vvEPA
United States
Environmental Protection
Agency
Municipal Environmental Research EPA-600/2-79-050a
Laboratory Ju|y -, 97g
Cincinnati OH 45268 y
Research and Development
Maximum
Utilization of Water
Resources in a
Planned Community
Executive Summary
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2, Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECH-
NOLOGY series. This series describes research performed to develop and dem-
onstrate instrumentation, equipment, and methodology to repair or prevent en-
vironmental degradation from point and non-point sources of pollution. This work
provides the new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-79-050a
July 1979
MAXIMUM UTILIZATION OF WATER RESOURCES
IN A PLANNED COMMUNITY
Executive Summary
by
William G. Characklis, Frank J. Gaudet
Frank L. Roe and Philip B. Bedient
Department of Environmental Science and Engineering
Rice University
Houston, Texas 77001
Grant No. 802433
Project Officers
Richard Field
Anthony N. Tafuri
Storm and Combined Sewer Section
Wastewater Research Division
Municipal Environmental Research Laboratory (Cincinnati)
Edison, New Jersey 08817
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
This report has been reviewed by the Municipal
Environmental Research Laboratory, U.S. Environmental
Protection Agency, and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental
Protection Agency, nor does mention of trade names or
commercial products constitute endorsement or recommen-
dation for use.
-------
FOREWORD
The Environmental Protection Agency was created
because of increasing public and government concern about
the dangers of pollution to the health and welfare of the
American people. Noxious air, foul water, and spoiled
land are tragic testimony to the deterioration of our
natural environment. The complexity of that environment
and the interplay between its components requires a con-
centrated and integrated attack on the problem.
Research and development is that necessary first step
in problem solution and it involves defining the problem,
measuring its impact, and searching for solutions. The
' Municipal Environmental Research Laboratory develops new
and improved technology and systems for the prevention,
treatment, and management of wastewater and solid and
hazardous waste pollutant discharges' from municipal and
community sources, for the preservation and treatment of"
public drinking water supplies, and to minimize the ad-
verse economic, social, health, and aesthetic effects of
'pollution. This publication is one of the products of
that:; research; a most vital communications link between
the researcher and the user community.
This project.focuses on methods maximizing the use of
water resources in a planned urban environment, while
minimizing,their degradation. Particular attention is
being directed towards determining the biological, chemical,
hydrological and physical characteristics of stormwater
runoff and its corresponding role, in the urban water cycle.
Francis T. Mayo
Director
Municipal Environmental Research
Laboratory
111
-------
ABSTRACT
An ecologically planned community (The Woodlands,
Texas) has adopted a unique water management plan designed
to avoid adverse water quality and hydrologic effects due
to urbanization while benefiting from the existing natural
drainage. The water quantity and quality were monitored
by a comprehensive sampling and analytical program to
evaluate the physical, chemical and biological effects of
its implementation.
Chemical parameters monitored include oxygen demand,
organic carbon, nitrogenous compounds, phosphate compounds,
dissolved oxygen, pH, specific conductance and pesticides.
Numerous indicator bacterial organisms and pathogenic bac-
teria were enumerated as were various aquatic and edaphic
algal species. Disinfectant demand and algal bioassays
were also conducted on stormwater runoff.
Relationships were developed between stormwater run-
off quality, land use and runoff quantities in an effort
to predict pollutant loads. The load-runoff relationships
were utilized in a modified version of the EPA Stormwater
Management Model (SWMM) to simulate stormwater runoff
quantity and quality for watersheds using the "natural
drainage" concepts at The Woodlands.
This report was submitted in fulfillment of Grant No.
802433 between the U.S. Environmental Protection Agency
and Rice University, Department of Environmental Science
and Engineering. This report covers the period July 16,
1973, to May 31, 1976, and work was completed as of
December 31, 1976.
IV
-------
CONTENTS
Foreword. . . ,
Abstract. . . ,
Figures . . . ,
Tables. . . . ,
Abbreviations .
Acknowledgment.
. 111
. iv
. vi
. ix
. xi
.xii
1.
2.
3.
4.
5.
6.
Introduction • t ,
Conclusions .*.".'!!!!." 4
Recommendations ! ! ! ! 8
Site Descriptions .....!! n
Sampling and Monitoring Programs ...!!!!*""" 23
Results and Discussion !!!!"""" 31
References,
130
-------
FIGURES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Location of study sites ...
Existing drainage network for The Woodlands ....
Schematic water balance for The Woodlands . . . t .
Hunting Bayou watershed land use . . , . t
Westbury Square watershed ....,
Location of sampling sites and rain ge.uges within
the Panther Branch watershed. ...
The Woodlands man-made lake system with locations
of stormwater monitoring sites . . . t
Site plan for the porous pavement parking lot . . ,
Temporal distribution of polychlorinated biphenyls
in The Woodlands . . »
Temporal distribution of polychlorinated biphenyls
in aquatic fauna in The Woodlands
Temporal distribution of Mir ex in The Woodlands
Temporal distribution of Mir ex in the Conference
Center Lakes (A and B)
Temporal distribution of chlordane in The Woodlands
Seasonal algal standing crops at P^IO in panther
Branch
Seasonal algal standing crops at p-30 xn panther
Branch
Generalized pollutographs observed for stormwater
parameters '....„....,.
12
15
17
21
22
25
28
30
46
47
49
50
51
61
62
67
-------
Number
FIGURES (continued)
17 Load-runoff relationships for Kjeldahl nitrogen
and total phosphorus. . .
" *
18
19
20
21
22
23
24
25
27
28
29
30
32
Page
69
7Q
71
Load-runoff relationships for nitrate and ammonia
Load-runoff relationships for total suspended solids
and total COD as related to the watersheds studied.
Load-runoff relationships for soluble chemical oxygen
demand (COD) and soluble organic carbon. ..... 72
Comparison of P-10 and P-30 temporal distribution of
streamflow, total suspended solids and total COD
for the storm event of April 8, 1975. ...... 77
Comparison of p-10 and p-30 temporal distribution
of streamflow, total Kjeldahl nitrogen and total
phosphorous for the storm event of April 8, 1975.
The Woodlands construction activity in relation to
the p-10 and p-30 sampling sites
Scalar approach to FC/FS patterns in different
land use areas . . .
Hydrograph and predicted solids loads for the p-30
hydrograph period of 10/28/74 to 4/12/75
26 Hydrographs and observed and simulated mass flow
curves for P-30 storm events. .
Fitted curves for storm runoff and pollutant mass
flows observed at p-10 on 4/8/75
Hydrograph and cumulative hyetograph at the Lake B
gauging station for the April 8, 1975 storm
event
78
79
81
85
89
Master programming routine in the SWMM. ...... 91
Subcatchments and drainage network in Hunting
Bayou ........
*
94
Predicted hydrographs for suspended solids concen-
trations at Hunting Bayou (5/8/75) ........ 96
31 predicted pollutograph and mass flow for suspended
solids at Hunting Bayou (5/8/75) ......... 97
107
-------
FIGURES (continued)
Number
33 Reduction of total suspended solids through The
Woodlands lake system 109
34 Comparison of survival-mortality characteristics
of fingerling catfish to chlorine exposure in a
static bioassay and in a flow through bioassay. . 113
35 Survival-mortality characteristics of fingerling
catfish to ozone exposure in a flow through
bioassay 115
36 Effect of short term chlorine exposure on the up-
take of 22Na by the gills of Ictalurus puneta-
tus 117
37 Effect of short term ozone exposure on the uptake
of ^Na by the gills of Ictalurus punctatus . . . 118
38 Optical densities of Selenastrum capricornutum
after incubation in water from Panther Branch and
Spring Creek 119
39 Growth of Selenastrum in stormwater runoff in re-
lation to hydrograph position at P-30 121
40 Growth of Selenastrum in stormwater runoff from
Hunting Bayou
122
41 Growth of Selenastrum in stormwater runoff from
Westbury Square 123
42 Water depth and rainfall for storm event of
3/6/76 to 3/8/76. 124
Vlll
-------
TABLES
Number
I
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Hunting Bayou Watershed Characteristics. ......
Westbury Square Watershed Characteristics. .....
Watershed Characteristics
Summary of LowHFlow Water Quality Parameters ....
Storm Event Hydrology Summary. ...........
Hydrological Definitions and Calculations
Rainwater Quality Analysis ....
Comparison of Rainwater and Runoff Quality in
Houston and at The Woodlands
Comparison of Groundwater Quality After Rainfall to
Runoff Water Quality
Runoff Water Quality Summary — Mass Flow and Weighted
Average
Summary of Westbury Runoff Bacterial Quality ....
Summary of Indicator and Pathogen Base Levels, Peak
Levels and Mean Levels at The Woodlands
Summary of Panther Branch Aquatic Algae
Summary of Edaphic Algae in The Woodlands
Water Quality for Stormwater Runoff, Untreated
Sewage, and Treated Sewage
Comparison of Stormwater and Wastewater Micro-
biological Analysis
Pollutant Load Ranking of the Pour Study Area
Page
21
22
26
32
34
36
37
38
39
41
5.2
54
60
63
64
66
73
-------
TABLES (continued)
Number
18
19
Annual Mass Loads from P-10 and P-30 Watersheds. r
Comparison of Stormwater Quality at P-10, p-30 and
Developing Areas During Storm #10. .......
Page
75
80
20 Comparison of simulated and Observed Results for
21
22
23
24
25
26
27
28
29
Tnree storms
Modeling Requirements by SWMM
Comparison of SWMM Predicted Results with Observed
Flow Measurements for Hunting Bayou Storm Events .
Water Quality Modeling Results for panther Branch
for Storm Event of 12/5/74
Modeling Results for Future Development Upstream
for Lake B
Summary of Water Quality Parameters for sites Lake A
and Lake B During the April 8, 1975 Storm Event. . .
Stormwater Sediment Removal at Lake Harrison ....
Coefficient of Friction
Summary of Stormwater Quality for Porous Pavement
Storm on 2/20/76
Noise Levels
87
92
95
101
103
110
111
125
126
127
-------
LIST OF ABBREVIATIONS
BOD Biochemical Oxygen Demand
COD Chemical Oxygen Demand
DDE 2,2-bis-(p-chlorophenyl)-l,l-dichloro ethylene
DO Dissolved Oxygen
FC Fecal Coliforms
FS Fecal Streptococci
g mass flow rate
HB Hunting Bayou watershed
LC50 Concentration of toxicant that is lethal to exactly 50%
of the test organism during.continuous exposure for a
specified period of time.
ortho P Orthophosphate
P mass of pollutant
PCB polychlorinated biphenyls
PDS slope of the load-runoff curve at some point in time
PS Pseudomonas aeruginosa
PVC Polyvinyl Chloride
Q volumetric flow rate
r rate of runoff
S total storage
SA Salmonella sp.
SOC Soluble Organic Carbon
ST Staphylococcus sp.
SWMM EPA Stormwater Management Model
TC Total Coliforms
TKN Total Kjeldahl Nitrogen
TOG Total Organic Carbon
TP Total Phosphate
TSS Total Suspended Solids
tp time to peak flow in a hydrograph
WB Westbury watershed
xi
-------
ACKNOWLEDGMENTS
This project was supported by the U. S. Environmental
Protection Agency, The Woodlands Development Corporation and
Rice University.
The Project Director expresses his sincere gratitude to
the EPA Project Officers, Anthony Tafuri and Richard Field and
to James Blackburn, Skip Christy, Peter D'Alessandro, Ralph
Everhart, Bill Kendricks, Ken Kirribrough, Robert Heineman, Plato
Pappas and James Veltman of The Woodlands Development Corpora-
tion for their time and valuable contributions .
Acknowledgments also go to Robert Gabrysch, Jim Hutchinson,
Steve Johnson, Emil Kamanski, and Robert Smith of the Water
Resources Division, U. S. Geological Survey, for collecting and
compiling essential hydrologic data at The Woodlands.
Credit is also due to the members of this multidisciplinary
project. They are as follows:
Rice University
P. B. Bedient
E. Birch
J. Bishop
K. Carter
W. G. Characklis
J. Coffey
j. A. Conner
M. Curtis
S. Davis
J. Delia
F. M. Fisher
G. Fortenberry
F. J. Gaudet
D. Gee
P. Graves
J. H. Hall
B. Hammond
T. D. Hayes
D. Harned
J. D. John
M. A. Kessick
J. M. King
J. LeBlanc
M. Lee
H. M. Liljestrand
W. L. Lloyd
K. Manchen
D. C. Marks
R. Morrison
L. P. Metzgar -
T. Miller
P. MeSherry
L. Price
F, L. Roe
J. B. Smith
C. Stagg
M. Walker
C, H. Ward
J. C. Weismiller
L. Wong
A. Yarletts
J. S, Zogorski
XII
-------
ACKNOWLEDGMENTS (continued)
University of Texas School of Public Health - Houston, Tx,
D. Casserly p. Mittlemark
E. M. Davis D. Moore
J. Greene H. Tamashiro
P. Mattox
Espey, Huston and Associates, inc. - Houston, Tx.
W. H. Espey, Jr. T. Remaley
E. Diniz F. Sofka
D. Holloway D. E. Winslow
S and B Engineers - Houston, Tx.
L. Chandler
W. Davis
J. Matson
Franklin Institute Research Laboratory - Philadelphia, Pa,
R. Hoilinger
E. Thelen
Kill
-------
-------
SECTION 1
INTRODUCTION
Non-point source pollution has become of greater concern
due to increasingly stringent point source effluent standards
and rapid development of urban areas. Urban stormwater runoff
is one of the major non-point pollution sources and innovative
methods must be developed to minimize the impact of stormwater
runoff on receiving waters. One such imaginative water resource
plan has been developed for The Woodlands, a planned community in
southeast Texas. The U.S. Environmental Protection Agency,
Woodlands Development Corporation and Rice University have com-
pleted a three year research/demonstration project to evaluate
the water resource system at The Woodlands and develop strategies
for maximizing the benefits to the community while minimizing the
effect on receiving waters.
The hydrological characteristics of a natural watershed
change with urbanization. Replacement of flow-retarding vegeta-
tion with impervious surfaces, such as roads and buildings, in-
creases the rate and amount of stormwater runoff. Removal of
the water is traditionally implemented by the use of an urban
drainage system consisting of storm sewers and/or deep, concrete-
lined drainage ditches, designed specifically for rapid drainage.
Increased runoff volumes and peak flow rates result, creating
problems of downstream flooding and channel erosion.
Infiltration of stormwater is a major groundwater recharge
source, however emphasis on surface removal minimizes the in-
filtration rate, resulting .in a lowered water table and possible
urban land subsidence problems. Water quality deteriorates be-
cause natural purification provided by infiltration is compro-
mised.
The urban environment, typified by high population density,
provides a major pollutant source for runoff waters (1). Recent
investigations recognize the significance and magnitude of pollu-
tion problems from urban stormwater runoff. In terms of specific
pollutants, the sediment yield problem is the most dramatic. Due
primarily to urban construction, urban sediment loads were found
to be as much as 75 times greater than loads in agricultural
regions (2, 3). Other runoff pollutants reported higher in urban
regions include dissolved solids (4), coliforms (5), biochemical
oxygen demand and chemical oxygen demand (BOD and COD), poly-
chlorinated biphenyls, heavy metals, pesticides and fer-
-------
tilizers (6-9). An increase in bacterial content is also re-
ported. Claudon (10) stated that agricultural and urban runoff
regularly contribute Salmonella sp. to recreational waters.
High fecal coliform an3fecal streptococcus numbers have been
found in urban runoff (11). Snowmelt and related agricultural
runoff in far northern climes in the continental U.S. have been
shown to contribute high densities of indicator bacteria and
pathogens to runoff (12, 13).
The addition of large quantities of inorganic nutrients,
particularly nitrogen and phosphorous, to freshwater lakes poses
a serious problem in lake management. Municipal sewage (14, 15,
16), agricultural drainage (17), managed forestland drainage (18)
and fertilization often accelerate the natural process of eutro-
phication, thus enhancing the growth of bacteria, algae and aqua-
tic vascular plants. Population densities of these organisms of-^
ten reach nuisance proportions and interfere with the aesthetic
qualities and recreational values of lakes. These "blooms" may
discolor, impart unsatisfactory tastes (19, 20) and excrete toxins
into the water (21, 22, 23). They can also clog treatment plant
filters (24), and, upon decomposition, produce foul odors (14).
Late summer "blooms"'can create anoxic conditions and cause the
death of fish. Accordingly, value of lake properties may depre-
ciate and there may be increased burdens on municipal water sys-
tems due to added costs of filtration and deodorization of the
water.
Although literature on eutrophication is extensive (25-29),
most limnological studies were conducted on lakes that were
^either eutrophic, or non-eutrophic at the time of study, and few
studies were continued long enough to follow changes in the
trophic status of lakes. Even fewer studies have been initiated
on a drainage system before the construction of lakes. Thus,
complete developmental histories of the water resources of par-
ticular areas are lacking. Also, it is evident that information
on more effective methods of removing mineral nutrients from ef-
fluents prior to release into natural waters is badly needed
(30).
OBJECTIVES
The overall goal of this research project was to evaluate
the water resource plan for The Woodlands and to make recommenda-
tions as necessary to maximize its effective utilization through
alterations in design or management.
One of the major objectives was to modify and expand the
capabilities of the EPA Storm Water Management Model (SWMM) to
apply to the "natural drainage" system designed for The Wood-
lands. SWMM has been expanded to include the following addi-
tional water quality parameters: total COD, total Kjeldahl
nitrogen (TKN), nitrates and phosphates. The model has been
used to evaluate the effectiveness of the "natural drainage"
-------
system in minimizing changes in storm runoff quantity and quality
and to assist engineers and planners in designing the drainage
system for future development phases at The Woodlands.
Since urban areas are typified by a variety of land uses,
including residential, commercial and industrial as the broadest
categories, determining effects of land use on stormwater quality
was conceived as another major project objective. The Woodlands
development plan could not provide sufficiently diverse study
areas during the project period. Consequently, urban watersheds
of similar physiographic and drainage characteristics in Houston
were studied to further relate storm runoff water quality to
urban land use.
In order to accomplish these two primary objectives, a mas-
sive sampling and monitoring program was established. Rainfall,
streamflow and over twenty-five water quality parameters were
monitored on a regular basis. Included in the water quality_
program were intensive programs concentrating on bacteriological
water quality, chlorinated hydrocarbons and phytoplankton identi-
fication and enumeration.
The sampling program included bacteriological tests to eval-
uate the traditional relationship between indicator organisms
(e.g. fecal coliform, fecal streptococcus) and pathogens in
stormwater runoff. Disinfection experiments were conducted to
determine relative effectiveness of Cl^, Br2 and 03 in untreated
stormwater runoff. Toxicological testing determined maximum
tolerable concentrations of disinfectant in the receiving stream
for maintenance of fish populations. Algal bioassays were con-
ducted to experimentally determine conditions, including nutrient
concentrations, necessary to prevent eutrophication in The Wood-
lands ' lakes.
Finally, the performance of a porous pavement was compared
to a conventional pavement with regard to runoff amount and
quality, wet skid resistance, hydroplaning and other charac-
teristics related to driveability.
-------
SECTION 2
CONCLUSIONS
Runoff Quality
From January 1975 to April 1976, seventeen storms were moni'-
tored at six different watersheds in the Houston metropolitan
area. The intensive sampling and analytical effort resulted in
the following conclusions:
1. Stormwater runoff from an undeveloped, forested watershed
is relatively low in pollutants and pollutant indicators. Typical
values are: total phosphorous (TP) 0.06 mg/1, total Kjeldahl ni-
trogen (TKN) 1.24 mg/1, total suspended solids (TSS) 36 mg/1,
total COD 42 mg/1, and dissolved oxygen (DO) 6 mg/1.
2. Development in forested watersheds significantly increases
suspended solids and nutrients in runoff. COD and other organic
parameters are not affected. Development increases surface water
turbidity. Increased nutrient loads from developed areas will
create algal and macrophyte growth problems in urban lakes.
3. Urban runoff contains higher nutrient and solids loads
than forest runoff. Nutrient concentrations (ammonia, TKN,
nitrate, nitrite, TP, ortho phosphates) are as much as 10 times
greater in urban areas. TSS concentrations are 4 times greater.
Higher concentrations, combined with increased runoff coefficients
(the amount of runoff for a given amount of rainfall), provide
receiving waters with heavy pollutant loads in urban areas. Sedi-
ment buildup and algal growth problems result.
4. A man-made lake serves as an effective trap for excessive
sediments transported by construction site runoff. During seven
separate storm events totaling 10.2 in. (26 cm) of rainfall, 180
tons (1.6 x 105kg) of sediment entering 110 ac-ft (13.56 ha-m)
Lake Harrison was reduced to 34 tons (3.08 x 104 kg) in the
effluent. This was an 81% reduction in sediment load.
5. A definite first flush is observed for urban and undevel-
ped watershed runoff, most commonly for TSS and turbidity para-
meters. The flush is related to transport of streambed sediments.
Urban drainage systems have increased transport potential and
therefore exhibit higher flush concentrations.
6. Rainwaters contain phosphates, nitrogen and COD which ac-
count for a significant portion of runoff pollutant loads.
-------
7. Undisturbed soils are capable of removing nutrients
found in rainwater. Disturbed soils in developing areas lose
this capability.
8. Municipal wastewater would require advanced treatment
to compare with nitrogen and phosphorous concentrations in storm-
water runoff. Secondary treatment of wastewater will lower TSS
and COD concentrations below that in stormwater runoff.
9. A linear relationship exists between total pollutant
loads and total stormwater runoff, which is useful in comparison
between watersheds and analytical prediction of stormwater pollu-
tant loads.
10. A statistical ranking of four watersheds, on a Ib/ac/in
of runoff basis, indicates that urban watersheds are clearly the
greatest producers of suspended solids and nutrient loads. Loads
from forested and developing watersheds are lower by as much as
an order of magnitude.
11. Load-runoff curves may be used to sequentially simulate
mass flow curves. Simulation of a six-month period, containing
three measured storm events, produced, reasonable comparisons of
of observed and simulated curves.
Eutrophication Potential
1. Algal associations encountered at The Woodlands were indi-
cative of oligotrophic or slightly mesotrophic waters.
2. Phosphorous is the limiting nutrient for algal growth in
the Conference Center Lakes and Panther Branch during low flow,
while nitrogen is more limiting in stormwater runoff samples.
Consequently, operation of the phosphorous removal process for the
treated sewage effluent entering the lakes is not necessary during
rainy periods. The savings could amount to $50,000/year.
3. Surface runoff from fertilized soilds serves as a source
of nutrients and troublesome algae. Fertilization of the golf
course, with subsequent increase in soil pH, results in larger
standing crops of blue-green algae and diatoms in the soil.
4. Urban stormwater runoff contains higher bacterial concen-
trations than runoff from forested areas. Substantial numbers of
bacteria, including pathogenic species, were found. Stormwater
runoff generally exceeded state recreational standards for fecal
coliforms (200 cells/100 ml).
-------
Disinfection
1. Excessive doses of chlorine or ozone are required for
effective disinfection of stormwater runoff. Chlorine and ozone
demands greater than 8 mg/1 and 32 mg/1, respectively, were mea-
sured, primarily a result of high TSS concentration.
2. Toxicity studies using channel catfish (Ictalarus punctalus)
in a flow-through bioassay system have established maximum safe
surface water concentrations at 7 yg/1 and 3 yg/1 for chlorine
and ozone, respectively.
Porous Pavement
A porous pavement parking lot was installed at The Woodlands,
Texas and its performance compared to a conventional, dense pave-
ment parking lot. The following conclusions resulted from the study:
1. Porous paving may be used effectively to store and release
stormwater which would otherwise cause erosion or flash flooding.
The quality of the released water is generally better than that
of runoff from standard paving.
2. Unacceptable lead concentrations suggest that stored
water under porous pavement should be prevented from contacting
drinking water supplies.
3. Porous pavement is comparable to conventional paving in
terms of driveability and safety. Mud and dirt from construction
activities in the vicinity can clog porous paving.
4. Periodic maintenance of the paving should be performed by
brush sweepers with vacuum followed by high pressure water washing.
5. A more durable porous pavement can be produced by using
lower penetration or stiffer asphalt than that at The Woodlands.
Stormwater Management Model
The SWMM release of February 1975, referred to in this report
as the original SWMM version, was extensively modified. The capa-
bilities of the modified version have been expanded to model run-
off and water quality from natural drainage areas. The study areas
where the new capabilities were tested are The Woodlands and
Houston, Texas. During the course of this study the following con-
clusions were reached:
1. After correction of errors in infiltration rate computation,
the modified SWMM predictions of observed peak and total discharge
were relatively accurate but the pollutant concentration predic-
tions, which are dependent on exact hydrograph replication, could
not be modeled for the study area.
-------
_ 2. Although runoff from natural drainage areas is of better
qualzty than that from areas with conventional storm sewer drain-
age, the effect of construction activity in both types of areas
could not be determined by the original SWMM version. The pre-
dicted values were always too low. But, modeling of erosion from
construction activities is now possibile by use of the modified
SWMM version.
3. Biochemical oxygen demand (BOD) data produced inconsis-
tent results and, therefore, BOD modeling proved unsatisfactory
Data for COD were consistent and were used to model oxygen demand.
4. The functional relationship between pollutant mass and
runoff volume can be linearized by the use of logarithmic trans-
froms. Resultant linear equations can be used to determine loading
rates and total pollutant transport from a watershed.
5. The exponential pollutant removal or decay coefficient
can be considered as a contstant in all geographical areas. The
modified SWMM allows for selection of the value of this coeffi-
cient by the user.
6. The modified SWMM can transport flow through natural
channels with a minimum of input data requirements. Each natural
channel is described by a Series of coordinates, and the program
now calculates the area-discharge curves which formerly had to
be input for each natural, channel.
7. The modified SWMM can determine the cost efficiencies in
the use of_natural drainage systems relative to those for conven-
tional drainage systems using either user supplied or default unit
cost estimates.
8. The modified SWMM can provide a detailed analysis of
storage_and flow into and out of porous (pervious) pavement systems.
The drain outflow, surface runoff, and storage volumes can be deter-
mined; but the lack of comprehensive data precluded the modeling
of water quality in porous pavements.
9. The modeling schemes developed during this study require
considerable input data preparation and, consequently, the modi-
fied SWMM, when applied to natural drainage systems, is more user
-------
SECTION 3
RECOMMENDATIONS
Sediment discharge from construction area stormwater runoff
should be controlled by sedimentation basins or more sophisti-
cated methods.
Research is needed to establish the relationship of air
quality to rainwater and stormwater runoff quality. Special
attention should be focused on effects of projected ambient air
quality criteria.
N
More standardized methods should be instituted to describe
storm events and watershed characteristics and to reduce data
resulting from a stormwater management program.
A water quality management program should be instituted to
ensure proper maintenance of the aquatic ecosy:3tems in The Wood-
lands. The first flush nutrients during storm events should be
captured or diverted and treated before release into the Con-
ference Center Lakes. Homeowners and golf cou::se managers
should be encouraged to select and apply fertilizers with cau-
tion, since excessive, long-term fertilization could result in
nutrient buildups in aquatic systems of The Woodlands.
The wet weather ponds and marshes in The Woodlands should be
managed since they often overflow into Panther Branch. Excessive
concentrations of algae and nutrients in these habitats would
ultimately affect the water quality of Panther Branch and its
downstream receiving waters, especially if Panther Branch water
is utilized to control water levels in preexisting and future
lakes.
Nutrient concentrations and detention times in the Confer-
ence Center Lakes should be controlled in order to prevent ex-.
cessive concentrations of algae. Treated sewage effluent and/or
well water could be used to periodically flush these lakes and
dilute algal nutrients.
The aquatic vascular plants should be managed in order to
prevent them from totally encompassing the Coniference Center
Lakes. This might be accomplished by increasing the slopes of
the littoral zones and/or by harvesting and reridval of the
plants.
8
-------
Future investigations on water resources for bacterial as-
sessment and management of stormwater runoff should establish
basef low _ diurnal variability in indicator bacteria concentra-
tions prior to evaluating storm events.
Further investigations should determine the effect of land
use on fecal coliform/fecal streptococci ratio.
•u-i Alternative methods must be developed to yield greater reli-
ability for bacterial indicator and pathogen quantification.
wif-Ti i Patterns of bacteria in impoundments
with relatively short detention times suggest that if disinfec-
tion is to be applied to a particular stormwater source, if at
aio J|aSlble' the equivalent of stilling basins should be em-
Examination of bacterial data for comparison with other
water quality parameters for analysis or resource management
purposes should not be accomplished using a simple arithmetic
approach because non-linear relationships exist between vari-
ables. Bacterial data should be transformed to Iog10 basis.
Before porous pavements are used for stored water release
to aquifers, the fate of lead salts in the percolated water
should be determined.
_ Studies are required for the design of porous pavement
parking lots in other than flat areas in order to avoid flood-
ing of low areas in a lot.
The water quality relationships developed in this study
should be verified with data from other watersheds across the
country to determine the possibility of data transfer. New or
revised coefficients may be developed for nonhomogeneous water-
sheds or different climatological regions. If linearization of
the log transforms of water quality data is found to be univer-
sally applicable, the prediction of water quality will be faci-
litated .
The use of the modified SWMM should be promoted so that
other users may utilize the vastly improved capabilities and
flexibility of the model. The use of baseflow recessions in de-
signing for low frequency or dry weather flows should be con-
sidered. Porous pavements should be evaluated, by means of the
modified SWMM, as urban runoff control facilities. The reduction
of data requirements for modeling natural streams should be em-
phasized and the use of the modified SWMM in modeling natural
drainage systems should be encouraged.
Water quantity and quality sampling at The Woodlands should
be continued at least until the watersheds have stablilized and
-------
erosion rates are reduced to pre-development or uniform levels.
Records of all construction activity should be maintained
so that the location and total area under construction during
a storm event will be known. This record will prove very helpful
in modeling water quality from that watershed.
The modeling or erosion in the SWMM needs to be refined.
The coefficients of the Universal Soil Loss Equation were de-
rived for agricultural areas and their applicability to urban
and forested areas is limited. Possibly, a new approach may
have to be considered. The significance of pollution from
construction activity has generally been underestimated.
10
-------
SECTION 4
SITE DESCRIPTIONS
Several sites in the Houston Metropolitan Area were chosen
to determine if stormwater runoff quality is dependent on land
use and development activities. The primary study site is The
Woodlands, Texas, a planned satellite city selected for a com-
prehensive investigation of runoff quality during all phases of
development. Two other watersheds were chosen to supplement data
collected from The Woodlands. Hunting Bayou is a developed
watershed with strong industrial influences and deteriorating
residential areas. Westbury Square is a middle class residential
area chosen because of the absence of construction in the water-
shed. .The locations of these study sites are shown in Figure 1.
Each watershed is comprehensively described in the following
text.
THE WOODLANDS
General Description
The Woodlands is a planned community being developed in
southern Montgomery County, Texas. The community is situated in
a heavily forested tract about 35 miles (56 km) north of Houston,
directly west of Interstate 45 (see Figure 1). The Woodlands en-
compasses 17,776 acres (7194 ha) and will be developed over a
twenty-year period beginning September, 1972. In contrast to a
residential subdivision, The Woodlands will contain all services
of a modern city, including facilities for social, recreational,
educational, commercial, institutional, business and industrial
pursuits. The community concept is committed to high standards
for environmental and lifestyle quality. The phased, long range
development places priority on ecological preservation and balance,
as well as social and habitational quality. This objective is to
be accomplished through a comprehensive environmental preservation
and management program, including planning and design controls.
The water resource system in The Woodlands, including its drain-
age system, is a good example of such planning and was the pri-
mary subject of this research.
Development Plan
The prime objective of The Woodlands is to provide the
finest urban environment in the Houston metropolitan area in
11
-------
INTERCONTINENTAL
AIRPORT
HUNTING I3AYOU
WESTSURY
SQUARE
Figure 1. Location of study sites.
(1 mi = 1.6 km)
12
-------
terms of physical setting and quality of human life and community
services. The basis for all aspects of development in The Wood-
lands is a unique ecological inventory conducted from 1971-1973
The geology, soils, drainage, water resources, weather, vegetation,
and wildlife endemic to The Woodlands were critically evaluated
for determining the most desirable location of roads, streets,
homes, offices, and other physical structures.
Land use allocation for The Woodlands provides for undevel-
oped areas. Residential areas will occupy 6,820 acres (2760 ha)
of The Woodlands site, while 1,699 acres (688 ha) are designated
for restricted industrial use. Additional area has been allocated
for retail, commerical, office, open space, and other land sales.
Approximately one-third (30.2%) of The Woodlands has been desig-
nated as open space. The majority of the' open space will be
located within the floodplain of Panther Branch and its ma-jor tri-
butary, Bear Branch.
Climate
The macroclimate of the Houston metropolitan area is domi-
nated by the Gulf of Mexico. Winters in the region are normally
mild, while summers are hot and humid. The daily winter tempera-
turl is
711_ Avfra9e yearly rainfall in The Woodlands totals about 46 in
(117 cm) and is evenly distributed throughout the year. April
May, November and December are usually the wettest months, while
March is the driest month. The quantity variation of rainfall for
specific storm events can be significant especially during summer
months. Most rainfall occurring during June, July, August, and
September is associated with thunderstorms. Precipitation during
these months is unpredictable and erratic. Frequenty an inch or
more (a few centimeters or more) of rainfall can be recorded in
one part of a watershed, while a very short distance away no pre-
cipitation occurs.
Plant Ecology and Wildlife
The _ tract of land upon which The Woodlands will be constructed
is comprised, almost completely, of forest. Eight major forest
ecosystems exist within the development. Some of these ecosystems,
including wet weather ponds, marshes, and grasslands, are extremely
unique and differentiated. The trees which are indigenous to the
PCS? c-ak
Because of the diversity of vegetation and varied food
oS/-?°St £f the £auna endemic to southern watersheds are found
the site. Common birds include blue jays, red-headed wood-
peckers, cardinals, Carolina chickadees, Carolina wrens and pink
warblers. Mammals frequently encountered are whitetail deerT
13
-------
raccoon, armadillo and ringtail cat. Numerous snakes and
other reptiles are also frequently observed on the site. The _
environmental impact of the development upon the wildlie and
vegetation has been estimated to be minimum (31-33) .
Soils
A comprehensive soil survey was performed at The Woodlands
development site by the USDA Soil Conservation Service, from
March 1972 to June 1973. The results of this survey are published
in a report entitled "Soil Survey of The Woodlands" (34). The
survey describes a total of 22 soil types in the area.
The soils on The Woodlands site characteristically contain a
zone of clay accumulation. The clay zone, which generally occurs
at a depth of from 18 in. (46 cm) to 6 ft (1.83 m) below the ground
surface, is relatively impermeable and creates a seasonal perched
water table. The high water table is an important 'factor in
maintaining the diverse vegetaion which occurs on the site. Of
major concern in the impact that development could have in re-
ducing recharge to this shallow reservoir due to soil compaction
and construction of impervious surfaces. If recharge is signi-
ficantly decreased, it could have a detrimental effect on the more
sensitive plant species on the site. Also, a reduced recharge
could have a significant effect on the base flow characteristics
of the streams draining The Woodlands site.
Existing Drainage
The natural drainage for The Woodlands community is shown in
Figure 2. Approximately 80% of the development is drained by
Panther Branch, a tributary of Spring Creek. The remaining portion
of the development drains directly into Spring Creek, which has a
total drainage area of 750 sq mi (1942.5 sq km). Because Panther
Branch and its major tributary, Bear Branch, represent the major
existing drainage for the development site, the hydrologic, morpho-
logic and transport characteristics of this stream are of extreme
importance.
Both Bear Branch and Panther Branch meander extensively along
well-defined, and low-flow channels, respectively, 9.0 (14.48 km)
and 14.6 mi (23.5 km) in length. Alluvial sediments, small
riffles, and slow moving pools are commonplace within Panther
Branch and Bear Branch. The width of the low-flow channel is highly
variable but is normally between 5 and 20 ft (1.5 and 6.1 m). When
the capacity of the defined channel is exceeded, storm runoff dis-
charges into a very broad, flat flood plain, presently covered
with heavy brush. Flood runoff is characterized by low velocities
and shallow depth because (a) a large land area is inundated,
(b) flow resistance is high, and (c) hydraulic slope is low.
Excluding those areas presently under construction, essentially no
evidence of any serious erosion can be found anywhere in Panther
Branch watershed.
14
-------
PANTHER BRANCH
DRAINAGE AREA
OUTLINE
SCALE IN MltES
SAWDUST ROAD
Figure 2. Existing drainage network for The Woodlands (numbers
shown indicate elevation above mean sea level).
1 mi = .1.6 km
15
-------
Flow rate measurements in Panther Branch indicate that water
velocities as low as 0.04 to 0.10 ft/sec (.012 to .03 m/s) are
typical for low-flow conditions. The slow travel rate is a
result of low channel slopes present throughout the drainage net-
work. The total change in channel elevation across the drainage
basin is about 120 ft (36.5m m), with an average rate of change
of 5 to 7 ft/mi (.95 to 1.33 m/km).
The impact of low channel and land slopes within The Wood-
lands is reflected in an extremely low stormwater runoff coeffi-
cient. USGS data for the 1973 and 1974 water years show that
only 23% of total rainfall resulted in surface runoff. The re-
maining 77% either evaporated, transpired or infiltrated into the
ground. It should be noted that rainfall was heavy during the 1973
and 1974 water years, respectively 77 in. (195.(i cm) and 51 in.
(129.5 cm). It is estimated that only 10 to 15% of rainfall will
run off during a year of average rainfall, 45 in. (114.3 cm) (35).
Runoff from the Panther Branch watershed is not evenly dis-
tributed throughout the year. During the summer months of May
through September little discharge occurs, exce:pt immediately
following an intense and prolonged rainfall. The average daily
low flow discharge at Sawdust Road, including summer months, is
1 to 2 cfs (.028 to .056 m3/sec). An average daily discharge of
100 cfs (2.8 m3/sec) at this site is exceeded only 5% of the time.
WATER RESOURCE SYSTEM OF THE WOODLANDS
The annual rainfall at The Woodlands is partitioned as run-
off into existing lakes and streams and infiltration into the
ground. Losses result from evapotranspiration, evaporation and
subsurface transport to streams (Figure 3).
The maintenance of a satisfactory groundwa.ter reservoir
above the perched water table is critical for the continued
growth of vegetation. Any drainage system for The Woodlands must
consider the detrimental consequences of disrupting the movement
of water within this shallow aquifer. Deeper a.quifers (1800 ft)
(549 m) are used for community water supply.
A series of wet weather ponds and variable volume lakes will
serve as recreational centers, wildlife preserves and, more im-
portantaly, storage for stormwater runoff. This system of water
reservoirs contributes to the maintenance of an adequate perched
water table for plant life. Lake water will be lost primarily
through surface evaporation and irrigation. Inflow to the lake
system will result primarily from stormwater runoff and reclaimed
water water from 2 or 3 sewage treatment plants. The projected
volume of waste water is 20 mgd (0.88 m^/sec). If necessary,
treated sewage effluents can be discharged directly into Panther
Branch.
16
-------
TOTAL
9 USD
CONSUMPTIVE USES
ft
WATER DEMAND OF
THE WOODLANDS
J20M6DJ
TRANSPORT AND
TREATMENT OF
SEWAGE (15-20 MGO)
EVAPORATION
IRRIGATION
SURFACE RUNOFF
PANTHER BRANCH
OR
SPRING. CREEK
Figure 3. Schematic water balance for The Woodlands.
17
-------
"Natural prainacre System"
The Woodlands Development Corporation has specified that tne
basic drainage system for their new community will utilize natu-
ral drainage" concepts. These design principles have been re-
ported in "Natural Drainage Systems: An Alternative to Conven-
tional Drainage Systems" by Winslow, Veltman and Espey (36) and
will not be described in detail herein. Rather, a brief summary
is presented on the "natural drainage" concept and reasons for
its selection in The Woodlands.
The normal procedure for disposing of stormwater runoff
within the Houston Metropolitan Area is to enlarge the natural
drainageways by deepening and widening existing stream channels
and providing supplementary lateral drains. In the City of
ISuston, thiS approach generally results in stormwater sewers for
the lateral drainage and deep, wide, concrete-lined ditches for
the major drainage. This solution to stormwater disposal, al-
though widely used and approved by the City of Houston, was in-
compatible with one of thS major criteria used in developing The
Woodlands— preserving and enhancing the natural environment.
"Natural drainage" cbncepts adopted by WDC are envisioned as a
method of providing ade
-------
stormwater runoff and serve as a sedimentation basin for the 337
acre (136.4 ha) watershed under construction at this time. Upon
completion it will include an 18 hole golf course meandering
through a residential area.
Lake A is a variable volume lake to be used for non-contact
recreation. Lake discharge is controlled by an outlet box at an
elevation of 121.8 ft (37.1 m) above sea level. Water from Lake A
irrigates The Woodlands Gold Course bordering the lake's eastern
and northern shores. In dry weather, the water level drops due
to evaporation and groundwater is pumped into the lake to com-
pensate. Lake water is not lost to groundwater recharge due to
a clay bottom which serves as an effective seal.
Porous Pavement
An experimental porous pavement parking lot has been con-
structed at The Woodlands as a possible solution to the problem
of excess runoff from impervious urban surfaces. The lot con-
sists of a permeable asphalt-concrete topping, overlaying a coarse
base and fill material. The pavement is designed to allow rain-
water to percolate through the asphalt and infiltrate into an ex-
isting groundwater reservoir. Besides reducing urban stormwater
runoff volumes, porous pavement has other benefits such as anti-
skid properties and better visibility of road markings.
HUNTING BAYOU WATERSHED
The Hunting Bayou watershed is located in northeast Houston
near the intersection of Highways 50 and 610. The 1,976 acre (800 ha)
watershed is characterized by low land slopes and impermeable
soils with high clay content. Primary drainage channels are
trapezoidal in shape and lined with vegetation which varies in
density from moderate to very heavy depending upon the season
and maintenance schedules. The majority of the secondary drain-
age is provided by roadside, grass-lined swales comparable to
the drainage design at The Woodlands. A fourth of the area is
drained by storm sewers. There are no known effluents entering
the drainage system, however, point sources are probable due to
the age of the residential districts (illicit sanitary sewer
connections to the storm system and the presence of industrial
influences). The area is poorly maintained and stream channels
are sometimes used as dumping areas for waste materials such as
oil and grease, old tires, and other refuse.
Land use in the watershed is mixed (Figure 4) with resi-
dential areas comprising the largest segment. Residences are
mostly single family dwellings of low value. Table 1 gives demo-
graphic information regarding the indigent population.
19
-------
Industrial activity in the watershed includes meat packing
and rendering plants, wrecking yards and mechanical contractors.
Various commercial establishments in the watershed support the
residential population. Construction in the watershed is cen-
tered around completion of Interstate Highway 610.
WESTBURY SQUARE WATERSHED
The "natural drainage" system utilized by The Woodlands is
an innovative method for controlling stormwater runoff and
preventing water quality deterioration. A Houston watershed
with a conventional drainage system was selected as a compara-
tive study site. The residential land use of Westbury is simi-
lar to that being constructed at The Woodlands.
The 210 acre (85 ha) watershed is located in Southwest
Houston and is comprised exclusively of single-family residen-
tial dwellings. No commercial or industrial influences are
present in this area. The watershed is completely developed and
contains no construction sites, empty lots, and no undeveloped
land. Figure 5 and Table 2 provide watershed information.
Demographic information from the 1970 census presents the
area as upper-middle class, median annual income of $19,000.
Population density is approximately 11 persons/acre (27 persons/ha),
or 4 persons/household.
The separate stormwater drainage system consists of lateral
drainage provided by concrete pipe, 18 in. to 54 in. (45.72 cm to
137.16 cm) diameter, connecting with a main collecting channel at
roadway intersections. The channel is a 10 ft (3 m) deep open
grass-lined ditch, often choked with vegetation in the summer
months. The ditch passes through culverts beneath roadways, and
at these points, ponding occurs upstream and downstream providing
slight storage of runoff and a reduction of flow velocity. Low
runoff coefficients can be expected due to the low land slope of
only 0.8%. Impervious cover is estimated at 35.4%. No dry weather
flow is present in this watershed.
Soils in the area are predominately dark clays and loams,
characterized by low permeability and high available water
capacity. The soils tend to be mildly alkaline or neutral.
Rainfall in the area averages 39.5 in/yr (1.0 m/yr).
20
-------
0)
CO
o
to
B
O
10
to
§
to
H
o
a
in
o
to
4J
U
a
n
9
to
e
ni m
s
i
•S
c
o
to
H
&
21
-------
TABLE 2. WESTBURY SQUARE WATERSHED
CHARACTERISTICS -
Total Drainage Area
Single Family Residential Areas
Impervious Cover
•Population Density
Family Median Income
Median School Years Completed
Home Values - Mean
Persons/Household
Rooms/Housing Unit
210 acres (SISha)
10054
35.4*5
7,040 persons/mile2 (2718 pers/km2)
$19,000/year
14.7 years
$29,000.00
3.8 persons
7.5 Rooms
* Data derived from 1970 Census: Census tracts 427 and
428 are averaged.
N
1 in. = 2.54
cm
Figure 5. Westbury Square watershed
(scale 1 in. = 1000 ft) .
22
-------
SECTION 5
SAMPLING AND MONITORING PROGRAMS
This section summarizes the field sampling and monitoring
programs initiated in September 1973. The sampling programs were
divided into two distinct segments. First, selected terrestrial
and aquatic environments within The Woodlands were sampled during
periods of no overland runoff (low-flow conditions) for the oc-
currence and concentration of selected physical, chemical and
biological constituents. These measurements comprise a low-flow
data bank from which (a) the effect of urbanization on the water
resources of The Woodlands can be quantified, and (b) the water
quality criteria of the lakes within The Woodlands can be estab-
lished to insure their usage for recreational and aesthetic pur-
poses. Sampling points were located for the comparison of un-
developed and developing areas within The Woodlands. Sampling
sites downstream in the drainage area of Panther Branch provide
data to establish the impact of The Woodlands development upon
the water quality of the receiving body (Spring creek).
The second phase of the sampling program concerned itself
with quantifying the chemical, physical, hydrological and bio-
logical characteristics of overland runoff and its subsequent
impact on the water resources of The Woodlands. The stormwater
sampling program involved the development of pollutographs for at
least 25 different water quality parameters per storm event. Not
every storm event was evaluated but, rather, selected storms were
monitored to define various hydrolog'ic and seasonal conditions.
A comprehensive hydrologic network was established within
the study area, focused around continuous discharge recording
stations operated and maintained by the Water Resources Division,
U. S. Geological Survey, Houston, Texas at the request of The
Woodlands Development Corporation. The purpose of this network
was to accurately delineate the movement of water, especially
surface flows, within the new community. For this purpose, a
weather station, rain gauges, streamflow stations and groundwater
observation wells were established.
DRY WEATHER PROGRAM
Dry weather field sampling programs were conducted within
the Panther Branch watershed, in the immediate vicinity of The
Woodlands development. These programs were directed toward
23
-------
gathering a basic data bank on the algal, bactezial, chemical,
hydrological and physical characteristics of thi.s area's surface
water resources. Water, soil, plant and animal samples were
collected on a regular basis for analysis.
The locations of aquatic sampling sites are*, shown in Figure
6 Stream water samples were collected at sites located from the
headwaters of Bear Branch and panther Branch downstream to Spring
Creek. Several ponds and marshes within the watershed were also
sampled. Sampling frequency was about once per month.
Specific analysis performed on water samplers included
temperature, dissolved oxygen (DO), pH, turbidity, total sus-
pended solids (TSS), soluble COD, total COD, soluble organic
carbon (SOC), total Kjeldahl nitrogen (TKN), total phosphorous
(TP), orthophosphate (ortho P), ammonia (NH3), nitrite (NO2),
nitrate (NO3), BOD, and specific conductance. The presence of
specific pesticides and chlorinated hydrocarbons; was also deter-
mined In addition, the following bacterial emimerations were
conducted: fecal coliforra, fecal streptococcus, total coliform,
total bacteria, Salmonella-Shigella sp., Pseudononas sp., and
Staphylococcus sp. Algae and macrophyte enumerations were also
conducted.
Other water samples were collected for experimental purposes
to determine limiting nutrients for algal growth and bacterial
disinfection characteristics.
Soil samples were collected from 21 sites located in the
study area and analyzed for bacterial and algal content.
Leachates from these samples were tested for pesticides and
nutrients.
Plant and animal samples were collected on a regular basis
and examined for the presence of pesticides and chlorinated
hydrocarbons. Aquatic animals collected include fish, shrimp
and crayfish.
STORM EVENT PROGRAM
The stormwater monitoring program used a hydrologic network
combined with intensive sampling to characterize runoff quality.
Runoff samples were collected at six sites foi:r _ locate^ within
The Woodlands, a fifth in Hunting Bayou, and a sixth at Westbury.
Characteristics of each watershed are summarized in Table 3.
The Woodlands
The Woodlands sampling locations and their designations are
listed below:
1. panther Branch at the Confluence
24
-------
e cj
B)
3
id
tn
.s-s
(d rC
M CO
M
H3 0)
fi 4J
(d (d
S
D]
5-8
•H CJ
,10
-------
^
£•
5
to
tr-
C 3
••H O
•W >
C IQ
s «
to
13
c a
10
rH 0)
T) AS
0 It
O r3
tn
3
3
rH 0)
s-s
3 V-
£
tn
3
3 o
•O 1
0 Oi
0
S
to
it) O
rH CO
•3 1
O ft
2
if
..
a a
C 4J
n) -H
CO
•o
01 O
£ C
|H~
0) C
-P E
m n
SJS
0
rH
CM
vo
r-
cn
rH
r-
ro
ro
f
a
T
O
in
0
vc
i—
VD
0
vo
^
IS
01
rl
1
a) 01
O^ 1
ffl 1
c
•H
10
a
m
ro
dP
rH
CM
dP
rH
•
t*.
dP
ro
H
dP
rH
V
rH
to
3
0
•rl
> a
IH 01
01 >-
H
dP
0
0
rH
do
CO
CM
O
o
o
o
E
0-0
4J 01
tn IH
01
10 s
o o
< w
1
CO
o
dP
rH
•
dP
ro
•
dP
fO
•
dP
O
CM
•
dP
vo
01
a
o
rH
W
•O
C
(C
rH
10
iH
-P dP
C 0
0) O
•rl
to
01
c:
CO CM ^1* VO
^T CO rH
II II II II
• • • •
C O IH rH
01 IH 4^ 01
•a 01 to >
•H e 3 (D
in E 'O 'O
ai o c c
K CJ w D
-dP •
C rH
1 O rH ro
• -H « e
01 4-1 "H O
rH O .P -H
.Q D C 4->
IB )H O> 10
•rl J-l >O 01
IH 10-rl rl
to c tn o
> o 01 ai
0 tS K
^ ,
C «•« H
1 O rH <0
•rl 10 rH C
fl) J-> -rl IIS 0
rH O -P -rl -H
J3 3 C O -M
10 )H 01 IH 10
•H 4J "O 01 01
IH 0) -H E IH
m c to E u
> O 01 O 01
CJ OJ CJ K
m
01
.D dp dP
(ij O> H3 rH
on a)
rH &
10 II 0 II
M rH
4-1 C > C
« 10 o m
2 rH Q rH
10 1
01 O
IH • 4J 3 •
O dP C IH dP
CH o CD 4J o
en g tn rH
rH a c
10 II O O II
ft rH O
3 10 01 e
4J C > tl O
m 10 ai c -H
•3 rH Q 10 4J
C
o
•H
01 4-J
(0 10
O O
•H
•O IH
C -rl
10 tn
>J tn
10
rH
D
to 01
tn H IH
B C rH O
rH O 10
U -rl II X.
JJ >itO
0) ro 4-> c
rH . rH -H O
"O 3 10 01
•a a. c IH
•H O 0) 01
E a. D a
a)
CNJ M
| fl rH 0
Q • O II CO
O -H " ~>-
c -u >> to
,H Ctf 4J fl
H -H O
IH 3 to to
g P. d rJ
g 0 D. in •
0) a m tn
•O S rH 0)
o o
•a « c
01 -4-1 01 01
C C rH *O
C 01 *O *rl
ic E 'O tn
rH CU-H 01
ft- 0 E M
1
rH IH
0) 01 0) <«
> a w
ai n. « to e
•O 3 rH 01 O
• O O -rl
•a •• n JJ
01 4J 01 01 B
c c rH -a 01
n 01 "o -H M
M O
Q
1
Q
in
o
II
o
26
-------
with Bear Branch
2. Panther Branch at Sawdust Road
3. Outflow of Lake Harrison
4. Inflow of Lake Harrison
P-10
P-30
Lake A
Lake B
A U.S.G.S. gauging station measured streamflow at each of these
sites .
Station P-10 is located on Panther Branch 200yds (182.9m) down-
stream of the confluence with Bear Branch (see Figure 6) The
watershed at this point measures 16,050 acres (6,495 ha) of pre-
dominantly undeveloped pine-oak forest. Data collected from P-10
represents runoff from a natural area devoid of urban influences
and, when compared to other sites, serves to determine the ef-
rS Of4.urba2, inf luences on r^noff quality. U.S.G.S. establish-
ed the streamflow gauging station at p-10 in July 1974.
4,-u P7?0.is located downstream on Panther Branch and includes
the P-10 drainage area in its 21,606 acre (8,744 ha) watershed
(Figure 6). The sampling site is in an advantageous position for
monitoring runoff quality from development areas immediately up-
™^?™ U.S.G.S. has operated a streamflow gauging and monthl?
sampling station at this site since April 1972. Because con-
struct ion at The Woodlands was a continuing process, water cmal-
ity changes were expected. H
Two stormwater sampling stations, shown in Figure 7 were
located at Lake Harrison to measure stormwater inflow and out-
flow. The upper station, situated in a swale at the head of
Lake B, sampled runoff from the major source of stormwater flow
into the lake system. During dry weather, no flow occurs in the
swale. The 337 acre (136 ha) watershed was under extensive con-
struction at the time of study. The Lake A gauging station
adjacent to the lake outflow box, was sampled to assess effects
of detention on runoff quality. The drainage area at Lake Harri-
son outflow is 483 acres (196 ha) .
Rainfall data was available at five locations in, or near
The Woodlands (see Figure 6) . The precipitation data collected
at these sites determined the average hyetograph for the Panther
Branch watershed during each storm event and defined the in-
tensity, duration, time period and total amount of rainfall
The rainfall network provided data on the spatial and temporal
characteristics of rainfall within the study basin.
• ^ samPles were analyzed for the same parameters listed
in the dry weather flow program. Chemical data were used in con
junct ion with flow data to derive mass flow relationships for
each constituent during each storm.
27
-------
4.1 ac (1.66 ha)
20 ac ft (2.47 ha-m)
Normal Water Level
el. 128 (39 m)
Average Depth
6 ft (1.83 m)
AREA SHOWN IN DETAIL
j.2.5 ac (54 ha)
90 ac ft (11.1 ha-m)
Normal Water Level
el. 122 (37.2 m)
Average Depth
8 ft (2.438 m)
• LOCATIONS OF U.S.G.S. GAUGING STATIONS
O 400 800 feet
SCALE
1 ft = .305 m
1 ac = .405 ha
Figure 7,
The Woodlands man-made lake system with
locations of stormwater monitoring sites.
28
-------
Hunting Bayou
The stormwater sampling program at Hunting Bayou was similar
to The Woodlands, except that only one sampling site was monitor-
ed. Stormwater samples were collected from the U.S.G.S. Hunting
Bayou at Falls Street gauging station (see Figure 4). precipita-
tion data were available from two recorders located south of the
drainage basin.
Westbury
Runoff from this watershed was sampled above the Atwell'
Street bridge crossing the primary drainage channel (see Figure
5). Flow measurements were determined using a pygmy f f ow meter, and
precipitation measurements were made using a portable volumetric
gauge located at the sampling site.
Porous Pavement
The porous pavement parking lot at The Woodlands Conference
Leisure and Commercial Center was finished October 1974. The
pavement consists of a porous asphalt-concrete topping overlaying a
gravel reservoir. Porous topping is produced from an asphalt bound
aggregate deficient in fine sizes. In comparison to conventional
toppings, increased amounts of asphalt are required to compensate
for loss of strength resulting from the lower contact area between
discrete particles of aggregate.
Soil permeability is poor at The Woodlands, so to simulate
a reasonably permeable subsoil, the site of the porous pavement
lot was excavated to a depth of 4 ft (1.2 m) and a permeable
soil/sand mixture was filled into the excavation to a depth of
33.5 in. (85 cm). Open aggregate base course, 12 in. (30.5 cm)
in depth was placed over the simulated subbase and 2.5 in. (6.6 cm)
of porous topping was put down. A French drain was constructed
to drain water from the soil/sand subbase.
The site plan for the experimental parking lot is shown in
Figure 8. Stormwater samples were collected from the_water depth
wells in both conventional and porous lots. The quality of
stormwater percolating through porous pavement was assessed and
compared to conventional pavement runoff for the following chemi-
cal parameters: TP, orthophosphate, NH3, N02, TKN, pH, conduc-
tance, soluble COD, lead, zinc, and total organic carbon (TOC).
29
-------
o
o
CD
-p
o
tn
fl
•r-l
-p
0)
en
3
O
i-l
o
a;
^1
4J
r-l
ft
0)
4->
•i-l
Cfl
00
0)
tn
-•-I
30
-------
SECTION 6
RESULTS AMD DISCUSSION
DATA SUMMARY
Dry Weather Monitoring
within The Woodlands was determined
m -K Periods to establish a baseline water qual-
. The baselme will be useful for determining (a) the effect
of urbanization on the water resources of The Woodlands over the
^no4° Yea^' ^ ^b) inmediate comparisons betweeSsLJmwater
Ji?ef ar^aindicaSfd ?10W ^ ^ity. Woodlands sampling
.sites are indicated in Figure 6. Dry weather, or low flow re-
fers to time periods when the stream stage is essentially COn-
f?ti • ?hf t1?? Peri°d foll™i*9 * storm event rJquiSdto
establish low-flow conditions depends on factors s2ch as antece-
dent moisture conditions, time since last storm, rainfall Surl-
tion and intensity, and groundwater elevation. in the Panther
'^1'™
Low flow water quality data for Panther Branch, Bear Branch
and Spring Creek are presented in Table 4. The headwaters in
the stream system are low in inorganic nutrients but signifi-
cant contributions from developing areas increase concentrations
below P-10. The primary nutrient input is from the golf course
^e^la^L?PStream °? P~30' Or9anic concentrations are high
nnm ?7 COD) consisting of relatively non-biodegradable (2 mg/1
BOD) leachate from decaying vegetation in the forest. COD dilu-
tion occurred downstream and the lowest concentrations were
observed in Spring Creek. TSS changed drastically as the
stream passed through developing areas where construction activity
af hiShr?W ?Jnn W^? located in the floodplain. Low flow TSS Y
as high as 1600 mg/1 were observed at P-30.
Storm Events
aSneA hydrolo*i<»l. Physical, and chemical
aspects of 43 distinct runoff events resulted from 17 selected
one fS L^ri? d^Wlth streamflow being sampled simultaneously at
v2rsi?v ?rL^ £he '!l°nitorin? stations established by Rice Uni-
versity (refer to Site Description, Section 4). The number of
31
-------
%
§
w
ta
E
a
3?
p £
^-p
OJ g
E-) ""
gj w
_ O
£H
/*** r •
VJ tj
5
I EH
fe fe;
O 13
^ ^
fa O
O 0
JH W
PJ O
rii yd!
s 2
H EJ
co »aj
.
<*
TI^I
fft
EH
C 0)
•H CD
M M
CO
tr>
C
0-H CD
CO rH 4J
cL E cn
S
(0
CO
o
• rH
H 1 0
CQ 04 CO
M C A<
CO Q)
•4J & C
C -P (0
03 0)
A< XI
O^
C
O -H CD
rH rH -P
1 Qi-rH
I > '_! *t
w € CO
m
CO
ssiio
HJ C rH
Pw n) i
M Pi
CD
M M >
RJ CD O
« -P «0
CM CO
CO CM rH Cn
CM l-H t- rH
0 0 0 O
in cn vo co
CO Kt* VO <^>
H O rH CM
O O rH 0
CO TJ«
o in co
rH 0 0 rH
O O rH O
CM
"•S" cn CM co
vo CM co cn
o o cn o
o o o o
r- in vo
vo CM in rH
0 O CO rH
o o o o
a*
o
ft H « JC
fc* O EH 2
in
rH
O
0
in
o
o
0
vo
o
o
o
^3*
in
0
o
0
vo
o
0
o
CM
O
2
cn
in
CO CO CO
O 00 O t^
CO JO ^«
rH
cn
m
o r* CM
0 0 rH o,
oo •«• o
cn •
CM
0 0 O
O CO VO CO
oo «a« m
*3*
*3*
CM
O VO f-
O CM rf VO
CM rH »*
VO
CM in
O^*O
O VO CM °
rH H m
^
JJ
>t -d
•«-» o 5
•H -rl "Jj
t3 m o
•rl -H 3
,Q U 'O
CO CO M Q) C
O co 3 fx O
2 B EH CO U
CM *»
cn vo
rH CM
in o
in rH
rH in
0 O
cn CM
rH in
cn o
cn o
rH in
O CO
CO ^J*
CM in
Q
O
O
rH
a
0 -P
o o
CO H
CO rji f- p- CM O
cn r-i vo vo en vo
rH H rH
rH
en r*- vo
^T H CO CM VO rH
!*» cs vo "«r co in
CO rH
O CO VO CO O O
O rH VD ^ O CM
^r CM
irt ,00 co *»
vo t- CM CM cn o
O rH VD in t*- CO
^T rH
vo
P* O *3< in O *3«
o CM VD in cn rH
*J* rH
o
U CD
Q) M
rH flj
•O • J3
3 Qi O
rH Q & CO
O O B O CO -H
W cq O* C B Q
•
"•T
EH °«.
FJ a)
^5
-P a
•a CD
•rl CU
Q S
M CD
f\
.-
•* CO
CQ -P
o'c
D
%
^n Cu
ft?
•SB,
CO
t _
•M •••
Q 6
o
\
•• to
•P O
g'-g
0 0
QJ O
•H
\s
0^ .
S ^i
-p
C-rl
•1*1
CO 4J
4J 0
C 3
CD «O
CD O
CO U
(fl -rl
CD
-------
runoff events monitored at each sampling site were as follows
Sampling Site Number of Runoff Events
P-30
P-10
Lake' A
Lake B
Hunting Bayou
Westbury Square
12
8
8
8
5
2
Hydrological Observations—
A summary of the hydrological data is presented in Table 5.
Hydrological parameters are specifically defined in Table 6.
Note that Total Streamflow is the sum of Baseflow and Runoff.
The number of storm events monitored was limited by two drought
periods, each of six month duration during the project.
Rainwater Quality—
Previous investigations indicate air pollution may contri-
bute to surface water pollution through rainfall and/or dry fall-
out, even to the extent of pollutants traveling via air from in-
dustrial and agricultural regions to be deposited in undeveloped
areas (37, 38). Samples collected in the Houston area and at
The Woodlands assessed relative contributions of rainwater qual-
ity to stream pollution. Results presented in Table 7 indicate
a substantial nutrient and COD content in rainwater at both sites.
A statistically significant increase exists between Houston and
The Woodlands rainwater in regard to NH3 and NO3 content. The
rainwater data are compared to stormwater data in Table 8. A
study of air quality at The Woodlands indicated high levels of
hydrocarbons, 7.6 ppm non-methane hydrocarbons, whose source was
attributable to vegetative emissions. These ambient air hydro-
carbons may contribute to the soluble COD in rainwater. The study
also found an absence of NOX at The Woodlands in contrast to
serious NOX air pollution problems in the Houston urban area (39).
At The Woodlands, rainwater nutrient concentrations were
greater than runoff water, while the opposite relationship pre-
valied in the urban watershed. Experiments were conducted to
determine the capacity of soils for stormwater nutrient removal.
Four samples of soil from various locations in The Woodlands
(see below) were dried and weighed. The samples were extracted
with demineralized water until no further NH3 was measured in the
extract, and then equilibrated with 30 ml portions of 1 mg N/l
(ammonium sulfate). After centrifugation, the supernatant was
analyzed for NH3 and then discarded. This was repeated until no
further adsorption was measured. Findings indicate low levels
of NH3 are definitely adsorbed by soils, with the greatest
33
-------
S
1
I
g
i
STORM
W
&
if
01
0 A
Ǥs
4J
a
0
0^
C 01
S3
a?
o o)
(4 it
U]
Eu
X 3
Q C
IT
to b
id K
m
U]
U
01 S
in o
OH
lid
i
rH
o in
in en
r-
o vo
in o
CM CM
in in
rH t-
o o
Hunting Bayou
Hunting Bayou
3/20/74
3/26/74
CM CO
CO
CO
O
O
cn
5
in
o
rH
cn
VD
o
•*r
rH
•a1
CM
rH
in
CO
O
Hunting Bayou
rH
rH
\
»
r~ co
co in
CM H
00 CO
rH 0
O 0
r~ o
in
rH
cn
CO •«*
in co
CN
en
r~ o
CO
CO
O
o co
CM VO
0 rH
CO 0
in r*
CO
cn
in vo
O CO
Woodlands P-30
Woodlands P-30
4/22/74
10/28/74
in vo
en rH
vo r~
^r r-
CM CM
0 O
0 0
CO O
O O
CO CM
CO CM
CM CO
rH
O O
CO CO
CO CM
cn H
CM rH
in H
T CM
o o
r--
CT1 CTI CM rH
rH 1—
rH
in cn rH rH
r- vo co co
o o o o
Woodlands P-30
P-10
Lake A
Lake B
3/13/75
cn
in m CM CM
CM CM CM CM
0 0 O O
O O O O
co in o o
vo cn co <*
in i" in co
o o •«• CM
VD O CO CO
CM H cn cn
CO VO
CM rH
o o -a- co
O r-- rH CM
H rH rH rH
H H
cn o
CM ^f
00
^3* O O O
0 rH O 0
cn
O OCO CM
cn ^j* co co
CM rH cn cn
CO VO
CN rH
vo co r- r^
r~ «a- cn cn
CM CM CO CO
Woodlands P-30
P-10
Lake A
Lake B
4/08/75
o
H
rH
•a-
H
o
r~ o
H in
vo
CO H
co r*-
CM
CM VD
r» co
CO
rH
in
vo o
•» 0
VD
cn rH
co r-
CN
rH in
co r-*
0 0
Hunting Bayou
Westbury
5/08/75
H
rH
•O
inue
4J
C)
B
O
S
m
a
a)
CO
ro
S
CO
CM
in o o o
• rH CO
CM • •
I! II II
II
co
pi O W M-l
•H cd m o
34
-------
•o
(I)
s
•H C
Q* *
•M 4-J
c c
0) O
1*
ox
0) 0)
•P Q)
<
r-
ro o>
CTl rH
H
CN ro
in ro co co
en CN r-« [—
rH CN
r~- co co ro
m m co
CO 00 Cn
CO CO O
CO
en
o
rH
«3* in fN CN
rHrH rHH
CO rH CO CO
in -ro r-- r~*
r** i*- o o
rH rH CM CN
Cn O^ CO CO
rH rH rH rH
co ro in in
O CO CN CN
CN rH ro ro
m in co co
o
• II .
cxi
4J ||
II MH
I CO
C O 14H
•H to a
35
-------
HYDROLOGICAL DEFINITIONS AND CALCULATIONS
Storm Event:
Total Rainfall;
Duration of
Rainfall;
Rainfall Intensity;
Type of Storm
Event:
Antecedent
Rainfall
Condition;
Total Discharge
or Flow;
Runoff;
Peak Flow;
Base Flow or
Groundwater Flow;
Time of Passage
Discrete period of rainfall producing
runoff monitored during the study.
Calculated via Thiessen method using total
precipitation data recorded over 2 separate
drainage areas, P-10 and P-30. The co-
efficients are calculated using these rain
gauges: Porous Pavement, Egypt, Confluence
and W. G. Jones.
P-30 Drainage Area
Porous P. - .125
Confluence- .328
Jones - .075
Egypt - .472
p—10 Drainage Area
Confluence- .258
Jones - .100
Egypt - .642
Time of
Concentrat ion;
Shortest time period during which 85% of
total precipitation occurred.
Total rainfall divided by duration.
Described type of rainfall - e.g. one
period of rainfall, 2 periods of rainfall,
etc.
Total precipitation in 1 week period prior
to storm event. Calculate same way as
Total Rainfall (see above).
Volume of storm event hydrograph.
4-
Overland flow volume. Equal to total dis-
charge minus base flow.
Maximum stream discharge.
Calculated by multiplying the instantaneous
discharge prior to the storm event by the
time of passage.
Time from first rise in stream to 0.1 of
the peak flow. If the hydrograph does not
return to this level due to successive
rainfall a hydrograph separation technique
is used to extend the recession limb to 0.1
of the peak. When runoff is minimal and
base flow greater than 0.1 of peak flow,
then the inflection point on the hydrograph
tail is assumed the end of passage.
Time from first rise in stream to peak
flow.
36
-------
4J
j;
(1)
r
4!
•r
X
W
c
r
3
^
(D
,_j
42
p
o
to
^*
o
cu
o
.d
M
o
CO
o
•a
*4*
tti
3
fl)
jJJ
CO
01
1 X
c
0)
t*
fi
01
IK
c
01
1 X
rH
Q> O
VO CM
t^
r» o
• •
co r-
rH . TJ<
« to •
rj< .in
rH S H
CM Q\
in «H
^j. VO
rH UTi
0 0
CM • a\
rH CO CO
o • o
53
rH O
rH CM
rH
vo r-
in rH
i •
0 0
in
o
CM " rH
in v co
• *
C3 O-i ^^
in en
00 rH
CM er»
rH O
^3 *5j* CD
m
o
*
rH CM
CO v CM
' o a o
CO
01
CM tJ
ss. g
O "8
01 - t3
3 0
O 0
(U
o
c -o
a) d
. • O «J
01 m -H
3) r^ iw H
rH \ iH IS
CUCM C O
grH &> -H
ta\ -H 01
01 CO 01 O
rH
l| | ^ t| | Q
O "I O -H
M\VO H m
O co r*^ c) 0)
g ^s.0 Q) -4-*
3 r* CM rH
G \ O
••CM in -P
II in »o
r- in • w
C\r^ C
rH\ 11 O
*r-t 0\ -H
c \rH cm -p
O r*» "v. >«o o a
i2 jj
01 'O nJ X<
TJ H C g -P
^ 0) T« -H -H
flj >rH X >
•O *H n3 O
C fi O ^ 01
<8
rH
^
•
•
52
*
02
IM
m
H
rH
O
•d
o
0
5
Q)
rH
en
fi
w
v
d
c
H
»,
H
rH
W C O
rH 01 01 rH rH 4J O C -H
04 04* COO
C! g g -P CQ
n^ RJ (0 -P II CP
fl) 01 01 C C 43
g
-------
TABLE 8. COMPARISON OF RAINWATER AND RUNOFF QUALITY
IN HOUSTON AND AT THE WOODLANDS
(ALL UNITS lN..mg/l)
Constituent
soluble COD
NH3
•N03
ortho P
Houston
Rain
14
.31
.52
.01
Runoff1
20
2.1
.38
.57
The Woodlands
Rain
15
.22
.31
.04
Runoff
44
.08
.05
.005
1 Avg. of stormwater data collected at Hunting Bayou
2 Avg. of stonrtwater data collected at p-10
adsorption occurring in undisturbed forest soil:
Soil Sample Total mg N/g Adsorbed
Golf Course
Roadside
Swale
Woods
0.027
0.022
0.017
0.043
1 mg N/g dry soil adsorbed in equilibrium with
1 mg N/l solution of ammonium sulfate.
The data suggest that ammonia nitrogen in rainwater is ad-
sorbed to a large extent in undisturbed forest soils where it is
metabolized by plants or nitrified and possibly denitrified by
soil microorganisms. In an urban area receiving the same am-
monia rainwater load, there is more impervious area resulting in
higher ammonia in the runoff. In addition, soils in an urban
area have a lower NH3 adsorption capacity and lower NH3 utilization
rate.
Of greatest significance is the effect of nutrient wash off
on lake eutrophication at The Woodlands. Ward and King (40)
have shown nitrogen to be the limiting nutrient for algal growth
in The Woodlands lakes during wet weather. As development con-
tinues, increasing nitrogen input from stormwater runoff can be
expected and lake enrichment may result.
38
-------
Groundwater Quality—
Surface water quality at The Woodlands is influenced by in-
filtration from the perched water table. To determine ground-
water quality, samples were collected from 17 wells located with-
in the watershed 3 days after a 2.8 in. (7.1 cm) rainfall. The
wells consist of 2 in. (5 cm) I.D. PVC pipe driven 2 to 5 ft,(0.6
to 1.5 m) into the soil. Results presented in Table 9 indi-
cate The Woodlands groundwater contains greater amounts of
nutrients but less organic material than runoff. Low pH values
are consistent with the naturally acid soil due to the presence
of humic and tannic components, high clay and low carbonate con-
tent.
TABLE "9. COMPARISON OF GROUNDWATER QUALITY AFTER
RAINFALL TO RUNOFF WATER QUALITY
Gr oundw a t er'
Constituent
Runoff
TKN
NHg
TP
Ortho P
pH
2.1
1.4
0.09
0.02
5.1
1.37
0.08
0.06
0.003
5.9
Specific
Conductance
Total COD
SOC
200
22
10
110
59
22
I m ^"•"™™-^""*™™
Data in mg/1 except specific conductance as micromhos/cm and
pH in pH units.
2
Mean concentrations of samples collected within 3 days of
2.8 inch rainfall on April 14, 1975.
Mean concentrations for runoff samples collected at P-10
during April 8, 1975 storm event.
Stormwater Runoff Quality—
Discrete water samples were collected for the purpose of
n *1"*0*?1 stormwater quality during a storm. Over 850
samples were collected and over 12,000 separate water
quality analyses were performed. A summary of the water qualify
f^\i%PreSented ** Table 10' inclu<3ing mass load and flow
weighted mean concentrations for each parameter. Mean concen-
trations are calculated from discrete sample results weighted
39
-------
according to instantaneous stream discharge at the time of col-
lection. Mass load represents total amount of constituent pass-
ing the monitoring site during the runoff event. Mass load is
incalculable where no discharge occurred, as with lake storage
of stormwater.
All water quality samples collected are representative of
the total streamflow volume (including baseflow) which is re-
ported in Table 10. In the majority of events, overland runoff
is approximately equal to total streamflow and samples can be
considered representative of runoff. For storm events producing
low runoff, water quality is influenced by bassflow.
Chlorinated Hydrocarbons
During the two-and-one-half-year study some 2500 biotic and
abiotic samples were collected in The Woodlands and analyzed for
halogenated hydrocarbon compounds. The major sampling emphasis
was directed toward surface waters and aquatic fauna, however
soil-sediment and plant samples were also analyzed.
Polychlorinated Biphenyls—
The major class of chlorinated compounds detected during the
study was the polychlorinated biphenyls (PCBs ) . These formula-
tions are used in a variety of industrial applications such as
varnishes, paints, inks, waxes, flooring tile, synthetic rubber
and asphalt; however, greatest usage is in the electronics indus-
try in the production of capacitors and transformers. PCBs have
been produced for over 40 years but only recently have they been
observed as a pollutant in the environment.
The temporal distribution of PCBs in The Woodlands surface
waters and soils is presented in Figure 9. In January 1974 a
sudden increase in PCB concentrations in soils was observed which
was followed by a rise in surface water concentrations some four
months later. The rates of decline of both of these peak concen-
trations resemble a first order curve. No further increase in
PCBs in water was observed throughout the remainder of the study.
By extracting larger amounts of soil and concentrating the com-
pounds on activated charcoal followed by elution and subsequent
analysis, a second minor peak, three orders of magnitude lower,
was observed in soil samples during the spring of 1975. Several
plant samples were found to contain trace amounts of PCBs, but
the minute amounts could have been due to surfcice contamination
since they coincided with the highest values observed in water
and soil samples. The level of PCBs in aquatic: animal samples
started rising in late 1973 and reached a peak in April of 1974
(Figure 10), which is coincident with the highest concentrations
in the water samples (Figure 9). The concentration in the
aquatic organisms is about 10 times that in the; water samples
but 1/10 that observed in the soil samples (Figure 9).
40
-------
TABLE 10. RUNOFF WATER QUALITY SUMMARY—
MASS FLOW AND WEIGHTED AVERAGE
Storm
»
I
2
3
4
5
6
7
S
9
10
11
Date
01/18/74
03/20/74
03/26/74
04/11/74
04/22/74
10/28/74
12/05/74
03/04/75
03/13/75
04/08/75
05/08/75
Site
The Woodlands P-30
Hunting Bayou
Hunting Bayou
Hunting Bayou
The Woodlands P-30
The Woodlands P-30
The Woodlands P-30
The Woodlands P-10
The Woodlands P-30
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
Hunting Bayou
Westbury
Streamflow
acre— feet
2334
2.51
42.1
12.4
5.80
937.
1267
833.
3.32
Stored
.135
119.
79.1
2.38
1.77
2829
1614
93.4
93.2
28.9
7.16
Ortho-PO4
x cone
Ib mg/1
7.73 .001
2.96 .437
45.1 .395
16.9 .503
0.70 .044
43.0 .017
33.7 .010
14.1 .006
0.23 .025
MA .008
.012 .033
3.15 .010
1.15 .005
.043 .006
.102 .021
85.2 .011
15.0 .003
3.75 .015
1.25 .005
40.4 .516
13.8 .710
TP
x cone
Ib mg/1
No Data
No Data
46.4 .407
30.3 .899
7.53 .478
256. .101
328. .096
171. .076
1.14 .127
NA .097
.077 .210
52.3 .162
19.7 .092
.893 .137
2.54 .530
675. .088
264. .060
26.2 .103
27.9 .110
84.2 1.08
14.2 .730
NH3
x cone
Ib mg/1
215. .034
16.5 2.44
287. 2.51
24.6 .732
5.27 .334
236. .093
355. .103
199. .088
.880 .100
NA .062
.238 .650
22.2 .069
19.2 .089
.760 .117
1.06 .221
1144 .149
339. ,077
39.6 .156
28.1 .110
94.4* 1.20
17.3 .890
NA - Not applicable - a mass loading was not calculable.
Stored - No discharge of atormwater from the lake system.
Notes ac-ft = .123 ha-m
Ib = .4536 kg (continued)
41
-------
TABLE 10 (continued)
N02
x cone
Ib mg/J
5.30 .001
.353 .052
7.42 .065
2.18 .065
.170 .011
12.5 .005
18.5 .005
.280 .000
.055 .006
MA .008
.016 .04*
1.08 .003
.750 .003
.065 .010
.113 .024
73.1 .009
27.1 .004
8.24 .032
2.18 .009
4.52 .058
.754 .039
N03
x cone
Ib rag/1
181. .029
2.96 .438
58.1 .509
11.4 .338
3.71 .235
94.2 .037
79.9 .023
27.6 .012.
.960 .107
NA .189
.131 .357
35.9 .111 -
22.1 .103
1.32 .203
1.74 .362
1181 .154
284. .065
70.4 .280
37.6 .150
40.1 .511
7.20 .371
TKM
Ib x cone
xlo mg/1
Ho Data
No Data
40.1 3.52
5.25 1.56
2.11 1.34
Ho Data
311. .905
187. .826
1.50 1.66
NA .829
.066 1.79
35.1 1.09
34.5 1.61
.401 .620
1.99 4.14
1065 1.39
600. 1.37
33.3 1.31
.47.1 1.86
25.5 3.25*
4.25 2.19
TSS
Ib x cone
xlO3 mg/1
1492 236.
.480 70.9
22.5 197.
4.10 122.
14.8 939.
775. 305.
296. 86.2
60.5 26.8
.426 47.0
NA 166.
.104 283.
28.8 90.0
14.4 67.0
.991 152.
13.8 2877
1312 171.
168. 38.5
61.9 245.
322. 1273
16.2 207.
.474 24.4
soc
Ib , x cone
xlO2 mg/1
No Data
2.59 38.3
18.5 16.3
18.6 55.3
2.74 17.4
455. 17.9
No Data
No Data
.802 9.00
NA 14.0
.027 7.35
69.8 21.6
48.2 22.5
.900 13.8
.705 14.7
1563 20.4
962. 22.0
34.4 13.6
40.8 16.2
13.7 17.5
3.13 16.1
Total_COD
Ib - x cone
xlO2 mg/1
No Data
5.18 76.6
114. 99.8
34.3 102.
14.7 93.1
No Data
1962 57.1
1400 62.0
3.02 33.4
NA 47.0
.181 49.2
192. 59.3
136. 63.4
2.47 38.0
5.90 123.
4037 52.6
2572 58.8
106. 41.8
161. 63.7
140.. 179.
10.4 53.8
Soluble COD
Ib , x cone
xlO1* mg/1
No Data
No Data
No Data
10.8 32.0
8.53 54.2
859. 33.8
1405 40.9
988. 43.7
2.88 32.0
NA 32.5
.120 32.6
143. 44.2
111. 51. 9
1.87 28.8
1.37 28.5
2981 38.8
1888 43.1
66.8 26.4
82.0 32.0
27.5 35.1
6.09 31.4
(continued)
42
-------
TABLE 10 (continued)
Storra
1
12
13
14
15
16
17
Date
06/30/75
09/05/75
10/25/75
03/07/76
03/08/76
04/04/76
Site
Hunting Bayou
Westbury
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
The Woodlands P-30
The Woodlands P-10
Lake A
Lake B
Streamf low
acre— feet
116.
11.2
9.82
.918
Stored
1.51
117.
57.1
18.9
No Data
14.9
11.2
Stored
.884
99.3
58.3
6.78
2.71
45.9
2.83
14.5
5.10
Ortho^PO^
x cone
Ib mg/1
212. .677
17.1 .560
1.67 .063
.051 .021
NA .011
.148 .035
12.0 .038
.342 .002
.068 .001
NA .027
1.18 .029
.039 .003
NA .005
.124 .052
7.92 .029
.217 .001
.131 .007
^.088 .012
5.38 .043
.027 .003
.148 .004
.077 .006
TP
,. x cane
Ib mg/1
403. 1.28
34.6 1.14
4.75 .180
.082 .033
NA . 050
.608 .148
46.7 .147
7.32 .047
1.35 .026
NA .081
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
18.7 .149
.468 .061
2.49 .063
1.07 .078
NH^
x cone
Ib mg/1
660. 2.10
4.40 .145
5.36 .201
.075 .030
NA .069
.567 .138
26.1 .082
8.34 .054
2.27 .044
NA . 108
2.58 .064
1.67 .055
NA .063
.089 .037
46.8 .174
3.61 .023
.723 .039
.247 .034
48.9 .390
1.21 .158
4.79 .121
1.66 .120
NA - Not applicable - a mass loading was not calculable.
Stored - No Discharge of stormwater from the lake system.
Note: ac-f t = .123 ha-m
Ib = .4536 kg
(cont inued)
43
-------
TABLE LO (continued)
N02
x cone
Ib mg/1
14.0 .044
.803 .026
.480 .018
.007 .002
NA .009
.047 .012
1.56 .005
.690 .005
.143 .003
NA .010
.593 .015
.321 .010
NA .011
.035 .014
3.70 .014
1.66 .011
.225 .012
.328 .045
2.13 .017
.052 .007
.205 .005
.163 .012
N03
x cone
Ib rag/1
117. .373
11.8 .388
8.13 .305
.075 .030
NA .009
.955 .233
47.0 .147
4.67 .030
1.19 .023
NA .113
10.2 .252
1.43 .047
NA .076
.393 .164
48.0 .178
3.69 .023
1.64 .089
1.31 .141
52.9 .420
1.04 .135
4.53 .112
4.92 .355
TKN
Ibs x cone
xlO mg/1
124. 3.94
4.51 1.48
1.16 .435
.024 .097
NA .172
.168 .410
9.69 .305
4.04 .261
.942 .184
NA .280
No Data
No Data
No Data
No Data
No Data
No Data
No Data ,
No Data
25.5 2.04
.516 .672
2.90 .735
1.29 .929
TSS
Ib x cone
xlO3 jag/1
51.0 182.
2,13 69.8
2.67 100.
.016 6.54
NA 24.0
6.70 1633
53.8 169.
1.18 7.60
1.85 36.2
NA 421.
9.55 237.
6.47 212.
NA 184.
1.77 738.
78.1 290.
20.6 130.
3.27 177.
3.07 419.
No Data
No Data
Mo Data
No Data
SOC
Ib , x cone
xlO^ mg/1
71.6 22.8
5.09 16.7
No Data
No Data
No Data
No Data
66.0 20.7
37.8 24.4
2.38 4.66
NA 20.6
5.55 13.7
6.72 22.0
NA 8.00
.322 13.4
42.4 16.0
41.4 26.0
t.ll 6.05
.962 13.1
18.5 14.8
1.27 16.5
3.98 10.1
•1.94 13.3
Total_COD
Ib _ x cone
xlO^ mg/1
252. 80.4
11.8 38.9
No Data
No Data
No Data
No Data
135. 42.4
70.0 45.1
10.9 21.2
NA 48.9
No Data
No Data
No Data
No Data
No Data
No Data
No Data
No Data
61.7 49.3
3.08 40.0
7.88 20.0
6.00 43.3
Soluble COD
Ib x cone
xlO2 mg/1
61.2 19.5
4.36 14.3
9.40 35.2
.852 34.5
NA 19.5
.878 21.4
77.0 24.2
58.9 38.0
7.21 14.1
NA 24.2
12.2 30.2
14.2 46.6
NA 15.4
.631 26.3
88.3 32.8
86.6 54.7
2.73 14.8
1.78 24.3
43.3 34.6
2.97 38.7
9.20 23.3
5.15 37.2
(continued)
44
-------
TABLE 10 (continued)
Storn
1 Date
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
01/18/74
03/20/74
03/26/74
04/11/74
04/22/74
10/28/74
12/05/74
03/04/75
09/13/75
04/08/75
05/08/75
06/30/75
09/05/75
10/25/75
03/07/76
03/08/76
04/04/76
Site
The Woodlands P-30
Hunting Bayou
Hunting Bayou
Hunting Bayou
The Woodlands P-30
The Woodlands P-30
The Woodlands P-30
P-10
The Woodlands P-30
Lake A
Lake B
The Woodlands P-30
P-10
Lake A
Lake B
The Woodlands P-30
P-10
Lake A
Lake B
Hunting Bayou
westbury Square
Hunting Bayou •
Westbury Square
The Woodlands P-30
P-10
Lake A
Lake B
The Woodlands P-30
P-10
Lake A
Lake B
The Woodlands P-30
P-10
Lake A
Lake D
The Woodlands P-30
P-10
Lake. A
Lake B
The Woodlands P-30
P-10
Lake A
Lake B
Turbidity pH
150
135
100
305
250
65
20
35
121
184
50
30
105
1200
130
28
165
330
110
18
140
14
100
7
19
830
120
12
24
325
74
12
137
138
S
-
120
70
7
53
110
6.5
6.5
6.5
5.7
6.0
5.2
7.5
7.5
8.1
6.6
6.1
7.5
7.8
6.8
5.9
7.54
6.9
7.6
7.3
7.2
7.3
7.05
6.6
8.01
7.6
7.4
5.8
8.1
7.4
_
_
-
_
_
7.3
6.9
7.9
7.2
Temp.
14.7
20
15
22
22
14
16
13.5
13.3
17
11.5
19
20
20
24
27
_
_
18.5
18
22
19
_
_
-
_
—
-
_
_
-
Spec.
Cond.
-
545
200
110
58
56
375
115
415
280
290
160
185
103
110
140
80
460
175
360
125
208
280
336
135
275
247
536
167
281
410
155
365
410
115
318
284
363
125
Total
DO Solids
9
2.5
4.6
2.25
3.5
8.3
7.8
7 6
6.4
7.2
7.7
7.4
7.8
6
7.65
7.1
3.2
3.34
3.3
4.6
_
_
7.5
6.6
8.2
7.6
-
_
_
-
_
-
-
-
450
• 385
110
_
~
^ .
_
-
_
_
-
-
-
_
_
-
_
—
-
-
_
-
_
BOD
17
22
6.1
-
_
_
_
-
_
_
-
-
-
-
—
—
-
-
'-
_
All measurements in rog/1 except:
Turbidity in JTU; pH in pH units, Temperature in
Centigrade and Specific Conductance In micromh,
.cromhos/cm.
45
-------
GO O
O
rQ
a
«
n.
q
cJ
o:
bJ
o
o
(M
QQcn
°
CL
o
o
0
co n
-z,
o
CO
CD
LL.
Q
O
CO
ti-
ro
O
CO
-P
rd
C
•H
M
O
rH
^3
u
O •
O. 10
o o
-H O
^
ra fi
•H -H
13
CO
rH H
ro >i
M Ci
O (U
p, jCj
S Pi
0) -H
CTi
(I)
$-)
3
tr>
46
-------
CQm on
Oft- on
CL&oa
o !P 9 £! o>
bi S- 10 cvj o
— , j- — r- | r~
*
•
J
^
SD
ED
CD
•
•
D
O
S3
• •
• n
OH •
a «o
«
• °
•
^
•
, , , . Q
Q
O
CO
^^r
m/
^^M /^^
2 "~
*^
2> •
ll.
"D
Q rf)
.^ i»w_
oa>
CO~"
o o o o o
^3- ro cvj —
u
•H
-P •
^( •». •
td • to
O in to
03 -H (0
tH M ,g
8,9 ri
5 U
^J fd —
•S^.c
& -H
H3 ID M-l
0) -H >i
J-> M-l (0
fd O M
-S-H ^
H to 0
&e
"o • ^
• to
M-l
O • CQ
01 -H
fi "D S
o H o
•H (3 a
-P i-) (U
-H O
jj IH
03 QJ -H
•0 fn W
(d -H CQ
W
|gD
-------
m The source of the PCBs in The Woodlands is not known. It is
of interest that the peak of PCBs in all components of the eco-
system appeared during a period- of intense ct.t and fill opera-
tions as well as utility installations. An abandoned land fill
with disposed capacitors or other electronic materials could have
affected the increase in concentration as could the use of road
oil contaminated with PCBs. Neither of these possibilities were
verified by observation.
Chlorinated Hydrocarbon Pesticides—
In the first year of investigation trace levels of DDE were
observed in samples of crayfish, mosquito and bluegills obtained
from The Woodlands aquatic ecosystem. After completion of The
Woodlands Golf course, the surface waters and soils in ponds and
adjacent ditches were examined for halogenatei compounds.
In the spring of 1975 Mirex, a chlorinated camphene, was
detected in water, soil and some aquatic organisms (Figure 11).
The highest values were found in mosquitofish and the lowest in
water samples with the residues in soil being intermediate be-
tween the fish and water, if there was any biological amplifi-
cation _ in the fish from this aquatic system it was limited to
approximately*a four-fold increase (Figure 11).
Water from the Conference Center Lakes (A and B) was used
for irrigation of portions of the golf course, since these man-
made impoundments were the potential recipients of both irriga-
tion and stormwater runoff, the lake water sediments and the
mosquito fish were examined for Mirex. The results are shown in
Figure 12. The highest level of pesticide was observed in raid-
to late summer (1975) and thereafter concentration in all these
components of the pond ecosystem diminished rapidly with the soil
residues showing the slowest rate of decline. Mirex concentra-
tion in Lake Harrison was at least one order of magnitude lower
than concentrations observed on The Woodlands Golf Course. The
lower concentration due to lake water dilutior. by groundwater
pumpage, runoff and treated wastewater.
During August of 1974 chlordane was also detected in the
golf course study. Residues were found in soil, water and aqua-
tic organisms (Figure 13). The highest levels of chlordane were
found in the crayfish from golf course ponds. Somewhat less was
observed in the same organisms from ditches adjacent to the
course. Similar, although not as pronounced, results were ob-
served in the concentration of chlordane in the waters from the
same areas (Figure 13).
Bacteriological Enumerations
Surface Water Characteristics—
Bacterial counts during dry weather were determined in
48
-------
60
50
40
MiREX
(PPB)
30
20
10-
0
o
ASONDJ FMAMJJASOND
1974 1975
Figure 11.
Temporal distribution of Mirex in
The Woodlands gold course.
•-• mosquitofish (Gambusia sp.);
O-O soil; o-o water.
49
-------
40 -
30
Ml REX
(PPB)
10
n n
ASONDJFMAMJJ A~S 0 N D
1974 1975
Figure 12.
Temporal distribution of Mirex in the
Conference Center Lakes (A&B).
d-D soils; •-• mosquitofish (Gambusia sp.)/
o-o water.
50
-------
40 -
30 h
CHLORDANE
(PPB)
20
10 -
Figure 13 .
ASONDJ FMAMJ JASOND
1974 1975
Temporal distribution of chlordane in The Woodlands
gold course A-A crayfish (Carribarus sp.)from pond^
A-A crayfish (Cambarus sp.) from ditch? o-o pond
water; .-. ditch water; mosouitofish (Gambusia sp )
from pond; a-O soil from pond. — sp';
51
-------
Panther Branch. Indicator bacteria numbers varied widely as
indicated below:
Total Coliform (TC)
Fecal Coliform (FC)
Fecal Streptococci (FS)
pseudomonas aeruginosa (PS)
Staphvlococcus sp. (ST)
Salmonella sp_. (SA)
Range (number/I00 ml)
200-10,000,000
10-214,000
10-1,580
10-53,600
1-23,400
1-5,800
The maximum values are comparatively high for a rural, forested
area receiving no direct sewage discharges.
Storm events monitored for water quality were also monitored
for microbiological content. The range and mean bacterial counts
observed during each storm event at The Woodlands are presented
in Table 12. All bacterial counts are reported as logarithms to
base 10. Substantial numbers of bacteria were observed in The
Woodlands storm runoff, including runoff from the undeveloped
watershed, site P-10. pathogenic species were identified at all
Woodlands monitoring sites in relatively high numbers. For ex-
ample, Salmonella sp_.were identified at a mean count of 77/100 ml
for storm event #10 at site P-10. In the majority of events the
maximum bacterial count occurred before or coincidental to the
hydrograph crest.
A summary of the three storm events monitored at Westbury
Square is presented in Table 11 (the storm event of 11/75 was
not monitored for water quality).
TABLE 1.1. SUMMARY OF WESTBURY RUNOFF BACTERIAL QUALITY
Storm
5/75
6/75
(n=12
TC
FC
FS
PS
ST
SA
TC
FC
FS
PS
ST
SA
log No./lOO ml
Min. Max.
8.70
4.60
4.61
4.40
4.30
Mean
7.48
4.34
4.12
3.88
3.91
(continued)
52
-------
TABLE 11 ("continued-!
log No./lOO ml
Storm
11/75
(n=5)
1
TC
FC
FS
PS
ST
SA
Number of
Min.
6.26
3.96
4.39
4.31
TNTC
2.00
Samples
Max.
6.67
4.50
4.68
TNTC2
TNTC
4.17
Mean
6.49
4.28
4.51
4.54
3.19
2
Too Numerous to Count
Stormwater bacterial counts at Westbury were aen«*r*Ti
greater than or equal to counts observed at^hJ WoolKSs
sample variations were common for P
-------
H
CO
tN
H
^i
o s
o
,-j
s
*0
tfS
SI
M
B
8
S
TJ
fl
CO
o|
0
• 4
a 5
m
o
5? c
ai
,.
4-
1
c
c
•r
n
a
co VD oo r~ in "=r
ro oo ^ oo ^j* CT»
in ro ro CN ro rH
CN OO -<3< O O CN
O CN CN O OO -^
vo *3< "3* co co CN
CO O O O O C>J
^< vo o o cTi in
•<3« CN CN CN CN rH
a K e s R a
r~
in o
\ ro
CN 1
rH ft
CN 00 rH
a\ vo vo i i
• ••III
T* ro ro
r~ CN o
ro ro ro I 1 1
• • • I I I
in in m
en r- «* i i
i i i
ro CN CN
o y co co EH rt
^
r>
oo o
•H ro
> a,
ro o in in r~ oo
rH CN VO rH r- 00
in ro CN CN CN rH
CN v£> vo t — ro CN
ro o *3* t — r~- in
VD ^ ro CN ro CM
O O O O O
o o o o o ro
-3« CN CN rH rH rH
V V
CJ CJ CO CO B f<
V.
m o
\ rH
CN 1
in CTI ro
rH 00 f- 1 1 1
• ••III
r- ro ro
ko ro ro CN
r- iH ro "* 1 I
• • • • 1 I
o o in oo
o CN en r- i i
• • • • i i
VD CN rH CN
o y co co EH <:
^
v^
CN 0
CN ro
\ 1
r- o r~ o o
rH o ro oo ^r
in ro ro H ro
rH CN ro o VD
VD ro o o H
in ro ^ CN «*
oo oo in oo en
vo in oo "* en
«* CN CN rH CN
y p co co EH
EH fa fa ft CO
m
\ o
•<3< CO
ro ft
oo ro ro vo r-
cn VD r- ** r~-
in ro ro ro ro
o in in in CN
en vo rH •ctf in
vo ^j* in ^i4 m
00 O O O O
r~ o o ro o
«* CN CN CN CN
E^ % S S &
f"**
CO
-------
•8
n
f4
-iH
-p
C
o
u
CN
i-H
11 1
§*
rH
6 >
2 fl
O
M 2
£
C
•i—
c
e
C
£
C
to
S
o s
o
rH
§3
m ^
t7> CJ
Q "jd
M S
3
«J
H
fi
I
B
CO
CTl
in
in
0
ro
•3*
%
in
£^
CN
rH
ro
CO
in
in
ro
o
•
vo
vo
ro
in
&J
I-O
[""••I
\
ro
ro
Cn
CN
3
ro
O
O
rH
V
K
o
H
ft
co
ro
ro
o
en
*
ro
O
vo
rH
fej
«
n
M
t~-
00
CM
ro
ro
o
0
rH
V
ft
rH
CN
•«*
00
ro
•
**
vo
ro
ft
in CN ro
in in in
rH CN rH
en o o
VO ^ rH
CN ro CN
000
o o ^
rH rH rH
V V
ft % $
CTi CN O
ro rH ^
ro «* rH
\/
ro in o
OO ^J* ^
ft ft ft
ro 'tf rH
V
r^ r^ o
VD VD ^
CN ro rH
V
ft & $
OO
rH
VO
in
t^
vo
0
vo
^3*
^
in
v^
CN
rH
ro
CN
in
•<*
0
r-
•*
in
00
ro
#
in
r-
^
ro
CN
ro
vo
ro
0
o
CN
V
K
CQ
(U
1
in
00
o
ro
•
rH
0
0
•
rH
V
&
<
"1
l-H
ro ro ^ in
in r- o CN
ro (N rf rH
ro co ro rr
en m CN CN
ro ro in CN
o o in o
0000*3*
CN CN CN rH
V V . V
CO CO EH <3
pn pt tO CO
VO CN ** O
rH in rH •*
• • • •
rO rH CN rH
V
CN ro •<* o
ro CN CN rH
V
O 0 0 0
vo o o «*
CN rH rH rH
V V
£0 CO EH rf
rH ^*
*^f* rH
en in
^1* rH
in CN
0 ' 0
O 0
^t* i~H
V
B K
in
\ ***
CN . vo
vo ro
00 O
r~ o
y &
in
r-
CN O
•H ro
ro" fl
CN
rH
ro
ro
in
ro
00
00
rH
r°
b
o>
CN
CN
cn
e
ro
o
0
*
__|
r^
S
r-
m
en
rH
o
o
r-|
V
s
CO
in
rH
in
CO
•
H
O
O
1
I— 1
V
ft
o r~~
m »*
rH rH
in o
on **
rH CN
O 0
o ^
rH rH
V
EH
-------
•a
0)
•H
4J
cs
o
o
r~*
pa
£
«
<
B
tii
o S
o
> .
1
o .
tr> c!
Si
1
Station I
rH C
Ei "Q
o S
o
rH
* •
& I
si
^
1
'•r
£
f-
|
|
u
VD ro o rH crj in
o o in VD CM ro
in ro ro 01 ro
Ol O VD VD VO O
vo vo o ro ro o
in ro "S1 ro ro oj
o o o o o rH
o o o o ro m
•<3< Ol 01 Ol 01 rH
v v
HKBRBtf
in *3
oo oo r^- r** vo CNJ
^* CO ^* VO C"*4 C^*
to ro ro oj ro r-H
0 0 01 rH in 04
VD ^ 'd' ro ro 01
t-~ O O O O rH
r~ o o o ro m
ro 01 01 01 01 H
p O en CO EH rf
EH pL4 P4 CM to c/j
in
r^
' \ 0
i c: 7
! > d
oo en 01 01 g o
o o r~ VD ro o
^D ro in «* "* r~i
•^i o cr> o o o
•<* o in o r- o
r> r) CO CO &•< ^
m
^ o
^ 2
crv CM
oo VD in in oo cr>
rH ro o ro co oo
• •••••
in ro ro 01 01 rH
CTi O "* VD in VD
r-- 'S1 04 oj -31 ro
vo ^ ** ro ro CM
Ol O O O O rH
ro o o o ro in
^r 01 01 01 01 rH
v v v
& & g S g 3
in
r-
"\ o
\ I
•* CM
56
O rH CT O O
ro *3* < — VD VD O
531 Ol CN rH rH r—1
CO CO f' O CO O
in r-- v£i cr> 01 o
IT) Ol Ol rH Ol rH
88-8888
tf "' v1 v
o o' 01 co fc fjS
u tj_j Jj,| p^ Ua uJ
m
S s
of cu
in ' r~- c"> rH r-- co
01 ro r-i co rH CM
in ro n 01 ro
o ch o r~ ro rH
co ro "V in oo in
m ^f T ro ro 01
O O CD O CO rH
O O O O ^J* in
•vf Ol CM Ol Ol rH
V
g B 12 B S-fii
in w
r~-
r^ w*
\ 5
•^ M
•H
fl
O
-------
ts
Q)
CJ
•rH
a
o
o
CN
m
PQ
^1
0
H
g >
MH
O
S3
c
•f-
"5
c.
H*
6
•r-
-i-
4j
to
5 (Q
o s
o
rH
,O X
5
?l
i-4 g
§
1
g
s
CO
i — 1 CO VO CT\ OO tf
in OO OO CN CN rH
CN rH O"i f» OO C —
O CO OO O^ n* rH
VO «vf -^ OO OO CO CO EH rt* CN CN
CN oo r- vo
• • • •
^* ^1* CO r-H
^ p*- to r**
*>3* ^3* LO ^^
•^j rH r-- o
OO OO CN rH
V
CO CO EH -f£
&4 Oj CO CO
•a
-------
rH C
6 ro
o s
0
rH
S
UH
O
Sc
i
M
B
8
1
5
CO
*I
o s
o
s d
52 j9
UH
0
Si
1 |
« T
8 ^
3 -1
H 4.
W 1
; i i i : '<
a s £ 3 o §
in co co co co rH
V
o oo oo in vo o
ro r~ rH rH CO O
*3< CM CO CO CM rH
B a a s K a
VD
\ o
r^ co
CO Ol
1 i ! i ! i
CM O (T> ^ rH O
CT> oo f~ o ^ r~
in •«* co co "sf CM
O «* CO O O O .
ro o CM co r~ o
in co co CM co CM
fi jjtj fe ^ 55 en
i m CQ
i r~-
' ^ S
1 CO 1-4
1
i i i i : :
,H ,-H CO in 0 0
rH cri oo en in o
in CO CO rH •si' CM
VD 00 00 0 Op 0
ro r- r~ o CM o
Ttf CM CM rH CM CM
o o cn cn EH rt
P K En fl< &3 03
VD CQ
s "
CO I-H
:!!:::
T* in ** o oo o
o ~l ^o
•«^ ro co CM co CM
o ^ o o in o
CM VD F- O <3" O
• •••*•
rjt fM CM i-H CM CM
O O W CO EH <
f) ftl |i| P-i UJ CO
vo
\ o
r- rH
ro- 04
0)
3
fl
•rl
rj
o
w
58
-------
!i
.0
•H
SI
° .
!
s
g
s
CO
rH. M
R (O
O JaJ
o
rH
O
si
Q) &
1 S
t) 'r-j
C! "g
O fl
CN 8
i
g &
M
i i i i i i
i i i i i
H oo rH o in o
m r~» rH r~- «* o
vo ro ro CM •TP CM
H o m o in o
oo en ** o "3" o
"* CM CM CM ro CM
V V
O O CO CO EH rt
u) pq
\ m
in ^
i i t i i i
i i i i i i
U3 vo ro rH oo o
<7I 00 O rH rf O
*>f CO "* CM »* CM
V
oo o vo o in o
o CM m o a\ o
•^ ro ro rH ro CM
V
^""
VD
\ O
§ 2
i i i i i i
ii i i i i
P** G^ G\ O C^ O
CTi ( — O CTi OO O
CO CM CO rH CM rH
V
VO OO CTi O O O
r- »* vo o o o
ro rH CM rH CM rH
V
RK**H*
VD <|
in ,x
( I I I i I
I I i I I I
^o o vo o oo o
CM in r~ ro CM o
• •••..
"* CM CM CM CM rH
CM O O O OO O
r~~ CT> ro o «3< o
rO rH CM CM rH rH
V V
U CJ CO CO EH f£
vo
\ o
S 2
-
i ! i ! ! !
r- o cr» o cr> o
t-» o m o in ro
m ^ ^j* ro ^3* CM
o o o o o o
o o o o o o
rO CM CM CM CM CM
V V V V
P fa fa CU W CO
\ 0
in ro
59
-------
TABLE 13. SUMMARY OF PANTHER BEANCH AQUATIC ALGAE
Division
Number of Species
Identified in Panther Branch
Chlorophyta
Cyanophyta
Chrysophyta
Euglenophyta
Pyrrophyta
92 a
15
16
1
1
43
8
12
15
1
79 Total number
a number of taxonomic species/division
Standing crops of algae at site P-10 and P-30 are presented
in Figures 14 and 15, respectively. On a seasonal basis, algal
cell numbers fluctuated more at site p-30 than at p-10. The
algal standing crops at both sites were dominated primarily by
members of the euglenoids (Euglenophyta), and green algae
(Chlorophyta) were minor components of the total algal popula-
tions. Numbers of blue-green algae (Cyanophyta) were compara-
tively low at both sites, even though they were more numerous at
site P-30. A survey of various collection sites along Panther
Branch (May, 1974) also indicated that standing crops in this
stream were dominated by euglenoids and/or diatoms (Chrysophyta),
while numbers of green algae and blue-green algae were compara-
tively low. Dominance of this type is indicative of slightly,
acid streams with high organic carbon content.
Soil Algae—
Fifty-two genera of algae were identified in soils collected
from disturbed and undisturbed sites in The Woodlands (see Table
14). Undisturbed soils had larger algal species diversities and
fewer algal numbers than disturbed soils. Green algae (Chloro-
phyta) genera were most numerous in soils from the forest, while
disturbance of soils favored the development of a more diverse
blue-green (Cyanophyta) flora and inhibited green algae diver-
sity.
The change in algae characteristics is probably a result of
increases in soil pH and nutrient content which accompanied soil
disturbance and fertilization. Undisturbed soils had the lowest
pH (6.1), compared to high pH values of 6.7 to 7.8 in disturbed
soils. The results confirm floristic surveys in other regions
of the United States which show that alkaline soils favor the
development of more luxuriant blue-green algal flora than do
acid soils (41, 42, 43). Higher algal numbers in disturbed soils
corresponded to the greater concentrations of nitrogen and phos-
60
-------
CHLOROPHYTA
CYANOPHYTA
CHRYSOPHYTA
EUGLENOPHYTA
TOTAL
NDJFMAMJJASO
MONTHS
Figure 14. ^Seasonal algal standing crops at P-10
in Panther Branch.
61
-------
I I I I I I I I I I I I
Chiorophyta
—— Cyanophyta
—x— Chrysophyta
Euglenophyta
NOJFMAMJJASO
TIME (mo)
Figure 15. Seasonal algal standing crops at
P-30 in panther Branch.
62
-------
phorus found in these soils. The increase in soil nutrient con-
centration is due primarily to fertilization.
TABLE 14. SUMMARY OF EDAPHIC ALGAE IN THE WOODLAEIDS
Undisturbed
Disturbed Soils
Division Forest Soil Golf Course Lawn
Chlorophyta
Cyanophyta
Chrysophyta
Euglenophyta
Total
27a
3853b
586
{' 11
1297
3
46
5736
7
39629
4
1446
4
13084
1
16
54159
4
17400
4
7195
3
12790
— —
11
37385
a numbers of
determined)
b , -
genera identified
(species
were not
Since terrestrial and aquatic ecosystems are connected
hydrologically, they cannot be considered as totally disjuncted
units. Surface drainage serves as a major component in this
hydrologic linkage and is thus an important ecological parameter.
Land usage in a watershed can determine the quality, as well as
the quantity, of surface runoff, thus influencing aquatic
habitats in the watershed. In addition to providing potential
nutrients for algal growth, surface drainage probably transports
algal cells to aquatic habitats from surrounding soils. Since
land use affects edaphic algal populations, it must also influ-
ence the diversity of algae which could be transported by sur-
face runoff. For example, a more diverse assemblage of blue-
green algae could potentially enter lakes in the study area by
surface drainage from disturbed, rather than undisturbed,•
soils.
Stormwater Comparison with Sewage
Table 15 compares stormwater runoff and sanitary wastewater qual-
ity. Stormwater constituents are lower than raw sewage with the
exception of suspended solids in runoff from Hunting Bayou and The
63
-------
fa
0
gj
§
CO
Q
W
EH
M
s
EH
p
§
X
O
rfj
s
fa
CO
Q
H
EH
gj
EH
J2j
P
fa
I
P**j
P
01
PS
EH
Si
§
O
EH^
%%
>H !3
EH H
H
, 1 rft
HH **••
gs
Ot ^
OS
fa Kl
EH J
& 5.
*
m
rH
ij
CQ
,-jj
EH
H
(d 'd
o c
•H (d
e g
0) CD
j; p
o
a
rH CD
(d tn
-P ^i
O &
EH 0
CO
(d o
4J ^f*
o a
EH CO
O
ft
^
^-4
(d
*d c!
i fli
n™i vy
0) Cr>
•n O
K SH
jj
•t**
rH -H
(d 3
I 1
•H
o
"
CD
rH >d CO
-P CO -H
O P|"~l
EH W O
3 CO
CO
H
fa
0
1
W
<
&
o
EH
W
co "*
VO in rH •<* rH in
1 I I I I I
ro o o\ co r*1* o\
^t1 ^J* ^ ro r**» ro
o^ o ro ^}* oo ^J1
o ro in rH CNJ H
O O 0 O H rH
1 1 t 1 1 1
co co i — i in i — i co
o H H o •<* r-
o o o o o o
rH rH 'J rH <* O\
VD ^ H CO CT> rH
H H ^ H CO CM
1 1 1 1 1 1
O VD O> t^ VD CO
rH o r- H m •*
CJ O rH O rH rH
H r- in t~-
r— CM co "5f o o
VO CO CM CM CM r-
1 1 1 1 1 1
in cri ro •<* I-H -sj1
• O CO CM t-« CM
VD rH CM
1
rH
CO
^
CO *-+ CD
P 0 }H
^-* (*) 3 id
o M i OS
rH CO ft ^1 D1
1 rg <_- (d CO
ft £ PQ
- G CT> M
CO CO g -H ,Q O
SnMOcXrXCco !5
O O (d (d 3 CO W
pti fa i-3 i-3 W |2 CO
o o m
in '* H
CO
O O rH
H rH
o o o
d o
CO fl C!
in o (d
•P u >
C 0 'd
p co <
CO
•P
fl
CO
CO
H
n
o
to
I>1
m ly
o c
CO
JH -rl
CO O
fit 1
It 1
c! fa
SH d
O CJi
II t „ 1
to
to d)
s p
•rl 1
•p
(d -P
»H d
•p rd
fl rH
CO ft
O
c a
0 0
O -rl
•p
G fd
fd g
(D (d
S o
I s
t^! M
tJi CO
•rl -P
f &
•* rn
of flo\
ored.
oodlandi
0) -P !3
c! C CD
id o ,£
& E EH
•H CM
64
-------
Woodlands developing areas. Advanced wastewater treatment yields
lower GODandTSS values than stormwater, however nutrients remain
higher. AS a result, sophisticated nutrient removal processes
would be necessary to achieve low concentrations observed in
stormwater runoff.
Ten samples of untreated and chlorinated secondary waste-
water effluent were obtained from two large, well-operated
sewage plants in Houston for bacteriological examination. The
data obtained is compared to the bacteriological quality of
stormwater runoff in Table 16. Bacterial numbers were lowest in
chlorinated effluent, higher in runoff and highest in untreated
wastewater. Urban runoff contained higher numbers than forest
runoff.
OBSERVED TEMPORAL AND SPATIAL VARIATIONS IN STORMWATER RUNOFF
QUALITY
Pollutograph Analysis
A pollutograph is defined as a plot of pollutant concentra-
tion versus time. In this sub-section it is plotted for storm-
f!6w quality. Temporal changes of water quality during runoff
events are important to the understanding of the impact of these
non-point sources on stream quality. The time-concentration rela-
tionship _ is also critical in consideration of stormwater treatment
alternatives. Pollutographs observed during the study exhibited
the five generalized patterns shown in Figure 16. These concentra-
tion patterns were common to all watersheds, although levels of a
particular parameter were site dependent.
Specific conductance of groundwater which feeds streamflow
is high due to dissolved minerals and as a result stormwater in-
flow decreases stream conductance similar to the second polluto-
graph shown in Figure 16.2. This dilution pattern applies to
other streamwater constituents, including7 some found in waste-
water effluents, which are concentrated in dry-weather flow. The
concentrations of SOC, soluble COD and total COD often increased
as runoff progressed, with highest concentrations observed at the
end of the runoff (Figure 16.4). Streamflow contributions from
interstitial and bank storage flow is greatest late in runoff
and could account for the pattern if enriched by contact with
soils serving as an organic carbon source. DO concentrations
in stormwater increased proportional to flow and assumed a
hydrograph-shaped pollutograph (Figure 16.5). Increased re-
aeration at greater streamflows accounts for this phenomena.
Several parameters observed at site p-10 remained at a constant
level throughout the hydrograph, including pH, NHo, NO, and
soluble COD (Figure 16.3). This pattern was not commonly ob-
served at the other watersheds where land use is diversified.
65
-------
CD
O
pa
o
oi
o
OS
w
E-
<;
3=
w
H
en
a
2
Pi r-H
w s
H
< O
S: O
OS
O
H
en
erf
m
CO
s
O
CO W
t-t t-t
erf co
< >>
cu j
S <
o ss
CJ <
VD
r-l
CO
E-
H 01
IS 4-»
Hunici
Was tew
Ji
4J
n
*
c
I
"1
P3
4J
C
3
S
M
P.
go
ra c
•00) ,
S3
0 O
fra*
c
I'St
l< 4J 0-0
Untreate
X 3
4J
• C
en a
ss.
x§
ss
WJ
* C
D» 3
S3
• C
II
4J
X §
S3
!"£
II
10 r-
•r co •
co r- «>
r-C iH 'r-(
XXX
IN O CO
0 O O O O C
X X X X X fc
ft *»
m in » CM w
CO •» •* ^ •» "O
X X X • X X tn
t- «> f> en m -u
tn in tn in ^H o
rt M M H H
CD
o
:!!!!!
« I 1 ! 1 i
n
n
H
» » » „ n
IM
X X X X X n
Ol
vo in \o n to
in ft ft f
000 0
* x « S « S
O^*O CD
n (M n H
*^a **o ^a *"o o
» r- H ^ «
o o o
H rt ^ ^ o
X X X w jo co
•-C ^ M
in ft r*t
f-l N H
L'
0) • «)
u i i nJ >H
E E 4J 0 Oln) pH
C M W-H E • -HI V
o o u o oJ >itn c
,H (M pH **-» ^ O T3 «1 j:]D Q
ni*ri n -H flj o 3 oJO E
4Ji-l U>H O O OH fljIU i~4*
es as £s ssl sis isa
OJ
c
0)
sewage treat]
I
I
i
•o
01
u
1
n
at
n
1. Averaged
n n n a
: 0) t" D^ DI
(0 0) (U 0> 0)
(N PI ^ tn
66
-------
O
s
cc
*-
Ixl
O
z
O
O
UJ
UJ
ac
4
UJ
o
tr
FIRST FLUSH
DILUTION
STABLE
INCREASING
HYDROGRAPH
TJME INTO STORM EVENT
Figure 16. Generalized pollutographs observed
for storntwater parameters.
67
-------
The "first flush" pollutograph pattern is characterized by
an abrupt rise in concentration early in the runoff event. At
Hunting Bayou, the "first flush" was observed for the greatest
number of runoff constituents including TSS, turbidity, ortho P,
TKN, TP, NH3, NO2 and N03. The Hunting Bayou watershed has a frac-
tion of impervious surface area and includes a diverse land use
mix including industrial and commercial activities. Turbidity and
TSS parameters exhibited the highest peak values over baseline,
while peak values for other constituents were,less pronounced.
Effect of Land Use on Stormwater Runoff Quality
Pollutant Load-Runoff Relationships—
Total pollutant loads were plotted against total runoff of
each storm event and regression lines were fitted to correspond
to Hunting Bayou, Westbury, P-10 and P-30 watersheds. Fitted lines
and associated correlation coefficients are shown in Figures 17-20
for the constituents TSS, total COD, TKN, TP, NC>3, NH3, soluble COD
and SOC. Correlation coefficients for a majority of the parameters
were greater than r = 0.8 (r = correlation coeff.), however, three
cases showed fair to poor correlation; P-30, NO3 (r =0.775),
Westbury, NH3 (r= 0.397), and Hunting Bayou, SOC (r =0.161).
N03, NH3, TKN and TP relationships, shown in Figures 17-18,
indicated that the urban watersheds produce nutrient loads greater
than the forested watersheds. In all cases Hunting Bayou nutrient
loadings are highest, followed in order by Westbury, P-30 and P-10.
TSS loads are highest in P-30 as a result of construction
activities in the watershed and urban runoff solids loading is
greater than forest runoff (Figure 19).
Runoff loads for nonspecific parameters ftota] COD,.soluble COD
and SOC are shown in Figures 19-20) are higher in forested water-
sheds than urban watersheds, with the exception of high total COD
loads from Hunting Bayou. The data suggest organic material in
runoff decreases with urbanization, however, insoluble pollutants,
sediments and oils will increase.
A ranking of the four watersheds, on a Ib/ac/in of runoff
basis, illustrates the relative conditions for each site and pollu-
tant. Table 17 shows these rankings for mean regression values
at one inch (2.54 cm) of runoff for all parameters and sites.
Confidence intervals (95%) are included in Table 17 to indicate
significant differences in pollutant loads at one inch (2.54 cm)
runoff. Confidence limits for Westbury show a particularly large
spread due to the small number (2) of storms monitored for that
watershed. Significant differences for the total COD, soluble COD
and SOC cases are indicated, with some overlap in the TSS case.
The patterns of nutrient response for the urban developing and
forested watersheds is distinctive, with the urban response pro-
ducing loads up to an order of magnitude larger. Hunting Bayou
ranks as the producer of the largest pollutant loads.
68
-------
1
o
k
LU
O
0.
<
i
o.
O -P
m a
-------
CE
IU
o
ov/san
u.
LL
O
£L
<
Q-
I
Q.
O
Ul
X
O
CO
o
(O
o
•3-
O
w
o
o
o
ro
W
m
ro
O
O -P
(U
to -H
(i O
•H -H
W M-)
Cl (1)
O O
-H O
-P
OJ -H
M^
11 r [
m
-------
I
00
UJ
X
o
to
o
a.
O
P
o
o
H
(0
-p
o
4->
d 'd
(0 0)
•H
tots
co 3
E-i -P
03
5-1 •—
O tn 4-1
""* (1) 0)
W rC -H
& 03 O
•H S-i -^
& cu m
tD 4->
4J fd
m rd Q^
O S-l
3 m o
MHO
m
O 03
en
rH
Q)
tr>
•H
e
o
in
CM
II
£
•H
ov/sai
71
-------
!£
(C
:D
(0
to
Id
*n5>
o
cr
i
a.
c
a.
O
u
o
CO
Q
O
U
H
•§
H
O
W
O 4J
W -H
O< O
•H-H
w m
c
-------
TABLE 17. POLLUTANT LOAD RANKING OF THE FOUR STUDY
AREA WATERSHEDS
RANK
2 3
ss
ECOD
3COD
soc
NO3
NH3
TKN
TP
Area
Upper CL*
1" Value
Lower CL
Area
Upper Cl
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
Area
Upper CL
1" Value
Lower CL
P30
61.59
43.48
25.36
HB
25.62
18.88
12.14
P10
-10.01
9.86
9.70
P10
5.045
5.00
4.95
WB
0.12
0.088
0.057
HB
0.58
0.48
0.38
HB
1.049
0.95
0.85
HB
0.40
0.28
0.16
HB
42.30
37.73
33.15
P10
14.13
13.50
1-2.87
P30
9.27
8.76
8.25
P30
4.78
4.54
4.30
HB
0.10
0.087
0.072
WB
1.28
0.069
-1.14
WB
1.50
0.40
-0.76
WB
0.90
0.24
-0.43
WB
88.02
13.66
-60.70
P30
12.72
12.09
11.46
HB
6.64
4.44
2.24
WB
4.48
3.75
2.66
P30
0.038
0.020
0.0031
P30
0.037
0.031
0.025
P30
0.36
0.30
0.24
P30
0.028
0.021
0.014
P10
10.16
8.19
6.21
WB
33.13
9.48
-14.17
WB
31.56
4.06
-23.45
HB
2.64
0.70
-1.25
P10
0.020
0.012
0.0051
P10
0 .020
0.018
0 .016
P10
0.38
0.28
0.18
P10
0 .017
0 .014
0 .011
* 95% Confidence Level. All confidence
gression values at 1 inch (2,. 54 cm) of
in Ib/ac.
P30 = Woodlands P30 Watershed
P10 = Woodlands P10 Watershed
HB = Hunting Bayou Watershed
WB- = Westbury Watershed
Note: inch x 2.54 = cm, Ib/ac x .184 = kg/ha
levels are for mean re-
runoff. All loads are
73
-------
The load-runoff relations developed from several storm
events can be extended in a useful way to estimate total annual
loads for selected pollutants. A measured or predicted annual
streamflow hydrograph is required along with average low-flow
concentration values. During storm events, the load-runoff re-
lation is used to predict the mass flow, while during intermit-
tent low flows, mass flow is estimated by the product of stream-
flow and concentration.
A comparison of annual loads for TSS, TP, NC>3 and total COD
is shown in Table 18 for the P-10 forested site and P-30 urbanizing
site at The Woodlands. The developed load-runoff relations were
used to calculate the storm generated mass flows. The urbanizing
watershed appears to be contributing greater loads of TSS on an
.annual basis compared to the forested site.
The procedure allows direct comparison of storm generated
pollutant loads from non-point sources with the low-flow contri-
butions, which are primarily of natural background or point source
origin. Consequently, non-point loads can be quantitatively deter-
mined as a function of land use patterns as more storm data becomes
available from other urbanizing watersheds.
The annual load calculation can be used in conjunction with
the U.S. Geological Survey grab sample method to calculate annual
sediment loads. Relative accuracy of the two techniques remains
undetermined.
Effects of Land Development on Runoff Quality—
Storm event #10—Stormwater quality monitored at site P-10
represents runoff from a forested, undeveloped watershed and ac-
cordingly serves as a baseline for assessing changes due to
urbanization. The P-30 sampling site, located 6.8 miles (11 km)
below P-10, monitors runoff from 5,500 acreas (2250 ha), in addi-
tion to the area monitored by the P-10 site. The additonal area
includes construction activity of The Woodlands Developed Corpora-
tion. A comparison of these two sites during storm event #10
illustrates the effects of construction activity on runoff quality.
Heavy rainfall over the Panther Branch watershed on April 8,
1975 produced large amounts of runoff sampled at P-10 and P-30.
Precipitation associated with the storm event began 'shortly after
midnight on April 8, 1975 and continued till noon the same day.
The storm featured 3 periods of intense rainfall at 4:30, 8:30
and 10:00 A.M. with interposing pauses of drizzle. The area rain
gages measured 2.00, 2.65, 3.42 and 3.97 in. (5.08, 6.73, 8.64
and 10.1 cm) of rainfall, upper to lower watershed gages, respec-
tively, with the Thiessen adjusted rainfall calculated to be
2.43 in. (6.17 cm) on the P-10 watershed, and 2.76 in. (7.01 cm)
on the P-30 watershed. Average rainfall intensity was 0.76 in/hr
(1.83 cm/hr) and antecedent soil moisture conditions were dry with
no rain recorded 7 days prior to the storm and 2.5 in. (6.35 cm)
the preceding month. Watershed runoff began after midnight April
8' 74
-------
VD
CO
in
(0
oo
II to
\
rQ' pQ
H <-{
rH H
75
-------
and ended three days later. As shown below, the volume of runoff
observed at site P-30 was greater than P-10 but peak discharge
was essentially the same.
Site
P-10
P-30
Streamflow Volume
acre ft. ha-m
1614
2829
(199)
(349)
•Runoff
in. (cm)
1 . 2
1.57
(3.05)
(4)
Peak Plow
cfs (m3/s)
1170
1100
(33.1)
(31.1)
Runoff
Coeff .
50%
57%
The greater runoff volume for P-30, almost twice that of P-10,
was a result of three factors: (1) larger drainage area, (2)
heavier rainfall in the lower basin, and (3) impervious areas in
The Woodlands development.
Figures 21 and 22 compare P-10 and P-30 pollutographs for TSS,
total COD, TP and TKN. Hydrographs are also presented in the
figures for flow rate and time references. Pollutograph analysis
should consider the following:
1. Those areas of The Woodlands developed or under
construction encompassed only 10% of the total
watershed. The majority of stormwater runoff
originated in undeveloped forest lands.
2. Developed or construction areas were located
adjacent to P-30, as shown in Figure 23. As
a result, runoff originating in these areas was
observed early in the storm event.
The pollutographs indicate high TSS loads at P-30 (Figure 21),
a result of sediments washed from easily eroded construction sites.
Although the developing area comprised 10% of the watershed, it
contributed as much as 80% of the TSS load at P-30. Total COD
(Figure 21) exhibited a "first flush" at P-30, probably due to asso-
ciated high TSS. The major portion (70%) of the total COD was
soluble. Significant increases in TKN and TP at P-30 resulted
from wash-off of ammoniated phosphate and urea based fertilizers
applied to the golf course in the developing area (Figure 22).
Development within the Panther Branch watershed has resulted
in stormwater runoff quality changes. TSS and nutrients have in-
creased although no significant change has been observed for
oxygen demand. Results for storm event #10 are summarized in
Table 19.
76
-------
656 TONS
84 TONS
l i
a
o
o
O
ixi
S
-------
0.4.
_j 0.3.
•^
s 0.2.
O.I .
0.0
675 LBS
264 LBS
z
t-
4.0.
3.0.
2.0.
J.OJ
0.0
til
O
£E
<
o
en
5
en
8
UJ
I
I
120.
2 90 J
x
CO
u.
o
60.
3O.
1065 LBS
6OO LBS
2829 AC-FT
1614 AC-FT
15 3O 45 6O 75 9O
HOURS
P-30
15 30 45 60 75 90
HOURS
P-IO
Figure 22. Comparison of P-10 and P-30 temporal distribution
of streamflow, TKN and TP for the storm event of
April 8, 1975.
1 Ib. = .4536 kg
1 ac-ft = .123 ha-m
1 cfs = .028 m /sec
78
-------
CONSTRUCTION AREAS
PROPERTY BOUNDARIES
1 mi = 1.6 km
1 ft = .305 km
FEET
Figure 23,
The Woodlands construction activity in
relation to the P-10 and P-30 sampling
sites.
79
-------
TABLE 19. COMPARISON OF STORMWATER QUALITY AT P-10, P-30
AND DEVELOPING AREAS DURING STORM #10
Forest
with
Development,
P-30
The Woodlands
Forest Development
P-10 (P-30) - (P-10)
Drainage Area (acres) 21,606
(ha) 8,750
Streamflow Volume (ac-ft) 2,829
(ha-m) 349
16,050
6,SCO
1,614
199
5,556
2,250
1,215
150
Average Concentration of Water Quality Parameters (mg/1)
Ortho-phosphate
Total Phosphorous
Ammonia
Nitrite
Nitrate
Total Kjeldahl Nitrogen
Total Suspended Solids
Soluble Organic Carbon
Total COD
Soluble COD
0.011
0.088
0.15
0.009
0.154
1.39
171
20.4
52.6
38.8
0.003
0.06
0.08
0.004
0.065
1.37
38.5
22.0
58.7
43.0
0.021
0.125
0.244
0.017
0.272
1.41
347
18.2
44.5
33.2
FC/FS bacteriological ratios and land use--An important com-
parison employed in examining data is the relationship between
FC and FS concentrations. A FC/FS ratio > 4 suggests the present
of human wastes. Between 2 and 4, the FC/FS ra.tio may suggest
human wastes mixed with other source materials, and 1 < FC/FS <
2 value represents an area of uncertainty. If values fall be-
tween 0.7 and 1.0, a predominance of livestock or warm-blooded
animal waste may be suggested. Following the latter range, FC/
FS values less than or equal to 0.7 strongly suggest a predomi-
nance of warm-blooded animal waste other than human wastes.
Mean FC/FS ratios for storm events, low flow and sewage de-
terminations are plotted in Figure 24. Note that FC and FS
bacterial numbers in The Woodlands stormwaters are greater than
low-flow waters, however the ratio remains the same. Urban run-
off at Westbury also exhibited a similar ratio although the
bacterial numbers were much greater. High ratios for the sewage
determinations confirm human waste contamination. Lake B runoff
contained the lowest ratios indicating stormwater impoundment
may have beneficial effects or that organisms attached to sus-
pended sediment. TC, FC, FS and ST bacterial species were all
observed to settle in quiescent water. Additional data have been
reported by Olivieri, et al., (44).
80
-------
81
-------
A least significant difference statistical analysis was ap-
plied to FC/FS ratios from all stations resulting in the follow-
ing sequence:
Location FG/FS (Geo. Mean)
Storms, Lake A
Storms, Lake B
Low Flow, Lake A
Chlorinated Secondary Sewage
Low Flow, P-10
Storms, P-30
Low Flow, Lake B
Low Flow, P-30
Storms, Westbury
Low Flow, Woodlands
Storms, P-10
Raw Sewage
Secondary Treated Sewage
0.17
0.53
0.58
0.73
0.92
0.97
0.99
1.26
1.47
1.68
2.11
2.42
13.30
Those groups which demonstrated no significant dif-
ferences between subsets are indicated by the verti-
cal lines on the side.
STORMWATER QUALITY MODELING
Several techniques are available for the prediction of water
quality responses in a watershed. The SWMM model has been
adapted for natural drainage conditions at The Woodlands in an
effort to simulate stormwater quality response. The model oper-
ates from a relation between runoff rate and pollutant load, but
the prediction of hydrographs. has been more successful than pol-
lutant response. The water quality procedure in the model is not
designed to simulate the response from natural drainage, and has
been updated for The Woodlands. New relationships between cumu-
lative load (Ib) and cumulative runoff volume (ft-*) have been
incorporated into SWMM for various parameters at The Woodlands.
In this way, concentrations can be predicted as a function of
runoff (45). Another water quality modeling approach is dis-
cussed below.
pollutant Load Modeling for Multiple Events
The load-runoff relationships presented above (Figures 17-
20) provide the'foundation for an uncomplicated, yet satisfac-
torvT model for runoff pollutant load simulation of multiple or
individual storm events. Given time increment values for runoff,
the model consults time-varying load-runoff relationships to
calculate mass flows during storm events.
82
-------
Variation in the average pollutant, concentration over time
is approximated by variation of the load-runoff line slopes
(Figures 17-20). These slopes represent the ratio of mass of
an.
kech watershed. Initially three parameters are defined for each
load-runoff relationship: the average slope, the initial slope, and
a factor which sets the range within which the slope can vary. The
average elope can be roughly determined from the cumulative rela-
tionship produced from field data. The initial slope value depends
primarily on initial conditions, and the range variable is deter-
mined by the spread in observed pollutant concentrations.
During dry periods the slopes are incrementally increased up
to but not above the pre-defined maximum. This corresponds to
the buildup of available pollutants on a watershed between storm
events. An increment chosen to increase the slopes is required as
input and is obtained primarily by calibration. During a storm
event the value of the slope decays exponentially by the same
means employed in both the SWMM (46) and the Storage, Treament, and
Overflow Model '(STORM, (47).
libs pollutant
washed off in
any time interval
Lbs
remaining on
the ground
or:
-dP
dt
= kP
(1)
which when integrated takes the form:
P0-P = PQ (l-e-kt)
(2)
Where P -P = Ibs washed away in time, t, and k is assumed to
vary in°direct proportion to the rate of runoff, r:
k = br
(3)
b can be evaluated given the assumption that 0.5 in. (!•.27 cm)
of runoff uniformly delivered in 1 hour washes away 90% of the
pollutants (22). As a. result the equation can be written:
(4)
83
-------
The equation used to decay the load-runoff line slopes is:
-4.6 rt
PDS = 1-e
(PDS),
(5)
Where PDS is the load-runoff curve slope at some point in time
during the storm event and (PDS) is the initial value.
A six month period of streamflow at site P-30 was chosen for
sequential simulation. This period dating from October 28, 1974
to April 8, 1975 includes storm events 5, 1 and 10 monitored
during the study. Storm events 8 and 9 were considered too small
for use in the simulation. Predicted solids loads and the ob-
served streamflow hydrographs are presented in Figure 25. Slope
parameters used were derived from the load-runoff relationship,
with the upper limit values found by trial and error.
Simulation results can be evaluated by comparing observed
and simulated mass flows for individual storm events. As shown
in Figure 26, the simulated curves compare satisfactorily to the
observed mass flow curves. Table 20 gives comparisons of simu-
lated to observed values for total pounds TSS, and peak magnitudes
for each of the three storms.
t
Unit Loadoqraph for Single Event Simulation
The form of the stormwater mass flow curves, obtained from
the product of instantaneous concentration and discharge, re-
semble the general shape of the streamflow hydrograph and provide
a more useful measure of runoff loadings than the concentration
curves. For a given watershed, it is possible to generate a unit
hydrograph based on the incomplete gamma distribution. Using the
theory of linear reservoirs, the resulting equation for the unit
hydrograph becomes (48)
Qn = kr(n)
(I) ^
exp(-tA)
(6)
where S = total storage (one inch (2. 54 cm)of runoff); k = constant;
n = outflow from n^*1 reservoir. The watershed is considered as n
serially arranged linear reservoirs, and it is possible to fit an
observed hydrograph by varying k and n.
The theory of linear reservoirs can be extended for mass
flow curves in order to develop a corresponding unit pollutoqraph
or unit loadograph for application to urban storm runoff. In
this way, mass flow curves can be generated in a similar fashion
to the hydrograph by varying appropriate constants. The water-
shed is considered as n serially arranged tanks with first order
decay, and the mass balance becomes
84
-------
S/9
sanos aaaN3dsns ivioi
SdO
Sn03NVlNVlSNI
85
-------
to
0) 0)
en
bO
o
Tf I
CJ
10 CJ
OI X S/9
sanos
TYIOI
10 ea —
?OI X SJO
SnO3NVlNVlSNI
10 O
f) 4J
O CO
fi
CO
•a 8
fd m
s-i
en to
o a)
M >
VO
OJ
<0
%
O
-------
g
CQ
04 Pi
CQ H
CQ H!
O CM
H
Q 05
S E^
CQ
EH vo
^*
CQ
< EH
.«
o 0
CQ
(tf
CO
eu
H
«
£4
^
1
^|
4->
rt
3
S
CQ
S CD
i-i
•P 3
CQ a
«5j<
in
o
o
in
CM
ro
0
o
vo
CM
T^ |
CM
in
iH
0
0
CM
m
ro
r-l
0
O
r-4
*3f
O
CM
CM
0
0
crv
CO
f^
CM
0
O
CT>
ro
in
. •
CO
r-4
in
f**.
r««
0
CO
00
vo
o
VO
CM
Crt
0
0
CM
CM
VO
CM
O
in
ro
CM
in
r-i
VO
0^
CM
O
^j1
ro
^
o
vo
o
&
ro
o
in
r^
CO
t^
o
cr>
ro
o
in
GO
in
ro
CM
r-l
ro
rH
ro
in
00
0
VO
ro
in
II
£1
H
H
W
•4J
O
c
CO
0) CM
d o
•H rH
05 ^
0) C
e o
•H -H
-P -P
(d
0) W
m o
cu tn
tl *"H
•H
10 C
•H G
-H
M CD
87
-------
~dt~
(7)
where MI = total mass; go = mass inflow; g1 = mass outflow;
ki = linear decay coefficient. By solving the equation for n reser-
voirs in series, assuming outflow from one is inflow to the next,
and assuming M = kg (analogous to S = kQ), the final resulting
equation for the unit loadograph, .de-fined as the ::nass flow in kg/sec
plotted vs time, becomes
M
gn =
(8)
The similarities of equation 8 and equation 6 are obvious, where
gn is mass flow (kg/sec), a equals (1 + kk1)/k, cind M is total
mass.
Hydrograph and mass flow simulations for Storm 10 on the
P-10 site are shown in Figure 27. The timing of the hydrograph
peak and the total volume compare well, but the recession rate
is predicted lower than the observed. The simulation of TSS and
TP mass flow curves (g/sec) yielded similar results, with good
peak and total mass definition, but a predicted recession rate
lower than observed. This storm yielded 1.2 in (3.05 cm) _of runoff,
and the unit response can be ovtained by dividing all ordinates by
1.2.
The application of this approach is in a preliminary stage
due to lack of significant storm runoff data (1 inch, 2.54 cm, or
greater) on Hunting Bayou of The Woodlands watersheds. As more
storm event data are collected from other watersheds in the area, it
will be possible to investigate relationships between the gamma
distribution shape parameters (n, k) and land USB or physiographic
factors in the watershed. In general, the time of peak of the unit
loadograph is related to n and k by the equation
t =
(9)
Urbanizing watersheds should have lower values of n and k than
undeveloped watersheds of the same size. As n is increased, k
must be correspondingly decreased in order to yield the same tp
value for a given watershed.
The unit loadograph can be used in the same manner as the
unit hydrograph. Once the unit response has been determined for
a watershed, storms of varying intensity can be analyzed by lagr
ging and superposition of the unit graphs. A unit pollutograph
(concentration vs time) is found by dividing the ordinates of the
88
-------
Ul
o
a:
< •
o
CD CM
5 g
°
01
2
12
9.
3.
HYDROGRAPH
tt - 6
K = .13
o
(O
(O
2.4
1.8.
1.2.
0.6.
0.0.
20 28 36 44 52 60 68
HOURS INTO STORM
OBSERVED FITTED
Figure 27.
Fitted curves for storm runoff and pollutant
mass flows observed at P-10 on 4/8/75.
1 cfs = .028 m3/sec
89
-------
unit loadograph by corresponding hydrograph flows. Because of
the linear load-runoff relationships which have been developed
tne linear assumption of unit response is further justified '
load°9raph approach suffers the same limitations as
hYfr°g^a?h1 method with re^rd to assumptions of uniform
and initial conditions, but it does offer a relatively
simple and useful technique for analyzing stormwater pollution
response as a function of land use, watershed characteristics
and hydrologic conditions.
Storm Water Management Model
SWMM is composed of five integral computation blocks as
shown in Figure 28. The Executive Block controls all activity
within the model because all input-output functions for other
blocks are programmed into the Executive Block. The Runoff
Block computes quantity and quality of runoff for a given storm
and _ stores results in the form of hydrographs and pollutographs
i*-J?«?t?1t° th| ?ain sewer system. The Transport Block sets up
initial flow and infiltration conditions and performs flow
quantity and quality routing to produce combined flow hydrographs
and pollutographs for the total drainage basin and at selected
^Jfm;diate PO1^3' Quantitv a^d quality of flow are stored and
SSS* ? Si6*?"1?3 criteria in the Storage Block. Diapers ioT
effects of the discharge in receiving waters are computed in the
JTXr-S* ^M^ B^°Ck- A m°re detailed description is available
in the User Manual - Volume in (46).
4.1, ~In ^eneral only one or two computational blocks, as well as
the Executive Block, are used in a run. However, all blocks may
be run together. The use of independent computation blocks
allows for examination of intermediate results. The necessity
for at least 35 OK bytes of core storage in SWMM leads to hiqh
run costs and limits the number of options to be analyzed.
The SWMM release of February 1975, referred to in this re-
port as the original SWMM version, was extensively modified The
capabilities of the modified version have been expanded to model
runoff and water quality from natural drainage areas. Study
Model Application —
Specific data required as input to the original SWMM are de-
scribed in Table 21. A quantified description of the watershed
provides a computational basis for the model and includes the
rainfall hyetograph for the storm to be modeled, a physical de-
scription of each subcatchment to be modeled including the drain-
age area, percent of impervious cover, ground slope, Manning's
roughness factors, estimated retention storage for both the per-
vious and impervious surfaces, and the coefficients to define
90
-------
g
o
ai
B
X
-Cp
u «
K O
o3
t-eo
._. r-^ §s
"« " VV^0"
vt c
1 s
lo£
(T2O
is
00
CNJ
0)
I
xo
So
91
-------
TABLE 21. MODELING REQUIREMENTS BY
Item 3.
Item 4.
Item 1. Define the Study Area
Land use, topography,
tract data, aerial photos, area boundaries.
Item 2. Define the System
tions of inlet structures.
Define System Specialties
Flow diversions, regulators, storage basins,
Define System Maintenance
Street sweeping (description and frequency). Catch-
basin cleaning. Trouble spots (flooding).
Item 5. Define the Receiving Waters
General description (estuary, river, or lak «>
sured data (flow, tides, topography, water quality).
Item 6. Define the Base Flow (DWF)
Measured directly or through sewerage facility °Pfr-
a?ing data Hourly variation and weekday vs. weekend.
^characteristics (composited BOD and TSSr results >.
Industrial flows (locations, average quantities,
quality) .
Item 7. Define the Storm Flow
nailv rainfall totals over an extended period (6 months
SrionqSr) encompassing the study events. Continuous
'
92
-------
Horton's soil infiltration equation. Input data also define
hydraulics for the storm sewer system in each subcatchment and
the main sewers or open channels in terms of gutter length,
slope, bottom width, and roughness coefficient, cross-sectional
area, side slopes, channel slope, and roughness factor. For
water quality modeling, a code defining the specific land use in
each subcatchment as well as the street-cleaning frequency, the
number of dry days prior to the storm event, the number of catch-
basins per unit area and the quality of their contents must also
be specified.
Horton's infiltration equation is used to calculate the in-
filtration rate of rainfall into the soil as a function of time
by Horton's relationship (49).
Manning's roughness coefficients were necessary for each
drainage element to describe the hydraulics of the drainage sys-
tem. Gutters and open channels were assigned an initial value
of 0.10, while a value of 0.03 was used for sewers. These
values were adjusted during model calibration to accommodate
higher peak flows. Combined sewers are not used in any of the
study areas and therefore all initial flows were zero.
A 10 minute time interval was the limit for modeling ac-
curacy at minimum cost, and all SWMM runs were made with regard
to this condition.
Hunting Bayou Modeling—
Input data for the Hunting Bayou drainage system were ob-
tained from existing engineering maps and site inspection. The
subcatchments and drainage network used as input data are shown
in Figure 29. The total drainage area of 1976.8 acres (499.86 ha)
is divided into 24 subcatchments ranging in area from 25 acres
(10.11 ha) to 138 acres (55.85 ha). Each subcatchment was
assigned a land use class for modeling water quality in SWMM.
The drainage system includes 23 gutters and pipes in the Runoff
Block and 44 manholes and conduits in the Transport Block.
Infiltration rates were originally estimated and then calibrated
by consecutive modeling runs.
Five storms were modeled initially. The rainfall data for
these storms were obtained from reports published by the U.S.
Geological Survey. All five storms occurred during 1968 and
1970, prior to the initiation of this project. Consequently,
only water quantity was modeled and no water quality data were
available. Comparisons of observed and computed hydrographs for
these five events, presented in'Table 22, indicate reasonable agree-
ment. The average absolute error in runoff volume was 26% (see
Table 22) of the observed value, while the average error in peak
flow prediction was 20% of the observed. The temporal agreement of
the hydrographs was very good. For instance, the times of peak
flow agreed within ten minutes in four of the five instances and
93
-------
LEGEND.
STUOV AREA BOUNDARY
SUBCATCHMENT DIVIDE
SEWER PIPE
OPEN DITCH
MANHOLE
MANHOLE NUMBER
1 ft = .305 m
Figure 29
•«*>*« *• ----
Subcatohments and drainage network in Hunting Bayou,
94
-------
average error was twenty-two minutes. However, computed values
tended to predict faster returns to low-flow conditions than were
actually observed.
TABLE 22. COMPARISON OP SWMM PREDICTED RESULTS
WITH OBSERVED FLOW MEASUREMENTS FOR
HUNTING BAYOU STORM EVENTS
Date of Storm
09/08/68
09/17/68
11/05/68
10/22/70
11/09/70
Total Runoff
(ft3 x 106)
Measured
4.84
9.06
4.40
12.69
1.52
Predicted
4.50
4.98
5.07
12.84
2.46
Peak Flow Rate
(cfs)
Measured
325
330
282
665
160
Predicted
303
302
337
549
205
NOTE: 1 cfs = .028 m3/sec
Three more recent Hunting Bayou storms on 3/26/74, 4/11/74
and 5/8/75 were sampled and the water quality and flow prediction
capability of the SWMM was tested. The 5/8/75 storm event will
be presented here as an example. Figure 30 shows the observed
and SWMM predicted hydrographs for this event.
Original SWMM predictive capabilities are based upon
dust and dirt accumulation data acquired in Chicago, and the
extrapolation of these data to natural drainage areas is a
limitation of the model, which resulted in poor water quality
predictions for this watershed. Consequently, a simplified
approach to water quality prediction in SWMM was developed
which does not consider pollutant buildup or input data on
dry days, street cleaning frequency, land use, or curb length.
Instead, pollutant availability loading rate at the beginning
of the storm is input. This information was produced by The
Woodlands project stormwater monitoring program. The user
determines effects of dry days, street cleaning frequency
and land use external to the model.
The new SWMM version was also run for the 5/8/75 storm event,
and the resulting predicted TSS pollutograph is shown in Figure 31
(left-hand figure). The loading rates used for this pollutograph
prediction are as follows:
95
-------
^_^
UJ
cc wj
in AT\
UJ ^4J
a
£0
!
!
«J
:
^
i
'
i
!
i
}
A
/
/
f j
i
./
r
r\i
i
i
i
j
i
i
i
i
i
^ i
N
.
'
i
\
\
\
N
^
i|
j
,
I
,
>
"S
b
^,
1
;
!
i
i
!
x 1
!
^j
1
•
!
!
•^. '
!
. 1
i
STREAM
FLOW
DISCHARGE
STATION
Ji-20
21 22 23 24 1 23456 7 8 9 10 M 12.
TIME (hrs.)
STORM OF 5/08/75
ui
o
O
o
OBSEUVIJ3
^OMPUTEl)
SUSPENDED
SOLilDS BY
ORIGINAL SWWM
VERSION
200 22 23 24 I
23 4567
TIME (hrs.)
8 9 D 1
STATION
M-20
Figure 30.
Predicted hydrographs for T55S concentratioinis
at Hunting Bayou ((5/8/75) ,
1 cfs = ,028 m3/sec
96
-------
[0 en i
£ CD -P
o -P M-I
H (d 0)
m M H
« £ is o
en
D§
g m fl
£
•d w Q
s w xi
(d fti CO
w
&
ft
in
vQ
LO
t-~
II
o\4J 's
-P co C J3
^"^ 0 ^r
H in O ;°
a^s-
0)
CD (0 CD
> « -P
Q) iJi ft
X! -H O
O 4-> O
0)
4J O
4-1 CD
CD -H
U CO (Q- (fl
•H EH CD
CD
CD
O CD
4-1 ^-1
fO
CD
M
3
tn
(l/fiui)
97
-------
Pollutograph
Prediction
TSS Loading Rate According to Land Use
m (Ib/acre)
Residential Construction Undeveloped
1-85 2.31 0.23
A removal coefficient of 21.2 min"
4.6 in the SWMM.
was used, as compared to
=T^T A_tr^l and error procedure was followed to determine the
above loading rates and removal coefficient combination that
would reproduce the observed pollutograph. Loading rate and
removal coefficients derived are valid only forttS s£rm Ssed
nd aPPlic*tion of results to othe? stSSL Is
Prevailing antecedent conditions and raXfall-
™ •s are similar for both storms and if study areas
are identical or at least homogeneous. ^uay areas
Although simulated pollutographs were curve-fit to
observed pollutographs, actual andcomputed masl ^ans^
*X2 not correspond. As shown in Figure 31 (right-hand fiou
SSFVi11^ P°°r reSUltS WSre outlined usingthe product o?'the
thJf Sn^^ograph -and pollutograph. it wa2 determined that
this condition resulted from a failure to properly predict the
timing of the peak of concentration. P-^UJ-C-C -cne
4-a«4- 5° improve the modeling of total mass loading of a pollu-
tant the new version was used to model pollutant mass flow rates
Again the loading rate and decay factor were adjusted to repro-
duce mass flow rates, using the following values:
TSS Loading Rate According
to Land Use (Ib/acre)
Mass Flow
Prediction
Residential
4.0
Construction
5.0
Undeveloped
0.5
A removal coefficient of 35 min-1 was used for this case.
Results, shown in Figure 31 (right-hand figure), indicate mass
flow rates can be accurately predicted using this method (dashed
line). The method was applied to other stormwater quality para-
meters, COD, N03, TP, with similar results.
Panther Branch Modeling—
The data input to the SWMM for the Panther Branch drainage
system was developed from existing engineering maps and numerous
site inspections of the watershed. The total drainage area for
Panther Branch of 21607 acres (874.40 ha) was divided into
57 subcatchments ranging from 21 acres (815 ha) to 1366 acres
(552.80 ha). The input parameter called "width ©f sub-
catchment" is defined as the width over which overland flow oc-
98
-------
curs. Values for this parameter were f-irst estimated by the
method described in the SWMM User's Manual (46). These values
were subsequently reduced by approximately 40% to achieve cali-
bration. The Panther Branch drainage system is made up of 57
"gutters" and 61 transport elements of varying characteristics.
A major drawback of SWMM at The Woodlands is that the area
below P-10 was in a transient state due to development. Con-
tinually changing land use affects the quality of runoff and
consequently P-10 is regarded as a control point. The area above
this gage is in a relatively stable condition and will give a
more accurate measurement of the pollutant loading due to the
land use rather than from a construction area. Stormwater runoff
from a construction area can vary in quality from storm to storm
depending on the stage of construction, and modeling proved diffi-
cult. Consequently, it is presumed that several construction
areas where the natural ground had been disturbed and stripped
of the protective vegetative cover contributed more TSS than SWMM
could predict from available input data.
Five storm events on Panther Branch have been modeled.
Similar to the data for Hunting Bayou, all subcatchment, gutter
and transport element data for Panther Branch were identical for
all runs. Infiltration rates were determined similar to
Hunting Bayou. The original SWMM was used to model both water
quantity and quality for two storm events on 10/28/74 and 12/5/74
and only quantity of flow for the remaining events (1/18/74,
4/8/75 and 11/24/75). Computed flow peaks and volumes agreed well
with the observed flows; the average absolute error in the volume
of runoff between observed and computed hydrographs was good
except for the storm events of 10/20/74 and 11/25/74 when the flow-
peaks between observed and computed hydrographs were approximately
three hours apart. Water quality modeling at P-30 was not
acceptable using the original SWMM. The 10/28/74 storm event at
P-30 was a multi-peaked hydrograph, and TSS modeling by the ori-
ginal SWMM was not accurate. The maximum observed TSS concentra-
tion of 1000 mg/1 during the second peak in streamflow was com-
puted to be 273 mg/1, a much lower concentration. Runoff quality
modeling of the 12/5/74 storm event was similarly too low in
concentration and total load.
The 12/5/74 storm event was modeled for both the upstream
gage, P-10, and the downstream gage, P-30. Since the entire
drainage area had the same land use before development began, most
differences between the upstream gage and the downstream gage can
be attributed to the changing land use in the developing area.
Using the original SWMM version, the computed peak concentration
of 142 mg/1 TSS at Station P-10 was in good temporal agreement with
the observed value of 130 mg/1. The falling limb of the observed
pollutograph occurred more rapidly than the simulation pollutograph,
resulting in'a difference of 11,330 kg (25,000 Ib) or a 40% error
99
-------
in computed total load. Observed TSS production at Station P-10
was about one-third that at Station P-30.
The modified water quality version of SWMM was also used to
model the 12/5/74 storm on Panther Branch for TSS, COD, N03 and
TP parameters. The pollutographs and mass flow curves were sepa-
rately computed using loading rates obtained by trial and error.
Simulation results are summarized in Table 23. Optimized polluto-
graph and mass flow curves corresponded well with the observed
data. However, mass flow rates calculated from the optimized pol-
lutographs were not as accurate as computed optimized mass flow
rates. TSS predictions at Station P-30 were compatible to observed
data with slight differences for occurrence of peak flow rates.
Modeling of TP at both P-10 and P-30 and NO3 at P-30 not entirely
satisfactory.
Swale 8 Modeling, Existing and Future Development—
Existing drainage and planning maps were used to develop the
input data for Swale 8, the watershed above Lake Harrison. Site
inspections to determine drainage area boundaries and extent of
construction were conducted on a periodic basis because this
watershed was in a transitional stage. During the project, the
channel was enlarged and construction of Lake C was underway
whiles Lakes A and B had already been filled.
The total drainage area for Swale 8 of 459.3 acres (185.87
hectares) was divided into 10 subcatchments ranging from 23 acres
(9.30 ha) to 66 acres (26.71 ha). Land uses, for the upstream
subcatchments were classified as open space, whereas the last
three downstream subcatchments were designated as multi-family,
residential and commercial. Seventeen drainage system elements
were used to model the entire area. Of these, two elements were
storage units, Lakes A and B, and all six channels were trape-
zoidal in shape as a result of channel enlargement.
Swale 8 storm even on 4/8/75 was modeled because the only
other observed storm event, 3/13/75, had a peak inflow into Lake
B of 0.06 m3/sec (2.0 cfs) from 0.81 in. (2.06 cm) of rainfall.
The transitional phase of development in Swale 8 gave rise
to several problems in modeling runoff. The most severe problem
is the lack of lake volume data. The topographic maps prior to
lake construction show the natural ground contours, but the
reservoir areas were used as borrow pits for fill material for
the dams as well as other construction at The Woodlands. Conse-
quently, the original storage capacity of the reservoirs was not
known and no subsequent reservoir surveys have been conducted;
therefore, the elevation-area-capacity data for these lakes were
100
-------
CO
&
w
ro
ro
o
H
X
•* Sf
CM CM
8
CM
CO
rt
0
K
y
W
vo
CO
«n <-4
OA
CN
«s
5
• • •
O\ i-l
CO r-l
«
ro
o
w
o
• •
g
CGCN
•a
en CM
° ' **
5
in
in
«s a
CO s^»
VO CM
o in
«d"
CO
CN CN
8
fa
8°
H
1-3
H
Q
fe
H
B
(4
H
EH
CM
H
J
1
« w
•y -««
ro
O
0) 03
w Q ^ w
CM r-l
en co
o
cr»
CM
NO
Is*
CM
0 0
vo
1 II
CO Q CM 'rj-
CO O O O
En CD 52; a,
V0
vo
i u
CO Q CM T
co o o o
t-l u ^ a,
O
(U
n
tn
in
G
•H
H
H
13
o
101
-------
only approximate. Also, groundwater was being pumped in Lake A
and pumpage rate was not recorded.
The outflow structure for Lake A is controlled by different
outlets at different water surface elevations. The outflow rat-
ing curve (discharge as a function of water surface elevation)
is composed of three segments, one controlled by the low-flow
orifice, the second controlled by weir flow through the flood
discharge outlet which in turn is limited at extreme flows by
the capacity of the outfall conduit and resulting in the third
segment of the rating curve. The SWMM is not capable of model-
ing this complex outflow scheme.
Under the conditions described above, the modeling of runoff
storage in the lakes proved to be difficult. Several attempts to
model the outflow from Lake A for the storm of 4/8/75 were un-
successful. The extent of assumed data was too large in magni-
tude to approximate the correct operation of Lakes A and B. As
a result, additional modeling of the watershed was conducted
only on that drainage area of Swale 8 upstream from Lake Harrison
at point £>-10 (Lake B gaging station) . The results of this
modeling effort are discussed in the following paragraphs.
Due to various external influences, urban development at The
Woodlands did not proceed as rapidly as expected. Site develop-
ment plans were available for Phase I, and in early 1976 a major
portion of the Swale 8 watershed was being planned for develop-
ment. Using these plans for the watershed, three development
scenarios were evaluated for modeling: (1) existing conditions,
(2) immediately developing conditions, and (3) future but not
ultimate conditions.
i
Water quality predictions by the modified version of SWMM
were attempted. Changes in land use and increase in impervious-
ness were computed from plat maps provided by The Woodlands De-
velopment Corporation and input to the SWMM. As described
earlier, the modified quality prediction version required the
input of loading rates for each pollutant. The initial loading
rates used were derived from the p-30 watershed modeling experi-
ence. Based on previously described experience with pollutograph
differences resulting from computed hydrographs, it was decided
that only mass, flow rates would be modeled. Runoff from the
storm on 4/8/75 was chosen for modeling; however, due to its
multi-peak complexity, only the first hydrograph peak was model-
ed. Results from the first computer run indicated that the load-
ing rates determined from the results of modeling at Station P-30
were too low. Observed peak mass flowof.TSS was three times the
peak mass flow computed from loading rates derived at Station
P-30. These differences are a result of the extreme effects of
lake and golf course construction, as well as channel improvement
concentrated in the Swale 8 watershed. Also the freshly sodded,
102
-------
EH
W
W
(0
H O 0 H
IT) en po CM
OCM
co co
OH
O CT» CTi (31
cr^ n
CM in
in co
vo
CO CM
VO CO
iH CM
CM
in CM VD H
i—1 VD iH CM
e
CO
J2;
8
i
en co
en ro
o H
• • t i
o in
00 CO
o
iH
CQ
o
£
H
8
§
ft
M
*""l!
p
• «
en oo
00 VD
CM
H
&
(0 CU
M CQ
CO P 4J O
CO O -H ,£)
EH U J3 ft
CQ
-------
developed areas severely eroded during the intense rainfall of
tnis period. Loading rates were revised for both construction
and developed areas. For Swale 8 the TSS loading rates from de-
veloped areas were 82% of the rate from construction areas. In
contrast, the same ratio at Station P-30 was 78%.
The modified version of SWMM was run for two future water-
shed development conditions described earlier, using the rainfall
on 4/8/75 to provide a basis for comparison between existing and
future conditions. AS anticipated, the larger proportion of area
under construction changes the pollutant loads: considerably; the
changes range from an increase of 77% for TSS to a decrease of 8%
for nitrates. '
After the construction phase of development has been com-
pleted, peak pollutant loads do not decrease as may be expected
but_the total mass of pollutant do decrease. These dramatic '
environmental effects of construction activities are listed in
"?abf^ 24C One reason for the increase in peak mass flow rates
is the change in the runoff hydrograph. After construction the
hydrograph peak is increased by approximately 40%. Another
reason is the increase in the input loading rates for developed
areas, which results in a doubling of peak mass flow rates for
the parameters N03, COD and TP. The 20% increase in the TSS
mass flow rate is a result of hydrograph modification due to
urbanization.
In summary, the modified water quality modeling version
greatly improved the capabilities of the SWMM. Water quality
modeling results are much more dependable, and observed events
can be adequately simulated. Each of the stores used to test
the new SWMM version was selected to present a range of flow
water quality and land use data; thus, the modal was tested over
a range of different conditions.
STORMWATER ALTERATIONS AT THE WOODLANDS
Water Quality Needs
Irrigation—
t Collected stormwaters are to be used for golf course irri-
gation at The Woodlands to supplement natural precipitation The
critical water quality parameter for irrigation is salinity. Ex-
cessive salinity affects plants by increasing osmotic pressure in
the soil which limits uptake of water by plants). However, this
is not the case at The Woodlands and, therefore, salinity will
not be a problem. Electrical conductivity measurements in storm-
water runoff is less than 3000 micromhos at Th<> Woodlands and is
excellent to good for most plants" (50). The presence of nutri-
ents in stormwaters slated for irrigation is not high and in this
case considered an asset rather than a pollutant. TSS concentra-
tion or large particulates could cause mechanical problems such
104
-------
as pump damage or clogging of sprinkler heads, but careful
placement of the intake structure will avoid these difficul-
ties. TSS concentration in Lake Harrison during low-flow
conditions is about 100 mg/1 and average particulate size is
estimated between 25 and 250 mg/1, levels acceptable for
pumping requirements. The velocity in the distribution system
will keep the solids in suspension. Particulate size criteria
will be set by the orifice size of the irrigation system.
Aesthetics—
The water quality level for aesthetics is presently met without
any stormwater treatment. The lake is devoid of floating debris or
objectionable odors and promises to support a wide variety of life-
forms. Superficially, it resembles early stages of other local
man-made lakes. High nutrient levels may promote macrophyte growth
and algal blooms, but macrophytes can be controlled by a regular
lake maintenance program and algal blooms can be prevented by
reducing the lake detention times and nutrient levels (40).
Recreation—
In a discussion of recreational water uses, two divisions must
be considered: contact and noncontact. The water quality require-
ment for contact recreation, which involves substantial risk of
ingestion, is more stringent than that of noncontact (51). Swimming
is the primary example of contact recreation and is prohibited in
The Woodlands' lakes.
A water quality criteria designed strictly for boating would
be similar to that for aesthetics, with the added requirement for
fecal coliform levels. It is recommended that fecal coliform
levels of 2000/100 ml average and 4000/100 ml maximum be observed
for "unofficial recreation" waters. Levels of 1000/100 ml average
arid 2000/100 ml maximum were suggested for official noncontact
waters (52). Fishing water criteria invoke an additional require-
ment that harvested species be fit for human consumption. Edible
fish species should be free of toxic chemicals and pathogenic bac-
teria or viruses. The data to determine the fulfillment of this
requirement is not presently available. The consumption of fish
from such waters has been practiced without harmful results.
Coliforms in the digestive systems of fish caught at Woodlands are
in higher concentrations than from other lakes but presumably do
not reach the edible portions of the fish (52).
Water Supply Uses—
Lake Harrison ranks as a poor raw drinking water source
because it would require a high level of treatment before use (50).
With groundwater, a less expensive and more reliable source is
easily obtained. Its use for this function is to be restricted to
emergencies.
105
-------
The Lake System
The man-made lakes at The Woodlands will serve as recrea-
tional centers, wildlife preserves and storage for stormwater
runoff. The lakes will contribute to the maintenance of a
perched water table necessary for plant life. The lake water
will also be used for irrigation of adjacent golf courses. In-
puts to the lake system are limited to stormwater runoff and
treated wastewater from The Woodlands Wastewater Reclamation
Plant with ultimate capacity of 6 mgd (..26 m3/sec>. Eor a
year of average rainfall, treated wastewater will comprise 75% of
the flow through the lake system. Design effluent quality for
the plant is indicated in Table 15 and suggests that the lakes
will contain clear waters with somewhat elevated nutrient con-
centrations as compared to existing surface waters. Lake de-
tention time during dry weather will be approximately six days.
Consequently, treated wastewater will be the dominant influence
on lake water quality at The Woodlands.
The lakes will serve as stormwater storage reservoirs and, in
so doing, will remove significant amounts of pollutants, primari-
ly due to sedimentation.
Lake Harrison Sedimentation—
Eight storm events were monitored at the lake system, ranging
from 0.26 in. (0.66 cm) to 3.97 in. (10 cm) of rainfall. In the
following paragraphs, the largest storm indicates the usefulness
of reservoirs for preventing release of construction site sedi-
ment washoff.
Rainfall associated with the storm event began shortly
after midnight on April 8, 1975 and continued until noon the
same day. An early morning cloudburst was followed after a four
hour pause by less intense rainfall totaling 3.97 in. (10 cm).
The hyetograph is shown in Figure 32. Runoff passing the Lake B -
gaging station, the major inflow to Lake Harrison, originated in
a watershed undergoing intense development at this time. Much
of the drainage system itself was being constructed under speci-
fications of the "natural drainage" system including Lake C,
656.2 ft (200 m) upstream of Lake B. Lake C was constructed to
serve as a wet weather pond and golf course water hazard.
Unfortunately, its low earthen spillway had yet to be sodded and
provided an erosion source within the drainage channel.
Lake Harrison inflow and outflow hydrographs are compared in
Figure 33. Characteristic of runoff response in a small water-
shed, the multi-peaked inflow hydrograph was a product of the
sporadic hyetograph (Figure 32). Intense stormwater flow deepen-
ed the inflow channel by 6 in. (15 cm) and obliterated bales of
106
-------
001 08 09
SJO '39MVHOS10
. 8
8
(O
to
o:
§
O
§
o
o
s
o
ffl •
4J
AJ
J (1)
-------
hay placed in the channel to act as flow c°^rol deviceJ- Storm-
water flow crested shortly before noon on April 8 at a recora
discharge of 123 cfs (3.48 m3/s). Bank storage and ponding
helped to prolong minimum flow in the channel for two days, con-
tributing to the total inflow runoff volume of 93 acre ft (114 m^).
TnelakegsyStem effectively damped inflow fluctuations The
hydrograph peak traveled through the lakes in a half hour.
«._*>_ _. —,—.*..•»—»—»v* *i ^r*\n o i TnS3.ri GOIlC"J.l L.J- d. L. A.\jj.+*y » x^-*- m _
ticie comparison ua- luccai* ^ *->,« outflow was nutrient enriched as
a'rSultNol ont o^a^omSinStion of two sources:
(1) unmeasured runoff from the fertilized --._- ---_
1 to the lakes and/or direct precipitation on the
lakes,
The ouality of water held in the lakes prior to
SS storm event. (Water impounded in the lake
prior to the storm event approximated the runoff
volume.)
ai£fe^=
proximated maximum concentrations.
sedimentation is a
of 2660 mg/1 at
a
(117 t)
the voume of the
assume<1.
mass
215 S
ing Panther Branch by the lake system.
Table 26 shows the reduction in stormwaterseaiment^oad by
^lete1^^!, 100,, is a
108
-------
2800 r '40
CO
Q
8
o
a
800
<
o
400
/ \
TOTAL
SUSPENDED SOLIDS
DISCHARGE
2800 r
,800
o
f-
400
100
(O
- UJ
o
CO
Q
OUTFLOW
J 4 8 12
HOURS INTO STORM
Figure 33. Reduction of TSS through The
Woodlands lake system.
1 cfs = .028 m3/sec
16
20
109
-------
TABLE 25. SUMMARY OF WATER QUALITY PARAMETERS FOR SITES
LAKE A AMD LAKE B DURING THE APRIL 8, 1975
STORM EVENT
Drainage Area (acres)
Runoff Volume (ac-ft)
Rainfall (inches)
OUTFLOW
Lake A
483
93.4
3.97
INFLOW
Lake B *
337
93.2
3.97
Concentration of Water
Quality Parameters:*
Ortho-P
TP
NH3
N02
NO3
TKN
TSS
SOC
Total COD
Soluble COD
Specific Conduc-
tance (micromhos)
Turbidity (JTU)
Avg.
0.015
0.10
0.16
0.032
0.28
1.3
245.
13.6
41.8
26.4
130.
Max.
0.048
0.19
0.26
0.046
0.32
2.
356.
19.
45
31.
215.
Avg.
0.005
0.11
0.11
0.009
0.15
1.86
1273.
16.2
63.7
32.
85.
Max.
0.013
0.36
0.15
0.054
2.1
3.1
2660.
22.
87.
45.
304.
160.
210,
375.
900,
1 ac = .405 ha
1 ac ft = .123 ha-m
1 in = 2.54 cm
*all concentrations in mg/1 except where indicated
110
-------
result of total stormwater storage by Lake Harrison and does not
preclude discharge at a later time.
TABLE 26. STORMWATER SEDIMENT REMOVAL AT LAKE HARRISON
TSS Load During Storm Event
Storm #
8
9
10
13
14
15 & 16
Ibs Input
(Lake B)
104
13800
322000
6700
115302
4840
Ib Discharged
(Lake A)
Flow stored
within lake
991
61900
Flow stored
1850
3270
% Load
Reduction
100%
93%
80%
100%
84%
32%
1 Ib = .453 6 kg
2
Estimated value (Lake B gage inoperative) calculated
using estimated 10.1 ac-ft (1.24 ha-m)inflow times
sample average concentration, 421 mg/1.
Disinfection—
The FC standard adopted by the State of Texas for contact
recreation is 200/100 ml (Iog10 = 2.3) (53). For noncontact
recreation the figures are one order of magnitude higher, i.e.,
2,000/100 ml (Iog10 = 3.3). Sixteen of 27 storms, considering
all stations, exceeded the noncontact recreation standard. All
the Westbury storm events exceeded the standard. Contact
recreation standards were met only for Lake A stormwater sampled
on 3/4/75 and 3/12/75.
Dry weather flow in Panther Branch met the contact recre-
ational standard at site P-10 but not at site p-30. The mean
FC value for p-10 low-flow data was 2.13 (log1Q basis) indicat-
ing that, on the average, the criteria for contact recreation is
satisfied. However, the value for stream water at P-30 was 2.38
and therefore unacceptable.
Ill
-------
Disinfection of stormwater is feasible, and it is generally
agreed that large dosages will be required to achieve adequate
reduction in indicator and pathogen densities. Chlorine or
chlorine dioxide has been reported to be the most effective dis-
infectant, and in many cases the least expensive (54, 55, 56).
Davis et al. (57, 58) discussed current disinfection research
and practices along with encountered problems which occur in
combined sewer disinfection and stormwater disinfection.
Samples were obtained from different locations during storm
events to determine disinfectant demand and effectiveness of
ozone and chlorine. Chlorine demand of Lake A stormwaters on
3/13/75 and 4/7/75 was 10 mg/1, and the ozone demand was in ex-
cess of 32 mg/1. These elevated demands were partially due to
suspended solids and oxidizable materials competing for the dis-
infectant. As a result, stormwater disinfection will be costly.
Chlorine and ozone toxicity—Water quality standards often
state that the final concentration of any waste in a receiving
water should not exceed 1/10 of the 96 hour LC5;) value (59). To
assume that 1/10 of even a true 96 hour LC^Q would not have
severe physiological effects and truly impair natural propagation
of all species is unrealistic. Tsai (60) reported that species
shifts occurred following the introduction of chlorine into a
Maryland river. Also, Arthur and Eaton (61) show that the re-
production of fathead minnows is drastically affected by exposure
to sublethal concentrations of chlorine. This evidence suggests
that sublethal concentration of chlorine is capable of producing
significant physiological impairment.
The approach taken herein was to establish a chlorine stand-
ard based on physiological responses of the test animal. The ef-
fect of chlorine and ozone exposure on the phys Lological function
of Ictalurus punctatus (the channel catfish) has been examined.
The physiological parameters evaluated were heart rate, blood
pressure, sodium uptake, ion excretion, and glomerular filtra-
tion. In addition, typical LC5Q bioassay tests were completed
for comparative purposes. Significant results are presented in
the following paragraphs.
Survival-mortality characteristics—The survival-mortality
characteristics of fingerling channel catfish to chlorine were,
examined and results are shown in Figure 34. A flow through
bioassay compared to a static bioassay results in a lower 96 hour
LC5Q. The LCcQ concentrations of 0.07 mg/1 for the flow through
bioassay and 0.45 mg/1 for the static bioassay were read directly
from the figure. A flow through or continuous ::low bioassay is
a more accurate measurement for 96 hour LC5Q of a highly reacting
toxicant like chlorine (62) . The chlorine concentrations in
Figure 34 are based on the measured amounts of total residual
chlorine added to the inflow and do not represent concentrations
during tests with fish. Thus the chlorine concentrations shown
112
-------
>
<
CO
(/»
<
o
CO
O
1MB*
A
O>
6
cn
•H C
C,
0)
>.
4- TJ
cn
T5
cn
cn
*!/ TO
O C. O
i"0 3 •*•"*
£j fft jQ
.
(U £4
C .C
GO
6
r-
•
o
03
6
10
0
*r
6
ro
6
CJ
Q
—
•v»
o>
E
UJ
z
oc
o
1
X
o
TJ Sj
p 0 3
ti «-H O
o x: --<
e o u-,
f-H O fl
(tf P
> c
> cn
3 ^-4 C
cn P fQ
OH O >
O TJ
oo cn
c c cn
O *r"l ^0
cn «-H o
• H C,-^l
t. a) jQ
(TJ CO
CU C 0
g -H -H
C •*-• P
O
-------
l v £ b? rePrese*tative at treatment outfalls
Samples taken from receiving waters should show chlorine concen-
X< s^lflcantly lower. Based on a 96 hour ie*ZfToT
and USing the "Aquatic Life Water Qualfgy criteria"
^^
noted that this concentration is below the liSit of Analysis?
The flow through bioassay was also used for ozone and the
survival-mortality characteristics of fingerling catfish are
, t.
reactive, it was decomposed in the presence of organics (fish).
In the presence of fish, ozone could not be detected in the
insensitivity of the analytical technique
^entrations were estimated from
by its odor (detectable odor
exceeds
chlorine^into the 10 liter recirculating system. Upon
??fe WaS a dr°P ±n bl°°d P^ssure, duS to ga
Following escapement from vagal inhibition, there
m;an.in=rease m blood pressure which in turn decreased
mmutes of exposure. This is thought to be due to an
e X vascular resistance, changes in heart rate
r be secondary compensations for the gill vascular
resistance chlorine ^xposure at levels approaching the 96 hour
LC50 (0.7 mg/1) were immediately detected as shown by a pro-
1111 ° in.blood Pressure. Continued exposure for 5 hours
increase in heart rate fr°ni 22 beats/minute to
At an exposure of 0.03 mg/1 chlorine there was
,.react:LOn shown b^ the bl°od pressure or heart rate
After two hours the fish appeared normal with no apparent SiJl
function. Thus, it is likely that with exposure to a chlorine
concentration of 0.007 mg/1 chlorine (1/10 of 96 hour LC I
fish will shown no .physiological response. 50
Tests to determine the effects of ozone on blood pressure
pulse pressure and heart rate indicate that ozone levels from
0.1 to 0.5 mg/1 for 10 hours had no apparent effect.
4-T, GJ1,1 S0d:i?m transPort~The influence of chlorine exposure on
the uptake of 22Na by gills of fish is shown in Figure 36. The
disappearance of 22Na from agueous phase (ordinateyaxis in Figure
114
-------
GO
C
P
t,
V JZ
GO GO
C 3
•H O
.C
M-i P
O
tn o
O rH
•H 4-1
P
tn 03
•H
p
o a>
f- 3
rfl 03
J2 O
u o.
X
>, QJ
P
•H 0)
fH C
m o
t. o
o
e o
i P
03 .C
>
03
tn tn
tn
03
3 03 -H
co o Xi
ro
0
fj
3
GC
•H
115
-------
36) is a measure of the removal of 22Na by the gills of the fish.
The addition of 0.1 mg/1 of chlorine at the start of the experi-
ment markedly reduced the ability of the gills to remove sodium
from aqueous solution. This reduction in uptake rate following
chlorine exposure suggests an impairment of normal physiological
function in the sodium transport system.
o p
The influence of ozone exposure on the uptake of Na by the
gills of fish is shown in Figure 37. Phase one of the experiment
utilized the fish as its own control. A one day recovery period
followed. The addition of 0.1 mg/1 ozone in the second phase of
the experiment markedly reduced the ability of the gills to re-
move sodium from the aqueous solution. This reduction in uptake
following ozone exposure also suggests an impairment of normal
physiological functions.
Eutrophicat ion—
The development of standing crops of phytoplankton in pan-
ther Branch is influenced to a large extent by streamflow rates.
High flow rates do not provide adequate detention times for de-
velopment of large standing crops at any given point along the
stream. However, reductions of flow rates and/or pooling in the
stream allow detention times suitable for standing crop develop-
ment, provided nutrients are not limiting for algal growth. De-
velopment of algal populations in this stream, as in other aquat-
ic ecosystems, is influenced by concentrations and availability
of various algal nutrients. This point is not only pertinent to
development of phytoplankton in Panther Branch but also to
aquatic systems which might receive water from this stream.
Thus, it is imperative to have some knowledge of which nutrients
stimulate algal growth in Panther Branch water. Therefore, algal
bioassays were conducted to determine whether nitrogen and/or
phosphorus were limiting for algal growth.
Low-flow water samples collected from sites on Panther
Branch and Spring Creek were used to determine the limiting-
nutrient for algal growth. Aliquots of stream water were inocu-
lated with the algae Selanastrum and spiked with nitrogen and/or
phosphorus as nutrients. The results presented in Figure 38 in-
dicate algal growth was increased by additions of both nitrogen
and phosphorus to water samples and phosphorus was the most im-
portant single limiting nutrient along Panther Branch. The in-
troduction of treated sewage effluent and agricultural runoff
into Spring Creek was probably responsible for the comparatively
larger algal yields in water from this stream. These findings
are in contrast to results derived from bioassays of stormwater.
Algal bioassays were also conducted with water collected
from Panther Branch at various time intervals during the course
of storm events. The stormwater runoff collected below the major
area of construction in The Woodlands (site P-^-30) seemed to
fluctuate in its ability to support the growth of algae (Figure
116
-------
CVJ
O
csi
CO
cr
•=>
o
X
t\J
III
S
00
Q
O
CJ
O
O
00
O
O
.a
(d
CN
M-i
O
^1 0)
«3 fi
-P -H
&H
D O
H
-------
P9
CD
W
co
to
cvj
00
cc
_ -3
O
•31
-
UJ
•v
«eE
__
CO h-
ro
p
3
CO
CD 3
P it!
P
C O
O C
3
CD O.
3 CO
CO 3
O SU
a 3
X >H
CD rd
CD O
C M
O
O O
e co
P< i — i
0) ^
P -H •
bO'—
P 0)
^ CD C
O ,C O
x; p N
CO C
UH il i— 1
O ^.
rQ bC
-P 2 E
OCM
CDOM r-i
4-4 U-t O
cj o >-'
1O
a
O
O
O
o
o
00
o
o
CO
0
o
^*
0
o
OJ
v
o
O
0)
bO
•H
3inNIW /SINOOD
118
-------
UJ
o
o
t
o
I-WATER SAMPLES - SPIKES
2-WATER SAMPLES * 1.0 ppm N
3-WATER SAMPLES * O.O5 ppm P
4-WATER SAMPLES * 1.0 N * 0.05 P
P-IO
P-20
P-3O
P-4O S-09
S-IO
COLLECTION SITE
Figure 38,
Optical densities of Selenastrum capricornutum
after incubation in v/ater from panther Branch.
and Spring Creek (May, 1974).
119
-------
39). Test samples collected early in runoff at P-30 demonstrated
the greatest growth in both control and spiked aliquots. These
samples corresponded to high nutrient stormwaters from the fer-
tilized areas of The Woodlands, not present at site p~10. As
runoff at P-30 progressed there was a decrease in algal cell
yields to similar levels observed for P-10 test samples. Algal
growth was increased by enrichment of stormwater samples with
both nitrogen and phosphorus; however, nitrogen was consistently
limiting for algal growth in stormwater runoff from The Woodlands.
Data from storm events at Hunting Bayou were similar to
those described above, indicating a stimulation of algal growth
with additions of nitrogen or both nitrogen and phosphorus to the
majority of water samples and slight or no stimulation with only
phosphorus spikes (Figure 40). The marked reduction in algal
growth, even with combined nitrogen and phosphorus, in the 4:30
sample could have been due to the presence of some toxic agent or
the absence of some essential trace element. Additional data ob-
tained from bioassays of stromwater from Westbury Square (Figure
41) also indicated that nitrogen was the limiting nutrient for
algal growth in stormwater runoff.
Porous Pavement
Rainwater Storage and Quality—
The water depth under the porous pavement in both the sand
subbase and the general storage layer during a period of rainfall
from 3/6/76 to 3/8/76 is shown graphically in Figure 42. The
depth of the sand layer is 33 in- (84 cm),, and the layer shows
saturation resulting from failure of the French drain to remove
water from the lower levels. The gravel layer responded to rain-
fall by storing water as shown by the increased water depth in
the gravel layer followed by a gradual decrease in depth as the
water was drained from the top of the sand layer. The height of
the stored water and the time to reach the peak height depend on
the quantity and intensity of the rainfall. This is best illus-
trated in Figure 42 in which seven separate rainfall events oc-
curred over a 60 hour time span. The first three events, total-
ing 1116 ft3 (31.6 m3), were spread over approximately five hours
and produced an even response with a peak of approximately six in.
The subsequent event at 21 hours involved only 283 ft3 (8m3) of
rainfall and increase in stored water height of 1.8 in. (4.6 cm).
For the two major storm events occurring at 28 and 48 hours, the
water level had not drained sufficiently and the increase in
stored water height exceeded the measuring probe length. The
time required for the majority of stored water to drain away was
approximately 10 hours.
The quality of the water in runoff from the standard pave-
ment and in the gravel and sand layers beneath the porous pave-
ment was monitored during six storm events. As expected, there
was a general flushing effect for the runoff water with initially
120
-------
*r~
1
*i
101
(Ml
* 1
TO]
OJ|
-1
*r
rol
—1
*c_
toT
M|
*E
Sevif^
5 o HT!
ff 0. 4. L-
— c> — toT"
^C
o o o p 9.
(Siuiooi/6uu) 1H9QM AHQ
— ^
2
• 1
UJ
J k
-------
CO
UJ
it:
CL
CO
i
_l
CL
^
CO
ce
ft
*<
•$•
i
CL
o-c
2 P C
g g- 4
a. °-
o. 10 2
/-\ O <~
2 § -
+ +• 4
_J -l -
Q. Q. a
2 2
< ^ <
CO. CO V
on on a,
UJ Ul U
• * i_
§ si
^ ^ ^
OJ K> ^
CMC
1
^•c
CVlf
^*
^
1
«•[
[Ml
*l
(Ml
Tl
CVJI
-
t
cad
•1
rol
-I
1
rol
— L
rol
— 1
i
— L
rol
i
1
rol
ro
__
•
•
:
srl
ro
•«•«•
CM
^^MNH
}
n
Q
CO
o
o
U>
S:
vC
O
CM
O
—
00
o
9
m
CM
en
LU
Q
a
8
UJ
CL
cn
^ VD
0) OJ
-P
rtJ .d
SC)
P* U
»-i ^1
M (d
O S
-P -—•
to
S O
fd
PQ
ti
-P
03
fd
C
Q)
Q)
CO
£ O
O G
M H
-t-i
CO
CO
1H9I3M
122
-------
O
O
C£
UJ
Q
7.0
6.0
5.0
4.0
3.0
2.0
1 - CONTROL
2- 1.0 N
.1. 3- 0.05 P
2
1
3
4
I
-
3
4- 1.0 N* 0.05 P
1
2
3
4
1
2
4
3
•••
2115-2215 , 2300-0045, OII5-O215, 0245-0315
WESTBURY SQUARE
TIME OF COLLECTION (hrs)
Figure 41. Growth of Selenastrum in stornwater
runoff from Westbury Square (May 8,
1975).
123
-------
UJUJUJ< •
>CE>Q.
<£LOtO
to
•»-
o
o
CO
IO
u>
CM
to
**
H-
N-
IO
•*-•
H-
to
CO
CM
to
•*"•
**-
JO
« w™
tLJ?
O m. O
— o
HH/S3HONI
UJO
cc
LU
<
CO
0.
o
o
(O
in
to
in
CO
a:
m rs
to o
o
to
in
CM
in
CO
-------
high concentrations of contaminants which then rapidly decreased.
The rapidity if change is not evident in the water underlying the
porous lot. Table 28 presents a summary of stormwater quality
data collected during a 0.44 in. (1.12 cm) rainfall on 2/20/76,
which is exemplary of the monitroing effort. SOC and soluble COD
are generally lower in the water under the porous lot than in the
runoff from the standard pavement, but the reverse is true for
conductivity, TKN, and NH3. This is apparently a result of the
failure of the French drain to completely drain the sand layer
beneath the porous lot. The lower COD and TOC values, together
with the high Kjeldahl and ammonia nitrogen, indicate than an
anaerobic digestion process is occurring in the standing water.
Lead values are generally lower in the water under the porous lot
and, in most cases, would be acceptable in aquifers. The few
cases where high lead values were found in the underlying layers
do indicate that before porous paving is used for recharge of
aquifers by percolation information must be obtained on the ability
of soils to remove lead from leachates.
Wet Skid Resistance—
A modification of the locked wheel test was used to deter-
mine the coefficient of sliding friction between a standard
passenger vehicle tire (size 6.50 x 13) and the test pavements,
both wet and dry. In the case of the porous pavement, tests were
made on a section of the original paving and on a repaired section.
Test results for three series are shown in Table 27,
TABLE 27. COEFFICIENT OF FRICTION
Test Date Paving and Condition
Reaction Force
(Ibs)
Coefficient
of Sliding
Friction (u,)
12-3-75
2-5-76
4-28-76
Old porous - dry
Old porous - wet
New porous - dry
New porous - wet
Standard - dry
Standard - wet
Old porous - dry
Old porous - wet
New porous - dry
New porous - wet
Standard - dry
Standard - wet
Old porous - dry
Old porous - wet
New porous - dry
New porous - wet
Standard - dry
Standard - wet
43.2
60.8
47.2
52.8
52.0
42.4
58.4
68.0
59.2
64.0
71.2
57.6
52.5
60.0
57.0
60.0
55.5
51.0
.605
.851
.661
.739
.728
.594
.818
.952
.829
.896
.997
.806
.735
.840
.798
.840
.777
.714
1 Ib . = .138 nt
125
-------
TABLE 28. SUMMARY OF STORMWATER QUALITY FOR POROUS
PAVEMENT STORM ON 2/20/76
Conventional
1
Constituent
pH
Sp . Cond .
Soluble COD
SOC
TKN
NH3
NO3
N02
TP
ortho P
Pb
Zn
Gravel
X
8.1
457
35.
15.
2.5
1.5
.12
.014
.10
.06
.05
.18
Porous
Layer
s
0.2
14.
5.8
4.4
.42
.41
.14
.017
.02
.02
.03
.39
Pavement
Sand
3c
8.0
542
35.
13.
2.7
1.33
.03
.004
.50
.48
.03
.27
Layer
s
.04
12.
3.7
2.8
.46
.28
.01
.002
.08
.06
.01
.18
Pavement Runoff
x
7.8
108
60.
30.
1.2
.15
.36
.013
.11
.06
.31
.34
s
.23
54.
35.
16.
.42
.09
.17
.009
.04
.02
.43
.21
n = 11
n = 11
n = 10
all measurements in mg/1 except
specific conductance in
2 _
in pH units and
= mean
s = standard deviation
n = number of samples
126
-------
In most cases the standard, dense paving exhibited dry skid
resistance better than or equal to the two porous pavements. The
reverse was true when wet paving was tested with the porous pave-
ments being clearly superior. For standard pavement, the coef-
ficient of friction under wet conditions was, as expected, lower
than under dry conditions. However, a consistent behavior pat-
tern for the porous pavements, for which there is currently no
verified explanation, is the increase of the coefficient of
friction on wet pavement over that on the dry pavement. One
possible explanation is that in the case of the wetted pavement
surface dust layers have been washed through the pavement and
the rougher surface .then comes into full play.
Noise Levels
Traffic noise levels (Table 29) were determined using two
different vehicles, both equipped with standard steel belted
radial tires. Measurements were made early in the morning to
avoid external noise interference.
TABLE 29. NOISE LEVELS 1 mi = 1.6 km
paving
Vehicle
Speed (mph)
Noise Level!(dB)
Old porous
New porous
Standard
Old porous
New porous
Standard
Old porous
New porous
Standard
1
1
1
2
2
2
2
2
2
15
15
15
15
15
15
20
20
20
A62
A57
A64
B68
B68
B71
B69
B69
B73
The porous pavings are less noisy than standard paving.
1 A scale consistent with human ear. B scale filters to
accentuate lower frequency sound.
Porosity
In situ porosity was determined on the porous pavements_by
grouting in place a 6 in. diameter tube having two scribed lines
5 in. apart. Water was placed in the tube and the time required
for the water level to pass the marks was measured. Results are
calculated in in./sec of water transmission. The section of new
porous pavement gave the best water transmission than that in the
original pavement. Water transmission rates in the new pavement
were normally 0.55 in./sec (1.4 cm/sec) while the original sec-
tions exhibited a normal porosity of 0.38 in./sec (0.96 cm/sec).
Sections of original paving were found to be clogged. These sec-
127
-------
tions were primarily near the curb or near the curbing at the
upper side of the lot opposite the curb. The clogging near the
curb resulted from mud and dirt from the wheels; of contractors'
trucks, while the clogging of the upper lot was; chiefly from
cement dust from curb construction and find sand and silt washed
from the grass plot before the grass gave adequate cover. The
partially clogged or completely clogged areas were marked for
studies of cleanability.
Cleaning and Maintenance
It was recognized early in the program that maintenance and
cleaning of porous pavement to prevent or alleviate clogging
would be a factor in the application of such peivements. Sections
of the porous pavement which were clogged were cleaned by various
methods. No method was satisfactory on fully clogged pavement
and only a superficially clogged section, showing a water penetra-
tion of 0.38 in/sec (0.96 cm/sec), could be restored to normal
operation. The best method for cleaning was brush and vacuum
sweeping followed by high pressure water washing of the pavement.
Vacuum cleaning alone, once the pavement is clogged, was found to
have no effect. The oils in the asphalt bind c'.irt and only
abrading and washing techniques are effective in removal. By
removing fractional thicknesses of paving, it was observed
the clogging to a depth of 0.5 in (1.3 cm) was sufficient to
prevent water penetration.
Damage to pavement porosity results chiefly from abuse dur-
ing the early life of the paving. Normally, pa.ving is carried
out while heavy construction and earth moving is continuing in
the area and is subjected to mud and dirt from contractor ve-
hicles for up to several months. Continual passage of these
vehicles serves to compact dirt into the pores. Porosity can be
retained only if the paving is cleaned daily by sweeping and high
pressure water washing.
Once a large area of porous pavement is fully clogged it
cannot be adequately cleaned and the paving must be removed to a
depth where the clogging is not evident and new porous paving
filled in. In extreme cases, the affected area of the porous
topping must be removed and new topping put down.
Both the standard and porous pavements showed signs of the
effect of heavy vehicle passage and of power steering damage.
Heavy vehicles tended to leave tire depressions in either lot on
hot days while power steering damage was evidenced by small cir-
cular depressions in the lot surface. The latter damage occurs
when wheels are turned when a vehicle is not in motion. The use
of a lower penetration asphalt for pavements should offset the
damage to paving in warm climates.
128
-------
Pavement Deflection
Pavement deflections in porous and standard surface lots
were measured on four different occasions. Measurements were
made of pavement acceleration and time duration of acceleration
as it deflected on passage of the test vehicle. Test results
indicate there are no real differences in the magnitude of the
deflections, but the time duration of the deflection was always
longer for the porous pavement than for the standard pavement
with correspondingly lower "g" levels. This long response may be
responsible for the lessened tire noise observed in the sound
level tests.
Pavement Survey
Visits were made to four locations where porous paving is in
use. The locations at the University of Delaware, Newark, Dela-
ware, Bryn Mawr Hospital, Bryn Mawr, Pennsylvania, and at The
Marine Sciences Consortium, Lewes, Delaware were in good con-
dition with little or no indication of pavement clogging. The
paving at a Travelodge in Tampa, Florida showed considerable
clogging from sand and silt caused by passage of trucks on the
nearby road and inadequate curbing to prevent surrounding soil
from washing onto the lot. Degradation of the surface was evi-
dent in the form of patches of loose stone and gravel.
It is evident from the damage at the Tampa lot and in the
lot at The Woodlands that a lower penetration asphalt should be
used in the topping, especially in warm areas. Again, it is felt
that clogging can be minimized by proper use of curbing to pre-
vent surrounding soil from washing onto the lot surface. Recent
installation of porous paving at The Franklin Institute parking
lot in Philadelphia, Pennsylvania has further pointed up the need
for close control of contractor vehicles on a newly installed
lot. It was necessary to sweep away caked mud from vehicle
wheels and then wash the affected area with high pressure water.
Traffic Paint Visibility
Two colors of traffic paint, white and yellow, with and
without addition of glass beads, were applied to the pavements.
They were photographed under equal incident lighting intensities
at night'under wet and dry conditions and the reflected light in-
tensity was determined from image densities on the negatives.
The white marking paint showed a slight superiority in visibility
over the yellow whether or not glass beads were added. No real
difference in the reflected light intensity from either pavement,
wet or dry, was evident. The test, however, did not evaluate
glare from oncoming vehicle lights which may obscure reflected
light from paint on a wet, non-porous pavement.
129
-------
1.
2.
3.
REFERENCES
Field, Richard, and Pauline Weigel. Urban Runoff and Com-
bined Sewer Overflow. JWPCF, 6,: 1108, 1973.
Wolman, M. G., and A. P. Schick. Effects of Construction on
Fluvial Sediment, Urban and Suburban Areas of Maryland.
Water Resources Research, _3_: 451, 1967.
Guy, H. P., and G. E. Ferguson. Sediment in Small Reser-
voirs Due to Urbanization. j. Hydraulic Div., ASCE, 88,
HY2; 27, 1962.
4. Anderson, D. W., and S. D. Faust. Changes in Quality of
Water in the passaic River at Little Falls, New Jersey.
USGS Professional Paper 525-D, 1965.
5. Geldreich, E. E., L. C. Best, B. A. Keener,'and O. J. Van
Donsel. The Bacteriological Aspects of Storm Water Pollu-
tion. JWPCF, 40; 1861, 1968.
6. Storm and Combined Sewer Overflows Section, Environmental
Impact of Highway Deicing, U.S. EPA Report # 11040GKK06/71.,.
1971
7. Bryan, E. H. Quality of Stormwater Drainage 'from Urban
Land. Paper presented at 7th American Water Resources Con-
ference, Washington, D. C., 1971.
8. Heaney, J. p., and R. H. Sullivan. Source Control of Urban
Water Pollution. JWPCF, 43: 571, 1971.
9. Sartor, J.D. and G.B. Boyd. Water Pollution Aspects of
Street Surface Contaminants. U.S. EPA Report # EPA-R2-72-081,
1972.
10. Claudon, D. G., D. I. Thompson, E. H. Christenson,. G. W.
Lawton, and E. C. Dick. Prolonged Salmonella Contamination
of a Recreational Lake by Runoff Waters, Appl. Microbiol.,
21. (5): 875-877, May, 1971.
11. Tafuri, Anthony N. Pollution from Urban Land Runoff. News
of Environmental Research in Cincinnati, U.S. Environmental
130
-------
Protection Agency, Cincinnati, Ohio, April 11, 1975.
12. Harms, L. L., P. Middaugh, J. N. Dornbush, and.j. R. Ander-
sen. Bacteriological Quality of Surface Runoff from Agri-
culture Land - Part II. Water and Sew. Works, 122 (11)-
71-73, November, 1975. ~r~~~
13. Harms, L. L., P. Middaugh, J. N. Dornbush, and J. R. Ander-
sen. Agricultural Runoff Pollutes Surface Waters - Part I
Water and Sew. Works, 122 (10): 84-85, October, 1975.
14. Edmondson, W. T. Eutrophication in North America, in:
Eutrophication: Causes, Consequences, Correctives, Anon
(ed.). Proceedings of NAS-NRC Symposium, Washington, D. C.,
1969 -
15. Sylvester, R. O. An Engineering and Ecological Study for
the Rehabilitation of Green Lake. University of Washington
Seattle, Washington, I960. ' '
16. Hasler, A. W. Eutrophication of Lakes by Domestic Drainacre
Ecology, 28; 383, 1947.
17. Biggar, j. w,, and R. B. Corey. Agricultural Drainage and
Eutrophication. In: Eutrophication: Causes, Consequences
Correctives, Anon (ed.). proceedings of NAS-NRC Symposium,
Washington, p. c., 1969. *
18. Cooper, C. P. Nutrient Output from Managed Forests. In:
Eutrophication Causes, Consequences, Correctives, Anon (ed )
proceedings of NAS-NRC Symposium, Washington, p. C., 1969.
19. Jackson, D. P., and H. F. A. Meier. Variations in Summer
Phytoplankton Populations of Skaneateles Lakes, New York
Assoc. Theor. Appl. Liranol., proa., 16: 173, 1966.
20. palmer, M. C. Algae in Water Supplies of the United States
In: Algae and Man, Jackson, D, F. (ed.). Plenum press.
New York, 1964.
21. Schwimmer, D., and M. Schwimmer. Algae and Medicine, in-
Algae and Man, Jackson, D. F, (ed.), Plenum Press, New York,
22. Rose, E. T. Toxic Algae in Iowa Lakes. Iowa Acad. Sci
Proc., 60.: 738, 1953. '
23. Olson, T. A. Water Poisoning - A Study of Poisonous Algae
Blooms in Minnesota. Am. J. Health, 50: 883, ij.960.
24. Moore, G.. T., and K. F. Kellerman. A Method of Destroying
or Preventing the Growth of Algae and Certain Pathogenic
131
-------
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Bacteria in Water Supplies. US.DA Bureau Plant Indus-
tries Bulletin 64, 1904.
Anon (ed.). Eutrophication: Causes, Consequences, Cor-
rectives. Proceedings of NAS-NRC Symposium, Washington,
D. C., 1969.
Kerr, P. C., et al_. The Interrelation of Carbon and Phos-
phorous in Regulating Heterotrophic and Autotrophic Popula-
tions in Aquatic Ecosystems. Proceedings of 25th1Purdue
Industrial Waste Conference, West Lafayette, Indiana, 1970.
Keuntzel, L. E. Bacteria, Carbon Dioxide, and Algal Blooms.
JWPCF, 41: 1737, 1969.
Lee, F. An Approach to the Assessment of the Role of Phos-
phorous in Eutrophication. Paper presented at American
Chemical Society, Los Angeles, California, 1970.
Shapiro, j.
1970.
A Statement on Phosphorous. JWPCF, 42 ; 772,
American Chemical Society.
Chemical Basis for Action.
D. C., 1969.
Cleaning Our Environment: The
Amer. Chem. Soc., Washington,
Poche, R. M. A Baseline Census and Development of a Monitor-
ing System for Important Animal Species on the Proposed
Woodlands Site,'Montgomery County, Texas. Consultant's Re-
port to The Woodlands Development Corp., April, 1973.
Maestro, R. Criteria for Residential Wildlife planning in
the New Community of The Woodlands, Montgomery County,
Texas. Consultant's Report to The Woodlands Development
Corp., Mary, 1973.
Hass, R. H., and M. C. McCaskill. Use of Large-Scale
Aerial Photography in Obtaining Vegetation Information for
Urban Planning. Consultant's Report to The Woodlands Deve-
lopment Corp., July, 1972.
Kendrick, W. W., and D. Williams. Soil Survey of The Wood-
lands. Consultant's Report to The Woodlands Development
Corp., July, 1973.
Water Resources Data for Texas, part 1 Surface Water Re-
cords, 1973 and 1974. U.S. Dept. of the Interior publica-
tion, prepared in cooperation with the State of Texas.
Winslow, D. E., J. A. Veltman, and W. H. Espey, Jr. Natural
Drainage Systems. Paper presented at Spring Session of the
Texas Section of the ASCE, Beaumont, Texas, March 29, 1974.
132
-------
37. Dugan, G. L., and P. H. McGauhey. Protecting Our Lakes:
Wastewater Treatment is Not Enough. Paper presented at
46th Annual Conference of W.p.c.F., Cleveland, Ohio,
September 30-October 5, 1973.
38. Schicht, R. j., and p. A. Huff, The Effects of Precipita-
tion Scavenging of Airborne and Surface Pollutants on Sur-
face and Groundwater Quality in Urban Areas. , Part II.
NSF # GK-38329, National Science Foundation, Engineering
Division, Washington, D. C., July, 197$.
39. Whitehead, L. W. Some Microclimate and Air Quality Impli-
cations of Urbanization in a Southern Coastal Forest. Dis-
sertation, University of Texas School of Public Health,
Houston, Texas, 1976*
40. Ward, C.H., and J. King. "Eutrophication Potential of Surface
Waters in a Developing Community, " Draft Final Report fpr
Maximum Utilization of Water Resources in a Planned Community,
EPA Research Grant #802433, October, 1976.
41.
42.
43.
44.
45.
46.
47.
MacEntee, F. J. A Preliminary Investigation of the Soil
Algae of Northeastern Pennsylvania. Soil Sci., 110s 313-
317, 1970. - ".
Hofstetter, A. M. A Preliminary Report of the Algal Flora
from Selected Areas of Shelby County. Jour. Tenn. Acad. of
Sci., 43; 20-21, 1968.
MacEntee, F. J., G. Schreckenberg, and H. C. Bold. Some
Observations on the Distribution of Edaphic Algae. Soil
Sci., 114; 717-719, 1972.
Olivieri, V., C. Kruse, K. Kawata, J. Smith. "Microorganisms
in Urban Stormwater, " EPA^600/2-77-087, MERL, Cincinnati,
July, 1977.
Diniz, E. V. and W. H. Espey, Jr. "Maximum Utilization of
Water Resources in a Planned Community; Application of the
Storm Water Management Model: Volume I," EPA-600/2-79-050c, 1979.
Metcalf & Eddy, Inc., et al., Storm Water Management Model, in
four volumes, "Final Report," (Vol. I: USEPA Report No. 11024
DOC 07/71, p. 352); "Verification and Testing," (Vol. II:
USEPA Report No. 11024 DOC 8/71, p. 172;) "Users Manual,"
(Vol. Ill: USEPA Report No. 11024 DOC 10/71, p. 249).
U.S. Corps of Engineers. Urban Runoff: .Storage and Treatment
and Overflow Model "STORM." Hydrologic Engineering Center
Computer Program 723-58-L2520, U.S. Army, Davis, California,
May, 1974. ^
133
-------
48. Nash, J. G. The Form of the Instantaneous Unit Hydrograph.
Int. Assoc. of Sc. Hydrol., Pub. 45, Vol. 3_: 114-121, 1957.
49. Horton, R. E. An Approach Towards a Physical Interpretation
of Infiltration Capacity. Soil Sc. Soc. of America, Proc.,
5.: 399-417, 1940.
50. McKee, J. E., and H. W. Wolf. Water Quality Criteria.
May, 1963.
51. Davis, Ernst. "Microbiological Quality of Stormwater Runoff
in The Woodlands," Draft Final Report for Maximum Utilization
of Water Resources in a Planned Community, EPA Research
Grant #802433, October, 1976.
52. Report of the Commission on Water Quality Criteria. FWPCA,
April, 1968.
53. Texas Water Quality Standards. Texas Water Quality Board,
Austin, Texas, February, 1976.
54. Bender, R. J. Ozonation, Next Step to Water purification.
Power, 114 (8): 58-60, August, 1970 (Abs.).
55. Glover, G.E. and Herbert, G.R. Microstaining and Disinfection
of Combined Sewer Overflows. U.S. EPA Report f 11023EVO
06/70, June, 1970.
56. Moffa, P.E., et al. Bench-Scale High-Rate Disinfection of
Combined Sewer Overflows: With Chlorine and Chlorine
Dioxide. U.S. EPA Report #670/2-75-021, June, 1970.
57. Davis, E.M., L.W. Whiteheard, and J.D. Moore. Disinfection.
Jour. Water Poll. Control Fedn., 46 (6): 1181-1191,
June, 1974.
58. Davis, E.M., J.D. Moore, D. Casserly, J. Petros, and W.
DiPietro. Disinfection. Jour. Water Poll. Control Fedn.,
47^ (6): 1323-1334, June, 1975.
59. Horton, R.K. Aquatic Life Water Quality Criteria. Envirn.
Sci. and Tech., 11; 888, 1967.
60. Tsai, Chu-fa. Effects of Chlorinated Sewage Effluents on
Fish in Upper Patument River, Maryland. Chesapeake Sci. 9,
2_: 83, 1968.
61. Arthur, J.W., and J.G. Eaton. Chloramine Toxicity to the
Amphiphod, Gammarus Pseudolimneaus and the Fathead Minnow,
Pimepephales promelas. Jour. Fish Res. Bd. Can., 28: 1841,
1971.
134
-------
62'
Los Angeles, California, June 27-28, 1972
135
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-600/2-79-050a
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
MAXIMUM UTILIZATION OF WATER RESOURCES IN A PLANNED
COMMUNITY
Executive Summary
5. REPORT DATE
July 1979 (Issuing Date)
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
William G. Characklis, Frank J. Gaudet, Frank L. Roe,
and Philip B. Bedient
8. PERFORMING ORGANIZATION REPORT NO.
I. PERFORMING ORGANIZATION NAME AND ADDRESS
Department of Environmental Science & Engineering
Rice University
P.O. Box 1892
Houston, Texas 77001
1O. PROGRAM ELEMENT NO.
1BC822 SOS 2 Task 02
11. CONTRACT/GRANT NO.
802433
12. SPONSORING AGENCY NAME AND ADDRESS
Municipal Environmental Research Laboratory—Gin, OH
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final 7/73-12/76
14. SPONSORING AGENCY CODE
EPA/600/14
15. SUPPLEMENTARY NOTES one in a series of volumes of one report. Project Officers:
Richard Field and Anthony N. Tafuri, Storm and Combined Sewer Section, FTS 340-6674,
(201) 321-6674 '
16. ABSTRACTgtormwater £rom four watersheds in the Houston area was monitored over a three
year period. Land use in the watershed included undeveloped forest, developing forest,
iully-developed residential and mixed commercial-residential.
Chemical parameters monitored included suspended solids, oxygen demand, organic carbon,
.itrogen, phosphorous, dissolved oxygen, pH, specific conductance and chlorinated hydro-
carbons. Indicator and pathogenic bacterial species were enumerated as well as aquatic
and edaphic algae species. Disinfectant demand and algal bioassays were also conducted.
Relationships have been developed between stormwater runoff quality, quantity and4land
use in an effort to predict pollutant loads. The appearance of a "first flush"' is depen
dent on the parameter measured and watershed characteristics. Rainwater quality contri-
mtes significantly to stormwater pollutant loads, especially in urbanized areas. Modi-
Eying effects of natural biological processes on nitrogen content in the runoff and ef-
fects of the hydrological regime on nutrient limitations were observed. The effective-
less of storage lakes, very positive in the case of suspended solids, were also observed
Ehe Storm Water Management Model (SWMM) was modified to describe the processes occurring
in the watersheds and allowed for (1) separate sewer systems, (2) effects of urbaniza-
tion on base flows, (3) performance efficiency and cost effectiveness of natural drain-
age systems, (4) four additional water quality parameters (COD, Kjeldahl nitrogen,
vibrates, phosphates), (5) hyduologie effects of
17.
KEY WORDS AND DOCUMENT ANALYSIS
a.
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Water quality, Pesticides, Pavements,
Disinfection, Urbanization, Chemical
analysis
Demonstration watersheds,
Hydrologic data, Hydrologi
models, Overland flow,
Fish toxins, Eutrophicatioi
Water sampling, Porous pav J
nent, The Woodlands, Storm
Water Management Model
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
TTNfTLASSTFIED
21. NO. OF PAG.ES
150
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (Rev. 4-77)
136
4 U.S. GOVERNMENT PBINIINO OFFICE: 1080 -657-060/5430
-------
-------
-------
-------
H
1
en
o
o
^^
K)
VD
\
O
Ul
o
p
g.3 =5
S S. c
0) & CX
^l O
S" ^ ^
5 ^ S.
^° S S"
5:52.
lYi "j
aa n>
H
o> a
~- c:
O
? r^.
3 ^
*"*' a
C3 O'
O)
a- "~-
2 ^
5. 1
?s
§
3 ^-
o. 5
3 w
!=:
O C5
»l
CD Ct>
03 £
§~
^ ^^
Cb CD
l!
s 1
• to
CD
§
^-
CO
0)
Cr-
•^
cb
.^
Q)
CB
m
D
O -n -n
-0 O o
H 3 OD
C < C
Z > 0)
3 m Z
< ._ m
5»tf
38
m °
0)
T3
(0
O
IE
Tl
CO O
O C
o 3.
7T 3"
i
O
w
(A
51 O m
" c? I
CD 33
o i
5' S
o w
" S
(/> to
CD 3
ffl O
3 Q. O
3" • H
§ D m
3 CD O
o> < -j
Up
^3 ?
CD 3 5*
-* G)
gT 5- m
o
c
(0
m
z
3J "D
i s
1-s
rn —j rn
? ? i
CO D D
8 S S
H m
m 0)
O -n
I 5
C)
m
Z
n
------- |