land use and area serviced. Infiltration trenches cost approximately 50%
more than dry ponds on small residential sites, but were clearly
uneconomical on larger or more intensively developed sites. Porous pavement
was generally competitive with pond BMPs when space requirements were not
limited. When available space was limited, however, porous pavement was
only economical for small developments sites.
*» 2.5 .
TOTAL BMP COST , ANNUAL PAYMENT (in thousands of dollar
-» M CO ^. Ol O> —
oooooo cn_«io
1 L _L.. 1 i ,
HI! DRV POND [ • \ IN BASIN I • • 1 POROU842
K'i'»/l EX.DET P E=i (-TRENCH
K\\| WET POND | | POROUS* 1
ONE ACRE SITES
cr
rr ' ::.v • r''
5 n •'•' ^
1 n lltV::
~- 11 ^ ''''L. >^
3
3
|
i
In
25 ACRE SITES
P^FIn
1 n
In 1 Ifl^FI
!ja
i
1 ;
' *1
1
\
_^J
1
|
in
—
i
Hi
i
1
~
=5
3
3
a
i
In
^y
13
3
3
3
=3
^
a
—4
3 :
IT
3
-j
'
Single Family
Residential
(2 DU/AC)
Townhouse
Development
(8 DU/AC)
Commercial/
Light Industrial
(85% Impervious)
Figure 1. COMPARATIVE TOTAL COST OF SEVEN DIFFERENT URBAN BMP OPTIONS IN THE
WASHINGTON METROPOLITAN AREA, 1984 DOLLARS. Top Panel: Cost Estim-
ates for One Acre Sites. Bottom Panel: Cost Estimates for 25 Acre
Sites. Total Cost Defined as Construction Cost + Contingenency Cost
+ Annual O&M Amortized at 8% for 20 years + Financing Cost.
P-III-C-6
-------
The higher cost of infiltration BMPs can be attributed to two factors.
First, unlike pond BMPs which only control the increase in peak discharge,
infiltration BMPs must have sufficient volume to accommodate the entire
increase in upland runoff volume. As a result, infiltration BMPs generally
have greater excavation requirements because of the larger storage volume
needed. In the case of infiltration trenches and porous pavement, even
greater storage volumes are required to allow for non-void space in the
gravel or coarse stone filter. The cost of filter materials can also be a
major factor which drives up infiltration trench and porous pavement
construction costs.
One advantage of infiltration BMPs is their lower cost for annual
operations and maintenance, relative to pond BMPs. Because of their
inherent design, many common pond O&M tasks, such as structural repair, and
sediment and debris removal, are either unnecessary or impossible to
perform on infiltration BMPs. Porous pavement, for example, was found to
be cheaper to maintain than any other BMP on commercial sites. Similarly,
infiltration trench maintenance typically was less than one-third as
expensive as annual O&M for dry ponds on residential sites. As might be
expected, dry ponds typically involved the lowest O&M cost for pond BMPs,
ranging from $6 to $60/DU/yr on residential sites, and $62 to $232/Acre/yr
on commercial sites. Extended detention pond O&M involved a modest
additional O&M cost of 5 to 7%, whereas wet ponds and infiltration basins
involved 16 to 70% more O&M cost when compared to dry ponds.
Physical Suitability of BMPs
The most frequently cited physical constraints to urban BMP application
which emerged in the suitability surveys included topography, soil
permeability, space requirements and site size. In general, infiltration
BMPs were found to be subject to the greatest number of site restrictions.
Current State of Maryland Infiltration Standards prohibit their use on
sites with soil profiles of low to moderate permeability ("C" and "D"
soils), slopes greater than 5%, and a depth to bedrock or groundwater of
less than 2 feet. Also, infiltration trench and porous pavement BMPs may
not be feasible on larger, more intensively developed sites, because of
their significant space requirements (which may often exceed 25% of the
total site area). However, the greater space requirements of porous
pavement may not be restrictive on some sites, particularly if significant
portions of the site are devoted to parking.
Application of extended detention ponds in the region is normally limited
to sites with a minimum size of about 20 acres. In smaller sites, required
orifice size in extended detention ponds becomes so small (<2 inches) that
the outlet is susceptible to chronic clogging. Wet ponds can usually only
be applied to development sites in excess of 10 acres, as a minimum
watershed size is needed to maintain a permanent pool of water during
extended periods of dry weather. Although pond drawdown does not adversely
affect a wet pond's runoff control or pollutant removal properties, it may
reduce or eliminate desired habitat, aesthetic or landscape benefits.
P-III-C-7
-------
Environmental Considerations in BMP Selection
Natural Environment
As urban BMPs nearly always represent a modification to the urban
landscape, it is often necessary to select or design BMPs to minimize their
impacts on the environment. Recent research has shown, for example, that
BMPs can have a substantial influence on the quality of wildlife habitat.
A field survey conducted in Columbia, Md. (Adams et al, 1983) suggested
that dry ponds had little habitat value. The same study also concluded that
wet ponds enhanced both aquatic and terrestrial habitat value. A variety
of waterfowl, marsh and shore birds, and terrestrial wildlife utilized wet
ponds, presumably attracted by the food and cover supplied by aquatic
vegetation. With proper design, it is likely that extended detention ponds
can also function to create valuable wetland habitat. Infiltration trench
and porous pavement probably have a negative influence on local habitat,
since they consume more area than pond BMPs, and are normally unvegetated.
Downstream habitat is also influenced by urban BMPs. For example,
infiltration BMPs often can sustain downstream flows during extremely dry
periods because of their superior groundwater recharge, thus reducing
seasonal stress on downstream aquatic life. Infiltration BMPs and extended
detention ponds also function to reduce the loss of riparian habitat caused
by downstream bank erosion (since these BMPs attenuate peak flows to a much
greater degree than either wet ponds or dry ponds). Some concerns have been
expressed that under some conditions wet ponds may produce thermal
discharge which could impact trout fisheries in cold-water streams.
Human Environment
An often overlooked aspect of urban BMPs is their role in the human
environment. Urban BMPs can function as a landscape amenity or as an
unsightly and discordant community feature. Similarly, urban BMPs can
support limited recreation use or be under-utilized and possibly dangerous
empty space, depending on the type of BMP employed, and its subsequent O&M.
As resident's perception of aesthetic value prominently figures in their
acceptance of and support for BMPs, these considerations can be important,
particularly if the same residents will be responsible for future
maintenance.
In this respect, wet ponds appear to be superior to other BMPs, if
properly designed and maintained. In a recent survey of resident attitudes
towards stormwater ponds in Columbia, Md. (Adams et al,1983), 75% of
residents surveyed preferred wet ponds over dry ponds. Most respondents
felt that wet ponds provided a more pleasant and natural environment, as
well more diverse recreational opportunities.
CONCLUSIONS
Comparative data on the cost, efficiency and physical/environmental
feasibility of six urban BMPs has been integrated into a summary screening
matrix shown in Table 3. The matrix is intended for use as a preliminary
screening procedure for rapid evaluation of available urban BMP options at
the site planning level.
P-III-C-8
-------
Table 3. Generalized Screening Criteria Model For
Selecting Urban Best Management Practices
SCREENING
CRITERIA
SITE SIZE
. 1-5 ACRES
£ 10 ACRES
» >25 ACRES
O SOILS
* "A" SOILS
C "B" SOILS
a "C" or "D" SOILS
g SLOPE
E 0-5%
w 6-20%
< >20%
U SPACE REQUIREMENTS
U3 < ^ C AT
Jj U < 5 %
g #< 6-15%
•?£: 16-25%
l£ 26%-75%
2 POLLUTANT REMOVAL
_ NUTRIENTS
f- ALGAL LIMITATION
«? SEDIMENT
j TRACE METALS
5 OXYGEN DEMAND
Z OTHER FACTORS
£ GROUNDWATER RECHARGE
Z WILDLIFE HABITAT
2 RECREATION
g AESTHETICS/LANDSCAPE
Z THERMAL DISCHARGE
LOW FLOW MAINTENANCE
a TOTAL COST OF BMP*
O within 10% of DP
C llto50%>DP
5 51 to 100% > DP
g over 100% > DP
53 O4M REQUIREMENTS*
Z less than DP
8 0 to 10% > DP
W llto50%>DP
5 ^lo/n^np
A ji tv^jjr
9 OTHER FACTORS
§ ECONOMIES OF SCALE
LIABILITY/SAFETY
RESIDENTIAL BMPs
(SINGLE FAMILY RESIDENTIAL -•• TOWNHOUSE)
DP XD WP IB IT
0 • • O O
o • o o o
0 O 0 O ®
0 0 0 O O
O O O 0 0
o o o • •
§ § 8 8 •
Or~\ n\ ^ ^
(J (£> w w
- - o o o
— — — — —
9 0 O O O
® o 8 8 8
• o o o o
• 0 0 O O
• ® 0 O O
* 0 O ® •
0 • O 0 •
® 0 O 0 0
o © • o o
• ® o o o
o • • •
-00*
A
~~~~ ~~ ~~~ ^0
• • • ©
o • • o
- 0 0 -
o o o © •
© © © 0 O
iirffNn n APPpriPBiATP f?\ OFTEN APPROPRIATE
LEGEND U APPROPRIATE (/_) W/DESION CHANGES
•NOTE: COMPARATIVE DP =• DRY POND (DETENTION BASIN)
BMP COST DATA HAS XD = EXTENDED DETENTION POND
BEEN COMPUTED WP WET POND (RETENTION BASIN)
USING DRY POND (DP) 1 B INFILTRATION BASIN
COSTS AS AN INDEX 1 T INFILTRATION TRENCH
BASE PP POROUS PAVEMENT
GS GRASSED ROADSIDE SWALES
PP GS
© 8
O 0
o o
0 0
0 9
: i
• 0
0 0
o -
o •
o •
o ®
o ®
o •
O 0
• 0
• ®
® 0
o o
o ®
• o
0 —
0 -
© o
o -
8 8
COMMERCIAL BMPs
DP XD WP IB IT PP
O • •000
O • O O ® 0
O O O 0 • 0
0 © © o o o
O O 0000
O O O 9 9 9
O O O O O O
O O O Q • •
O O O • • •
O O O 0 • •
_ _ — © © •
- - - O O O
• • 0000
• • O 0 0 0
• O O 0 O O
• O O 0 0 0
• © © 0 O O
• ® 0 O O O
0 • O 0 • •
® 0 00®®
O 0 9 0 U 0
• ® O 0 O O
- o • • • ®
- O O O • 0
— — 00
— — — — O O
- • • • o ®
- O • • - 0
- - O O - 0
§O 0 0 • •
© 0 0 O O
® ™PR.ATE • NOTAPPROPR.ATE
"A" SOILS = SANDS, LOAMY SANDS
"B" SOILS - SANDY LOAM, LOAMS
"C" SOILS - SILT LOAM, SANDY CLAY LOAMS
'.'D" SOILS - CLAY LOAMS TO CLAY
P-III-C-9
-------
Use of the Screening Matrix
As an example, suppose a site planner is reviewing a proposal for a ten
acre shopping mall development situated on a site of moderate (>5%) slopes
and relatively permeable "B" soils. Apart from controlling the peak
discharge from the two year design storm, the planner is required by local
regulations to install a BMP capable of effectively removing sediment and
nutrients, and having, at a minimum, no adverse impact on the local
environment. At the same time, the site planner wishes to minimize the
costs and space requirements facing the builder.
Consulting the commercial development column in the screening matrix, the
site planner can quickly eliminate the use of extended detention due to
site constraints. Likewise, the dry pond and infiltration trench options
can be rejected because of poor pollutant removal and prohibitive costs,
respectively. If space is limited at the site, porous pavement can also be
ruled out on the basis of large costs and space requirements. However, if
the site will have a large area devoted to parking, porous pavement may be
considered. Wet pond and infiltration basins are also indicated to be
feasible BMP options. Both have similar cost and O&M requirements.
However, the matrix indicates that wet ponds would have a smaller space
requirement and greater overall habitat and amenity value. As this example
illustrates, the screening matrix enables a site planner to select the best
overall BMP option given diverse and often competing objectives.
Implications for Developing Effective Urban BMP Strategies
The matrix also offers several useful insights for developing low-cost,
effective urban nonpoint source control strategies. For example, it is
clearly evident that there are no single, "all-purpose" BMPs that can be
applied over the entire range of site conditions. Rather, each urban BMP is
only feasible and cost-effective in a rather narrow range of development
situations. As a consequence, it would appear that stormwater management
programs should be both flexible and site-specific, as broad prescriptive
BMP policies will rarely be sufficient to cover the entire spectrum of
development activities encountered in urbanizing areas.
For example, if high overall pollutant removal is a desired objective in a
stormwater management program, wet ponds would appear to be the BMP of
choice on most development sites greater than 10 acres in size. If nutrient
control is not a major concern, then extended detention ponds may also be
attractive on larger (20+ acre) sites. Infiltration BMPs generally provide
the greatest overall mix of pollutant removal, cost and feasibility on
smaller sites (1-10 acres). In particular, porous pavement appears most
appropriate in small commercial development sites, whereas infiltration
basins or trenches may be more suitable on small residential sites. When
infiltration BMPs are not possible because of site constraints (a frequent
occurrence in the Washington metropolitan area), there appear to be no
other BMP options available which possess a significant nonpoint source
control capability.
The screening matrix reveals that there are several other combinations of
site conditions, environmental or cost restrictions for which no effective
BMP can be applied for nonpoint source control. One implication is that
P-III-C-10
-------
the existence of these situations could substantially curtail success in
achieving basin-wide urban nonpoint source loading reductions. However, it
is possible to achieve some measure of control at these sites by other
methods. For example, some local jurisdictions have adopted off-site
regional stormwater management programs which make it possible to pool
urban runoff control needs from several small sites in a single, larger wet
pond or extended detent-ion pond further downstream. Developers who cannot
build an effective BMP on-site are required contribute to the financing of
the off-site facility.
A second approach might be to separate stormwater runoff control from water
quality protection needs. Thus, low-cost detention BMPs could be used to
control runoff from the two year design storm, while infiltration BMPs,
designed to accommodate much smaller runoff volumes, could be applied to
achieve pollutant removal for smaller and more frequent runoff events.
Because of their reduced size, these infiltration BMPs would presumably be
subject to fewer cost and site restrictions.
Future Directions
While the matrix can only evaluate a combination of nine general
development scenarios at present, future research is being directed to
develop an interactive software package capable of analyzing BMP options of
any individual development site, given simple input parameters readily
available from site plans. Concurrently, research is being conducted to
better define the more uncertain elements of the matrix, particularly in
regard to infiltration costs. By making better BMP screening tools
available to site planners and developers, it is anticipated that wider
implementation of effective urban BMPs can be achieved, and consequently, a
greater degree of urban nonpoint source control can be attained.
REFERENCES
Adams, L.W., L.E. Dove, D.L. Leedy and T.M. Franklin. (1983) Urban Wetlands
for Stormwater Control and Wildlife Enhancement. Urban Wildlife
Research Institute. 64 pp.
Maryland Department of Natural Resources.(1984) Standards and Specifications
for Infiltration Practices. Stormwater Management Division, Maryland
Water Resources Administration. 180 pp.
Metropolitan Washington Council of Governments. (1983a) An Evaluation of the
Costs of Stormwater Pond Construction and Maintenance. Contract Report
for the U.S. EPA Nationwide Urban Runoff Program. 98 pp.
___. (1983b) Urban Runoff in the
Washington Metropolitan Area: Final Report Washington DC Area Urban
Runoff Project. U.S. EPA Nationwide Urban Runoff Program
. (1984) Potomac River Water
Quality 1983: Conditions and Trends in Metropolitan Washington. Water
Resources Planning Board. 94 pp.
U.S. Soil Conservation Service. (1975) Urban Hydrology for Small Watersheds,
Technical Release No. 55. U.S. Dept. of Agriculture. 75 pp.
P-III-C-11
-------
IMPLEMENTATION OF URBAN NONPOINT SOURCE
WATER POLLUTION ABATEMENT PROGRAMS
Robert P. Biebel
Chief Environmental Engineer
Southeastern Wisconsin Regional Planning Commission
Waukesha, Wisconsin
and
David B. Kendziorski
Principal Planner
Southeastern Wisconsin Regional Planning Commission
Wa u ke s ha, Wi sc ons in
Abstract
Urban runoff management strategies have shifted from exclusive flooding and
drainage concerns to an increased emphasis on water quality and comprehensive
stormwater management. Recent studies, especially those conducted under the
Nationwide Urban Runoff Program (NURP), have improved the technical basis for
implementing urban nonpoint source pollution abatement programs. However,
there are still questions regarding the practicality of providing adequate
urban runoff pollutant control in existing urbanized areas. Implementation of
an urban nonpoint source control program should ideally include the prepara-
tion of a stormwater management plan.
P-III-D-1
-------
INTRODUCTION
Historically, most urban runoff concerns were related to flooding and drainage.
The concept of the full-flowing storm drain designed by the Rational Method
has been the mainstay of urban storawater drainage practice in this country
since the late 1800s. Its objective of the rapid removal of surface water did
not fully consider water quality effects, as well as downstream water quantity
effects, through the construction of culverts, storm sewers, and open channels
designed to accommodate the increased flow rates from urban areas. The effect
of this approach has been to increase the velocity of runoff, reduce the
natural storage of water in the watershed, and transmit the Increased runoff
more quickly downstream. It is well known that the impervious surfaces of an
urban area contribute to these increased flow rates: within a typical resi-
dential area, streets and parking areas may contribute from 45 to 85 percent
of the runoff volume, with roof surfaces contributing an additional 10 percent.
Over the past 20 years, however, water quality has become an increasing con-
cern. In the 1960s, studies by the U.S. Public Health Service and the Federal
Water Quality Administration Indicated that urban runoff caused water quality
problems in receiving surface waters and posed a human health hazard. In the
1970s, water quality management, or 208, plans were prepared by regional or
county planning agencies in most major urban centers. Many of the 208 studies
concluded that urban stormwater runoff was a water quality problem, but the
208 studies lacked sufficient data on:
o Sources of pollutants
o Washoff and transport mechanisms
o Receiving water Impacts
o Effectiveness and costs of urban nonpoint source control measures
In the 1980s, the EPA funded the Nationwide Urban Runoff Program (NURP)
studies, and state and local nonpoint source planning efforts—such as a
Milwaukee Harbor Estuary Study and the Milwaukee River Priority Watersheds
Planning Program in Southeastern Wisconsin—also began to address control of
urban nonpoint source pollution.
In order to understand how these recent studies have helped in the implementa-
tion of urban nonpoint source control measures, it is necessary to briefly
review the technical findings of these studies.
CHARACTERIZATION OF URBAN STORMWATER RUNOFF
Compared to rural streams, streams draining urban areas frequently have similar
concentrations of solids and nutrients, and higher concentrations of metals
and bacteria. No only, as stated above, are runoff volumes higher from highly
impervious land uses, such as commercial and industrial land areas, but pollu-
tant concentrations are also higher from these land uses. Metals in urban
runoff which often violate acute (short-term) and chronic (long-term) criteria
include copper, zinc, lead, chromium, and cadmium. Within the Menomonee River
in southeastern Wisconsin, studies have indicated that metal concentrations in
the water column, as well as in the bottom sediments, Increase substantially
as the River flows into the Milwaukee metropolitan area.
P-III-D-2
-------
URBAN NONPOINT SOURCES
Street surfaces are the predominant source of pollutants In urban runoff. In
typical urban areas, street surfaces may account for 70 to 80 percent of the
total solids load during average storm events, While pervious surfaces gener-
ally account for less than 20 percent. Pervious surface loadings become more
significant during larger storm events. Traffic emissions and atmospheric
loadings may each account for up to 10 percent of the total loading.
The NURP studies indicated that pollutant loadings on streets generally
increase as the tmperviousness of the land area Increases. With respect to
pollutant loading, the impervious surfaces which are directly connected to a
storm sewer system are considered to be most important. The composition of
the street loadings also varies for different land uses: transportation and
industrial land uses generate higher concentrations of metals, whereas resi-
dential areas have the highest concentrations of fecal coliform.
URBAN NONPOINT SOURCE CONTROL MEASURES
Street Sweeping
In general, the NURP studies indicated that street sweeping is not an effective
measure for achieving high levels of nonpoint source control. Street sweeping
was not very effective because:
o Pollutant loading removals from the street surface by the sweeper
were often less than 50 percent.
o Sweepers effectively remove the larger particles but do not generally
remove the smaller particles—which contain higher pollutant concen-
trations.
o Street loads accumulate rapidly following sweeping or a storm event.
o There is too much variability related to wind, traffic, street, and
rainfall conditions.
Nevertheless, street sweeping programs can be optimized, providing a modest—
up to 20 percent—level of nonpoint source control, particularly in commercial
and high density residential areas, by:
o Increased sweeping in spring and fall
o Increased sweeping of highly impervious areas
o Improved parking regulations
Stormwater Storage
The NURP studies Indicated that atormwater storage could be an effective way
to achieve relatively high pollutant removals. The studies indicated that:
o Properly designed wet basins can remove more than 50 percent of the
incoming particulate loadings, whereas, dry basins generally do not
remove significant amounts of pollutants.
P-III-D-3
-------
o Some dry basins may be modified to increase their detention time and
pollutant removal effectiveness.
o Off-site basins, draining large drainage areas of perhaps 100 to 700
acres (0.4 to 2.8 km ), are more cost-effective than onaite basins
draining 5 to 20 acres (0.02 to 0.08 km ).
IMPLEMENTATION
The NURP and other related studies have provided the information on the tech-
nical basis for the development of implementation programs for urban nonpoint
source control. It may be concluded from these studies that there is not a
wide choice of urban control measures that can achieve relatively high levels
of reduction in pollutant loadings. Generally, storrawater storage in properly
designed wet basins appears to be the only practical method available which
can achieve removal levels exceeding 30 percent for several pollutants. How-
ever, there are potential limitations and problems associated with the use of
stormwater storage in many urban areas. These limitations include:
1. Maintenance problems related to:
a. Lack of adequate access
b. Clogging of outlets
c. Litter accumulation
d. Weed control
2. Safety problems related to:
a. Stagnant pools in urban areas which attract mosquitoes, rodents,
and other pests;
b. High water velocities during storm events;
c. Attraction to children.
3. Inadequate flood control, if hydraulic design and analysis is not
considered. There frequently is a large difference between the
volume of storage and outlet capacity needed to provide adequate
flood control and that needed to provide optimal water quality pro-
tection.
4. Aesthetic problems, since these basins accumulate large amounts of
pollutants, are frequently sediment-laden, and have high concentra-
tions of nutrients which can support weed and algae growths.
5. Lack of suitable available open land sites. This is probably the
most significant limitation for the use of stormwater storage basins,
particularly in developed urban areas. Alternative storage facili-
ties, such as subsurface storage and stormwater treatment systems,
which may be more suitable in developed urban areas, are also very
expensive.
P-III-D-4
-------
While many of these limitations to the use of stormwater storage facilities
can be mitigated by proper planning and design, and by the development of
effective management and maintenance programs, these factors, nevertheless, do
limit the wide application of stormwater storage as a water quality management
measure.
Other urban nonpoint source control measure which can achieve limited effec-
tiveness include:
1. Street sweeping, which as noted above can often be designed to remove
an additional 10 to 20 percent of the loadings of various pollutants
in some areas.
2. Construction site erosion control, which can effectively prevent high
loadings of sediment.
3. Public education programs, which could particularly be helpful in
voluntary nonpoint source control programs.
4. Pet waste control ordinances, which can help reduce high bacterial
loadings especially in residential areas.
Finally, urban nonpoint source control programs, especially those that require
extensive design and investment, are best implemented under the guidance of a
comprehensive stormwater management plan. A stormwater management plan should
address both water quantity and quality concerns and contain the following
elements:
1. Hydrologic/hydraulic analysis of urban runoff quantity and quality
under existing and future conditions.
2. Water quantity and quality objectives and design criteria.
3. Development and evaluation of the costs, hydraulic effects, and water
quality impacts of alternative stormwater conveyance and storage
measures.
4. Selection of a recommended plan which is economically feasible, which
provides for the wise use of water resources, and which meets, where
practical, the objectives of the plan.
P-III-D-5
-------
NFS POLLUTION CONTROL:
A CONSULTANT'S PERSPECTIVE3
by
Stuart G. Walesh.1'
INTRODUCTION
Since the early 1970's, much of the non-point source (NPS) pollution manage-
ment activity in the U. S. has been driven primarily by government regulatory
and funding programs. Examples of these programs, which have been at all
levels of government from federal to local, are the federal 208 Program and
local (county, city, village, and special, district) erosion and sedimentation
control ordinances. Government programs have apparently been based on the
conviction that NPS problems exist or will develop.
In other words, NPS management activities to dale have been primarily in a
"top-down" mode, that is, government directing individuals and entities to
mitigate or prevent NPS pollution problems. The historic "top-down" approach
to NPR management is characterized by emphasis on research and development,
large-scale studies, broad recommendations, and, in general, absence of
specificity and little implementation.
The historic "top-down" approach contrasts with a "grass-roots" approach.
With the latter, individuals and entities undertake NPS pollution management
because they have NPS or NPS-related problems that must be solved now. Some
consulting engineers have had opportunities to apply NPS pollution management
measures on a problem resolution basis completely outside of, but consistent
with the intent of, government programs.
PURPOSE
Using a case study approach, this paper briefly describes several NPS pollu-
tion control projects motivated largely by local need, not by government
regulatory and funding programs. The case studies are intended to stimulate
thinking concerning the value of encouraging and supporting a "grass roots"
approach to NPS pollution management. In the absence of, and perhaps unfet-
tered by, government controls, innovative, cost-effective NPS pollution
control measures have been designed and are being implemented. Experience
gained on these and similar "grass roots" projects is used to draw conclu-
sions and make certain recommendations concerning NPS pollution control.
a
Presented at the Non-Point Pollution Symposium, Session on Efficiency and
Feasibility of Best Management Practices for NPS Control, Milwaukee, Wis-
consin, April 23-25, 1985.
Head-Water Resources Engineering, Donohue & Associates, Inc., Milwaukee
Division, 600 Larry Court, Waukesha, WI 53186 (414/784-9200).
P-III-E-1
-------
CASE STUDY 1: SEDIMENTATION BASINS IN SERIES WITH A DETENTION FACILITY
Physical Setting
The setting for this project, as shown in Figure 1, is a 640 acre almost
completely urbanized watershed in the Madison, Wisconsin area. The watershed
contains parts of three government entities—the City of Madison, the Town
of Madison, and the Town of Fitchburg—plus the University of Wisconsin
arboretum. The arboretum is a sanctuary for native plants and wildlife and
is used for research and teaching.
Problems
Two growing problems were of concern to area residents several years ago
at the beginning of the engineering phase of this project. First, flooding
of increased frequency and severity was occurring in the peninsula of single-
family residential development which protrudes into the arboretum from the
east, as shown in Figure 1, Watershed runoff from the entire watershed
passing through and over this area caused surface and basement flooding.
Second, debris and sediment carried from the urbanizing watershed were being
deposited in and causing damage to meadows, wetlands, and lagoons in the
arboretum.
Solution
Working together, the three government entities and the Arboretum Committee
agreed to a surface water management system consisting of two sedimentation
basins and a sedimentation facility as shown in Figure 2. The sedimentation
basins are located on the upstream side of the arboretum and trap objection-
able sediment and attached pollutants, floatables, and other debris before
it enters the arboretum. The detention facility, which is normally empty,
temporarily impounds stormwater in the arboretum and thus provides flood
protection to the downstream residential area.
The 2,200 foot long berm required for the detention facility was designed
with a curvilinear alignment and irregular cross-section to minimize its
visual impact on the aesthetically sensitive arboretum. The sedimentation
basin-detention system was constructed in 1982 at a total of cost of
$103,000 and is in operation (Donohue, 1981; Raasch, 1982).
CASE STUDY 2: SEDIMENTATION BASIN IN SERIES WITH RESTORED WETLAND
Physical Setting
McCarron Lake in Roseville, Minnesota receives runoff from a fully urbanized
1.6 square mile watershed. The 72 acre lake is heavily used for fishing,
swimming, and boating and provides a setting for a county park.
Problems
A one-year diagnostic study confirmed that McCarron Lake is eutrophic.
Hypolimnetic oxygen depletion releases nutrients into the water column when
the lake is stratified and there are abundant nutrients in the water column
throughout the year. Surface water runoff into the lake is the principal
contributor of the troublesome nutrients and of sediment.
P-III-E-2
-------
FinURK 1
WATKKSHKD MAP
'I »>^Av """» •"""«•
thid I I V7/'\ ^S< \ Ml Milt MltAl
^N^^^v'' *H"
Source: Raasch, 1982
P-III-E-3
-------
FIGURE 2
SEDIMENTATION BASINS-INTENTION FACILITY SYSTEM
.... J I ._.. J L_J t ..... -. J L ________
J L ...... ____ I UH, ,H J I _ ...... _
Wtbl
SfcOIMtNTAIIOH - -
PONO
N
Source: Raasch, 1982
P-III-E-4
-------
The solution to the nutrient and sediment problem has two components. The
first, as shown in Figure 3, is a sedimentation basin in series with a
restored wetland on a drainageway that controls 80 percent of the runoff
into the lake. The 1.5 acre sedimentation basin bounded on three sides by
8.5 acres of available, but normally dry, detention storage area is located
at the upstream end of the system. The sedimentation basin and related
detention area will trap sediment and control flow into the restored wet-
land. The wetland would be restored to a size of about 8.0 acres by building
a series of low berms across the existing channel, as shown in Figures 3
and 4. Vegetation such as cattails, sedges, and willows will grow in the
restored wetland. The sedimentation basin-restored wetland is expected to
remove about 75 percent of incoming total phosphorus and suspended solids
and have a total cost, including engineering and administration, of $250,000.
The second component of the NFS pollution control system is two large stomps
to control large street runoff from the remaining 20 percent of the water-
shed. The circular sumps, one of which is shown in Figure 5, would be
large--both are 10 feet deep and one is 18 feet in diameter and the other
24 feet in diameter. As suggested by Figure 6, the sumps are simple struc-
tures designed to remove the suspended solids and buoyant materials. The
sumps, which are expected to achieve 65 percent reduction in suspended solids
and a significant reduction in adsorbed or absorbed phosphorus, have a total
cost of $75,000.
The sedimentation basin-restored wetland-sump system is under construction
and will be completed by late summer 1985 (Donohue, 1983).
CASE STUDY 3: JiUTRFACK AND SIJBSIIRFACE DETENTION IN AN URBAN AREA
Physical Sett i
The 1,200 acre Skokie, Illinois Howard Street Sewer District (HSSD) , as shown
in Figure 7, is a combined sewer service area. The HSSD is completely and
densely developed and over 80 percent of the district is single-family resi-
dential with the rest being commercial and industrial. The long, narrow
HSSD is very flat having an overall slope of only 0.2 percent.
The combined sewer system serving the HSSD discharges at the east end of the
district to an interceptor paralleling the North Shore Channel. The start-up
of the Chicago area Tunnel And Reservoir Plan (TARP) in 1985 will mitigate
wet weather overflow to the channel.
Problem
Athough solution of the combined sewer overflow problem is imminent, serious
basement flooding will continue to occur in many of the over 5,000 single-
family residences and other buildingy ia the HSSD. TARP will improve the
outlet conditions for runoff events from the district, but it will not
resolve deficient sewer capacities throughout most of the HSSD. Detiign,
construction, arid operation of a conveyance or detention system to take full
advantage of TARP is ueceusary to solve the flooding problem and is the
complete responsibility of the: Village of Skokie. Although a relief sewer
system, such as new separate storm sewers, is technically feasible, the cost
is prohibitively high.
P-III-E-5
-------
FIGURE 3
SEDIMENTATION BASIN IN SERIES WITH RESTORED WKTl.AWu
10 ACRE SEDIMENTATION BASIN
INCLUDING I5ACRE PERMANENT POND
' / x
y f, \
•»—- / i
NORMALLY 'V
DRY AREA / N
PERMANENT
POND
SEDGE,WILLOWS ft CATTAILS
LOW LEVEL BERMS
SEDGE a WILLOWS
OUTLET CONTROL
MIXED VEGETATION
SEDGE a WILLOWS
STREET DITCH
MC CARRON LAKE
LEGEND
LOW LEVEL BERM
7 SURFACE STREAM MONITORED (ID NUMBER)
. - •* DIRECTION OF WATER FLOW THROUGH WETLAND
Source: Donohue, 1983
P-III-E-6
-------
FIGURE A
LONGITUDINAL SECTION THROUGH SEDIMENTATION BASIN AND RESTORED WETLAND
SEDIMENTATION
BASIN BERM
SEDIMENTATION
BASIN ^---''
;,
FILL
TYPICAL SECTION OF LOW BERM
NOTE LOW BERMS TO BE CONSTRUCTED AT ELEVATION
AND SPACING TO PROVIDE A WETLAND SLOPE
OF LESS THAN 0.002 FT/FT
LOW BERM OVERFLOW SECTIONS PROTECTED
FROM EROSION BY PERMANENT PLASTIC
REINFORCING MESH IN TOPSOIL.
NO SCALE
WETLAND
EXIT BERM
Source: Donohue, 1983
-------
FIOURE 5
SUMP LOCATION
MC CARRON LAKE
EXISTING TEMPORARY CURBING
PROPOSED PERMANENT
EXISTING ASPHALT SPILLWAY
PROPOSED STORMWATER SUMP
FLOW PATTERN OF
STREET RUNOFF
Source: Donohue, 1983
P-III-E-8
-------
FIOURF. 6
SECTION THROUGH SUMP
FLOW OF STORMWATER RUNOFF
PROPOSED STORMWATER SUMP
MC CARRON BLVD. S.
MC CARRON
LAKE
Source: Donohue, 1983
P-III-E-9
-------
Solution
After extensive field and office investigations spanning several years,
the Village of Skokie has begun implementation of a runoff control system
(RCS) in which stormwater detention will be retrofitted into the existing
stormwater drainage and combined sewer system (Donohue, 1982, 1984a, 1984b,
1984c; Walesh and Schoeffmann, 1984). The concept is shown in schematic form
in Figure 8 for existing and proposed conditions.
Stormwater will be intentionally ponded in streets in a controlled fashion
and then gradually released to the combined sewer system at a rate that can
be conveyed by the combined sewer system without surcharging. On those
streets where insufficient surface storage is available, supplemental subsur-
face storage tanks will be used. More specifically, the recommended RCS
consists of:
0 431 flow regulators functioning in conjunction with 210 street
berms.
0 Eight subsurface storage tanks.
0 9,500 feet of 30-irich to 72-inch separate relief sewer for the
commercial, downstream end of the HSSD.
The estimated 1985 cost for the recommended RCS, including engineering, legal,
and administration fees, is about $8.5 million or less than 20 percent of the
cost of conventional relief sewer system.
Most of the temporary street detention will be accomplished by installation
of simple flow regulating devices in existing catch basins as shown in
Figure 9. Each catch basin contains a sump which traps both settleable and
buoyant material. In addition, each subsurface tank will, as illustrated
in Figure 10, contain a large sump to trap soluble and buoyant material.
Therefore, essentially all stormwater will, while in route to the combined
sewer, pass through and be detained in a sump where settleable and buoyant
material, along with adsorbed and absorbed potential pollutants, will be
removed. These sumps will be regularly cleaned and maintained by Village
personnel.
Design of flow regulators, street berms, and tanks is almost complete. Flow
regulators installation is underway and will be completed this year. Tank
and berm construction will begin soon and be completed in 1985. Relief sewers
will be designed in 1985 arid are scheduled for construction in 1986.
Although control of suspended solids and other potential NFS pollutant is
not a primary objective of the Skokie RCS, this system or a variation on it
has the potential to control NFS pollutants near their source because of the
widespread detention occurring and the presence of numerous sumps and traps.
An engineered RCS built into an new separately sewered development or retro-
fitted into an existing separately sewered or combined sewered area could
have the following water quality benefits:
P-III-E-10
-------
FIOURE 7
STUDY ARKA: HOWARD STREET SEWER DISTRICT IN SKOKIK, ILLINOIS
STUDY
HOWARD ST
r> Mi
Source: Waleeh and Schoeffmann 198A
P-III-E-11
-------
FIGURE 8
STREET PONDING, REGULATED CATCH BASIN AND UNDERGROUND TANK
i
-TI
DOWNSPOUT
CONNECTED
TO SEWER
DISCONNECTED
DOWNSPOUT
TEMPORARY K>NDING
TANK OPTION
HOUSE
VtLLACE COMBINED SEWER
USED WHERE STREET
PONDING CAPACITY IS
INADEQUATE
EXISTING CONDITIONS
RUNOFF CONTROL SYSTEM
Source: Walesh and Schoeffmann, 198-
-------
FIGURE 9
FLOW REGULATOR AND SUMP IN CATCH BASIN
- CATCH BASIN
OUTLtT PIPE
-ORIFICI RESTRICTOR
Source: Wnleah and Schoeffraann, 1984
P-III-E-13
-------
h: 10
UNDERGROUND TANK
V /
-- f
A
/
^
tXOAVAMON
1. 1 Ml IS
COMhlM.lj---
SE Wfft
K t UW
HEGULATOR
Ht I NTOHCI-C>
CONCHL 1C tlOX
Source: Walesh and Schoeffmann, 1984
P-III-E-14
-------
0 Reduced frequency and volume of combined sewer overflows to the
surface water system.
0 Control of NFS pollution near its source for subsequent removal
as a part of sump cleaning and maintenance.
0 Reduced peak flows at a wastewater treatment plant.
OBSERVATIONS
The following observations are made based on experience, such as the examples
presented in this paper, with engineering of NFS controls at. the "grass
roots" level in response to particular environmental problems:
1. Suspended Sediment is Primary Target: In addition to being a
potential pollutant, suspended sediment can be a tranporter of
potential pollutants such as phosphorus, pesticides, heavy metals,
bacteria, and oxygen-demanding materials. Therefore, the success-
ful control of suspended solids should indirectly result in sub-
stantial control of other non-point source pollutants.
2. Small Incremental Expense Concept: In terms of land acquisition
and construction costs, relatively little additional expenditures
are likely to be required to add effective water quality control
components to planned stormwater detention/retention facilities.
That is, control of the quality of stormwater can often be achieved
for a small incremental cost above that needed to control the quan-
tity of stormwater.
3. Need for Design-Oriented Manual: Research results, performance
data, and design tools and techniques should be integrated into
a design-oriented manual. Such a manual could present tools and
techniques to be used by engineers in the planning, design, and
operation of temporary and permanent NFS pollution control facili-
ties such as sedimentation basins, detention/retention facilities,
enhanced wetlands and sumps.
^- Stringent Inspection and Maintenance Requirement^: The principal
purpose of NFS control measures is removal and concentration of
potential pollutants from stormwater runoff. At minimum, sedi-
ment and other debris will be collected in NFS control facilities.
These facilities must be regularly cleaned and otherwise maintained
if they are to function as intended. Designers should clearly
indicate the type and frequency of inspection and maintenance and
provide cost estimates.
5. Resist Mandating Specific NFS Measures: Given the limited, but
growing, knowledge concerning the design and operation of NFS
pollution control measures, local, state, and federal governmental
units should "go slow" in mandating specific means of controlling
NFS pollutants. Such agencies should focus instead on statements
of intent and performance. Premature rule-making and regulation is
P-III-E-15
-------
likely to result in "action" but little progress. More is likely
to be accomplished in advancing the state-of-the-art of achieving
significant control of NFS pollutants by (a) funding additional
research and development projects and pilot studies and (b) by
encouraging the control of NFS pollutants but not dictating the
means.
Acknowledgements
,'i.
The author gratefully acknowledges the ideas and information provided by
colleagues and drawn from project experience at Donohue & Associates, Inc.
However, the author is solely responsible for the paper's content.
P-III-E-16
-------
CITED REFERENCES
Donohue & Associates, Inc., Lake Forest Storm Water Detention Facility -
Preliminary Engineering. 1981.
Donohue & Associates, Inc., Me Carron Lake Diagnostic - Feasibility Study -
Roseville, Minnesota. November, 1983.
Donohue & Associates, Inc., Preliminary Engineering - Runoff Control Program -
Howard Street Sewer District - Skokie, Illinois, July, 1982.
Donobue & Associates, Inc., Monitoring - 1983 - Runoff Control Program -
Howard Street Sewer District - Skokie, Illinois, January, 1984a,
Donohue & Associates, Inc., Flow Regulator Pilot Study - Runoff Control Pro-
gram - Howard Street Sewer District - Skokie, Illinois, February, 1984b.
Donohue & Associates, Inc., Preliminary Engineering - Addendum - Runoff Con-
trol Program - Howard Street Sewer District - Skokie, Illinois, March, 1984c.
Raasch, G.E., "Urban Stormwater Control Project in an Ecologically Sensitive
Area," Proceedings - 1982 International Symposium On Urban Hydrology, Hydrau-
lics and Sediment Control, University of Kentucky, Lexington, Kentucky,
July 27-29, 1982, pp. 187-192.
Walesh, S.G. and M.L. Schoeffmann, Surface and Sub-Surface Detention in
Developed Urban Areas: A Case Study, presented at the American Society of
Civil Engineers Conference "Urban Water - 84," Baltimore, MD., May 28-31,
1984.
P-III-E-17
-------
SUMMARY
Panel TV
Organization and Institutional Arrangement of
Nonpoint Source Control Programs
PANEL MEMBERS
Moderator: Jim Bauman, Nonpoint Pollution Section, Wisconsin Department
of Natural Resources, Madison, WI.
Assistant Moderator: Dean Massey, Economic Research Service, U.S. Depart-
ment of Agriculture. Madison, WI.
Panelists:
Floyd Heft, Soil Conservation Society of America, Columbus, OH.
William Horwath, National Association of Conservation Districts, Stevens
Point, WI
William Katz, Professor, Marquette University, Milwaukee, WI.
Fred Madison, Associate Professor, University of Wisconsin, Madison, WI.
Glen Stout, Professor, University of Illinois, Urbana, IL.
Recorder: Gary Korb, Southeastern Wisconsin Regional Planning Commission,
Waukesha, WI.
INTRODUCTION
Sources of nonpoint pollution are numerous and diversified and their control
would presently be the responsibility of a number of federal, state and
local agencies and private landowners and developers. The present approach
to nonpoint pollution control relies on voluntary compliance of polluters
with little financial assistance available from federal and state sources.
The oversight of the nonpoint pollution control programs are carried out
by U.S. Department of Agriculture through its soil and water conservation
programs; by the U.S. Environmental Protection Agency which has very limited
resources to carry out any meaningful nonpoint pollution abatement programs,
and by states that have enacted nonpoint pollution control and abatement
laws. The questions this panel was faced with were as follows:
1. Is there an adequate organizational structure in place which would
enable to carry out nonpoint pollution abatement programs?
2. If the present organizational structure could be used how could it
be improved?
3. If the present organizational structure is not adequate what new
organizations and institutions are needed to carry out the programs?
DISCUSSION
Panelists agreed that the present organizational structure and institutions
are not adequate however the panel did not reach a conclusion whether a
new organizational structure is needed or whether present organizations and
institutions should be improved.
P-IV-1
-------
There are a number of drawbacks and gaps in the present organizational
structure which were identified by one of the panelists as follows:
1. Inadequate funding.
2. Lack of joint planning for soil erosion and water quality control among
the agencies.
3. Lack of public accountability due to large number of agencies involved.
4. Institutional goals are not clear.
5. Dilution and shallowness of educational programs due to voluntary
approaches.
6. Uneven treatment and understanding of the nonpoint pollution problem
across the state lines.
7. Targets and goals of the programs have not been clearly specified.
8. U.S.D.A. agency roles and roles of Soil and Water Conservation Districts
have not been clearly defined.
Another member of the panel pointed out the legislative limits that have
been imposed on present Soil and Water Conservation Districts. In 13 fiorth
Central states there are great differences in the regulatory powers of the
SWCD's, and subsequently, there are substantial differences among the states
as to:
1. Ease of adoption of ordinance.
2. Recommendations for soil conservation practices.
3. Administrative procedures.
4. Enforcement provisions.
Some stateshave so called permissive regulatory powers, otherstateshave
mandatory regulatory powers. States with permissive regulatory policies
may have difficulties in getting necessary regulations adopted.
The availibiltyof fundsand funding for NFS control programs was also ex-
tensively discussed. Some panelists and discussers had reservations against
subsidies since they could make the polluters wait for the subsidy and do
essentially nothing. In general, financial aid (cost sharing) and subsidy
should be given primarily to existing farms that have initiated a nonpoint
pollution problem. New farms and fanning entreprenuers should consider
the NFS control as a part of their cost.
An innovative financial arrangement was suggested by one panelist for lake
management districts whereby lake users (water quality beneficiaries) pro-
vide subsidies to upstream farmers to implement nonpoint pollution control
programs.
P-IV-2
-------
CONCLUSIONS AND RECOMMENDATIONS
1. The panel concluded that the present lack of funding and somewhat
inadequate organizational structure should not preclude the initia-
tives to limit nonpoint pollution inputs. The state-of-the-art of
the technology and existing organizational structure are satisfactory
for an initial effort to curb the nonpoint pollution in many watersheds.
2. Lack of state and, primarily, federal funding has been cited as the
major obstacle for successful nonpoint pollution abatement programs.
3. The role of existing agencies should be clearly identified to avoid
overlaping and unnecessary competition.
4. Cost sharing and subsidies should be used with caution in order for
them not to become an obstacle and a factor causing slowdown of the
efforts. Cost sharing and other subsidies should be limited to
existing farms and based on their potential to degrade water quality.
New farming operators and construction developers should incorporate
the nonpoint pollution control in their cost.
5. New organizational structures may be needed if the present existing
organizations and institutions are unable to accomplish the goals.
The future most optimal organizational structure is not known.
6. The organizations and institutions should have some regulatory back-up
if voluntary approaches fail.
P-IV-3
-------
ORGANIZATION AND INSTITUTIONAL ARRANGEMENTS
OF NFS CONTROL PROGRAMS
William J. Horvath
North Central Regional Representative
National Association of Conservation Districts
Stevens Point, Wisconsin
ABSTRACT
Under S304(4)(1)(A,B,C) of Public Law 92-500, the Federal Water Pollution
Control Act Amendments of 1972, EPA was directed, after consulting with .
appropriate state and Federal agencies and other interested persons to:
"issue to appropriate Federal agencies, the States, water
pollution control agencies, and agencies designated under
S208 information including guide lines for identify-
ing and evaluating the nature and extent of nonpoint sources
of pollutants resulting from
(A) agricultural and silvicultural activities,
including runoff from fields and crop and forest
land;
(B) mining activities, including runoff and siltation
from new, currently operated and abandoned surface
and underground mines;
(C) all construction activity, including runoff from
the facilities resulting from such construction. . ."
Both the Federal Government and state governments have struggled with the
complexity of nonpoint source pollution control. Control raises complex
legal problems by the very definition of it. Standards must be set and
identification of source becomes a burden to the legal framework of
enforcement.
More importantly—the number of actors necessary to control nonpoint
pollution becomes difuse and often confusing. Institutional and legal
remedies can not be wrapped up in neat packages. There are shortcomings
that can only be addressed below the Federal level.
P-IV-A-1
-------
ORGANIZATION AND INSTITUTIONAL ARRANGEMENTS
OF NFS CONTROL PROGRAMS
BACKGROUND
It is well established that until the Federal Water Pollution Control
Act (FWPCA) was revised by PL 92-500 in 1972, water quality efforts dealing
with nonpoint pollution were fairly limited to research. About the only
programs that dealt with nonpoint sources did so on an indirect basis.
Regular soil conservation programs and watershed protection activities of
SCS, cost-sharing programs such as AGP administered by ASCS, and pesticide
registration and labeling programs of ARS were the bulk of activities.
State efforts were generally conducted as an adjunct to these efforts.
Soil and water conservation districts and their associated agencies only
indirectly dealt with the problem in what was thought of as best a bootleg
operation.
PL 92-500 redirected the nation's efforts in water pollution abatement
efforts. These new efforts focused attention on controlling both point and
nonpoint pollutants at their source. Obviously, nonpoint took a secondary
seat to the massive efforts undertaken to control point sources. The
problem of point was more definable, the technology to a degree in place,
and the institutional arrangements more clearly defined.
It was difficult to define nonpoint sources as well from the legal
standpoint. The first consideration is the definition of pollution itself.
Legislation is designed to maintain water quality by controlling pollution
and that depends upon limiting contamination below some established
standard.
Then what about identifying the pollutant itself? If you identify it
does it become a source point and if you can't how can you legally impose
sanctions on the polluter?
Such was the struggle in simplistic terms, but as the entire water
quality program gained experience, it became clear that pollution
contributed from nonpoint sources could easily equal or exceed efforts in
point source control.
Nonpoint was never defined by statute but defined more by exclusion in
generally three categories.
(1) It was not susceptible to treatment at the "end of a
pipe" however controllable by other changes generally in land
management;
(2) It was convened by a natural process such as a storm or
seepage rather than through a deliberate controlled effort; and
(3) It was generally generated by a diffused set of man's
activities and not identified with any particular one.
P-IV-A-2
-------
Refining was done to this identification for a number of reasons
including the well known Natural Resources Defense Council v. Train which
narrowed the definition of nonpoint sources. That called for regulation
including the permit system to include smaller feedlots, irrigation return
flows and storm sewers in urban areas and certain activities of
silviculture such as rock crushing, gravel washing, log sorting and
storage. A howl of protest and a flurry of activity arose over this
ruling. Subsequent rules by EPA in 1976 softened some of the requirements
for irrigators through the issuance of general permits if in conformance
with 208 plans.
Farm and some closely aligned organizations like NACD spent a
considerable amount of time trying to educate the actors to be involved.
First to educate EPA and others aligned with them that nonpoint source had
to be handled differently than point source. A permit and enforcement
system just wasn't feasible for the bulk of nonpoint problems. It would
not work to place a policeman on every farm in America.
Secondly, an educational process was needed in EPA as well as state
water quality agencies that a mechanism was in place to handle a good
portion of the probem since it became clear that soil movement was a major
culprit in nonpoint sources. Nor was it as easy as establishing a single
set of laws or ordinances to handle problems associated with agriculture,
urban storm water, construction site run-off, mining run-off and
silviculture.
NACD through a series of grants from EPA, conducted a number of
nonpoint source pollution projects including a series of 45 state sediment
control institutes, a series of nonpoint pollution control workshops and a
number of publications including a series of 48 Nonpoint Notes published
over a period of years designed to inform all who were involved in the
nonpoint source areas. The Soil Conservation Service placed over 30 staff
people on IPA's in Washington and EPA regional offices to assist them in
establishing a nonpoint program. Cooperative Extension placed people on
IPA in regional offices or contracted for conducting educational
activities.
The Chicago office of EPA, because of the implications of nonpoint in
the Great Lakes Basin, effectively utilized Section 108 of the Water
Quality Act to fund several water basin projects to test certain
approaches. The Red Clay Project along Lake Superior tested new
methodology to combat sediment in red clay soils, the Washington County,
Wisconsin project was designed to take a look at institutional
arrangements, and the Black Creek, Indiana project was designed to look at
a variety of best management practices (BMPs) and determine methods for
measuring the results of BMPs against an improved water quality.
The appropriate roles of state and other management agencies received
their most significant direction by the issuance of SAM-31, a program
guidance memorandum establishing implementation criteria for EPA approval
of NPS elements of state water quality managements. This guidance document
P-IV-A-3
-------
stated that the appropriate level of government and the form of a
regulatory program depended on existing laws, programs, and governmental
working relationships. Conditional approval could be given by EPA if the
state could develop changes in legislative authorites or administrative
capabilities.
Fully approved NFS regulatory programs had to include the following
elements:
-enforcement authority
-a designated management agency to implement the program
-authority to use the control tools necessary to implement the program
-monitoring authority
-authority to require application of BMPs
-authority to control the activity> pollutant or geographical area
to be assigned
Sam-31 also provided for approval of non-regulatory NFS control
programs but such programs had to include
-a schedule for the program including implementation,
monitoring, and progress evaluations
-annual progress reporting
-adequate technical and financial assistance
-effective educational programs
-identification of BMPs
REMAINING INSTITUTIONAL AND ORGANIZATIONAL GAPS
While a decade has passed since the public attention was first turned
to nonpoint pollution, there remains a number of organizational and
institutional impediments to completing the task.
Lack of. reliable funding sources both at the state and Federal level,
Even though it is recognized that nonpoint source is the remaining
significant problem in water quality, Congress has been unable to
appropriate the funds necessary to attack the problems. At the state level
it is a mixed bag primarily because major emphasis in the last
several years has been on soil erosion control. Water quality and offSite
benefits are secondary in concern.
Lack of combined planning and implementation for water quality and soil
erosion.
Planning for soil erosion and water quality at the state level as well
as the Federal level is done in different agencies which have different
objectives. They are not necessarily coordinated at the state level and
use a separate set of disciplines.
P-IV-A-4
-------
Lack £f enforcement and standards leads to the lack of accountability.
With diffusion of best management agencies in the process including
Federal agencies, state agencies and local agencies, the public has a
difficult time of pin-pointing anyone accountable for program results or
expenditure of funds.
The institutional role for water quality control agencies in nonpoint is
not clear.
They must rely heavily on other conservation agencies that may be
reluctant to play supporting roles in the overall effort.
A diffusion of. educational effort combined with voluntary application is; _a_
problem.
Education of those necessarily involved in the process including those
landowners who apply them are diverse and numerous. On a national scale
this involves literally millions of people.
Unequal ability between states t_o adequately develop a_ state mechanism is
still a_ problem.
A river basin such as the Great Lakes or Susquehanna River has a bevy
of state capabilities which often do not mesh with the severity of the
problems.
An over dependency on data and data that may be flawed still exists.
This unecessarily complicates implementation of water quality
projects.
The external influcence of factors in controlling nonpoint are ever
present.
Weather, changes in technology in farming, state budget problems, all
are major factors in planning.
Conflict between old concepts £f treating every problem equally with some
effort rather than targeting the project i^s still a political problem where
responsibility is diffuse.
The relationship between USDA agencies and conservation districts which
most often provide technical assistance and coordination and state water
quality agencies i^s not well defined.
The providers and implementers have different expectations.
P-IV-A-5
-------
Overall, the institutional arrangements are not as clear and concise
as we might like because of the nature of the problem which is diffuse and
requires a wide variety of approaches. However, the funding for water
quality programs dealing with nonpoint supercede all problems. When
society funds these programs most of the implementing problems will
disappear.
P-IV-A-6
-------
NORTH CENTRAL STATES CONSERVATION DISTRICTS:
EFFECTIVENESS OF REGULATORY POWERS
Dean T. Massey
Agricultural Economist, Economic Research Service
U.S. Department of Agriculture
Law School, University of Wisconsin
Madison, Wisconsin, USA
ABSTRACT
Soil and water conservation districts are the only local agencies primarily
responsible for controlling soil erosion and reducing sediment. The 13 North
Central states have such districts or an equivalent. Districts in 9 of these
states are authorized to regulate land use. Kentucky, Nebraska, North Dakota
and Wisconsin provide for permissive regulatory powers; other states require
more mandatory controls. Iowa and Ohio regulations are based on soil loss
tolerances; Michigan's regulations are based on land-disturbing activities,
and South Dakota's, on conservation standards. Illinois has both permissive
regulations and mandatory standards for land-disturbing activities. A
district's effective abatement of nonpoint pollutants depends on the ease of
adopting regulations, conservation practices to be included, procedures for
administering regulations, and enforcement methods.
Adopting land use regulations in states with permissive regulatory powers can
be difficult, since proposed regulations must often be approved by a high
percentage of voters. These states permit districts to include adequate con-
servation practices in erosion control regulations for agricultural land.
None of these states specify how the regulations are to be administered or by
whom, although most provide for enforcement power.
In states with more mandatory controls, state agencies have more control over
nonpoint pollution. Regulations are handled in three ways: (1) districts
are required to adopt regulations and have them approved by a state agency;
(2) districts are required to adopt conservation standards and regulations
based on state agencies' programs and guidelines; or (3) state agencies are
required to adopt statewide control programs and regulations that are admini-
stered and enforced by state and local agencies.
Only Illinois, Iowa and South Dakota conservation districts administer land
use regulations. Permits are required for land-disturbing activities in
Michigan and South Dakota. Agricultural land in Illinois, Iowa and South
Dakota is deemed to comply with regulations until a complaint is made; the
districts then provide for enforcement. No enforcement provisions exist in
Ohio under either agricultural or urban pollution control rules.
Keywords: Agricultural runoff pollution, conservation practices, land use
regulations, nonpoint pollution, nonpoint pollution abatement,
regulatory powers, soil erosion, soil conservation districts
P-IV-B-l
-------
INTRODUCTION
Nonpoint source water pollution is a major environmental problem in much
of the United States. Sediment from erosion is the major nonpoint pollutant,
and agriculture contributes more than half of this, both by mass and volume.
Almost half of the agricultural sediment comes from cropland, and nearly half
of all cultivated cropland consists of soils with a high potential for ero-
sion.
Controlling sources of nonpoint pollutants is considered the best method
of preventing sediment and plant nutrients from reaching water resources.
Possible institutional tools for controlling sources of nonpoint pollutants
include prohibiting certain land uses at particular locations and regulating
uses on land at the source. Most of the responsibility for controlling non-
point pollutants rests at the local level, with federal and state governments
establishing water quality standards, guidelines, and soil loss limits, and
providing technical and financial assistance. Soil and water conservation
districts are the only local agencies whose primary responsibility is con-
trolling soil erosion and reducing sediment and other resultant nonpoint
source pollutants.
Section 208 of the Clean Water Act provides for the identification of
agricultural, urban and construction-related nonpoint sources of pollution,
and for the development of areawide water quality management plans containing
regulatory procedures to control pollutants from these sources (Iowa L.R..
1977). (Citations to applicable federal and state statutes and state agency
administrative regulations are provided in the references.) One or more
state, regional or local agency must be designated to implement the areawide
plans, and a possibility may exist for designating soil and water
conservation districts as management agencies (Davey).
This paper deals solely with the 13 North Central states: Illinois,
Iowa, Indiana, Kansas, Kentucky, Michigan, Minnesota, Missouri, Nebraska,
North Dakota, Ohio, South Dakota and Wisconsin. It describes the regulatory
powers of soil and water conservation districts in these states, and evalu-
ates the effectiveness of these regulatory powers in abating nonpoint sources
of pollution in both rural and urban areas; it finds that the effectiveness
of regulations depends on ease of adoption, conservation practices that may
be included, administrative procedures, and enforcement ability. District
regulatory powers are also evaluated for their effectiveness in providing the
local regulatory programs required to implement Section 208 areawide water
quality management plans.
DISTRICT ORGANIZATION AND REGULATORY POWERS
All 13 of the North Central states have soil and water conservation dis-
tricts or an equivalent. Nebraska eliminated conservation districts and
created 24 natural resources districts that assumed their functions.
Wisconsin abolished soil and water conservation districts and transferred
their functions to county boards, requiring them to create land conservation
committees to implement soil conservation programs.
Soil and water conservation districts and their equivalents, except in
P-IV-B-2
-------
Indiana, Kansas, Minnesota and Missouri, have some type of land use regula-
tory powers to control soil erosion and sediment. Illinois, Kentucky,
Nebraska, North Dakota and Wisconsin give districts or counties permissive
power to adopt land use regulations. More mandatory controls are established
in Iowa, Michigan, Ohio and South Dakota, and in Illinois (under a second
statute).
OPERATION OF LAND USE REGULATIONS
Adoption Procedures
District supervisors in four of the states with permissive regulatory
powers—Illinois, Kentucky, Nebraska and North Dakota—may develop and adopt
land use regulations after a public hearing and after the proposed regula-
tions have been approved by the required number of voters on a referendum.
Approval in the four states requires a favorable vote of between 75 and 90
percent of the qualified voters voting in the referendum. In Wisconsin, the
county land use conservation committee develops the land use ordinance and
submits it to the county board, who may adopt it after a public hearing.
However, the ordinance adopted by the board does not become effective until
approved by a majority of the voters on a referendum in the affected area.
Even though district supervisors or county boards in these five states
are authorized to adopt land use regulations, the requirements for approval
on a referendum as a prerequisite to adoption make adoption almost impossible
in these states, except Wisconsin. Requiring a referendum and a favorable
vote of more than 51 percent can block the enactment of regulations, espe-
cially if voting is limited to landowners within the district (Holmes).
Because of such adoption procedures, these states cannot effectively provide
the regulatory programs necessary for nonpoint pollution control under
Section 208 of the Clean Water Act.
Adoption of the more mandatory regulations in Iowa, Michigan, Ohio and
South Dakota, and under Illinois' second statute, involves a joint effort by
state agencies and local governments and soil and water conservation dis-
tricts (Davey; EPA, 1978). Illinois amended its original statute requiring
the State Department of Agriculture to adopt guidelines, and requiring dis-
tricts to adopt erosion control programs in accordance with state guidelines.
Iowa district commissioners have classified land in the state's 100 districts
and established and implemented soil loss limits for each classification.
Illinois and South Dakota require state agencies to prepare guidelines
for erosion and sediment control, and require conservation districts to adopt
regulations or standards based on the state guidelines (Holmes). The
Illinois Department of Agriculture and South Dakota Conservation Commission
have developed comprehensive statewide erosion and sediment control programs
and guidelines to be implemented and administered by the conservation dis-
tricts. Each district in Illinois had until April 1982 to develop and adopt
soil erosion and sediment control programs and standards consistent with the
state program and standards; the department reviewed and approved district
programs and standards, and also developed appropriate programs and standards
for districts that failed to develop their own. All except two conservation
districts in South Dakota have developed and adopted conservation standards
P-IV-B-3
-------
in cooperation and consultation with counties and municipalities; the commis-
sion reviewed these standards prior to adoption.
State agencies adopt land use regulations that are enforced by state
agencies and local governments in both Michigan and Ohio. The Michigan
Department of Agriculture, with the assistance of the conservation districts,
prepared and submitted a unified statewide soil erosion and sedimentation
control program for approval by the Water Resources Commission. The commis-
sion, with the assistance of the Department of Agriculture, prepared and
adopted rules for the unified program that became effective on January 1,
1975. The Ohio Division of Soil and Water Districts, with the approval of
the Soil and Water Conservation Commission, adopted standards to abate soil
erosion and degradation of state waters from agricultural, urban and live-
stock feeding activities (Holmes).
Conservation districts in the five states with more mandatory controls
come closer, as adoption procedures are concerned, to meeting the program re-
quirements of Section 208 water quality management plans. Illinois, Iowa and
South Dakota give districts the most authority in land use regulation: dis-
tricts in these states adopt their own soil erosion and sediment control pro-
grams, regulations or standards and have them approved by a state agency to
assure compliance with statewide programs. Problems would exist, however,
for districts in Michigan and Ohio, because rules and standards are adopted
at the state level, with districts playing only an advisory role.
Conservation Practices to be Included inRegulations
Legislation should be sufficiently broad to enable state and district
regulations to include a variety of soil management practices from which can
be selected the most appropriate combination for each situation or condition
(EPA, 1973). Criteria for selecting management practices include their abil-
ity to manage pollutants generated from nonpoint sources, compatability with
water quality goals, effectiveness in preventing or reducing pollutants,
practicability, and technical capability of preventing or reducing runoff.
Illinois, Kentucky, Nebraska, North Dakota and Wisconsin, with their
permissive powers, do not require the inclusion of specific provisions in
their land use regulations. Enabling legislation in these states, except in
Wisconsin, generally provides that district regulations may contain any one
or a combination of the following: (a) requirements for performing necessary
engineering operations; (b) requirements for observing particular methods of
cultivation; (c) specifications for cropping programs and tillage practices
and (d) provisions for removing highly erodable areas from cultivation
(Holmes).
Enabling legislation in these states is agriculturally oriented and
applies to soil erosion control (Davey; EPA, 1973). Wisconsin's statute has
the strongest provision prohibiting land uses and management practices that
cause excessive erosion, sedimentation and nonpoint source pollutants. Land
use regulations adopted by conservation districts in the other four states
would probably be too narrow to fulfill the requirements of the Section 208
areawide management plans. Such regulations, if adopted, would adequately
manage only nonpoint pollutants derived from agriculture.
P-IV-B-4
-------
Conservation practices that may be included in regulations of the five
states with more mandatory programs vary. Iowa's program is based on soil
loss limits, and Michigan's is based on land-disturbing activities. A combi-
nation of soil loss tolerances and standards for land disturbances forms the
basis of the Illinois, Ohio and South Dakota programs.
Landowners in Iowa are required to employ soil conservation and erosion
control practices on their land, but district commissioners may not specify
particular practices to be employed as long as the owner voluntarily complies
with established soil loss limits. Commissioners also may not require the
employment of erosion control practices on land used in good faith for only
agricultural purposes and may not prohibit operators from fall plowing.
The Michigan Water Resources Commission adopted rules based on land-
disturbing activities, implementing the Department of Agriculture's recom-
mended unified statewide program that included provisions for land use plans
and permits. Under the rules, all persons engaged in earth changes that re-
quire a permit must design, implement and maintain acceptable erosion and
sedimentation control plans that effectively reduce erosion. Permits are re-
quired when disturbing one or more acre of land, or if the disturbance is
within 500 feet of a lake or stream. In addition, approved soil and water
conservation district erosion and sedimentation control standards and speci-
fications must be followed by those requiring a permit. Agricultural prac-
tices were included in the rules in 1979, and now agricultural earth distur-
bances, except for plowing and tilling, or unless the farmer has an agreement
with the district, are subject to permit and plan requirements.
Conservation practices incorporated in land use regulations in Illinois,
Ohio and South Dakota are based on a combination of soil loss tolerances and
conservation standards for land disturbances. Under Illinois Department of
Agriculture guidelines, all conservation systems and practices applied to
agricultural land must seek to reduce soil losses over a period of years to
levels at or below the soil loss tolerance ("T" value) established by the
Soil Conservation Service. A "T" value is the average annual soil loss, in
tons per acre, that a given soil may experience and still maintain its pro-
ductivity over an extended period of time. Policies and specifications for
various erosion and sediment control devices, structures and practices ap-
plied to agricultural land were also given in the guidelines. Programs and
standards for various soils and land uses adopted by the districts contain
criteria, guidelines, techniques and methods for controlling erosion and
sediment from land-disturbing activities (Holmes).
District conservation standards or soil loss tolerance limits in South
Dakota are to be consistent with the State Conservation Commission's guide-
lines. The guidelines contain recommended soil loss limits and suggested
conservation practices and methods for controlling erosion and sediment
(Holmes; EPA, 1978). Counties adopt ordinances incorporating district stan-
dards relating to agricultural and non-agricultural land-disturbing activi-
ties; non-agricultural activities subject to permit requirements must adhere
to the district's soil loss standard and to a soil erosion and sediment con-
trol plan (Holmes).
The rules adopted by the Ohio Division of Soil and Water Districts to
P-IV-B-5
-------
abate agricultural nonpoint pollutants provide for the achievement over a
period of years of the applicable soil loss tolerance or permitted soil loss
values. Farmers are required to practice conservation and to follow a man-
agement system, so that under given cropping and management conditions, the
predicted soil loss (using the Universal Soil Loss Equation) from erosion is
equal to or less than the "T" value for that specific soil series, as speci-
fied in the Ohio Erosion Control Guide or the SCS Technical Guide. Farmers
may not use earth-disturbing practices, including tillage, on land immediate-
ly adjacent to water, unless practices are constructed or implemented in ac-
cordance with proper engineering designs.
Conservation practices that may be included in the land use regulations
of the five states with a more mandatory approach are generally broader than
those of the states with permissive regulations. All are based on land-
disturbing activities for both agricultural and non-agricultural activities,
and all include sediment control in addition to erosion control. One draw-
back to the Iowa and South Dakota regulations is that they do not apply to
agricultural land unless the erosion exceeds specified soil loss limits and a
complaint is filed. However, considering the management practices and land-
disturbing activities that may be included in the content of land use regu-
lations, the statutes and regulations of these five states are probably broad
enough to fulfill the regulatory requirements of Section 208 areawide water
quality management plans.
Administrative Procedures
Nothing in the Illinois, Kentucky and North Dakota statutes states how
land use regulations are to be administered or who administers them.
Nebraska statutes provide only that districts have power to administer the
rules and regulations; Wisconsin provides that the county board prescribe
administrative procedures in the land use ordinance. In these five states,
enabling legislation concerned with administrative procedures is probably in-
adequate to effectively provide the necessary regulatory programs for non-
point pollution in the Section 208 areawide plans. Such legislation should
provide who is to administer regulations, the method for administering,
whether permits are to be issued, and, if so, who issues them, and if plans
and specifications are to accompany permit applications. The powers and du-
ties of the administrator should be specified along with the permit, plan,
and specification requirements.
Of the five states with more mandatory regulatory powers, only Michigan
and South Dakota provide much detail on administrative procedures. Regula-
tions are administered by districts in Illinois and Iowa, by counties in
Michigan, by local permit-issuing authorities in South Dakota, and by the
State Division of Soil and Water Districts in Ohio.
Illinois does not require persons engaged in land-disturbing activities
to apply to soil and water conservation districts for permits or to submit
erosion and sediment control plans. Districts, upon request, make adequate
information and technical assistance available to persons engaged in land-
disturbing activities. Neither are permits or plans required for land-
disturbing activities in Iowa.
P-IV-B-6
-------
The Ohio Division of Soil and Water Districts established procedures for
the districts to administer agricultural and urban pollution abatement rules
and to enforce animal waste management rules. The division also recommends
criteria and procedures for approving urban sediment pollution abatement
plans and issuing permits. Under the agricultural pollution rules, the divi-
sion will inform the districts of state standards, criteria and procedures
for pollution abatement; enter into agreements with districts for implemen-
ting agricultural pollution abatement programs; provide administrative guid-
ance for implementing and administering such programs; and implement agricul-
tural pollution abatement programs in districts failing to negotiate an
agreement with the division. Permits, plans or specifications are required
for land-disturbing activities in Ohio.
Counties in Michigan administer the state-adopted rules; county boards
designate a particular agency responsible for enforcing the rules. Each
county enforcing agency adopts the district's standards and specifications.
Persons engaging in earth changes must obtain permits from appropriate county
enforcing agencies, and must submit soil erosion and sedimentation control
plans with permit applications. District standards and specifications that
have been approved by county enforcing agencies must be followed and utilized
as they apply to specified earth changes.
Conservation standards adopted by districts in South Dakota are adminis-
tered by local permit-issuing authorities; neither the State Conservation
Commission nor conservation districts may issue permits. Each permit-issuing
authority within the district must include provisions in its permit procedure
to ensure that any proposed action is in compliance with district conserva-
tion standards. Persons engaged in agricultural or minor land-disturbing
activities need not prepare a conservation plan, secure a permit, or report
their activities to the conservation district unless they violate the dis-
trict standards.
Not only are Michigan and South Dakota alone in providing much detail on
administrative procedures, they are also the only states requiring the issu-
ance of permits for land-disturbing activities. Basically, only Michigan
statutes provide sufficient detail on who administers the conservation pro-
gram and how it is administered to be adequate for Section 208 water quality
management plans. Administration in Michigan, however, is not by the soil
conservation districts, but rather by counties.
Enforcement Procedures
A variety of methods are available in the states with permissive powers
to enforce the land use regulations. Of these five states, only Nebraska
does not provide for enforcement powers. Wisconsin goes further by providing
that county ordinances should prescribe administrative and enforcement proce-
dures. District supervisors have the authority in Illinois, Kentucky and
North Dakota to enter upon any lands affected by the regulations to determine
compliance.
Illinois and Kentucky districts may provide in their regulations that
persons damaged by someone violating regulations may recover from the viola-
tor. Wisconsin ordinances may be enforced through civil forfeitures or by
P-IV-B-7
-------
injunctions in court actions initiated by the county board or the land con-
servation committee. There are no criminal penalties or forfeitures in
Illinois, Kentucky or North Dakota. District supervisors in Illinois,
Kentucky and North Dakota may initiate a court action compelling violators
to perform the work, or permitting supervisors to perform the work themselves
if the violators failed to perform in a specified time. Conservation dis-
tricts or counties in Illinois, Kentucky, North Dakota and Wisconsin have
adequate available enforcement powers over their land use regulations to per-
mit regulation by such agencies in a Section 208 areawide management plan.
Enforcement authority also varies in the five states with a more manda-
tory approach. Agricultural land in Illinois, Iowa and South Dakota is
deemed to comply with regulations unless a complaint is made; the conserva-
tion districts are then responsible for providing at least the initial phases
of the enforcement process. Counties and a state agency provide the enforce-
ment in Michigan, while there are no enforcement provisions for the Ohio
agricultural or urban pollution rules.
Any person in Illinois engaging in land-disturbing activities is encour-
aged to comply with district standards for erosion and sediment control.
Complaints may be filed with the conservation district by any person, the
district, or the Department of Agriculture if it is believed that a serious
problem exists. If the district finds that standards have been violated, it
notifies the violator and seeks voluntary compliance, gives a deadline for
compliance, and suggests practices enabling the person to comply with the
standards.
Landowners in Iowa are required to employ either conservation or erosion
control practices. An owner or occupier of land being damaged from sediment
files a written complaint with district commissioners, charging that soil
erosion is occurring in excess of the district limits. If the commissioners
find sediment damages and excessive soil erosion, they are required to issue
to the violator an administrative order stating the extent to which soil ero-
sion exceeds the district's soil loss regulations; if the land is agricultur-
al, the order states when the needed work is to be commenced and satisfactor-
ily completed. Persons failing to comply with the order may be held in con-
tempt and fined or imprisoned.
South Dakota does not require persons engaged in agricultural or minor
land-disturbing activities to prepare a plan or obtain a permit unless the
district determines that a particular land disturber is violating district
standards. Once such a violation has been determined, the land disturber is
ordered to prepare an erosion and sediment control plan, to have it approved
by the district, and to implement the plan. Either the district or permit-
issuing authority may initiate a court action for an injunction or other
appropriate relief to enforce the order.
County or local enforcing agencies in Michigan are responsible for en-
forcing the state land use regulations. Persons who fail to get a required
permit for earth-disturbing activities are guilty of a misdemeanor. County
or local enforcing agencies notify the Water Resources Commission of all
violations of the state regulations or local land use ordinances. If the
p-IV-B-8
-------
commission finds a violation has occurred, it proposes an appropriate correc-
tion agreement to the violator. If the violator does not agree to the pro-
posal, an order requiring correction, enforceable by the court, is issued by
the commission. The county or local enforcing agency may also issue cease
and desist orders and revoke permits upon finding violations of regulations
or approved soil erosion and sedimentation control plans.
The statutes giving the Ohio Division of Soil and Water Districts author-
ity to adopt rules for agricultural pollution and urban sediment pollution
abatement do not provide enforcement powers. If, however, the conservation
districts receive complaints on violations of the standards, they will attempt
to find a solution through voluntary cooperation.
Soil and water conservation districts are involved in enforcement in all
these states except Michigan. However, soil erosion control in Illinois and
Ohio is voluntary; there are no means of enforcing compliance with the stan-
dards, except through cooperative efforts. Only in Iowa and South Dakota do
soil and water conservation districts meet the regulatory agency requirements
of a Section 208 areawide water quality management plan from an enforcement
standpoint.
CONCLUSIONS
All 13 North Central states have soil and water conservation districts
or similar local governmental units. Districts in nine states either possess
or are involved with some type of nonpoint source pollution regulatory power.
Adopting land use regulations in the five states with permissive regulatory
powers is a difficult process and involves stiff requirements for approval on
a referendum. Enabling legislation in these states permits districts to in-
clude adequate conservation practices and measures in the land use regula-
tions for controlling soil erosion on agricultural land. None of the permis-
sive regulatory states specify who is to administer land use regulations or
how they are to be administered. All the states except Nebraska provide for
enforcement powers.
State administrative agencies in states with more mandatory regulatory
powers — Illinois, Iowa, Michigan, Ohio and South Dakota—have more authority
in the nonpoint source pollution control process than do state agencies in
the states with permissive regulatory powers. Conservation practices and
methods are based on soil loss limits, conservation standards and land-
disturbing activities, and they apply to both agricultural and non-
agricultural uses of land. Only in Illinois, Iowa and South Dakota are soil
and water conservation districts responsible for administering land use regu-
lations. Permits are used for land disturbances in Michigan and South Dakota.
Agricultural land in Illinois, Iowa and South Dakota is deemed to comply with
the regulations unless a complaint is made; the districts then provide for
enforcement. Counties and a state agency provide enforcement in Michigan,
and there are no enforcement provisions in Ohio for either agricultural or
urban pollution control rules. Only districts in Illinois, Iowa and South
Dakota have sufficient powers to be considered adequate management agencies
for implementing the Section 208 areawide water quality management plans.
P-IV-B-9
-------
REFERENCES
Davey, W.B. (1977). Conservation Districts and 208 Water Quality Management.
National Association of Conservation Districts, Washington.
Holmes, B.H. (1979). Institutional Basis for Control of Honpoint Source
Pollution Under the Clean .Water Act--Hith Emphasis on Agricultural
Nonpoint Sources. U.S. Environmental Protection Agency WH-554, Washington.
U.S. Environmental Protection Agency (1973). Methods and Practices for
Controlling Water Pollution from AgriculturalNonpoint Sources. EPA-430/9-
73-015, Washington.
U.S. Environmental Protection Agency (1978). Alternative Policies for
Controlling Nonpoint Agriculture Sources of Water Pollution. EPA-500/5-
78-005, Washington.
Federal and State Statutes and Administrative Regulations
United States Code, Title 33, Sec. 1238 (1982).
Illinois Annotated Statutes, Chap. 5, Sees. 128 to 130 (Smith-Hurd 1975).
(permissive)
Illinois Annotated Statutes, Chap. 5, Sees. 111(8), 138.3 to 138.9 (Smith-
Hurd Cumulative Supplement 1984-1985). (mandatory)
Illinois Department of Agriculture. Illinois Erosion and Sediment Guidelines
(April 18, 1980).
Iowa Code Annotated, Sees. 467A.42 to 267A.51 (West 1971 and Cumulative
Supplement 1984-1985).
Kentucky Revised Statutes, Sees. 262.350 to 262.520 (1981 and Cumulative
Supplement 1984-1985).
Michigan Compiled Laws Annotated, Sees. 282.104 to 282.117 (1979 and
Cumulative Supplement 1984-1985).
Michigan Administrative Code, Rules 323.1701 to 323.1714 (1979).
Nebraska Revised Statutes, Sees. 2-3244 to 2-3250 (Reissue 1983).
North Dakota Century Code, Sees. 4-22-27 to 4-22-39 (1975).
Ohio Revised Code, Sec. 1515.30 (Page Supplement 1983).
Ohio Administrative Code, Chapts. 1501:15-1, 1501:15-5.
South Dakota Codified Laws Annotated, Sees. 38-8A-1 to 38-8A-21 (1977 and
Supplement 1984).
South Dakota State Conservation Commission and South Dakota Department of
Agriculture. Suggested Guidelines for Local Erosion and Sediment Control
Programs (1977).
Wisconsin Statutes, Sec. 92.11 (1981-1982).
P-IV-B-10
-------
EMERGING INSTITUTIONS FOR NFS CONTROL
By
Glenn E. Stout
Water Resources Center
University of Illinois
2535 Hydrosystems Laboratory
208 North Romine Street
Urbana, IL 61801
The management of our water and land resources require an
interdisciplinary approach involving the many users of the resources. For
fifty years, agriculture has been attempting to reduce nonpoint pollution,
NFS, by various programs which had some success in the fifties, but, due to
new technologies, the grass waterways approach has disappeared. New
cultivation techniques are helping to reduce the NPS, but the users of the
land are often unappreciative of the impact of the loss of soil and its
consequences throughout the watershed and the subsequent deposition in lakes
and deltas. We need a certain amount of erosion in order to maintain the
coastal estuaries and delta regions. But, an excess amount is unnecessary for
the natural ecosystem. The city dweller and the rural land manager has not
been aware of the consequences of NPS as it affects the rest of the land
resources, lakers and the ecosystem. Soil erosion fills in drainage ditches
and causes lakes to lose their storage capacity and subsequent transformation
into a wetland.
In the late 70s, residents of Mattoon approached the Water Resources
Center for assistance in restoring their man-made lake which was rapidly
filling with sedimentation. It was rather obvious that the problem was on the
watershed and that there was a need for the lake owners to talk and
communicate with the farmers in the area. This was a very difficult task
since it involved a municipality, residents around the lake as well as the
users of the land on the watershed. In fact, the city officials were
reluctant to take any action because of the possible incresed taxes that might
be invoked upon their city. Finally, after numerous meetings and publicity in
the area, the region developed a concensus that there was a need to work
together.
The city of Mattoon, who owned the lake, changed its attitude and began
to support erosion control program on the Lake Paradise watershed. Farmers in
the area received free water from the city, but, at the same time, were not
aware that their soil was creating a water quality and supply problem for
future users. Everyone was finally awakened to the issue and now there is a
general concensus in the area that erosion is a problem for everyone. In
order to maintain a high-quality water supply, the problem has to be resolved
at the source.
The Water Resources Center has held three meetings on lake and watershed
management (1, 2). Last fall, 200 people gathered in Springfield for a
meeting which involved the Association of Soil and Water Conservation
Districts, the EPA, owners and managers of lakes and scientists. As a result
P-IV-C-l
-------
of this meeting and other efforts in the state, there have been several new
cooperative programs developed involving the integration of lake and watershed
management.
Several years ago, David Daily, a student of soils at the Southern
Illinois University at Edwardsville, undertook a personal project to determine
the soil losses on the watershed of 4,000 acres for Holiday Lake, a nearby
private lake, which was experiencing a heavy siltation (3). He went to the
management of the lake and pointed out to them that the source of the problem
was the watershed. The owners were receptive to his concerns, since they
would have to spend $45,000 per year for dredging. Mr. Daily then explained
the potential solutions to the landlords on the watershed. ASCS was
approached for matching funds and the local Madison County Soil and Water
Conservation District agreed to administer a sediment reduction plan
regardless of the ASCS backing. As a result, the Holiday Lake Homeowners
Association are providing funds to combat the erosion on critical portions of
the watershed. This effort was initiated outside of government and is an
excellent example of how people can work together in forming a local
institution to solve local problems.
In October at the North American Lake Management Association meeting in
Geneva, several papers are being presented which will describe additional
recent examples of cooperative local efforts involving the Association of Soil
and Water Conservation Districts, the Illinois EPA and the private sector.
The big issue in solving local problems is the compatibility of the
individuals--everyone has to be unselfish and interested in the welfare'of the
entire community. Although there will be personal biases, these must be
minimized in order to create an atmosphere whereby an institutional
arrangement is established to solve local problems. Universities and state
agencies can provide technical assistance, but they should not be involved in
the legislation and demands of extravegant plans to resolve these problems.
Many of the local erosion NFS programs could easily be solved by small units
of government or individuals working together to protect the environment for
the future of mankind.
REFERENCES
Water Resources Center. (1980). Proceedings of a conference on restoring
man-made lakes in Illinois - February 19, 1980.
Water Resources Center. (1981). Proceedings of a round table on reclaiming
and managing lakes inIllinois.
Water Resources Center. (1984). Proceedings of the Illinois Conference on
Lake and Watershed Management - November 8-9, 19B4.
P-IV-C-2
-------
Technical Reports
-------
APPLICATIONS OF GEOGRAPHICAL INFORMATION SYSTEM AND HYDROLOGIC
MODELING TO AN AGRICULTURAL WATERSHED IN ILLINOIS
Ming T. Lee, Hydrologist
and
Rodolfo Camacho, Research Assistant
Illinois State Water Survey
Champaign, Illinois, USA
ABSTRACT
The state-of-the-art appraisals of water quality impacts induced by non-point
sources are difficult due to many factors. Lack of existing research data
base, the diffusive characteristics of non-point source pollutants, and the
technical difficulties relate the instream water quality to the topographic
features, soil characteristics, land use, land cover, soil moisture condi-
tions, and the management practices in the watersheds. In order to overcome
these difficulties, the comprehensive monitoring and evaluation program was
established in Highland Silver Lake Watershed.
The basic techniques may be divided into three categories: 1) field monitor-
ing, 2) geographical information system, and 3) watershed modeling. These
three components are mutually dependent. The field monitoring is described
in an accompanying paper. The geographical information system consists of
two basic types of data: cartographic coordinates which identify the points,
lines, and polygons and attributes data which describe the characteristics of
these features. The computer oriented GIS performs manipulation and analysis
which include map overlay, buffer generation, dissolve boundaries (regroup
and reclassify), tabular analysis and network analysis. The watershed model-
ing utilizes the data from field and GIS to evaluate the effects of watershed
management practices. The MULTSED and AGNPS models were used. Results of
the modeling effort are discussed.
Keywords: Nonpoint source pollution, geographical information system,
watershed modeling, Best Management Practices, agricultural
watershed, sediment load, runoff, water quality.
T-I-A-l
-------
Introduction
Watershed modeling is a tool for hydrologic system synthesis, prediction, and
optimization. Hydrologic modeling requires simplification or abstraction.
There are two ways to pursue this abstraction: lumped or distributed systems
(Chow, 1964; Haan et al., 1982). In lumped systems, non-uniform parameters
such as rainfall, soils, land use, and topographic characteristics are aver-
aged to obtain representative values for the entire drainage basin. On the
other hand, a distributed system has the built-in capability of assembling
the non-uniform parameters and simulating the interaction of these parame-
ters. Thus, a distributed system should be more flexible for describing
spatial-temporal dynamics of overland and channel flows. However, this type
of model requires more detailed input data.
One way to overcome these difficulties is to use a geographical information
system (GIS). A brief description of 6IS is as follows.
Geographical Information System
In general, a geographical information system consists of two basic types of
data: cartographic coordinates which identify the points, lines, and poly-
gons; and attribute data, which describe the characteristics of these fea-
tures. The computer-oriented GIS such as ARC/INFO (ESRI, 1984) can perform
sophisticated manipulation and analyses which include map overlay, buffer
generation, dissolve (regroup and reclassify), tabular analysis, and network
analysis. The system contains a series of computer mapping and display
capabilities for generating high quality cartographic displays. The user can
specify size and scale to produce the desired map outputs.
Highland SiTver Lake Project: An Example
The Highland Silver Lake Project is one of thirteen experimental projects for
the National Rural Clean Water Program in the United States. The project ha^
been in operation since September 1980. The project consists of field data
collection and evaluation of different land management practices applied in
the watershed. Since land management practices consist of different tillage
practices and different crop sequences on different soil and topographical
conditions, the collected field data alone will not be enough to determine
the effects of management practices in the entire project area. Therefore,
utilization of a watershed model is needed to evaluate the impacts of land
management practices on water quality and on stream and reservoir sedimenta-
tion, as well as the costs and benefits of the program.
Data Base of Highland Silver Lake Watershed
In the Highland Silver Lake Project, the watershed data base consists of a
hydrographic map (stream networks and water bodies), soil maps, land use
data, and slope map. The hydrographic map was digitized from a standard U.S,
Geological Survey 7-1/2 minute topographic map with the recent aerial photor
graphs as supplemental data sources. The soil map was digitized from pub-
lished county soil survey maps. The land use data were obtained from recent
aerial photographs. The slope map was developed on the basis of topographic
maps published by the U.S. Geological Survey. Watershed and sub-basin boun-
daries and locations of monitoring stations were also coded in the GIS.
T-I-A-2
-------
Agricultural Non-Point Source Pollution Model (AGNPS)
This model was developed by the North Central Soil Conservation Research
Laboratory, Agricultural Research Service, USDA (Onstad and Young, 1983).
AGNPS is a single event, deterministic, and distributed model. It configures
the watershed into small square subareas called "cells." It is at this level
that all characteristics are established and calculations made.
Figure 1 shows a subwatershed of the Highland Silver Lake watershed. The
model is composed of the following parts: (1) runoff component, (2) channel
flow, (3) erosion, (4) sediment transport, (5) nutrient transport, and
(6) feedlot.
The basic output from AGNPS includes a hydrology component which estimates
both volume and peak runoff; a sediment component which estimates upland
erosion, gully erosion, and sediment yield; and a nutrients component which
estimates nitrogen and phosphorus yields in terms of concentration and load.
The output can be examined on a cell basis or for an entire watershed. Sedi-
ment analyses are broken down into particle-size classes.
The second model is the Multiple Watershed Storm Water and Sediment Runoff
Simulation Model (MULTSED) (Simons et al., 1978 and 1981).
MULTSED Model
This model is also a single event, distributed, deterministic model. The
model contains two basic components: a hydrologic and hydraulic routing
component and a sediment component. The calculations are conducted in three
basic units: (1) two-plane and single channel, so called Wooding-plane sub-
watershed as shown in figure 2 (Wooding, 1965); (2) channel unit; and (3) a
single-plane unit. The number and size of these units for a specific water-
shed can be selected by the users.
Use of GIS for Input Data Preparation
The two selected models that have been described represent two types of model
configurations. AGNPS represents the grid-cell system. MULTSED represents
the irregular configuration which reflects the hydrologic units such as
Wooding-plane, single overland flow, and single channel.
The input data for the AGNPS model consist of: (1) cell identification
number; (2) receiving cell identification number; (3) SCS curve number;
(4) land slope in percent; (5) land slope shape factor; (6) field slope
length in feet; (7) channel slope in percent; (8) channel side slope in
percent; (9) Manning's roughness coefficient for the channel; (10) soil
credibility factor; (11) cropping factor; (12) practice factor; (13) surface
condition constant; (14) aspect; (15) soil texture number; (16) fertilization
level; (17) availability factor; (18) point source level; (19) gully source
level; (20) chemical oxygen demand factor; and (21) impoundment factor.
In order to compile these input data from the GIS data base, the first step
is to construct a "template" which delineates the watershed into sub-units.
In the case of the AGNPS model, the grid-cell system is the template. The
T-I-A-3
-------
Figure 1. An example of grid-cell system of AGNPS model
T-I-A-4
-------
Line E
(a) Original Subwatershed Topographic Map
(b) Wooding Plane Representation
Figure 2. Watershed configuration of MULTSED model
T-I-A-5
-------
computer software is able to retrieve the attribute data and the sub-units
from the data base. However, few items in the input data require field
inspection to acquire the necessary data. Similarly, the data can be re-
trieved for the MULTSED model.
Results
AGNS was run for field site 5 of the Highland Silver Lake watershed using a
10-acre grid system as shown in figure 1. A 4.4-in. rainfall storm was
considered which represents an extreme storm for this watershed in the last
three years. Table 1 shows the results given by the AGNPS and MULTSED model
for the same event.
Table 1. AGNPS and MULTSED Model Results for a 4.4-in. Rainfall Event
in Field Site 5 of the Highland Silver Lake Watershed
AGNPS MULTSED
Runoff volume (in.) 2.8 3.6
Runoff ratio (%) 64.0 82.0
Peak runoff rate (cfs) 382.0 283.0
Sediment yield (tons/acre) 0.327 0.089
The primary results indicated that runoff volumes and runoff ratio show
higher values by using MULTSED model, but peak runoff rate and sediment yield
show higher values by using AGNPS model. Further verification with field
observation is being conducted.
Three events during 1982 and 1983 were selected for the calibration ofthe
MULTSED model. The events were selected so that they fairly represent the
average characteristics of the events for these years. The events considered
were: 1) May 31, 1982, with a rainfall of 0.88 in. and a duration of 4
hours; 2) June 4, 1982, with a rainfall of 0.4 in. and a duration of 11
hours; and 3) March 18, 1983, with a rainfall of 1.34 in. and a duration of
16.5 hours.
The parameters used in the calibration were: 1) the hydraulic conductivity,
2) initial soil moisture, 3) rainfall splash detachment coefficient, and 4)
roughness coefficients. These parameters were adjusted by doing sensitivity
analysis on their value ranges until the simulated and measured runoff and
sediment concentration hydrographs were matched. Other parameters such as
ground and canopy cover, porosity, average suction head, final soil moisture,
interception by ground cover or canopy, and soil temperature were kept con-
stant. Table 2 shows the results of the three selected events. A selected
runoff and sediment hydrographs are shown in figure 4.
Conclusions
1. Application of the MULTSED model for a portion of the Highland Silver
Lake watershed has been performed. Satisfactory simulation of runoff and
sediment hydrographs was obtained. Very little is known on the relationships
between the detachment coefficients and the land use and watershed management
practices in the MULTSED model. An attempt to find these relations will be
done when the entire basin is studied.
T-I-A-6
-------
Table 2. Runoff and Sediment Yield for Field Site 5 and Gaging
Station 3 of Highland Silver Lake Predicted by MULTSED Model
Event
Subarea A
May 31, 1982
June 4, 1982
March 18, 1983
Measured Computed Runoff Measured Computed
Rainfall runoff runoff Error ratio sediment sediment Error
(1n.) (in.) (in.) (%) (%) (tons) (tons) (%}
0.88
0.40
1.34
Subarea B
May 31, 1982 0.88
June 4, 1982 0.40
March 18, 1983 1.34
0.198
0.093
0.352
0.213
0.053
0.550
0.204
0.097
0.355
0.222
0.053
0.521
3.04 22.5
3.85 23.3
0.64 26.3
3.99
0.06
-5.2
24.2
13.3
41.0
N.A.
0.55
2.44
38.32
4.60
80.37
1.15
0.48
2.23
27.10
4.61
80.58
-13.5
-8.7
-27.0
2.1
0.25
N.A. = not available
T-I-A-7
-------
LEGEND
SW = Subwatershed Unit
PL = Plane Unit
CH = Channel Unit
A = Drainage Area of FS5 (SW1)
B = Drainage Area of SyV2, PL1, PL2
GS3
Figure 3. Gaging Station 3 Sub-watershed of Highland Silver Lake,
Highland, Illinois
T-I-A-8
-------
024
<£ °
|i
< 0.5
oc
150
". 125
$ 100
K
<
I 75
50
25
Time, hours
2000
a
^ 1500
o
z
o
<
£ 1000
LU
o
O
o
o
LU
I/)
5 10
TIME, hours
15
500
GS3
5 10
TIME, hours
15
• Measured Concentration
Measured
Predicted
FSB Field Site No. 5
GS3 Gaging Station No. 3
5 10
TIME, hours
Figure 4. Runoff and Sediment Hydrographs of March 18, 1983 of
Highland Silver Lake Watershed, Highland, Illinois
T-I-A-9
-------
2. A methodology illustrates that a geographical information system can
provide an efficient way to compile necessary input data for complex dis-
tributed hydro!ogic models.
3. The development of the data base in a geographical information system is
independent of a specific model. The data can be retrieved from GIS to meet
the requirement of different geometric configurations of watershed models.
4. The sensitivities of various geometric delineations and hydrologic proc-
esses in a watershed can be evaluated by utilizing a geographic information
system.
References
Chow, V. T. (ed.) Handbook of Applied Hydrology. McGraw-Hill Book Company*
1964.
Environmental System Research Institute, ARC/INFO User Manual, Version 2.3,
Redlands, California, 1984.
Haan, C. T., H. P. Johnson, and D. L. Brakensiek, Hydrologic Modeling of
Small Watersheds, American Society of Agricultural Engineers, Monograph
Series, 1982.
Onstad, C. A., and R. A. Young, A Procedure for Prioritizing Water Quality
Problem Areas, Paper No. 83-2156, presented at the 1983 Summer Meeting,
American Society of Agricultural Engineers, held at Montana State Uni-
versity, Bozeman, Montana, June 26-29, 1983.
Simons, D. B., R. M. Li, and B. E. Spronk, Storm Water and Sediment Runoff
Simulation for a System of Multiple Watersheds - Vol. I. Water Routing
and Yields, CER77-78DBS-RML-BES47, Colorado State University, April
1978.
Simons, D. B., R. M. Li, W. T. Fullerton, and T. R. Grindeland, Storm Water
and Sediment Runoff Simulation for a System of Multiple Watersheds -
Vol. II. Sediment Routing and Yield, Colorado State University, October
1981.
Wooding, R. A., A Hydraulic Model for the Catchment-Stream Problems^
I. Kinematic Wave Theory, Journal of Hydrology, Vol. 3, No. 3/4,
pp. 254-267, 1965.
T-I-A-10
-------
Delineating Sources of Agricultural Nonpoint Source Pollution
W.T. Dickinson*, R.P. Rudra*, and G.J. Wall»»
* Professor and Assistant Professor, respectively, School of Engineering,
University of Guelph, Guelph, Ontario, Canada, N1G 2W1
** Regional Director, Environment Canada, 55 St. Clair Avenue East, Toronto,
Ontario, Canada, M4T 1M2
T-I-B-1
-------
INTRODUCTION
Fluvial sediment and associated nutrients and chemicals are now
acknowledged to be a major cause of contamination in river and laKe systems
throughout North America (7, 8). A prime source of these contaminants is
sheet and rill erosion from agricultural cropland (19). Such soil erosion is
widely distributed over the landscape, but the rates at which soil erodes and
yields sediment to streams can be highly variable from field to field and from
watershed to watershed (15). Therefore, although soil erosion has been termed
a nonpoint source of pollution, most basins have more or less specific
locations which yield much of the sediment and associated contaminants. The
design and selection of cost-effective non-point source pollution control
strategies requires delineation of key source areas.
Recent erosion/sedimentation research studies at Guelph have led to the
development of a method for estimating sources and rates of sediment yield in
small agricultural watersheds (2, 6). Examples of the application of the
method to two agricultural watersheds in Southern Ontario are presented in
this paper, to reveal the utility of watershed modelling for delineating soil
erosion and sediment yield source areas.
LITERATURE REVIEW
The notion of sediment contributing areas, i.e. that prime sources of
sediment in agricultural regions constitute an area less than and in some
cases much less than the total watershed area, has received considerable
attention in recent years (3» 13» 1b» 17). Much in the erosion/sedimentation
literature, although not specifically focussed on the definition of sediment
sources, provides useful background for the location of significant sediment
sources. Some field-scale models provide estimates of annual erosion (22),
and others predict annual sediment loads directly (9, 13» 22). Another
category of models predicts annual amounts of erosion and sediment on the
T-I-B-2
-------
basis of the Universal Soil Loss Equation (23) and a functional relationship
for watershed delivery ratio or field transport factors (10, 14, 16, 17, 1$,
21).
Single storm event and continuous erosion/sediment/chemical models have
also been developed for both the field scale (11) and the watershed scale (1).
This group is more descriptive of the physical and chemical processes involved
and can be used to delineate key sources. However, the large volume of input
data required for these models, some of which is not always readily available,
is often a constraint on their use.
Therefore, although there is now an extensive literature on soil erosion
and fluvial sedimentation which provides useful background and detailed
distributed models, there has not been a method developed with the prime
purpose of delineating key source areas. Further, although there has been
considerable evidence that stream sediment loads are usually highly variable
throughout the year (2, 5, 12, 19, 20), the variation of seasonal sediment
sources has not been addressed. GAMES - the Guelph model for evaluating the
impact of Agricultural Management systems on Erosion and Sedimentation - has
been developed for the purpose of predicting on a seasonal basis the magnitude
and location of soil loss by water erosion and subsequent sediment yield
METHODOLOGY
Several criteria have been kept in mind during the development of GAMES.
The model was to take account of seasonal variations in soil loss and sediment
transport mechanisms, but was not to become encumbered with the data
requirements of an event on continuous model. The basic computational land
units were to be field size, with the intent of applying the model to small
agricultural watersheds. Although a predictive capacity was deemed to be
desirable, the prime purpose of the model was to distribute sediment yields
T-I-B-3
-------
throughout a watershed on a quantitative and objective basis.
The resulting model, GAMES, has been suitably constructed for the
computation of seasonal potential soil loss (the soil loss component) and
seasonal sediment yield (the delivery component) for field-size cells in small
agricultural watersheds. The soil loss component is based on the Universal
*
Soil Loss Equation (23) has been modified for seasonal application. That is,
^3 - RsCsKaLS.P3
where AS = seasonal soil loss per unit area, Rs = seasonal rainfall factor,
Cs s seasonal land use or cropping factor, Ks = seasonal soil credibility
factor, LS = slope gradient and slope length factor, and P = seasonal
erosion control practice factor. The sediment yielded by each field unit to
downstream field and stream units is determined from the product of the
seasonal potential soil loss and the seasonal delivery ratio for the unit.
The seasonal delivery ratio, or percentage of the potential soil loss which
can be expected to be delivered downstream in a given season, is determined
from the expression,
DR3 =a
1
- . S . Hc<
L.ns
where 0 £ DRS <. 1J n3 = seasonal surface roughness coefficient, as indexed
by Manning's n; S = cell slopej Hcs = seasonal hydrologic coefficient, an
index regarding the probability and depth of overland flow in a season; Lg s
seasonal length of overland flow path; and a, 3 = calibrated parameters.
Application of the soil loss and delivery components initially involves
the development of a composite overlay of land use, soils, and land slope to
divide the watershed area into land units each of which is characterized by a
single land use, a single soil type, and a single slope class. Then values of
the various variables in the USLE (23) and DR (2) expressions are determined
T-I-B-4
-------
and assigned to each unit. Determination of the a and 3 parameters for a
watershed involves an optimization routine to compare the accumulated water-
shed sediment load calculated by means of GAMES for assigned values of and a
and 3 with a measured sediment load or one estimated by another means. Output
for a calibrated application is generated on a unit by unit basis for each
season considered, and includes potential soil loss, potential soil 3,033 per
hectare, delivery ratio to the adjacent downstream unit, delivery ratio to the
nearest stream, sediment yielded to the stream, and sediment yielded per hectare.
With the selection of soil loss and sediment yield tolerance levelsf J.t
is possible to extend the application of GAMES into the area of problem
identification (4). The framework shown in Table 1 has been suggested on the
basis of distinct definitions for erosion and sediment problems. An erosion
problem is considered to be fundamentally a source problem, the loss of soil
creating hardships such as decreases in agricultural productivity, loss of
tillable land, damage to buildings, roads, equipment etc. A sediment problem
involves downstream situations e.g. siltation of streams and reservoirs,
pollution of fish spawning areas, increased water treatment costs. Areas of a
watershed which fall into problem category I may be expected to exhibit both
high soil erosion rates and high sediment yield rates. Areas in category II
have high erosion rates, but contribute relatively little soil to downstream
locations. Category III includes those areas which have relatively low soil
loss rates, but from which most of the eroded soil moves into the stream
channels contributing significantly to sediment problems. Category IV is the
"no problem" category.
For the sake of illustrating the utility of a model for delineating soil
loss and sediment yield source areas, GAMES has been applied to two small
agricultural watersheds located in the Thames River Basin of Southern Ontario.
T-I-B-5
-------
Table 1: Framework for soil erosion and sediment
problem identification
Problem
Category
I
II
III
IV
Erosion*
Problem
X
X
.
"
Sediment**
Problem
X
-
X
"
* Agricultural productivity is adversely affected
** Sediment yield and associated nutrient contribu-
tion to the stream is excessive.
T-I-B-6
-------
The Stratford/Avon Watershed is comprised of 434 hectares of rolling upland
soils, cropped predominantly in continuous corn with fall ploughing and few
if any conservation practices. The Big Creek Watershed is 2141 hectares in
extent on quite flat lowland soils. It is almost totally in row crops (e.g.
continuous corn, beans, vegetables) which are fall ploughed.
SAMPLE RESULTS
Sample output maps from the GAMES application are presented in Figures 1
through 6. Figures 1 and 2 reveal the spatial pattern of spring potential
soil loss quantities for the upland and lowland watersheds, respectively. The
estimated average spring gross erosion per unit of watershed area is not large
for either basin, being 2.5 tonnes per hectare for the lowland basin and 3.5
tonnes per hectare for the upland basin. However, the spring erosion
estimates are quite variable in the upland Stratford/Avon Watershed, ranging
from essentially zero to 25 tonnes per hectare. The range is much smaller
across the lowland Big Creek Watershed, varying only from 0 to 3 tonnes per
hectare. It is also evident from Figure 1 that a major portion of the eroded
soil in the upland watershed moves within a small percentage of the basin.
For example, 92 percent of the spring sheet and rill erosion volume is
estimated to occur in 4? percent of the watershed area. The areal
distribution of soil loss is more uniform in the lowland basin (Figure 2).
About 56 percent of the landscape contributes 64 percent of the sheet and rill
erosion.
Sample spatial distributions of sediment yield from the upland and
lowland agricultural basins are presented in Figures 3 and 4. Similar to the
erosion picture presented above, but more pronounced, the great majority of
the spring sediment load leaving the upland watershed is estimated to emanate
from a small percentage of the watershed area. For example, 83 percent of the
load is generated in 14 percent of the basin. Indeed, such a watershed is
T-I-B-7
-------
Tonnes Per Hectare
Figure 1. Spatial variation of soil erosion on upland watershed for
spring conditions,
T-I-B-8
-------
Tonnes Per Hectare
Q<2.5
Figure 2.
so11 eroslon
T-I-B-9
-------
Tonnes Per Hectare
Figure 3. Spatial variation of sediment yield on upland watershed for
spring conditions.
T-I-B-10
-------
Tonnes Per Hectare
Figure 4. Spatial variation of sediment yield on lowland watershed for
spring conditions.
T-I-B-11
-------
Problem Type
Figure 5. Spatial distribution of soil erosion and/or fluvial sediment
areas on upland watershed for spring conditions.
T-I-B-12
-------
Problem Type
Figure 6. Spatial distribution of soil erosion and/or fluvial sediment
areas on lowland watershed for spring conditions.
T-I-B-13
-------
characterized by rather distinct sediment sources. On the other hand, the
lowland watershed is seen to generate its sediment load rather uniformly
across the basin, and sediment yield hotspots are not evident (Figure 4).
Figures 5 and 6 present maps of the four soil erosion and sediment
problem categories on the basis of a spring erosion tolerance level of 2.5
tonnes per hectare and a spring sediment yield tolerance level of 1.0 tonnes
per hectare. A quantitative summary relating to these figures is given in
Table 2.
The sample results regarding erosion and sedimentation problem
identification reinforce the results discussed previously, but also reveal
additionalsignificant points. It is clear from Table 2 that 4? percent of
the upland watershed and 75 percent of the lowland watershed contribute
significantly to erosion and/or sediment problems. In and upland
Stratford/Avon watershed 14 percent of the basin exhibits almost half of the
soil loss and is the source of more than three quarters of the sediment load.
Only 4? percent of the basin is estimated to have spring erosion rates greater
than the defined tolerance level, accounting for 92 percent of the total basin
soil loss; while only 14 percent of the basin is estimated to serve as a prime
sediment source, contributing 63 percent of the downstream loads. More than
half of the landscape of this basin exhibits potential spring soil loss and
sediment yield less than the defined erosion and sediment yield tolerance
rates.
The lowland Big Creek watershed, typical of many level lowland watersheds
in Southern Ontario, has more than half of its area in erosion problem
categories. Fifty-six percent of the basin has spring erosion rates greater
than the defined tolerance level. About one third of this basin is estimated
T-I-B-14
-------
Table 2. Quantitative identification of soil erosion and sediment
problems in selected agricultural watersheds
Problem
Category
I
II
III
IV
Comparative i
Percentages
% of Area
% of Soil Loss
51 of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
% of Area
% of Soil Loss
% of Sediment Load
Stratford /Avon
Watershed
14
45
83
33
47
15
0
0
0
53
8
2
Big CreeK
Watershed
12
13
18
44
51
34
19
17
30
25
19
18
T-I-B-15
-------
to provide a prime source of sediments contributing to half of the downstream
loads. It should be noted that 19 percent of the basin exhibits potential
soil loss less than the defined erosion rate tolerance. Nonetheless, the same
area contributes about one third of the downstream sediment load at sediment
yield rates above the defined sediment yield rate limit.
CONCLUSIONS
A watershed soil erosion and fluvial sedimentation model, GAMES, has been
shown to provide a very useful tool not only for evaluating the seasonal and
spatial variability of field potential soil loss and sediment yield in an
agricultural watershed but also for delineating spatially the distribution of
erosion and sediment problems. For selected soil loss and sediment yield
tolerances, the approach affords in map and summary statistics form
information regarding the location and extent of areas contributing to (i)
both erosion and stream sediment problems, (ii) only erosion problems,' (iii)
only sediment problems, and (iv) no problems. Such description and
delineation of agricultural nonpoint sources is proving to be of great
assistance in the planning and targetting of nonpoint pollution control
programs in rural areas.
ACKNOWLEDGEMENTS
Sincere appreciation is expressed to all members of the very large team
of researchers associated with the various phases of this project. And in
addition to the funds provided by Environment Canada through the Department of
Supply and Services, the financial support of the Natural Sciences and
Engineering Research Council of Canada, the Ontario Ministry of Agriculture
and Food, the Thames River Implementation Committee, and the Stratford/Avon
River Environment Management Program is gratefully acknowledged.
T-I-B-16
-------
REFERENCES CITED
1. Beasley, D.B., L.F. Huggins and E.J. Monke. 1980. ANSWERS: A model for
watershed planning. Transactions of the ASAE, 23:938-944.
2. ClarK, D.J. 1981. An expression for determining sediment delivery ratio.
M.S. Thesis. Univ. of Guelph, Guelph, Ontario, Canada.
3. DicKinson, W.T. and G.J. Wall. 1977* Temporal and spatial patterns in
erosion and fluvial processes. Proc. 5th Guelph Symp. on Geomorphology;
"Research in Fluvial Geomorphology." Geo Abstracts Ltd., East Anglia.
4. Dickinson, W.T. and R. Pall. 1982. Identification and control of soil
erosion and fluvial sedimentation in agricultural areas of the Canadian
Great Lakes Basin. Final Rep. to Supply and Services, Canada. School
of Engr., Univ. of Guelph, Guelph, Ontario, Canada.
5. Dickinson, W.T., A. Scott and G.J. Wall. 1975. Fluvial sedimentation in
Southern Ontario. Can. J. Earth Sci., 12(11):1813-1819.
6. Dickinson, W.T., R. Pall and G.J. Wall. 1984. GAMES - A model for identi-
fying sources and amount of soil erosion and fluvial sediment. Can.
Hydrol. Symp., Quebec City, P.Q., Canada.
7. Duttweiler, D.W. and H.P. Nicholson. 1983- Environmental problems and
issues of agricultural nonpoint source pollution. From "Agricultural
Management and Water Quality", edited oy F.W. Schaller and G.W. Bailey.
Iowa State Univ. Press, Ames, Iowa, 3-16.
8. International Joint Commission. 1980. Annual report of the International
Joint Commission. I.J.C., Windsor, Ontario.
9. Jansen, J.M.L. and R.B. Painter. 1974. Predicting sediment yield from
climate and topography. J. of Hydrol., 21:371-386.
10. Kling, C.P. and C.W. Olson. 1974. The sediment transport computer model.
Cornell Agri. Mimeo 74-11, Dept. of Agro., Cornell Univ., Ithaca, N.Y.
T-I-B-17
-------
11. Knisel, W.D. Jr., Ed. 1981. CREAMSJ A field scale model for chemical,
runoff and erosion from agricultural management systems. U.S. Dept. of
Agri., Cons. Res. Rep. No.26.
12. Mildner, W.F. and R.C. Boyce. 1979. Monthly variation in soil loss and
sediment yield. ASAE Paper No.79-2528. Amer. Soc. Agri. Engr.,
St. Joseph, Ml.
13« McPherson, H.J. 1975. Sediment yield from intermediate-sized stream
basins in Southern Alberta. J. of Hydrol., 25:243-257.
14. Onstad, C.A. and C.R. Foster. 1975. Erosion modelling on a watershed.
Transactions of the ASAE 20:89-95.
15. Pall, R., W.T. Dickinson and R. McGirr. 1982. Climatic and soil
conditions significant to soil erosion and sedimentation in Southern
Ontario. NAR-ASAE Paper No. 82-201. Amer. Soc. of Agri. Engrs., St.
Joseph, MI.
16. Piest, R.F. 1965. The role of the large storm as a sediment contribution.
Proc. 1963 Federal Inter-Agency Conference on Sedimentation. U.S.Dept.
of Agri., Misc. Publ. 970:97-108.
17. Renfro, G.W. 1972. Use of erosion equations and sediment delivery ratios
for predicting sediment yield. Sediment-Yield Workshop. U.S. Dept. of
Agri. Sedimentation Lab., Oxford, Miss. 33-45.
18. Robinson, A. R. 1977. Relationship between soil erosion and sediment
delivery. IAHS Symp. on Erosion and Solid Matter Transport in Inland
Waters. IAHS Publ. No.122.
19. van Vliet, L.J.P., G.J. Wall and W.T. Dickinson. 1978. Erosion losses
and sediment delivery ratios for agricultural watersheds. I.J.C. Tech.
Rep. re: Great Lakes Pollution from Land Use Activities, Windsor,
Ontario, Canada.
T-I-B-18
-------
20. Wall, G.J., W.T. Dickinson and L.J.P. van Vliet. 1979. Agricultural
sources of fluvial suspended sediment. Prog. Wat. Tech., 11(6):U8l-M99.
21. Williams, J.R. 1972. Sediment yield production with Universal Soil Loss
Equation using runoff energy factor. Sediment-Yield Workshop, U.S.
Dept. of Agri. Sedimentation Lab., Oxford, Miss., 244-252.
22. Williams, J.R. and H.D. Berndt. 1976. Sediment yield prediction based on
watershed hydrology. ASAE Paper No. 76-2535. Amer. Soc. Agrl. Engr.,
St. Joseph, Ml.
23. Wischmeier, W.H. and D.D. Smith. 1965. Predicting rainfall-erosion
losses from cropland east of the Rockey Mountains. Agri. Res. Ser., U.S.
Dept. of Agri. Handbook No. 262.
T-I-B-19
-------
COUPLING NOMPQINT POLLUTION ME. WATER QUALITY MODELS:
AN EXAMPLE FOR THE GJJEEJJ. BAY - FjQX. RIVER WATERSHED
Thoaas M. Heldtke, Associate Professor, Department of Civil
Engineering, Wayne State University, Detroit, MI 48202, Martin T.
Auer, Associate Professor, Department of Civil Engineering, Michigan
Technological University, Houghton, MI 49931, Raymond P. Canale,
Professor, Department of Civil Engineering, University of Michigan,
Ann Arbor, MI 48109, and Theodore A.D. Slawecki, Computer Engineer,
Limno-Tech, Inc., Ann Arbor, MI 48103
ABSTRACT
Green Bay is a major gulf located in the northwest corner of Lake Michigan
which has been cited as one of the major water quality problem areas in the
Great Lakes. Historically, interest in water quality degradation has been
limited to the lower Pox River and extreme southern Green Bay, the site of
municipal and industrial pollutant discharges. Recent surveys indicate
dissolved oxygen depletion far out into the bay, beyond the immediate
Influence of the Fox River's organic loading. Attention is now focused uppn
nonpoint sources of phosphorus as stimulants for primary production and the
ultimate cause of oxygen depletion problems. The Fox River contributes 78$ of
the annual tributary load of total phosphorus to Green Bay. Of this amount,
only 3% is attributable to the Green Bay Metropolitan Sewerage Authority, the
region's largest point source discharger of phosphorus.
Nonpoint source loads of total phosphorus are estimated using a microcomputer
model which considers land use, soil texture, and the phosphorus loading
associated with each location in the land use/soil texture matrix (UALs, unit
area loads, kgP/ha'year). Land use classifications and soil textures are
developed specifically for the Fox River watershed and UALs are derived from
recent studies of nonpoint source loadings (PLUARG, 208 Areawide Water Quality
Management Planning Program). Annual total phosphorus loads to Green Bay from
the Fox River are calculated for baseline (= existing) conditions and six
additional land use scenarios as the sum of the products of the land use/soil
texture and unit area load matrices for each component sub-basin in the Fox
River watershed.
Load estimates are input to a steady-state mass balance phosphorus model which
calculates the phosphorus concentration along the major longitudinal axis of
Green Bay. Trophic levels corresponding to points along the phosphorus
gradient are established using a trophic index. A marked gradient in trophic
conditions exists along the bay which is sensitive to variations in land use
practices impacting nonpoint phosphorus loads. The hypothetical scenarios
help demonstrate the utility and of interactive, coupled nonpoint pollution
and water quality microcomputer models.
Keywords: Nonpoint source pollution model, water quality model, abatement of
pollution from agriculture, phosphorus, eutrophication, Green Bay,
Lake Michigan, Great Lakes
T-I-C-1
-------
INTRODUCTION
Modeling approach
Microcomputer models for the estimating of nonpoint source pollutant
loadings and receiving water quality response may be coupled and used to
examine the impact of land use activities on water quality. This manuscript
focuses on nonpoint source loads of phosphorus to Green Bay, Lake Michigan and
their relation to the trophic state gradient within that system. Seven
hypothetical land use managment scenarios are described to illustrate the
application of interactive planning and management tool. The methodology is
versatile and may be modified to address other substances and their impact
through construction of pollutant-specific unit area load tables and
quantification of sources and sinks for the parameter of interest.
Green Bay and the Fox River
Green Bay is a gulf approximately 160 km long and 22 km wide, located in
the northwest corner of Lake Michigan (Figure 1). The bay has a mean depth of
15.8 m, a volume of 67 km^ and a residence time of 6 years (Mortimer, 1978).
Historically, industrial and municipal discharges of oxygen-demanding
substances and plant nutrients have led to severe dissolved oxygen depletion
during summer in the lower Fox River and extreme southern Green Bay. Recent
evidence of hypolimnetic oxygen depletion over much greater portions of the
bay has resulted in the current interest in nonpoint sources of algal
nutrients, specifically phosphorus. The Fox River is the single greatest
hydrologic and pollutant source for Green Bay, contributing ^5% of the annual
tributary flow and 78$ of the annual tributary total phosphorus load
(Roznowski and Auer, 1984). Strong longitudinal gradients in trophic status
are set up in response to the Fox River phosphorus loading — hypereutrophic
conditions exist in the southernmost region near the river's mouth, while
oligotrophic conditions prevail in the northern reaches near the junction with
Lake Michigan.
NONPOINT SOURCE MICROCOMPUTER MODEL
Model Structure
The microcomputer model used to assess average annual total phosphorus
loadings from nonpoint sources within a given drainage system is based on
three critical assumptions:
1 ] The total phosphorus load generated by surface runoff from a given
site may be represented by a unit area load (UAL) — the mass of
phosphorus contributed per unit area of land per unit time.
2] Under identical climatological conditions, the UAL for a given area
is principally determined by the predominant land use and soil
texture within that area.
3] UALs are derived to reflect nonpoint source loadings for an "average"
year of wetness (total precipitation).
T-I-C-2
-------
Lake
K Michigan
rLake
f Winnebago
I
N
I
FIGURE 1. Location of Green Bay and major morphological features.
T-I-C-3
-------
To evaluate the annual nonpoint source total phosphorus loading, the area
under study is divided into a set of sub-basins or sub-watersheds representing
the major hydrologic units within the system. Each sub-basin is then further
divided into a set of cells which represent the area comprised of a specific
land use/soil texture combination. The area of each cell is multiplied by its
appropriate UAL and the products summed to arrive at the total annual nonpoint
source load for any sub-basin. The cumulative load for the entire drainage
system is simply obtained as the sum of the contributions for each of the
major sub-basins as indicated by Equation 1 below:
n n
TP = y TP. = y
L, i Li
i=l i-1
m
UAL.
c=l
(1)
where TP: annual nonpoint source load for the entire
drainage system
TPi: annual nonpoint source load from sub-basin i
(i = 1, 2, ..., n)
UALj k: annual nonpoint source total phosphorus load
per unit area of land use j and soil texture k
(j = 1, 2, .... m; k = 1, 2, ..., 1)
*i i k: area °f land use j and soil texture k within
'J> sub-basin i
The microcomputer model allows a user to quickly conduct a preliminary
macroscopic nonpoint pollution loading analysis on any drainage area of
interest. The accuracy or goodness of the analysis is a direct function of
the user's ability to accurately represent the land use, soil texture and
hydrologic characteristics of the system under study.
The model is comprised of two major sub-programs: (1) SETUP, which
prompts the user to define the land use and soil texture matrix for each sub-
basin in the drainage system and to provide the UAL matrix for the pollutant
of interest, and (2) LOAD, which integrates the data provided to SETUP with
new information on land use changes or pollutant controls. Output from LOAD
includes tabular and graphical displays of nonpoint source pollutant loadings
for each major sub-basin and for any set of land use management/pollutant
control scenarios under consideration.
Application to Fox River/Green Bay
Land Use/Soil Texture Description
The nonpoint source pollution model is applied in a preliminary
assessment of annual total phosphorus loads to Green Bay from the Fox River
drainage basin. Information on land use and soil classifications within the
area were obtained from the Fox Valley Water Quality Planning Agency (FVWQPA)
and Soil Conservation Service offices for the five counties comprising the Fox
River watershed. For the purposes of this demonstration, soil classifications
were grouped into three major soil texture categories ~ coarse, medium, and
fine; eight land use categories were considered in the analysis:
T-I-C-4
-------
1] Residential 5] High Tillage Cropland
2] Commercial 6] Low Tillage Cropland
3] Industrial/Transportation 7] Pasture
4] Institutional/Government 8] Woodland/Residual
The Fox River watershed was divided into three sub-basins (Lower Fox
River, Lake Winnebago, and River-Lakes) as shown in Figure 2. Acreages for
each soil association in a given sub-basin were calculated from county soil
survey maps. Soils in the River-Lakes (97$ fine) and Lake Winnebago (99$
fine) sub-basins are assumed to be comprised of 100$ fine textured soils. The
Lower Fox River sub-basin contains 35$ coarse soils, 20$ medium soils, and 15$
fine soils.
Land use areas within the three sub-basins are dominated by cropland (64-
69$), residual (17-25$), and residential (4-8$). The residual category
includes woodlands, wetlands, and grasslands. Because UALs for any areal cell
are dependent on the predominant land use and soil texture, it is necessary to
apportion land use areas to the three soil texture classifications noted
above. For this demonstration, the acreage of each land use corresponding to
a given soil texture category is approximated using a simple weighting ratio
of soil texture area to total sub-basin area. The acreage within each sub-
basin corresponding to a specific soil texture and land use is defined by the
resulting land use/soil texture matrix.
High Tillage vs. Low Tillage Cropland
The land use analysis conducted for this demonstration did not
distinguish between low tillage and high tillage cropland areas. High tillage
crop management practices are defined here as conventional or deep tillage
methods consisting of complete soil inversion following crop harvest. Farming
areas utilizing conservative tillage management systems, including minimum
tillage (e.g., noninversion soil conditioning with chisel plows) and no till
methods, have been grouped together and defined here as low tillage cropland.
For the purposes of baseline calculations, it is assumed that the high tillage
and low tillage cropland areas are equally represented in the three sub-
basins .
Unit Area Load Matrix
Unit area loads used to estimate nonpoint source total phosphorus
contributions within the Fox River watershed were derived from storm event
monitoring data and literature values reflecting similar climatological, land
use and soil texture conditions. Among the major information sources utilized
in developing UALs were: a] surface washoff measurements from 21 test
watersheds in northern Virginia (NVPDC/VPI, 1978), b] monitoring of 12 test
watersheds in metropolitan Washington, DC (NVPDC, 1983)f c] data from 11 test
watersheds in the Chesapeake Bay drainage basin (Hartigan et al., 1983), d]
pilot watershed measurements in the Great Lakes basin (PLUARG, 1978), e]
methodologies of the water quality assessment procedure for toxic and
conventional pollutants (Tetra Tech, 1982), and f] simulations of water
quality and hydrology for the Occoquan Reservoir, Virginia (Office of
Comprehensive Planning, Fairfax County, Virginia, 1982).
The UALs applied in this demonstration for the Fox River and Green Bay
T-I-C-5
-------
FOX RIVER WATERSHED
SUB-BASINS
RIVER LAKES
SUB-BASIN
LOWER FOX RIVER
SUB-BASIN
LAKE WINNEBAGO
SUB-BASIN
FIGURE 2. Component sub-basins of the Fox River watershed.
T-I-C-6
-------
are presented in Table 1. Total phosphorus contributions to the Fox River and
Green Bay from interf low/basef low water are not included in the UAL values
presented in Table 1; these loads are treated as constant at a rate of 0.018
kg/ha*year over the entire watershed. The UALs applied in this demonstration
are intended to represent annual total phosphorus contributions from surface
runoff events within the Fox River watershed during an "average" year of
wetness or total precipitation. The current analysis has not attempted to
quantify and compare climatological and physiographic factors among the areas
from which UAL information was compiled.
Table 1. Total Phosphorus UAL Matrix (kg/ha*year)
Land Use
Coarse
Soil Texture
Medium
Fine
Residential
Commerical
Industrial/Transportation
Institutional/Government
High Tillage Cropland
Low Tillage Cropland
Pasture
Woodland/Residual
0.06
0.28
0.18
0.09
0.55
0.22
0.06
—
0.13
0.28
0.37
0.18
0.74
0.29
0.07
0.02
0.18
0.28
0.55
0.28
0.92
0.37
0.09
O.OM
Average annual total phosphorus loads are calculated for each sub-basin as the
product of the land use/soil texture matrix and the unit area load matrix.
Land Use Scenarios and Loading Estimates
The nonpoint source microcomputer model is used to generate estimates of
average annual total phosphorus loads to the Fox River for seven hypothetical
land use scenarios. These test cases are not intended to represent expected
land use trends within the drainage area; rather, they demonstrate the value
of the model in obtaining a rapid, macroscopic estimate of nonpoint loadings
under a diversity of land use conditions. In addition, the test cases show
how the model may be used to examine potential load reductions and
concommitant water quality improvements resulting from implementation of
remedial action within given portions of the watershed.
The seven scenarios include a baseline (=existing) condition, a "best-
case" condition (all basin acreage in woodland), a "worst-case" condition (all
basin acreage in high tillage), and three intermediate scenarios (baseline
conditions plus combinations of high and low tillage). The estimated nonpoint
source total phosphorus loadings to Green Bay from the Fox River are presented
in Table 2.
Model Output and Analysis
An examination of model-estimated loadings points to the importance of
the cropland contribution to the overall load — an expected result
considering the high UAL assigned to cropland areas and the high acreage of
cropland present within eachof the Fox River sub-basins. The average annual
total phosphorus load for Scenario 1 — baseline conditions — is estimated at
T-I-C-7
-------
Table 2. Land Use Scenarios and Annual Phosphorus Loads
Scenario
1. Baseline (includes 50$ High Till)
2. Woodland
3. High Till Cropland
4. Baseline/100$ High Till
5. Baseline/75* High Till
6. Baseline725$ High Till
7. Baseline/100$ Lo Till
« Watershed total; includes baseflow
Total Phosphorus Load
(kgP/year)§ (kgP/day)«
861,725
91,993
1,675,647
1,186,948
1,024,338
699,114
536,499
2361
252
4591
3252
2806
1915
1470
861,725 kgP (2361 kgP/day). Almost 90$ of that total is attributable to
runoff contributions fom cropland. The influence of cropland on nonpoint
loadings is further evidenced by the significant increases and decreases in
the expected loads which accompany changes in the percent of cropland
distributed between the high tillage and low tillage categories. The water
quality implications of these hypothetical land use scenarios are now examined
by coupling output from the nonpoint source model to a microcomputer-adapted
trophic response model.
WATER QUALITY MICROCOMPUTER MODEL
Model Structure and Baaelina Conditions
The water quality microcomputer model is a steady-state, 12-cell mass
balance model for phosphorus; details of model development, calibration, and
verification are provided elsewhere (Auer and Canale, 1985). The model
considers sources (tributary loads) and sinks (net settling, mass transport)
for total phosphorus in Green Bay and calculates the steady-state phosphorus
profile for each of 12 model cells oriented along the major (NE/SW) axis of
Green Bay. A comparison of model output with summer average total phosphorus
concentrations for 1982 is provided as Figure 3. Trophic conditions for each
model cell are evaluated using the methodology of Chapra and Dobson (1981); a
total phosphorus concentration <11.0 ugP/L Indicates oligotrophy, 11.0 - 21.7
ugP/L — mesotrophy, and >21.7 ugP/L — eutrophy. The gradient in trophic
state corresponding to the phosphorus levels in Figure 3 (Baseline + 50$ High
Till) is presented as Figure 4. Under baseline conditions, the four
southernmost model cells (approximately 20 km) are eutrophic, the mid-bay
region is mesotrophic and the outer reaches are oligotrophic. It should be
noted that significant oxygen depletion occurs in the hypo limnetic waters of
the mid-bay region — an area currently classified as mesotrophic.
Response to
Scenarios
The microcomputer nonpoint source loading model estimates a baseline
annual total phosphorus loading of 2361 kgP/day; this value compares well with
the five-year average (2174 kgP/day, range 1240-2716 kgP/day) calculated from
monitoring data and discharge permit reports by Roznowski and Auer (1984).
T-I-C-8
-------
wo-
Green Bay Summer Average -1982
Data-mean*S.D.
— Mode! Output
20 40 60 80 100
DISTANCE FROM FOX RIVER MOUTH (km)
FIGURE 3. Steady-state model output compared with measured summer average total phosphorus con-
centration at 12 stations in Green Bay; bars (standard deviation, n = 18) indicate temporal
variation at each station.
T-I-C-9
-------
Baseline
Scenario I
TROPHIC STATE
OLIGOTROPHIC
MESOTROPHIC
EUTROPHIC
Woodlands
100% HiTill
Scenario 2
Scenario 3
Base* 100% Hi
Base •»-75% Hi
Scenario 4
Scenario 5
Base-i-25% Hi
Base+100 %Lo
Scenario 6
Scenario 7
FIGURE 4. Trophic status for Green Bay under baseline conditions and 6 hypothetical land use scenarios.
T-I-C-1C
-------
The microcomputer water quality model calculates summer average total
phosphorus concentrations based on the summer average tributary total
phosphorus load. Scenario-specific summer average loads for the Fox River are
calculated according to Equation 2:
W = '"seen / wbase> * wsum <2>
where:
W = scenario-specific summer average total phosphorus load (kgP/day)
Wscen = scenario-specific annual total phosphorus load (kgP/day)
= baseline annual total phosphorus load (kgP/day)
Wgum = baseline summer average total phosphorus load (kgP/day)
The baseline (1982) summer average total phosphorus loading to Green Bay from
the Fox River is 959 kgP/day (Auer and Canale, 1985) and the baseline annual
total phosphorus loading to Green Bay (Table 2, Scenario 1) is 2361 kgP/day --
thus the summer load accounts for approximately 41$ of the annual load.
Illustrations of the gradient in trophic conditions for each land use
scenario are presented in Figure 4. The most dramatic change in trophic
status is that for the hypothetical "best case" condition (Scenario 2)
corresponding to 100 percent woodland where oligotrophic conditions prevail
over the entire length of the bay. The load yields a summer average Fox River
total phosphorus concentration of 11.3 ugP/L, only slightly above the limit
for oligotrophy, and approaching the Lake Michigan boundary condition of 8
ugP/L. Output from this scenario may reflect conditions prior to cultural
development in the basin. Only slightly less dramatic is the hypothetical
"worst case" condition (Scenario 3), where the entire basin is characterized
by high tillage cropland. In this case, eutrophic conditions persist for
approximately 65 km out the bay and overlay the regions most susceptible to
hypolimnetic dissolved oxygen depletion. These results indicate that a wide
range in water quality conditions is possible for Green Bay -- and that the
current gradient in trophic state is strongly related to land use practices.
The "worst-" and "best case" scenarios demonstrate the sensitivity of the
system to land use practice, but are not realistic or viable management
options. Scenarios 4-7 examine the impact of variation in high tillage/low
tillage ratios under the baseline scheme for land use. The output for
Scenario 5 (Baseline + 75$ High Till) is not appreciably different from that
of the baseline condition. Expansion of high tillage practices to baseline
plus 100$ high till (Scenario 4) extends the region of eutrophy by
approximately 10 km. Similarly, reduction of high tillage practices to
baseline plus 25$ high till (Scenario 6) or baseline plus 100$ low till
(Scenario 7) reduces the region of eutrophy to the southernmost 10-15 km of
the bay. Implementation of 100$ low tillage practices offers the additional
advantage of extending the oligotrophic zone into the mid-bay region where
oxygen depletion has been reported under baseline conditions. Specific
water quality impacts may be evaluated for each scenario through the
application of indices relating phosphorus and transparency or chlorophyll.
T-I-C-11
-------
CONCLUSIONS
Microcomputer models for nonpoint source phosphorus loads and water
quality are coupled to examine the impact of variation in land use practices
on trophic state in Green Bay. Trophic response to various land use
management scenarios is dramatic because of the large contribution O8555) by
cropland to the annual Fox River nonpoint total phosphorus load. Model output
indicates that land use management, especially low tillage practices, offer an
opportunity for improved water quality in the bay. The utility of this
interactive planning and management tool has been demonstrated for several
hypothetical land use strategies in the Fox River/Green Bay system.
REFERENCES
Auer, M.T. and R.P. Canale. 1985. A phosphorus budget for Green Bay. Report
to U.S. Environmental Protection Agency, ERL-Duluth, MN (In Preparation).
Chapra, S.C. and H.F.H. Dobson. 1981. Quantification of the lake trophic
typologies of Naumann (surface quality) and Thienemann (oxygen) with
special reference to the Great Lakes. J. Great Lakes Res., 7(2): 182-193.
Fairfax County Office of Comprehensive Planning. 1982. Occoquan Basin Study.
Fairfax County, Virginia, 162 pp.
Hartigan, J.P., Quasebarth, T.F., and E. Souther land. 1983. Calibration of NFS
model loading factors. J. Env. Eng. Div., ASCE, 109(6): 1259-1272.
Mortimer, C.H. 1978. Water movement, mixing, and transport in Green Bay. p.
10-56, In Green Bay Workshop Proceedings, Univ. Wisconsin Sea Grant Pub.
No.HIS-SG-78-234, Madison, WI
Northern Virginia Planning District Commission and the Virginia Polytechnic
Institute and State University. 1978. Occoquan/Four Mile Run Nonpoint
Source Correlation Study: Final Report. Metropolitan Washington Water
Resources Planning Board, Washington, DC.
Northern Virginia Planning District Commission. 19&3. Washington Metropolitan
Area Urban Runoff Demonstration Project: Final Report. Metropolitan
Washington Council of Governments, Washington, DC.
Pollution from Land Use Activities Reference Group. 1978. Environmental
Management Strategy for the Great Lakes System. Final Report to the
International Joint Commission, 115 pp.
Roznowski, D.M. and M.T. Auer. 1984. Tributary loadings to Green Bay: A mass
balance approach. Report to U.S. Environmental Protection Agency, ERL-
Duluth, MN, 74pp.
Tetra Tech, Inc. 1982. Water Quality Assessment: A Screening Procedure for
Toxic and Conventional Pollutants — Part 1. U.S. Environmental Protection
Agency, Report No. EPA-600/6-82-004A, 570 pp.
T-I-C-12
-------
MODELING VERTICAL FLUX OF PESTICIDES WITH CREAMS
R. A. Leonard, Soil Scientist; W. G. Knisel, Research Hydraulic Engineer;
D. A. Still, Agricultural Engineer
U. S. Department of Agriculture
Agricultural Research Service
Southeast Watershed Research Laboratory
Tifton, Georgia
A. W. Johnson, Research Plant Pathologist
U. S. Department of Agriculture
Agricultural Research Service
Nematodes, Weeds, and Crops Research Unit
Tifton, Georgia
ABSTRACT
Mathematical models to assess nonpoint source pollution and evaluate the
effects of management practices are needed to respond to water quality legis-
lation. The USDA model, Giemicals, Runoff, and ^rosion from Agricultural
Management ^ystems (CREAMS), which has "Keen used successfully in modeling pes-
Ticides in surface runoff from field-sized areas was modified to simulate pes-
ticide flux and persistence in the soil plant root zone. The modified version
of CREAMS uses the daily rainfall option, and the soil erosion submodels in
the original CREAMS. The modified version considers pesticide adsorption
characteristics, degradation rates, and soil characteristics by horizon. Pes-
ticide distribution in the root zone is computed with time and depth, as well
as the quantities leached below the root zone. Pesticide concentrations in
water and sediment fractions of surface runoff are also computed. Up to 10
different pesticides and multiple applications of each can be considered
simultaneously. Preliminary tests comparing measured persistence and movement
of the nematicide fenamiphos (Ethyl 3-methyl-4-(methylthio)phenyl(l-methyleth-
yl)phosphoramidate) with simulation results appears promising.
Utility of the model in long-term simulation was demonstrated. Application of
the model in comparing relative effects of management, climate, pesticide
properties, and soil characteristics on potential movement of pesticides to
groundwater is envisioned.
Keywords: Groundwater pollution, water quality, pesticide leaching, runoff,
erosion
T-I-D-l
-------
INTRODUCTION
Mathematical models to assess nonpoint source pollution and evaluate the
effect of management practices are needed to adequately respond to the water
quality legislation implemented since about 1970. Nonpoint source pollution
from agricultural areas must be assessed, specific problems identified, and
alternative practices developed to reduce or minimize problems. Sediment and
chemical losses from surface runoff and chemical losses from the root zone by
percolation must be addressed. Monitoring every field or farm to measure pol-
lutant movement is impossible, and landowners and farm operators need to know
benefits and costs before they implement alternative practices.
Although models are available for surface runoff and chemical transport, the
burden is upon the model user to select from several potential models the one
that will best represent the conditions, practices, and desired results for
specific problems. The U. S. Environmental Protection Agency has sponsored
development of several models; e.g. PRT (Crawford and Donigian, 1973), ARM
(Donigian and Crawford, 1976a), NPS (Donigian and Crawford, 1976b), culmina-
ting in HSPF (Johanson et al. 1980) which is a comprehensive model with the
capability of continuously simulating the dynamics of river basins and system
response to implementing basin-scale water quality management decisions.
Other field-scale pesticide models were described by Haith and Tubbs (1981)
and Steenhuis and Walter (1980). In 1978, the U. S. Department of Agricul-
ture, Agricultural Research Service, began a national project to develop math-
ematical models for evaluating nonpoint source pollution. The first develop-
ment was the CREAMS model, published in 1980 (Knisel, 1980).
With reports of widespread contamination of groundwater (Pye and Patrick,
1983), concerns have mounted for the protection of this valuable resource.
Agriculture may be an important contributor of organic contaminants such as
pesticides (Cohen et al. 1984; Marti et al. 1985). In most of the reported
incidences of groundwater contamination by pesticides, contamination has
occurred because of combinations of several conditions and prerequisites con-
ducive to pesticide transport. These may have been heavy applications of per-
sistent and mobile compounds applied to soils with little adsorptive capacity,
spills, shallow water tables, and high rainfall with annual aquifer recharge.
As with surface water problems, models are needed to assess the potential for
transport to groundwater and evaluate corrective measures. The PESTANS model
(Enfield et al. 1982) was developed to provide a one-dimensional projection of
vertical chemical movement through the unsaturated zone. The Pesticide Root
Zone Model (PRZM) by Carsel et al. (1984) was developed for evaluating the
movement of pesticides within the root zone and the lower unsaturated zones.
This model incorporates a daily rainfall/runoff method similiar to that in
CREAMS and a numerical solution for the chemical transport equation. The PRZM
model is currently undergoing extensive verification and testing.
The USDA CREAMS model is widely used (Knisel and Svetlosanov, 1982; U. S. Soil
Conservation Service, 1985) and has received rigorous testing and evaluation
by the developers and other scientists (Leonard and Wauchope, 1980; Lorber and
Mulkey, 1982; Nutter et al. 1984). Although CREAMS routed plant nutrients
through the root zone, pesticides were not; leaching only below the surface 10
mm zone was considered.
T-I-D-2
-------
Users of CREAMS have expressed considerable interest in an addition to CREAMS
for pesticide leaching in the root zone. This paper describes such an addi-
tion presently called CREAMS 1.7R. This addition is in keeping with the orig-
inal CREAMS structure and simplicity. A more comprehensive model, CREAMS2, is
currently being tested and will also consider pesticide leaching through the
sol urn (Leonard and Ferreira, 1985).
THE MODIFIED CREAMS MODEL
The original CREAMS model consists of three major components: hydrology,
erosion/sedimentation, and chemistry. The hydrology component estimates run-
off volume and peak rates, infiltration, evapotranspiration, soil water con-
tent, and percolation on a daily basis. If detailed precipitation data are
available, infiltration may be estimated at histogram breakpoints. The ero-
sion component estimates erosion and sediment yield, including particle-size
distribution at the edge of the field. Storm loads and average concentrations
of sediment-associated and dissolved chemicals in the runoff are also esti-
mated as well as nitrate-nitrogen leached from the root zone.
The daily hydrology model of CREAMS accounts for soil water by dividing the
root zone into seven soil layers and estimating flow through each layer using
a storage-routing technique. CREAMS 1.7R uses this daily hydrology option ex-
clusively and routes pesticides through the same seven layers using pass-files
of water content and flux generated from the hydrology model. Plant nutrients
are not simulated in CREAMS 1.7R to simplify input requirements. Also, the
streamlined Soil Conservation Service version of pesticide input files and
their crop rotation aspect (U. S. Soil Conservation Service, 1985) was incor-
porated into CREAMS 1.7R for user convenience. For long-term simulation runs,
the irrigation option of DelVecchio et al . (1983) was added so that adequate
soil moisture was maintained for crop growth without having to manually speci-
fy irrigation days and amounts.
A schematic of the pesticide submodel in CREAMS with the CREAMS 1.7R addition
is given in Fig. 1. Pseudo first-order functions are assumed for pesticide
dissipation from foliage and soil after pesticide application. Pesticide in
the runoff active soil layer is partitioned between the solution phase and the
soil phase by the following relationship:
(CWQ) + (CSM) = A Cp
and
where Cw is pesticide concentration in runoff water; Q is the volume of
water per unit volume of surface active layer; Cs is pesticide concentration
in soil equilibrating with runoff; M is mass of soil per unit volume of active
surface layer; A is an extraction ratio specifying the soil to water ratio in
the extraction zone; Cn is the gravimetric concentration of pesticide resi-
due in the soil; and Kj is the coefficient for partitioning the pesticide
between sediment and water phases. Before runoff computation, a portion of
the pesticide mass is moved below the surface 10 mm by the function:
T-I-D-3
-------
PESTICIDE
APPLICATION
FRACTION ON SOIL
ADD
PREVIOUS RESIDUES
COMPUTE
CONCENTRATION
OF RESIDUE
ADJUST FOR
DOWNWARD MOVEMENT
COMPUTE
AVAILABLE
RESIDUE
FOR STORM
RAINFALL,
RUNOFF, SEDIMENT
(HYDROLOGY AND
EROSION MODELS)
FRACTION ON FOLIAGE
ADD
PREVIOUS RESIDUES
COMPUTE
MASS
OF RESIDUE
WASHOFF
FRACTION
CONCENTRATIONS
IN WATER
AND SEDIMENT AND
TOTAL MASS
Fig. 1. Schematic of CREAMS 1.7R model
T-I-D-4
-------
Z = Z0e
F RF-RO-S "1
[^ UKd (1-p) + p J
where Z is the pesticide mass remaining in the surface layer after infiltra-
tion; Z0, the initial pesticide mass; RF, rainfall; RO, runoff; S, initial
abstraction; D, soil particle density; and p, soil porosity. In routing pes-
ticide through the root zone, the mass (Z0-Z) is added to the adjacent soil
layer below, along with the volume of water percolate. Pesticide concentra-
tions in the water and soil phases of this hypothetical reservoir are then
computed using the relationship Cs = K(jCw. In the model, Kj is calculated
from K
-------
Four replicate soil cores were removed from each plot to depths of 1 m. Each
core was subdivided into 0-10, 10-20, 20-30, 30-40, 40-50, 50-75, and 75-100
cm depth intervals for pesticide analysis. Soil samples were removed at 5
time intervals, spanning a period of 1-40 days after pesticide application.
Groundwater wells were sampled weekly after first determining water table
depth and evacuating to remove stagnant water in the well casing. Rainfall
and other meteorlogical data were available from a station located adjacent to
the study area.
Several pesticides were included in this study, however, only data on fenami-
phos in 1983 will be presented here. Fenamiphos (Ethyl 3-methyl-4-(methyl-
thio)phenyl(l-methylethyl)phosphoramidate) is a systemic nematicide giving
broad spectrum nematode control for a variety of crops. A granular formula-
tion of fenamiphos was applied in 1983 at a rate of 6 kg/ha active ingredient
in 30-cm bands- in each row and incorporated 7.5 cm deep. The 6 kg/ha is
broadcast equivalent, i.e. 6 kg/ha if broadcast uniformly, however, since only
one third of the soil was treated (90-cm row spacing), actual mass applied was
2 kg/ha on an areal basis.
Soil samples (Ca. 300 g) were air dried, thoroughly mixed and stored under re-
frigeration pending analysis. Samples were extracted by shaking for two hours
with 50% chloroethanol. After extraction, fenamiphos was converted to fenami-
phos sulfone by oxidation with magnesium sulfate and potassium permanganate,
partitioned into methylene chloride, and determined by gas chromatography
using a thermionic detector operated in the nitrogen mode. Recoveries by this
method were >95% with detection limits of 0.005 ug/g soil.
Fenamiphos concentrations found in the root zone are plotted in Fig. 2(a) with
respect to time and depth. Points plotted are at midpoints of each sampling
depth interval and are averages of four replicates. Movement of fenamiphos
below the incorporation zone with time after application is indicated; how-
ever, concentrations rarely exceeded 0.005 ug/g below 50 cm. Apparently, most
of the fenamiphos had dissipated by degradation in 34 days.
Using appropriate soils data and measured rainfall/irrigation inputs, fenami-
phos degradation and movement was simulated as shown in Fig. 2(b). However,
no model calibration to the specific site was done. A degradation constant
based on a 14-day half-life was estimated from previous work of Johnson et al.
(1982). A KOC value of 160 was assigned by averaging measurements made in our
laboratory on 20 Coastal Plain soils. Incorporation depth was set at 7.5 cm
and all other parameter values estimated as recommended (Knisel, 1980). In
the model runs, a root zone of 46 cm was assumed and divided into approximate
increments of 0-1, 1-7, 7-15, 15-22, 22-30, 30-38, and 38-45 cm. Computed
concentrations are plotted in Fig. 2(b) and are at the midpoint of each depth
increment with respect to time. Comparing Fig. 2(a) and Fig. 2(b), a reason-
able correspondence between measured and simulated results is measured consid-
ering that depth increments and times were not identical. Model output dates
were chosen to correspond to rainfall events, whereas actual sampling dates
were somewhat arbitrary. The estimated 14-day half-life provided good repre-
sentation of observed fenamiphos persistence. Based on field-collected data,
fenamiphos appeared to be more mobile than simulated by the model. Note con-
centration maxima at lower depths and broadening of the concentration profiles
in the observed vs simulated. However, several variance factors should be
T-i-D-6
-------
PESTICIDE IN SOIL (pg/g)
1234
PESTICIDE IN SOIL (pg/g)
1234
60
50
60 -
(b)
FENAMIPHOS, SIMULATED
DAYS AFTER APPLICATION
O 3
a 6
O 14
• 21
O 40
Fig. 2. Comparison of measured and simulated fenamiphos concentration in
Bonifay sand
T-I-D-7
-------
considered. Numerical dispersion may be created by the sampling and simula-
tion depth increments chosen. The model assumes a constant 7.5 cm incorpora-
tion depth; however, in the field, the incorporation implement cannot be regu-
lated exactly and the depth of incorporation may have varied considerably.
Establishment of row beds at planting and soil subsidence with rainfall also
creates surface roughness and distorts the surface reference.
In model application, a soil organic matter profile of 1.0, 1.0, 0.7, 0.5,
0.3, 0.2, and 0.1% was assumed for the seven depth increments. Organic matter
contents were not available on the specific site, only an approximate value of
£ 1% was available for the surface horizon. Lower organic matter content than
assumed would have contributed to an underprediction of pesticide movement.
During the model simulation period depicted in Fig. 2(b), simulated loss below
the 46 cm root zone was 10.2 g/ha. Computed percolation below the root zone
during this period was 9.6 cm giving an average fenamiphos concentration of
0.011 mg/1 in the percolate. At saturation, this solution concentration plus
that adsorbed at equilibrium by the soil organic matter would give a concen-
tration in soil of 0.007 ug/g on a dry weight basis. The practical limit of
detection for fenamiphos by actual analysis was 0.005 ug/g, so it is obvious
that comparisons are made in a range of uncertainty in the actual data. Using
Fig. 2(a), an interpolated fenamiphos concentration at the 45-cm depth of
0.082 ug/g produces 0.2 mg/1 in the soil solution at saturation. At this con-
centration, if 3.7 cm of water percolated through this plane (predicted by
model simulation), 74.2 g/ha fenamiphos could have been transported. This is
considerably greater than the 10.2 g/ha simulated. However, within the'limits
of detection no fenamiphos was transported below the 55 cm depth. Also, none
was detected in sampling wells or tile outflow. For the entire simulation
period, however, about 10 g/ha could have been leached and would not have been
detected if concentrations remained below 0.005 ug/g soil. Clearly, we are in
a concentration range that does not give a complete test of the model, and
illustrates the difficulty in obtaining field test data for pesticides ranging
from the nonadsorbed to moderately adsorbed (such as fenamiphos). For this
and other reasons, models are best used to make relative comparisons and not
absolute predictions.
An indication of model sensitivity to pesticide Koc is provided for two hypo-
thetical pesticides in Fig. 3. Model runs were performed using identical in-
puts as in simulations depicted in Fig. 2(b) except that in test A, Koc = 500
and in test B, Koc = 5 were assumed. A degradation half-life of 50 days for
both and identical application rates of 1.7 kg/ha were also assumed. Com-
paring these simulations demonstrates that significant movement of very weakly
adsorbed pesticides would have occurred below the root zone under rainfall
patterns observed in 1983. Of the total amount applied, about 31% of pesti-
cide B moved below the root zone compared to 0.01% of pesticide A.
In the fenamiphos simulation for 1983 (Fig. 2[b]), 0.74% of the application
was estimated to move below the 46-cm root zone during the entire year.
Annual losses will, however, be highly dependent on rainfall occurring within
a few weeks after application. Field experiments are usually limited to 1 to
3 years and cannot assess the year-to-year variation expected in leaching
losses. Using daily rainfall records for the period 1958-77 at Tifton, 6A, a
20-year simulation of fenamiphos leaching at the experimental site was per-
T-i-D-8
-------
0.25
PESTICIDE IN SOIL (>ig/g>
0.50 0.75 1.00
1.25
1.50
4Q
50
60
PESTICIDE A. SIMULATED
DAYS AFTER APPLICATION
O 14
• 21
Q 40
K-v • 500
PESTICIDE IN SOIL (>ig/g)
0.25 0.50 0.75 1.00
1.25
DAYS AFTER APPLICATION
- 3
- 6
-O 14
— 21
-Q 40
Knr • 5
60
Fig. 3. Comparison of mobility in Bonifay sand of two hypothetical pesticides
differing only in K . Rainfall and other inputs same as used in
simulation shown
QC.
if Fig.
2.
T-I-D-9
-------
formed, and results are presented in Fig. 4. Annual losses ranged from near
zero to about 5%. The 5% loss occurred in 1973 when 20.3 cm of rainfall oc-
curred within a 14-day period after application. These results demonstrate
the value of long-term simulations in assessing expected leaching below the
plant root zone and serves as a useful tool in extending field observations.
SUMMARY AND CONCLUSIONS
Based on preliminary evaluations, a modification of CREAMS which provides sim-
ulation of pesticide concentrations in the root zone, and masses leached below
the root zone appears promising. Testing and verification efforts will con-
tinue. Model utility in long-term simulations was demonstrated. Model appli-
cations in evaluating relative effects of management decisions, differing soil
and pesticide properties, and weather patterns on potential pesticide loading
to groundwater are envisioned.
REFERENCES
Carsel, R. F., C. N, Smith, L. A. Mulkey, J. D. Dean, and P. Jowise. 1984.
Users Manual for the Pesticide Root Zone Model (PRZM) Release 1. U. S. En-
viron. Prot. Agency. Athens, 6A.
Cohen, S. Z., R. F. Carsel, S. M. Creeger, and 6. G. Enfield. 1984. Poten-
tial for pesticide contamination of groundwater resulting from agricultural
uses. In: R. F. Krueger and J. N. Seiber (eds.). Treatment and Dispersal of
Pesticide Wastes. Amer. Chem. Soc. Washington, DC. pp. 297-325.
Crawford, N. H. and A. S. Donigian. 1973. Pesticide transport and runoff
model for agricultural lands. EPA-660/2-74-013. U. S. Environ. Prot. Agency,
Washington, DC.
DelVecchio, J. R., W. G. Knisel, and V.
irrigation on pollutant loads. Proc.
Amer. Soc. Civil Eng., Jackson, WY. pp.
Crawford.
1983. The impact of
Drainage Conference,
Donigian, A. S., Jr. and N. H.
trients on agricultural lands.
Technology Series. Office of
Agency, Washington, DC. 317 p.
Donigian, A. S., Jr., and N. H.
from the land surface. U. S.
3-76-083.
Enfield, G. G., R. F. Carsel, S. Z.
Approximating pollutant transport to
Haith, D. A. and L. J. Tubbs. 1981.
ricultural nonpoint source pollution
Univ., Ithaca, NY.
A. Ferreira.
Irrigation and
113-123.
1976a. Modeling pesticides and nu-
EPA-600/2-76-043. Environmental Protection
Research and Development, U. S. Environ. Prot.
Crawford. 1976b. Modeling nonpoint pollution
Environ. Prot. Agency, Athens, GA. EPA-600/
Cohen, T. Phan, and D. M. Walters. 1982.
groundwater. Ground Water 20:711-721.
Operational methods for analysis of ag-
. Search: Agriculture 16:20 pp. Cornell
H. Davis.
1980. User's
U. S. Environ.
Manual for the
Prot. Agency,
Johanson, R. C., J. C. Imhoff, and H
Hydroldgic Simulation Program-FORTRAN (HSPF).
Athens, GA. EPA-600/9/80-015.
Johnson, A. W., W. A. Rohde, and W. C. Wright. 1982. Soil distribution of
phenamiphos applied by overhead sprinkler irrigation to control Meloidogyne
incognita on vegetables. Plant Disease 66:489-491.
Knisel, W. G. (ed.). 1980. CREAMS: A Field-Scale Model for Chemicals, Run-
off, and Erosion from Agricultural Management Systems. U. S. Dept. of Agric.,
Sci. and Educ. Adm., Conserv. Research Rep. No. 26.
T-I-D-10
-------
UJ
8
i-
S
!•
z
o
§2
O I
t-
So
FENAMIPHOS INCORPORATED IN
BONIFAY SAND FOR NEMATODE
CONTROL
2 4 6 8 10 12 14 16 18 20
YEAR IN 20-YR. SIMULATION PERIOD
1958 - 77
Fig. 4. Results of 20-year simulation of fenamiphos leaching below 46 cm root
zone in Bonifay sand using rainfall record. 1958-77, at Tifton, GA
T-I-D-ll
-------
Knisel, W. G. and Svetlosanov, V. (eds.). 1982. European and United States
Case Studies in Application of the CREAMS Model. International Institute for
Applied Systems Analysis, Laxenburg, Austria. Collaborative Paper CP-82-S11.
Leonard, R. A. and V. A. Ferreira. 1985. CREAMS2. The chemistry submodel.
Proc. Natural Resource Modeling Symposium. U. S. Dept.of Agric., Agric. Res.
Service. In Press.
Leonard, R. A. and R. D. Wauchope. 1980. The pesticide submodel, ch. 5, pp.
88-112. In: Vol. I. Model documentation, CREAMS: A Field Scale Model for
Chemicals, Runoff, and Erosion from Agricultural Management Systems.W. G.
Knisel (ed.). U. S. Dept. of Agric., Conservation Res. Rep. 26. 643 pp.
Lorber, M. N. and L. A. Mulkey. 1982. An evaluation of three pesticide run-
off loading models. J. Environ. Qua!. 11:519-529.
Marti, L. R., J. De Kanel, and R. C. Dougherty. 1985. Screening for organic
contamination .of groundwater: Ethylene dibromide in Georgia irrigation wells.
Environ. Sci. Tech. jn Press.
Nutter, W. L., T. Tkaus, P. B. Bush, and D. G. Neary. 1984. Simulation of
herbicide concentrations in stormflow from forested watersheds. Water Re-
sources Bulletin 26. In Press.
Pye, V. I. and R. Patrick"!1983. Ground water contamination in the United
States. Science 221:713-718.
Steenhuis, T. S. and M. R. Walter. 1980. Closed form solution for pesticide
loss in runoff water. Trans. ASAE 23:615-620, 628.
U. S. Department of Agriculture. 1972. Soil Conservation Service, SCS Na-
tional Engineering Handbook, Sec. 4, Hydrology, 548 p.
U. S. Soil Conservation Service. 1985. An SCS user's guide for CREAMS: A
Field Scale Model for Chemicals, Runoff, and Erosion from Agricultural'Manage-
ment Systems. Tech. Release No. 72. In Press.
T-I-D-12
-------
NON-POINT POLLUTION ARATFMFNT IN TAMPA BAY
Bernard E. Ross, Professor and hark Ross Research Assistant
Civil Engineering and Mechanics Department •
University of South Florida
Tampa, Florida, USA
ABSTRACT
This paper presents the most recent evaluation of the effect of pollutlonal
loadings upon Tampa Bay.
In the years between 1970 and 1984, the Tampa Bay Region has been the
subject of A major pollution abatement programs. In each case urban runoff
has been evaluated. In three of the programs all point and non-point sources
were evaluated.. Loadings of pollution parameters as determined for the year
2000 in all of the programs displayed a remarkable similarity. The
projections were prepared by divers groups of professionals over a long time
span using totally different techniques. Thus loadings determined for the
year 2000 for the wet and the dry seasons can be accepted as good estimates.
This report presents the methods used to evaluate the effect of the pollutants
upon the water quality of Tampa Bay.
A data file of water quality parameters dating from 1969 through 1983 were
subjected to multiple regression analysis.
An improved water quality model was developed calibrated and verified to
more closely represent the nutrient storages and flows In Tampa Bay. In
particular the effect of phytoplankton growth on light distributions in the bay
were addressed. Field measurements were conducted to determine pertinent
rate coefficients for the model. Benthic processes were found to be dominate
in Hillsborough Bay and some portions of Upper Old Tampa Bay.
Keywords: Tampa Bay, Nonpoint pollution.point sources pollution, Statistical
study, Ecological models.water quality, hydraulic models, benthic
processes, eutrophcation
T-I-E-l
-------
INTRODUCTION:
Tampa Bay is a shallow well-mixed bifurcated estuary on the west central
coast of Florida-see Figure 1. It is beset by all of the problems attending the
rapid development of the area. The University of South Florida (USF), Center
for Mathematical Models has been involved in numerous studies relating to the
water quality of Tampa Bay, dating from 1971 through the present. The 1984
Wasteload Allocation Program (WLA) sponsored by the Florida Department of
Environmental Regulation (FDER) is the occasion of this paper.
From the first major study of a portion of Tampa Bay conducted by the Federal
Water Pollution Control Administration IFWPCA 1969], to the present
studies, the fundamental problems of the bay remain. The consistent
problems are
A Excessive nutrient levels which encourage algae growth,
B. Excessive benthic oxygen demand which produces oxygen
depletion in the benthos and oxygen stratification in the
water column.
C. The gradual disappearance of sea grasses.
The WLA program was designed to address each of these problems in
attempting to allocate whatever resources the bay had to assimilate
non-point and point source loadings. A program was devised to review all
existing data, photographic and archival. Field tests were conducted to
provide rates for the mathematical model, calibration and verification data
were obtained. The existing data and the new data were subjected to
statistical analysis.
This paper concentrates on the tasks performed by USF.
PERTINENT FACTS REGARDING TAMPA BAY
Tampa Bay is approximately 35 miles long (67.3 km), and 12 miles (15.5 km)
wide at the widest point. The bottom area is 8.6 billion square feet (1.14
xlO7 km2). The volume at an average high tide is 116 billion cubic feet (5.6
xlO9 m3). The normal tidal prism is 13.5 billion cubic feet (6.5 x 108 m3).
Tides are semi diurnal. Point source input to the bay is approximately 26.6
mi 11 ion cubic feet per day (1.27 x 106 m3). Wet weather nonpolnt source
input is approximately 260 million cubic feet per day (1.27 x 109 m3).
Pollutants enter Tampa Bay at the 39 locations shown in Figure 2. Pollutant
loads for this study were arrived at by a statistical separation of low and
T-I-E-2
-------
Figure 2
Nutrient Discharge
Points
T-I-E-3
-------
high flows by durations, COM ( 1 984). Runoff from ungaged areas were
determined by use of U. S. Geological Survey empirical relations developed in
the region ,USGS (1983). Parameter concentrations were determined by use
of empirical relations from the same source. Yearly total loads determined in
this study were within a few pounds of those determined in 1971 - for the
year 2000, TBRPC ( 1 973). Agreement of the year 2000 estimates for the
1976 determination and the present prediction was also within a total of
2200 poundsJBRPC ( 1 976). Natural phosphorus loadings are higher for the
Tampa Bay region since this is an area of natural deposits and mining.
STATISTICAL ANALYSIS RESULTS
Samples of waters in Hillsborough Bay subjected to bioassay always indicate
a nitrogen limitation to growth .City of Tampa ( 1 983). Statistical analysis
of the 1 980- 1 98 1 data shows no correlation of chlorophyl a and nitrogen in
Hillsborough Bay but a positive correlation in other parts of the bay. Strong
correlation existed between BODu and light extinction everwhere. A
correlation between chlorophyl a and ortho phosphorus existed everywhere.
The latter finding was surprising inasmuch as phosphorus concentrations are
high in Tampa Bay .
In brief, the statistics showed that during the wet season planktonic growth
for the most part is limited by some factor other than nitrogen or phosphorus,
possibly light. The relation between chlorophyl a and phosphorus is the
result of the spatial proximity of the two since concentrations of phosphorus
were hundreds of times the half saturation concentrations required for
chlorophyl a.
THE HYDRAULIC MODEL USED IN THE STUDY
The hydraulic model used in the study is the one developed at USF in 1970.
The model comprises the vertically integrated equations of motion and
continuity in two dimensions, Ross( 1 984). The basic equations used are
given:
_gDati + x -
dt D dx D dy p dx dx D2
* Y -
dt Ddy D dx p dy dy D2
T-I-E-4
-------
^i*t~«««j««^"" \jf dK"j
'•;—tcs-^ir .rJT.:::
• r-t^r^I^rtlii!.1.1.* .•£.•::**•<
T-I-E-5
-------
Figure 5
Net Velocities
USF
mnpn ant. n.
0-3D02HI-ND PS
IUIIII
HMD
oau
txua
(Licnr
isniLi
OIL
anus
cic
OIF
DIN
lty P«r«wtcr« m Mm U.S f. Eo.logiool Bay Had.I
n) I lions uF povndE in ttn toy
eofc*ntraHon In th» boy
T-I-E-6
-------
OISS.
Basic Equation:
PG
f 1KC
Parameter
ALGAE(A)
CBOD(C)
ORGP(P1) Kp
ORTHO-P(P2) 0
ORGN (N1) KN
AMMONIA-N (N2) f2
NITRATE-N (Nj) 0
Table l Ecological nodel Equations \
Decay Rate(K) Source/Slnk(R)
K0
0
RC
- apUA
onpA
Sp,
" flKCC
Terns:
PG -
P
oc
ON
aOG
°OR
°ON
• grazing rate
« respiration rate • 0.1 p'
- algae settling rate
- local specific growth rate (see below)
» deoxygenation rate
- 0X/(0X+0.01)
- settling rate of CBOD
- reaction rate of Organic P
- settling rate of Organic P
» ratio of phosphorus to biotnass
» benthic source rate of Ortho-P
- hydrolysis rate of Organic N
• settling rate of Organic N
- ratio of nitrogen to bioraass
• nitrification rata
- Ox/(Ox+0.5)
- settling rate of ammonia
• 0.9N,/(0.9N2+0.1N3)
• bentnic source rate of ammonia
» benthic source rate of nitrate
• reaeration rate
- DO saturation
» photosynthic oxygen production rate
» oxygen consumption rate by respiration
• oxygen uptake per unit nitrification » 4,
Supplementary Equations:
U "
where:
M1
KN
Kp
L
iNj+Nj+XjjyvrjT^
a
b
IA
OCL
maximum specific growth rate
half saturation value for inorganic V
half-saturation value for inorganic P
f%(exp(-ab) - exp(-b))
2?718...
photoperlod (fraction of day)
extinction coefficient » fa£ + 0.0S2A
algae-free secchi depth
exp (-KeD)
IA/IS
avg light intensity over daylight hours
saturation light Intensity
11.6101 TJ°-969D-l-673+KowS
wind coefficient for reaeratlon
wind speed
chlorophyll-a concentration • oc^A
• ratio of chTorophyll-£ to bioraass
57
T-I-E-7
-------
Table 2 Calibration Constanta
I B.S7U IkacUntiM
§>Ty "• AVttM^B %Amfft 3cv.J
•
CM. a-iq «OgBl irioKUB !•_) A.U Mr dH/og Ua
O.D994&V
Ai|«nic MBwboBM decmr tK f 9.005 day*
UllJinan bfAMdyclB xat* (R ) fl.fl«» J«y~
w
l«» «te OLJ .. 9.3 W1
iCMBMttllfli SKS *rj S.3'
Ocg. BMutaam flDct±iA) tam i« } o.ttli 2r/tUf
p.oJ
*J D.4 W»
la alv*l MXBKDV (Hjjl P.wi* i
u co ilcQ ^riir** bi_) O.ffli i
a- paaveilca r«r alfal gcoMtft 1«C m«
M imMHini QV Jdgal reontotiim 3.0 09 Qytaf Itoai
T-I-E-8
-------
dt dx dy
These equations expressed in finite difference form are solved explicitly on a
grid representing Tampa Bay. The resulting model was calibrated once in
1970 by the use of sodium f luoroscene dye which was released on target and
photographed repeatedly from the air. The model has been verified in seven
major studies including the study of this report, Ross (1984). Verifications
in elude tide heights and velocities. A graphic comparison of net transports
in Tampa Bay is given in Figures 3 and 4. Figure 3 is the results from the
US6S model of Tampa Bay and Figure 4 is the results from the USF model with
the same input data ,Goodwin and Ross( 1984). Comparison of the results from
these two very different models indicates that major circulations do exist in
Tampa Bay. Net velocities in Tampa bay are shown in Figure 5 ,Ross( 1973).
It has been shown that suspended material tends to congregate in velocity
gyres, Ross(l985).
For this study the driving functions for the model were obtained by fitting the
observed tide data with an astronomical harmonic series, this allowed
extensions beyond the observed period. Hydraulic calculations were
performed for each one minute period for a real time of 14.4 days (a repeating
tidal period). Transports, velocities, depths,and accumulated transports were
recorded for each 10 minute interval. These data are the input for many other
auxllliary models including a new Ecological model developed for this study.
THE ECOLOGICAL MODEL USED IN THIS STUDY
The ecology of Tampa Bay is shown in Figure 6. Instead of the open system
suggested by Lewis (I960), a semi-closed system Including the recycling
from the carnivores is the basis for the new model.
The model is based upon the vertically Integrated equations of conservation
of mass including sources and rate processes. The basic equation and the
applicable rates and parameters are shown in Tables I and 2.
An example of the output of the program is given in Figure 7. The form of the
output shown is of invaluable aid in the calibration and verification process.
The entire cycle is visible on one page. The rates used in the model were set
at default rates as given by Ztson( 1978). A box model created to summarize
the ecologic system (part of the main program) is useful in the calibration
process. Average values of the concentrations were used to adjust
approximate rate coefficients. This can be done in minutes. A long run
shows that other changes may be necessary.
T-I-E-9
-------
During calibration and verification, time varying hydraulic conditions were
utilized in the hydraulic model as discussed. The ecotogic mode) was
initialized close to fact then time varying loads are introduced and the model
run for a 40 day period. For discharges from the 39 sources, independent
tests discovered the dynamic equilibrium time of 40 days for Tampa Bay.
PARTIAL RESULTS OF THE STUDY
Some of the important results of study are summarized in six figures. Figure
8 and Figure 9 show year 2000 with nonpoint source pollution only - present
bottom conditions. Figures 10 and 11 show nonpoint sources with best
practical treatment- present bottom conditions. Figures 12 and 13 show the
results for nonpoint sources with BMP -and benthic uptake reduced by SOX.
CONCLUSIONS :
The six figures from 8 through 13 show that reducing nonpoint sources by 30%
by the use of the Best Management Practice only slightly reduces the spatial
extent of poor water quality, however, slightly reducing the benthic uptake
of nutrients greatly improves water qual i ty.
The computer simulations also verified that the phytoplankton in Hillsborough
Bay are nitrogen limited and that they grow to a density at which they become
light limited. Self shading increases the oxygen demand and detritus
settling. The uptake of nutrients from the bottom contributes to the algae
growth tending to fix the location of the high density. The process spreads
slowly with increased deposition. This is the classic pattern of
eutrophication. While the model does not calculate the vertical distribution
of dissolved oxygen, it is safe to assume that in areas of high benthic demand
, anoxic conditions exist in the lower water column much of the time even
though the average concentrations throughout the water column may meet
minimum standards.
T-I-E-10
-------
- BtMlM •/
»ltod*l! TMJ> MM «ltl ••
CMtrvl* ••* •» n -
«•!•» DO •* ttifk flow
Figure 8
Minimum D.O.
Yow2000
Nonpolnt Sources Only
No BMP
*f
iM*n TM* mt
OMtr»I» «n4 M n
Figure 9
Year 2000
Average Chlorophyl a
Noo point Sources Only
NO BMP
T-I-E-11
-------
- Ceeolta ol
tat CsSelt Year »CCO tdtb
Orbea ecd Aoricultcral t»»
ej fO - nalca £3 at
Fiqure 10
Year 2000
Nonpoint Sources Only
With BMP
Minimum DO
Baetdta of
Orteo end AQtltttlterol
r s4 K, pg .
Dleb Pica
o e»
Figure '11
Average CMoro.o
Yocr2000
NOT point eourcoo only
WithBHP
- Catele* e» Ittyt
sees o* nj. c.i
c^li-iii:
V •' d
Figure 12
Minimum DO •
Nonpoint Sources only
With Benthf c demand
Reduced
Day
o*
to M
CaAnod Bcatble
« a« Qlfib Pica
Figure 13
Average Chloro. a
Nonpoint Sources Only
With BonthfcOomend
Reduced
T-I-E-12
-------
REFERENCES
1. Rates, Constants, and Kinetics Formulations in Surface water Quality
Models. Zison, S.W., W.B. Mills, D. Deimer, C.W. Chen of Tetra
Tech, Inc. for USEPA (ORD) EPA-600/3-78-105; December 1978.
2. "Results of Phytoplankton Bioassay Experiments". City of Tampa
Department of Sanitary Sewers, June 1983.
3. "Tributary Streamflows and Pollutant Loadings Delivered to Tampa
Bay". Camp, Dresser & McKee, Inc. January 1981.
1. "A New Geographies Description of the Boundaries and Subdivisions of
Tampa Bay". Lewis, R.R. Ill and R.L. Whitman. BASIS Symposium,
May, 1982.
5. "Courtney Campbell Causeway Tidal Flushing Study". TBRPC with B.E.
Ross and M.W. Anderson for FOOT. 69pp. February, 1983.
6. "Problems and Management of Water Quality in Hillsborough Bay,
Florida". Hills. Bay Technical Assistance Project, SE Region
FWPCA, December, 1969.
7. "Water Quality Characteristics of Urban Runoff and Estimates of
Annual Loads in The Tampa Bay Area, Florida 1975-80. USGS, 1981.
8. "Area Wide Water Quality Management Plan". TBRPC, 1976.
9. "Water Quality Management Plan Tampa Bay Basin". TBRPC, 1973.
10. "Waste Load Allocation". Vol. I - Hydraulic Model, B.E. Ross, 1981.
11. "Waste Load Allocation". Vol. Ill - Model Validation, B.E. Ross,
1981.
12. "Circulation Patterns in Tampa Bay". B.E. Ross, 1985.
T-I-E-13
-------
POLLUTION ABATEMENT IN THE RUHR RIVER BASIN,
WEST GERMANY
D.R. Albrecht and K.R. Imhoff
1. The Ruhr System
Nonpoint pollution abatement has been an integrated part of the
comprehensive water management in the Ruhr River basin since
72 years. Innovative approaches to water management in this area
developed over seven decades ago as increasing pressure on a very
limited supply of water by the massive urban-industrial complex in
the Ruhr Industrial Area required decisive action to resolve severe
water related problems. Special legislation created two organizations
which were empowered, respectively, to care for water quantity and
water quality. A joint directorship of the two was formed 47 years ago,
thus institutionalizing the fact that management of water quantity and
water quality are inextricably intertwined. The Ruhr River water
management associations are characterized by compulsory membership
for all major water users and suppliers, i.e. communities, industries
and water works. A comprehensive water management in the Ruhr
catchment is all the more necessary because approximately one-third
of the average seasonal low flow is exported as drinking water to other
river basins in the industrialized area (Fig. 1).
Northrhine-Westphalia is one of the most industrialized states in West
Germany. In the industrial district - known under the name "Ruhr District" -
water consumption and the volume of wastewater are seven times as high per
km2 as in the average of West Germany. The principal source of water
supply in the district is the Ruhr River which at present has to meet
70 % of the whole demand. The Ruhr River has a drainage area of
4,488 km2. It flows at the southern border in parallel to the large urban-
industrialized district from east to west. The long term average precipitation
varies within the river basin from 800 to 1,400 mm per year. Due to seasonal
factors, the river flow varies between a minimum of 4 m3/s and a maximum
of 2,200 m3/s. The present permanent export rate of drinking water
abstracted from the Ruhr basin is 12 m3/s. Thus, in dry periods a
T-I-F-1
-------
To supply cities and industries %
annually 410 million m3
are pumped out of the Ruhr Valley
320 Mill m3 to the Emscher
82 Mill m3 to the Lippe
6 Mill m3 to the Wupper
and 2 Mill m3 to the Ems Basin
c±i Water Works
Fig. 1: Water export from the Ruhr to adjacent river basins
considerable water deficit must be compensated for by the operation of a
system of reservoirs. In total a volume of 471 million m3 of storage capacity
exists to augment low flow.
Besides providing sufficient water for water supply the Ruhr River also
has the task of carrying the wastewater of about 2 million people living
in its drainage area and of various industries located in the basin. Thus,
effective wastewater treatment is necessary. The Ruhr River Association
operates a system of 118 treatment works, 40 of which are vested with
tertiary treatment by polishing lagoons.
One of the first tasks of the association was to assist the municipalities
to design and construct sewer systems, in order to improve their hygenic
conditions. At this stage the former nonpoint pollution of communities
was changed into point pollution for the major rivers. In many cases
smaller tributaries of the Ruhr were integrated into the sewerage system.
The following period was characterized by the increasing control of point
T-I-F-2
-------
pollution and increasing degrees of wastewater treatment. At
present, the major load of point pollution is controlled by treatment
efficiencies of more than 90 % in respect to BOD. In this situation the re-
maining nonpoint pollution has become the largest residual fraction of
organic load to rivers. Sewage and unpolluted water of creeks are
presently separated and many tributaries in the cities are now "renaturized".
2. Purification Lakes
According to the innovative design of the first director of the Ruhr River
Association, Dr. Karl Imhoff, a system of five river impoundments has
been constructed in addition to the system of treatment works to compensate
for residual pollution from treatment plants and nonpoint pollution above
major water works intakes [1], (Fig. 2). Both concepts, polishing lagoons
and river impoundments provide facilities for sedimentation and extended
biological activities. River water polishing by impounded lakes has proven
to be a feasible measure to improve river water quality and to control
nonpoint pollution. The always available large amount of fresh water has
compensated for the considerable oxygen demand of storm water runoff at
any time. In contrast to other German rivers, a fish kill was never observed
in the impounded Ruhr after heavy rainfalls.
Dortmund
0 A 8 12 16 20km
Impounded
Lake
Hengstey
Harkort
Kemnade
Balden ey
Kettwig
year of
impounding
1929
1930
1979
1932
1949
average
depth
(m )
1,90
2,30
2,40
3,25
2,60
surface
(km2)
1,70
1,38
1,25
2,60
0,55
capacity in
the year of
impounding
(mio. m3)
3,3
3,2
3,0
8,7
M
N
Fig. 2: River impoundments in the Ruhr valley
T-I-F-3
-------
The selfpurification effect in river impoundments is mainly provided
by sedimentation. In Lake Hengstey 100,000 m3 of sediment have
accumulated and in Lake Harkort some 350,000 m3. During 50 years
the volume of Lake Baldeney has diminished by 1.5 x 106 m3. Subsidence
caused by coal mining until 1970 is estimated to be of the same order.
Consequently, some 3 x 106 m3 have been accumulated, resulting in a
sedimentation rate of 60,000 m3 per year [2]. This rate depends on the
flow characteristic of the river and on: size and morphology of the
impoundment. For example, in Lake Baldeney maximum sediment accumulation
in terms of m3 per day is observed at flows of the order of three times
long term average flow, despite the then considerable turbulence. At
lower flows circumstances for sedimentation are more favourable, but the
input of settleable solids is then limited. At flows equal to ten times long
term average, deposits start to be mobilized, and only flows exceeding
15 times average flow flush deposits partly into downstream reaches.
Table 1 indicates analytical data of the sediments in Ruhr lakes.
Table 1: Average composition of sediments in the Ruhr lakes
Parameter
Dry solids
volatile solids
Fe
Cu
Zn
Ni
Cr
Pb
Cd
POn
N total
Dimension
%
% of dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
mg/kg dry sol.
Pur
Hengstey
39.4
12.3
69,200
1,240
4,030
261
440
450
18
14,400
3,750
ification La
Harkort
45.8
11.0
59,900
830
3,120
250
280
480
29
11,110
2,840
ke
•
Baldeney
45.6
14.0
50,100
730
3,540
312
400
525
36
17,700
6,820
A detailed research program on the influence of the largest Ruhr River
impoundment, Lake Baldeney, on the river water quality was performed
from 1979 throughout 1982 [3]. Some results will be described below:
T-I-F-4
-------
On average, 30,000 tons of suspended solids are retained
annually in Lake Baldeney, which is 20 % to 50 I of the input.
The average difference between BOD5-input and output is
2,200 tons per year, being equivalent to a purification capacity
of 100,000 population equivalents. 20 % to 30 % of the BOD5
input is retained. Fig. 3 represents the decrease of BOD5-load
in dependency of flow, distinguished for winter and summer
half years.
10 20 50 100 200 500 1000
Flow (m3/s)
Fig. 3: Impounded Lake Baldeney: percentual decrease
of BODs-load in dependency of flow
By evaluating 284 influent and effluent samples a detailed
phosphorus balance was performed. Fig. 4 indicates the
retention of the P-load in dependency of flow. At summer
low flows 22 % or 70 tons of dissolved phosphorus are retained
by algal development and subsequent sedimentation of phyto-
plankton or grazing zooplankton. During the winter half year
16 % or 40 tons of dissolved phosphorus are eliminated. In
addition, 170 tons of phosphorus is settled per year with solid
particles.
T-I-F-5
-------
50
100
Flow (m3/s)
150
Fig. 4:
Decrease of dissolved phosphorus in the
impounded Lake Baldeney
Due to geochemical reasons and because of a considerable
number of metal finishing factories in the Ruhr basin, the
heavy metal content of the sediments is ten times as high as usual.
The average elimination rates for heavy metals in Lake Baldeney
are as follows:
Element
Fe Pb Cr Cu Cd Zn Ni
Elimination
rate %
50 46 38 33 33 31 10
A separate investigation on origin and fate of heavy metals
in the Ruhr basin has indicated that a fraction of 41 % of
the total heavy metals in Ruhr water is of geochemical origin,
i.e. from nonpoint sources (Fig. 5) [4],
T-I-F-6
-------
measured in one year without floods
I nput _ Output+ Retention
Percentage of the
catchment area
;j Discharge at',';
'' river mouth \'<
283 t/a!;
!; Retention in !;
;>sediment and!'
:', subsoil |.
iXiXoxo.1:1:0,
'/Exported with •
'drinking water!
'.over the
; watershed
50 t/a
iX.ov.'ViV.':'
Sum: 483 t/a
Sum: 483t/a
H partly estimated
Fig. 5: Balance of heavy metal loads in the
Ruhr River basin
Bacterial counts are reduced in the lake by agglomeration
at particles and sedimentation to the order of 90 %.
Nitrification occurs with considerable lower rates in the lake
as in free flowing reaches of the river. Oxygen consumption
by nitrification processes is therefore not important in the
impounded reaches of the Ruhr [5].
A potential secondary pollution is created in the impoundments
by mass algal development in summer. Consequently, oxygen
depletions may result and have to be compensated for by instream
aeration [6].
In order to keep sufficient conditions for boating and other recreational
uses a minimum depth between 2 and 2.5 m is required. Sediment deposits
in the Ruhr impoundments are observed annually by echo sounding
monitoring. By hydraulic dredging 1.1 million m3 of sediments were
removed from Lake Baldeney in 1983/84. The total cost including
sedimentation lagoons and land use was 11 million DM. As supernatant water
from the lagoons flows back into the river (Fig. 6) its quality has been analyzed,
T-I-F-7
-------
Essen-Bredeney
boat harbour
Heisingen
sediment removal
tills2.3m depth
Essen-Werden
areas of sediment deposit
fixed sludge pipeline
ditches for
supernatant
water
Essen-Kupferdreh
1 km
Fig. 6: Sediment removal from Lake Baldeney
The following results were obtained:
BOD5 8.4 mg/l
COD 23 mg/l
NH4 8 mg/l
Fe, Mn <0.1 mg/l
Cu, Zn, Ni, Cr, Pb, Cd <0.01 mg/l
Half a year after finishing sludge dredging wild herbs, bushes, and
trees have grown already without any human aid to a height of half
a meter on the surface of the sediment lagoons. Some rare species are
among them. According to German legislation a detailed plan for land-
scaping of the lagoons had to be designed before starting the measure,
but it is doubtful if this will be performed in future, as a new probably
also valuable "substitute" habitate develops.
3. Stormwater Treatment
After completion of the system of biological wastewater treatment plants
and purification lakes stormwater of the mostly combined sewerage systems
T-I-F-8
-------
remained as the now major pollution to rivers. According to a basic study
for the German Federal Government, the load of organic pollution from storm
overflows without any treatment represents on average 12 % of the per capita
organic load of domestic wastewater, which is shown as 0.12 pollution units
(pu) per capita in Fig. 7.
Although the accurate figure is strongly dependent on the particular
circumstances of a drainage area, i.e. topography, size, urban structure,
and others, 0.12 pu/capita give a general idea of the relative order of
magnitude of the waste load resulting from storm water. Fig. 7 indicates
schematically the waste load to rivers which can be reduced by 50 % if in
accordance to the German guideline A 128 [7] storm water flow is retained
by storage and later is discharged to the municipal treatment plants, where
up to twice the dry weather flow is treated biologically. The dimensioning
rules of this guideline consider especially the storage of the highly polluted
first flushing flood and ensure the transport of up to 90 % of the storm water
flow into the treatment plant at critical conditions in the receiving water by
calculating a critical storm water flow rate for any particular case. In
dependency of the ratio between average low flow of the receiving water
and the total wastewater flow including the storm water the critical .rainfall
rate is between 7 and 15 l/(s • ha). It is assumed that herewith 75 % of the
annual organic load from storm water can be removed.
waterwork
sewerage system
for 5,000-1 0,000
inhabitants
untreated
stormwater
0.12 pu/capita
wastewater
treatment plant
stormwater retention and
biological treatment
0.03 pu/capita
0.05 pu/capita
Fig. 7: The new storm water treatment concept results in
more than 50 % load reduction
T-I-F-9
-------
The German Federal Law on Wastewater Charges includes a charge for
storm water discharges in addition to the charge for dry weather
effluent of treatment works. According to State Water Laws the storm
water charge can be reduced or even cancelled if storm water retention
tanks of appropriate size are existent. Thus, a double incentive is
given to construct storm water retention facilities from the financial and
the water pollution control point of view. A new 100 million DM program
to fight this part of pollution has therefore been approved for the Ruhr
basin.
Some of the stringent State requirements were incentive to the development
of sophisticated models to determine the pollution load from storm overflows.
It seems, however, that this ambitious approach has failed because of the
most complex interrelationships. As already the accurate determination of
the discharged quantities from the particular systems is a very difficult
approach, the Ruhrverband has decided to develop a practical hydraulic
model. Recent discussions indicate the tendency of changing sophisticated
State regulations on storm water charges to more simple calculations.
4. Particular Strategies
To avoid detrimental effects for the Ruhr water works from accidences
on highways a detailed field study has been performed in the early sixties [8].
It resulted in design recommendations for retention tanks to be constructed
at highways in the Ruhr valley. The effluent from highways thus passes
in any case a tank where oil and other hazardous substances can be retained and
abstracted before reaching the Ruhr River. The structures are simple and
not equipped with automatical devices. They are designed for a storm water
flow rate of 100 l/(s • ha) and a retention time of 10 minutes. The hydraulic
surface loading is 12 m/h. The tanks are operated by highway maintenance
services and observed by personnel of the Ruhr River Association.
A new French investigation has indicated that from double laned roads
230 to tOO kg COD/(km • a) result. For lead 0.9 to 1.3 kg/(km • a) were
observed. The pollution load from roads is comparable to that of domestic
wastewater related to the surface area. Only few heavy rainfalls may flush
30 % of the annual load into the environment [9].
Other strategies have to be applied in the case of major agricultural districts.
T-I-F-10
-------
It is an aim to educate farmers to save artificial fertilizers, to apply
manure properly, and to avoid erosion for example by appropriate
ploughing in hilly areas, especially in the catchments of reservoirs,
which are even more sensitive to eutrophication as river impoundments.
5. Summary and Conclusions
In the past, nonpoint pollution has been transferred into point pollution
by the construction of sewerage systems. Consequently, this point
pollution was more and more decreased over the decades. The greater
the progress in this area of water pollution control the larger the
percentage of residual nonpoint pollution as a part of the total river load.
For the Ruhr basin a system of river impoundments has proven to be a
feasible measure for river water polishing and nonpoint pollution reduction.
As impoundments on the other hand may create disadvantages by forming
secondary pollution from eutrophication processes, the task remains to
find an optimum between impounded and free flowing reaches in a river
system.
A new storm water treatment concept will lead to further reduction of
residual pollution to rivers.
T-I-F-11
-------
References:
[13
Imhoff, K,
Die Stauseen im Ruhrtal. Wasserkraft und
Wasserwirtschaft, Vol. 26 (1931), pp 85-89.
[2]
Imhoff, K.R,
The Design and Operation of the Purification
Lakes in the Ruhr Valley. Conference Paper No.8,
Annual Conference of The Institute of Water
Pollution Control, September 5 - 8, 1983, England,
Edwards The Printers Ltd., Coventry, England.
[3] Koppe, P.
et al.:
Untersuchungen iiber den Einfluft der Stau-
haltung eines Flusses auf die Wasserbeschaffen-
heit. Forschungsbericht BMFT - FB 02 WT 854,
December 1983.
[4] Imhoff, K.R., Heavy Metals in the Ruhr River and their Budget
Koppe, P. and in the Catchment Area. Progress in Water
Dietz, F.: Technology, Vol. 12, pp. 735 - 749.
[5] Klopp, R. and Abhangigkeit der Ammoniumkonzentration
Koppe, P.: des Ruhrwassers von den hydrologischen
und meteorologischen Bedingungen. Forum
Stadtehygiene (in press).
[6]
[7]
Imhoff, K.R. Instream Aeration in the Ruhr River. Progress
and Albrecht,D. :in Water Technology, 1978, Vol. 10, p. 277.
Abwassertech-
nische Vereini-
gung (ATV):
Arbeitsblatt A 128: Richtlinien fur die Bemessung
und Gestaltung von Regenentlastungen in Misch-
wasserkanalen, 1977.
[8]
Imhoff, K.R.:
Olabscheidebecken fur FernstraBen im Ruhrtal.
Gas- und Wasserfach, Vol. 108 (1967), pp.
43 - 45.
[9] Balades, J.D. Chronic Pollution of Intercity Motorway Runoff
et al.: Waters. Wat. Sci. Tech., Vol 17 (1984), pp. 1165-1174,
T-I-F-12
-------
USING LINEAR OPTIMIZATION FOR URBAN NONPOINT
SOURCE POLLUTION WATER QUALITY MANAGEMENT
J. G. Garland III
Master of Science Civil Engineering Degree Candidate
F. S. Tirsch
Assistant Professor of Civil Engineering
C. A. Markowski
Assistant Professor of Management Information Systems/Decision Science
C. C. Churn
Assistant Professor of Civil Engineering
Old Dominion University
Norfolk, Virginia, USA
ABSTRACT
Urban nonpoint source pollution is a principal cause of water quality
problems in the United States. This pollution can be reduced by implementing
management practices. One factor influencing management practice selection
is the spending level at which diminishing marginal returns occur.
This study demonstrates diminishing marginal returns for spending to reduce
nonpoint source pollution in Hampton, Virginia. A linear optimization program
selected the best combination of management practices to minimize the cost of
nonpoint source pollution control. Management practice options were grass
swale roadways, porous pavers, detention basins, ponds and fertilizer
management. Pollutants evaluated were total nitrogen, total phosphorus,
suspended solids, fecal coliforms and five-day biochemical oxygen demand.
Additionally, a sensitivity analysis was conducted for total phosphorus to
determine how sensitive overall nonpoint source pollution control costs were
to key management practice data used in the linear optimization problem.
Part of the sensitivity analysis examined the impact of using more site-
specific cost data.
Significant diminishing marginal returns occurred above 40.0 percent removal
for nitrogen and 50.0 percent removal for phosphorus, suspended solids and
fecal coliforms. Overall pollution reduction cost was most sensitive to the
pollution removal data for grass swale roadways and ponds. Using more site
specific cost data in the sensitivity analysis lowered the spending level of
diminishing marginal returns.
Keywords; Urban runoff pollution, nonpoint source pollution, linear
optimization, water quality models
T-I-G-1
-------
INTRODUCTION
In a recent United States Environmental Protection Agency report to
Congress on nonpoint source (NFS) pollution, six of ten Agency regions
specified NFS pollution as the principal cause of water quality problems
(USEPA,1984). This report also listed urban NFS pollution as one of the most
cxxnmonly cited problems. Urban NFS pollution presents a special challenge to
water quality managers, because the many available pollution removal and
treatment options {management practices) have cost and pollution removal
characteristics which vary with land-use. This complicates selecting optimum
treatment practices and spending levels. The focus of this study is on a
factor which influences decisions on spending levels — diminishing marginal
returns. Diminishing marginal returns occur at the point where less pollution
removal occurs per dollar spent.
This study's objectives were to find out if diminishing marginal
returns would occur for spending on management practices in Hampton, Virginia
and to evaluate how sensitive the point of diminishing marginal returns was
to the management practice and land-use data in the problem. The sensitivity
of the results to the data was evaluated to determine the importance of data
accuracy. This information is essential to evaluate the reliability of the
results and the effect of data assumptions.
METHODOLOGY
To achieve the diminishing marginal return objective, a linear
optimization problem was created for each pollutant. These problems included
the cost and pollution removal characteristics of every management practice
in every land-use area. The data for the allowable pollution was set at the
desired level and the problem was solved. The problem solution contained the
overall cost of achieving the pollution level and a cost breakout, i.e., a
list of the management practice area recommended in each land-use that
contributed to the overall cost. The allowable pollution was made more
stringent in 10.0% increments from 10.0% to 60.0% pollution reduction. This
systematic adjustment provided cost and management practice data for a range
of pollution removal which was used to analyze diminishing marginal returns.
These results are reported in a graphic presentation of percent pollution
removal and cost. The results are also presented in the form of cost per
percentage pollution removal.
The sensitivity of the problem to the data was evaluated using a
single pollutant. The approach was to change one data element and solve the
linear optimization problem again at each of the pollution removal
increments. The difference between the new (modified data) cost and the
previous cost was the result of the data element change. These differences
are reported for 50.0% pollution removal in the results. Another form of
sensitivity evaluated was the sensitivity of the point of diminishing
marginal returns to the level of detail of the land-use data. In this
analysis a comparison was made between two linear optimization problems for
the same city subbasin. A consolidated problem divided the subbasin into
8 areas and a parcel problem divided the subbasin into 33 smaller areas. The
affect on the level of spending at which diminishing marginal returns occurs
is shown.
Linear optimization has been used as a water quality management tool
in the past. Anderson and Day (1968) minimized the regional operating costs
T-I-G-2
-------
of conventional point source waste treatment with linear optimization.
Revelle et al. (1968) compared linear programming results to dynamic
programming to show both techniques yielded essentially the same result.
Smith and Morris (1969) optimized water-use goals and management options
using point sources as the modified management practice and dissolved oxygen
as the water quality constraint. Watershed Handbook (USEPA,1981) illustrated
how to use linear programming to minimize the cost of reducing point and NPS
pollution flowing into a water body.
Linear optimization model
The linear optimization approach was to minimize the cost of NPS
pollution control subject to land-use and water quality constraints. The
land-use constraints required the sum of all the management practices and the
land where there was no recommended management practice to equal the total
area in each land-use category. Water quality constraints set limits on the
NPS pollution permitted in each land-use area. Without water quality
limitations, no management practices were necessary and all the urban NPS
pollution reached the receiving water. Making the water quality constraints
more stringent, i.e., requiring larger reductions in pollutants, caused the
program to select a land-use area and management practice for implementation
to reduce pollution.
The management practice options were grass swale roadways (GSR),
porous pavers (PP) — a composite management practice including porous
pavement and modular pavement, detention basins (DB), ponds (PO) and
fertilizer management education programs (EM). The land-use categories were
commercial strip, heavy industry, light industry, high density residential,
multi-family residential, central business district, low density residential
and open land.
The objective function (Equation 1) minimized the total cost (C) of
implementing area in each of the management practice options (Xij) with
individual costs (Cij). The management practice options were represented by
subscript j, and the land-use categories were represented by subscript i.
m n
Equation 1. Minimize C = Y" V" C. .X. .
i=L T^l 1-' -1
The land category constraints (Equation 2) required the sum of the
management practices in each land-use area i to equal the area in i (Ai).
Equation 2. ) X.. = A. for i = l,2,...,m
3=0 ^ 1
Each hectare (ha) of management practice Xij was assigned a fixed
amount of NPS pollution that would be generated yearly (Lij). The sum of
LijXij for each land-use area was a water quality constraint equation
(Equation 3). The right side of the water quality constraint (w.q.) was set
at fixed increments. Non-negativity constraints required Xij be greater than
zero (X. . - 0) because negative area was not feasible. A detailed formulation
of the problem is described by Garland (1985).
T-I-G-3
-------
n
Equations. T L. .X..-w.q. for i = l,2,...,m
The result of these equations was a linear programming problem with
40 variables and 16 constraints for each pollutant.
Management practice costs
The assigned cost of each management practice ($/ha) was based on the
management practice cost and the different characteristics of each land-use
area (Table 1). Land-use influenced cost because most management practices
could only be applied to a fraction of the total land-use area. Other land-
use factors which affected cost were length of roadway, percentage of
impervious area and dwelling unit density. These factors influenced GSR, PP
and FM costs respectively.
GSR for example, can only be applied to roadways in certain land-use
areas because they require wider right-of-ways and are not feasible on main
highways with a high number of access roads. This led to the assumption that
GSR were not applicable in commercial strip and central business districts,
and they were only feasible on 20% of industrial land. GSR were not
considered a management practice option for open land, because limited
roadways would only provide very limited overall pollution reduction. The
cost ($/meter) for GSR was obtained from the Hampton Roads Water Quality
Management Agency (1978). The length of roadway in each hectare (m/ha)'for
each land-use was multiplied by the cost ($/m) and feasible area (e.g., 0.20
for industrial land) to calculate the GSR cost per hectare.
Management practice effectiveness
Both the rate of pollution generation and the ability of the
management practice to remove the pollutant for each land-use area were
needed to create the water quality constraint. Pollution loading data was
obtained from a 1978 STORM (Storage Treatment Overflow Runoff Model) analysis
of the Hampton area conducted by the Hampton Roads Water Quality Agency
(1978).
Management practice effectiveness for this linear programming format
was expressed in units of kg pollution loading/ha. Literature values
provided figures for how effective each management practice was, but the
literature did not detail how extensively the management practice could be
applied in each land-use hectare. For example, porous pavers (PP) are 100%
effective at removing surface runoff in a parking lot by infiltration if
properly designed. However, PP are not 100% effective at removing NPS
pollution on a hectare of any land-use category because they can not be
applied to the entire area. The PP treatable percentage of each land-use
category had to be estimated based on the impervious area of the land-use
categories. The effectiveness data is displayed in Table 2. For example, in
Table 2 the pond (PO) management practice in commercial strip land-use will
remove 50% of the total P, SS and FC; 40% of the total N and 20% of the BODj.
Both PO and DB were assumed to be applicable to all of the area in
every land category. Pollution removal effectiveness was based on the
removal rate of three PO systems in the local area examined by Anderson and
T-l-G-4
-------
TABLE 1
MANAGEMENT PRACTICE COST SUMMARY IN 1984 DOLLARS PER HECTARE
Land-Use
Commercial
Strip
Heavy
Industry
Light
Industry
High Density
Residential
Multi-Family
Residential
Low density 5
Residential
Central Bus.
District
Open Land 2
SOURCES:
a
Areaa
(ha)
811
333
249
700
109
,760
67
,663
Hampton
Management Practice
($)
Grass Porous
Swales Pavers
44,702
911 38,935
911 27,399
911 38,935
319 12,256
190
47,587
• • • • • •
h
Pond0
11,577
4,787
4,532
22,519
22,373
15,875
75,582
1,820
Roads Water Quality Management
h
Basin
36,964
15,284
14,472
71,894
71,427
50,679
241,312
5,812
Plan (4)
Fertilizer
Management
• • •
• • •
• • •
141
469
32
• • •
• • •
and
Tourbier and Westmacott (10).
NOTE: Costs include equipment, material, installation, capital and
maintenance factors. For a complete discussion of cost assumptions see
Garland (3). All cost data converted to January 1984 dollars using the
Department of Commerce Composite Cost Index. Infeasible combinations "...".
aThe city-wide land-use data was from Hampton, Virginia. Hampton is
a 10,753, ha city located on a peninsula. ^
Volume.,determined by inflow to area ratio of 53 m /inflow ha for
ponds and 170 m /inflow for detention basins. Price increased by a real
estate factor in each land-use category.
T-I-G-5
-------
TABLE 2
MANAGEMENT PRACTICE POLLUTION REMOVAL EFFECTIVENESS
(Percent)
.Land-
Use Poll.
Comm.
Strip
Heavy
Ind.
Light
Ind.
High
Density
Resid.
Total P
Total N
BOD.
ss-
FC
Total P
Total N
BOD.
SS
FC
Total P
Total N
BOD.
SS
FC
Total P
Total N
BOD.
SS
FC
Mgt.
GS PP
• *
10
13
15
15
15
10
13
15
15
15
50
65
73
77
77
62
62
62
62
62
54
54
54
54
54
38
38
38
38
38
54
54
54
54
54
Practice3
PO DB FM
50
40
20
50
50
50
40
20
50
50
50
40
20
50
50
50
40
20
50
50
80 ..
57 ..
63 ..
70 ..
70 ..
80 ..
57 ..
63 ..
70 ..
70 ..
80 ..
57 ..
63 ..
70 ..
70 ..
80 10
57 10
63 ..
70 ..
70 ..
Land-
Use
Multi-
Family
Resid.
Low
Density
Resid.
Cent.
Bus.
Dist.
Open
Land
Poll.
Total P
Total N
BOD
SS
FC
Total P
Total N
BOD
SS b
FC
Total P
Total N
BOD
SS
FC
Total P
Total N
BOD
SS
FC
Mgt.
GS PP
50
65
73
77
77
50
65
73
77
77
* •
* *
• *
. .
* •
17
17
17
17
17
• •
• *
66
66
66
66
66
* *
• *
Practice
PO DB FM
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
50 80
40 57
20 63
50 70
50 70
10
10
* •
10
10
. .
* *
• *
• •
• *
SOURCE: Anderson et al. (1982).
NOTE: For a complete discussion of effectiveness
assumptions see Garland (1985). Infeasible combinations "..".
GS — grass roadway swales, PP — porous pavers, PO —
ponds, DB — detention basins and FM — fertilizer management.
T-I-G-6
-------
coworkers (1982). These researchers showed PO needed a 53 m /inflow
hectare ratio in order to achieve a 50% SS reduction. A 50% removaKrate for
SS for DB corresponded to a storage volume to inflow ratio of 170 m /inflow
hectare. FC were assumed to behave as particles, because most bacteria are
attached to suspended solids.
FM programs were assumed to be applicable only in the residential
areas since this management practice is directed toward the reduction of P
and N pollution caused by excessive fertilizer application practices. The
effectiveness assigned to this management practice was a 10.0% reduction in
the total N and total P runoff.
RESULTS
The plots of the city-wide data for cost and percentage of pollution
removal are shown in Figures 1, 2 and 3. A visual inspection of the plotted
data revealed linear segments in the relationships between cost and
percentage pollution removal. For BODj., the linear range extended from 10.0%
to 60.0%. Total P, SS and FC appeared linear between 10.0% and 50.0%, and
linear at a new slope between 50.0% and 60.0%. Total N appeared linear from
10.0% to 40.0% and then linear at a new slope above 40.0%.
A best fit line was developed with linear regression and the slope
used to calculate the cost per percentage pollution removed ($/!%). This data
is shown in Table 3. There was a perfect correlation of the data (one) when
the points being measured represented costs from the same set of recommended
management practices. When the correlation dropped slightly, e.g., to 0.9999,
this indicated a region where a new management practice was added to the
treatment set or one of the management practices was no longer recommended.
No feasible solution was possible above 57.0% removal for total N
because none of the management practice options would remove more than 57.0%
of the pollution load for industrial and open land.
BOD- was the most expensive pollutant to remove in the 10.0% to 50.0%
removal range, costing about 1.5 times the removal cost of the other
pollutants. BOD5 cost more to remove due to the low effectiveness of PO.
Because PO were only capable of removing 20.0% of the initial BOD<-, the DB
management practice had to be used in some land categories for removal rates
starting at 20.0%. DB cost between three and four times as much per hectare
as PO creating the high cost for BOD^ pollution removal.
Above 40.0%, total N removal costs also increased due to the switch
in recommended management practices to DB instead of PO and GSR in industrial
areas. For SS, FC and total P, DB were not selected until above 50.0%
pollution removal.
The extremely high cost of total P removal above 50.0% was caused by
the requirement to implement DB in residential areas while all the other
pollutants were still removed by GSR management practice. DB cost from 70 to
270 times the cost of GSR per hectare in residential areas.
As Figure 4 shows, linear segments also occurred in the subbasin
analysis by parcels and by consolidated land-use areas. The first segment was
from 10.0% to 50.0% pollution removal. In this range, the parcel cost per
percentage pollution removed was $47,825/1% removed and the consolidated cost
was $48,679/1% removed. The second segment was 50.0% to 60.0% pollution
removal, where parcel analysis cost was $2,097,985/1% removed and the
consolidated cost was $1,231,235/1% removed.
The subbasin analysis showed that more detailed examination lowered
T-i-G-7
-------
M
O
I
oo
60.0-
500-
40.0-
Percent
Pollution 30.0-
Reduction
20.0-
10.0-
0.0
0
10
BODc
*SS "
20
30
10~5 Dollars
50
60
FIGURE 1. CITY-WIDE POLLUTION REDUCTION COSTS FOR BOD5
AND SS IN 1984 DOLLARS.
60.0-1
50.0-
40.0-
Percent
Pollution 30.0-
Reduction
20.0-
10.0-
0.0
0
30
60
~90~
10~6 Dollars
120
150
180
FIGURE 2. CITY-WIDE POLLUTION REDUCTION COSTS FOR TOTAL
P IN 1984 DOLLARS.
60.0-
50.0-
40.0-
Percent
Pollution 30.0-
Reduction
200-
10.0-
0.0
10
» Total N
FC
20
30
10~6 Dollars
40
50
60
FIGURE 3 CITY-WIDE POLLUTION REDUCTION COSTS FOR FC
AND TOTAL N IN 1984 DOLLARS.
60.0-j
50.0-
40.0-
Percent
Pollution 30.0-
Reduction
20.0
10
0.0
Total P Subbasin
• Consolidated
A Parcels
12
16
20
24
10~6 Dollars
FIGURE 4. SUBBASIN AND PARCEL ANALYSIS POLLUTION
REDUCTION COSTS FOR TOTAL P IN 1984 DOLLARS
-------
TABLE 3
CITY-WIDE DATA ANALYSIS
Pollutant
Pollution
Removal
Range
Cost/Percent
Pollution
Removed
(1984 Dollars)
BOD,-
SS
PC
Total N
Total P
10-60
10-50
50-60
10-50
50-60
10-40
40-60
10-50
50-60
992,603
427,460
1,996,428
429,938
2,006,944
526,510
2,364,271
592,273
12,735,077
TABLE 4
SENSITIVITY ANALYSIS SUMMARY FOR 50% TOTAL P
REMOVAL AND CITY-WIDE LAND-USE DATA
Mgt. Practice/
Variable Change to all Land-use Areas
Total Cost
% Increase
Grass Swale
Roadways
Ponds
Ponds
Open Land
Porous Pavers
Grass Swale
Roadways
Open Land
Fertilizer
Management
Lower effectiveness 10%
Lower effectiveness 10%
Increase cost 10%
Raise area 20% and decrease
other areas fractionally
Increase cost 10%
Increase cost 10%
Raise area 10% and decrease
other areas fractionally
Raise cost 10.0%
Lower effectiveness 10.0%
154.38
19.56
13.57
1.45
0.86
0.62
0.53
0.00
0.00
T-I-G-9
-------
costs at lew pollution removal levels and raised costs above 50% pollution
removal. This effect was caused by using 33 different prices (one for each
parcel for DO and DB) rather than 8 different prices (one average figure for
PO and DB for each land-use category). Below 50.0% pollution removal in the
parcel analysis, some PO and DB parcel prices were inexpensive enough to
cause their selection while they were not selected in the consolidated
problem which used the average figures. Above 50.0% pollution removal, all
of the inexpensive PO and DB parcel area had been used. Additionally, there
were no other management practice options which could achieve the desired
pollution removal. The result was large overall cost increases in the parcel
problem because more expensive PO and DB area was recommended.
The parcel approach had the additional advantage of providing more
information on where to implement the recommended management practices. For
example, the consolidated problem recommended 9.9 ha of GSR in low density
residential area, while the parcel problem specified GSR in low density
residential area in the following parcels: 0.0 ha in parcel A, 3.2 ha in
parcel B, 0.8 ha in parcel C, 5.6 ha in parcel D, 0.3 ha in parcel E. The
disadvantage of the parcel problem was the problem size — 165 possible
management practice and land-use combinations and 66 constraints.
The cost sensitivity analysis results are shown in Table 4. At the
total P removal levels examined, the greatest overall cost change was caused
by changing PO cost. FM was not selected at 50.0% pollution removal, so the
change in FM cost had no impact on overall cost. Below 50.0% pollution
removal, a 10.0% increase in FM cost created a change in overall cost of
approximately 1%, e.g., at 10.0% pollution removal a 0.65% overall cost
change occurred as the result of a 10.0% increase in FM cost. Increasing PO
costs had a greater effect than other cost sensitivity changes because the PO
fraction of the overall cost was greater. For example, the area recommended
for PO in all the land-use categories (4,106 ha) accounts for 62.74% of total
costs at 50.0% pollution removal compared to 6.20% of the overall cost for
the GSR area (6,444 ha).
Decreasing management practice effectiveness caused an increase in
pollution removal costs due to the increase in management practice area which
had to be implemented to achieve the desired pollution reduction. In some
cases the decrease in management practice effectiveness made it necessary to
switch practices. The increases in overall cost caused by the drop in the
management practice effectiveness of PO and GSR (19.56% and 154.38%
respectively) at the 50.0% pollution removal level was caused by a switch in
management practices. For example, at the 50.0% pollution removal level PO
capable of removing 50.0% of the pollution could be implemented on all the
land-use area and achieve the desired pollution reduction. When PO
effectiveness is reduced to 45.0%, another more effective and more expensive
management practice (DB) must be selected in order to achieve 50.0% pollution
removal. When DB was selected to augment PO or GSR at the 50.0% pollution
level the cost increased greatly for each percentage of pollution removal.
An additional sensitivity analysis raised the effectiveness of GSR in
residential areas to 80.0% for total P. The primary reason for this analysis
was to simulate the lack of diminishing marginal returns demonstrated by BOD5
with another pollutant. One of the main differences between BOD^ and the
other pollutants was GSR removal effectiveness. GSR are 73.0% effective for
removing BOD^ in residential areas. Improving the GSR effectiveness to 80.0%
for total P eliminated the need to recommend any other management practices
in the residential land categories. This switch to one management practice
T-I-G-10
-------
for the residential areas did cause a dampening of diminishing marginal
returns between 40.0% and 50.0% pollution removal. The results were similar
to those shown for BOD^.
The fraction or the open land land-use area was examined to see how
sensitive the results were to errors in land-use data collection caused by
using zoning data. An open land area data error would occur if a 10 ha area
zoned commercial strip was in fact 8 ha of commercial strip and 2 ha of open
land.
Increasing open land area increased removal cost in the 10.0% to
50.0% pollution removal range and decreased removal cost above 50.0%. This
occurred because the average cost of treatment of open land below 50.0%
pollution removal exeeded the average cost of treatment in other land-use
areas. However, above 50.0% removal DB and PO, the recommended practices on
open land, were less expensive than the management practices recommended on
other land-use areas.
CONCLUSIONS
Significant diminishing marginal returns did occur for all pollutants
except BOD5 at spending levels between $20 and $28 million. This spending
level is too high to effect Hampton's expected NFS pollution control program,
because it represents approximately 25% of the city's total expenditures.
Pollution removal costs were the sum of the costs of the management
practices in each land-use category. As additional acres of a set of
management practices were added to meet the more rigid water quality
criteria, a linear relationship was formed between cost and pollution
reduction. Diminishing marginal returns occurred when the cost of removing an
additional percentage of pollution increased.
A new set of management practices was selected when the old set would
no longer achieve the desired amount of pollution removal. This occurred at
40.0% pollution removal for total nitrogen and 50.0% removal for total
phosphorus, suspended solids and fecal coliforms. New management practices
were introduced causing a change in the cost and pollution removal
relationship. Each new set of management practices was capable of removing
less pollution per dollar spent than the previous management practice set
thereby creating diminishing marginal returns.
The five-day biochemical oxygen demand cost per percentage removal
remained approximately the same from 10.0% to 60.0% pollution removal levels.
A sensitivity analysis showed this situation was caused primarily by the
assumptions making grass swale roadways inexpensive and highly efficient at
removing biochemical oxygen demand through 73.0% removal levels.
In general, pollution removal effectiveness was the most sensitive
input variable, then management practice cost and then the percentage of open
land. The point of diminishing marginal returns was also affected by the
level of detail of the analysis. The parcel analysis showed more site
specific management practice costs reduced the spending level of diminishing
marginal returns. By using more detailed cost data for ponds and detention
basins, this analysis showed the spending level needed to encounter
diminishing marginal returns for total phosphorus in the subbasin decreased
by $201,436.
While this study focused on spending levels for treating urban
nonpoint source pollution, linear programming is a potentially valuable tool
for selecting the optimum combination of acreage for application in each
T-I-G-11
-------
management practice option. The optimum set of management practices under the
assumptions of the study were similar for all pollutants. Ponds and detention
basins were recommended for commercial strip. Porous pavers were recarmended
for central business districts. For industrial areas, grass swale roadways
were recommended below 10.0% pollution removal. Between 10.0% and 50.0%
pollution removal, a combination of ponds and grass swale roadways was
recommended. Where five-day biochemical oxygen demand was the pollutant,
detention basins were recommended instead of ponds in the 10.0% to 50.0%
pollution removal range to compensate for poor biochemical oxygen demand
removal by ponds. A combination of ponds and detention basins was recommended
for all pollutants above 50.0% pollution reduction. The management practice
recommendation below 50.0% removal for residential areas was grass swale .
roadways. Above 50.0% pollution reduction, a combination of detention basins
and grass swale roadways was recommended. Open land treatment was by ponds
below 50.0% pollution removal and a combination of ponds and detention basins
above 50.0% pollution removal.
REFERENCES
Anderson, G. F., Neilson, B.J.; and Campbell, D.H. (1982). Management Practice
Evaluation for Urban Areas in Hampton Roads Vicinity. Gloucester Point,
Virginia: Virginia Institute of Marine Science.
Anderson, M. W. and Day, H.J. (1968). Regional management of water quality —
a systems approach. J. Water Poll. Cont. Fed.,40, 1679-1687.
Garland, J. G. (1985). "Selecting optimum urban nonpoint source pollution
management practices using linear programming computer modeling." M.S. 'thesis,
Old Dominion University, Virginia.
Hampton Roads Water Quality Agency (1978). Hampton Roads Water Quality
Management Plan. Virginia Beach, Virginia.
Northern Virginia Planning District Commission (1979). Guidebook for Screening
Urban Nonpoint Pollution Management Strategies. Final report prepared for
Metropolitan Washington Council of Governments, Wash., D.C..
Revelle, C. S. (1968). Linear programming applied to water quality management.
Water Resource Research,4, 1-9.
Smith, E. T. and Morris, A. R. (1969). Systems analysis for optimal water
quality manaaement. J. Water Pol1. Cont. Fed..,41, 1635-1646.
Tourbier, J. T. and Westmacott, R. (1981). Water Resources Protection
Technology; A Handbook of Measures to Protect Water Resources in Land
Development. Urban Land Institute, Wash., D.C..
U.S. Environmental Protection Agency (1984). Report to Congress; Nonpoint
Source Pollution in the U.S.. U.S. Government Printing Office, 436-672/879.
U.S. Environmental Protection Agency (1981). Watershed Handbook
EPA-905/9-84-002.
T-I-G-12
-------
DESIGN OF URBAN DETENTION BASINS FOR
J Bryan Ellis
Urban Pollution Research Centre
Middlesex Polytechnic
Queens,. Enfield, EN3 ASF, UK.
ABSTRACT
Engineering effort has been traditionally placed on improvements in the
hydraulic design and management of urban storm drainage detention
facilities with quality control considerations of stormflow performance
receiving relatively little attention. Nevertheless/ the pollutional
impact of such routed flood discharges on receiving streams in urban
catchments can be significant. The paper examines pollutant
settleability and removal efficiencies of flood detention basins and
considers the effects of particle size, pollutant form and timing on trap
efficiency.
It is shown that timing and partitioning of pollutants during stormflow
events are not amenable to the quality control function of detention
basins as best management engineering practice. The significance of
particle size and settling velocities on pollutant removal within storage
basins are demonstrated through sedimentation column tests and
application of trap efficiency formulae. Efficiencies of between 18 to
62% were found for various pollutants although these values were
considerably reduced if adjustments are made to allow for turbulent flow
routing through the basin.
Ke^words^ Urban runoff pollution, pollution retention, trap efficiency,
particle size and contaminant controls, detention basins.
T-I.-H-1
-------
NOMENCLATURE
V = particle setting velocity
L. = basin length
D
v = mean flow velocity in basin
y = basin depth
n = manning roughness coefficient for basin floor
g = gravitational constant
v = shear velocity
n = efficiency function
INTRODUCTION
It is well recognised that urbanisation causes permanent changes in the
storm runoff regime and leads to increases in both the magnitude and
frequency of peak flows discharged to receiving streams. It is equally
clear that urban developments lead to a marked deterioration in the water
quality of these discharged stormflows despite the enhanced volume
available for dilution. (Ellis, 1982).
A common low-cost engineering means of limiting the effects of
development on the hydrological response of a catchment and of
maintaining the predevelopment downstream river flow regime, is to
provide temporary storm storage facilities within the urban catchment.
In effect, a detention pond represents an attempt to replace the natural
storage capacity lost through urban development, although since the
latter was distributed throughout the catchment and the former is located
at a single site, the comparison is not totally appropriate.
The design of such detention basins has been conventionally based on
hydrological and hydraulic criteria with their size and capacity
predetermined from selected peak flow rates for a specified recurrence
probability. The design methods provided by Davis (1963) or Hall and
Hockin (1985) allow the routed peak storm inflows to be delayed and
attentuated through a restricted outlet or overflow capacity.
Unfortunately, the design of what Whipple (1979) has referred to as
dual-purpose detention basins has to resolve the conflicting requirements
of water quantity and quality. There is a limited amount of post-project
appraisal work available to evaluate the hydraulic design performance of
these control structures (ASCE, 1982) but little consideration has been
given to their trap efficiency in water quality terms. It is not known
if the designated storm for any particular basin serves to retain the
pollutant inflow loadings through sedimentation or if the release time is
sufficient to regulate and minimise receiving stream quality impacts.
The concept of a "settleable design storm" expressed in terms of specific
pollutant parameters for flood detention basins remains to be formulated
and tested.
POLLUTANT RETENTION IN FLOOD DETENTION BASINS
Whilst short term retardation and attenuation of urban flood flows by
T-I-H-2
-------
storm detention basins may represent a best management practice in
hydraulic terms, it does not necessarily provide the most effective means
of retaining or reducing associated pollutant flows. Urban detention
basins are normally designed to balance short-duration, high intensity
storm events since these are the cause of the most frequent downstream
flood activity in the receiving stream. These basins release the
detained peak inflows over a period of a few hours at most, although
gated and throttled outlets can retain water for rather longer periods.
However, it can be argued that it is the aggregate effect of storms of a
return period less than that of the design storm for flood control that
is of most importance in terms of water quality. Although the major
flood damage may result from the larger more infrequent storms, the bulk
of the contaminant loading is carried by the small and medium sized storm
events.
In addition, the proportion of pollutants incorporated within the "first
flush" of the design flow that will be detained by the flood storage
basin can be extremely variable. Fig. 1 demonstrates the
runoff-pollutant timing curves for a flood flow of 10 year return period
within a 350 hectare separately sewered urban catchment in Hendon, NW
London.
100
Total
Pollution
load
80
60 -
40
20
• Heavy Metal*
— —Suspended Solid*
Total Nitrogen
Total Phosphate
20 40 bO BO 100
% Total Runoff Volume
Figure 1. Runoff - Pollutant Timing Curves
Metals exhibit the greatest propensity towards a first flush effect and
this might be expected given their known affinity to solids. The •
greatest deviations from peak flow are shown for dissolved constitutuents
although delayed 'last flushes' for solids, cadmium, zinc and
hydrocarbons are not uncommon (Revitt et al, 1982; Morrison et al,
1984) and the random variability of pollutant parameters in urban runoff
is well documented. (US EPA, 1983.) Therefore, the initial two hours of
the first flush design inflows to the urban detention basin can contain
T-I-H-3
-------
between 60 to 80% of the total storm pollutant mass. (Ellis, 1979).
However, the partitioning of pollutants between solid and dissolved
states implies that those which remain in the latter phase will not be
amenable to control through any solids retention and this may
substantially reduce the notional retention percentages quoted. Sizeable
fractions of nitrogen (60 to 75%) and phosphorous (30 to 55%) have been
shown by Grizzard and Randall (1978) to exist in the dissolved phase of
urban stormflows. Runoff studies by Morrison et al (1984) in NH London
and Gothenburg, Sweden have shown that as much as 63%, 77%, 60% and 32%
of total zinc, cadmium, lead and copper respectively are held in the
dissolved or easily exchangeable form. Therefore, the form or speciation
of the pollutant in storm flows needs to be ascertained to adequately
evaluate the retention efficiency and design performance of the detention
basin as a quality control structure.
A further and fundamental complication to the quality control performance
of urban flood detention basins is in the particle size association of
the storm runoff pollutants. Ahtough only some 8 to 25% of the total
solids loading to the basin will be of particle size less than 150
microns, it is this fine size fraction which accounts for much of the
pollution potential. This size range can typically include 30 to 50% of
algal nutrients, 30% of heavy metals, 50 to 60% of hydrocarbons, 25% of
the oxygen demand and 20 to 40% of total coliforms. (Sartor et al, 1974;
Ellis, 1976). Given the low settling velocities of such fine
contaminated particles and the strongly turbulent nature of both inlet
and outlet flows, it is highly probable that such material will not be
retained within the basin but will be routed in suspension directly
through the structural facility. The high flow rates maintained during
flood flows also stimulate disaggregation of organically bound floes in
the stormwater suspension (Ellis et al, 1982) and this further
exacerbates the quality control problem.
Despite the anomalies presented by partitioning and size association,
analysis and appreciation of the pollutant retention efficiency of flood
detention basins must proceed through considerations of sedimentation
rates and sediment related quality parameters.
TRAP EFFICIENCY OF FLOOD DETENTION BASINS
The average concentration of total suspended solids (TSS) in urban storm
runoff varies between 100 and 600 mg I (Ellis, 1982; US EPA, 1983)
although individual low return storm flows can average as much as 1700 mg
I . Settleable solids greater than 150 microns, as determined by
settling through a 1.83m column of 0.3m diameter for 24 hours, comprise
between 60 and 90% of TSS and possess specific gravities ranging between
0.8 and 2.6. The settling velocities of stormwater runoff solids_as
determined under these static conditions varied from 0.001 cm sec~ for
fine silts/clays up to 8.5 cm sec for coarse sand/grit particulate of
1/2 mm. The modal stormwater suspended sediment diameters of 15 to 20
microns (Ellis et al, 1982) and the 150 micron cut-off definipg
settleable solids possess settling velocities of 0.01 cm sec
respectively. These theoretically determined values do require some
adjustment to allow for the relatively turbulent flow regime conditions
obtained in the detention basin. It has been suggested that effective
settling velocities, for the computation of net pollutant removal, may be
as low as one-ninth of those determined under quiescent laboratory
conditions. (Northern Virginia Planning District Commission, 1980.)
T-I-H-4
-------
A number of methods are available to predict the pollutant trap
efficiency of detention basins of which the oldest and simplest is that
of Camp (1945). This method, like the more recent but complex modelling
approach of Ferrara and Hildick-Smith (1982), utilises the geometry and
hydraulics of the basin together with computed settling velocities. The
three parameter sediment removal function of Camp is
r) = function
v y
vsy
1/6'
v n g
and expresses removal as a function of surface loading with a turbulence
function equivalent to V^/V . The removal efficiencies can be read
from the graph (Fig.2) following calculation and insertion of the
appropriate values.
**•' 60
Efficiency
40
0
0.01
0.10
1.0
10.0
Figure 2. Detention Basin Trap Efficiency
Table 1 shows pollutant settleability and trap efficiencies for a number
of test runs and storm events; the latter determined by the Camp method
using data for a small urban detention pond draining 14.9 hectares of the
larger separately sewered, residential catchment in NW London mentioned
earlier.
The tabulated results illustrate the potential utility of the Camp
methodology, which in conjunction with adequate hydraulic criteria can
provide a dual purpose approach to detention basin design. The design
also emphasise the relatively low efficiencies of existing structural
designs in providing a quality control, particularly when subject to
turbulent flow regime conditions. This conclusion is confirmed by the
field observations and from independent studies (Ferrara and Witkowski,
T-I-H-5
-------
TfBLjM
SETTEJBIUTY_ANO.TRJP.EFFlCI|NCIES.OF_OETENTION_B*SIN.£OLLyTANTS
Pollutant
Phosphate
Total
Coliforms
(1)
Imhoff
Settleability
(Average
X Removal)
(3)
2 Hour
Removal as
X of 24 hour
Removal
(3)
X Trap
Efficiency
"•isq..'
(4)
X Trap
Efficiency
(¥.1M/9)
(5)
Observed
X Trap
Efficiency
TSS
BOD 5
Hydrocarbons
Lead
Cadmium
Total
73
32
67
62
26
46
85
34
13
18
30
15
20
60
62
48
60
59
18
41
42
28
35
42
2
10
6-84
(249) - 86
2V - 48
abed
,cde
20 - 60
10 - 40
(16) - 70
(48) -
cd
10 - 78
20 - 70'
abf
7-73
NOTE: Numbers in parentheses indicate negative trap efficiencies; data from following references:-
*Ferrara and Witkowski, 1983
bGrizzard et al, 1982
C0ally et al, 1983
dtfliippte and Hunter, 1981
*EUis et al, 1985
Ferrara «nd Hildick-Smith, 1982
1983; Grizzard et al, 1982; Dally et al, 1983), although some recently
derived modelUng approaches would indicate very much higher trap
efficiencies in respect of TSS. (Amandes and Bedient, 1980; McCuen,
1980.) However, application of the Camp method with settling velocities
for. particle sizes within the range 20 to 60 microns produce extremely
low or event negative trap efficiencies and the validity of this is borne
out from the extreme ranges observed in the field. Therefore, the
particle size association of pollutants is of importance to the removal
efficiency, despite assertions occasionally found in the literature that
pollutant removal shows little correlation with sediment removal (Whipple
and Hunter, 1981; Kuo and Ni, 1984). Nevertheless, the low computed and
observed trap efficiencies would imply that factors other than the
immediate storm hydra lies and pollutant settleability are operating
within the detention basin to determine the resultant outflow quality.
The most obvious of these would be the exchange of pollutants between the
deposited and reducing or anaerobic basal sediments to the overlying
water phase of the detention basin. It is well documented that trace
metals nutrients and hydrocarbons accumulate in detention and flood
storage basin sediments, (Wignington et al, 1983; Hvitved-Jacovbsen et
al, 1984; Ellis et al, 1985) and during storm plug inflows these
pollutants can be released and mobilised from zones of maximum
accumulation such as occur near outlets which experience the longest
periods of standing water. This resuspension and mobilisation phase will
also transfer some of the pollutant load to the dissolved phase,
particularly in the case of pollutants such as cadmium, zinc or nitrate
which have high exchangeable potential. It may be that the short
retention time of materials in the basin during storm events would imply
that substantial transformation of particulate to soluble form is not
likely. However, the high degree of turbulence would facilitate exchange
and in any case this argument could not be applied to 'wet1 basins where
flushing of soluble material generated during the preceeding dry period
T-l-H-6
-------
would inevitable take place.
CONCLUSIONS
Considerations of the potential pollutant removal efficiency of urban
storm detention basins require a coupling of their particle size and
settling velocity characteristics. If this relationship can be
successfully and reliably determined it can be used in conjunction with
hydraulic and geometrical criteria to provide a dual purpose procedure
for detention basin design. The operational quality performance of any
basin, however, will also be influenced by post-design criteria which
depend on complex biogeochemical exchange mechanisms between pollutants
in the accumulated basal sediments and the overlying water phase. Proper
and regular maintenance, including sediment removal, is therefore
essential if the projected quality design limits are to be achieved.
Whilst it may be arguable to conclude that urban flood detention basins
may have only a limited pollutant control function, it must be admitted
that with proper maintenance they can only help improve receiving stream
quality. In addition, they can and do provide substantial intangible
aesthetic, as well as occasional direct recreational community benefits
despite and irrespective of any adverse water quality conditions.
REFERENCES
Amandes, C.B. and Bedient, P.B (1980). "Stormwater detention in
developing watersheds". Journal Environmental Engineering Division4
ASCE, 106, 2, pp. 403-4197
American Society Civil Engineers. (1982). "Proceedjngs^of^Conference^on
Stormwater Detention Facilities". Urban Hater Resources Research
CouncU~~ASCE7~New"York7
Camp., T.R. ((1945). "Sedimentation and the design of settling tanks".
IlSDSSSi J2Q5-£2£ll£j?D-§2£l£iy._PJyiL,§n.9iDe.e.£§z PP 895-936 .
Dally, L.K., Lettermaier, D.P., Surges, S.J. and Benjamin, M.M. (1983).
Technica Report 79, Hater Resources Series, Dept. Civi Engineering,
University of Washington, Seattle.
Davis, L.H. (1963). "The hydraulic design of balancing ponds and river
storage ponds". Chartered_Mun2cJBa_l_Engjneer, 90, pp 1-7.
Ellis, J.B. (1976K "Sediments and water quality of urban Stormwater".
ysifr_§sryj£fs^ 80' PP- 730-734.
Ellis, J.B. (1979). "The nature and sources of urban sediment and their
relation to water quality". In: Hollis, G.E. (Edit): "Man^s^Imgact_on
i!i§-!J^£2i29i£2i_£^£i§_lD_ii!§_yi!$i" PP- 199-217., Geobooks, Norwich.
Ellis, J.B. 7?9827. "Benefits and blights of urban Stormwater quality
control". In: Featherstone, R.E. and James, A. (Edits): "Urban_Drajnage
Systems", pp. 39-51, Pitmans, London.
ElITs7"j.B., Hamilton, R and Roberts, A.H. (1982). "Sedimentary
characteristics of suspensions in London Stormwater". Sedimentary
GeoJ-ogy., pp. 147-154.
ITIis, J.B., Revitt, D.M. and Gavens, A. (1985) "Polyaromatic
hydrocarbons in sediments of an urban catchment". lDi££D3iJ2D§i..:i2y£D2i
§Qyi!2D2§Qi!iJ5D§i2£i£2i_£t£2!iii£y.' ^, (In Press).
Ferrara, R.A. and Hildick-Smith, A. (1982). "A modelling approach for
Stormwater quantity and quality control via detention basins". Water
Resources Bulletin, 18, 6, pp. 957-981.
T-I-H-7
-------
Ferrara, R.A. and Hitkowski, P. (1983). "Stormwater quality
characteristics in detention basins". Journaj._|nvjronjnentaJ._Engjneer2Qg
£iyi2i2D^.£§£l' 109' 2' PP- 428-44.
GrTzzard,~T7j7 and Randall, C.H. (1978). "Occogujn/Four_MUe_Ruj}_Runoff
P.2iiyi22Q_§ly2^" • Report/ Dept. of Civil Engineering, Virginia
Polytechnic Institute, Blacksburg, Virginia.
Grizzard, T.J., Heand, B.L. and Randall, C.H. (1982). "An evaluation of
Stormwater management ponds for the control of urban runoff pollution".
In: Featherstone, R.E. and James, A. (Edits): "Urban_Drainage-S^stems"i.
pp. 135-148, Pitmans, London.
Hall, M.J. and Hock in, D.L. (1980). "§ujde_to_the-des2gn>_of-storage_Bonds
i2£-f i222L£2Dl£2i_iQ-B2£iIy..y£2§Qi;J£9^£i!i£il!!!§D£_§£§2§ri Techni ca I Note
TOO, Construction Industry Research & In?ormation Association, Bristol.
Hvitved-Jacobsen, T., Yousef, Y.A., Hanielista, M.P. and Pearce, D.B.
(1984). "Fate of phosphorous and nitrogen in ponds receiving highway
runoff". §cjence_Totaj._Env2ronment, 33, pp. 259-270.
Kuo, C.Y. and~N:i7"'wTY7""(T9847T'"""Pollutant trap efficiency in a detention
basin". In: Balmer, P., Malmquist, P-A and Sjoberg, A. (Edits): "Urban
§£2£E-££iiD52£" • PP- 21-28, Chalmers University of Technology,
Gothenburg.
McCuen, R.H. (1980). "Hater quality trap efficiency of Stormwater
management basins". Hater_Resources_BuJUetij2, 1/ pp. 15-21.
Morrison, G.M.P., Revitt, D.M., IITis, J.B., Balmer, P., Svensson, G.
(1984). "Variations of dissolved and suspended solid heavy metals
through an urban hydrograph". Environmental Technology Letters^ 7, pp.
313-318. .........
Morrison, G.M.P., Revitt, D.M., Ellis, J.B., Svensson, G., Balmer, P.
(1984). "The physico-chemical speciation of Zn, Cd, Pb and Cu in urban
Stormwater". In: Balmer P., Malmquist, P-A. and Sjoberg, A. (Edits):
"y£25D-§i2£2-2£2iD29S" PP- 989-1000, Chalmers University of Technology,
Gothenburg.
Northern Virginia Planning District Commision. (1980). "Gujdebook_for
§££££DiD9-H£2§D-D2D;E2Di_B2ii!dii2D_!D§D§9£!D£Di-§i£§i£9l£S-i Met ropo I i tan
Hashington Council of Governments, Hashington, D.C.
Revitt, D.M., Ellis, J.B. and Oldfield, F. (1982). "Variation in heavy
metals of Stormwater suspended solids in a separate storm system". In:
Yen, B.C. (Ed it): "U rban_S t or mwa t e £_2uajlit^z>_Management_and_P J. annjng"^
pp. 49-58, Hater Resources~PuBncatTons'"Ltd.7 LittIeton,""Colorado.
Sartor, J.O., Boyd, G.B. and Agardy, F.J. (1974). "Hater pollution
aspects of street surface contaminants". J2y£D2i_y§i££-£2iiyii2D
££^§£J?ii2G/ ^/ PP- 458-467.
US~ EPA~7T 983 ) Fi^na]. _ Regg r t ^of _t h e^Na t |on wide_U r ban^Runof f __Pr og r am .
Vol.1. US Environmental Protection Agency, Hashington, D.C.
Hhipple, H. (1979). Dual-purpose detention basins. Journ^Water
""~
Higinton Pj,RandalcH.andGrzzardTj (1983. "Accumulation
of selected trace metals in soils of urban runoff detention basins".
Hater Resources Bulletin, 19, 5, pp. 709-718.
T-I-H-8
-------
USING THE UNDILUTED EVENT MEAN CONCENTRATION
TO DETERMINE RUNOFF LOADS
Lei and L. Harms and Marsha Smith, Professor and Research Chemist
Department of Civil Engineering,
South Dakota School of Mines and Technology, Rapid City, SD
ABSTRACT
A three year urban runoff study was conducted on Rapid Creek near Rapid
City, SD. A new parameter, the undiluted event mean concentration (UEMC), was
introduced in order to directly relate water quality changes to the quality of
runoff water. Thirty rainfall runoff events were monitored with about 20 data
sets being complete enough to allow the calculation of UEMCs.
Six water quality sampling stations were installed to monitor water quality
changes in Rapid Creek resulting from urban stormwater runoff in the Rapid City
area. Five of the stations were in-stream locations on the creek while the
sixth station was located on the drainage channel from a 2,000 acre (809 ha)
watershed. All sampling sites were gaged for flow. Precipitation for the
sampled events ranged from 0.08" (0.20 cm) to 2.99" (7.59 cm) and the urban
discharge ranged from near zero to about 470 acre-feet (579 x 10^ m^).
Concentrations of instream constituents varied widely throughout the study, with
the majority of the constituents increasing in concentration at the downstream
stations.
The UEMC is calculated for a particular event by subtracting out the
upstream load. To obtain valid UEMCs for various parameters, careful deter-
mination of baseline flow and quality is needed as well as accurate deter-
minations of flow and water quality during the runoff event. A statistical
analysis of UEMCs will be presented for ammonia, chloride, COD, total lead,
total phosphorus, and suspended solids. Results show that data are positively
skewed with an error in the UEMC of about 5 percent.
T-II-A-1
-------
THE STUDY AREA
General
The study area lies within the physical boundaries of Rapid City, South
Dakota. Figure 1 indicates the locations of the six sampling stations. Rapid
City is located at the foothills to the Black Hills in western South Dakota and
is the county seat for Pennington County. Rapid City is the largest community
in western South Dakota and is the business center for much of the surrounding
area.
Between 1940 and 1960, Rapid City experienced a dramatic growth rate
resulting in a 1960 population of 42,399, up from 13,844 in 1940. A more
stable, much slower growth occurred in the next two decades, giving a 1980 cen-
sus figure of 46,492.
Sampling Locations
The six sampling stations are indicated on Figure 1. Ideally, a sampling
station should have been located just upstream of all urban development, and
just downstream from the city. Station 1 essentially satisfies the upstream
station, but some urban influence will manifest itself because of an increase in
density of dwellings and upstream highway traffic. It was not hydraulically
possible to establish a sampling station downstream from the study area before
other non-point sources of runoff would enter Rapid Creek. Consequently, a
theoretical station, No. 56, was developed which mathematically establishes
downstream values based on mixing the loads present at Sta. 5 (the last station
on Rapid Creek) and Sta. 6 (the last major urban drainage into Rapid Creek).
Drainage areas ranged in size from 33,730 acres (13,650 ha) at Sta. No. 1
to 1,610 acres (650 ha) at Sta. No. 5. Land usage varied from 96 percent non-
urban at Sta. No. 1 to 19 percent non-urban at Sta. No. 5. Complete land use
information can be obtained from a document by Harms et al. (1983).
TECHNICS
Field Methods
The majority of the water samples were collected using automated equipment
but some samples were collected manually, primarily at the upstream stations.
Baseline samples were taken at normal flow conditions periodically. Manual
methods were used to obtain these samples. Even with the automated equipment
in-pi ace, every effort was made by the field crews to be on site during the
runoff events.
Manually collected water samples were obtained by wading the streams and
collecting depth integrated aliquots using between 10 to 15 verticals across the
channel. In the first year of the study, manually collected baseline and precip-
itation runoff samples were collected directly into new, one-gallon (3.8 L)
plastic milk containers. During 1981 and 1982, samples were collected by use of
DH-77 depth-integrating samplers. After collection the samples were transferred
to new, one-gallon (3.8 L) plastic milk containers. All sampling equipment and
T-II-A-2
-------
H
I
CO
r
I
1
\ Q-
P ^ i
\ " © r /
^ ^^ ^^_Z C4)
LIME ( ^ ^^
CREEK ^.^ U WATER DQWN ^
Xj^ PLANT TOWN \^
r* r-----xy._.
\ A
\ r \ -' ^^ A
CANYON X^ r--1" r\
CO
*:
o
oc
<
Q.
w IRRIGATION
DITCH
xfe""
RIM ROCK V ^< J \ ,-^~-*~S ,^' \ tx
HKHWAY ^jtf^-S / f— __^ ,' S^
N
— >
\/ w^ ^ -^
r (T^BFISH ' ^^--^^--
^\ /\ HATCHERY ^^
RAPID CREEK
\
LEGEND
DRAINAGE DIRECTION
' SAMPLING STATION
Figure 1. •• Map of Rapid Creek and Major Drainage Basins within Rapid City Proper; Sampling Site Locations.
-------
containers were rinsed three times with native water prior to taking the first
aliquot. The sample was immediately cooled to 4°C. Date, time, and stage were
recorded immediately before and after sample collection.
The automated sampling equipment, referred to as an urban hydrology moni-
toring system by USGS, incorporates a microprocessor based system control unit
to receive a record on site and to control the automated water sampling device.
The system would switch on to storm mode when a select stage, corresponding to a
definite discharge, was reached in-stream during a storm event. The system
would make continuous recording data (time, date, stage, accumulated rainfall,
and sequential sample number) at a pre-determined time interval ranging from 30
seconds to 1 hour. If a rapid rise in stage occurred, the timing sequence was
over ridden and additional samples were collected. Samples were pumped into
new, one-gallon (3.8 L) plastic containers which were stored in a refrigerator
unit, at 4°C, and later, as soon as possible, transported to the South Dakota
School of Mines and Technology (SDSM&T) environmental engineering laboratory.
Under most conditions, samples were delivered to the lab within three hours of
collection. A sketch of this equipment is shown in Figure 2.
Laboratory Methods
All sample preservation and preparation were done in the laboratory. Flow
weighted composites were normally made for each station using the appropriate
aliquot from each discrete sample. Volumes for each aliquot were determined by
calculating the volume under the hydrograph that each sample represented,
baseflow included. A computer program was developed to compute the correct ali-
quots, Harms and Smith (1983).
Composites were produced using a USGS cone splitter, USGS (1980). The
splitter splits any sample into 10 equal aliquots to the nearest one percent.
The cone splitter was evaluated for accuracy prior to use. After some practice
the error averaged about 3 percent which was the sum of the cone splitter and
analytical error. After the composite was properly consituted, the cone
splitter was used to split the composite sample into individual containers for
storage and/or preservation.
A quality assurance (QA) plan was developed and approved by EPA. Standards
(both known and unknown), duplicates, blanks, and spikes were routinely ana-
lyzed. Values which did not satisfy the QA requirements were not entered into
the data base.
UNDILUTED EVENT MEAN CONCENTRATIONS
The undiluted event mean concentration, UEMC, was selected to indicate the
quality of runoff which entered the stream. The UEMC was calculated for each
parameter from laboratory data for a particular event by subtracting out the
portion associated with the baseline flow, as follows:
QBL CBL + QRO UEMC = q EMC
and, _
UEMC = q EMC - qBL
QRO
T-II-A-4
-------
RAIN
GAGE
REMOTE
RAIN
GAGES
Figure 2. •• Automated sampling station.
-------
where,
= Base flow preceeding runoff events, cfs
= Assumed base flow concentration for the
parameter under consideration. Value is
assigned from baseline quality obtained
from samples collected when runoff was not
affecting the quality, mg/L.
= Mean runoff flow, equal to in-stream mean
flow recorded during a runoff event - base
flow (Q - QBL), cfs.
UEMC = Undiluted event mean concentration, a
calculated approximation of the mean con-
centration of a parameter in the runoff,
mg/L.
T) = Mean in-stream flow during a runoff event,
flow weighted; i.e., (Q)(Runoff duration)
Volume of water passing a station during
event, cfs.
EMC = Event mean concentration. Obtained by
laboratory testing of a flow weighted com-
posite sample, mg/L.
(Note: cfs x 2.8317 x 10"2 = m3/s)
A simple example may best illustrate the UEMC:
Given: Station No. 3
Runoff Event No. 5
Base flow = QBL = 86 cfs (2.44 m3/s)
Mean in-stream flow = Q = 110 cfs (3.15 m3/s)
Instream suspended solids = EMC = 60 mg/L
Base flow suspended solids = CBL = 8
To Find: The concentration of suspended solids in
the runoff above Sta. No. 3, UEMC.
UEMC = H EMC - QBL CBL = (110)60-86(8)
_ 110-86
QRO
= 246 mg/L ans.
The above procedure is not applicable at Sta. 56 (the theoretical
downstream location) because the event duration was not exactly the same at
Station 5 and Station 6. The UEMCs for Sta. 56 were calculated using the dura-
tion recorded at each station, as shown on the following page:
T-II-A-6
-------
UEMC56 =| (UEMC)(Vol R0)| 5 + | (UEMC)(Vol ROj 6
Vol
+ Vol
where:
UEMC = as described above, mg/L
Vol RO = Volume of runoff (base flow subtracted)
at each station, acre-feet (1.2335 x 103 m3).
Because of the inherent error in sampling and measuring procedures. Some of
the calculated UEMC values for Stations 1 through 5 yield a negative value.
Obviously this is impossible as a negative 3.5 mg/L of chloride could not be
present in the runoff, for example. Statistically, however, these negative
values are just as valid as positive values and they were not discarded during
subsequent calculations. Generally negative UEMC values were viewed as an indi-
cation that that particular constituent was very low in the runoff of the event
under consideration.
Frequently, only a minor change in the measured concentration or flow would
have produced a positive value. For example, the UEMC for Station 1, Storm 1
shows a -21.5 mg/L for a chloride concentration. If the baseflow used in the
UEMC calculation was in error by only 3% (a very acceptable field measuring
error), the UEMC would calculate to be a positive 14.2 mg/L. Thus, even though
some of the UEMC values compute as negative quantities; they appear to give
reasonable concentrations. Approximately 3.7% of the UEMC values for Sta. 1
through 5 yield negative concentrations.
TABLE 1 — Log Normal mean values for UEMC, mg/L
Sta. No.
1
2
3
4
5
6
56
0
0
0
0
0
0
0
NH3.
N
.26
.12
.09
.12
.11
.18
.15
COD
12
52
91
113
95
190
119
ci-
5.7
4.8
4.3
5.5
5.8
5.5
5.9
TKN
0.56
1.30
1.14
2.42
2.19
3.87
2.56
N03
0.35
0.38
0.35
0.60
0.46
0.52
0.51
Total
P
0.04
0.25
0.29
0.62
0.55
1.48
0.80
Diss.
P
0.01
0.03
0.03
0.06
0.04
0.08
0.05
SS
22
289
350
693
595
2,443
1,038
Tot
Res.
240
533
552
771
621
2,786
1,161
VSS
3
33
35
81
67
228
104
A summary of the log normal mean values for the UEMCs is presented in Table
1. Although mean values can be misleading statistics, the data clearly show an
increase above normal runoff in the stream (Sta. No. 1) as urban runoff enters
Rapid Creek. When values from Sta. No. 5 are compared with the theoretical con-
centrations downstream at Sta. No. 56, the impact of the pollutants from the
Robbinsdale watershed (Sta. No. 6) are very apparent.
A statistical analysis was done for the UEMC data sets for each parameter.
Summaries of the statistics for the computed UEMC values for ammonia and chemi-
cal oxygen demand (COD) are presented in Tables No. 2 and 3. Ammonia is repre-
T-II-A-7
-------
sentative of a dissolved species and shows some negative values while the COD is
representative of a parameter associated with solids in suspension and does not
show any negative values.
TABLE 2 -- Summary of UEMC for Ammonia-Nitrogen
Parameter
Mean, mg/L
Log mean, mg/L
Median, mg/L
Min., mg/L
Max., mg/L
No. samples
No. negative
values
1
0.60
0.26
0.26
-0.11
3.80
15
2
2
0.36
0.12
0.19
0.02
1.96
18
0
3
0.18
0.09
0.16
0.02
0.78
19
0
Station
4
0.23
0.12
0.18
0.02
1.32
19
0
5
0.22
0.11
0.13
0.02
0.62
21
0
6
0.26
0.18
0.16
-0.02
1.00
20
2
7
0.22
0.15
0.13
0.26
0.61
19
0
TABLE 3 « Summary of UEMC for COD
Parameter
1
Station
4
Mean, mg/L 53 69 102 137 112 214 135
Log mean, mg/L 12 52 91 113 95 190 119
Median, mg/L 2.5 50 78 106 85 171 100
Min., mg/L 2.5 11 47 33 21 87 51
Max., mg/L 359 276 251 377 339 599 367
No. samples 15 17 19 18 21 21 20
No. negative
values 0 0 00 00 0
Runoff Loads
Rainfall runoff loads for each event were calculated by using the UEMC for
each parameter. The UEMC was multiplied by the volume of runoff to give the
total load to the stream in pounds (0.454 kg).
Obviously, there is a dramatic increase in material entering Rapid Creek as
the stream moves through Rapid City. Table 4 shows these increases as a percen-
tage of the load entering Rapid City (at Sta. No. 1). The higher percentages
are associated with those constituents which fluctuate with suspended solids,
the most predominate being total lead. Lead was almost entirely tied to
particulates, and very little dissolved lead was detected.
T-II-A-8
-------
TABLE 4 - Increase in Runoff Loads
Parameter
Percent Increase
Sta. 1 to Sta. 56
Parameter
Percent Increase
Sta. 1 to Sta. 56
Ammonia-Nitrogen
COD
Chloride
Total Kjeldahl N
Total Lead
Nitrate + Nitrite
633
24,400
12,900
2,530
104,000
2,300
Total P
Dissolved P
Sodium
Suspended Solids
Volatile SS
Total Residue
14,000
800
5,980
63,500
61,400
4,730
An important consideration is that these loads are deposited in the stream
within a relatively short time-frame. Downstream from the Rapid City urban
area is the municipal wastewater treatment facility. Assuming that the facil-
ity treats its sewage to normal levels, the runoff loads show that it would
take three months for the treatment plant to discharge the same amount of
solids that a moderate storm event would cause to be washed into the creek. It
would take almost three years for the effluent to contribute solids equal to
those deposited in Rapid Creek in one day by a 2.5 inch (6.35 cm) rain (Event
No. 21).
CONCLUSION
If used in conjunction with accurate flow measurements and representative
water samples, the undiluted event mean concentration is an effective parameter
for computing runoff loads.
REFERENCES
Harms, L.L., Smith, M. and Goddard, K. (1983). Urban Runoff Control in Rapid
City, South Dakota. Sixth District Council of Local Governments, Rapid City,
SD.
U.S. Geological Survey (1980). Equipment and Supplies - New Sample Splitter
for Water Quality Samples, Technical Memorandum No. 80.17, USGS, Reston, VA.
Harms, L.L. and Smith, H., Jr. (1983). A compositing Program for Water Quality
Sampling. Water Engr. and Management. 130, 12, pp. 39-40.
T-II-A-9
-------
UET WEATHER IN-STREAM WATER QUALITY EVALUATION
John J. Warwick, Ph.D., P.E.
Assistant Professor of Environmental Sciences
The University of Texas at Dallas
P.O. Box 830688, Mail Station BE-22
Richardson, Texas 75083-0688
Contributing Authors: David Getz and David Kibler
ABSTRACT
The magnitude and duration of non-point source discharges are stochastic in
nature. For this reason, in-stream water quality standards should be based
upon an acceptable frequency of violation rather than an absolute level which
should never be exceeded. The effect of various pollution abatement
methodologies could then be evaluated based on their ability to meet the
prescribed water quality criteria and associated acceptable frequency of
violation. This type of approach requires characterization of receiving water
responses to a large number of storm events.
A project was conducted for the city of Altoona, Pennsylvania, to quantify the
degradation in water quality of the Little Juniata River as a result of
combined sewer overflows. Dissolved oxygen was chosen as the principal
indicator of in-stream water quality. Runoff flows and associated biochemical
oxygen demand (BOO) loadings were calculated for the urban and surrounding
rural areas by the Army Corps of Engineers' Storage, Treatment, Overflow,
Runoff Model (STORM). A three year period from 1975 through 1977, containing
209 individual storms events, was chosen to characterize the statistical
behavior of in-stream dissolved oxygen concentrations.
A one-dimensional, steady state, dissolved oxygen model was utilized to
calculate spacial variation of dissolved oxygen concentration. Model input
included STORM output (first flush flow and BOD loadings) for each storm
event. First flush values were chosen to simulate worst possible conditions.
The dissolved oxygen model calculated the critical or minimum dissolved oxygen
value occurring within the river study area for each storm. A cumulative
frequency distribution of minimum in-stream dissolved oxygen values was then
calculated via a simple ranking procedure.
Results indicate that the seasonal operation of CSO facilities may be a viable
option, with real economic savings. The impact of supplying various levels of
detention storage for the urbanized area of Altoona was also investigated.
Diminishing returns (smaller decreases in the frequency of violation) can be
expected with increasing levels of detention storage.
Keywords: Urban runoff pol 1ution, nonpoint pol1ution, nonpoint pol 1ution
abatement, stormwater quality, stormwater models, water quality models
T-II-B-1
-------
INTRODUCTION
Combined sewer overflow (CSO) discharges pose a serious threat to
receiving quality in many urban drainage basins. The combined discharge of
untreated sanitary "dry weather" flow and urban runoff can introduce
significant amounts of pollutants, resulting in severe degradation of
receiving water quality.
In 1981, the Pennsylvania Department of Environmental Resources (DER)
directed the City Authority of Altoona to analyze the effects of combined
sewer overflows in the wastewater management plan being developed for the
city. Altoona, a city of approximately 60,000 people, serves as the headwater
region for two rivers: the Little Juniata River and the Beaverdam-Frankstown
Branch of the Juniata River. The respective watershed areas are 343 and 395
square miles. Water quality in the two rivers is very sensitive to the impact
of CSO discharges since the discharges constitute a significant portion of the
total flow in upstream reaches of the rivers during periods of heavy rainfal 1
or snowmelt.
This report will focus on the study performed in the Little Juniata
River. The study was undertaken to satisfy two principal objectives: 1) to
estimate the frequency of violation of dissolved oxygen standards as a result
of existing combined sewer overflows; and 2) to evaluate the potential
effectiveness of alternative CSO abatement measures. Mathematical models were
used to simulate the generation of CSO discharges in Altoona and their effect
on in-stream dissolved oxygen concentration. The models were used to-eval uate
abatement measures such as expanded sewage treatment plant capacity and the
utilization of stormwater detention basins. The effectiveness of an abatement
measure was expressed in two interrelated ways: 1) by its ability to reduce
the biochemical oxygen demand (BOD) loadings attributed to combined sewer
overflows; and 2) by its ability to reduce the frequency of violation of
established dissolved oxygen standards during wet-weather events.
MODEL FORMULATION
The mathematical modeling effort was divided into two phases: a land
phase and an in-stream phase. In the land phase, the Army Corps of Engineers'
Storage Treatment, Overflow, Runoff Model (STORM) was used (6). STORM
performs a continuous simulation of the rainfal 1-runoff and snowmelt-runoff
processes on the land surface. It calculates the quantity of runoff generated
in a drainage basin and the quality of the runoff in terms of 5-day BOD
concentrations. The STORM model was modified in two minor areas: (1) to
better represent street cleaning operations, changes in the sweeping
efficiency function were made; and (2) the closure action of shear-gates at
each diversion structure was incorporated in the model. The STORM model is
wel 1-documented for its runoff and pollutant loading calculations and no
further discussion is provided.
The second model used was the Distributed Oxygen Analysis Program (DOAP),
a model developed specifically for this project. In this model, the f 1 ow
generated by STORM is mixed with the background flow in the river, and the
total flow is tracked downstream. The model calculates in-stream dissolved
oxygen concentrations along the length of the river using steady-state
equations that perform an oxygen balance. The oxygen balance calculations
T-II-B-2
-------
provide a simulation of the consumption of in-stream oxygen by the bacterial
breakdown of organic material present in the combined sewer overflow and other
waste inputs.
The oxygen mass bal ance equation used in the stream model is the wel 1 -
known steady-state form:
0 = .2.L [expjt . S.i expgt] + DQ expgt (1)
k2-K1 S2
where:
U2
j = - [1 - Sj] (2)
2E
U2
g = -- [i - s2] (3)
2E
4 k,E
- 1 + --«- (4)
U2
4 k2E
and So = 1 + ----- (5)
U2
The variables present in Eq. (1) to (5) are defined as: D = resultant DO
deficit (mg/1); D0 = initial DO deficit (mg/1); E = longitudinal dispersion
coefficient (mi 1es2/hour); kj = deoxygenation coefficient (I/hour); ko =
reaeration coefficient (I/hour); L0 = ultimate BOD (mg/1); t = elapsed flow
time (hours); and U = flow velocity (miles/hour).
The effects of photosynthesis, benthal oxygen demand, and plant
respiration are not included in Eq. (1). It is expected that little
photosynthetic activity occurs during runoff events due to cloudy weather and
turbid water conditions. Similarly, plant respiration may have an
insignificant effect because of the increased volume of flow in the river.
The potential effect of benthal oxygen demand on DO concentrations can be
quite significant, however. During wet weather, high flow velocities can act
to scour sediements that can include 1arge quantities of organic material.
This phenomenon and its corresponding sediment oxygen demand are not
represented in the present study. There is no evidence that benthal oxygen
demand plays a significant role in depleting the DO in the Little Juniata
River.
T-II-B-3
-------
Controlling parameters in the DOAP model are: (1) ki, the in-stream
deoxygenation coefficient; and, (2) k2, the reaeration coefficient. Because
all DO depletion was assumed to take place through biological oxidation, the
deoxygenation coefficient, k]_, was a sensitive parameter. It was estimated as
a function of flow rate using a formulation by Wright and McDonnell (7):
kl - \ Q Y2 (6)
where kj = in-stream deoxygenation coefficient at 20° C (I/day); Q = steady
state flow rate (cfs); Y]_ and Y2 are empirical coefficients to be evaluated by
calibration against observed DO data from in-stream measurements during a
storm. These were primary parameters in fitting the model to the Juniata
River system.
The reaeration coefficient, k2, was estimated by use of an equation
developed by Owens et al . (5) for shallow streams:
21.7 U°'67
(7)
where ko = reaeration coefficient at 20°C (I/day); U = mean stream velocity
(ft/sec); and H = mean stream depth (ft). Both k} and k2 were adjusted for
temperature using standard correction formulas with 9 values of 1.068 and
1.024, respectively.
Since many inflows occur along the rivers, the in-stream flow rate,
velocity and depth will vary greatly with distance. In the case of a
headwater receiving stream, this variation can be several times greater than
the base discharge during dry weather periods. Consequently, it is important
to represent the spatial distribution of flows and pollutant discharges to the
headwater receiving stream. In the present study, this has been accomplished
by developing depth discharge and velocity-discharge relations for each reach.
With changes in flow rate at each inflow point, the estimates of mean flow
depth and velocity are updated. These revised stream values in turn are used
to re-calculate kj, k2 and the flow time for each sub- reach (distance between
inflows). A sub-area definition sketch is shown in Figure 1 for the 343
square mile Little Juniata River portion of the total Juniata River watershed.
FREQUENCY ANALYSIS
The strategy of analyzing the frequency of occurrence of low DO levels to
evaluate CSO abatement alternatives has been introduced previously by Medina
(3). The DO frequency analysis performed in DOAP is possible due to the
model's ability to generate DO profiles for a large number of separate runoff
events. One DO value from each event is used in the frequency analysis. This
is the minimum in-stream DO concentration present, irrespective of its
location in the river. Thus, the magnitude of the DO sag is considered, but
its location is not. The minimum DO value from each event is then stored for
subsequent frequency analysis. This set of DO minimums is then ordered and
T-II-B-4
-------
H
I
Figure 1: Little Juniata sub-area definition sketch,
-------
assigned a plotting position using the formula: P = lOOm/N, where P is the
percentage of events exceeding a specified DO level, m is the rank in
descending order, and N is the total number of events.
MODEL APPLICATION
The period of analysis was 1975 through 1977 -- a three-year period
experiencing above-normal rainfall distribution with the watershed in a
condition close to that projected over the life of any CSO abatement project.
The definition of an individual "event" is based on the return to zero storage
as employed by STORM. The timing of individual "events" is thus dependent on
watershed storage and sewer system capacities, as well as on the hourly
precipitation record. To reduce the dependence between events, a minimum
inter-event time of six hours was used for separation purposes. Using this
event definition criteria, STORM showed that 209 CSO events occurred in
Subarea A (urban) during the three-year simulation period. The number of
events in Subareas B-H (non-urbon) varied from 120 to 183, depending on
watershed characteristics. Obviously, larger amounts of precipitation are
required to produce runoff in the more highly pervious rural watersheds.
Thus, many smal 1 storms that cause runoff in Altoona (Subarea A) produce no
runoff in other subareas. The first flush or first hour flow rates in the 209
events ranged from 11 cfs to 389 cfs. BOD5 concentrations in the CSO ranged
from 9 mg/1 to 302 mg/1. The wet-weather runoff from Subareas B-H varied
between zero and several thousand cfs during the simulation period, with
typical BODg concentrations of !S mg/1 to 10 mg/1.
Model parameters in STORM were adjusted for runoff volume and pollutant
(BODc) load by calibration against nine individual storms on three urban sub-
areas where gaging instrumentation was installed. The details of STORM
calibration and extension to the Little Juniata system in Figure 1 are
presented elsewhere (1,2). The in-stream DO model, DOAP, was then structured
to simulate point source loading from each subarea. Base flows were estimated
from long-term quarterly averages. These were assumed to be at the same
temperature as the wet-weather inflows and also were assumed to be DO
saturated. Sewage treatment plant discharges were obtained and introduced in
DOAP at their respective locations.
In-stream DO samples collected during wet-weather events were then used
to calibrate DOAP. Samples were collected at different times during an event
at several accessible locations. During calibration, efforts were made to fit
the calculated DO profile to the observed DO data. Calibration was
accomplished by adjusting the values of YI and Y^ in the deoxyyenation
coefficient equation (6). Since an attempt is made to simulate the worst in-
stream condition during each storm event, the DO profile was fitted to the
minimum DO concentration measured at each sampling station.
The wet-weather event that occurred on July 27, 1982 was selected for
model calibration since the DO data collected on that day clearly illustrated
the impact of combined sewer overflows. The best possible fit of the DO sag
point was obtained for values of YJ and Y£ set at 14.5 and -0.37,
respectively. Several other wet-weather events sampled in 1982 were also used
for model verification purposes. Calibration and verification of the DOAP
model was successful in terms of matching the magnitude of the dissolved
oxygen sag. The DOAP model was not very successful in matching the entire
T-TI-B-6
-------
measured dissolved oxygen profile. Discrepancies of up to 2.0 mg/1 were
observed at downstream locations.
The discrepancy between the calculated profile and the measured data is
caused by several factors. First, due to problems in tracking CSO discharges
downstream, the samples col lected during an event may not reflect the true
minimum DO concentration that occurred at a specific location. Secondly, the
results from DOAP depend upon accurate wet-weather flow volumes and BOD
loadings as calculated by the STORM model. During calibration of STORM it was
observed that the predictions from STORM may not be reliable for any
individual event, although the model matched observed values reasonably well
over a series of events. Variation in the velocity-discharge relations is
another possible source of error. Thus, matching the magnitude of the DO sag
is believed to be compatible with the planning-level application of the DOAP
model in this study.
RESULTS
The mathematical models, STORM and DOAP, were used in tandem to simulate
the effectiveness of various measures in reducing the water quality impacts of
combined sewer overflows on the Little Juniata River. The results are
presented in this section for: (1) existing conditions; (2) selected CSO
storage treatment alternatives; and, (3) seasonal analysis of DO levels. It
should be noted that the results described below are for first-hour flow rates
and BOD loadings. This initial hour of each event was adopted as the most
severe loading period in an effort to represent first flush impacts on the
receiving stream.
Existing Conditions
As a means of establishing a reference or bench-mark for comparison of
alternative CSO abatement alternations, existing conditions were analyzed.
Results show that under existing conditions (existing diversion, no storage,
and a 6.77 M6D wet-weather treatment capacity) combined sewer overflows
frequently cause water quality conditions in violoation of the DER dissolved
oxygen standard. During the simulation period, in-stream conditions during 65
of the 209 events violated the standard established for the Little Juniata
River. The cumulative frequency curve for the 209 events under existing
conditions is shown as a dashed line in Figure 2.
Storage Treatment Alternatives
A number of storage/treatment/diversion combinations were selected for
analysis with the DOAP model to determine their effectiveness in protecting
water quality in the Little Juniata River. Figure 2 shows the effectiveness
of different storage capacities with existing flow diversion and existing
treatment capacity. Diminishing returns (smaller decreases in the frequency
of violation) is shown with increasing levels of detention storage. Similar
analyses were completed for modified flow diversion and a doubled treatment
capacity. The results from all of the analyses are summarized in Table 1.
Results for each storage/treatment/diversion combination are expressed as an
annual frequency of violation of the DER dissolved oxygen standard. The
T-II-B-7
-------
0
ci
Q
LU
u.
o
y
Q.
CD
z
Q
y
Ld
o
X
Ul
z
Li
Ul
u.
0
2 3 4 5 6 7 8 9 10 11 12 13 M 15
MINIMUM DISSOLVED OXYGEN CHG/L)
E « Existing Condition
.01 » Storage Capacity (Inches)
Figure 2: Minimum dissolved oxygen cumulative frequency curve.
T-II-B-8
-------
Table 1: Frequency of Violation of DO Standard Under Various
Storage/Treatment Conditions, Little Juniata River
Treatment Storage Treatment
Type of Capacity
Diversion (MGD)a (in)b (MG)C
Existing 6.77 0.00
0.01
0.05
0.10
0.20
Modified 6.77 0.00
" " 0.05
0.10
Modified 13.54 0.00
0.05
0.10
0.00
0.29
1.47
2.94
5.87
0.00
1.47
2.94
0.00
1.47
2.94
No. of Violations
Per Yeard
21.7
19.7
11.3
7.0
4.3
19.0
11.7
7.7
17.7
9.7
7.3
= million gallons per day; 1.0 gal. = 3.785 liter
°1.0 in. = 25.4 mm
CM6 = million gallons; 1.0 gal. = 3.785 liter
DO value below 5.0 mg/1 from February 15 to July 31; any DO value below
4.0 mg/1 during the remainder of the year.
present frequency of violation, 21.7 events per year, is significantly reduced
by the larger CSO storage capacities. Modified flow diversion and expanded
treatment capacity are effective in combination with small storage facilities,
but do not significantly affect the frequency of violation when storage
capacities are large.
Seasonal Effect
In an effort to observe seasonal trends in the minimum DO concentrations,
each of the three years in the simul at ion period was divided into quarters.
Results from the seasonal analysis are presented in Figure 3. It is evident
that wet-weather water quality conditions are worst during the third quarter
T-II-B-9
-------
o •««
. IvVr
Q :
* Qft-
Q ai!j~
UJ :
r" flcu
UL QV^
o :
lil 7ft-
111 /g-
Q. :
w ^
0. 5CU
(/) ^*r
2 i«:
y l0:
UCL;
u.
0 (
A
I ^
S
^
) t
•~-~.
2
vxv
<
kM
I
V
.3
'N
A
• ' !•— •
\?
•«.
]
\
*•%_
s
T
"^
"s,
-.
[
\
\
'N-^
ie
~\
\
\
i.
^v
V
it
*-s
>
>•».
N —
12
\
\
\
\
13
1
N
\
i
—.1 .
i
\
t
i
t
"-^->
H
t
i%
*»
^
IS
MINIMUM DISSOLVED OXYGEN CM6/L)
1 - Quarter 1: January 1 - March 31
2 - Quarter 2: April 1 - June 30
3 - Quarter 3: July 1 - September 30
4 - Quarter 4: October 1 - December 31
Figure 3: Seasonal analysis of minimum dissolved oxygen cumulative
frequency curve.
T-1I-B-10
-------
(July 1 to September 30). In general, CSO discharges have little effect on DO
levels during the first and fourth quarters, although a few violations of the
DO standards are observed in the first quarter. The seasonal operation of CSO
facilities may provide adequate water quality protection while significantly
reducing operating expenditures.
CONCLUSIONS
Several conclusions can be drawn from the results obtained in the study.
Combined sewer overflows from Altoona presently introduce substantial amounts
of biochemical oxygen demand in the Little Juniata River. During warm
weather, the BOD loading is sufficient to cause frequent violations of the
established dissolved oxygen standards. Due to the predominantly rural nature
of the study area downstream of Altoona, wet-weather inflows dilute the
impacts from combined sewer overflows. Combined sewer overflow abatement
measures that include storage facilities, expanded wastewater treatment
capacity, and modified CSO diversion have been shown by the modeling results
to be effective to varying degrees. In general, storage facilities show the
most promise in protecting receiving water quality.
The results from this study suggest that the Pennsylvania DER address the
issue of modifying its water quality standards to include provisions for wet-
weather conditions. Appropriate wet-weather provisions must be based on the
significance of the frequency, magnitude, and duration of DO depletion in
regard to the health of aquatic life. Meaningful cost-benefit abatement
analyses will depend on a better understanding of the significance of low in-
stream DO levels during wet-weather periods. This is of critical importance
to the selection of alternative CSO abatement facilities.
ACKNOWLEDGMENT
The investigation reported here was conducted under funding from the EPA
Municipal Wastewater Treatment Works Grant Program. The authors wish to
acknowledge Gwin Dobson & Foreman, Consulting Engineers, Altoona, PA for
permission to publish the study results.
REFERENCES
1) Gwin, Dobson and Foreman Inc., 1980. Combined Sewer Overflow Analysis,
Altoona, Pa., Altoona, Pa.
2) Gwin, Dobson and Foreman Inc., 1983. Combined Sewer Overflow impact on
Dissolved Oxygen in the Little Juniata River and the Frankstown Branch of
the Juniata River, Altoona, Pa.
3) Medina, Miguel A., Jr., 1980. "Continuous Receiving Water Quality
Modeling for Urban Stormwater Management," Proceedings of the national
Conference on Urban Stormwater and Combined Sewer Overflow Impact on
Receiving Water Bodies, EPA, pp. 466-501.
4) Medina, Miguel A., Jr., 1979. Level III: Receiving Water Quality
Modeling for Urban Stormwater Management, EPA-600/2-79-100.
T-II-B-ll
-------
5) Owens, M., Edwards, R., and Gibbs, J., 1964. "Some Reaeratlon Studies in
Streams," Int. Journal of Air and Water Pollution. Vol. 8, pp. 469-486.
6) U.S. Corps of Engineers, 1977. The Hydrologic Engineering Center,
Storage, Treatment, Overflow, Runoff Model (STORM) Users Manual, Davis,
Ca.
7) Wright, Raymond M., and McDonnell, AJ., 1979. "In-stream Deoxygenation
Rate Prediction," Journal of the Environmental Division, ASCE, Vol. 105,
No. EE2, pp. 323-33TI
T-II-B-12
-------
Potential and actual loading in a small mixed-use belgian basin.
Yves DELEU, attach^ at the Institute for Hygiene & Epidemio-
logy, Radioactivity section.
This paper emphasizes the result of a study of the load
of the Dyle River in the center of Belgium. This little basin
presents a high population density, and the ground is used for
farming, industry and urban purposes. The population is main-
ly concentrated along the riversides. The first step of the
study was a potential load approach, on the basis of two-year-
ly population, culture and crop, cattle, sheep, pig and poult-
ry census and non point source loading values in the lite-
rature. The second one was the work on four automatic sampling
stations, on time-mixed sample mode, for one and half year,
in the upper basin of the Dyle river. So it has to be consta-
ted a sharp contrast between the relatively constant potenti-
al loading and the quite variable weekly mixed samples para-
meters values, emphasizing the preponderant role of the potenti
al loading for the gross average load value and this of the
short time meteorological conditions for the variations.
T-II-C-1
-------
Introduction.
The study of the relationship between pollution potential and
actual loads seems to be a greatly difficult question, becau-
se of a lot of either less controlled or even unknown parame-
ters implicated in their estimation.The socioeconomic and de-
mographic changes in the given area complicated the approach.
A great discrepancy exists in the estimation of the unit po-
tential loads as computed by Cmernik ( 1976 ) and Volleweider
( 1968 ), who attempted to find a conrmon denominator between
the many and diverse values of particular potenial load of
grass, cattle, corn and beet field, poultry, industry, depen-
ding on the slope and the pedology of the ground, a.s.o., ...
A review of the literature over more than 10 years, suggested
me that, despite the revealed works of Cmernik and Vollenwei-
der, the particular potential loads census is far from termi-
nated, because the descriptive parameters of the many possi-
ble situations ( some of them are numbered above ) existing
for a sole parameter potential effect, as nitrate, are not
well systematically defined yet. A second discussion element,
is the applicability of the literature data, which are rrost
from an american source, to the european coastal continental
theatre, of which Belgium is a part. For this latter scene,
Vollenweider showed typical values of potential loads (1968).
The american ground use seems to be of relatively the same
intensive effect on the soil as the european continent one is.
This is despite the particular history of these soils, the
difference of climatic intensities, and other extrinsic fac-
tors differentially happening on the both sides of the Atlan-
tic Ocean.
We had the opportunity to study a little drainage basin for
three semesters in the years 1978-1979, with a small prototy-
pe sampling network system, using standard industrial refri-
gerated samplers, but with a rrodification, for remote working
control by phone lines. So, from week to week, a complete
historical data serie has been recorded for approximately one
and a half year, at four locations ( a, b, c and d on the map
figure 1 ). The data were 2-hour, day and week means ( Glaze-
makers and Dele, 1980, Dele 1984 ). Independent quantitati-
ve and rain data completed the data set. A full work on the
pedological, socio-economical and agricultural factors of
this basin has been performed, on the basis of statistical
data,with the aim to compare the measured loads with the pic-
tures resulting from the census.
The Upper Dyle basin. ( figure 1 )
The Dyle Basin is a little drainage basin at the east ofBrus-
sels, with three main affluents: the Senne, where Brussels
stands, the Demer with Hasselt as main town and the Dyle it-
self, with Louvain and the quasiconurbation Wavre-Ottignies-
T-II-C-2
-------
Fig. 1: The Higher Dyle Basin.
1 . GENAF'PE
2. LASNJXS
3. WAT1XR1 .OC)
4. LA HU1..PE
r5 . SI NT-GENES'I US RODE
6. 11 OP! 11..A ART
7. OVERUSE
8. HULDENDERG
9. SINT-JORIS-WEERT
10. GRKX DOICI.'JMJ
11. CHAUMONT-GISTOrX
12. WAL.HAIN
13. CllA.S'lKl-S
14. VJL.I FRS LA Vll.l.F.
1 ~>. COURT SA1 N'l ET J EN NE
16 . MONT SA I NT Gil I BURT
17. OTTJGNIKS l.OUVAJN !.A N'HUVE
18. RIXENSART
19. WAVKK
T-II-C-3
-------
Louvain-la-Neuve-Court-Saint-Etienne, and then a lot of sub-
sidiaries and municipalities of diverse importances. Its sur-
face is some 2700km2, with which this study concerns 599km2,
or 22 5K, called Upper Dyle basin, from the spring to the vil-
lage of Sint-Joris-Weert. The pedology of the upper basin
shows the predominance of calcarous sands ( table 1 ), cove-
red with loamy sand, then supporting the culture horizon. The
landscape is various, with hills and valleys, and not truly
plain surface. Some part of this country is called " Ardennes
Braban\onnes ". An estimated 385K of the above mentionned sur-
face is covered with forestry. The basin is inhabited by some
165000 inhabitants, with activities varying along the basin,
from predominently farming at the South to equally agricultu-
ral and residential at the North of the partial basin. So,the
dwelling density differs from 100 in the predominently agri-
cultural zones to 1500 in the rrostly residential ones: Rixen-
sart and Waterloo ( //// 3 and 18 on the map figure 1 ), for e-
xample. The potential population is estimated as 180000 to
200000 inhabitant-equivalents ( IE ) ( Dufour 1975, De Schep-
per 1976 ). Some industrial activity exists, but this momen-
tarily figure of some 50000 IE's ( De Schepper 1976 ) is now
different from the present one. Some industries are typically
bound to farming and then seasonal as sugar industry at the
end of each year; some others have now disapppeared ( as pa-
per and metal ) where they were traditional and new ones are
appearing in quite different zones. From this point of view,
this area is in constant evolution. The presented pictures a-
re thus valuable for the years 1978 and 1979. The communica-
tion network is relatively dense, with any sorts of ways from
the truck path to the great runways, as this one from Brussels
to Namur ( E41 ) and a dense railway. Thus, in a word, this
area is typically a mixed-use basin, with high dwelling and
use density, but on a few dozen of kilometers ( 20 to 30 mi.)
along the river axis. As another singularity of this area as
other ones in the neighbourhood of great cities, a great num-
ber of people travel daily for work to Brussels, Namur, Cha-
rleroi or Gembloux so that certain municipalities can be seen
as "sleeping zones" of these main attractive towns.The poten-
tial load computations don't take into account these daily
transhumances.
Material and methods.
The sampling program was started at four points in the area
( a,b,c and d on the map figure 1 ), and performed with stan-
dard industrial Buehler PRP/12T type samplers, with standard
refrigeration at 4[C and a 12 2.5 liter (2/3 USGal) high den-
sity PE bottles distributor.
T-II-C-4
-------
table 1 Pedogeology on the Upper Dyle Basin underground.
Age Nature of the formation Tickness
Pleistocene Loamy sand (agr.soil) 20m (60ft)
Brusselian Coarse quartz sand 25m (80ft)
Ypresian Clayed sand 40m( 130ft)
Landenian Fine glauconiferous sand 70m( 250ft)
Masstrichtian tuff 15m (50ft)
Senonian mamas 15m (50ft)
Silurian schists and quartzophyl lades 30m(100ft)
Loy W ( 1978 ).
-------
For particular technical reasons (Deleu 1981), it was prefer-
red to sample in function of time, in place of in function of
flow. For the same reasons, no conservation reagent has been
added; only cold at 4[C is used, though it was reconmanded by
Standard methods ( 1978 ) and Belgian standards (1981), among
many others.The analysis delay has been of one week after the
last sampling. The samples were analyzed according to the me-
thods of the Institute of Hygiene and Epidemiology (IHE 1976)
for a total of 16 parameters. From this set, 6 are discussed
here, because they seems to be the mostly referred in the li-
terature. They are N03-N, total N, Total P, O-PO4-P, SS and
ODD. Flow data came from the l£R telemetred hydro logical net-
work jointly operated by the former Noyau Administratif de
1'Eau and by the Rural Engineering Department of the Faculty
of Agronomy,Catho lie University of Louvain at Louvain-la-Neu-
ve. ( Persoons 1979, Bazier, 1980 ). The Meteorological data
came from the Royal Institute of Meteorology ( IRM ) ( Van
Diepenbeek 1980 ). The data are treated on a Hewlett -Packard
HP97 computer ( Hewlett-Packard 1977 ) and on a Tandy - Radio
Schack TRS80 model 100 portable computer ( Radio-Shack 1983).
For the estimation of the potential loading, it was seeking
to different sources, either already written studies over the
subject ( De Schepper 1976, Laurent 1980, Oovaert 1983 ), in-
formations graciously given by the Ministry ofEconomicAffairs
( 1979,1980 ) and the National Institute of Statistics (1979,
1980 ) or still, the known composition and quantitites of the
main spreaded fertilizers and manures during the asolement
period . Thus, there exist numerous approaches of the poten-
tial pollution loading estimation and comparison. But because
the statistical informations are far more complete and giving
6 months pictures, and so permitting to see the possible va-
riations in the potential loads along 2 years, then the re-
sults and discussion about potential pollution mainly will
concern these statistical informations.
Results.
Figure 1 is a contour map with the administrative limits of
the municipalities in the upper Dyle Basin.The pictures refer
to the name listing of these municipalities. The small I.S.O.
letters indicate the four sampling locations.
Figure 2 is an analysis of the different 6 month censuses for
the population and human activities in the basin (figure 2 t)
and for the culture ( figure. 2b-), and the information is
treated according to Berlin ( 1977 ). The activities are re-
censed following the river axis and following the values of
T-II-C-6
-------
NACE coded industrial activitie
o o o o o o
FIGURE 2 : human acMvities in the Upper Dylo Basin
-------
the pollution load unit conversion coefficient for industrial
used waters ( Moniteur Beige 1974 ).
The lisibility of such pictures is somewhat particular. Each
subfigure owns its own scale. The full black square expresses
100% value (in which the true value may be either inserted or
drawn on the upper side ), the other values are proportional
to the width of the band. Under 5%, which is pictured with
either a sharp line or a blank, the values are expressed in
the closest round fraction of a disk, as 7/8, 3/4, 4/8, 1/2,
3/8, 1/4, 1/8,... A fraction of a thenth or less is indicated
with a little sharp vertical upper line.A fraction of a twen-
tieth less is drawn with a little sharp vertical lower line.
When it is zero, the figure is O. The occasionnally seasonal
activity ( which is expressed with a 25 % variation between
two consecutive 6-month actvities ) is shown by partitioning
the square figure in two triangles ( upper: surnner, June, lo-
wer: winter, december) with a diagonal,in wich both triangles
the above mentionned code is also valuable. The industrial
and service activities are distributed according to the N/*CE
code,into categories numbered from 0 ( farm and natural acti-
vities ) to 9 ( services ) ( Gewestelijke Ontwikkelingsmaat-
schappij voor Vlaamse Brabant 1979 ). The relative quantities
(in percent) for a given N/°CE category is written in front of
the concerned municipality. Contrary to the other comparison
items, showing a vertical comparison line between municipali-
ties, along the river axis, the distribution is here horizon-
tal for the same municipality. By this way,the repartition of
the human activities and the residential zones becomes quite
well seen, with the dynamics of the changes along the river
axis and the year.
The industrial activities are also taken into account as po-
tential pollution load on a twin basis according to the N/CE
distribution on one side, and according to the standardized
industrial equivalent-inhabitants (HE), which is deduced on
the legal definition of the equivalent-inhabitant. ( Moniteur
Beige 1974 ). The definition of the IE has been discussed el-
sewhere ( Deleu 1983, Vanderborght 1981 ). Unfortunately, the
lawmaker didn't use the HflCE. industrial activity distribution
code and considered a little amount of the classical indus-
try, with an arbitrary insistance over unattended details.The
distribution of the recensed activities is then dependent for
a given part to the subjectivity of the user.But the numerous
chosen N/"CE categories are sufficient enough to minimize the
possible estimation errors. The resulting IE are converted to
potential pollution loads according to the I E legal defini-
tion. It becomes then apparent that such detailed approach of
the potential polluting EH, even if misestimated for the abo-
vementioned reasons, is much higher than the preceeding esti-
mations ( Dufour 1975, De Schepper 1976, Oovaere 1981 ).
Thus, owing to the loss of data due to non recensed ground u-
T-II-C-8
-------
ses (ways, aso.,...) and the uncertainty of some of the in-
dustrial activities, the potential pollution estimates must
be considered as partial ones, even if this part represents
the great majority of the true estimation.
Figure 3 shows the variations of potential ( a) computed from
the literature minimum values, b) computed from the literatu-
re maximum values, c)conrputed from the non industrial equiva-
lent living population, including the recensed cattle, d)com-
puted from the industrial equivalent population, as recensed
following the NAZE code ) and actual 6 months loadings (e) at
the four sampling sites and for the six analyzed parameters.
The values are expressed in metric ton/6 months. The lecture
code is the same as for figure 2. Here, the seasonal activity
is visible by itself, because the shown values are 6-monthly
ones.
Figure 3 presents data corrected with respect to the length
of a semester in weeks, e.g.26 weeks/haIf-year, and converted
from kg/s to metric tons/6 months, with a time conversion co-
efficient from 86400 second a day to week of 7 day and a se-
mester of 26 weeks. For each partial basin,the picture is the
expression of a specific income load in the river for the
concerned parameter over 6 months.
Figure 4 presents a distribution of the actual loads between
the runoff ( white area ), the hypodermic (delayed) dischar-
ge ( hatched area) and the ground discharge (black area).
The picture is drawn on the same scheme as the preceding one.
The symbols are the same as above excepted one more : a squa-
red one indicates insufficient data.When a particular dis-
charge is estimated equal to zero, the figure shows a O. If
the discharge value is under 5%, a as above fractionned disk
is drawn.over the zone,with the same drawing code of the dis-
tribution.The square hatched zone of the disk is without mea-
ning.
The values are expressed with respect to the percent!le dis-
tribution; 100 % is the value of each squared subfigure.
Discussion.
It is first worth saying that every measured values conside-
red here are mean values, not point values. The total grab
sampling number is within the 36000/year.station. The results
and discussion are dramatically affected with such data col-
lection because the majority of the past discussions are re-
T-II-C-9
-------
H
I
PARAMETER
POTENTIAL POLLUTION
LOWER LOAD ESTIMATE
T /6 months
CULTURES
B) POTENTIAL POLLUTION
HIGHER LOAD ESTIMATE
T /6 months
CULTURES
C) POTENTIAL POLLUTION
LOAD ESTIMATE
10* T /6 months
NONINDUSTRIAL EQUIVALEN
POPULATION (people-<-cattle
D) POTENTIAL POLLUTION
LOAD ESTIMATE
T/6rponths
INDUSTRIAL EQUIVALENT
POPULATION.
ACTUAL POLLUTION
LOAD ESTIMATION
T/6 months
NN N N
NN N N
NN N N
NN N N
NN N N
NN N N
NN N N
NN N
FIGUi
-------
H
I
n
NO-N
1978
1979
N
n
total-N
1_978
1
1979
N
N
N
n
total- P
1978
1979
N
N
EC
n n
o-PO-P
1976
jMm
1979
N N
N
N
DD
SS
ma
samp
ing sites
FIGURE4: percentile distribution of actual loads between ground discharge, hy-
podermical discharge and runoff
-------
ferring to grab samples, which are taken at different time
frequencies, or with a more or less complex relation with the
flow.
On the other hand, by the census recensed surfaces are lower
than the above cited picture of 599 km2. It remains about 85
km2 ( 1* % ) out of the census, which are taken back as not
recensed, for example some routes and forestry.
The pathways and the forestry are taken into account by them-
selves for the actual load computation, but in the potential
load estimation not, excepted some forestry, as rural ones
( National Institute for Statistics 1979, 1980 ).
On the figure 3, the effect of the variations from a semester
to the other is remarkable, for the six concerned parameter,
even in the case of the actual pollution loads. The alternan-
ce in the loads is quite apparent, giving an impression of
predominance of the seasonal activities on the otherones.This
same variation character of both potential and actual loads
suggests that the potential pollution pressure also ownsfrac-
tal properties, as the actual ones have (Pearsons coefficient
values up to more than 150 % ). By the way, it was shown else
where ( Dele 1984 ), that the quality variations follow the
laws of fractal curves, or the Mandelbrot theory ( Mandelbrot
1975 ) from which the most remarkable property is the conser-
vation of the standard deviation variations at every scale
these variations are examined. The load pressure was assumed
constant because the normal way of life of the populations
and the indefatiguable industrial activity. Actually, the po-
tential loading is quite more dynamic and variable than the a
priori computations can show. Thus, it seems to be worth wor-
king the calculations at smaller scales than the abovementio-
ned 6 months, for greater informational precision.
The second remarkable element is, shown in the figure 3, the
great difference between potential ( figure 3 a minimum, 3 b
aximum, 3c non industrial EH, and 3d industrial EH ) and ac-
tual ( figure 3e ) loads estimations, in the order of 1000 to
1.
The import from population and livestock appears much more im-
portant than the culture import. The actual load appears clo-
ser of the higher culture load estimations and a few tenths
of percents of the population and livestock E H and its legal
conversion.
Furthermore, the potential estimation is but partial,because
not including the non recensed sources,though the actual es-
timation concerns everything passing through the Dyle. Logi-
cally, both are estimated with the same amount of population
and ground use, minus what wasn't recensed in the potential
estimate.
The most frequent ( more than 95 % in frequency ) polluting
load import is concormuttant with occurrence of rain,superfi-
cially appearing to have the central role in load transport
and then which could explain the alternance from a semes-
ter to the another. But the few percents ( 5 to 30 % ) of the
T-II-C-12
-------
actually transported loads by net rains in the case of some
parameters,forbit to consider its role as of the most im-
portant. In other wards, the role of the rain seems less being
the cause of transportation than its throttle. In non rai-
ning conditions, the ground discharge withdraws the greatest
part (from 80 up to 100 %) of the total charge coming up to
the river, despite of the lower flow, with stagnation and se-
dimentation in the waste pipes. When raining,the runoff with-
draws a great part ( from 30 up to 70 % ) of the total charge
and the resulting sedimented loads are pushed down with the
runoff and then become active in the river water, giving the
well known pollution peaks. But, an analysis of the base line
of these events shows a great contribution of the hypodermic
discharge to the polluting load (up to 80 %), with a time de-
lay of about one week.The remaining of the charge (up to 75%)
is taken in charge with the ground discharge and constitutes
the basis of the loading. Sometimes, for nitrate-N,the ground
discharge does not exist.
The effect of a seasonal activity on the water quality as the
yearly sugar campaign ( Muller 1966 ), may be quite expressi-
ve and dramatically change the parameter values even for are-
latively long time length: two months,e.g.one third of semes-
ter ( Glazemakers and Deleu, 1980 ). But as it could be seen
on the figure 3e, this local effect remains invisible in the
whole load mass, despite the time rate.
The dramatic difference between the measured charges and the
potential load computation results will not lead to conclude
that the earlier work must be repeated. It evently could be
computed according to the unit values of Vollenweider (1968),
but the parametrization of these unit values is not known yet
for the Upper Dyle Basin. This difference overall underlines
a great difference in the approaches in measuring and in
computing.
On the other side,it could also be the sign of a great reten-
tion potency of the polluting load in the basin. This reten-
tion could be either physical ( long time sedimentation, at-
mosheric elimination of NH3 ( Loehr 1974 ) for example, ) or
depending on a long delay discharge, or degradative ( natural
purification ), or still the three phenomena together as Be-
noit ( 1972 ) showed in the case of natural waters. The form
of the purification processes could be rrore general and comp-
lex than assumed by Benoit, it is to say that they occur not
only in the watercourses. Then, at the next step, the measu-
red load is only a small fraction of complex origin of the
initial load computed as potential load. Further, it appears
that this purification process is of quite high efficiency,
on a time delay on only a few days in the river for the run-
off ( Bazier 1981 ), one week for the hypodermic discharge.
On quantitative basis, the known low values of runoff coeffi-
cient ( .1 to .32 following De Pelsmaker ( 1980 ), .01 to .15
following our own less rigourous estimations ) are assumed to
T-II-C-13
-------
fill a part of the process conditions. In river water,the in-
tensity of these degradation for NH4 is high ( 80 to 95 %
oxidation )( Curtis, Durrant and Harman 1975 ). Joining these
two values, 0.01 for the runoff coefficient and .05 for the
degradation, one can estimate the remaining NH4-N load to the
.05 % of the initial potential load.This is within the actual
load of the NH4-N relative parameter N03-N.The final explana-
tion lies thus in a combination of degradation and quantity.
It could maybe remain elsewhere,but with greater uncertainty.
Conclusions.
The above presented results indicate an unexpected great dif-
ference between potential and actual loads.This great diffe-
rence between room calculations and field measurements indi-
cates a retention potency of the polluting charge along the
time and resulting in a quite efficient loss of load. A begin
of complete explanation can be presented in a combination of
degradation chemistry and hydrology.
The alternance of the results from 6 to 6 months, indicates a
seasonal effect on both potential and actual pollution loads
showing the possible short term variation of the potential
load like the actual load variations.
During this study, the use of high density data, as automatic
permanent samplers can collect, appear quite potent in the
approach of long term complex water quality phenomena.
Acknowledgements: The autor is quite indebted to Mr E. Glaze-
makers, Technical Engineer, who was in charge of managing the
sampling network, to SSrs 3. Sadones and C.Van Eerdenbrugghe,
Technical Engineers, who helped the work and accomplished pa-
rallel programs on the Dyle R.
He wants too to thank Mr. Bouquiaux, Chief, Environmental De-
partment, who supported the project, Mr. Ir. E. Laurent, for-
mer Director of the NAE, for the command of the project and
for the quantitative values, Mr Van Diepenbeek, Attach6 at
the Royal Institute for Meteor logy,for the meteorological da-
ta, Mrs.Bavin and Mr. Coevelier,of the National Institute for
Statistics, for the numerous census listings, Prof. L.De Bac-
ker, former Dean, Prof. E. Persoons and Mrs G. Bazier, Assis-
tant, of the Rural Engineering Department, Faculty of Agrono-
my, University of Louvain la Neuve, for critical discussions
and practical helps, Mrs. Braun and Mr. Bastin of the IBW for
helpful discussions and data, Dr. P. Lejeune, Head, Radioac-
tivity section, for his encouragement, Mrs.Martur-De Vre for
the revew the manuscript, and Tandy computer Centre for his
kind help for printing the manuscript. This project was sup-
ported as an extern program of the N/°E,Dyle Corrrnission,Minis-
T-II-C-14
-------
try for Public Health and Family.
Bibliography.
Bazier G ( 1980 ) in Monographic du Bass in de la Dyle, E.
Laurent ed., Ministere de la Sant£ Publique, Noyau
Administratif de 1'Eau, Corrmission Dyle, Brussels, 1980 (
Dyle Basin Monograph, E. Laurent ed., Ministry for Public
Health and Family, Water Administration Cell, Dyle
Conrmission 1980 )( in French ).
Belgian standards (1980) Echantillonnage des eaux - ^
Principes ge"ne"raux NBN T 91-051, Belgian Insitute for
Standardization ( Water sampling - Fundamentals ), Brussels
1980 ( in French ).
Belgian standards (1980) Echanti1lonnage des eaux -
Reconmandations g£n£rales pour la conservation et la
manipulation d'<§chanti lions NBN T 91-052, Belgian Insitute
for Standardization ( Water sampling - General
Recommandations for samples conservation and manipulation )
Brussels 1980.
Benoit RJ ( 1972 ) Self-purification in Natural Waters, in
Water and Water Pollution Handbook, L. Ciaccio ed. M. Dekker
New York 1972.
Bertin 1977 La graphique et le traitement graphique de
1'information, Flamnarion, Paris 1977. ( in French ).
Buehler PRP/12T type sampler ( in German ).
Curtis, Durrant and Harman ( 1975 ) Nitrification in rivers
of the Trent Basin, Water Research 1975, 9, 268.
Deleu Y ( 1981 ) L'<§crianti llonnage appliqu6
1'e'chanti llonnage automatise" special licence memo ire,
Louvain-la-Neuve 1981 ( in French ).
Deleu Y ( 1983 ) L'expression du bilan massique du bassin
moyen de la Dyle en termes d'Equivalents-Habitants (
manuscript ) ( The net load yield expressed in Inhabitant-
Equivalent units )( in French ).
Deleu Y ( 1984 ) Le caractere fractal des courbes de
qualite" de 1'eau Tribune du CEBEDEAU 1984, 37,n[489, pp90-
101 ( Fractal curves in water quality. ) ( in French ).
De Pelsmaker ( 1980 ) Ame"nagement hydraulique du Bassin de
la Dyle (Dyle Basin Hydraulic Management , a monograph,
Agricultural Hydraulics ), Ministry for Agriculture, Mons
1980 ( in French ).
T-II-C-15
-------
De Schepper et al. ( 1976 ), The chemische en
fysicochemische verontreiniging van de Dijle, IHE, Ministry
of Public Health and Family, Brussels, 1976 ( The chemical
and physicochemical pollution of the Dyle )( in Dutch ).
Dufour 3 ( 1975 ) La collecte des eaux use"es dans la Valise
de la Dyle, in Le Cycle de 1'Eau Louvain-la-Neuve, 19 mars
1975, Socie"te" de Ge"nie Rural, Louvain-la-Neuve, 1975, ( Used
Waters collection in the Dyle Valley, in The Water Cycle at
Louvain-la-Neuve, March 19th, 1975, Society for Rural
Engineering ).
Gewestelijke Ontwikkelingsmaatschappij voor Vlaams Brabant:
Bedrijven in Vlaams Brabant adresboek 1979 Gewestelijke
Ontwikkelingsmaatschappi] voor Vlaams Brabant, Vilvoorde
1980 ( Regional Development Association for Flemish Brabant
: Trade and Industry in Flemish Brabant addressbook 1979
RegionaT Development Association for Flemish Brabant )( in
Dutch ).
Glazemakers E and Deleu Y ( 1980 ), unpublished results.(
in Dutch )
Hewlett-Packard ( 1976 ) Stat Pac I,co Hewlett-Packard
Singapore 1976.
Hewlett-Packard ( 1976 ) Standard Pac, co Hewlett-Packard
Singapore 1976.
IHE ( 1976 ) Metoden voor de fysicochemische analyse van
water, Institute for Hygiene and Epidemiology, Ministry for
Public Health and Family, December 1976, Brussels (
Physicochemical analysis methods for waters )( in Dutch ).
Laurent E. ( 1980 ), in Monographic du Bass in de la Dyle,
E. Laurent ed., Ministere de la Sant£ Publique, Noyau
Administratif de 1'Eau, Conrmission Dyle, Brussels, 1980 (
Dyle Basin Monograph, E. Laurent ed., Ministry for Public
Health and Family, Water Administration Cell, Dyle
Comnission 1980 )( in French ).
Loehr RC ( 1974 ) Characteristics and comparative magnitude
of non point sources JWPCF 1974, 46, 8, 1849.
Loy W ( 1978 ) Hydrogeological guide-book - The Dijle
Valley. Hydrographica 1978, p25.
Mandelbrot B. ( 1975 ) Les objets fractals: forme, hasard et
dimenesion, Flanrmarion Paris 1975 ( The fractals Things:
Shape, Randomness and Dimension, Flanmnarion, Paris )( in
French ).
Ministry of Economic Affairs ( 1979 ) Occupation des
terrains @ 1'usage de 1'industrie, de 1'artisanat ou des
services, situation au 31.12.1978, Ministry of Economic
Affairs, Administration of industry, economic expansion and
foreign investments, Brussels 1979 ( Soil occupation and
industry, manufacturing and servicing, situation at
12.31st.1978 )( in Dutch and French ).
T-II-C-16
-------
Ministry of Economic Affairs ( 1981 ) Occupation des
terrains 1'usage de 1'industrie, de 1'artisanat ou des
services, situation au 1.1.1980, Ministry of Economic
Affairs, Administration of industry, economic expansion and
foreign investments, Brussels 1981 ( Soil occupation and
industry, manufacturing and servicing, situation at
1.1st.1980 )( in Dutch and French ).
Moniteur Beige ( 1974 ) Arrt£ Royal du 23 Janvier 1974
fixant les minima de charge polluante preVus aux articles 9
™ 2, 3 et 20 b de la loi du 26 mars 1971 sur la protection
des eaux de surface centre la pollution. ( King Arretee of
January, 23th, 1974 stating, the minimal polluting loads,
foreseen in articles 9 ] 2, 3 and 20 b of the Law of March
26th, 1971, upon the surface water protection against the
pollution,
Moniteur Beige ( 1974 ) Arrt£ minist<§riel du 15 fe"vrier
1974 determinant en execution de 1'article 4 "2, de la loi
du 26 mars 1971 sur la protection des eaux de surface centre
la pollution , la valeur des coefficients de conversion en
unite's de charge polluante pour les eaux provenant des
entreprises industrielles ou autres. ( Minister Arretee
determining, for the accomplishment of the Law of March
26th, 1971, upon the surface water protection against the
pollution, the conversion coefficients values in pollution
loading units for used either industry or others waters ) (
in Dutch and French ).
Muller ( 1966 ) Treatment of mixed domestic sewage and
industrial wastewaters in Germany CECD report Pris 1966.
National Institute of Statistics ( 1979 ) population census
1978 -municipalities of Sint-Genesius-Rode, Hoeilaart,
Overijse, Huldenberg, Genappe, Villers-la-Vilie, Chastre,
Court-Saint-Etienne, Walhain, Ottignies-Louvain-la-Neuve,
Mont-Saint-Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-
3oris-Weert, Lasnes, Waterloo, La Hulpe, Rixensart ( in
Dutch and French ).
National Institute of Statistics ( 1980 ) population census
1979 -municipalities of Sint Genes ius-Rode, Hoeilaart,
Overijse, Huldenberg, Genappe, Villers-la-Vilie, Chastre,
Court-Saint-Etienne, Walhain, Ottignies-Louvain-la-Neuve,
Mont-Saint-Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-
Joris-Weert, Lasnes, Waterloo, La Hulpe, Rixensart ( in
Dutch and French ).
National Institute of Statistics ( 1979 ) agricultural
surface uses census 1978 ( listing ) - municipalities of
Sint-Genesius-Rode, Hoeilaart, Overijse, Huldenberg,
Genappe, Villers-la-Vilie, Chastre, Court-Saint-Etienne,
Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-Guibert,
Grez-Doiceau, Chaumont-Gistoux, Sint-Joris-Weert, Lasnes,
Waterloo, La Hulpe, Rixensart ( in Dutch and French ).
National Institute of Statistics ( 1980 ) agricultural
surface uses census 1979 ( listing ) - municipalities of
Sint-Genesius-Rode, Hoeilaart, Overijse, Huldenberg,.
T-II-C-17
-------
Genappe, Villers-la-Ville, Chastre, Court-Saint-Etienne,
Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-Guibert,
Grez-Doiceau, Chaumont-Gistoux, Sint-3oris-Weert, Lasnes,
Waterloo, La Hulpe, Rixensart ( in Dutch and French )
National Institute of Statistics ( 1979 ) cattle, porcine,
horses, and farm yard census 1978 ( listing ) -
municipalities of S in t -Genes ius-Rode, Hoeilaart, Overijse,
Huldenberg, Genappe, Villers-la-Ville, Chastre, Court-Saint
Etienne, Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-
Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-Joris-Weert ,
Lasnes, Waterloo, La Hulpe, Rixensart ( in Dutch and French
National Institute of Statistics ( 1980 ) cattle, porcine,
horses, and farm yard census 1979 ( listing ) -
municipalities of Sint -Genes ius-Rode, Hoeilaart, Overijse,
Huldenberg, Genappe, Villers-la-Ville, Chastre, Court-Saint-
Etienne, Walhain, Ottignies-Louvain-la-Neuve, Mont-Saint-
Guibert, Grez-Doiceau, Chaumont-Gistoux, Sint-Joris-Weert,
Lasnes, Waterloo, La Hulpe, Rixensart ( in Dutch and French
Cmernik 3M (1976) The Influence of Land Use on Stream
Nutrient Levels, EPA-600/3-76-014, US Environmental
Protection Agency, Washington DC.
Oovaert A. et al. ( 1983 ) Optimal isatie van de
water zuiver ing door mi dd el van kaarten met de biologische
kwaliteit van de water lopen - Bekken van de Dijle tot
Wrechter, Groep voor Toegepaste Ekologie vzw, IHE Brussels
1983 ( Optimization of water purification with mean of
watercourses biological quality maps - Drainage basin of the
Dijle River at Wrechter ( Belgium - Working Group for
Applied Ecology )( in Dutch ).
Persoons E. ( 1979 ) Application du systee LGR en Belgique,
in Problemat ique et gestion des eaux int^rieures, Calembert
ed. Derouaux, Liege 1979 ( LGR systems applications in
Belgium in Inland Water Problematics and Management )( in
French ).
Radio-Shack ( 1983 ) TRS-80 Model 100 Portable computer
User Book, Tandy Corporation, Fort Worth 1983.
Standard methods ( 1978 ) for analysis of waters and
wastewaters APHA-AWW 1978.
Van Diepenbeek ( 1980 ) meteorolgical data for Beauvechain,
Heverlee and Ottignies for years 1978 and 1979 ( listing )
Vollenweider ( 1968 ) Recherches sur rame"nagement de 1'eau
- Les bases scientif iques de 1 'eutrophisation des lacs et
des eaux courantes sous 1 'aspect partiel du P et de N comme
facteurs de 1'eutrophisation, OCDE Paris 1968 - Scientific
fundamentals of the eutrophication of lakes and flowing
waters, with particular reference to nitrogen and phophorus
as factors in eutrophication. CECD Paris 1968.
T-II-C-18
-------
MODELING SEDIMENT DELIVERY FROM FIELDS TO CHANNELS;
A CASE FOR FIELD SCALE DATA AND HIERARCHAL DATA STRUCTURES
Kenneth Baun
Environmental Specialist
Wisconsin Department of Natural Resources
ABSTRACT
There is a broadscale awareness of and concern regarding the water quality
problems caused by nonpoint source pollution. There is also a growing
committment to abating this type of pollution. In order to direct nonpoint
pollution control activities in the most cost-effective manner, it is
necessary to understand the erosion, sedimentation and sediment transport
processes in a watershed. This understanding is best developed through the
use of a mathematical model.
There are several watershed models available today that can be used to
evaluate erosion, sedimentation and sediment transport. They vary
considerably in their input requirements and output abilities. These models
generally require that a watershed be subdivided into many small, square,
cells. The unit of data is a grid cell. Data for the model is then
collected, entered into the computer, and analyzed on a cell by cell basis.
From the modeling experiences associated with the Wisconsin Nonpoint Source
Program, it appears that this grid cell technique is not the best way to
divide, collect and analyze watershed data.
An alternative modeling approach was subsequently defined and is presented.
This approach incorporates field scale data and hierarchal data structures.
Field scale data means that the unit of data is the farm field. Data is
collected on a field by field basis. The use of field scale data allows for
more homogeneous cell attributes and more applicable field analysis and
management plans.
The use of field scale data also necessitates the use of hierarchal data
structures. Hierarchal data structure refers to the manner with which data is
accessed and processed in a computer program. Hierarchal data structure
indicates a data storage technique by which all cells are accessed only by
means of their relationship to other cells. The use of hierarchal data
structures parallels the downslope movement and aggregation of runoff, and is
the most appropriate data structure for the process being modeled. Most
significantly, the use of field scale data and hierarchal data structures
enables a model to be used to estimate the delivery of eroded sediment from a
field to a channel.
Key Words: sediment delivery, sediment yield, delivery ratios, sediment
trapping, nonpoint pollution.
T-II-D-l
-------
INTRODUCTION
The Wisconsin Fund Nonpoint Source Program was initiated in 1978 to identify
and rectify nonpoint pollution problems in the state. This program, with a
current annual budget of $5,600,000, provides state cost-share dollars for
nonpoint controls in targeted areas of designated Priority Watersheds. The
Nonpoint Source Section of the Department of Natural Resources ("the
Department") has the responsibility for developing a watershed plan for each
Priority Watershed. Each plan includes an identification of the nonpoint
pollution problems in the watershed, the water quality objectives, the
critical sources (eligible areas) of nonpoint pollution, and the applicable
best management practices. Each year there are four to six new watersheds
added to the program, each of which is usually 250 to 750 square kilometers
(100 to 300 square miles) in area. The analytical tools used in the
identification of critical areas, the selection of remedial measures, and the
estimation of resultant water qualtiy impacts, must be appicable to large
areas in a timely and cost-effective manner.
Extensive land use iventories are conducted on each watershed. Currently, all
fields within a quarter-mile of a channel are inventoried for location,
landowner, soils, topography, land cover, rotation and other management
practices. These inventories are conducted on a field by field basis by local
county agencies under contract with the Department. An average annual rate of
erosion is derived for each field using the Universal Soil Loss Equation
(USLE). In the most recent round of watershed plans, one of the watersheds had
over 8500 fields for which the USLE was calculated. Following these soil loss
calculations, the watershed planner, in conjunction with the local agencies,
selects a target soil loss rate and a prioritized list of best management
practices applicable to the watershed. The computer then "applies" these
practices one at a time to each field with a soil loss rate above the target
level, recalulating the soil loss rate after application of each practice,
until the rate is either below the target level or, infrequently, all
practices have been applied. The fields are then aggregated by landowner.
Landowners are in turn prioritized for eligibility in the program based upon
the anticipated total reduction in soil loss.
This method has several advantages: The data base is similar to that used by
several county and state resource management agencies. This sometimes allows
for shared data and reduced data gathering time. It enables broadscale data
collection to be carried out over many large areas simultaneously and requires
no sophisticated data gathering tools or precise geolocator techniques. It
also gives the local agencies a significant role in the planning process,
which gives them a large measure of "ownership" in the plan. This method has
two strong disadvantages however. It does not distinguish between differences
in sediment delivery between fields, nor is it capable of estimating watershed
sediment yield. Rather it treats all inventoried fields identically,
regardless of the ultimate delivery of eroded sediment from each field to
receiving waters.
T-II-D-2
-------
Consequently, it does not distribute limited money for nonpoint source
controls in the most cost-effective manner. In lieu of a better tool, it
relegates this phase of the program to a function of soil erosion control
rather than water quality control.
Recognizing this, in 1983 the Department procured a Federal grant to
incorporate the ANSWERS model into it's planning and evaluation activities.
ANSWERS is a distributed parameter hydrologic and sediment detachment and
routing model, intended for primarily agricultural watersheds (Beasley and
Muggins, 1982). A distributed parameter model, unlike a lumped parameter
model, divides a watershed into many discrete cells. ANSWERS, like several
other distributed models (MODANSW [Park and Mitchell, 1983] and AGNPS LBOSCH
et al,1983]) is a grid based model. A square grid is placed over the
watershed, and a row and column locator is assigned to each cell. For the
ANSWERS model, the location (row and column), slope, direction of slope, soil
type, land cover and channel class (if any) of each cell must be designated.
Several variables are required for each soil type, including total porosity,
field capacity, infiltration descriptors, antecedent soil moisture, and USLE
"K" value. Several more variables are required for each land cover, including
rainfall interception capacity, surface cover, surface shape descriptors,
Manning's N, and USLE "C" and "P" values. Several of these variables
fluctuate seasonally, requiring several data bases for different crop stages.
ANSWERS, like other distributed models, is event specific, i.e., it models
runoff and sediment concentrations for single events. It also operates in a
stepped time mode, i.e., under event simulation, the conditions of each cell
are calculated and updated every few minutes for the duration of the event.
Because of its distibuted nature and stepped time mode, ANSWERS requires large
amounts of data and is computationally intensive for large watersheds.
The primary advantages of a distributed model like ANSWERS is that it can
estimate net sediment erosion/deposition on each cell, total watershed
sediment loadings, and the overall impact of changing land management on
various cells on sediment loadings. However, there are two drawbacks to a
distributed model like ANSWERS that, indirectly, are the subject of the rest
of this paper. The first is the difficulty and inapproprlateness of
collecting and processing land based data, especially for large areas, in a
grid based format. The second is the inability of a computer internal grid
based data structure to estimate the eventual deposition or delivery of
sediment from individual fields to receiving waters (as opposed to the net
sediment movement on each cell). In response to these drawbacks and to the
needs of the Wisconsin Nonpoint Source Program, an outline of a proposed
sediment delivery model, incorporating field scale data and hierarchal data
structures, is presented.
T-II-D-3
-------
DISCUSSION
Field scale data
The data format that has most frequently been used in watershed modeling of
sediment erosion and delivery has been the grid cell approach. Using this
method, a grid is placed over a map of the area being modeled, which is then
divided into numerous identically sized cells. The cells are thereafter
identified by row and column numbers and they become the unit by which data is
collected, entered into the computer, and processed. The utility of the grid
cell approach can be examined in light of its attributes relative to those of
a field scale data approach.
To its credit, the grid cell method enables an easy locator to be associated
with each cell, i.e., any given row/column designation can be readily located
on a map. Therefore, the input or output of the model can be easily graphed
and visually displayed. Further, in the case of ANSWERS and MODANSW, it
allows for a more accurate determination of runoff direction from each cell,
i.e., the direction of flow is highly variable and can be proportioned between
two adjoining cells. In the case of AGNPS, the direction of flow must be
designated to one of four directly adjacent cells or four diagonally adjacent
cells. Lastly, because of the rigid and simple relational locators, it makes
programming and modifying grid based models a relatively simple task.
There are of course a few major drawbacks in using the grid cell approach.
One relates to the size and homogeneity of the cells. The size of the cells
should be a function of the diversity and the size of the area being modeled.
The smaller the cells, the more homogeneous their contents, the greater their
number.
Tradeoffs will often have to be made, sacrificing the homogeneity of smaller
cells in favor of the practicality of larger ones. In the 727 hectare
watershed that the Department calibrated and evaluated the ANSWERS model on, a
cell size of one hectare was employed. Using this same cell size on the
Priority Watersheds would require generally 25,000 to 75,000 cells for each
watershed, obviously a prohibitive number on any account. Using larger cells,
perhaps ten hectares, would alleviate that problem but would result in another
one. In much of Wisconsin's diverse topography, ten hectare cells in a grid
system would yield very high diversity in the many cells that encompased two
or more very diverse land covers, slopes, soils, etc.. The authors of the
AGNPS model (Bosch et al, 1983) suggest using a four hectare (ten acre) cell
size for watersheds less than 800 hectares (three square miles), and 16
hectare (40 acre) cell size for larger watersheds. Given the diversity found
in most of Wisconsin's landscape, cells of the latter size would prove far too
heterogeneous for accurate representation of their contents.
In a recent related study by the author, the soil diversity of a 1024 hectare
area of Dane County, Wisconsin was examined. The soils of the area had
previously been digitized, and this analysis was performed by computer. A
grid matrix composed of one hectare cells was superimposed over the area, and
T-II-D-4
-------
the number of different soil classifications in each cell was counted. Only
153 cells had just one soil classification within them. Of the other 871
cells, 447, 313, 79, 28, and 4 cells contained, respectively, 2, 3, 4, 5, and
6 different soil classifications.
A second problem related to using grid cell data involves making field
management decisions based on grid cell model outputs. The entire purpose of
running a model is to make field management decisions related to water quality
objectives. Yet how are management decisions made when there are cells
straddling more than one field, or multiple cells in a field indicating
divergent net soil movement? Field management is not typically based on or
applied to a small portion of a field, but rather it is based on and applied
to the entire field. Grid cells are less than ideal for making field scale
management decisions.
A third formidable problem of the grid cell method relates to referencing the
landscape to the grid cell pattern that was established. While it is a simple
though tedious task to overlay a grid on a topographic map and determine
elevations, slopes and flow directions of cells, it is not a simple task to
accurately locate grid cell boundaries on unrectified air photos, soil maps,
or worse yet, actual landscapes. Though the problem is not insurmountable, it
presents significant difficulty in application.
There are alternatives to using a grid cell pattern. One very attractive
alternative is using actual field boundaries as the basis for or unit of data
collection and processing. The unit of data becomes the farm field. Data
collected at this scale is, not surprisingly, termed field scale data. For
these purposes, a field may be defined as an area of homogeneous land cover,
managed as a unit by the same individual(s), without significant topographic
or soil differences within it. There are several attributes which make field
scale data collection and processing attractive.
The primary advantage of using field scale data is that it involves readily
identifiable, relatively homogeneous, management oriented data units. There
is little ambiguity about the boundaries of a field. They can easily be
identified on air photos, soil surveys, and "in the field". Most
significantly, land cover and land management are homogenous by definition.
Further, field boundaries are often laid out following the contours and
channels of topography and drainage and, as such, they tend to have more
homogeneous slopes and soils then similar size grid cells placed on the
landscape irrespective of naturally occuring features. Where fields do have
significant topographic or soil differences within them, it may be in both the
modeler's as well as the land owner's interest to subdivide them. Field size
also tends to follow the diversity of the landscape. Where slopes are
uniform, fields tend to be larger, and vice-versa. This relation coincides
with the inherent preferences of the modeler, i.e., smaller scale data where
needed and larger scale data where warranted. And of course, by definition,
field scale data is most appropriate for making field management decisions.
The land operator is already operating at this scale. Lastly, this is a
common unit of data gathering, and several other agencies and organizations
T-II-D-5
-------
collect this type of data in a similar format. This common data requirement
and existence of common data files provides for the opportunity of shared use
of common resources.
The use of field scale data has many appealing attributes but, as we shall
see, it necessitates the use of hierarchal data structures.
Hierarchal data structures
The term data structures refers to the method by which data is organized and
processed within a computer model. There are two types of data structures to
be considered here: array and hierarchal.
An array data structure is (in this application) a two dimensional
representation of data with X-Y or row-column coordinates. The array
parallels the grid cell data format, (although a grid cell format, within
which runoff is passed to one adjoining cell rather than two, may as well be
represented in the computer by an hierarchal data structure). The array is
processed sequentially, proceeding row by row, column by column, through each
iteration. Runoff and sediment is routed out of each cell to adjoining cells.
There are two problems with array processing of runoff and sediment routing
data. First and foremost, array processing makes tracking of sediment eroded
out of one cell impractical to separate and track independently as it moves
downslope and merges with sediment from other cells. Instead, the net
sediment movement on each cell is tracked (deposition plus incoming sediment
minus detachment minus outgoing sediment). A simple tally of net increase or
decrease in sediment on each cell is kept, not the eventual deposition or
delivery of sediment from each cell.
The second problem is one of efficiency or waste of computer resources. The
size of arrays must be declared (allocated) before the computer program can
run. Either the arrays must be declared large enough to accomodate the
greatest anticipated cell numbers (how big is the watershed and how small are
the cells?), or the program must be modified repeatedly to fit the individual
watersheds, or the watershed must be subdivided to fit the program. Recall
also that there are several parameters for each cell (slope, land cover, soil,
flow direction, etc.), and that this problem is compounded by the number of
parameters. Wasted allocation can rapidly limit the size of a watershed which
a particular computer can accomodate. ANSWERS cleverly transforms its data
into several one dimensional arrays, which alleviates the wasted allocation
which results from watersheds not being square, i.e., not completely filling
the corners of two dimensional arrays.
An alternative to the array data structure is the hierarchal data structure.
Hierarchal data structure is a method of data storage and access by which each
cell is accessed by its location relative to other cells. It does not have
such absolute designations such as "row and column" or defined locations in an
array. Instead, each cell is accessed only through access to a previous
cell. The location in computer memory of each record is kept in and a part of
another computer record.
T-II-D-6
-------
The use of field scale data necessitates the use of hierarchal data
structures, because the fields obviously are not uniformly rectalInear.
Fields have a variety of shapes and sizes. They do not conform to a grid
pattern either on the landscape or in the computer. Rather than being
accessed by row and column in the computer, they are accessed by their
position relative to other fields.
There are a couple of drawbacks to using hierarchal data structures however.
Probably the most significant disadvantage is that it requires that all runoff
from a field flow to a single adjoining downslope field or channel. Without
some very complicated programing, which would in execution severly hinder
model performance, it is impossible to split the flow of runoff from one field
to two or more adjoining fields or channels. In designating a flow route,
this requires making a selection based on dominant flow, worst case
conditions, or some other criteria. A second less critical drawback is that
field attributes (the input and output of the model) can not readily be mapped
by the computer.
While the use of hierarchal data structures has some drawbacks, there are
several advantages of this type of data structure that makes it the prefered
choice in watershed modeling. To understand why this is so requires a more
in-depth look at hierarchal data structures.
Hierarchal data structure is perhaps more commonly refered to as a tree
structure, because the internal representation of the data approximates the
structure of a tree (Figure 1.). A cell which has other cells flowing into it
is called a root. Each root has one or more cells flowing into it called
branches, each of which in turn may have one or more branches flowing into
it. Each cell is a branch of the root to which it flows, and the root of the
branches which flow into it. All of the cells that flow either directly or
indirectly into a root cell are called daughter cells of that root. In Figure
1., cells B, G and H are branches of cell A, and cell A is the root of cells
B, G and H. All cells other than A are daughter cells of cell A.
Figure 1. Pictoral Representation of a Hierarchal Data Structure
T-II-D-7
-------
This tree structure closely mimics the natural flow pattern of surface runoff,
aggregating from field to field into small intermittent channels, and those in
turn into larger and larger perennial channels.
In a hierarchal tree structure, in order to proceed down the tree, each cell
must have one data record associated with it indicating the cell to which it
flows. This record must be determined beforehand and be included in the basic
data file. In addition, each cell may have one or more cells flowing into
it. To proceed up the tree then, each cell must have a list of which cells,
if any, flow into it. This list does not have to be developed beforehand but
can by built in the computer based on the root record of each of the cell's
branches.
In addition to paralleling natural flow patterns, a significant advantage
using hierarchal data structures may be obtained if the model is programed in
a computer language called Pascal. This advantage results from a feature of
Pascal, known as dynamic data allocation, which allows computer space to be
allocated as needed, eliminates wasted allocation and maximizes computer
storage potential. Incidentally, another outstanding feature of Pascal is its
ability to do something called recursion, which allows a subroutine
(subsection) of a program to reference (call) itself. This feature is a
tremendous asset when using hierarchal data structures.
There is one other strong advantage of using hierarchal data structures, which
is the ability to track sediment from a field to u channel. This feature may
best be described in an overall discussion of modeling sediment and runoff,
using the combined strengths of field scale data and hierarchal data
structures.
Modeling using field scale data and hierarchal data structures
Natural flow patterns may be subdivided into and modeled as two distinct but
integrated subsystems. There can be one hierarchal subsystem representing the
channel segments, and another one representing fields flowing to the channel
segments. In the channel subsystem, a channel segment must flow to another
channel segment, but it may have either fields, channels or both flowing into
it. On the other hand, a field can flow to another field or channel, but it
can only have fields flowing into it, not channels. Clearly the dynamics of
sediment routing operate differently in both subsystems. Also the distinction
serves the purpose of identifying the extent of eroded sediment from an upland
field that is delivered to a channel (as opposed to the mouth of the
watershed). The need for and practicality of this limit will be made clear
shortly.
When modeling a watershed using hierarchal data structures, one begins at the
root of the entire watershed, i.e., the mouth of the main channel. For each
iteration or time segment, the model proceeds up the tree, stopping at each
cell to calculate sediment detachment, transport and deposition, routing the
excess back down to its root, then proceeding up each of its branches, in turn
repeating this process. The cells of Figure 1. for example would be processed
T-II-D-8
-------
1n alphabetical order. Therefore, for a main channel outlet that has two
branches coming into it, one branch, and all of the watershed draining into
it, will be processed during each interation before the analysis proceedes up
the other branch. If the branch in question is a channel, then a group of
paramaters describing that channel, and equations governing channel transport,
can be called in. If it is a field, then the parameters must describe the
field, and the equations governing overland detachment, deposition and
transport must be used. In either case, the excess runoff and sediment is
routed to the cell's root, be it a field or channel, and the cell's branches
(if any) are in turn processed, be they fields, channels, or a combination of
both. By looping through several iterations over time, it is possible to track
the net sediment dynamics in each of the field and channel segments.
If the analysis ended at this point, the major differences between this and
current models would be the units of the input data, the manner of cell
processing, and the units of the output data. The output of the model could,
in a parallel fashion to others, indicate net sediment gain or loss on each
field and channel (not cell), as well as total watershed sediment loading.
However, using the components of this model, it is possible to improve on this
output significantly. With field scale data and hierarchal data structures it
is possible to track the amount of sediment eroded off of a field that is
delivered to the channel network. This is possible because the length of
overland flow prior to channelization is generally relatively short. The SCS
(1972) indicates that it is seldom more than 1000 feet. In the small
watershed that ANSWERS was calibrated on, overland flow was seldom across more
than two or three fields before channelizing. In that watershed, the greatest
number of fields of consequtive overland flow was five. These five fields were
narrow, contour cropped, and on the side of a long gently sloping hill.
This rather rapid channelization of overland flow means that the tree
structures representing fields in the computer are typically quite short,
i.e., there are a lot of them but they do not branch very far. As a
consequence, it is practical to do two things in the computer that would allow
a program to track sediment downs!ope. The first is to, for every source
field, keep a linked list (a one dimensional tree) representing the sediment
in suspension on each field over which overland flow from the source field
must traverse before reaching a channel. Fields adjoining a channel would
have only one record in their linked list. Fields with two intervening fields
between themselves and receiving channels would have three records in their
linked list. Sibling cells (branches with the same root) would have the same
number of records in their linked list. Each of their records would, in
parallel, refer to the sediment originating from their respective cells in
suspension on coinciding downslope cells. The sediment in suspension on each
field at any one time would be equal to the sum of the sediment in the
coinciding cell of the linked list of each of its daughter cells. Again,
using Pascal and dynamic data allocation, the required storage capacity would
be minimized.
The second component necessary to track sediment downslope is a process, not
an item. Whenever sediment dynamics on a given field are calculated, a
T-II-D-9
-------
traversal through each of the cell's daughter cells is required. The purpose
of this traversal is to modify the record on each of the daughter cell's
linked list corresponding to the field in question. On the record of the
linked list corresponding to the field, it would proportionally decrement the
sediment load for deposition (if any) and downslope transport. Then it would
increment the adjoining downslope record for inflow from the upslope record.
When the root cell flows to a channel (as opposed to another field) the
downslope transport represents delivery to the channel. By this means
sediment delivery from a field to a channel can be tabulated. The reason that
this is possible but limited to overland flow is that the computer storage and
execution effort required with each cell to perform these calculations
increases linearly with the number of daughter cell it has. A field seldom
has more than a few daughter cells, but a channel can have hundreds to
thousands.
The above description outlines a proposed flow of control for a suggested
sediment delivery model. References to specific methods for calculating
overland and channel flow and sediment transport are noticeably lacking, but
they can be selected from several in existence.
The data that would be necessary to run the proposed model includes two
detailed types of inventory (field and channel), as well as the selection of
appropriate design events. The field inventory would require field by field
descriptions, including all of the data presently collected in the Piority
Watersheds to estimate erosion using the USLE, plus two additional data
records. One would be the overland flow distance of the field. This, in
conjunction with the field acreage, would be needed to estimate field
dimensions, as flows are usually calculated in unit volume per unit time per
unit width. The second additional data record would be a designation of the
field or stream to which each field flows. Anticipating that at some time in
the not too distant future the Department may procure such a model, the
Nonpoint Source Section is evaluating a method by which these additional
pieces of information may be obtained.
The second part of the inventory consists of a channel inventory. The channel
inventory would require a channel sepent identifier, a physical description
of the channel segment (width, profile, gradient and vegetation if any),
length of the segment, and a designation of the channel segment that this
segment flows to. This would not be difficult for perennial streams, but
could become tricky when applied to small intermittent streams, grassed
waterways, roadside ditches, etc.. The present channel inventory that is
undertaken in Priority Watersheds, mainly to evaluate habitat and biotic
potential, would have to be upgraded significantly to accomodate this data.
CONCLUSIONS
The Wisconsin Nonpoint Source Program requires a model capable of estimating
the sediment load from fields to streams, and sediment transport within
streams, under current and projected field management conditions. This model
T-II-D-10
-------
must accept data that can be collected for large areas in a reasonable amount
of time and with a high degree of reliability. Lastly, the output of the
model must be appropriate for making field scale management decisions based on
water quality objectives. It is argued that the proposed model could
demonstrate all of those attributes. Finally, it is suggested that
appropriate agencies consider funding the implementation of such a model.
REFERENCES
Beasley, D. B. and Muggins, L. F. (1982). ANSWERS Users Model.
EPA-905/9-82-001.
Bosch, D. D., Onstad, C. A., and Young, R. A. (1983). A Procedure For
Prioritizing Water Quality Problem Areas. A presentation at the 1983 Summer
Meeting of the American Society of Agricultural Engineers, Montana State
University, Bozeman, Montana, June 26-29, 1983.
Park, S. W. and Mitchell, 0. K. (1983). MODANSW (A Modified ANSWERS Model)
Users Guide. Dept of Agricultural Engineering, University of Illinois at
Urbana-Champaign.
Soil Conservation Service (1972). National Engineering Handbook, Section 4,
HYDROLOGY.
3622U
T-II-D-11
-------
EUTROPHICATION PROBLEMS IN NORTH CAROLINA AMD
MANAGEMENT APPROACH FOR NONPOINT SOURCES
William A. Kreutzberger, George T. Everett, and Alan Klimek
North Carolina Department of Natural Resources
And Community Development
Raleigh, North Carolina
The State of North Carolina has been experiencing
eutrophication problems in several coastal rivers for the past
decade. These problems are characterized by extensive surface
blooms of blue-green algae during June through September. Two
recently filled multi-purpose reservoirs in the lower piedmont
have also been exhibiting signs of eutrophication although only
minor blooms of nuisance algae have as occurred yet. Extensive
monitoring and research efforts have been conducted in the
watersheds of these eutrophic systems to characterize the
magnitude of nonpoint sources of nutrients and sediment from
urban, agricultural and forested watersheds. The effectiveness
of "best management practices" (BMP's) has also been evaluated.
In an effort to mitigate the occurrence of nuisance blooms
one coastal river and prevent the occurrence of these problems on
the piedmont reservoirs, the State has classified the watersheds
of these systems as "Nutrient Sensitive Waters" (NSW). This
classification provides the regulatory authority to limit the
input of nutrients from point sources. However, agricultural and
silvicultural activities are explicitly exempted from this
regulatory authority. Although urban areas are not exempted from
this regulatory authority, there are limited approaches for
controlling nonpoint nutrient inputs from existing urban areas.
Despite the limited regulatory authority, the State has used
the NSW classification to educate the public, local government
officials, and the General Assembly about nonpoint problems. The
threat of point source nutrient controls has been used to force
local governments to plan and control runoff from new urban
expansion. It was also a major factor in motivating the General
Assembly in 1984 to provide $2.5 million for the voluntary
implementation of cropland and animal operation BMP's in NSW
watersheds. These efforts in NSW watersheds are the first test
case for the effectiveness of nonpoint source planning and
control measures on a large scale in North Carolina.
Keywords; Eutrophication, nuisance algal blooms, nutrient
sensitive waters (NSW), nonpoint source pollution control,
agricultural runoff, urban runoff, future development.
T-II-E-1
-------
INTRODUCTION
Freshwater sections of several coastal rivers in North
Carolina have been experiencing severe eutrophication problems
since the early 1970's. The Chowan River (Pig. 1) was the first
of these rivers to bring public and regulatory attention to
eutrophication problems with the development of extensive surface
mats of blue-green algae. These surface blooms have occurred
nearly every year since 1970 and in 1972, 1978 and 1983 surface
mats covered nearly all of the 30 km long lower river during part
of the summer and fall (N.C. DNRCD, 1982a).
Fig. 1 Nutrient Sensitive Watersheds in N.C.
N. C. inland lakes and reservoirs have a wide range of
trophic conditions. Trophic state classification surveys of
nearly 80 bodies of waters conducted in the mid 1970s and early
1980s indicate that 30 to 40 percent of these lake/reservoirs are
eutrophic or hypereutrophic (Weiss and Kunzler, 1976; N.C. DNRCD,
1982b). Extensive surface blooms of blue-green algae have not
occurred in lake/reservoir systems. However, many water treatment
plants using these systems as raw water supplies periodically
have to modify treatment to minimize taste and odor problems.
Two recently impounded reservoirs, B. Everett Jordan Lake in the
Upper Cape Fear River Basin and Falls Lake in the Neuse River
Basin (Fig.l), are already considered the most eutrophic
reservoirs in N. C. These reservoirs are planned as water
supplies and recreational areas for the rapidly growing
Raleigh/Durham/Chapel Hill area and there is considerable concern
over the impact of severe eutrophication on uses of the lakes
(N.C. DNRCD, 1983a).
The N. C. Division of Environmental Management (DEM)
has developed a watershed classification entitled "Nutrient
Sensitive Waters" (NSW) as a regulatory means to deal with
eutrophication. The NSW classification provides the
regulatory authority to limit inputs of nutrients to
background levels, if necessary. The authority is effective in
controlling nutrient inputs from point source discharges, however
the North Carolina General Assembly has passed legislation which
has exempted agricultural activities (other than concentrated
T-II-E-2
-------
feed-lots) and silvicultural activities from this of the
authority. Urban runoff is not exempted from the DEM authority
provided by the NSW classification but there are limited
approaches for effectively reducing nutrient inputs from existing
urban areas.
The Chowan River watershed has been classified as NSW since
1979 while the watersheds to Falls and Jordan Lakes were
classified NSW in 1983. Point source nutrient controls (nitrogen
and phosphorus in the Chowan River watershed and phosphorus
controls in Falls/Jordan watersheds) are currently being
implemented. Nonpoint source control implementation has
initially been slow, however a special funding program provided
by the N. C. General Assembly is making progress on agricultural
Best Management Practices (BMP) implementation in all three NSW
watersheds. Also, a unique local/state government cooperative
approach seems to be making progress in minimizing runoff
problems from urban expansion as well as showing some potential
to control some existing urban runoff problems in the
Falls/Jordan watersheds.
CHARACTERIZATION OF EUTROPHICATION PROBLEMS
Chowan River
The Chowan River is located in Northeastern North Carolina
(Fig. 1). It is formed by the confluence of the Blackwater and
Nottway rivers near the Virginia/N.C.state line and stretches.
about 80 km south to Albemarle Sound. The watershed is
approximately 12,700 km with 73 percent in the Piedmont and
inner Coastal Plain of southeastern Va. The watershed in North
Carolina is entirely in the coastal plain and the Chowan is
tidally influenced over its entire N. C. length. Despite the
tidal influence, the Chowan River is essentially a freshwater
system except during periods of extreme drought.
The Chowan River can be divided into two district regions.
The upper portion of the river, from the Virginia state line to
Holiday Island (48 km) is a relatively narrow (90 to 1200 m in
width), sluggish river with depths up to 12 m. The lower river
from Holiday Island to Albemarle Sound (32 km) ranges from 1200
to 3000 m. wide and behaves like a shallow (mean depth 4.0 m),
rapidly flushing lake. Surface mats of blue-green algae occur
primarily in the lower river due to the large surface area,
shallow depth, and lake-like tendencies (N.C. DNRCD, 1982a).
Detailed monitoring by DEM has been used to estimate nutrient
inputs to the lower Chowan River from 1977 to 1981 (N.C. DNRCD,
1982a);
Annual Flow Total Nitrogen Total Phosphorus
Lt^/sec) (kg x 10^/yr) (kg x 10^/vr)
Ave Range Ave Range Ave Range
133.7 34-242 40.3 6.8-75.1 4.3 0.8-7.5
T-II-E-3
-------
These estimates indicate that the N:P ratio is in a range
where either nitrogen or phosphorus could be limiting. Detailed
research studies on phosphorus and nitrogen uptake, nitrogen
fixation, and sediment nutrient recycling confirmed that nitrogen
and phosphorus limitations varied both seasonally and as a
function of the phytoplankton community. Phosphorus was
identified as the primary nutrient requiring control since the
major sympton of the eutrophication problem was surface blooms of
nitrogen fixing blue-green algae in the lower river (Kuenzler,
et.al., 1980). A 30 to 40 percent reduction in phosphorus inputs
was recommended based on modelling analyses by DEM (N.C. DNRCD,
1982a) and confirmed by research results (Paerl, 1982). A 15 to
20 percent reduction in total nitrogen inputs was also
recommended since nitrogen was considered a major factor in
occasional blooms of Microcystis spp..a non-nitrogen fixing
blue-green, in the upper portion of the Chowan (Witherspoon and
Pearce, 1982).
The magnitude of blue-green blooms seems to be related to
hydrologic conditions (Pig. 2). The most severe blooms seem to
occur after above average spring tributary flows and moderate
summer flows. If high river flows are sustained through the
summer, the flushing action minimizes blue-green growth.
Likewise, if stream flow is extremely low, insufficient nutrients
are delivered to the lower river to sustain algal growth. For
example, during 1980-81, a 15 month drought limited blue-green
growth in 1980 from a lack of nutrients. By the summer of 1981,
saltwater had moved up the Chowan River above (Holiday Island)
completely eliminating the growth of nuisance species of
blue-green algae (N.C. DNRCD, 1982a).
«•
Ml
* "-
id
S
5 31.
3
S. II.
a oj
SI IS 101 12S ISO IIS 200
CltWM RIVER flOW ( i3/SEC)
Pig. 2. Relationship between flow, residence time,
nutrient delivery, and algal bloom magnitude in the
lower Chowan River(N.C. DNRCD, 1982a).
Palls and Jordan Lakes
Falls and Jordan Lakes are located near the edge of the lower
Piedmont and lower Coastal Plain in N. C. (Figure 3). Falls Lake
T-II-E-4
-------
is a 5000 ha impoundment of the Neuse River which9was filled in
February 1983 and has a watershed area of 2000 km . The reservoir
is 39 km long with a mean depth of about 4 m. Jordan Lake is an
impoundment of two major tributaries of the Cape Fear River - Haw
and New Hope Rivers and was filled in September 1981. The
watershed to Jordan Lake is 4464 km2, the lake area is about 5500
ha, and the mean depth is approximately 5 m.
UNE
SCUE
Fig. 3. Falls and Jordan Lake Watersheds.
There has been considerable controversy over both
impoundments. Both reservoirs are U. S. Army Corps of Engineers
projects for purposes of flood control, recreation, water supply,
wildlife protection and downstream water quality control.
However, due to extensive urban, industrial and agricultural
development in both watersheds the trophic state of the lakes was
a major issue. Also, the suitability of Jordan Lake as a water
supply is still being examined due to the presence of 170
permitted wastewater dischargers as well as extensive urban and
agricultural runoff in the watershed (N.C. DNRCD, 1985).
Water quality monitoring conducted during the first few
years after filling the reservoirs has confirmed some of the
predictions of lake quality. Jordan and Falls Lakes are the most
eutrophic reservoirs in N.C. although the trophic status has not
yet impaired uses of the lakes (N. C. DNRCD, 1984b). Chlorophyll
a, levels in both lakes have at times exceeded 200 to 300 mg/m
However, this has been the result of blooms of phytoplankton
which are not nuisance species (e.g. Chrysoch roroulina spp^.) (N.C.
DNRCD, 1984a)., Chlorophyll a. levels have been averaging between
40 to 80 mg/m during the summers of 1983 and 1984. There seems
to be a trend towards more dominance of some blue-green species
in Jordan Lake, but more time is needed to confirm this trend.
These reservoirs have relatively short average retention times of
7 and 270 days in the Haw and Mew Hope arms of Jordan Lake,
respectively, and 120 days in Falls Lake. Already there have been
several occasions where extreme flows have flushed substantial
phytoplankton population from the lake (N.C. DNRCD, 1983a; Weiss
et.al., 1984; Francisco et.al., 1984).
Table 1 summarizes inflow information and estimated
nutrient inputs to the lakes for mean annual flow conditions.
This information indicates N:P ratios in the range of 5:1 which
T-II-E-5
-------
indicates more nitrogen limitation than in the Chowan River.
The general abundance of nutrients and nitrogen limited
conditions are major reasons for the concern over future problems
from nuisance nitrogen fixing blue-green algal species.
Table 1. Estimated Nutrient Inputs to Falls and Jordan Lakes
Prom Instream Monitoring Results (N.C. DNRCD, 1983) .
Mean Annual Total Nitrogen Total Phosphorus
Watershed Flow (m^/sec (ka x IQ-Vyr) (kg x IQ-Vyrl
Falls 22.5 77.9 13.5
Jordan - total 48.6 227.0 55.6
Haw River Arm 39.3 181.8 45.7
New Hope River
Arm 9.3 45.2 9.4
NUTRIENT SOURCES
Chowan River
The Chowan River watershed is predominantly rural in both
N.C. and Va. It's watershed is 80 percent forested and wetland
areas, 19 percent agricultural and 1 percent urban. The total
population in the watershed was a little more than 200,000 in
1980. Municipalities are generally quite small. The largest
dischargers in Va. are Emporia (0.6 MGD), Blackstone (0.5 MGD),
and Franklin (0.8 MGD), while in N.C. Ahoskie (0.8 MGD) and
Edenton (1.0 MGD) are the largest municipalities. There is one
large industrial discharger in the Va. watershed. Union Camp is
a pulp and paper facility located near the state line which
discharges in excess of 380 x 10bl/day (100 MGD) from holding
ponds only during the months of December through March when the
Chowan River can assimilate the waste. In N.C., there was a
large fertilizer manufacturer (C.F. Industries) in operation
until 1983 which was initially blamed for the occurence of
nuisance blooms in the early 1970s. C.F. Industries did
discharge significant quantities of nutrients in the late 1960s
and early 1970s. This discharge ceased by 1975 and thereafter,
this company contributed nutrients to the river primarily through
air fallout, site runoff and contamination of groundwater. (N.C.
DNRCD, 1982a).
Figure 4 depicts nutrient budgets for the Chowan River
watershed in N.C. (N.C. DNRCD, 1982a). Agricultural runoff
(50%), forest and wetlands (22%), direct precipitation (-5%) and
urban runoff (0.5%) account for about 78% of the phosphorus
inputs from the N.C. watershed. Nonpoint source contributions of
nitrogen are similar, 80 percent.
T-II-E-6
-------
rim
NITIICIN
1,411, Oil RC-VI
FIIESI 1GIIC
IPS NP $
28.3*5,9.1%
FIIEST ACIIC
NPS IPS
22.1% 49.1%
rim
PIISPIMIS
11$, Ml 1C-VI
Fig. 4. Land use/point source nutrient budget for the
Chowan River Watershed in N.C.(M.C. DNRCD, 1982a).
Falls and Jordan Lakes
In contrast to the Chowan River watersheds, portions of the
Neuse and Cape Fear River basins draining to Falls and Jordan
contain some of the most rapidly developing areas in N.C. The
Jordan Lake watershed includes the urban centers of Greensboro,
Burlington, Chapel Hill and the southern part of Durham with a
total population in 1980 of about 475,000 people. There are 10
major municipal dischargers in the watershed releasing nearly 300
x 10 I/day (80 MGD). In addition, there are several major
industrial dischargers and a total of 170 permitted dischargers
in the watershed. Urbanization is continuing at a rapid pace in
this watershed with the projected population for the year 2000 to
exceed 600,000 people.
Development in the Falls Lake watershed is not currently as
extensive as in the Jordan but the area is also growing rapidly.
Major urban areas include Hillsborough and northern Durham while
the northern portion of the Raleigh metropolitan area is also
pushing into the watershed. The 1980 population was about
150,000 people and is estimated to exceed 210,000 people by the
year 2000. There are two major municipal dischargers and a total
of approximately 60 permitted dischargers in the watershed.
Based on analyses of land use and point source data nonpoint
sources account for approximately 45 percent of the nitrogen and
40 percent of the phosphorus into Jordan Lake and 70 percent of
the nitrogen and 57 percent of the phosphorus into Falls Lake
(Figure's 5 and 6). Nutrient inputs seem to be readily
controllable through point source actions. However, the lakes
currently have such high level of nutrients that the systems may
remain eutrophic even if all point source inputs of phosphorus
were eliminated (N.C. DNRCD, 1983a; Francisco et.al., 1984).
Future development in the watershed could also dramatically
T-II-E-7
-------
increase nutrient contributions from urban runoff which are
already substantial.
urn
NITIICEN
IU.MI it -i i
PHUT
SltttCE
30-5%
FIIEST ACIIC
IPS IPS
23.7% 29.1%
Tin i
POUT URI»I PRECIP
SOURCE UPS IPS
42.6% 22.6% 1.8%
X
FBIEST URIC
IPS IPS
7.2% 25.8%
U2,flflfl«6-ri
Fig. 5. Land use/point source nutrient budget for the
Palls Lake Watershed (N.C. DNRCD, 1983a).
TITU
NITINEI
MIMIIIC-VI
P0IIT IRMI PRECIP
SOURCE IPS IPS
54.9% 6.2% 1.4%
FIIEST «Giic
IPS IPS
13.0% 24.5%
T8TU
PIISPNOIBS
(30,001 KG YR
PIIIT URIAI PRECIP FIREST IfiRIC
SIIIICE IPS DPS IPS IPS
59.5% 12.6% 0.6% 4.2% 23.1%
Fig. 6. Land use/point source nutrient budget for the
Jordan Lake Watershed (N.C. DNRCD, 1983a).
NONPOINT AND POINT SOURCE MANAGEMENT
Chowan River
Nonpoint source control initiatives began slowly. When the
Chowan was classified as NSW in 1979, agricultural agencies were
unwilling to admit to any contribution to eutrophication problems
of the Chowan. This unwillingness along with legislation
limiting the ability of the DEM to require nutrient reductions
T-II-E-8
-------
from agricultural and silvicultural operations severely limited
options to reduce nutrients in N.C.
With these restrictions, the State took two major
initiatives; 1) Actions were started to eliminate or minimize all
point source contributions of nitrogen and phosphorus and 2)
Research/education efforts were begun to verify agricultural
contributions to the nutrient enrichment problems, identify the
effectiveness of various best management practices, and to offer
technical assistance for BMP implementation.
Point source efforts were relatively straight-forward in that
all the municipal facilities were small and required upgrading.
Facilities were promoted on the priority list for construction
grants and all facilities began to conduct non-discharging to
land application, wastewater disposal. These systems will all be
operational by 1987. Small industrial discharges were required
to meet total nitrogen and phosphorus limits of 3 and 1 mg/1,
respectively, and several enforcement steps were taken to limit
nutrients in runoff and seepage from the abandoned C.F.
Industries fertilizer manufacturing site. These actions have
eliminated point source nutrient contributions as an issue in the
North Carolina watershed, and have focused attention on
agricultural contributions (N.C. DNRCD, 1982a).
A three-year research study from five watersheds from 200 to
2800 ha in the Chowan basin in N.C. was conducted from 1979 to
1982 (to.examine agricultural runoff) (Humenik et.al., 1983),
The study was not able to quantify before and after effects from
BMP implementation due to the short research time. However,
valuable survey information on existing management practices and
producer attitudes towards nonpoint source controls were
obtained. The information was utilized in developing supportive
publicity, educational and technical assistance programs for the
entire Chowan watershed. Estimated annual yields from the
forested (control) watershed and the four agricultural watersheds
(averaging 50 percent cropland) are shown on Table 2.
Despite the failure of the research to demonstrate BMP
effectiveness, these studies in combination with other State
action helped to involve agricultural agencies in the development
of a nonpoint control program. Also, the study indicated that
estimates of nutrient contributions from agriculture shown in
Fig. 4 may have been somewhat conservative. In addition, survey
results helped to target cropland requiring control measures and
animal operations requiring waste management systems. BMP
effectiveness was demonstrated as part of research in other parts
of N.C. and demonstration farmswere set up to demonstrate
structural BMPs , provide education on fertilizer application,
serve as advocates for soil testing, and demonstrate animal waste
management systems.
T-II-E-9
-------
Table 2. Sediment and Nutrient Yields from Five Watersheds in
the Chowan River Watershed (Humenik et.al., 1983).
Annual Yield (kg/km-^-
Parameter Forested Watershed Agricultural Watersheds
Aye. Range
Suspended Sediment
Total Nitrogen
Ammonium-Nitrogen
Nitrate Nitrogen
Organic Nitrogen
Total Phosphorus
13,700
135
8.8
17
101
13
71,940
627
28
275
303
114
4972-400,400
110-1331
1.1-73
6.6-672
84-1071
19-351
At the end of 1982 and early 1983, this put the agricultural
control program in a position to begin implementation. Based on
the overall nutrient reduction goals of 30 to 40 percent for
phosphorus and 15 to 20 percent for nitrogen from watershed areas
in N.C. and Va. as well as the reductions being obtained from
point sources, goals of 30 percent effectiveness for phosphorus
and 20 percent effectiveness were set for the agricultural
control program (N.C. DNRCD 1982a). However, there was little
or no special funding to encourage implementation of BMPs so the
program was entirely voluntary. The implementation program had
to rely on existing staff from the N.C. Agricultural Research
Service and N.C. Soil and Water Conservation Districts to provide
technical assistance. The capability of the N.C. Department of
Agriculture to greatly expand soil testing in the Chowan
watershed was also insufficient. Cost sharing funds from the
ASCS also had to be relied upon for structural BMP
implementation.
Falls and Jordan Lakes
When Falls and Jordan Lakes were classified as NSW in October
1983, this clearly indicated the intent of DEM to regulate point
sources of nutrients beginning with phosphorus. Initial policies
for the lake were to limit all new facilities to a total
phosphorus limit of 1 mg/1 and encourage land application
wastewater treatment systems whenever possible. This applied
primarily to rural subdivision developments on both watersheds
but also is applicable to any expansion of existing wastewater
facilities. Existing facilities in the New Hope arm of Jordan
Lake were issued compliance schedules to meet a 1 mg/1 phosphorus
limit in 3 years. All other facilities were issued letters
indicating that the watershed was NSW and nutrient limits might
be established in the future.
Clearly substantial phosphorus reductions to both lakes could
be accomplished based on budgets in Figure 5 and 6 by requiring
phosphorus removal at point sources. However, contractors
studying the lake for the Corps of Engineers have expressed
doubts over the estimated nutrient contributions from point
T-II-E-10
-------
sources in the Jordan watershed, the effectiveness of phosphorus
reductions for controlling eutrophication, and the need to
control eutrophication since no uses are currently being
impaired. (Francisco et. al., 1984). This opinion was considered
in developing the initial control strategy for point sources.
There has been little disagreement that actions are necessary to
encourage proper planning to minimize future nonpoint source
contributions of sediment, nutrients and toxicants. All three
pollutant categories are a major concern relative to water supply
usage of the lakes.
Public concern over the future of these lakes put state
agencies in a favorable position to get cooperation from local
governments to protect the lakes. The threat of point source
controls on phosphorus has also assisted in getting cooperation
from government groups in the upper part of the watershed which
do not get recreational and water supply benefits from these
lakes. Committees were formed with state and local officials
and, with assistance from DEM staff and local/regional planning
groups, guidance was developed for local actions to protect the
lakes from future nonpoint source inputs. Essential elements of
these action plans are as follows;
Local sedimentation and erosion control plans,
County soil and water plans for high erosion areas
(cropland),
- Requirement of storage of first 1/2" runoff from new
development,
50 foot stream buffers,
Designation of water quality critical areas near the lakes
6% impervious limit
Restricted commercial/industrial development
No sewer extentions,
12% impervious limit for unsewered, non-critical areas,
- 30% impervious limit for sewered, non-critical areas,
Designation of limited industry areas.
In addition to these efforts to minimize nonpoint source
impacts from new development, DEM is continuing to investigate
means to reduce runoff pollutants from existing urban areas.
Using loading rate information from the Nationwide Urban Runoff
Program study in Winston-Salem, N. C. and nationwide, portions of
the urban areas have been targeted for controls(N.C. DNRCD,
1983b) . Stormwater detention basins for the 30 percent of urban
areas with highest pollutant yields could reduce metals inputs
from urban runoff by as much as 50 percent and nutrient inputs by
20 to 30 percent (N. C. DNRCD, 1985). This information is being
provided to local planning agencies and funding is being
considered for urban detention basin demonstratJon projects.
NSW LEGISLATION
In June 1984, the N. C. General Assembly approved special
funding for the implementation of nonpoint source controls in NSW
watersheds. Approximately $2.5 million/year was provided to the
county soil and water conservation districts through the N. C.
Division of Soil and Water. The bulk of this funding will be
T-II-E-11
-------
used for 75 percent cost-sharing of animal waste management
systems and cropland BMP's. A portion of the funding was also
used to hire an additional staff member for each of the counties
in NSW watersheds to provide technical assistance.
During the first six months of its existance this program
concentrated on publicity and getting agricultural producers
interested in the program. A good portion of the cost-sharing
funds were used for animal waste management. Specific targeting
of BMPs for high erosion rate cropland has not occurred during
the first year since it takes some time to get producers involved
in this type of voluntary/cost incentive program and it is
important for continued funding to utilize all funds provided.
Targeting of highly erodable cropland for BMPs is planned for the
second year of implementation.
In addition to agricultural cost-sharing funds, additional
funds were provided for enforcement of state sediment and erosion
control laws. Also, special research funding is available to
local governments to investigate means to control nutrients from
municipalities. Research will likely center on biological means
to reduce phosphorus at existing wastewater facilities. However,
these funds may be used in the future to investigate the
effectiveness of detention ponds for urban nonpoint source
pollution control.
FUTURE DIRECTIONS
The NSW watersheds are the first test case for the
effectiveness of nonpoint source planning and control measures on
a large scale in North Carolina. At present, initial nutrient
control needs have been established and a program has been
developed for agricultural land use. A program is being further
developed to control urban runoff. For the Chowan River Basin,
the State of Virginia has just developed a plan for point and
nonpoint source controls for protection of their watershed and is
beginning to be implemented.
The future success of the nonpoint control programs in NSW
watersheds depends on continued cooperation between state and
local agencies, continuing legislative funding, and
research/study efforts.
LITERATURE CITED
Francisco, D.E., C.M. Weiss and P. H. Campbell (1984). Jordan
Lake Water Quality in 1984. Presentation at the N. C. Water
Control Federation and American Water Works Association
Conference. Raleigh, N.C. November 13, 1984.
Humenik, F.J., B.A. Young, and F.A. Koehler (1983). Agricultural
Nonpoint Source Case Studies in N.C. Ill: Chowan River
Priority Watershed. Dept. of Biol. and Agric. Engineering,
North Carolina State Univ., Raleigh, N.C.
Kuenzler, E.F., S. Mozley, H. Paerl, and A.M. Witherspoon (1980).
Chowan River Nuisance Algal Bloonms and Probable Causes - A
Position Paper Prepared for the N.C. Dept. of Natural
Resources and Community Development. UNC WRRI, Raleigh, N.C.
T-II-E-12
-------
N. C. DNRCD (1982a). Chowan River Water Quality Management Plan.
Division of Environmental Management, Raleigh, N.C.
N. C. DNRCD (1982b). North Carolina Clean Lakes Classification
Survey. Division of Environmental Management, Raleigh, N.C.
N. C. DNRCD (1983a). Water Quality Discussions of Falls of the
Neuse and B. Everett Jordan Lakes. Division of Environmental
Management, Raleigh, N.C.
N. C. DNRCD (1983b). Nationwide Urban Runoff Study in
Winston-Salem, N.C. Division of Environmental Management,
Raleigh, N.C.
N. C. DNRCD (1984a). 1983 Annual Summary of Phytoplankton in B.
Everett Jordan Reservoir. Division of Environmental
Management, Raleigh, N.C.
N. C. DNRCD (1984b). Ambient Lakes Monitoring Report - 1983.
Division of Environmental Management, Raleigh, N.C.
N. C* DNRCD (1985). Toxic Substances in Surface Waters of the B.
Everett Jordan Lake Watershed. Division of Environmental
Management, Raleigh, N.C.
Paerl, H. (1982). Environmental Factors Promoting and Regulating
N9 Fixing Blue-Green Algal Blooms in the Chowan River, N.C.
UNC WRRI, Report No. 176, Raleigh, N.C.
Witherspoon, A.M. and R. Pearce (1982). Nutrient and
Multi-species Criteria Standard for the Chowan River, N.C.
UNC WRRI, Report No. 187, Raleigh, N.C.
Weiss, C.M. and E. J. Kuenzler (1976). The Trophic State of
North Carolina Lakes. UNC WRRI, Report No. 119, Raleigh,
N.C.
T-II-E-13
-------
REDUCING BACTERIAL NON-POINT POLLUTION
IN TILLAMQOK BAY. TILLAMOOK. OREGON
James A. Moore
Associate Professor
and
J. R. Miner
Professor
Agricultural Engineering Department
Oregon State University
Corvallis, Oregon, USA
ABSTRACT
High coliform counts in the waters of Tillamook Bay closed oyster
harvest. A review of possible sources of pollution revealed that sewage
treatment plants, failing on-site domestic septic systems and the
county's 18,000 dairy cows were the primary candidates.
Extensive water sampling confirmed all three contributed. Major
improvements were made in the sewage treatment plants and the on-site
systems but the nonpoint contribution from the 100 plus dairy operations
proved to be challenging.
A computer program was developed to evaluate management practices and
predict the movement of organisms from the cow to the Bay. This
assisted dairy operators in selecting the specific component to improve
or add to their unique operation to make the greatest positive impact on
water quality. State and federal agency personnel also used the model
to identify those operations with the greatest need and distribute
federal Rural Clean Water Program cost share monies to assist in
upgrading facilities.
Several local citizens' groups, clean water committees, state and
federal agencies worked with the oyster growers and the dairymen toward
the common goal of improved water quality. The program is still
underway but recent sampling suggests an improvement in Bay water
quality.
T-II-F-1
-------
INTRODUCTION
There are many documented water quality problems in the United States.
Over the past 15 years nutrients, primarily nitrogen, have received the
greatest attention as pollutants. Currently several cases of bacterial
pollution from non-point sources have drawn attention to these problems
and to the general lack of research effort in this area.
Organisms have been responsible for closures of oyster harvesting areas
in a number of bays and estuaries. One such closure occurred in
Tillamook Bay in 1977 (1). Tillamook Bay is on the Pacific Coast in the
northwest corner of Oregon (Figure 1).
This closure brought attention to the problem and focused the efforts of
several state and federal agencies and local groups. Water quality
sampling conducted by the federal Food and Drug Administration (FDA) (2)
and the state Health Division (HD) and Department of Environmental
Quality (DBQ) during the seventies showed that bay water quality did not
meet standards of the National Shellfish Sanitation Program (NSSP) (6)
(3). In general these studies showed that a potential hazard did exist
as high levels of total and fecal coliform bacteria were present in
oyster beds regardless of weather and tidal conditions.
From reviewing the studies and input from involved agency personnel, •
three major sources of bacterial pollution have been identified. These
are the five sewage treatment plants in the watershed, failing on site
domestic sewage systems and livestock wastes. A two step plan was
proposed to address the pollution from these sources. First to reduce
the bacterial input by implementing technical and institutional options
such as improved sewage treatment, land zoning, and installing best
management practices on the dairy farms (5). The second part was to
develop a bay closure scheme when failing equipment or heavy rainfall
would provide conditions that would excess the sanitation standards set
by the NSSP.
The remainder of this paper will address the efforts to reduce those
organisms originating from the dairy operations. There are some 18,000
dairy cows on 118 farms in the watershed. These dairies support the
Tillamook cheese plant which converts 25% of all the milk produced in
Oregon into cheese. The coastal marine climate brings 150 inches of
rainfall to the upper part of the watershed and 90 inches of
precipitation in Tillamook. Most of the rain falls during the winter
while the summers are dry. Within the project area there are 23,500
acres of agricultural lands. The dairies which occupy almost all of the
agricultural land are located on the tidal flood plains and river
terraces. These flat uplands are dedicated to pastures that are utilized
by grazing during the dry 6 months of the year. Cows' access to the
streams for watering purposes provide opportunity for fecal organisms
contamination during the summer pasturing season. Manure storage and
spreading are the waste management practices of concern during the winter
season, when cows are confined in and around the buildings.
T-H-F-2
-------
H
I
u>
111
o
o
SEWAGE TREATMENT PLANT LOCATIONS
0123
-*. I
Figure 1. Sewage treatment plant locations on the five rivers in
Tillamook Bay, Tillamook, Oregon.
-------
Personnel front the Soil Conservation Service (SCS) and Oregon State
University (OSU) provided leadership in a group which developed a list of
best management practices. Many of the practices on the list have been
proven effective in reducing nutrient pollution but were only speculated
to be effective in reducing bacterial pollution.
A research project was funded to assist in evaluating manure management
practices and help operators select effective practices to reduce
organism movement and escape. This research developed a computer model
to evaluate manure management practices. The evaluator was developed
using coefficients in the literature to estimate die-off, losses and
movement in runoff. This model follows organisms on a daily basis,
from the cows through manure handling to land spreading and possible
escape in runoff.
A flow chart of the model is shown in Figure 2. The dairyman inputs the
cow numbers, manure collection techniques and other farm specific
information. Organims die-off begins in storage and varies slightly
between dry and liquid storage systems. If the operator elects to spread
manure on his pasture land the application method and rate influences the
number of organisms available to runoff. Seasonal environmental
conditions influence the die-off rate of bacteria both in the storage
system and on the land surface.
Twenty-nine years of weather records were evaluated to generate a wet,
dry and average year in order to evaluate different management
practices. Utilizing the daily rainfall and irrigation as hydraulic
input and interial drainage as outflow, a moisture balance is kept for
lands receiving manure. Once a precipitation event occurs the antecedent
moisture conditions and field soil characteristics are utilized to
calculate if and when soil saturation will occur and runoff will begin.
A percentage of available bacteria are moved into the soil with
infiltration and are considered removed from the system in terms of
impact on surface water quality. When saturated soil conditions occur a
first order kinetic relationship is utilized to calculate the bacteria
that are moved overland in the runoff water. If the operator is using a
buffer strip a reduction of the transported bacteria occurs. The reader
is directed to a paper entitled "Modeling Dairy Waste Management Systems'
Influence on Col i form Concentration in Runoff" for details of the model
Using the model on a large number of farms generated the following
guidelines for reducing the bacterial pollution potential of dairy
operations and land applied wastes:
(1) Storage of wastes will increase management options and reduce
pollution of surface waters.
(2) Installing tile will improve soil drainage, reducing runoff
events and the movement of bacteria.
(3) Save well drained soils for waste receiver sites during high
rainfall periods.
T-II-F-4
-------
Steps
animal wastes
defecated &
collected
storage bac-
teria die-off
wastes with-
drawn & spread
land surface
bacteria die-
off
infiltration of
bacteria
quantity of bacteria
in runoff unchanged
Input
Condition/Management
input herd
size & ragmt.
practice
application rateI
spreading method |
function of
seasonal en-
vi ronmental
factors
characteristics
antecedent.
soil moi K t u r e
slope and
crop ruiidit ion-
quantity oi bacteria
in runoff reduced
Figure 2. Flowchart of daily bacterial transport model.
T-II-F-5
-------
(4) Utilize buffer strips between land spreading areas and streams.
(5) Use low spreading application rates over large areas when
possible.
(6) Reduce barnyard runoff by minimizing lot area, guttering
buildings and diking. This source is the greatest pollution
potential. This means keep the clean water clean by rerouting
and collect and hold dirty water.
Obviously management is a critical ingredient of any successful water
pollution reduction program. Utilizing windows or dry weather periods
when they occur to spread manure is perhaps the most important management
choice made by dairymen.
The many positive actions of those involved in the project; agency
personnel, local clean water committees, oystermen and dairymen were
instrumental in presenting a successful grant application to the federal
Rural Clean Water Program. This brought over 2 million dollars of cost
share funds into the program to assist dairymen in constructing and
installing best management practices.
Some 69 operations have signed contracts to implement best management
practices as spelled out in their farm plan. By the end of November .1984
some 60 dairy operations had spent over $3,690,000 in construction costs
to install new pollution control facilities. Approximately two-thirds of
these funds were cost share funds provided by the federal government.
The model and a broad based evaluation system was used to evaluate every
dairy which signed up for assistance and cost share funds in order to
establish where the funds should be spent to make the greatest positive
impact. The model was also utilized to select which practice(s) or
component(s) should be installed to have the greatest contribution in
reducing bacterial movement to surface waters. Table 1 shows how each
component can be ranked when evaluated using the computer model. As
suspected, the precipitation level influences the effectiveness of
management components. That is, drain tile will be more effective during
a high rainfall year, etc.
On site farm visits and educational programs were conducted by members of
the DEQ, SCS and Extension Service personnel. Publications have been
written specifically for dairymen in this project area. Once the project
was underway tours of newly installed practices and facilities were
conducted. Local, state, and national press were utilized to tell the
story of the efforts and results of the project.
The program is a success in terms of moving the dairymen to improve their
facilities. The program began in 1980 and while much construction has
been accomplished, the entire program is not expected to be complete
until 1990.
T-II-F-6
-------
Table 1. Model comparison of different management procedures (each examined using a 100 acre field of pasture during the winter
with the specified characteristics).
H
I
M
M
I
"ij
Example
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 •
17
18
19
20
Storage
(days)
0
0
0
20
20
20
40
0
0
0
0
0
0
0
0
0
0
20
20
0
Management
System
D=dry;
S=semi ;
L=liquid
D
S
L
D
S
L
D
0
D
S
L
D
0
S
L
D
L
D
L
D
Application
Rate
(tons/ac)
10
10
10
10
10
10
10
40
10
10
10
10
10
10
10
10
10
10
10
400
Buffer
Strip
(yes.no)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
Y
N
Drainage
Tile
(yes.no)
N
N
N
N
N
N
N
N
N
N
N
N
Y
Y
Y
V
Y
Y
Y
N
Soil Type
(loam, clay)
L
L
L
L
L
L
L
L
C
C
C
L
L
L
L
C
C
C
C
C
Avg. ppt.
Net Runoff
F£
(xlO9 org)
1,848.5
1.388.2
1,829.9
440.0
330.3
453.2
154.3
7,393.8
2.128.8
1,598.6
2.157.8
4,261.2
838.3
616.8
696.3
1.697.9
1,624.9
440.5
439.7
212.873.0
Dry ppt.
Net Runoff
FC
Rank (xlO9 org) Rank
15 1,186.8 3
10 891.1 1
14 1,188.1 2
4
2
6
1
19
16
11
17
18
9
7
8
13
12
5
3
20
Heavy ppt.
Net Runoff
FC
(xlO9 org) Rank
2.685.6 3
2,016.8 1
2.647.0 2
-------
The ultimate success of this program will be measured by sampling the
water quality in the bay. A comprehensive sampling program has been
planned to begin in the fall of 1985. Routine samples to date indicate
an improved quality in the streams and bay.
T-II-F-8
-------
REFERENCES
Anonymous (1979). Tillamook Bay bacterial study; work plan. Department
of Environmental Quality, Water Quality Division. 75 pp.
Anonymous. (1981). Tillamook Bay Drainage - Oregon Basin Fecal Wastes
Management Plan. Oregon Department of Environmental Quality, Waste
Quality Division.
Carr, V. E. and staff. (1976). Tillamook Bay, Oregon: Pollution source
evaluation with classification and management considerations - May 1976.
Prepared by: Northeast Technical Services Unit, Davisville, R.I., for The
Department of Health, Education and Welfare, Public Health Service and
the Food and Drug Administration - Shellfish Sanitation Branch. 66 pp.
Carr, V. E., Furfari, S. A., and Miescier, J. J. (1978). Sanitary
survey of shellfish waters - Tillamook Bay. Oregon. November-December
1977. Prepared by: Northeast Technical Service Unit, Davisville, R.I.,
for The Department of Health, Education and Welfare, Public Health
Service and the Food and Drug Administration - Shellfish Sanitation
Branch. 143 pp.
Moore, J. A., Grismer, M. E., Crane, S. R. and Miner, J. R. (1983). .
Modeling Dairy Waste Management Systems' Influence on Coliform
Concentration in Runoff. Transactions of the ASAE. 26. (4), pp. 1194-1200.
Oster, Dennis (1975). Land and water use guidelines for development of
the Tillamook Bay estuary. Tillamook Bay Task Force and Tillamook County
Board of Commissioners. 252 pp.
T-II-F-9
-------
Abatement of Nonpoint Pollution of Semi-arid
Streams Using Livestock Grazing Management, Vegetation,
Instream Flow Structures, and Beaver
Quentin D. Skinner, Jerrold L. Dodd, J. Daniel Rodgers, Michael A. Smith
Associate Professors, Department of Range Management, University of
Wyoming, Laramie, Wyoming, USA
ABSTRACT
An important but often ignored contribution to nonpoint pollution in the
semi-arid western United States is sediment. Accelerated erosion and
sediment in stream flow is often thought caused by livestock grazing on
rangeland in less than excellent condition. Erosion of watersheds is
often recognized by stream channel condition. Poor condition streams
are often downcut and show active bank rilling, bank sloughing, and
piping. Lateral movements of gullies away from downcut streams through
upland ecosystems are, in addition, contributing sources for increased
sediment pollution. Equally important, when downcut streams occur,
associated riparian vegetation is often reduced in areal distribution,
disappears, or is replaced by pseudoriparian plant species. Loss of
riparian vegetation and the hydrologic support of these limited but
valuable ecosystems may further reduce stream channel stability and
sediment trapping efficiency of nonpoint source pollution transported to
the stream by overland flow and ground water.
This paper addresses theory and methodology being applied in two
research programs on perennial and ephemeral streams designed to: 1)
promote water storage and vegetation yield, 2) control nonpoint source
pollution, 3) reverse desertification of riparian zones and, 4) advance
the state of knowledge for management of stream side ecosystems
subjected to user pressure.
Livestock grazing treatments, instream flow structures, willow
management, and beaver activities are being used to manipulate stream
flow velocity. Reduced velocity causes sediment deposition on channel
banks where it is then stabilized by vegetation. Encroachment of stream
banks should cause aggradation of stream channels, promote water
spreading over or under dry flood plains, and increase area and health
of riparian zones.
Key words: Nonpoint Pollution, Sediment, Riparian, Desert, Streams, Best
Management Practices
T-II-G-1
-------
INTRODUCTION
Riparian Zones
Riparian zones are wetlands supported by a high water table because of
proximity to surface or subsurface water. They are characterized by
distinct soils and plant communities with high productivity as well as
species diversity. They normally occur as an ecotone between more xeric
lowland communities and aquatic ecosystems (Brown 1978). Strictly
speaking, riparian zones are dominated by plant species that depend on
sub-irrigation throughout their growing season. When sub-irrigation
does not occur riparian communities are replaced by more xerophytic
communties.
Importance
The riparian zone has a high level of use by a wide variety of
interests as noted by Busby (1978), Johnson (1978), Tubbs (1980), Thomas
et al. (1978), Haugen et al (1980), and numerous others. Apparent
causes for concentration of multiple uses in riparian zones are the
vegetation species diversity, productivity, and proximity to open water.
High species diversity in the riparian zone is reported by Campbell and
Green (1968), Brown et al. (1978), Ewel (1978), Haugen (1980), Speck
(1981), and Kauffman et al. (1983a). Diversity is associated with zones
of variation in soil moisture (Miller 1979). Higher plant biomass
production due to increased soil moisture is described by Thomas (1978),
from research by Minore (1970), Minore and Smith (1971), and Miller
(1979). Flowing water wetlands have higher rates of gross primary
productivity and net biomass production than still water wetlands (Brown
et al. 1978).
Riparian areas provide a buffering ecotone between the aquatic and
dryer lowland and upland ecosystems. The beneficial effect of this
ectone to stream condition, water quality, and fish populations has been
reported by Odum (1978), Jahn (1978), and Haugen et al. (1980). Water
quantity and quality as well as physical features of streams also depend
on characteristics of the watersheds drained by each stream (Jahn 1978).
Characterization of watersheds and riparian zones is therefore of
importance in evaluating user impacts.
_User_
A complex interrelationship exists between soils, water, and
riparian vegetation (Miller 1979). Impacts compacting the soils, such
as livestock grazing and recreation, may reduce infiltration, lower
water tables, and increase erosion (Peterson 1950, Schmidly and Ditton
1978, Meehan and Platts 1978, and Thomas 1978). Methods to increase
water harvest on channelized streams will also affect the soil moisture
regime and vegetation (Peterson 1950, Jahn 1978, Campbell 1970, McCall
and Knox 1978) through dewatering the riparian zone. Such impacts as
loss of plant vigor and changes in species composition in plant
communities result from livestock grazing and trampling, roads, mining,
and recreation (Peterson 1950, Gunderson 1968, Martin 1978, Thomas 1978,
Schmidly and Ditton 1978, and Speck 1981). The loss of capacity to bind
T-II-G-2
-------
soil from a decrease in vegetation increases the potential for erosion,
loss of stream channel stability, and results in sedimentation within
channels.
Peterson (1950), Busby (1978), Meehan and Platts (1978), Thomas
(1978), Roath (1982), Kauffman et al. (1983b), and Kauffman and Krueger
(1984) indicate an increase in erosion has resulted from livestock
grazing, water development, and recreational use of riparian zones.
Channel instability indicated by sloughed stream banks results in wider
channels, more shallow water depths and altered flow characteristics
(Platts 1981). Gunderson (1968), Duff (1978), Platts (1981) and
Kauffman et al (1984b) reported channel widening due to impacts of
livestock grazing. Haugen et al. (1980) included mining, water
development, and timber harvest as user impacts that altered stream
channels. Sediment from erosion results in a loss of habitat for
aquatic organisms such as invertebrates, reduces fish production, and
alters water quality (Thomas 1978, Duff 1978, Gunderson 1968, and Platts
1981). Thus impacts to streams caused by users of riparian zones have
been well documented.
This paper addresses two research programs designed to improve
riparian zones of an ephemeral and a perennial desert steppe stream by
1) promoting water storage and vegetation yield season long, 2)
controlling nonpoint source pollution and, 3) advancing the state of
knowledge for management of stream side ecosystems subjected to user
pressure.
Theory for Reclamation of Cold Desert Steppe Streams
Reclamation of degraded streams should be initiated on hydraulic
units where treatments will most likely reverse impacted channel
conditions. Criteria for selecting hydraulic units should allow
experimental replication on similar stream reaches or between streams.
Our study sites are selected based on seven geomorphologic and hydraulic
criteria. The criteria used are as follows: 1) the stream reach
gradient must be low, 2) the stream reach should have a mature meander
pattern, 3) the channel morphology of the stream should have a developed
or developing flood plain, 4) stream flow should be losing to
surrounding alluvium during high flow, 5) channel damming within the
selected stream reach should cause maximum response upstream, 6) a high
potential exists for water spreading to adjoining flood plains as
reclamation progresses and, 7) the stream reach selected should be
typical of those found and similarly managed for grazing in the
immediate area.
_S tream_ Grad ient
The success or failure of reclaimation of degraded stream channels
and adjoining riparian zones depends on deposition of sediment on stream
banks or bottoms if dams are constructed. Channels are usually wide
compared to mature natural conditions. Streams with high sediment load
usually cross areas with steep gradients but lose sediment in areas of
low gradient because of a drop in flow velocity. If one wishes to
achieve maximum deposition of sediment for reclamation of stream
T-II-c-3
-------
channelsi reclamation should start on reaches with the lowest gradient
along the stream course.
Stream Maturity
Mature stream reaches are those which meander from valley wall to
valley wall through a developed flood plain of deposited alluvium.
Maturity is associated with low stream gradient. Alluvium is present
because of a decrease in flow velocity and meandering is a natural
physical flow response across low gradient areas. Because of
meandering, valley flood plain width is maximum. Meandering also
increases the length of stream per lenght of valley floor. This
increases the contact of water with banks and the length of time of
exposure of water to a valley reach. Reclamation of degraded streams on
reaches where meandering occurs should subject more land area to
flooding and promote riparian habitat.
Flood Plain
Flood plains along a stream course reflect response of flow regime
to channel width. Degraded desert streams that have been downcut tend
to widen because of wet-dry cycles that cause bank slumping. Piping and
bank erosion also add sediment to the channel bottom. Slumped banks and
eroded sediment deposited on the channel bottom are flushed downstream
during periods of high flow. This flushed sediment is often more than
is contributed to stream flow by scouring of channel banks (Leopold et
al. 1964) As downcut channels widen, streams during low flow begin to
meander within the confinement of high banks and deposition of sediment
occurs on low banks.
Stream dynamics and channel morphology related to sediment
transport and delivery are generally reviewed in Leopold et al. (1964),
Morisowa (1968), Graf (1971), ASCE Task Committee (1975) and Simons and
Senturk (1977). Hydraulics of sediment transport in stream flow is
further reviewed by Parker (1978a,b). Andrews (1983) studied bedload
transport and deposition within a stream in western Wyoming and
clarified the relationship between flow regime and bank deposition of
sediment. Shumm (1963) was able to show trapping of fine sediments
by vegetation on banks in eastern Colorado. Andrews (1983) work related
to Shumm's by showing that build up of bank sediment occurs before
stream bottoms rise. During periods of low flow sediment is first
deposited on banks. Built up banks cause the channel to become
narrower. Bed load is then moved dowstream by increased flow velocity
because of the new narrow channel. If aboveground and belowground
biomass does not solidify the new bank sediment before the next high
flow event, bank accumulation will be transported downstream as
suspended sediment or bed load. This response to high flow maintains
the channel at a given width and depth corresponding to a yearly average
flow regime. The width and depth of a stable channel however
corresponds to the mean average bank full flow (Andrews 1983). Stable
banks are not eroded because high flows overtopping stable banks and
dissipated over a well vegetated flood plain reduces flow velocity.
This hydraulic process is explained in part by Parker (1978a,b).
Andrews (1983) points out that although a stream maintains an average
T-II-G-4
-------
width, it will move laterally from year to year thus fitting Leopold and
Lanbein's (1966) description of meandering streams. The presence of
undercut banks along stable stream systems are evidence of lateral
movement of meanders.
Bank first sediment deposition, vegetation encroachment into the
new bank sediments, and reduced flow velocity by dewatering or ponding
are important points to consider when reclaiming degraded streams.
Vegetation which is allowed to become well established within newly
deposited sediment on banks creates channel bank stability and a
roughness factor to decrease flow velocity during flood stages. Reduced
flow velocity will cause flood waters laden with sediment to deposit
some of its load. Sediment dropped on vegetated banks will be further
stabilized by regrowth of plants. This process will continue to narrow
a channel until it adjusts to a mean width sufficient to meet the annual
flow regime and is often observed in desert streams within the
confinement of high banks. When this happens the channel is mature with
a developed and stable flood plain.
Reclamation of downcut streams in areas where flood plains are
developed and well vegetated can most effectively be accomplished in low
gradient stream reaches. Meander patterns can further be broken down
into straight and meander reaches. Manipulation of stream flow by
darning to decrease flow velocity should be implemented on straight
stream reaches. This reduces the chance of losing dams because of
channel lateral movement along the meander's concave curve. In
addition, belowground and aboveground biomass offer resistance to
erosion and flow velocity when instream flow structures are utilized to
create ponding and entrapment of bed load.
Lpsing_Strearns
Stream channel transmission loss of water downstream during high
periods of flow to surrounding alluvium should cause a decrease in flow
velocity and consequently a loss in sediment. Sediment deposition should
be maximized in mature stream reaches in areas of low gradient where
meandering increases time in travel of flow downstream. Transmission
loss of stream flow in desert streams has been shown by Lane (1970).
Loss of stream flow should be maximum near the mouth of a drainage basin
or in larger basins of losing desert streams because of increased travel
time of water during any one runoff event (Glymph and Holtan 1969).
Location of manipulative practices to reclaim degraded streams based on
loss of flow and sediment deposition should therefore be placed to
maximize water travel time after considering the previously discussed
criteria.
JDamming
Damming by instream flow structures, like check dams or trash
collectors, and biological damming by beaver or constrictive channel
dams caused by encroaching banks and riparian zones may cause: 1)
reduced flow velocity, 2) stable bed load and 3) storage of water in
banks proximal to the dam. Heede (1978 and 1982) discusses reclamation
of gullies by raising a local base level of an ephemeral stream reach to
decrease gradient slopes upstream using a variety of check dams. The
T-II-G-5
-------
lower gradient reduces sediment transport. Deposition occurs upstream
in a wedge shape until reaching an elevation change upstream equal to
the elevation of the dam. Following Heede's 1978 and 1982 research,
dams should be placed downstream just above a tributary junction. To
achieve restoration of riparian habitat however, the dam should also be
located on a stream reach having a low gradient where meandering occurs
and a stable floodplain exists. The dam will then cause bank deposition
of sediment and maximum channel filling within the upstream drainage
network.
Water spreading across flood plains formerly left dry because of
stream downcutting and occupied by pseudoriparian vegetation decreases
flow velocity. Shrubby type vegetation may cause sediment to be
deposited in rows downstream from woody plants parallel to the main
stream channel. These rows of deposited sediment may form many smaller
channels (braiding) which may carry water during high flow events.
Braiding may also occur in wide channels when vegetation establishes on
the channel bottom and sediment deposits downstream. Once braided
channels become stable because of vegetation, multiple check dams can be
established to maximize water spreading over larger areas.
Water spreading increases riparian zone area by promoting return
flow as runoff and ground water to the main channel. Where degraded
streams have downcut, widened, and formed a mature flood plain within
high banks, water spreading is limited. However, constrictive damming
by encroached riparian zones into a main channel, instream flow
structures, and beaver dams often cause peak flows to flood over high
banks upstream thereby creating desired riparian habitat. Wider
alluvial valley floors are desirable to maximize water spreading.
jGrazing JJ.espj)nse_
Knowing how to maintain riparian zones at desired conditions when
used by livestock and wildlife necessitates accumulating verifiable
information on: 1) how behavior of animals using uplands and riparian
zones under any one grazing management strategy vary, 2) which season of
use best maintains riparian habitat, and 3) how different stocking rates
affect vegetation and stable banks. Information gathered about these
three questions provide input for developing sound strategies to manage
livestock in cold desert regions where the area of riparian zones is
small in comparison to surrounding uplands.
Grazing by livestock on degraded downcut streams that have a mature
flood plain confined within high banks may be beneficial. High banks
channelize stream high flow causing an increase in velocity and
prevent water spreading. Hoof action by livestock may round or flatten
high banks by trampling. Hoof imprinting may cause depression storage
and increase infiltration.
Approach Taken for Reclaiming Cold Desert Steppe Streams
Two field research facilities were selected to study reclamation of
degraded desert streams and associated riparian zones. One is located
T-n-G-6
-------
on an ephemeral stream (15 Mile Creek) in North Central Wyoming near the
town of Worland. The second, a perennial stream (Muddy Creek), is
located in south central Wyoming between the towns of Rawlins and Baggs.
Both study areas were selected by following the basic criteria
previously discussed. High concentrations of sediment are present in
flow of both streams.
Ephemeral^ Channels-Facilities
The 15 Mile Creek facility is divided into 5 pastures. Pasture 1,
downstream, is going to be used for the study of watershed improvement
practices designed to deposit sediment and improve riparian habitat.
Pastures 2, 3, and 4, located upstream are being used to study the
effect of spring, summer, and fall grazing on stream channel morphology
and vegetation.
Pasture 5 serves as a control and is located upstream from the
other four pastures. Livestock behavior studies are being conducted
inside grazed pastures as well as outside to document seasonally
selective use of upland and riparian habitats. Livestock use studies in
a mature riparian zone and utilization of associated riparian plants is
being conducted in two separate pastures placed on a second stream near
to the 15 Mile Creek facility. Utilization levels on plant species in
relation to season of use and stocking rate are being determined.
Epjietaeral_ Channel Mea_su_remen£s^
Change in stream channel morphology is being monitored using cross
section techniques. Permanent cross sections placed on meander and
straight stream reaches are located downstream, within, and above the
pasture facility. Changes in channel depth and in cross-sectional area
are being determined.
Comparative vegetation analysis are being conducted upstream and
downstream from the pasture system. Within the pastures, permanent
vegetation transects located at channel cross section locations are
established to monitor change in production and shrub density.
Vegetation transects extend from uplands to within 50' of the low
channel bank, from 50" to the low bank, from the top of the low bank to
the interim channel, and along the edge of the interim channel. Degree
of encroachment of vegetation (above and belowground) across the interim
channel is also being measured. Core samples on low banks are also
being taken to determine belowground biomass.
Monthly monitoring of soil moisture using neutron scattering
techniques follows vegetation transects from upland to low bank edge in
all pastures. In addition, soil moisture data is being collected across
meanders and straight sections. Water table depth is recorded and
precipitation monitored continuously during spring, summer, and fall.
Substrates on cross-sections have been core-sampled and analyzed for
particle size distribution and textural classification. We intend to
use these data to characterize sediment deposition on banks and with
belowground biomass examine bank resistivity to erosion.
T-II-G-7
-------
Willow propogation and planting trials have been initiated within
the interim channel of the downstream pasture (1). Survival relative to
planting depth and irrigated-nonirrigated treatments are being
evaluated. The purpose of the planting trial is to study the
feasibility of forming a channel plug of willows to filter sediment from
flow and raise bed load level.
Perennial Stream-Facilities
The Muddy Creek facility is divided into six hydraulic stream reach
units including areas above and below designated treatment areas. The
first hydraulic unit upstream demonstrates stream headcut problems
encroaching into riparian zones formed by beaver dams and good riparian
zone habitat management. Because of waterspreading and constrictive
damming, this riparian zone has caused sediment deposition and
maturation of the stream side zone upstream. Existing differences in
livestock grazing management within this unit illustrate negative
impacts of inappropriate cattle grazing on willow communities and beaver
damming activity as well as the subsequent deterioration in bank
stability.
Downstream, (Unit 2) below active headcutting, channelization has
occurred but meandering flood plains are confined within high banks.
Willows are present but limited and need rest from grazing to stabilize
flood plains. In addition, recent sediment deposited should be planted
with willows to encourage beaver damming as well as provide potential
sites for installing instream flow structures to raise water tables,
cause deposition of sediment on flood plains, and stabilize bed load.
This channelized stream reach, hydraulic unit number 2, is planned as a
research area for documenting the hydrologic response affiliated with
reclamation of cold desert perennial streams.
Hydrologic data will be collected for three years to develop a
water balance of this degraded stream and riparian zone. Three more
years are also planned to stabilize flood plains using willows. Four
additional years will be devoted to using instream flow structures and
beaver to raise the water table and bed load profile over a distance of
approximately three miles and six feet change in elevation. Pertinent
questions to be answered are:
1) How much water is stored underground by natural damming and
riparian zone improvement practices?
2) What change in downstream flow regime occurs in the form of
prolonged release of water season long?
3) What water is lost to downstream users during periods of need
and right?
4) What water balance and bank stability response occurs because of
watershed improvement practices like brush control (burning,
spraying) and range fertilization?
5) What is the significance of good riparian zone management to
abatement of nonpoint and point source pollution?
6) What water balance, bank stability, and vegetation response
occurs because of livestock and wildlife grazing on improved
riparian zones because of watershed improvement practices?
T-H-G-8
-------
7) What are the up and downstream effects of riparian zone and
watershed improvement practices?
8) What are the economic costs and benefits of improving degraded
riparian zones?
Research methodology initiated after three years of hydrologic
monitoring of this three mile stream reach will be based on pilot
research being conducted downstream. Water leaving the designated
research facility (Unit 2) flows into the third hydraulic unit. The
downstream portion of hydraulic unit three, approximately three miles of
a 6 mile stream reach, is characterized by filling with sediment caused
by waterspreading and beaver damming downstream in the fourth hydraulic
unit.
Hydraulic unit number three is in various stages of plant
succession and dissected by braided stream channels caused by flooded
shrub vegetation and deposited sediment. Mud flats are evident.
Willows are virtually absent, perhaps caused by historic early spring
livestock grazing. This reach has been fenced to manipulate spring
grazing and promote willow establishment. This unit will serve as an
area for initiating riparian watershed improvement practices on a
response area caused by good riparian zone management downstream. Grass
and willow planting trials will be initiated on exposed sediments.
Hydraulic unit number four is a stream reach approximately ten
miles long and has been managed to maintain riparian habitat. The main
stream has one diversion to spread water and return flow enters the
braided main channel. Channels are dammed by beaver. Grass has been
established to stablize sediment and provide livestock forage. Willows
are prominent and grazing by livestock occurs during late fall and
winter.
Next downstream, hydraulic unit number five, approximately three
miles long, is downcut and has an immature flood plain constrained
between high banks. This unit may reflect response to flow that has
been stripped of sediment by the improved riparian zone upstream. Flow
in this unit, because of less sediment, should have more power to erode
the existing channel. No work is currently planned in this area.
Hydraulic unit number six changes from a stream channel with an
immature flood plain to a channel with mature floodplains abruptly below
the confluence with an ephemeral stream. Perhaps this change in flood
plain development is caused by an influx of sediment to Muddy Creek
during flow events from the ephemeral channel. Willow growth on the
floodplains of hydraulic unit number six grades from light just below
the confluence of the ephemeral stream channel to heavy downstream.
This entire unit (three miles in length) has been fenced to demonstrate
the use trash collectors as a tool for reclaiming riparian zones.
Fencing will allow control of livestock grazing trials on streamside
vegetation during different grazing seasons and a wildlife exclosure
will provide information on effects of livestock and wildlife vegetation
on riparian zone vegetation and channel morphology. Grazing trials have
begun coincident with installation of trash catchers and fences.
T-II-G-9
-------
Trash collectors have been Installed on straight stream reaches
within three pastures to be used for livestock grazing during spring,
summer, and fall followed by winter wildlife use. They have also been
installed in the wildlife exclosure and in a pasture downstream
adjoining the wildlife exclosure where wildlife grazing but no livestock
grazing will occur. A control reach is downstream from all fenced areas
and has no trash collectors. This entire three mile stream reach is now
in a ponded condition during low flow due to trash collectors. Beaver
have now used three trash collectors for constructing dams and winter
food storage.
Hydraulic unit number six was selected based on the criteria
discussed previously for testing trash collectors as a tool to reclaim
degraded streams and riparian zones. It is hypothesized these
structures will: 1) promote bank storage of water, 2) cause deposition
of sediment on channel flood plains thereby causing channel narrowing,
3) stablize and raise bed load levels and 4) cause upstream flooding
because of constrictive damming of high flow events and thereby sediment
deposition on upstream banks. Willow establishment caused by increased
height of watertables and water spreading should promote beaver damming.
This would eliminate the need for further installation of trash
collectors in hydraulic unit number 6. Planting of willows on new
floodplains in hydraulic unit number five would promote channel
stabilization and sustain trash collector improvement practices upstream
if the hypothesized effects in unit six do occur.
Perennial Stream Measurements
Measurements recorded follow those discussed for the 15 Mile Creek
study. Changes in stream channel morphology will be evaluated using the
cross section technique. Species composition, production, and
utilization of the channel, flood plains and riparian zone vegetaion are
being monitored. Bank-stream interflow and flood plain soil moisture in
hydraulic unit number 6 is ready for monitoring to evaluate change
caused by trash collectors. Willow propogation in planting trials will
be evaluated in hydraulic units numbers 1, 3, and 6. Trash collector
stability and new installations will be evaluated for collecting
bedload, causing bank flooding upstream, and any rise in water tables.
Installation of equipment to construct a water balance is planned for
hydraulic units number 2 and 2 through 6 inclusive.
T-II-G-10
-------
LITERATURE CITED
ASCE Task Committee. 1975. Sedimentation Engineering. ASCE, Manuals
and Reports on Engineering Practices, No. 54, American Society
Civil Engineers, New York, New York.
Brown, Sandra, Mark M. Brinson, and Ariel E. Lugo. 1978. Structure and
functions of riparian wetlands. Proc. of Symposium - Strategies
for Protection and Management of Floodplain Wetlands and Other
Riparian Ecosystems. U.S.D.A. For. Ser. GTR-WO-12: p. 17-31.
Busby, Frank E. 1978. Riparian and stream ecosystems, livestock
grazing, and multiple-use management. Proc. of Forum-Grazing and
Riparian Streams Ecosystems. Denver, Colo. p. 21-30.
Campbell, C.J. and W. Green. 1968. Perpetual succession of stream-
channel vegetation in semiarid region. J. Ariz. Acad. of Sci. p.
86-98.
Campbell, C.J. 1970. Ecological implications of riparian vegetation
management. J. Soil and Water Cons. 25. p. 49-52.
Duff, Donald. 1978. Riparian habitat recovery on Big Creek, Rich
County, Utah - A summary of 8 years of study. Proc. of Forum -
Grazing and Riparian Stream Ecosystems. Denver, Colo. p. 91-92.
Ewel, Katherine Carter. 1978. Riparian ecosystems: conservation
of their unique characteristics. Proc. of Symposium - Strategies
for Protection and Management of Floodplain Wetlands and Other
Riparian Ecosytems. U.S.D.A. For. Ser. GTR-W012: p. 56-62.
Glymph, L.M. and H.N. Holton. 1969. Land treatment in agriculture
watershed hydrology research. In: Effects of watershed changes on
streamflow, Water Resources Symp. No. 2, Univ. Texas Press.
Austin, p. 44-68.
Graf, W.H. 1971. Hydraulics of sediment transport. McGraw-Hill Book
Co., Inc. New York, NY.
Gunderson, Donald R. 1968. Floodplain use related to stream morphology
and fish populations. J. of Wildl. Mgt. 32(3): 508-514.
Haugen, Gordon, et al. 1980. Management and protection of western
riparian stream ecosystems. Amer. Fisheries Soc. Western Div.
publ. 24 p.
T-II-G-11
-------
Heede, B.H. 1978. Designing gully control systems for eroding
watersheds. Environ. Manage. 2(6):509-522.
Heede, B.H. 1982. Gully control: Determining treatment priorities for
gullies a network. Environ. Manage. 6(5):441-451.
Jahn, Laurence R. 1978. Values of riparian habitat to natural
ecosystems. Proceedings of Symposium - Strategies for Protection
and Mgt. of Floodplain Wetlands and Other Riparian Ecosystems.
U.S.D.A. GTR-WO-12: p. 157-160.
Johnson, R. Roy. 1978. The lower Colorado River: A western system.
Proc. of Symposium - Strategies for Protection and Management of
Floodplain Wetlands and Other Riparian Ecosystems. U.S.D.A. For.
Ser. GTR-WO-12: p. 41-55.
Kauffman, J. B., W.C. Krueger, and M. Vavra. 1983a. Effects of cattle
grazing on riparian plant communities. J. Range Manage.
36:685-691.
Kauffman, J.B., W.C. Krueger, and M. Vavra. 1983b. Impacts of cattle
grazing streambanks in northeastern Oregon. J. Range Manage.
36:683-685.
Kauffman, J.B. and W.C. Krueger. 1984. Livestock impacts on riparian
ecosystems and steamside management implications. Arevien. J.
Range Manage. 37:430-438.
Lane, L.J., M.H. Diskin, and K.G. Renard. 1970. Input-output
relationshps for an ephemeral stream channel system. J. Hydrology
13:22-40.
Leopold, L.B., M.G. Wolman, and J.P. Miller. 1964. Fluvial Processes
in Geomorphology. W.H. Freeman and Company, San Francisco.
Leopold, L.B. and W.B. Langbein. 1966, River meanders. The Physics of
Everyday Phenomena, Sci. Am. 79-9287:28-38.
Martin, S. Clark. 1978. Evaluating the impacts of cattle grazing on
riparian habitat. Proc. of Forum - Grazing and Riparian Stream
Ecosystems. Denver, Colo. p. 35-38.
McCall, James D. and Robin F. Knox. 1978. Riparian habitat in
channelization projects. Proc. of Symposium - Strategies for
Protection and Management of Floodplain Wetlands and Other Riparian
Ecosystems. U.S.D.A. GTR-WO-12: p. 125-128.
Meehan, William R. and William S, Platts. 1978. Livestock grazing and
the aquatic environment. J. Soils Water Conservation: 33(5):
274-278.
T-II-G-12
-------
Miller, Robert C. 1978. Relationship between the vegetation and
environmental characteristics associated with an alluvial valley
floor. M.S. Thesis. University of Wyoming, Laramie, WY. 135 p.
Minore, D. 1970. Seedling growth of eight northwestern tree species
over three water tables. U.S. For. Serv. Res. Note PNW-115, Pac.
Northwest For. and Range Exp. Stn., Portland, Oregon, p. 8.
Minore, D. and C.E. Smith. 1971. Occurrences and growth of four
northwestern tree species over shallow water tables. U.S. For.
Serv. Res. Note PNW-160. Pac. Northwest For, and Range Exp. Stn.m
Portland, Oregon, p. 9.
Morisowa, M. 1968. Streams their dynamics and morphology. Earth and
Planetary Science Series, McGraw-Hill Book Company, New York.
Odum, Eugen P. 1978. Ecological importance of the riparian zone.
Proc. of Sym. - Strategies for Protection and Management of
Floodplain Wetlands and Other Riparian Ecosystems. U.S.D.A. For.
Serv. GTR-WO-12: p. 2-4.
Parker, G. 1978a. Self-formed straight rivers with equilibrium banks
and mobile bed. Part 1. Th/e sand-silt river. J. Fluid Mecb. 89:
109-125.
Parker, G. 1978b. Self-formed straight rivers with equilibrium banks
and mobile bed. Part 2. The sand-silt river. J. Fluid Mech. 89:
109-125.
Peterson, H.V. 1950. The problem of gullying in western valleys. In;
Applied Sedimentation. John Wiley & Sons, Inc., New York. p. 407-
434.
Platts, William S. 1981. Effects of sheep grazing on a riparian -
stream envrionment. U.S.D.A. For. Serv. Res. Note Int. 0-307: 6
Roath, L.R. and W.C. Krueger. 1982. Cattle grazing influence on a
mountain riparian zone. J. Range Manage. 35(1): 100-103.
Schmidly, David J. and Robert B. Ditton. 1978. Relating human
activities and biological resources in riparian habitats of western
Texas. Proc. Symp. - Strategies for protection and management of
floodplain wetlands and other riparian ecosystems. U.S.D.A. For.
Serv. GTR-WO-12. p. 107-116.
Schumra, S.A. 1963. Sinuosity of alluvial rivers on the great plains.
Geological Soc. of Amer. Bui. 74:1089-1100.
Simons, D.B. and F. Senturk. 1977. Sediment transport technology.
Water Resources Publications, Fort Collins, Colo.
T-II-G-13
-------
Speck. J.E. 1981. A comparative study of the effects of continuous
season-long and deferred rotation grazing on mountain riparian
vegetation and water quality. M.S. Thesis. University of Wyoming,
Laramie. 149 p.
Thomas, Jack Ward, Chris Maser, and John E. Rodiek. 19878. Riparain
zones in managed rangelands- their importance to wildlife. Proc.
of Forum - Grazing and Riparian Stream Ecosystems. Denver, CO. p.
21-30.
Tubbs, Allen A. 1980. Riparian bird communities of the Great Plains.
Proc. of Workshop - Management of Western Forests and GRasslands
for Non-game birds. U.S.D.A. For. Ser. Tech. Rep. Int-86: p.
419-433.
T-II-G-14
-------
MANAGEMENT ALTERNATIVES FOR URBAN STORMWATER
by Robert Pitt, and Roger Bannerman
Department o-f Natural Resources
Madison, Wisconsin
A. Introduction
This paper is a brie-f example o-f how suitable
stormwater management alternatives can be designed using
available monitoring information. Milwaukee area urban
runoff has been extensively studied by the Wisconsin
Department of Natural Resources, the U.S. Geological Survey
and the Southeastern Wisconsin Regional Planning Commission
(SEWRPC), among others, for about the past ten years. Two
major research projects were conducted for the International
Joint Commission (IJC) on Great Lakes Research (Bannerman,
et al, 1979) and for the Nationwide Urban Runoff Program
(NURP) for the U.S. EPA (Bannerman, et al, 1983). The IJC
project involved detailed urban runoff monitoring of many
single land use and mixed land use sites, while the NURP
project examined urban runoff at eight single? land use sites
for two years. This information, in conjunction with other
regional and nationwide urban runoff characterization and
control data, was recently used by the Nonpoint Source and
Land Management Section of the DNR to prepare a Construction
Site and Stormwater Management Plan and Model Ordinance
(Pitt 1985). This paper uses this summarized information in
several related examples to show how stormwater management
controls can be designed to meet both flood control and
urban runoff quality objectives.
B. Required Inventory Information
Eiefore an evaluation of suitable control measures can
be made, certain information about the study area must be
obtained. This inventory information pertains to the
importance of the various source areas in contributing
runoff flow or pollutants to the outfall. Land cover
information is needed to describe the percentage of the
study area covered by various surfaces, such as streets,
roofs, landscaping and parking areas. This information is
usually obtained by direct measurement from aerial
photographs. Additional information is needed to describe
the aspects of these areas that modify their pollutant or
flow generating capabilities. Roof drain connections,,
pavement texture and condition and the type of drainage
system are examples. This information requires field
investigations of representative areas for each land use
subcategory. Table i is an example of land cover information
T-II-H-1
-------
for typical medium density residential and industrial areas
recently studied (Pitt and McLean 196S5) . Landscaping
comprises the most common surface cover in the single family
residential area, while rooftops and paved parking and
storage areas are the most common surfaces in the industrial
area. These major differences can result in significantly
different runoff characteristics.
Figures 1 through 4 are based on extensive runoff
monitoring of these two study areas and show how the
different areas contributed to the outfall total solids and
lead yields for different rain volumes (Pitt and McLean
1985). Rain has a much greater effect on the relative
contributions for the residential area. The larger amount of
impervious surfaces in the industrial area reduces the
variations observed. Table 2 summarises this information for
a selection of pollutants for large rains (fairly stable
source contributions occur with large rains because of the
decreased importance of pervious area rain infiltration).
Roof and paved parking/storage area runoff contribute most
of the runoff volume and most of many of the pollutants in
industrial areas. Flow and pollutant contributions are much
more diverse in residential areas. This information is used
in the following example to estimate the effectiveness of
the different control measures in reducing runoff flows and
pollutant contributions.
C. Infiltration and Detention Basin Combined Effects in
Areas Having Poor Soil Percolation
1. Hypothetical Site Characteristics,,
A series of calculations were made? to investigate
possible alternative flow and pollutant reduction options
available for a hypothetical study area. Options
investigated included dry and wet detention basins, roof and
parking/storage area infiltration, and grass waterway
drainages. The benefits examined included flow volume and
rate reductions along with pollutant concentration and yield
reductions.
These calculations are intended to generally describe
the benefits of alternative urban runoff controls available.
The calculation procedures, along with suitable design
guidelines, are described in the recently prepared
Stormwater Management Plan (Manual of Practice) (Pitt 1985).
As stated above, these calculations are based on many field
monitoring projects, including several conducted during the
past ten years in the Milwaukee and Menomonee River
watersheds (Bannerman, et al, 1979 and 1983).
2. Flow Estimates.
Table 3 shows the land use areas in the example
watershed, divided by hypothetical soil percolation
characteristics. Table 4 shows the calculations used to
T-II-H-2
-------
estimate the average urban runo-f-f flow rates for a variety
of large storm conditions. These flow rate estimates are
based on the monitoring results of several hundred Milwaukee
area urban runoff events studied by Banner-man,, et al <1979
and 1983). Statistical analyses of these runoff data also
indicated that peak runoff flow rates for large urban runoff
events are about 6.5 times the average flow rates. The peak
flow rates generally only occur for several minutes during
the event.
3. Dry Detention Basin Design.
Dry detention basin storage requirements were estimated
for" the different land uses. The basins should drain within
a few days to enable control of storms likely to fallow in a
short time. For the critical 100 year storm., basin outflow
rates of between 100 and 200 cfs would result in reasonable
drainage times. Table 5 shows required basin sizes for
various outflow rates. Total basin areas of about three
percent of the watershed area may be required to control the
critical storm. Alternative basin criteria can also be used
to determine the required basin sizes.
Alternative detention basin design objectives and
associated required basin sizes are summarized on Table 6.
The minimum basin size would reduce the expected peak flow
rate to the assumed channel carrying capacity and would
require about 0.3 percent of the watershed area. The maximum
basin size is designed to allow basin drainage in three days
to obtain a maximum flow rate reduction and would require
about three percent of the watershed area.. A reasonable
basin objective is intermediate to these two sizes and would
reduce the peak flow rates to the average flow rates. A
basin size of about 0.5 percent of the watershed area would
be required. Even though these dry detention basins would
substantially reduce the peak flow rates, they do not change
the total volume of water or the amounts of pollutants
entering the creek and may not benefit flood control or
water quality in downstream receiving waters.
4. Wet Detention Basin Design.
Many monitoring projects have demonstrated substantial
water quality benefits associated with well designed wet
detention basins (EPA 1983). Wet basins can be used in
conjunction with dry basins (and/or other urban runoff
control devices) to help meet both flood control and water
quality objectives. Wet basins also do not reduce the total
volume of water discharged to the receiving water, but can
remove much of the settleable solids.
High control levels of pollutants in wet basins are
only reasonably possible for the more frequent rains. Water
quality is not usually of much concern during very large
infrequent events, such as the 100-year storm. The most
cost-effective benefits can be relealizecl during common,
smaller events. A wet basin about 0.3 percent of the
watershed area, having a wet depth of about four feet and a
T-II-H-3
-------
•freeboard height of about 12 -feet would produce about 75
percent, solids control. This -freeboard requirement is needed
to reduce the wet pond outflow velocity to desired values
and may be larger than desired. It would be possible to
reduce this freeboard requirement with the use of a surge
dry basin or upland controls.
5. Upland Infiltration Design.
Upland infiltration can be used to reduce the flow
rates to downstream detention basins, but more importantly,
they reduce the total volume of water and amounts of
pollutants discharged. Upland infiltration devices recharge
the local groundwater and increase critical low flows during
summer dry months, further benefiting the receiving water.
Care must be taken to protect significant groundwaters, of
course. Upland infiltration can also be used only in areas
with appropriate soil percolation conditions. The soils in
the hypothetical watershed are of two percolation classes:
poorly drained and ,/noderatly well to well drained. The
following calculations only consider the application of
upland infiltration in the better drained soils.
Table 7 shows the maximum effects that roof and
parking/storage area infiltration may have on flow rate and
volume reductions for several large rains in residential and
industrial areas. These reductions can be substantial,
especially for industrial areas where about SO percent of
the water originates from roofs and parking/storage areas.
These improvements must be reduced because of the poor soils
in the assumed study area. Roof runoff infiltration
performance in the residential area would be reduced to
about ten percent, roof runoff infiltration performance in
the industrial area would be reduced to about 15 percent,
parking/storage area infiltration performance in the
industrial area would be reduced to about five percent and
roof plus parking/storage area runoff infiltration
performance in the industrial area would be reduced to about
20 percent.
6. Grass Waterway Design.
Grass waterways may also significantly reduce flow
volumes through infiltration. Table 8 shows the effects of
grass drainages, instead of underground sewerage. Flow rates
and volumes during the large 100 year storm may be reduced
by as much as 20 percent by these devices. Table 9 shows the
total expected watershed flow rates using these controls.
7. Revised Dry Detention Basin Design Considering Upland
Infi1trati on.
Table 10 shows the revised dry basin storage
requirements with these upland controls. The area
requirements have been slightly reduced, to about 0.4
percent for the 100 year storm and for peak shaving, but the
outflow rates have been reduced substantially for many
conditions. The revised wet basin area and freeboard
T-II-H-4
-------
requirements are slightly reduced (down to about 10 -feet
over about 0.3 percent o-f the watershed area).
D. Summary.
Table 10 presents a summary o-f the expected water
quality and quantity improvements -for the various
alternatives investigated. No water quality or water volume
improvements will occur -for the dry basins alone, but
substantial reductions in -flow rates,, water volume, and
pollutant yields and concentrations will likely occur -for a
combination o-f control devices.
In conclusion,, a reasonable amount, of the watershed
land area can be used to effectively control urban runoff
quantity and quality. Dry basins comprising about 0.4
percent of the? watershed area (about six feet deep) can be
used with wet basins requiring about 0.3 percent of the area
(having a wet pond de?pth of about four feet and a freeboard
elevation of about ten feet) and upland infiltration devices
(in areas having suitable soils) to achieve significant flow
and pollutant reductions. Similar control devices have been
monitored and proven effective during many urban runoff
research projects. Special restricting conditions (such as
poor percolating soils) can be considered in the design of
these control devices.
E. References
Bannerman,, R. , J.G. Konrad, D. Becker., G.v1. Simsiman,
G. Chesters, J. Goodrich-Mahoney, and B. Abrams. "The IJC
Menomonee River Watershed Study",, U.S. Environmental
Protection Agency, EPA-905/4-79-029-C, Chicago, 111., 1979.
Bannerman, R., K. Baun, M. Bohn. P.E. Hughes, and D.A.
Grac2 y k. "Evaluation of Urban Nonpoint Source Pollution
Management in Milwaukee County,, Wisconsin", U.S.
Environmental Protection Agency, PB 84-114164., Chicago,
111., 1983.
Pitt, R., and J. McLean. "Toronto Area Watershed
Management Strategy Study: Number River Pilot Watershed
Project", Ontario Ministry of the Environment, Totonto,
Ontario,, 1985 (Draft).
Pitt, R. "Construction Site Erosion and Stormwater
Management Plan and Model Ordinance", Wisconsin Department
of Natural Resources, Madison, Wisconsin, 1985 (Draft).
T-II-H-5
-------
H
I
33
u
THISTLEDOWNS TOTRL SOLIDS SOURCES
Rein (MI)
Vve.Ua-> lift
-------
F\
H
I
a
i
EMERY TOTRL SOLIDS SOURCES
-------
H
I
a
00
THISTLEDOWNS LERD SOURCES
Rein (MI)
-------
EMERY LERD SOURCES
H
a
K
-------
Table 1. Industrial and Medium Density Residential Land Covers (%)
Land Cover Industrial Residential
Streets 5.6% 11.4%
Swales 0 3.1
Grass strips 0 9.2
Concrete walks 0.4 3.7
Front landscaping 10.2 16.3
Paved driyeways 2.2 6.4
Paved parking/storage 19.9 0>1
Unpaved parking/storage 15.4 0
Sidewalks 0.1 0.7
Back landscaping 0 24.5
Undeveloped 12.4 0
Railroad 2.6 0
Connected roofs 31.1 4.9
Roofs draining to driveways 0 8.7
Roofs draining to lawns 0 10.9
Total 99.9 99.7
Total Impervious 47.4 36.4
Total Pervious 30.4 47.6
Total Disconnected Impervious 22.2 16.0
Source: Pitt and McLean 1985
T-II-H-10
-------
Table 2. Sources of Pollutants During Large Events
Industrial Areas:
Constituent
Front Drive-
Streets Walks Lands, ways
Paved Unpaved Open Connected
Park. Park. Space Roofs
H
I
I
SB
I
Total solids
Suspended solids
Phosphorus
Kjeldahl N
Phenolics
COD
Fecal colif.
Copper
Lead
Zinc
Flow
Area
Mixed Residential
Total solids
Suspended solids
Phosphorus
Kjeldahl N
Phenolics
COD
Fecal Colif.
Copper
Lead
Zinc
Flow
Area
2%
3
1
1
2
1
2
3
2
1
9
6
Area:
Streets
6
8
3
8
3
8
8
14
12
3
23
10
1%
1
1
1
1
1
3
1
1
1-
1-
1-
Walks
6
4
6
24
17
27
59
4
12
5
3
4
27,
10
10
6
2
2
6
6
3
2
4
10
Front
Lands.
20
25
24
3
1-
4
7
4
2
1-
8
23
3%
4
2
2
5
3
2
4
3
4
2
2
Drive-
ways
8
8
3
2
2
2
1
14
10
2
6
5
26&
42
20
30
58
37
21
50
35
51
30
21
Paved
Park.
14
18
8
6
4
3
2
26
22
5
8
5
20%
20
8
3
8
8
25
12
12
8
3
15
Open/
Rear
Lands .
26
32
29
6
1-
7
14
6
3
1-
15
31
10%
14
10
5
1
3
5
4
2
2
3
15
Connect .
Roofs
14
3
20
36
14
34
6
21
26
59
19
8
36%
6
48
52
23
45
36
20
42
32
49
31
Roofs
to
Drive.
6
2
6
13
56
13
3
10
12
24
9
7
Roofs
to
Lawns
1-
1-
1
2
4
2
1-
1
1
2
2
7
Source: Pitt and McLean 1985
-------
Table 3. Example Land Use by Soil Type (acres)
Residential
Industrial
Commercial and Institutional
Open Space
Total
Poorly Well to Total
Drained Moderately
Soil Well Drained
1125 1445 2570
435 235 670
110 110 220
125 85 210
1795 1875 3670
Table 4. Average Flow by Land Use (cubic feet per second)
Return
Freq .
(yrs)
1
1
1
1
1
1
1
10
100
Rain
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24
Total
Rain
(in.)
0.9
1.1
1.3
1.5
1.7
2.0
2.3
3.9
5.5
Rain
Int.
(in/hr)
1.80
1.10
0.65
0.50
0.28
0.17
0.10
0.16
0.23
Residential
0.83
0.46
0.27
0.23
0.14
0.081
0.045
0.083
0.13
Institutional,
Commercial, or
Industrial
1.1
0.59
0.36
0.30
0.18
0.10
0.056
0.10
0.15
Open Space
0.06
0.04
0.03
0.03
0.01
0.01
0.01
0.02
0.05
T-li-H-12
-------
Table 5. Alternative Dry Basin Designs for Hypothetical Example
1) Minimum design to reduce peak runoff to assumed channel capacity
(1500 cfs):
Storm Peak
Reqd.
(24 hr) (cfs)
(1.)
1 yr 1100
10 2000
100 3100
(acre-ft)
0
21
66
Reqd.
Freq. Inflow Storage Surface
(acres)
(2)
0
3.5
11
Reqd. % Pond
of 3700 Drain
acre basin (days)
0%
0.1
0.3
0
1.3
2.1
2.) Maximum design to allow pond drainage in three days:
Storm Ave.
Freq. Inflow
(24hr) (cfs)
Reqd.
Outflow
(cfs)
Reqd. Reqd.
24 hr Surface
Storage (acres)
(acre-ft) (2)
lyr 170 57
10 300 100
100 480 160
220
400
630
37
67
105
Reqd. 2
of 3700
acre basin
1.0%
1.8
2.8
3.) Reasonable design to reduce short-term peak to ave.
(peak shaving):
Storm Peak
Freq. Inflow
(24hr) (cfs)
(1)
flow
Ave.
Inflow
(cfs)
Reqd.
30-min
Storage
(acre-ft)
Reqd.
Surface
(acres)
(2)
Reqd . %
of 3700
acre basin
1 yr
10
100
1100
2000
3100
170
300
480
38
70
108
6.3
12
18
Footnotes:
0.2%
0.3
0.5
(1) Peak flow duration is assumed to be 30 minutes
(2) Dry pond depth is six feet
T-II-H-13
-------
Table 6. Effects of Upland Infiltration on Runoff Rates (cfs per acre)
Residential Area
Storm
Freq.
(yrs,)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24
Ave . Flow
With No
Infilt.
0.83
0.46
0.27
0.23
0.14
0.081
0.045
0.083
0.13
Ave . Flow
With Roof
Disconnect •
0.67
0.38
0.23
0.19
0.12
0.065
0.037
0.071
0.11
Percentage
Ave . Flow
Reduction
19%
17
15
17
14
20
18
15
15
Industrial Area:
Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs)
0.5
1
2
3
6
12
24
24
24
Ave . Flow
With No
Infilt.
1.1
0.59
0.36
0.30
0.18
0.10
0.056
0.10
0.15
Roof Dis. :
Hew %
Ave . Reduc
Flow
0.54 51%
0.30 49
0.19 47
0.15 50
0.094 48
0.053 47
0.031 45
0.060 40
0.098 35
Parking
New
Dis.
%
Ave . Reduc .
Flow
0.90
0.50
0.31
0.25
0.15
0.088
0.048
0.090
0.13
18%
15
14
17
17
12
14
10
13
Roof and Parking
New %
Ave. Reduc•
Flow
0.37
0.21
0.13
0.11
0.064
0.038
0.022
0.048
0.080
66
64
64
63
64
62
61
52
47
T-II-H-14
-------
Table 7. Effects of Swale Drainages on Runoff Rates (Percent)
Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs.)
0.5
1
2
3
6
12
24
24
24
Percentage Flow Reducrion for:
Residential Industrial
(3 ft. swales) (7 ft. swales)
15-
15-
15-
15-
19
27
50
33
23
14-
14-
14-
14-
14
23
35
18
14-
Table 8. New Runoff Volumes with Upland Infiltration (cfs)
Storm
Freq.
(yrs)
1
1
1
1
1
1
1
10
100
Storm
Dur.
(hrs.)
0.5
1
2
3
6
12
24
24
24
Ave. Flow
With No
Infilt.
3100
1700
1000
860
520
300
170
300
480
Ave . Flow
With
Infilt.
2760
1520
900
760
425
230
120
240
380
Peak Flow
With No
Infilt.
NA
NA
NA
NA
NA
NA
1100
2000
3100
Peak Flow
With
Infilt.
NA
NA
NA
NA
NA
NA
750
1600
2500
Percentage
Flow
Reduction
11%
11
11
12
22
22
32
21
21
T-II-H-15
-------
Table 9. Alternative Dry Basin Designs for Hypothetical Example
With Upland Infiltration
1) Minimum design to reduce peak runoff to assumed channel capacity
(1500 cfs):
Storm Peak
Freq. Inflow
(24 hr) (cfs)
(1)
1 yr 750
10 1600
100 2500
Reqd. Reqd. Reqd. % Pond Pond
Drain Acres
n (days) Saved
0 0
1.1 2.8
1.7 4.2
Storage
(acre-ft)
0
4
41
Surface
(acres)
(2)
0
0.7
6.8
of 370i
acre b;
0%
0.02
0.2
2.) Maximum design to allow pond drainage in three days:
Storm Ave.
Freq. Inflow
(24hr) (cfs)
lyr 115
10 240
100 380
Reqd.
Outflow
(cfs)
Reqd.
24 hr
Storage
(acre-ft)
Reqd.
Surface
(acres)
(2)
Reqd. %
of 3700
acre basin
Pond
Acres
Saved
38
80
130
150
320
500
25
53
83
0.7%
1.4
2.2
12
14
22
3.) Reasonable design to reduce short-term peak to ave.
(peak shaving.):
flow
Storm Peak
Freq. Inflow
(24hr) (cfs)
(1)
1 yr 750
10 1600
100 2500
Ave.
Inflow
(cfs)
115
240
380
Footnotes:
Reqd. Reqd.
30-min Surface
Storage (acres)
(acre-ft) (2)
26
56
88
4.4
9.4
15
Reqd. %
of 3700
acre basin
0.1%
0.25
0.4
(1) Peak flow duration is assumed to be 30 minutes
(2) Dry pond depth is six feet
Pond
Acres
Saved
1.9
2.6
3.0
T-II-H-16
-------
Table 10. Pollutant Control Alternatives (Approximate Percentage Control at Outfall)
Control Total Sus.
Option: solids solids
H
M
M
a
i-1
•vj
1
1
1
2
2
2
3
3
3
4
5
6
6
6
7
7
7
8
8
8
9
9
9
10
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
lOyr
lOOyr
lyr
0
0
0
0
0
0
0
0
0
90
75
34
25
20
34
25
20
34
25
20
34
25
20
84
0
0
0
0
0
0
0
0
0
60
50
28
19
14
28
19
14
28
19
14
28
19
14
64
Phos
0
0
0
0
0
0
0
0
0
55
45
34
25
20
34
25
20
~ 34
25
20
34
25
20
64
. TKN
0
0
0
0
0
0
0
0
0
35
30
44
35
30
44
35
30
44
35
30
44
35
30
61
Phen.
0
0
0
0
0
0
0
0
0
50
40
36
27
22
36
27
22
36
27
22
36
27
22
62
COD
0
0
0
0
0
0
0
0
0
55
~45
44
35
30
44
35
30
44
35
30
44
35
30
69
Fecal
colif .
0
0
0
0
0
0
0
0
0
70
. -6Q .....
30
21
16
30
21
16
30
21
16
30
21
16
72
Cu
0
0
0
0
0
0
0
0
0
70
60
37
28
23
37
28
23
37"
28
23
37
28
23
75
Pb
0
0
0
0
0
0
0
0
0
70
60
40
31
26
40
31
26
40
31
26
40
31
26
76
Zn
0
0
0
0
0
0
0
0
0
55
45
54
45
40
54
45
40
54
45
40
54
45
40
75
Flow
volume
0
0
0
0
0
0
0
0
0
0
0
32
20
19
"32
21
21
32
21
21
32
21
21
32
Ave.
flow
rate
0
0
0
67
67
67
0
0
0
12
0
32
20
19
32
21
21
32
21
21
32
21
21
32
Peak
flow
rate
0
25
52
95
95
95
85
85
85
86
74
32
20
19
0
6
40
97
96
96
90
88
88
90
Pond
area
(%)
0
0.1
0.3
1.0
1.8
2.8
0.2
0.3
0.5
0.7
0.3
0
0
0
0
0.02
0.2
0.7
1.4
2.2
0.1
0.25
0.4
6.7 ~
Codes:
1. Minimum dry detention to meet channel carrying capacity
2. Maximum dry detention to drain in three days
3. Peak shaving dry basin
4. Wet basin for 90% solids control for 1 yr, 24 hr storm only
5. Wet basin for 75% solids control for 1 yr, 24 hr storm only
6. Upland infiltration and swales
7. Minimum dry basin plus upland infiltration and swales
8. Maximum dry basin plus upland infiltration and swales
9. Peak shaving dry basin plus upland infiltration and swales
10. Peak shaving dry basin plus small (75%) wet basin plus upland infiltration and swales
-------
NONPOINT SOURCE POLLUTION OF THE VENICE LAGOON:
PERSPECTIVES OF LONG-TERM ABATEMENT
F. Zingales
Cattedra di Chimica, Facial ta r- dii. Ingegneria, Universita* di
Pad ova, via Mar sola 9, 35131 Pad ua (1ta1y).
A. Marani
Dipartimenta di Spettroscapia, Elettrochimica e Chimica Fisica,
Universita" di Venezia, D.D. 2137, 30123 Venice? (Italy).
G. Bendoricchio
Cattedra di Chimica, Facolta" di Ingegneria, Universita" di
Padova, via Marzolo 9, 35131 Padua (Italy).
A. Rinaldo
Istituto di Idraulica "Giovanni Poleni", Universitan di Padova,
via Loredan 20, 35131 Padua (Italy).
ABSTRACT
The paper adresses the major research topics and achievements
related to nonpoint source pollution (NPSP) abatement of the
Lagoon of Venice (Italy) aiming at. source evaluation -for the
2000 km2 wide mainland and water quality forecasting of the
receiving Lagoon.
The relevance of the work lies in the forthcoming need of setting
control policies . for the planned man-made flood barrages
regulating the exchange between the sea and the Lagoon. In fact,
control strategies should account for general health of the city
of Venice, of the natural environment and of the macro-- and
micro-economics. Hydrological, chemical, hydraulic and physical
aspects are worked out in a tentatively synoptic framework.
Detailed aspects are considered for the control of urban and
agricultural NPS, for nutrient transports through the Lagoon
system and for possibility of controlling algal blooms via
Catastrophe Theory.
KEYWORDS
Non Point Source Pollution, Nan Paint Source Abatement,
Quality Modeling, Tailoring Runoff Models to Recei vi rig
Models, Catastrophe Theory, Algal Blooms, Lagoon of Venice.
T-II-I-1
Water
Water
-------
INTRODUCTION
The paper is centered on a discussion of the research
progress on nutrient Nan Point. Source Pollution (NPSP)
abatement of the Lagoon of Venice (Italy).
The first part of the paper summarizes the physical setting
af the system and the main results of the hydrological,
hydraulic and chemical research, aimed at selection ofs
- source distribution and magnitudes
- coding of overall mass transports and balances:!
••••• modeling aspects.
The second part puts forth some new ideas on forecasting of
the eutrophication of the Lagoon, with special emphasis on
modeling of algal blooms via application of combined mass
transport models and Catastrophe Theory.
The third part briefly
long term abatement,
texture of the problem*
outlines some likely possibilities of
given the social and environmental
THE LAGOON SYSTEM
The Venice Lagoon (Figure 1) is a complex tidal embayment
which consists of interconnected sloughs, marshes, often
meandering natural and man-made channels,, Unlike other
similar tidal water bodies, t h e L a g o o n s y stern c a n b e
represented by one basic tidal type, that of wel 1--mixed
shallow waters. The salinity, in the whole reach, ranges
•from the salinity value of sea water near the three mouths,
to that of nearly fresh water only along a confined belt at.
the tributaries outlet. Nevertheless salinity concentrations
are marttained at values close to that of sea water for-
al most the entire length span. At the boundary, minor-
salinity stratification may occur in winter as a result of
runoff during periods of heavy rainfalls otherwise the
Lagoon water is isohaline due to low fresh water inflows and
strong wind-- and tide-induced mixing (Bhetti , 1979),,
The Lagoon receives drainage waters -from a large basin at its
north-western reach, whereas the ratio between drainage area
•four (2000 Km2 / 50O
its input of fresh
c ample;-; n e t w a r k o f
agri culture, whereby
and the Lagoon surface adds up to about
km2). The system receives about 40X of
water from pumped outlets draining a
lowland areas, mainly devoted to modern
large and steadily growing urban runoff pours into the
system (Rinaldo, 1982; Cossu et al., 1984; Bendoricchio et
al., 1984; Bendoricchio et al.', 1985).
No major tributary flows into the Lagoons in fact this unique
T-II-I-2
-------
natural environment has been preserved through the centuries
by diverting the estuaries of two rivers, Bile and Brenta
(the works lasted from 1630 and 1850 ca.) once flowing
within the reach of the system. The actual tributaries
(Zero, Dese, Marzenego, Naviglio Brenta) peak at about 20-f6O
m3/s without major solid transport occurences.Input of fresh
water, besides the mentioned urban and agricultural
components, is completed by other small tributaries (up to
27 overall delivery locations), sewage treatment plants and
industrial discharges surroundings the Lagoon proper
(Figure 1) (Rinaldo ancJRinaldo, 1983; Cossuet al. , 1984).
Lagoon is ruled by the
s o m e instances 10-5-15 m
shoals alongside. The
the forcing tide at the
altered with respect to
made navigation channel
Tidal propagation throughout the
hydraulics of deep channels (in
deep) connected to broad shallow
phase delay of tidal elevations to
three marine openings (substantially
natural conditions because of a man
lined through the mid Lagoon) ranges from about 1 hour at
the city of Venice to 2,5 hours at the northern reach
(Goldmann et al., 1975), The tides in the Lagoon do not show
properties of standing waves, underlining the importance of
both inertial and dissipative effects (Di Silvio and
D'Alpaos, 1972).
Water circulation in the Venice Lagoon system is a
combination of tide-driven and wind-driven currents, whereas
taaroclinic density-driven currents seem negligible
(Ghetti,1979). Total water volumes stored within the system
range from 160 to 510 millions of cubic meters (Mcum), where
the dominating tidal exchange (for a gauged M2 tide) over-
half period may amount to 330 Mcum (through the Lido (130
Mcum), Malamocco (136 Mcum) and Chioqgia (64 Mcum) mouths)
(Ministero dei Lavori Pubblici, 1979).
The further lack of coastal currents in the outer Adriatic
sea induces further uncertainty to environmental studies: it
has been inferred, in fact, from some experimental evidence
that some 15-5-30% of the water volumes ebb-released by the
system are not dispersed and reenter the mass balance of the
Lagoon for a typical M2 tide (Cossu et al., 1984).
NPSP SOURCES
The task of predicting overall non-point
nutrient for the actual purposes) into the
is demanding due to the geomorphologic
complexity of the drainage watershed
agricultural and urban components.,
sources (mainly
receiving Lagoon
and hydraul :l c
both in its
Urban runoff tied to about. 1,000,000 inhabitants within the
whole mainland, is noteworthy and mainly non-point in
nature. Only 250,000 units are collected through sewage
T-II-I-3
-------
systems and sewage treatment plants. The remnants load
directly the drainage network and -flow into the Lagoon»
Furthermore, the inhabitants of the city of Venice and of
the islands plus seasonal presences due to intense tourism
amount to 100,000 -further equivalent inhabitants, untreated
at present stage. Hence, total urban IMPS Nutrient load has
been estimated in about 4,000 t/year of total Nitrogen and
1,500 t/year of Phosphorus (Bendoricchio et al.,1985).
As far as other NPSP components are concerned, the nature of
the watersheds and the range of the land uses have called
•for a thorough model selection phase, in view also of the
difficulties of calibration. The possibility of use of
sera-order approximation . (like empirical
runoff-concentration curves) has been discarded, in view of
their conceptual inadequacy (Zingales et al.,1984). In fact,
significant differences between rising and falling stage
discharge/concentration relations can'be fully explained by
quite simple mass transfer relationships referring to
interactions with channel bottom materials and soil
adsorption mechanisms. In this pattern, pollutant
load/discharged total water volume relations seem fit to
linear correlation scheems, as proved also by the results of
smoothing of experimental data gathered in a testing
watershed in environment of the Lagoon, for which two annual
daily data collections had been carried out (Zingales et
al., 1982).
An important result of such conceptual mass transfer and
balance relations consists of the validation of the
first-order approaches introducing reference weighted
concentrations in the runoff, or loading functions for
long-term, or event-based, time scale (Haith and Dougherty,
1976; Haith, 1982$ D i c k er of f De 1 w i. c h e an d Ha i. t h , 1983).
In view of these facts, screening model have been deemed to
the point for the preliminary phase? of the case study at
hand, the flow-averaged concentrations being realistically
linear with respect to total water quantities at. large time
scales. Incidentally, this fact holds true also for other
indices, like BOD pollution loadings (Whipple and Hunter,
1977) and, even more notably, for solid transport (Haith,
1982).
A second-order class of modeling approaches, suitable to
simulation of runoff and dissolved nutrient losses, consists
of the development of unit-mass response functions (UPIK'F)
for the source area. Such approach is either of experimental
nature (based, for istance, on parameter indentification via
constrained linear search) (Jolankai , 1.983) or of conceptual
nature (Bendoricchi o and Rinaldo, 1981; Zingales et. al . ,
1984), or of statistical derivation (Marani and
Bendoricchio, 1984). Such approaches, which portray in depth
most features of time development of pollution loadings,
T-II-I-4
-------
rely on hydrologies! and chemical parameters which require
calibration and hence prove suitable to large scale
prediction
predictive
excessi ve
placed in
frequently
given the
the rapid
upon -field-size calibrations. In fact, although
errors of screening or UMRF models may seem
if compared to observed values, they should be
perspectives: errors in model predictions are
comparable to sampling and analysis errors,
difficulty to accurately measure NPSP and, e.g.,
transformations (changes in crop practices;
urbanizations) of the drainage areas.
At this stage, there seem no point in trying to implement
distributed parameter modeling approaches (among the several
outstanding contributions in the field) to the actual case
study.
either of
Phosphorus,
area due to
over a t i me;
nature and
(Haith, 1982).
The model adopted, as preliminary screening phase of source
areas, is a modification of loading function approach suited
to the available data. The waste loading,
solid-phase or dissolved (or total) Nitrogen and
is defined as a nutrient export from unit source
runoff event on i~th day, eventually integrated
span. Combined effects of crop type, soil
hydrology are accounted far in the scheme?
Computational sequences for dissolved and suspended chemicals
are separated at small time scales, reflecting the different
mechanisms of leaching from croplands (runoff-erosion);
nevertheless both phases can be treated jointly for
long-term total nutrient losses evaluations. Examples of
typical average concentration ranges in the runoff, as from
the U.S. and European literature and of the wide variability
of these parameters have been discussed in the literature
(Jolankai,1983)„
Table 1 - Yearly urban (URB) arid agricultural (AGR) loads of
N and P split into Lagoon subsystems (after Bendo~
ricchio et al.,1985).
NUTRIENT
N
P
SOURCE
URB
AGR
URB
AGR
LOADS ON LAGOON SUB-BASINS
(tc
Northern
2 , 200
1 , 900
850
350
Dins/year )
Central
1 , 300
1 , 750
500
300
Southern
750
1 , 500
300
300
LAGOON LOADS
(tons/year*)
4 , 250
5 , 200
1 , 650
950
The overall results of source magnitude evaluation are?
T-II-I-5
-------
reported in Table i -for comparison among the single
contributions. A detailed description of the results is
shown in Figure 2» The discussion of modeling and physical
implications are reported elsewhere (Zingales et 'al.,1984;
Bendoricchio et al„, 1985).
As a sole remark, the discretization of input load delivery
locations is relevant to the receiving water model. In fact,
the limited number of NFS yields a lowered priority for
spatial transport modeling, at least. in the scale needed for
evaluation of residence time within the Lagoon. On the other
hand, integration of runoff and receiving water models
needs a time discretisation which forces the user into
considering tidal effects on'dispersion of the water body
(Rinaldo and Marani 1984). Therefore, once long-term
nutrient discharge sites are pointed out via screening
models, the further step is related to the need of shrinking
the time scale of event-based responses (which, for the
mainland of the Lagoon, range from one day to dozen of days)
to, at least, inertidal scales. Such a time scale is deemed
necessary for water quality modeling of the receiving water
body (Rinaldo and Marani, 1984). Experimental results have
allowed calibration of conceptual UMRF approach which fits
the level of detail required for prediction of lagoonal
conditions (on the order of the hour) (Zingales et al„,
1984).
In this pattern, the concept of reference weighted
concentration of pollutants in the runoff has been linked to
the determination of parameters of conceptual UMRF (Zingales
et al., 1984) portraying mass transfer kinetics and flow
rates generation. Even though any assumptions concerning
extrapolations of calibrated parameters seem, in principle,
arbitrary, it is certainly passible to tailor UMRFs to total
quantities predicted via screening models.
NFS POLLUTANT TRANSPORT
One of the foremost issues of long term NPSP forecasting and
control is related to feasible? integrations of runoff and
receiving water models. The need of tailoring the tools
suitable to reliable mass balances for large contributing
areas and complex receiving water bodies is manifest: in
fact, the conceptual chain of. events to be simulated is
const!tued by the array of sources, the mechanisms of
transport, plus the net exchange with a larger water body
(in the actual case, the Adriatic Sea) which is, somehow
questionably, considered as a vessel of infinite capacity,.
Hence, NPSF' model selection is affected by the
characteristics and the quality standard of the receiving
water body. In fact, the output of NPSF1 model simulations
serves as input for the water quality model, whose level of
sophistication needs be tailored to foreseable predicting
T-II-I-6
-------
ability and seal ing-timing of NPSP models. Quality and extent
of the data base, either for the source areas or- -for the
hydrodynamic regimes of the water body, also play a major
role in model selection.
Remarkable efforts have been put, in the recent past, in
establishing the theoretical grounds for simulation of the*
hydrodynamic and environmental regimes of the Venice Lagoon.
Hydraulic approaces (summarized in thorough state-of-the-art
contributions: Ghetti, 1979; Di Silvia and Fiorillo, 1980)
range from zero order lumped mathematical models to complex
formulations of the water quality model. For instance
(Ghetti, 1979; Ghetti et al., 1971; Rinaldo and Marani,
1984), the simpler models of tidal behavior of the Lagoon
work satisfactorily with respect, to prediction of water-
levels and global water balances, while detailed prediction
of local features (velocities and concentrations) needs
further specializations (Ghetti, 1979; Di Silvio and
Fiorillo, 1980; Leendert.se, 1967; Cheng and Walters, 1.982).,
Detailed modeling of movements and transformations of
intensive parameters (Dejak et al., 1981; Jorgensen, 1983;
Leendertse and Gritton, 1971) like temperature, salinity or
nutrient concentrations, induced by the decay or reaction
properties of given tracers, coasts further uncertainty to
the prediction ability of the theoretical tools,,
All of the dispersion models considered are, in some way,
based on mass, momentum and energy balances, ranging from
microscopic transport to macroscopic intertidal phenomena.
Albeit perhaps, the dispersion type of description is locally
good for the Lagoon, it need not be true at the scale of the
entire Lagoon. In other words, the traditional equations of
turbulent convection-diffusion mass balance? might not be
useful for describing fluid motions that are on a scale
comparable with the system itself (Himmelblau and Bishoff,
1968). In view also of these fact, the representation of net-
long term mass transport in a model is inseparably related to
the spatial structure of the physical system (Shanahan and
Harleman, 1984) and needs be resolved via spatial lumping of
appropriate level (Rinaldo and Marani, 1984),,
Furthermore the net transport of mass by time-varying
currents like those tide- or wind-driven within the Lagoon
of Venice, must consider properly velocity fields other than
the mean velocity in fixed vertical sections. Stokes'
drifts (dependent on the time and the distance scales of
the fluctuating currents) must also be accounted for in
tidal period-averaged net transports, according to the
Eulerian framework (Longuet-Higgins, 1969). Quite recently
direct calculations of Lagrangian residual circulations have
also been shown to be feasible -for large tidal water bodies
similar to the Lagoon of Venice (Cheng and Casul1i, 1982).
Lagrangian random walk-type simulations have also been
proposed for far-field dispersion (Chin and Roberts, 1985).
T-II-I-7
-------
Furthermore:
a) -while the mechanism of leakage of urban and industrial
point sources is not. affected by the hydrodynamic
conditions and is characterized by relatively high local
concentrations, NFS tracers are characteristically
discharge dependent as far as either concentration or
total quantities are concerned. The higher dissolved or
solid-phase loads are carried, in fact, by the floods,
which at the same time, yield important, dilution of the
loading. Low concentrations, difficult. simulations of
movements and transformations of its components, time
variability (dependent upon hydrologic and hydraulic
parameters) and large total quantities involved in mass
balances are therefore the main .features of NPSPj
b) - to simulate long-term concentration variations, a lagoon
water quality model must consider the influence of net mass
transport. Long-term lagoon water quality modeling should
therefore implement fast and efficient algorithms, and
subtidal time discretization. Certain discontinuities at.
the sea-lagoon interface need be accounted for, because
gener rally the outflowing water holds higher nutrient.
concentrations than the inflowing ones,, It seems that
dispersion models, as the one already implemented for
environmental studies on the Lagoon (Di Silvio and
Fiorillo, 1980$ De j ak et al., 1981), are tailored to
simulations of current and tracer patterns in a specified
tidal or wind condition only for relatively short simulation
pattern.
At the present stage, it is therefore clear the difficult in
meeting all requirements of prediction/simulation
characterized by short, time step and long time span.
In view of the fulfilling of such twofold objective, the
rules for the modeler seem to follow some speculations:
1) - detailed studies of simplified schemes of two--
dimensional tidal propagations have been carried out
(Rinaldo and Marani, 1984). The most competitive tool for
lumping hydraulic fluxes into any level of aggregation is
related to filtering of velocity arid elevation fields
calculated by advanced finite element techniques (F'utti et.
al., 1985);
2) ~ multiple-vessel spectral models serve at best the
overall purposes of long-term forecasting of dispersion at.
spatial scales like that, of the Lagoon and time scales like
those forced by NPSP inputs and by eutrophication processes
(Rinaldo.and Marani, 1984);
3) - the spatial discretisation of the system of cells
(vessels) simulating the water quality variations over the
horizontal landscape can be linked to the finite element,
grid of hydrodynamic simulation (Putti et. al „ , 1985);
4) ~ theory of signal processing and finds of numerical
spectral analysis induce important simplifications in the
T-II-I-8
-------
mathematics and in the computational loads.
Due to limited field data available -for the Venice Lagoon
system, to date hybrid modeling techniques (mainly referring
to model calibration via more sophisticated models runs)
have been implemented.
EUTROPHICATION OF THE LAGOON
Eutrophication phenomena imply the most, ostensibile and
undesirable facets of phytoplanktonic growth and often
represent an ultimate concern for the environmentalist.
Albeit the phenomenon has long been known within the Lagoon
(as some toponyms referring to algae might suggest) alarming
frequencies and spatial spreading of algal blooms have been
occurring only in recent years. Among the noteworthy
spinoffs of the modified trophic conditions, fish
death-blows (either in fisheries or within the tidal
system), odors and often limitation to recreational
activities are perhe^ps to be mentioned.
While the "cause" in plankton-based systems (certainly due
in most part to the increase in nutrient release from the
drainage area) seems amenable to numerical simulation, model
selection and application for eutrophication effects is
demanding, due**the lack of continuous data on a complex
system which is further characterized by tidal oscillations
of the parameters and marked hydraulic: advection (Rinaldo
and Marani, 1984).
In the actual case of Venice, for which the scheme of
homogeneous well-mixed interconnected vessels proved
satisfactory, a feasible approach for simulation has turned
out to refer to Catastrophe Theory (Zeeman, 1978; Poston and
Stewart, 1978). The foremost advantages of such an approach
are:
— step by step refinements of the qualitative description;
- effective representation of the processes and overall
reliable forecasting of blooming phenomena.
Catastrophe, within environmental systems, means a jump from
an initial equilibrium state to another one,, usually far
from the former, induced by slow changes of the control
variables. Analytical manipulations of the discontinuity
allow application of the theory if certain requirements are
met. Such constraints seemingly fit eutrophication
phenomena. The -recent 1i t er a ture? h as, i n fact, proved
noteworthy successful in application to the field of algal
population growth (Dubois, 1979; Duckstein et al., 1979; Van
Nguyen and Wood, 1979; Kempf et al ., , 1984),.
Catastrophe models may be applied to bimodal (Voinov and
Svirezhev, 1984), piecewise continuous, hysteretical or
T-II-I-9
-------
divergent processes provided that control variables can be
clearly pointed outs algal levels in coatstal or estuarine
waters are amenable to control via Chlorophi11-a (a-Chl)
concentration while, if the bloom only is to be modeled, the
control parameters may be cast into a growth factor,
specialized by an index of nutrient content (Liebig's
limiting factor concentration in most applications to date),
plus an index combining . the effects of temperature and
insulation. During the bloom, in fact, bacterial reactions,
net .zooplankton assumptions and other biological events do
not affect substantially the overall processes,,
As far as phytoplanktonic mass growth is concerned, a typical
bimodal trend is shown to exist. This behavior is
characterized by a state of survival at low concentrations
(10-5-30 ppb of a-Chl) and a. further one at. high concentrations
< 100-5-300 ppb of a-Chl) which is proper to algal blooming. The
back-jump from high to low conditions is patterned on paths
other than that labelling the blooms. Furthermore, evolution
patterns of the phenomena show typical hysteretical trends of
catastrophes.
The only catastrophic surface portrayed by two control is
Whitney's cusp, i.e. a cubic equation dependent upon the
catastrophic variable. In this context, the catastrophic
dependent variable is the phytpp 1 ankton mass measured a.s
concentration of a-Chl; the splitting variable is
concentration of limiting nutrient; and the normal variable
is a lumped parameter of light intensity and water
temperature.
Albeit Whitney's model does not seem very sophisticated
indeed if compared to other schemes (Swartzman and Bentley,
1979), in simulating all phenomena connected to the algal
blooms, nevertheless its structure allows a certain degree
of accuracy in portraying most, critical events and a simple
graphical layout.
As experimental evidence supporting these ideas, a clear
bimodal trend and an overall catastrophic oecurenee, with
change of trend on the 24-th day, is shown by the results
of Figures 3 and 4. A field experiments campaign of chemical
and biological nature (Zingales et al.,1983) has, in
•fact, been carried out aiming at pointing out the
parameters controlling the eutraphi cation of the Lagoon of
Venice. In particular, attention has been focusing on the
water bodies adjacent the industrial sites of larger
size and marked by limited tidal circulation. Existing
temperature gauging have also been checked. The campaign,
performed during spring time/due to its potential embedding
of algal blooms, lasted up to 25 days and considered as
potential indices: pH; turbidity; dissolved oxigen;
salinity; water temperature; and sylicon, ortophosphate,
nitrogen (in various forms) and Chlorophyl1-a concentrations.
T-II-I-10
-------
Algal blooms have been detected only at the station quite
apart -from the industrial sites, due to the presence of
Ammonium-ion levels toxic to algal life. The -foremost data
(temperature, total Nitrogen and Ohlorophyl 1 --a) gaged at the
two stations involved in blooming phenomena are reported in
Figure 3. Figure 4 further illustrates, -for the station near
Tessera, the plot of sequences of equilibrium paths on a
cusp surface.
The normal control variable (water temperature) undergoes a
five-degree excursion connected with the deeper insulation
and wind-calm, Nonethel ess, the high de?gree of mixing
experienced by lagoonal canals, river outlets or the
openings to the sea, where tidal convection is strong, yields
a generally critical determination of the parameters unless
•for shallow low-current sites.,
The second control variable turned out to be total Nitrogen.,
Concentration changes have, in fact, experienced a noteworthy
increase on the 10-th day (right at a 30 mm rainfall event.
in the area) released most likely by croplands just
fertilised. A decrease in salinity and an increase in
turbidity have also been observed,.
Algae populations, the -sawdepedent •-'ar3 ab 1 e of the mode 1 are
formed by BOX diatoms (the most common species) plus
chlorophiceae and cyariophyceae in much smaller proportions.
Certain correlation exists between maxima of Chi orophyl 1 --a
and limiting n u t r i e n t c o n <::: e n t r a t i o n s „ I n f a c t, a ••• • C h 1
concentrations of 250 ppb against maximum Nitrogen
concentrations of 3 ppm are shown At blooming stations
(Tessera) whereas otherwhise (S. Siuliana) the maxima reach
120 ppb (a~-Chl) and 2 ppm (Nitrogen),, The amplitudes of
Chiorophyl1-a concentration oscillations may be further
related to tidal convection and, partially, to the
analytical method,,
PERSPECTIVES OF LONG-TERM ABATEMENT
Nutrient NPSP pollution abatement of the Lagoon of Venice i.B
feasible (may be at low cost, practices) provided that major
sources be pointed out and controlled.. The logical step
forward is hence related to a detailed knowledge of:
nutrient release mechanisms; nutrient movement and
trasformations within the drainage network and within the
L a g o o n proper; a u t a d e p u r a t i ve p rop er t i es of the La g oa n
system. The tool is primarily a continuous data acquisition
system focused on the environmental aspects of the
phenomenon which is, to date, still discussed. Top-priority
is therefore to be granted to creation and maintenance of
such data base and on the related verification of the
theoretical progresses, namely generation, propagation, near--
T-II-I-11
-------
and far-field mixing and dispersion model'
abatement has also been
; upon completion of
f i r a t a n d s e c o n d
total domestic NPS
Nitrogen and 1,500
overall abatement
Domestic pollution
(Bendoricchio et al., 1985)
sewage systems (and via
wastewater treatment), the
out to be 3,400 t/year of
Phosphorus. In practice, the
about, 20% lower levels. Sea outfall for industrial
point sources are being planned, with the further
consi dered
the planned
stages of
load turns
t/year of
will range at
and urban
aid of a
realised by-law reduction in the content of polyphosphates
within commercial domestic detergents from 87. to 5%,,
Urban runoff pollution abatement, albeit not yet considered
in the array of studies in progress, may benefit from the
parallel r e s e a r c. h p r • o g r e s s in the field,,
*
Agricultural releases may be controlled via a set of
techniques of increasing diffusion, mainly referring to'slow
release -fertilisers and controlled drainage (Bkaggs and
Billiam, 1984). In fact, 50% abatements in Nitrate losses
have been measured without losses of production through
Nitrate reduction via s u ta m er g ed dr a ins, an increasing1y
papular practice in the zone,, These techniques added to the
results of the oncoming studies of the Best Management
Practices commissioned by the Regions del Veneto (the local
Government) should guarantee consistent improvements,,
Aids to reduction of .eutrophication phenomena may also
yielded by modifications of tidal circulation within
suffering shoals via revitalization of dead water zones,,
The basic idea underlined by the actual work is to single
out the relative importance of each elementary mechanism
affecting the eutrophi cat. i an of the Lagoon and to
establish the theoretical grounds for an .assessment. of
control policies for the foremost causes. Hence, the
mentioned abatement approaches, efficient in a cost/benefit
perspective, have been outlined.
The study is focusing on the fate of nutrients in surface
waters as a major contributor to the actual decay, since
toxic substances of industrial ,, a g r i c u 11 u r a 1 a n d u r b a n
source may just requires a .thorough control even in
commercial routes™
As a final comment, a noteworthy environmental aspect, as
far as .either nutrients or toxic substances are concerned,
is r e1 at ed to lagoonal sed i men ts« They ha ve n ot been
mentioned insofar because they might require ad-hoc
interventions on low-current zones and mechanical remotion
in some instances.
Through the complex set. of planned interventions, among which
T-II-I-12
-------
only a part already operational, the complexity of the
problems can be inferred and hence the fascination to
researchers, professionals and, may be, politicians.
ACKNOWLEDGEMENTS
The researches are supported by funds granted by the National
Research Council (CNR), the Ministry of Public Instruction
(MPI), and the Regione Veneto.
T-II-I-13
-------
REFERENCES
Bendoricchio G. , and A. Rinaldo, Un rnodello matematico del
dilavamento di sostanze chimiche da terreni agricoli,, Atti &
Memorie del 1 'Accademia Patavina di SS.LL.AA., XCIII, 137-5-154,'
1982.
Bendoricchio B. , C. Camis, and S« Alessandrini, Evoluzione
temporale del 1 *inquinamento da sorgenti diffuse di origine
agricola. Inquinamertto, 25 (12) , 61*64., 1984,,
Bendoricchio S., L. Agostinettq, and S. Alessandrini, • Nutri-
ent! di origine diffusa provenienti- dal bacirto scolante in
Laguna di Venezia. Acqua Aria, 2, 177-5-184, .1.985.
Cheng R.T., and V» Casulli, On Lagrangian Residual Currents
With Applications to South San Francisco Bay, California.
Water Resour. Res., 1.8, 1652*1662, 198.2.
Cheng R« T., and R. A. Walters, Modeling of Estuarine Hydro
dynamics and Field Data Requirements. In: Finite Elements in
Fluids, Vol. 4, Edited by R. H. "Gallagher, J. T. Helen, C.
Taylor, and D. C,, Zienkiewicz, 'pp.' 891108, .J. Wiley ?< Sons,
New York, N.Y., 1982.
Chin DuA., and J.W, Roberts, Model of Dispersion in Coastal
Waters. J. Hydr. Eng., Ill, '12128, lc?85.
Cdssu R., E. de Fraja Frangipane, R. Donazzolo, and A. A.
Orio, Elementi per il bilancio di nutrient.! nella Laguna di
Venezia. Ingegneria Ambient-ale, 13,. 80191, 1984.
Dejak C. , 1. Mazzei Lalatta, and M. Mol in, Dispersions di in--
quinanti nella Laguna di Venezia (modello bidimensionale
avvetti vo~-di f f usi vo) . Proceedings of the Seminar "Process! di
mescol amenta in arnbienti lagu'nari e lor a model 1 &z i one" , Roma,
(1), 25155, 1981.
Dickerhoff Delwiche L, L.,and D. A,, Haith, Loading Functions
for Predicting Nutrient Losses From Complex Watersheds.. Water
Resour. Res., 19, 9511959, 1983.
Di Silvio, G. , arid L. D'Alpaos, II comportamento del la Laguna
di Venezia esaminato con il metodo propagatorio unidimensio-
nale. Istituto Veneto di Scienze, Lettere ed Art! - Rapporti
e Studi,
-------
Regulation o-f Rivers. State of th
ing, 7, Copenhagen, 7511-758, 1979.
the Art in Ecological Model
Duckstein L. , J. Cast! , and J, Kempf , Modeling Phytoplankton
Dynamics Using Catastrophe Theory. Water Resour. Res., 15,
1189*1194, 1979.
Ghetti A., L. D'Alpaos, and R. Dazzi, La regolazione delle
bocche della Laguna di Venesia per 1 'attenuazione delle acque
alte indagata con metodo statico. Atti e Memorie del 1 * Accade-
mia Patavina di SS.LL.AA., XVII, 11-5-41, 1971.
Ghetti A., Etudes concernant les problemes hydrodynamiques de
la Lagune de Venise. Proceedings o-f the XVII Congress o-f
IAHR, Cagliari (Italy), (6), 56-5-94, 1979.
Goldmann A., R. Rabagliati, and P. Sguaszero, Characteristics
of the Tidal Wave in the Lagoon of Venice. IBM Tech. Rep.
CRV009/5 13-3539, 47, 52 pp. , 1975.
Haith D. A., Development and Testing of Watershed Loading
Functions for Nonpoint Sources. Proceedings of Thirteenth
Annual Pittsburgh Conference on Modeling and Simulation.
Part 4, 1463-5-1468, 1982.
Haith D. A., and J. V. Dougherty, Non-Point Source Pollution
from Agriculture Runoff. Journal of the Environmental Engi-
neering Division, ASCE, 102, 1055-5-1069, 1976.
Himmelblau D. M. , and K. B. Biscohoff, Process Analysis and
Simulation: Deterministic Systems. J. Wiley & Sons, New York,
N.Y. , 348 pp. , 1968.
Jolankai G. , Modelling of Nonpoint Source Pollution. In: S.E.
Jorgensen (Editor), Application of Ecological Modelling in
Environmental Management, Part A., Elsevier Scientific Pub-
lishing Co., Amsterdam (N) , 283-5-385, 1983.
Jorgensen S. E. , Application of Ecological Modelling in Envi-
ronmental Management, Part A. Elsevier Scientific Publishing
Co., Amsterdam (M) , 735 pp., 1983.
Kempf J., L. Duckstein, and J. Casti , Relaxation Oscillations
and Other Non-Michael i an Behavior in Slow-Fast Pyt op lank ton
Grow Model. Ecological Modeling, 23, 67*90, 1984,.
Leendertse J. J., Aspects of a Computational Model for Long-
Period Water-Wawe Propagation. Rep. RM-5294-PR, Rand Corp.,
Santa Monica, California, 1967.
Leendertse J. J., and E. C. Gritton, A Water-Quality Simula-
tion Model for Well-Mixed Estuaries and Coastal Sea, 2,
Computation Procedures, Rep. R-708-NYC, Rand Corp., Santa
Monica, California, 1971.
T-II-I-15
-------
Longuet-Higgins M. S. , On the Transport of Mass by Time-
Varying Ocean Currents, Deep-Sea Res. , 16, 431-5-447, 1969.
Marani A. , and G. Bendoricchio, Models of Statistical Distri-
butions for NPSP Concentrations. Proceedings of International
Conference on "Agriculture and Environment 1984". Venezia,
11-5-15 June, 1984.
Ministero dei Lavori Pubblici, Le correnti di marea nella La~-
guna di Venezia. Istituto di Idraulica del 1'University di.
Padova, Padova, 1979,
Poston, T. and I. Stewart, Catastrophe Theory and Its Appli-
cations, Pitman, London, 408 pp.,. 1978.
Putti M. , A. Rinaldo, and A, Marani, Long Term Water Quality
of Tidal Water Bodies Via Mixed Spettral-FE Methods. Forth
coming, 1985.
Rinaldo A., Manpoint Source Pollution from Agricultural Run-
off: a Large Scale Application of Mathematical Modeling.
Proceedings of Thirteenth Annual Pittsburgh Conference on
Modeling and Simulation. Part 4, 1570*1.574, 1.982.
Rinaldo A., and A. Rinaldo, Determinations dei singoli reca--
piti delle quantita di fosforo .e asoto che provengono annual---
mente in Laguna di Venezia dal proprio bacino scolante rela-
tivamente ai territori agricoli. Atti del Convegno di Studi
"Lagunai, fiumi, lidi; cinque secol i di gestione delle acque
nelle Venezie", Ministero dei Lavori Pubblici, Magistrate al-
le acque, Venesia, 10-5-12 giugno 1983.
Rinaldo A., and A. Marani, Runoff .and Receiving Water Models
for NPS Discharges into the Venice Lagoon. Proceedings of
International Conference on "Agriculture and Environment
1984". Venezia, llflS June, 1984.
Skaggs R. W. , arid J, W. Gilliam, Modeling Subsurface Drainage
and Water Management Systems to Alleviate Potential Water
Quality Problems. Proceedings of International Conference on
"Agriculture and Environment 1984". Venezia, 11-5-15 June,
1984.
Shanahan P., and R. F. Harleman, Transport in Lake Water
Quality Modeling. J. Environ. Eng., ASCE, 110, 42-f57, 1984.
Swartzman G.L., and R. Bentley, A Review and Comparison of
Phytoplankton Simulation Models. ISEM Journal, 1, 30*82,
1979.
Van Nguyen V., and E. Wood, On the Morphology of Summer Algae
Dynamics in Non-stratified Lakes. Ecological Modeling, 65
117-5-131, 1979.
T-II-I-16
-------
Voinov A.A., and Yu.fl. Svirezhev, A Minimal Model of
Eutrophication in Freshwater Ecosystems., Ecological Modeling,
23, 2777-292, 1984.
i»
Zeeman E. C., Catastrophe Theory: Selected Paper 1972-1977,
Addison-Wesley Publishing, Reading, Massachusetts, U.S.A.,
700 pp., 1978.
Zingales F., S. Alessandrini, G. Bendoricchio, C. Comis, A.
Marani, F. Pianetti, A. Rinaldo, and C. Sartori-Baratto, A
Model of Nanpoint Source Pollution in the Agricultural Run-
off. Proceedings of Thirteenth Annual Pittsburgh Conference
on Modeling and Simulation,, Part 4, 1575-5-1580, 1982.
Zingales F. et Al., Indagini Idratermodinamiche e Biologiche
per la valutazione dei riflessi ambientali del funzionamento
a piena potensa della Centrale. Technical Report--E. N.E. L. ,
Venesia, 1983.
Zingales F., A. Marani, G. Bendoricchio, and A. Rinaldo, A
Conceptual Model of Unit-Mass Response Function -for Nanpoint
Source Pollutant Runoff. Ecological Modeling, 26, 285v311,
1984.
T-II-I-17
-------
I
»-*
1
»-*
CD
Figure 2 - Mainland draining into the Lagoon o-f Venice: spatial
distribution o-f nutritent loads.
-------
ucow «r ima
UONH V HMO
-------
rsj
o
Figure 1
THE VENICE LAGOON SYSTEM
In block, largo navigation channels, tha Chioggia,
Malamocco, Lido mouthsj cantor tha city of Venicoi
towards tho mainland boundary olougho and marches.
-------