CRWR Online Report 97- 5
USE OF VEGETATIVE CONTROLS FOR TREATMENT
OF HIGHWAY RUNOFF
by
Patrick M. Walsh
and
Michael E. Barrett, P.E.
Joseph F. Malina, Jr., P.E.
Randall J. Charbeneau, P.E.
Principal Investigators
October 1997
CENTER FOR RESEARCH IN WATER RESOURCES
Bureau of Engineering Research ? The University of Texas at Austin
J.J. Pickle Research Campus ? Austin, TX 78712-4497
ii
DISCLAIMERS
The contents of this report reflect the views of the authors, who are responsible for
the facts and the accuracy of the data presented herein. The contents do not necessarily
reflect the official views or policies of the Texas Department of Transportation. This report
does not constitute a standard, specification, or regulation.
There was no invention or discovery conceived or first actually reduced to practice in
the course of or under this contract, including any art, method, process, machine,
manufacture, design or composition of matter, or any new and useful improvement thereof,
or any variety of plant, which is or may be patentable under the patent laws of the United
States of America or any foreign country.
NOT ENTENDED FOR CONSTRUCTION, BIDDING, OR PERMIT PURPOSES
Joseph F. Malina, Jr., P.E. (Texas No. 30998)
Research Supervisor
ACKNOWLEDGEMENTS
The authors would like to express their thanks to the TxDOT project director, Kristie
Denton, for her many helpful suggestions.  We would also like to those that preceded her in
that role, Peter Smith and Jay Vose.
Research performed in cooperation with the Texas Department of Transportation and
the U.S. Department of Transportation, Federal Highway Administration.
iii
iv
TABLE OF CONTENTS
DISCLAIMERS.....................................................................................................ii
LIST OF TABLES ..............................................................................................vii
LIST OF FIGURES ...........................................................................................viii
SUMMARY............................................................................................................ x
Chapter 1 Introduction......................................................................................... 1
Chapter 2 Literature Review ............................................................................... 3
2.1 Factors That Affect Vegetated BMP Efficiency....................................... 3
2.2 Vegetated Buffer Strip Treatment Effectiveness ...................................... 4
2.3 Effect of Metals Deposition on Vegetated BMPs..................................... 7
2.4 Grassed Swale Treatment Efficiency........................................................ 8
Chapter 3 Channel Swale Experiments ............................................................ 11
3.1 Introduction............................................................................................. 11
3.2 Methods and Materials............................................................................ 11
3.2.1 Setup ........................................................................................... 11
3.2.2 The Cocktail................................................................................ 14
3.2.3 Experiment Procedure................................................................. 16
3.3 Experiment Philosophy........................................................................... 19
3.3.1 Water Depth................................................................................ 20
3.3.2 Season ......................................................................................... 20
3.3.3 Length of Swale and Groundwater Quality ................................ 21
3.3.4 Schedule and Experimental Conditions...................................... 21
3.4 Efficiency Calculations........................................................................... 22
3.5 Channel Experiments Results ................................................................. 25
3.5.1 General Results........................................................................... 25
3.5.2 Effect of Water Depth on Swale Removal Efficiency................ 26
3.5.3 Effect of Swale Length on Removal Efficiency ......................... 31
3.5.4 Effect of Seasons on Swale Removal Efficiency........................ 33
v
3.5.5 Underdrain Water Quality........................................................... 43
3.5.6 Summary of Channel Swale Results........................................... 47
Chapter 4 Field Experiments ............................................................................. 49
4.1 Introduction............................................................................................. 49
4.2 Methods and Materials............................................................................ 49
4.2.1 Site Selection .............................................................................. 49
4.2.2 Site Descriptions......................................................................... 50
4.2.3 Sampling/Monitoring Setup........................................................ 54
4.2.4 Sampling/Monitoring Procedures ............................................... 59
4.2.5 Numerical Analysis..................................................................... 60
4.2.6 Grab Samples.............................................................................. 63
4.3 Field Results ........................................................................................... 64
4.3.1 Runoff Coefficients..................................................................... 64
4.3.2 Concentration and Loading Reductions...................................... 66
4.3.3 Grab Sample Results................................................................... 72
4.3.4 Other Monitoring Results ........................................................... 73
4.4 Effects of Metals on Vegetated Areas .................................................... 75
4.4.1 Concerns Regarding Metals Deposition on Vegetated Areas..... 75
4.4.2 Use of Part 503 Regulations to Assess Environmental Risk ...... 76
4.4.3 Justification of Use of 503 Regulations for Stormwater............. 77
4.4.4 Metals Limitations Placed by the 503 Regulations .................... 78
4.4.5 Metals Risk Analysis Results and Discussion ............................ 79
4.5 Summary of Field Study Results ............................................................ 81
Chapter 5 Conclusions and Recommendations................................................ 82
5.1 Channel Swale Conclusions.................................................................... 82
5.2 Field Study Conclusions ......................................................................... 83
5.3 Recommendations................................................................................... 84
BIBLIOGRAPHY ............................................................................................... 85
APPENDIX A ...................................................................................................... 88
vi
APPENDIX B....................................................................................................... 91
APPENDIX C .................................................................................................... 110
vii
LIST OF TABLES
Table 2.1.  Summary of previous filter strip studies. ................................................................ 5
Table 2.2.  Summary of grassed swale removal efficiencies. ................................................... 9
Table 3.1.  Ingredients in the cocktail. .................................................................................... 15
Table 3.2.  Analytical methods for sample analysis................................................................ 19
Table 3.3.  Summary of lab experiments. ............................................................................... 22
Table 3.4.  Removal efficiencies calculated for the channel swale at different water depths. 26
Table 3.5.  Required swale length for total suspended solids removal. .................................. 33
Table 3.6.  Average removal efficiency for constituents based on underdrain water quality. 47
Table 4.1.  Vegetative composition of Walnut Creek median (October 1996)....................... 52
Table 4.2.  U.S. 183 at MoPac Vegetation Composition (October 1996)............................... 54
Table 4.3.  Vegetated buffer strip description summary......................................................... 55
Table 4.4.  Schedule for taking samples during storm events................................................. 59
Table 4.5.  Reductions in concentrations observed at two vegetated buffer strips. ................ 67
Table 4.6.  Constituent loadings with and without treatment by the vegetated buffer strip. .. 68
Table 4.7  Comparison of filter strip performance with three sand filtration systems............ 70
Table 4.8.  Annual metals loading rates, in comparison to the 503 Regulations. ................... 79
Table 4.9.  Site lives based upon metal deposition limitations. .............................................. 80
viii
LIST OF FIGURES
Figure 3.1. Cross-section of channel swale............................................................................. 12
Figure 3.2.  Overview of PVC Pipe Locations........................................................................ 13
Figure 3.3.  General flume apparatus. ..................................................................................... 13
Figure 3.4.  Addition of sediment and clays for creation of highway runoff cocktail. ........... 16
Figure 3.5.  Appearance of the swale during an experiment................................................... 18
Figure 3.6.  Confirmation of relatively constant influent concentrations in the swale. .......... 25
Figure 3.7. Effect of water depth and swale length on TSS removal efficiency..................... 27
Figure 3.8. Effect of water depth and swale length on COD removal efficiency. .................. 28
Figure 3.9. Effect of water depth and swale length on nitrate removal efficiency. ................ 29
Figure 3.10. Effect of water depth and swale length on TKN removal. ................................. 29
Figure 3.11. Effect of water depth and swale length on total phosphorus removal................ 30
Figure 3.12. Effect of water depth and swale length on zinc removal efficiency................... 30
Figure 3.13. Effect of water depth and swale length on iron removal efficiency. .................. 31
Figure 3.14.  Seasonal comparison of TSS removal (water depth = 4 cm)............................. 34
Figure 3.15.  Seasonal comparison of TSS removal (water depth = 7.5 cm).......................... 35
Figure 3.16.  Seasonal comparison of zinc removal (water depth = 7.5 cm).......................... 36
Figure 3.17.  Seasonal comparison of zinc removal (water depth = 4 cm)............................. 37
Figure 3.18.  Seasonal comparison of COD removal (water depth = 7.5 cm) ........................ 38
Figure 3.19.  Seasonal comparison of nitrate removal (water depth = 4 cm). ........................ 39
Figure 3.20.  Seasonal comparison of nitrate removal (water depth = 7.5 cm). ..................... 40
Figure 3.21.  Seasonal comparison of total phosphorus removal (water depth = 4 cm)......... 41
Figure 3.22.  Seasonal comparison of total phosphorus removal (water depth = 7.5 cm)...... 42
Figure 3.23. Removal of TSS during infiltration in channel experiments. ............................. 44
Figure 3.24. Removal of turbidity during infiltration in channel experiments. ...................... 45
Figure 3.25. Removal of total phosphorus during infiltration in channel experiments. ......... 45
Figure 3.26. Removal of zinc during infiltration in channel experiments. ............................. 46
Figure 4.1.  Mopac at Walnut Creek filter strip. ..................................................................... 51
Figure 4.2.  Vegetated buffer strip at U.S. 183 site................................................................. 53
ix
Figure 4.3.  Compound V-notch weir for flow measurement. ................................................ 58
Figure 4.4.  Runoff coefficient of the filter strip drainage area at Walnut Creek. .................. 65
Figure 4.5.  Initial calculation of runoff coefficient of filter strip drainage area at U.S. 183. 65
Figure 4.6. TSS concentrations along the center of the median for 5 storm events................ 73
Figure 4.7.  Erosion at the Walnut Creek vegetated buffer strip............................................. 74
x
SUMMARY
Pollutants found in runoff from highways may cause toxic responses in receiving
waters for some conditions and are obstacles to obtaining water quality goals in the United
States.  This study investigated the capability of two vegetative controls, grassed swales and
vegetated buffer strips, to treat highway runoff.
A grassed swale was constructed in an outdoor channel to investigate the impacts of
swale length, water depth, and season of the year on removal efficiency.  Results indicate that
swale length and water depth affect the removal of runoff constituents by swales, and the
removal efficiency can vary with the season of the year.  Two vegetated strips treating
highway runoff in the Austin, Texas, area also were monitored to determine removal
capabilities.  The filter strips removed most constituents effectively and consistently, and the
inclusion of filter strips is recommended in future highway design if conditions are
appropriate and right-of-way is available.
1
Chapter 1 Introduction
Stormwater runoff from highways can contain pollutants, including suspended solids,
nitrogen and phosphorus, organic material, and metals.  Concern regarding the harmful
effects of these constituents on receiving waters has grown since the 1970s.  The results of
bioassay tests of organisms from streams and lakes receiving highway runoff have shown
that highway runoff, though it may not demonstrate acute toxicity, may cause toxic responses
for some conditions (Barrett et al, 1995b).  In addition, highway runoff can add to existing
runoff problems in urban areas.  Today, sources of urban runoff, including highways, are
considered ?formidable obstacles to achieving water resource goals? in the United States
(U.S. EPA, 1993).
Regulatory requirements reflect the need to protect the environment from urban and
highway runoff effects.  Approval by regulatory agencies is required before development is
begun in urban areas, including the construction highways.  On a national level, for example,
a stormwater discharge permit is required for highways in urban areas, as part of the National
Pollutant Discharge Elimination System (NPDES) enforced by the U.S. Environmental
Protection Agency (EPA).  In addition, state or municipal rules may apply.  For example, the
Texas Natural Resource Conservation Commission (TNRCC) requires a stormwater
management plan before development is allowed over the environmentally sensitive recharge
zone of the Edwards Aquifer in Austin, Texas.
Hence, both environmental response and regulatory reasons indicate the need for a
stormwater management plan for highways.  The Texas Department of Transportation
(TxDOT) manages highways in Texas and funded this study for these reasons.
The BMPs investigated in this study are permanent vegetative controls: grassed
swales and vegetated buffer strips.  Grassed swales are shallow, grass-lined, typically flat-
bottomed channels that convey stormwater at moderate slopes.  In grassed swales, treatment
occurs as the water flows in deep flow down the swale.  Vegetated buffer strips, also known
as filter strips, are not channels, but are relatively smooth vegetated areas at moderate slopes
that accept highway runoff as overland sheet (shallow) flow.  The mechanisms of removing
constituents in runoff for the two practices are the same: filtration by grass blades or other
2
vegetation, sedimentation, infiltration into the soil, and biological activity in the grass/soil
media.
The objectives of this study are:
1. determination of the effectiveness of grassed swales and vegetated buffer strips for
treating highway runoff;
2. determination of the factors that affect the removal efficiency of grassed swales and
vegetated buffer strips; and
3. evaluation of the potential risk to human health and the environment caused by the
deposition of metals on grassed swales and vegetated buffer strips.
The work involved in this study consisted of two parts.  First, a study of grassed
swales was completed in an outdoor channel.  This swale provided a controlled environment
that allowed for an evaluation of the effects of swale length, water depth, and season of the
year on the capability of a swale to remove constituents from simulated highway runoff.  The
second portion of the study involved monitoring of two vegetated buffer strips that receive
highway runoff.  Monitoring demonstrated the effectiveness of filter strips at removing
constituents from highway runoff.  It also provided constituent concentrations necessary to
accomplish the third objective of this project: an evaluation the environmental effects of
metals deposition on vegetated BMPs.
3
Chapter 2 Literature Review
2.1 Factors That Affect Vegetated BMP Efficiency
Factors that affect the removal efficiency of vegetated BMPs treating urban runoff
include vegetation type, slope, flow velocity, flow depth, season, and length.  Only one
previous study was designed specifically to understand the extent of the impacts and the
relative importance of the various factors.  Other insight into factors was gained incidentally
while researching the effectiveness of BMPs.
Glick et al (1993) investigated the effect of vegetative cover and several other factors
on filter strip effectiveness in an urban area.  Four different vegetated covers were compared:
wooded areas, wooded areas cleared, native grasses unmowed, and native grasses mowed.
The forested areas produced the highest concentrations of pollutants, and the mowed and
unmowed areas generally had the lowest concentrations.  Thus, grassed areas were found to
be more effective at removing pollutants than forested areas.  Vegetative composition was
found to have a significant impact on filter strip effectiveness.
Schueler (1987) reports that vegetation type is an important factor in filter strip
performance.  He reports that forested filter strips have greater pollutant removal capability
than grassed filter strips, due to faster nutrient uptake and longer nutrient retention in forest
biomass.  The report suggests that a forested filter strip should be twice as long as a grassed
one, however, because less vegetative cover is available in the forest strip.
Yousef et al (1985) also commented on vegetative cover in a grassed swale.  In their
study, a thick grass cover (80% grass, 20% bare soil) was found to have reduced nutrient
removal efficiencies when compared to a thin grass cover (20% grass, 80% bare soil).  This
was attributed to increased decay of organic matter where thick grass cover was available.
Some studies commented on the effects of season on vegetated BMP effectiveness.
Barrett et al (1995b) cites one study that expresses concern over reduced grassed swale
effectiveness during times of summer drought, when vegetation can die or become dormant.
The Seattle Engineering Department (1993) attempted to investigate the effects of season on
4
a grassed swale, but insufficient data was collected to determine seasonal variations in
removal.  Yousef et al (1985) attributed a decline in removal effectiveness of organic
nitrogen in a swale to increased organic debris that exists during periods of grass growth.
Glick (1993) investigated the effect of vegetated buffer strip width, or length of the
strip in the direction of flow, on pollutant removal.  Increased width was found to increase
pollutant concentrations, rather than decrease them, as other researchers have reported.  The
increased concentrations were attributed to detachment of pollutants contained in the strip.
The Municipality of Metropolitan Seattle (1992) consolidated the effects of factors
such as swale slope, width, length, flow velocity, and contributing watershed area by
recommending a swale hydraulic residence time.  Two swale configurations were
investigated, one with a 4.6 minute residence time, and one with a 9 minute residence time.
The study suggests that a swale with 4.6 minute hydraulic residence time is not adequate to
assure adequate removal of constituents, but that the 9 minute configuration resulted in more
consistent removal efficiencies, on the order of 83% for TSS.  The study recommends further
investigation before residence times shorter than 9 minutes can be used with confidence.  No
laboratory studies have been performed to carefully identify the effects of various factors,
such as season, length, and water depth, on vegetative BMP efficiency.  This study uses a
controlled environment for measuring these effects.
2.2 Vegetated Buffer Strip Treatment Effectiveness
Most research on vegetated buffer strips has focused on the removal efficiency for
filter strips in agricultural situations.  Only in a few recent studies has their ability to treat
urban runoff been documented, and the results vary.
Schueler et al (1992) cites only two monitoring studies of filter strips in urban areas.
The studies indicates that filter strips do not trap pollutants efficiently in urban areas due to
high runoff velocities; one of these studies indicated a removal rate of 28% for TSS.  The
Schueler report does say that filter strips can effectively remove sediments, organic material,
and trace metals in areas where runoff velocity is low to moderate.  It recommends a
5
maximum flow velocity of 0.76 m/s.  The ability of filter strips to remove soluble
constituents, such as nutrients, is reported as variable.  Design guidelines include minimum
filter strip width of 15 meters, use of a level spreader device to distribute flow evenly, regular
removal of accumulated sediment, and slopes less than five percent.
Yu et al (1995) report removal efficiencies for a vegetated buffer strip treating
highway runoff as 64% for TSS, 59% for chemical oxygen demand (COD), -21% for TP, and
88% for zinc.
Young et al (1996) cites a 1994 study which reports 70% TSS, 40% particulate
phosphorus and zinc, 25% lead, and 10% nitrate/nitrite removal efficiencies for a filter strip.
It recommends that slopes of filter strips be less than 15 percent to prevent the formation of
gullies in the strip, use of a level spreading device for even distribution of runoff, and dense
vegetation.  Furthermore, the report cites a 1995 study which recommended filter strips only
for roadways with a maximum of 2 lanes and roadway average daily traffic (ADT) of less
than 30,000 vehicles/day.
Table 2.1.  Summary of previous filter strip studies.
Study Notes Removal Efficiencies
Schueler et al (1992) recommended velocity
<0.76 m/s, length>15 m
28% TSS
Yu et al (1995) specifically hwy runoff 64% TSS; 59% COD;
-21% TP; 88% zinc
Young et al (1996) efficiencies from cited study 70% TSS, 40% P, Zn;
25% Pb; 10% NO
3
/NO
2
Previous research on vegetated buffer strips used to improve highway runoff quality
is sparse.  Important conditions such as climate, vegetation, size and geometry of the filter
strip, size of the highway, and soil type vary from study to study, making results from one
investigation difficult to extrapolate to other conditions.  Additional research is necessary to
6
determine identify the expected removal efficiencies for vegetated filter strips treating
highway runoff under a variety of conditions.  The conditions of this research that might
make it notable from other urban filter strip research include the following:
? Climate.  Austin, Texas, has hot summers and mild winters, with moderate
average rainfall (83 cm/yr);
? Land use.  The source of runoff for the filter strips in this study is restricted to
highways only.  A highway provides a small watershed area in comparison to the
watersheds for urban-area filter strips in other studies;
? Vegetation.  The vegetation of the filter strips used in this study are common to
Texas, and in particular, are commonly used by TxDOT for seeding of roadside
areas.  In addition, the two monitored filter strips have different vegetation types
(one mixed, one mostly Buffalo grass);
? Geometry.  The two monitored filter strips are the sides of V-shaped highway
medians that were not originally designed for water quality enhancement.  These
filter strips are relatively short, with average length from pavement to median
center of 7 to 9 meters.
? Extent of monitoring.  Often, studies of BMPs present removal efficiencies from
individual storms or average removal efficiencies from perhaps three to five
storms.  The high variance in constituent concentrations and other conditions from
storm to storm can unfairly bias results for shorter studies.  This study finds
average removal efficiencies over a period of at least 14 months, with multiple
storms (34 total events over all collection sites and a minimum of 19 storms
monitored at any one sampling location).  Monitoring of many storm events
ensures the reliability of results by minimizing effects of data outliers that can
strongly influence removal efficiency calculations in stormwater studies.
7
2.3 Effect of Metals Deposition on Vegetated BMPs
Several previous studies have shown that most metals in urban runoff are primarily
found in a particulate, insoluble form.  Barrett et al (1995b) refers to one study where the
particulate fractions of lead, copper, and cadmium in urban runoff were respectively 90%,
75%, and 57%.  Wiginton et al (1996) found that less than 2% of cadmium, lead, copper, and
zinc in urban runoff were leachable and that much of the total mass of metals in urban runoff
is sorbed onto soil components such as clays, organic matter, and hydrous oxides.  Hence,
only the small soluble portion of metal mass deposited onto vegetated buffer strips is likely
to pose a risk to plants, animals that eat the plants, and groundwater resources.  A large
fraction of metal mass deposited on a vegetated buffer strip is bound to solids in the runoff
and deposited in nearby soils in an insoluble form.
Previous research on metals accumulation in roadside vegetative areas has focused on
identifying increases in metals concentrations in soil and in plant and animal life near
highways.  It is clear from numerous studies that atmospheric deposition results in elevated
concentrations of metals including lead, zinc, cadmium, and chromium in roadside soils
(Lagerwerff and Specht, 1970; Gish and Christensen, 1973).  Only a few studies, however,
have examined an increase in metal concentrations near roadways as a result of runoff, rather
than by atmospheric deposition.  In general, the studies indicate significant accumulation of
metals in soils near the surface.  Howie and Waller (1986) found elevated levels of iron, lead,
and zinc in the first foot of soil in a swale accepting runoff from a highway.  Gish and
Christensen (1973) found levels of Cd, Ni, Pb, and Zn in earthworms and soils at one of their
sites that were elevated beyond that which could be attributed to atmospheric deposition.
They attributed the elevated concentrations to metals-rich runoff from several roadways
which drained over and deposited metals at the site.  Wigington et al (1986), however,
concluded that there was no statistical evidence of metal accumulation due to urban runoff
above that deposited by air pollutants at the highway site studied.
Barrett et al (1995b) summarized the results of numerous studies that looked at the
impact of highways on metals accumulation in groundwater.  The Barrett study concluded
that highway runoff can have a significant impact on groundwater in some situations, but
8
natural processes occurring in soils will attenuate metals in highway runoff prior to reaching
the groundwater.  One cited study found zinc concentrations in groundwater wells near a
highway as high as 220 ug/L but concentrations in wells further from the highway were
almost always below 50 ug/L.  Another study cited in the Barrett report, however, found high
concentrations of metals, including 1000-6600 ug/kg lead and 490-2400 ug/kg iron, in the
top 15 cm of soil underneath a highway swale, but nearby groundwater was unaffected.
Lagerwerff and Specht (1970) and Waller et al (1984) found decreasing metal concentrations
with increased soil depth near highways and expected limited downward movement of metals
in soils.
Some studies have documented metals accumulation in roadside areas due to
deposition from highway runoff.  None, however, have assessed the risks to human health
and the environment associated with such deposition.  More investigation is required to
assess these risks and to understand whether metals deposition in vegetated BMPs can cause
environmental or health problems.
2.4 Grassed Swale Treatment Efficiency
The benefits of roadside grassed swales for improvement of runoff water quality and
prevention of erosion were recognized in the early 1980s.  Yousef et al (1985) studied two
grassed swales, 53 and 90 m long, for the removal of nitrogen, phosphorus, and heavy metals
in highway runoff.  Results showed the swale had moderate to high removal efficiencies (29-
91%) for metals, but nitrogen and phosphorus concentrations were often higher after runoff
had passed through the swale.  When infiltration of pollutants into the soil was considered,
however, less mass of both metals and nutrients reached receiving waters because of the
swale.
Schueler et al (1992) reported varying removals of sediments and metals in urban
runoff by grassed swales.  However, the study states that a well-designed and maintained
swale could be expected to remove 70% or total suspended solids (TSS), 30% of total
phosphorus (TP), 25% of total nitrogen, and 50-90% of trace metals.  Swales were
9
recommended as a BMP to be used in conjunction with other BMPs.  Cost and maintenance
requirements were stated as low.
The Municipality of Metropolitan Seattle (1992) conducted an extensive study on a
60 meter grassed swale which treated runoff from a residential area.  The swale showed 83%
reduction for total suspended solids (TSS), 63-72% for metals, 65% for turbidity, and 74%
for oil and grease.  Moderate (up to 40%) to negative removals were seen for nitrogen and
phosphorus, and a high variation was seen for removal of fecal coliform bacteria.
The Seattle Engineering Department (1993) studied a 173 meter long swale which
also treated runoff from a residential area.  The study showed that concentrations of TSS and
most metals at the swale effluent were 60-70% less than influent levels, but nutrient
concentration reductions were less than 40% and fecal coliform reductions were negative.
Table 2.2.  Summary of grassed swale removal efficiencies.
Study Notes Removal Efficiencies
Yousef et al (1985) 53 to 90 m swale; hwy
runoff
29-91% metals; N, P conc. increased
in swale
Schueler et al (1992) Expect 70% TSS; 30% TP; 25% TN;
50-90% metals
Mun. Met. Seattle
(1992)
60 m swale; residential
area
83% TSS; 63-72% metals; 65%
turbidity; 74% O&G
Seattle Egr. Dept.
(1993)
173 m swale;
residential area
60-70% TSS, metals; 40% nutrients;
neg. bacteria
Yousef et al (1985) recommended swales of minimal slope to increase contact time;
soils with high infiltration rates, for maximum reduction of pollutant loadings to receiving
waters; earthen cross barriers to increase retention and infiltration; and, removal of grass
clippings and debris from the swale.  Dorman et al (1988) prepared extensive design
guidelines for grassed swales and filter strips for the Federal Highway Administration
(FHWA), based on vegetation development and expected flow rates.  Scheuler et al (1992)
10
warned that swales cannot control runoff effectively if flow velocity exceeds 0.46 m/s.  The
report also recommended long contact times, minimum grass height of 6 inches, and regular
mowing of the swale.  The Municipality of Metropolitan Seattle (1992) suggested that
pollutant removal in a swale is fundamentally dependent on the residence time in the swale,
thus combining the effects of factors such as swale width and length, flow depth, volumetric
flow rate, slope, and vegetation characteristics.  The study recommended a 9 minute
minimum residence time in a swale to achieve 80% removal of suspended solids.  The study
also recommends a maximum velocity less than 0.27 m/s, slope between 2 and 4 percent,
water depth less than one half the height of the grass, and regular mowing.
11
Chapter 3 Channel Swale Experiments
3.1 Introduction
Construction of a grassy swale in the laboratory was deemed an ideal method for
investigating the influence of individual parameters on swale efficiency.  The swale allowed
for repeated experiments with one varying parameter, thus demonstrating the effect of that
factor on swale efficiency.  The effect of water depth, season, and length of swale were
investigated in this manner in these experiments.  Overall efficiency of the laboratory swale
was also investigated.
3.2 Methods and Materials
A grassed-lined channel was constructed at the Center for Research in Water
Resources (CRWR) located on the J.J. Pickle Research Campus of the University of Texas in
Austin, Texas.  The soil and grass were installed during May and June of 1996 in a steel
flume that was constructed in the 1960s.  Eleven experiments were performed in the swale
from October 1996 to May 1997.
3.2.1 Setup
The steel flume has a U-shaped cross-section with square corners (Figure 3.1).  The
flume bottom is 0.76 m wide and its walls are 0.61 m tall.  The flume contains 7.6 cm of soil
and gravel, a layer of plastic sheeting, an underdrain pipe, 5.1 to 7.6 cm of clean gravel, a
fiberglass screen, and 15 to 17.8 cm of topsoil that was sodded with Buffalo grass.  Buffalo
grass is common in the Austin area and has been used by TxDOT along highway medians in
Austin.  This grass is a short, hardy, turf grass that is drought tolerant and requires little
mowing.  The grass sod was approximately 1.3 cm thick and the height of the grass was
approximately 8 cm high at the time of planting. A perforated PVC pipe was laid as an
12
underdrain along the length of the flume, lying along the swale centerline and on top of the
plastic sheeting.  The plastic sheeting was placed with a V-shaped cross-section that forced
water to the underdrain.  The fiberglass screen supports the topsoil and prevented soil from
entering the underdrain.
61 cm
76.2 cm
7.6 cm
5.1 to 7.6 cm
15 to 17.8 cm
1.3 cm
7.6 cm
Existing dirt/gravel Plastic sheeting
Perforated PVCClean gravel
Fiberglass screening
Topsoil Grassy sod
Y max
Figure 3.1. Cross-section of channel swale.
The swale was 40 m long and the average slope was 0.44%.  Holes were drilled in the
swale bottom at the swale influent (0 m) and at 10, 20, and 30 m along the length.  ?? PVC
pipes were installed vertically through these holes to the sod surface.  Ball valves were
installed at the end of the pipes (Figure 3.2).  These pipes allowed for easy sampling of water
passing over the grassy swale at any time at 10, 20, and 30 m from the inlet.  At 40 m, a steel
barrier was anchored to the flume to keep the gravel, topsoil, and sod in place.  A weir is cut
into the center of the barrier to allow discharge for the swale effluent.  Water collected at the
weir represented water quality at 40 m.  The underdrain extends through the barrier through a
90 degree elbow for easy sampling for water quality analyses.  Rulers were fastened to the
side of the flume at 0, 10, 20, and 40m for monitoring of water depth.
13
10 m 10 m 10 m 10 m
90 degree elbow
Steel barrier Flume
Vertical PVC pipes
for taking samples
Figure 3.2.  Overview of PVC Pipe Locations.
Simulated highway runoff flowed down the length of the swale during experiments.
Water for the runoff originated in an open brick-lined common reservoir at CRWR (Figure
3.3).  The water was continually pumped during experiments to a constant head reservoir.
Overflow from this reservoir returned to the common reservoir.  The discharge from the
constant head reservoir then entered the first of two steel basins.  A valve regulated flow to
this basin.  Water flowed from the first basin over a V-notch weir into a mixing basin.  Flow
was monitored by reading the depth of water behind the weir.
Figure 3.3.  General flume apparatus.
The 61 cm x 51 cm (plan view) mixing basin was continually mixed using an
approximately 30 cm blade.  A mixture of synthetic, concentrated highway runoff (?cocktail?
14
described below), was continually pumped to the water that discharged over the weir and into
the mixing basin.  In the basin, the water was completely mixed with the cocktail, and the
water exiting the mixing basin effectively simulated highway runoff.  Water exited the basin
through a perforated baffle into the channel.
The influent water flowed over 1.22 m of plastic sheeting covered with 8-10 cm rocks
to create an evenly distributed flow.  The grass area begins immediately after the plastic
sheeting, where the first vertical sample pipe is located.  Occasional weeds were allowed to
grow among the Buffalo grass.  The grass was not mowed or weeded throughout the
experiments.  During a cold spell, 12 lightbulbs were suspended along a PVC frame just
above the grass.  This frame and the channel were wrapped in several layers of clear plastic
in an effort to prevent freezing of the suspended flume during the winter.  The wrapping was
kept on only a few weeks; the frame and lightbulbs were left in place for the duration of the
experiments.
3.2.2 The Cocktail
The highway runoff ?cocktail? is synthetic highway runoff prepared onsite.  The
cocktail was made in a concentrated form that, when diluted with the appropriate amount of
water, was representative of the average water quality of runoff from highways in Austin.
Dulay (1996) developed this cocktail.
The post-dilution desired concentrations of the added constituents, as well as the mass
of constituents used in these experiments for dilution in 5000 gallons of well water, are listed
below in Table 3.1.
15
Table 3.1.  Ingredients in the cocktail.
Necessary Post-
Dilution
Concentration
Mass Required for
dilution in 5000
gallons
Mass Used After
Experiment 3
Constituent Added mg/L g g
detention pond sediment 500 20 lb 10 lb
Gleason clay 40 800 400
Velvacast kaolin 60 1200 600
coarse clay 20 400 200
Pb(NO
3
)
2
0.16 3.03 3.03
Cu(NO
3
)
2
 3H
2
0 0.113 2.16 2.16
Zn(NO
3
)
2
 6H
2
0 0.91 17.22 17.22
Na
2
CO
3
0.9 17.04 17.04
The constituent masses listed in Table 3.1, when diluted, approximate the suspended
solids, nutrients, and metals in highway runoff in the Austin, Texas area.  The following
items should also be noted about the cocktail:
? The sediment was collected from the bottom of a local detention pond used solely for
treating highway runoff and only that portion which passed through the 250 micrometer
(mesh #60) sieve was used.
? Constituents such as chemical oxygen demand (COD), total organic carbon (TOC), and
total phosphorus (TP) were not added separately but were associated with the detention
pond sediment or were present in the reservoir water.
? Na
2
CO
3
 was added to provide the appropriate distribution of small, medium, and large
particles, that are contained in highway runoff.
? Iron nitrate (Fe(NO
3
)
2
 9H
2
0), though prescribed in the original cocktail recipe, was not
added in any of these experiments since sufficient iron was provided from rust in the
basins and tanks prior to the swale.
? After experiment 3, the dose of detention pond sediment and all three clays was halved in
16
order to lower the TSS concentration to levels seen in the field.  This reduction is noted in
Table 3.1.
3.2.3 Experiment Procedure
The concentrated highway runoff cocktail was prepared by mixing continuously
several gallons of untreated well water while the detention pond sediments, Na
2
CO
3
, and
metal nitrates were added (Figure 3.4).  The stirring was continued for at least ? hour.
Figure 3.4.  Addition of sediment and clays for creation of highway runoff cocktail.
The cocktail was kept continuously stirred during the experiment, and the cocktail
bucket was turned regularly during the experiment to prevent sediment buildup in bucket
corners opposite the stirrer.  Weather conditions and a description of the grass appearance
were reported prior to each experiment.
A pump was used to deliver the concentrated cocktail to the mixing basin.  Reservoir
water was pumped into the constant head reservoir and flowed into the first basin.  The
cocktail pump was started when the reservoir water flowed over the weir between the first
17
basin and the mixing basin.  The pump was calibrated and was set at a rate such that the
cocktail would be used up when 5000 gallons of water had passed over the weir.  The depth
of water behind the weir had been decided upon prior to the experiment, depending on the
desired water depth in the swale.  This depth was constant throughout the experiment.  The
flow rate is
Q = 365 h
2.43
where
Q = flow rate (L/s);
h = head on the weir (m).
One to three sets of samples were taken simultaneously along the length of the swale
to determine changes in concentration along its length (Figure 3.5).  Sample sets were
collected at 5 minute intervals during an experiment.  Water was flushed through the vertical
sample pipes prior to sampling.  To avoid variations during the initial flow, no sample was
collected until the flow reached a quasi steady-state.  Steady-state was determined by
monitoring the water depth at 30 or 40 m, or by monitoring the distance the water flowed.
Steady state had been reached when either remained constant.  Water depth was recorded
using the fixed rulers at 0, 10, 30, and 40 m after steady state had been reached.
18
Figure 3.5.  Appearance of the swale during an experiment.
Time was also recorded during each experiment.  The moment that water exited the
mixing basin and entered the swale was considered time = 0.  The time of each water depth
measurement, weir height measurement, and sample set was recorded.
Underdrain flow also was monitored for some experiments after steady-state
conditions were reached.  Underdrain flowrate was measured using a volume-calibrated
bucket and a stopwatch.
Each sample was collected in 4 separate bottles, and preserved for the analyses to be
performed.  A total of 109 samples were collected during the 11 experiments.  The samples
were logged and preserved in the laboratory at CRWR.  All laboratory analyses were
performed at CRWR.  The constituents that were analyzed for all experiments were total
suspended solids (TSS), turbidity, fecal coliform, fecal streptococcus, chemical oxygen
demand (COD), total organic carbon (TOC), nitrate (NO
3
), total Kjeldahl nitrogen (TKN),
total phosphorus (TP), zinc (Zn), lead (Pb), iron (Fe), and copper (Cu).  The analytical
methods used for determining sample concentrations are listed in Table 3.2.  Note that a
bacterial analysis was not done for the channel experiments, but was included in the field
19
experiments.
Table 3.2.  Analytical methods for sample analysis.
Constituent Method Identification Holding
Times
Preservative
TSS Std. Methods 18
th
 ed. 2540 B 7 days None
Turbidity Std. Methods 18
th
 ed. 2130 B 24 hours None
Fecal coliform Std Methods 18
th
 ed. 9222 D 24 hours None
Fecal strep Std Methods 18
th
 ed. 9230 C 24 hours None
COD Std Methods 18
th
 ed. 5220 D 3 months H
2
SO
4
TOC Std Methods 18
th
 ed. 5310 B 28 days H
2
SO
4
Nitrate Std Methods 18
th
 ed. 4500-NO
3
-D 24 hours None
TKN EPA 351.4 28 days H
2
SO
4
Phosphorus EPA 365.3 28 days H
2
SO
4
Metals ICP Method 6010 6 months HNO
3
3.3 Experiment Philosophy
The channel allowed for investigating five aspects of grassy swales: the effect of
water depth, season of the year, and swale length on removal efficiency, the effect of
highway runoff on groundwater, and the capability of the swale to reduce constituent
concentrations in highway runoff.  These five aspects were chosen for investigation because
water depth, season, and length could be varied easily in the channel, and underdrain water
quality should reflect the ability of soil to treat highway runoff after travel through
approximately 24 cm of soil and gravel.  Effects of the length of swale were evaluated
concurrently during water depth and seasonal experiments.
20
3.3.1 Water Depth
Water depth can hinder the mechanisms of removal of constituents from runoff that
flows over grassed swales.  Filtration by the grass, impedance and increased sedimentation,
and biological activity on grass blades were expected to be less effective at removing
constituents in deeper water.
Four water depths were used: 3 cm, 4 cm, 7.5 cm, and 10 cm to cover the range of
depths observed in swales in the field.  Infiltration at water depths less than 3 cm prevented
water in the swale from reaching the 20 m sampling tube.  The four water depths were
associated with four different flowrates as measured by the depth of the water behind the
weir in the pre-mixing basin.
The data analysis for determining the effect of water depth utilized at least two
sample sets at each depth, and with the exception of the 10 cm depth, each water depth was
investigated in at least two separate experiments.  Seasonal effects were assumed to have a
negligible effect on the water depth analysis.  For example, experiments 6 and 11 were
performed at the same water depth at different seasons, and results for that water depth are
averaged over both experiments.
3.3.2 Season
The effect of season on the grassed swale?s removal efficiency was investigated.
Examples of potential seasonal changes in the swale?s characteristics include increased grass
blade density during growth seasons, increased nutrient uptake rate during growth seasons,
decreased nutrient and organic removal during plant decay, and increased infiltration during
dry seasons due to an increase in permeability and decrease in soil saturation.
The seasonal analysis was performed by comparing the constituent removal capability
of the swale during the dormant and growing seasons of the Buffalo grass. Experiments
during the dormant season were begun with experiment 4 on December 13, 1996, during
which time the grass appeared greenish brown to brown and dry.  Experiments were given
the ?growing? season designation once green, healthy grass from the new growing season
21
had sprouted in significant number and density.  This occurred in mid to late March 1997,
and experiments 8-11 thereafter were considered growing season experiments.  Experiment 7
occurred during the transition period between dormancy and growth and is not included in
the seasonal analysis.  The effect of season was investigated at two water depths, 4 cm and
7.5 cm, by repeating experiments at those water depths in the dormant and growing seasons
and comparing the effectiveness of the swale during those seasons.
3.3.3 Length of Swale and Groundwater Quality
Investigating the efficiency of the swale for various lengths, as well as sampling
underdrain water quality, was performed for every experiment.  Sampling from the
underdrain was useful because underdrain water quality simulates water quality of recharge
to groundwater from swales treating highway runoff in the field.  Also, patterns in underdrain
water quality through multiple experiments after construction of the swale may reflect
changes in the capability of soils at field projects to filter infiltrated runoff after construction
phases are completed.
3.3.4 Schedule and Experimental Conditions
A summary of the schedule, water depths, season, and furthest sampling distance for which
samples could be taken (a function of how far the runoff traveled in the flume before
infiltrating completely) are provided in Table 3.3 below.
22
Table 3.3.  Summary of lab experiments.
Water No. of Sample Furthest Sample
Depth Sets Taken Along Distance
Exp. No. Date cm Season Swale Length m
1 10/16/96 10 * 1 40
2 10/23/96 10 * 1 40
3 11/20/96 10 * 1 40
4 12/13/96 7.5 dormant 2 40
5 1/22/97 7.5 dormant 1 40
6 1/31/97 4 dormant 2 40
7 3/13/97 3 * 3 20
8 4/30/97 7.5 growth 2 40
9 5/13/97 3 * 3 10
10 5/19/97 10 * 2 40
11 5/22/97 4 growth 2 20
* indicates the experiment was not included in determination of dormant or growing season
removal efficiencies.
3.4 Efficiency Calculations
Removal efficiencies for a constituent were calculated with respect to the
concentration of that constituent sampled at 0 m.  The following equation was used to
calculate removal efficiency:
E = %100
0
0
?
?
C
CC
x
where
E = removal efficiency (%),
C
x
 = concentration of constituent sampled at distance x down the swale (mg/L or
NTU),
23
C
0
 = concentration of constituent sampled at the 0 m sampling tube (mg/L or
NTU).
Analyses of multiple sample sets and multiple experiments were required to calculate
average removal efficiencies.  An average removal efficiency at a specific water depth was
calculated using the following steps:
1. For each experiment at a water depth, the average C
0
 was calculated by
averaging all sample concentrations taken at the 0 m location, if more than
one was taken.
2. Removal efficiencies for each sample were calculated using the average C
0
 for
that experiment from step 1 and the equation above.
3. The average removal efficiency at each sampling distance was calculated by
an average of all removal efficiencies for samples at that distance from step 2.
The average removal efficiency for a water depth at a particular distance along
the swale is from all experiments at that water depth.
For seasonal analysis, the average removal efficiency during dormant and growth seasons
were calculated for two water depths.  A season's average removal efficiencies were
calculated using the following steps:
1. For each experiment during a season at a particular depth, the average C
0
 was
calculated by averaging all sample concentrations taken at the 0 m location.
2. Removal efficiencies for each experiment at each sampling location were
calculated using the average C
0
 from step 1 for that experiment and the
equation above.
3. The removal efficiency at each sampling distance for a season and water depth
was calculated by an average of all removal efficiencies found at that distance
from step 2 over all experiments at that water depth and during that season.
24
The removal efficiencies reported for the swale in this report represent the reduction
in pollutant concentrations that occurred in runoff along the swale.  To calculate a reduction
in pollutant mass, rather than concentration, infiltration of contaminants into the soil must be
taken into consideration.
The concentrations of constituents at the swale influent must remain constant
throughout the experiment in order to insure meaningful results.  For example, if the cocktail
pump was clogged temporarily during an experiment, this would prevent some influent water
from receiving the appropriate amount of constituents.  Hence, water sampled at some
locations might be cleaner than expected and falsely indicate that removal had occurred.  In
order to verify that the influent concentration was constant, multiple samples were taken at 5
minute intervals at the influent sample pipe during experiment 3 and experiment 5.
Experiment 3 results were inconclusive; TSS levels in the 4 samples taken at 0.0 m were 440,
624, 474, and 678 mg/L.  A more extensive, 6-sample test was done during experiment 5
after adding a small filter on the end of the cocktail influent tube to prevent grass from
entering the cocktail pump.  These results showed that the influent concentrations to the
swale are relatively constant, as shown in Figure 3.6.
The constituent concentration was assumed to be equal to the detection limit when
detection limits were encountered in the data.  This policy was chosen because it tended to
give the most conservative removal efficiencies; higher concentrations in the sampled runoff
result in the lower, or more conservative removal efficiencies.
Data from experiments 1-3 were not used because the mass of cocktail ingredients
changed.  The use of less solids in experiments 4 through 11 could bias the calculated
removal efficiencies, especially by increasing removal efficiencies for experiments 1-3 in
which more sediment and clays were used.  This would render a comparison between
experiments 1-3 and subsequent experiments impossible.  A preliminary analysis of the data
observed in experiment 1 through 3 was performed and the removal efficiencies in these
experiments were higher than in subsequent experiments with comparable water depths.  An
exception to this rule was made for analyzing underdrain water quality, for which results
25
from all experiments were used; changes in sediment used likely has a small impact on
percolate water quality.
A table with data for all constituents for all experiments is provided in Appendix A.
0
50
100
150
200
250
0 5 10 15 20 25 30
Elapsed time, min
TSS conc., mg/L
Figure 3.6.  Confirmation of relatively constant influent concentrations in the swale.
3.5 Channel Experiments Results
3.5.1 General Results
Suspended solids and metals demonstrated the highest removal efficiencies in the
swale, with reduction in constituent concentrations varying from 51 to 86 percent after 40 m
of treatment.  Removal of COD was ranged from 25 to 79 percent, and removal of nitrate,
26
TKN, and TP ranged from ?26 to 45 percent, after 40 m of treatment.  The ranges of
pollutant removal efficiencies for all constituents are listed below in Table 3.4.  Ranges
represent efficiencies observed during experiments at different water depths.  The calculated
removal efficiencies agree well with grassed swale field results reported by other researchers
(Barrett et al, 1995b; Municipality of Metropolitan Seattle, 1992).
Table 3.4.  Removal efficiencies calculated for the channel swale at different water
depths.
Constituent 10 20 30 40 Underdrain
TSS 35-59 54-77 50-76 51-75 73-87
COD 13-61 26-70 26-61 25-79 39-76
Nitrate (-5)-7 (-5)-17 (-28)-(-10) (-26)-(-4) (-8)-(-10)
TKN 4-30 20-21 (-14)-42 23-41 24-41
Total phosphorus 25-49 33-46 24-67 34-45 55-65
Zinc 41-55 59-77 22-76 66-86 47-86
Iron 46-49 54-64 72 76 75
Distance along swale, m
3.5.2 Effect of Water Depth on Swale Removal Efficiency
Average removal efficiencies at different water depths for the monitored constituents
are presented in Figure 3.7 through Figure 3.13.  The data in the graphs indicates that
constituent removal efficiencies were reduced as water depth was increased, with the
exception of nitrate and TKN.  No trend is obvious for the relationship of water depth and
removal efficiency for nitrate and for TKN.  The data presented in Figure 3.7 indicates that
removal of total suspended solids increased with decreased depth of water.  However, the
difference in average removal efficiencies for TSS at 20 m for different water depths was not
statistically significant among all adjacent (3 and 4 cm, 4 and 7.5 cm, etc.) water depths at
27
the 90% confidence level.  This statistical analysis was performed for the data at 20 meters
because the runoff at a water depth of 3 cm did not reach the 30 meter sample tube.
0
10
20
30
40
50
60
70
80
90
10203040
Distance along swale, m
Removal efficiency, %
WD=3cm
WD=4cm
WD=7.5cm
WD=10cm
Figure 3.7. Effect of water depth and swale length on TSS removal efficiency.
The increase in removal efficiency of TSS with decreased water depth confirms
expectations, since the filtration action of the grass blades is expected to be more significant
for smaller water depths.  However, the flow velocity in the swale was higher during
experiments at deeper water depths.  It is likely that the increased removal efficiency in
shallower water is influenced both by the water depth and its velocity.  Thus,
recommendation of a maximum water depth for a swale based upon desired removal
efficiency requires a simultaneous limitation on runoff velocity within the swale.  These
results do indicate that a grassed swale which treated slow moving, shallow (3-4 cm) runoff
will achieve higher removal efficiencies for most constituents of concern than swales with
deeper (7.5-10 cm) runoff at higher velocities.  The trend between water depth and removal
efficiency for COD, nitrate, TKN, total phosphorus, zinc, and iron are presented in Figure 3.8
28
through Figure 3.13.  For COD, total phosphorus, and iron, the trend of increased removal
efficiency with decreased depth is apparent; for nitrate, TKN, and zinc, the trend is not
certain.  The solubility of nitrate and TKN decreases the swale?s capability for increased
filtering action at lower water depths.  The lack of a trend for zinc, however, is difficult to
explain.  Zinc is associated with sediments in the runoff, and its removal efficiency often
simulates trends in sediment removal.  Thus, the inverse relationship between removal
efficiency and water depth would be expected for zinc; however, this trend is not apparent.
More experiments could explain this result.
0
10
20
30
40
50
60
70
80
90
10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
WD=7.5cm
WD=10cm
Figure 3.8. Effect of water depth and swale length on COD removal efficiency.
29
-35
-30
-25
-20
-15
-10
-5
0
5
10
15
20
10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
WD=7.5cm
WD=10cm
Figure 3.9. Effect of water depth and swale length on nitrate removal efficiency.
-20
-10
0
10
20
30
40
50
10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
WD=7.5cm
WD=10cm
Figure 3.10. Effect of water depth and swale length on TKN removal.
30
0
10
20
30
40
50
60
70
0 10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
WD=7.5cm
WD=10cm
Figure 3.11. Effect of water depth and swale length on total phosphorus removal.
0
10
20
30
40
50
60
70
80
90
100
0 10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
WD=7.5cm
Figure 3.12. Effect of water depth and swale length on zinc removal efficiency.
31
0
10
20
30
40
50
60
70
80
90
10203040
Distance along swale, m
Removal efficiency, %
WD=3m
WD=4cm
Figure 3.13. Effect of water depth and swale length on iron removal efficiency.
3.5.3 Effect of Swale Length on Removal Efficiency
Figure 3.7 through Figure 3.13 above can also be used to evaluate the effect of swale
length on removal efficiency.  The data in the graphs shows that removal efficiency increases
with length, but the increment of increased efficiency diminishes as runoff proceeds further
down the swale.  This trend is especially evident for total suspended solids, chemical oxygen
demand, total phosphorus, and metals.  The majority of total removal occurs in the first 20
meters of flow over the swale for these constituents.  The removal of total suspended solids
after 20 meters accounts for 92%, 80%, and 105% of the total removal observed at 40 meters
at water depths of 4, 7.5, and 10 cm, respectively.  The 105% at the 10 cm water depth
indicates that removal at 20 meters was actually higher than the removal observed after 40
meters of flow.
The diminishing increases in removal efficiency observed after 20 m indicates that
swales longer than 20 m may not be cost-effective.  This is particularly true in situations
where construction or maintenance of a swale longer than is costly, such as in areas where
32
land is expensive or where considerable excavation or landscaping is required for swale
construction.  If expected water depths in the swale are 7.5 cm or greater, however, 30 to 40
meter-long swales are necessary for TSS removals of greater than 60%, assuming a thick
vegetated cover exists on the swale.
The diminishing increases in removal efficiency observed as swale length increased
confirms intuition.  Many constituents in highway runoff are attached to sediments and clays
that settle out or are filtered out quickly once the runoff enters the swale.  More soluble
constituents and constituents attached to smaller particles which do not settle quickly are not
removed effectively in the swale?s initial 20 meters, as demonstrated by the removal data for
nitrate and TKN.  Three visual observations from the channel study are testament to this
phenomenon.  First, sediments accumulated on the blades in the first 10 meters of the swale.
The coating was obvious in the first 3 meters of grass, and could still be observed after 10
meters, but no sign of the coating was found at 20, 30, or 40 meters.  Second, layers of
sediment formed on the plastic sheet that covers the first meter of swale after runoff exits the
mixing basin.  The heaviest sediments fell out of suspension after less than one meter in the
swale, before any grass was reached, and formed these layers.  Finally, the height of the soil
surface with respect to the walls of the flume rose substantially after several experiments at
the 0 meter distance.  Deposited sediments raised the soil surface level at the swale influent
by approximately 2 centimeters after eight experiments.  No noticeable increase in soil
surface height occurred at distances of 10 to 40 meters.  The State of Maryland (1985)
recommends the periodic manual removal of sediment deposits to preserve the infiltration
capacity of the soil and to prevent ponding.  Removal of sediments also may prevent burying
of grass blades which can cause the grass to die and encourage channelization (Municipality
of Metropolitan Seattle, 1992).
The TSS removal efficiencies observed in these experiments may be used for design
purposes.  The data in Table 3.5 presents the length of swale necessary for a desired TSS
concentration removal efficiency at an expected water depth in the swale, assuming the swale
has a slope of 0.44% and thick, even vegetated cover with height of at least 10 cm.  Longer
swale lengths are necessary for swales of slopes greater than 0.44% or swales without thick,
33
even vegetated cover with vegetation height of at least 10 cm.
Table 3.5.  Required swale length for total suspended solids removal.
Expected
Water Depth Desired TSS Concentration Reduction
cm 30% 40% 50% 55% 60% 65% 70% 75% 80%
3 10101010 20 202020>20
4 10101020 20 203030>40
7.5 10 10 10 20 20 30 40 >40 >40
10 10 20 20 >40 >40 >40 >40 >40 >40
3.5.4 Effect of Seasons on Swale Removal Efficiency
Removal efficiencies for total suspended solids were greater in the growth season
than in the dormant winter season.  The growth season removal efficiencies for total
suspended solids are greater than dormant removal efficiencies at every sampling length, for
both water depths (Figure 3.14 and Figure 3.15).
34
0
10
20
30
40
50
60
70
80
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.14.  Seasonal comparison of TSS removal (water depth = 4 cm).
35
0
10
20
30
40
50
60
70
80
90
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.15.  Seasonal comparison of TSS removal (water depth = 7.5 cm).
The TSS concentrations observed for the two seasons were compared statistically.
The comparison shows that TSS removal efficiencies for the two seasons are significantly
different from each other at 40 meters for the 7.5 cm water depth and at 20 meters for the 4
cm water depth at the 90% confidence level.  This suggests that suspended solids is better
removed during the growing season.  On the other hand, zinc, which is often attached to
sediments in runoff, was demonstrated higher removal efficiencies during the winter season
for the 7.5 cm water depth (Figure 3.16).  There are no definitive seasonal differences for
zinc at the 4 cm water depth (Figure 3.17).
36
0
10
20
30
40
50
60
70
80
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.16.  Seasonal comparison of zinc removal (water depth = 7.5 cm).
37
0
10
20
30
40
50
60
70
80
90
100
10203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.17.  Seasonal comparison of zinc removal (water depth = 4 cm).
The higher removal efficiency for sediments in the growth season may be attributed
to the increased density of grass blades in the growth season.  During the growing season,
new green Buffalo grass grew alongside the dead, brown grass from the previous season.
The dormant Buffalo grass was shorter than the new growth of grass, and this dead grass
continued to shrink and decay throughout April and May of 1997.  The dead grass
nonetheless contributed to the overall grass blade density, thereby increasing the filtration
capability of the grass.  Some of the dormant undergrowth was no longer attached to the soil.
Much of the dead grass, however, was still anchored to the soil presumably by a remaining
root structure.  The previous generation of grass was still approximately 7.5 cm tall by the
end of April, at the beginning of the growing season experiments.  The new grass was 10-
12.5 cm tall at that time.  The shrinking dormant grass was still approximately 2.5 cm high
by the last experiment on May 22.
The decaying grass also may contribute nitrogen and phosphorus and organic
38
compounds to runoff passing through the swale.  Previous recommendations to remove grass
clippings from mowed swales are directed at reduction in nitrogen and phosphorus loads
(Municipality of Metropolitan Seattle, 1992).  Removal of the clippings prevents them from
decomposing in the swale.  Indeed, lower removal efficiencies for organic material, as
indicated by COD data, were observed in the growing season than the dormant season at the
7.5 cm water depth (Figure 3.18).  Analysis of COD was impossible at the 4 cm water depth
due to loss of a runoff sample.  However, neither nitrogen (Figure 3.19 and Figure 3.20) nor
phosphorus (Figure 3.21 and Figure 3.22) demonstrated lower removal efficiencies in the
growing season.  In fact, total phosphorus removal increased during the growing season.
0
10
20
30
40
50
60
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.18.  Seasonal comparison of COD removal (water depth = 7.5 cm)
39
-15
-10
-5
0
5
10
15
20
25
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.19.  Seasonal comparison of nitrate removal (water depth = 4 cm).
40
-60
-50
-40
-30
-20
-10
0
10
20
30
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.20.  Seasonal comparison of nitrate removal (water depth = 7.5 cm).
41
0
10
20
30
40
50
60
70
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.21.  Seasonal comparison of total phosphorus removal (water depth = 4 cm).
42
0
5
10
15
20
25
30
35
40
45
50
010203040
Distance along swale, m
Removal efficiency, %
Dormant
Growing
Figure 3.22.  Seasonal comparison of total phosphorus removal (water depth = 7.5 cm).
Runoff flowed further down the swale for some dormant season experiments than for
growth season experiments at the same water depth (Figure 3.21, etc.).  Increased grass blade
density may have the slowed the runoff down, which allowed the runoff to infiltrate at a
greater rate during growth season experiments than in dormant season experiments.
However, warmer, dryer weather may have dried out the soil in the spring, also encouraging
infiltration.  It is possible that the increased blade density in the spring enhanced detention of
the runoff, encouraging infiltration and removal of constituents from the surface runoff.
These results indicate that swales with sodded with Buffalo grass are effective at
removing runoff constituents during the dormant and growth seasons.  The shift to dormant
season did not have any obvious effect on the stiffness of the Buffalo grass blades.  The grass
blades continue to maintain height and some stiffness in the dormant season, even though the
grass was brown and dry.  Buffalo grass blade density does increase during the growing
season because of dormant grass remaining from the previous season.  This answers concerns
43
by a researcher cited in Barrett et al (1995b) regarding reduced efficiencies during vegetation
dormancy.  In fact, it may be during the growing season, when the previous season?s
vegetation is decaying, that removal efficiencies for organic compounds and nitrogen and
phosphorus are at their lowest.  Other grasses may lose their density and stiffness to greater
extents than Buffalo grass during dormant seasons.  If this is the case, seasonal impacts on
removal efficiency can be expected to be greater for these vegetation types.  A more
extensive study would be required to discover the seasonal impacts for various kinds of
grasses.
3.5.5 Underdrain Water Quality
The simulated highway runoff reached the underdrain by percolating through a top
layer of grass sod, 16 cm of topsoil, and 6 cm of gravel before entering the underdrain pipe.
Underdrain water quality was sampled for all 11 experiments except for experiments 7 and 9.
The underdrain analyses focused on two aspects of the underdrain water quality.
First, changes in underdrain water quality with time were investigated.  During
construction of the swale, the layers of soil were compacted by wetting the grass thoroughly
and walking over the sod several times.  However, a slow, additional compaction and settling
of topsoil likely occurred in the channel as a result of the percolation of water during the
experiments.  In addition, grass roots may have grown into the soil, filling cracks and pores
in the soil and taking up nitrogen and phosphorus and other constituents from the runoff as
the roots establish.  These changes may simulate similar changes that occur after construction
at sites in the field.  The compaction and root development can have an impact on the quality
of the underdrain water over time.
Secondly, average removal efficiency for water that entered the underdrain was
measured.  The underdrain water quality demonstrates the filtering capability of the soil, and
reflects water quality of recharge for groundwater in situations with shallow soils.
A steady decrease in the concentration of TSS in the water sampled from the
underdrain was observed through the first five experiments (Figure 3.23).  This reduction in
44
TSS concentrations may have been caused by an increase in the filtering capability of the
soil, and suggests that soil compaction may have occurred during the first five experiments.
This trend of increasing percolate water quality ends, however, after the first five
experiments, indicating that further compaction by the infiltrating water was minimal.
0
10
20
30
40
50
60
70
80
90
100
123456789101
Experiment number
Removal efficiency, %
Figure 3.23. Removal of TSS during infiltration in channel experiments.
There are two implications of this trend in suspended solids removal by the topsoil.
The first is that construction, which disrupts soil matrix by replacing a settled, stable soil
with loose, disjoint soil, can decrease groundwater quality by reducing the filtering capability
of the soil.  These effects have been documented by other researchers (Barrett et al, 1995b).
The second implication is that groundwater quality may increase significantly in the first 5
storm events after construction activities cease.  Constituents other than TSS, however, did
not demonstrate a decrease in concentration in underdrain water during the first five
experiments.  Turbidity (Figure 3.24) and total phosphorus (Figure 3.25), for example,
showed no recognizable trend in filtering capacity of the soil.  Zinc, whose removal is often
linked to removal of sediment, showed relatively constant removal via soil filtration over the
first 5 experiments (Figure 3.26).  This may indicate that construction has little effect on the
45
filtration capacity of soils for pollutants that are heavily associated with smaller particles,
such as many metals (Barrett et al, 1995b), or pollutants that are soluble.
0
10
20
30
40
50
60
123456789101
Experiment number
Removal efficiency, %
Figure 3.24. Removal of turbidity during infiltration in channel experiments.
0
10
20
30
40
50
60
70
80
90
100
123456789101
Experiment number
Removal efficiency, %
Figure 3.25. Removal of total phosphorus during infiltration in channel experiments.
46
0
20
40
60
80
100
120
123456789101
Experiment number
Removal efficiency, %
Figure 3.26. Removal of zinc during infiltration in channel experiments.
The underdrain water was used to calculate average removal efficiencies for the soil
during infiltration.  These removal efficiencies are listed in Table 3.6 below.  The average
removal efficiency of the soil was calculated using an average of removal efficiencies for
each constituent over all experiments, with the following exceptions.  Only experiments 5-11
were used to calculate a representative removal efficiency for TSS.  Also, data for metals
other than zinc is restricted to experiments 2 through 7 because of difficulties with analytical
equipment.
With the exception of nitrate, the removal of constituents during infiltration was at
least 37%.  The underdrain water quality was higher than the surface runoff after 40 meters
of treatment by the grassed swale.  The primary mechanism of removal for the percolated
runoff is filtration by the soil.  It is likely that a thicker layer of topsoil than the 16 cm of soil
used in these experiments would result in greater attenuation of pollutants.
47
Table 3.6.  Average removal efficiency for constituents based on underdrain water
quality.
Average Removal Efficiency
Constituent %
TSS 78
Turbidity 42
COD 49
NO
3
-45
TKN 37
Total phosphorus 65
Zn 80
Pb 41
Fe 74
3.5.6 Summary of Channel Swale Results
A grassed swale constructed in a steel channel removed over 50% of suspended
solids, zinc, and lead after 40 meters of treatment by the swale.  COD concentrations
decreased 25 to 79 percent after 40 meters of treatment, and reduction of nutrient
concentrations varied from negative to 45%.  In general, the majority of pollutant removal
occurred in the first 20 meters of swale.  Increasing the water depth and velocity of surface
flow of runoff in the swale reduced the removal efficiency of the swale.
More suspended solids were removed in the channel swale in the growing season than
in the dormant season.  During the growing season, new grass stood alongside dormant grass
which increased the grass blade density in the swale.  This increase in removal is attributed to
the combined filtering capacity of the dead material and live grasses.  The removal of
nutrients and organic material may decline in the growing season, when decay of vegetation
48
from the previous season contributes to the constituents in the runoff.
The concentrations of constituents in runoff that had percolated through the soil in the
swale were generally lower than the concentrations in surface runoff after 40 meters of
treatment by the swale.  However, the impact of swales on groundwater quality in the field
will vary with thickness of soil to groundwater, permeability of the soil, and the constituents
in the highway runoff.
49
Chapter 4 Field Experiments
4.1 Introduction
A primary objective of this study is measurement of the efficiency of vegetated buffer
strips for removing constituents in highway runoff in the Austin, Texas area.  The efficiency
of a vegetated buffer strip was determined by measuring concentrations of pollutants in
samples of the runoff directly off the road and after highway runoff passes through the filter
strip.  Efficiency was calculated based on the changes in the average concentrations in the
runoff samples at these locations.
Two filter strip sites were monitored in this study.  Four hundred twenty-three (423)
samples were collected over approximately thirty-four (34) storm events at the two sites.
Two sites were selected to investigate the potential for variation in performance between
vegetated buffer strips.  Also, monitoring of two sites under different conditions offers a
comparison that might provide insight into the factors that affect the removal efficiency of
filter strips.
4.2 Methods and Materials
4.2.1 Site Selection
Field sites were selected from existing highway medians or other grassy areas near
highways in the Austin area.  The primary criteria that were used in the selection of field sites
include:
? configuration of the drainage system at the site allowed for sampling of runoff from the
highway and from the vegetated buffer strip, i.e., the road and filter runoff were not
contaminated with water from other areas;
? the drainage to the vegetated buffer strip originated from a highway and did not include
50
runoff from other areas.
Secondary criteria included choosing two sites with different characteristics (e.g., vegetation
and slope), proximity to the research facility, safety of the personnel, and security of the
equipment.
Two sites were selected for monitoring.  The first vegetated buffer strip is located in
the median of MoPac (Loop 1) where the highway crosses Walnut Creek in northwest
Austin. The Walnut Creek site was monitored during a previous study (Irish et al, 1995), and
some data from the prior research was utilized in this study.  This site was monitored over the
period of April 1994 to May 1997.  However, only data collected from the period from
February 1996 to May 1997 was used to describe runoff from the road because the sampling
system was modified.
The second of the two filters is located in the median of U.S. 183 immediately north
of MoPac.  The U.S. 183 site is also in northwest Austin.  This site was monitored from
March 1996 to May 1997.
4.2.2 Site Descriptions
Walnut Creek
The vegetated buffer strip at Walnut Creek is a 1055 m section of highway median
which collects runoff from the northbound and southbound lanes of MoPac just south of
Walnut Creek (Figure 4.1 below).  The median was designed originally as a hydraulic
conveyance and not as a vegetated buffer strip.  The median cross-section is V-shaped with a
rounded bottom.  Runoff from the highway flows as sheet flow down the sides of the grassy
slope.  The runoff then flows along the center of the median into 4 drop inlets situated along
the centerline of the median.  The drop inlets discharge to a 1.22 m concrete storm drain that
conveys the runoff to Walnut Creek.  This storm drain collects runoff from the road and
median, as well as from several grassy shoulder areas.  The total drainage area of the storm
drain is approximately 10.46 ha (104,600 m
2
).  Approximately 38% of the drainage area is
paved with asphalt.
51
Figure 4.1.  Mopac at Walnut Creek filter strip.
Runoff from either the southbound or the northbound lanes of MoPac flows to the
median at any location along its length, since the cross-sectional slope of the highway
changes in this section.  The southern half (approximately 500 m) of the median receives
runoff from the 3 southbound lanes only, and the northern 500 m of the median receives
runoff from the 3 northbound lanes.  Lanes not feeding to the median drain to grassy
shoulder areas, which eventually drain to the 1.22 m storm drain.
The side slopes of the median vary from approximately 6.3 to 12.4%, with an average
grade of approximately 9.4%.  The total width of the median varies from 15.5 m to 16.2 m.
The distance from the pavement edge to the lowest point in the median, or the treatment
length of the filter strip, varies from 6.7 m to 8.2 m.  The median drains northward with the
exception of the northernmost 150 m, which drain southward to the northernmost drop inlet.
Slope of the median along the centerline varies from approximately 0.75% to 2.9%, with an
average grade of 1.7% along the northward-draining section.
The vegetation cover in the median is a mix of bunch grass and sod grass.  A
summary of the vegetation transect of the site performed in October of 1996 is shown in
52
Table 4.1.
Table 4.1.  Vegetative composition of Walnut Creek median (October 1996).
Species Name
Percent
Composition
Bermudagrass 30
Illinois Bundleflower 30
Medow Dropseed 19
Little Bluestem 10
Florida Palpalum 7
Indiangrass 2
Bare ground 2
Prairie Buffalo grass <1
The median was planted originally in Sideoats Grama, Green Sprangletop, Switchgrass,
Little Bluestem and Buffalo grass about 1989.
Water from the Mopac bridge over Walnut Creek drains to pipes which open to the
creek below.  The drainage area is paved with asphalt, thus providing an ideal source for
water quality sampling of the road at this site.
Approximately 47,000 vehicles per day traveled on the 3 northbound and 3
southbound lanes along this section of MoPac in April 1995.  The hourly traffic ranged from
100 to 3600 vehicles.
183 at MoPac
The vegetated buffer strip monitored at U.S. 183 at MoPac is the 356 m of grassy
median of U.S. 183 just north of MoPac.  This median was designed originally for hydraulic
conveyance.  Only the 3 southbound lanes of 183 drain into the median; the northbound lanes
53
drain to a curb-and-gutter storm drain.  The cross section of the median is V-shaped with a
rounded bottom.
Figure 4.2.  Vegetated buffer strip at U.S. 183 site.
The side slope of the median varies from 10.3% to 15.3%, and has an average slope
of approximately 12.1%.  The distance from the edge of the pavement to the lowest point in
the median, or the treatment length of the filter strip, varies from 9.1 m to 7.3 m.  The median
drains southward with an average slope of 0.73%, varying from approximately 0.60% to
0.83% along its length.  The northern edge of the drainage area of the median begins at a
drop inlet that collects runoff from areas further north.  The median ends at a drop inlet 356
m down gradient.  This drop inlet connects to a 0.61 m concrete storm drain.  The drainage
area of the drop inlet consists only of the southbound lanes of 183 and the median itself.
This area is 13,000 m
2
, approximately 52% of which is paved.
The vegetative cover of the filter strip is primarily Prairie Buffalo grass, which was
installed as plugs of sod in 1991. The vegetative composition of the median is summarized in
Table 4.2.  The high percentage of bare ground was caused by a brush fire that occurred
54
sometime around July 1996 in the median.  All signs of the fire disappeared within several
months.
Table 4.2.  U.S. 183 at MoPac Vegetation Composition (October 1996).
Species Name
Percent
Composition
Prairie Buffalo grass 76
Cedar Sedge 6
Texas Frogfruit 2
Illinois Bundleflower 1
Bermudagrass 1
Bare ground 14
A curb and gutter system drains the northbound lanes of U.S. 183 at this site.  All of
the runoff collected in these gutters originated from the highway.  The gutters drain to an
0.46 m concrete storm drain, providing an appropriate location for sampling road water
quality at this site.  The 1995 annual average daily traffic along U.S. 183 at this site was
111,000 vehicles.
Site Description Summary
A summary of the characteristics of the two vegetated buffer strips is given in Table
4.3.
4.2.3 Sampling/Monitoring Setup
The monitoring of both sites included the following tasks:
1) sampling of runoff from both the road and the grassy median;
2) measuring amount of flow from both the road and the median; and
3) measuring rainfall.
55
Table 4.3.  Vegetated buffer strip description summary.
Characteristic Walnut Creek U.S. 183
Centerline length (m) 1055 356
Width of entire median (m) 15.5 to 16.2 14.9 to 19.5
Filter strip treatment length (m)
Average median side slope
7.8 to 8.1
9.4%
7.5 to 8.8
12.1%
Average centerline slope 1.70% 0.73%
Cross-sectional shape V, rounded bottom V, rounded bottom
Drainage area (m
2
) 104,600 13,000
Vegetation mixed mostly Buffalo grass
Average Daily Traffic 47,000 111,000
Filter drainage area % paved 38% 52%
Road drainage area % paved 100% 100%
4.2.3.1 Equipment
Two Isco 3700 samplers, one Isco 674 rain gauge, and two Isco 3230 bubbler flow
meters were installed at each site to sample runoff, measure rainfall, and measure flow,
respectively.  Two samplers and flow meters were needed in order to monitor both the road
and the vegetated buffer strip.  A 12 volt battery recharged by a solar panel powered the
equipment.  The samplers, flow meters and battery at both sites were kept in a closed steel
housing.  Other equipment, such as pipes, tubing, and weirs also were used and are described
in sections below.
The bubbler flow meter measures flow by measuring the pressure required to force air
out of a tube.  This pressure indicates the height of water above the tube.  The height of the
water is converted to flow using equations reflecting the characteristics of either the pipe
(i.e., smoothness and slope of the pipe), the weir (i.e., type and angle of the weir), or other
56
characteristics depending upon the type of flow measuring device.
The sampler, when triggered by the flowmeter, pumps water from the area being
sampled through a plastic tube and into sample bottles (see Sampling/Monitoring Procedures,
page 59).  The Isco 3700 samplers contained 24 bottles each holding 350 mL of sample.  The
rain gauges are tipping gauges with increments of 1/100 inch.
Flow and rainfall data was relayed to the flow meter, where it was stored.  This
information was periodically downloaded onto a laptop computer for analysis.
4.2.3.2 Walnut Creek Setup
Vegetated buffer strip
Samples from the vegetated buffer strip discharge at Walnut Creek were collected
from the outfall of the 1.22 m storm drain.  The runoff sample tube was fastened to the inside
of the pipe several feet from the outfall to Walnut Creek.  The flow meter bubbler tube was
fastened to the pipe several feet further inside along a joint between two pieces of the pipe.
Flow in the storm drain was calculated using Manning?s equation for pipe flow.  The
following is Manning?s equation:
n
SAR
Q
2/13/2
1000
=
where:
Q = flow rate (L/s),
A = cross-sectional area of flow (m
2
),
R = hydraulic radius (m),
S = slope of the pipe (m/m), and
n = roughness coefficient of the pipe (n = 0.013).
Flowlink software was used to analyze flow data.  Inputs were pipe slope, roughness and
diameter, and the measured water height.  The flow was calculated automatically.  The
flowmeter was calibrated by capturing discharge in a bucket over a measured time.  The
slope was adjusted so that flowrate calculated by Flowlink matched the measured flow.
57
Road
Runoff from the MoPac bridge over Walnut Creek drains to vertical openings in the
road surface which drop water to the ground below.  A 10.2 cm PVC pipe was installed to
connect one of these openings to a wooden collection box at ground level.  The box was 1.85
m long by 1.22 m wide by 0.61 m tall.  Runoff from the road entered the box via the pipe,
and discharged over a weir.  The end of the sample tube from which runoff was collected
initially was placed in the bottom of the box; however, the tube was moved to inside the PVC
pipe to prevent sampling of resuspended sediment that had settled in the box.  The flow
meter bubbler tube was fastened to the bottom of the box and flow was measured from the
road by gauging the height of water behind the weir.
The weir in the collection box was a compound V-notch weir.  The weir has 3
sections; the bottom portion is 20.1 cm tall and has an angle of 30 degrees; the middle
portion is 4.8 cm tall at a 90 degree angle, and the upper portion is rectangular with height
5.3 cm.  In these experiments, the height of water in the weir rarely exceeded 20.1 cm;
therefore flow was calculated with the assumption that a 30 degree weir was used.  Flowlink
software calculates the flow over the weir using built-in formulas for flow over a 30 degree
V-notch weir.  The rain gauge for the Walnut Creek site was located several feet from the
1.22 m outfall to the creek.
58
5.3 cm
4.8 cm      90?
20.1 cm       30?
Figure 4.3.  Compound V-notch weir for flow measurement.
4.2.3.3 183 at MoPac Site
Vegetated buffer strip
Discharge from the vegetated buffer strip at the U.S. 183 site was sampled from the
storm drain which collects runoff from the filter.  The end of the sampler tube was fastened
to the pipe approximately 60 feet from the drop inlet.  No storm drain connections conveyed
additional water to the drain prior to this spot, i.e., 100% of the sampled water had passed
across the filter.  The flow meter bubbler tube was located several feet upstream of the
sampler tube end.
Road
Runoff from the road at U.S. 183 was sampled from a storm drain which collects
water from a curb and gutter draining the northbound lanes of U.S. 183.  The sampler tube
end was fastened to the bottom of this drain and the flow meter tube was fastened several feet
upstream of the sampler tube.
59
Flow from the filter strip and road were calculated using Flowlink software. The
flowmeter for the 183 filter strip was also calibrated using a bucket and stopwatch.  The
slope adjusted so that the flowmeter was accurate, similar to the calibration at the Walnut
Creek filter strip flowmeter.  The road was not calibrated; however, the road flow
measurements were accurate relative to other road flow measurements.  This relative
accuracy was needed to weight the sample concentrations against each other so that weighted
mean concentrations for the road runoff could be calculated.  The rain gauge at the U.S. 183
site was located at the downstream end of the median, approximately 32 m from the
downstream drop inlet.
4.2.4 Sampling/Monitoring Procedures
The flowmeters triggered the samplers during a storm event when the water level at
the monitoring location reached a designated height.  Once this water height was reached,
samples were collected on a programmed timed schedule that varied for each location.  These
schedules are listed in Table 4.4.  The schedules were dependent upon the duration and size
of the storm peak and tail typical for each location.  The samplers filled 4 bottles per sample;
thus, 6 samples were possible from the 24-bottle samplers before the sample bottles required
replacement.  No more than 6 samples were taken for most storms.  During the storm, flow
and rainfall were recorded every 5 minutes.
Table 4.4.  Schedule for taking samples during storm events.
Location Elapsed time between samples (minutes)
Walnut Creek road 30, 30, 60, 60, 60
Walnut Creek filter 15, 30, 30, 60, 60
183 road 15, 15, 30, 30, 60
183 filter 30, 30, 30, 60, 60
60
Sample bottles were collected immediately after daytime storms; however, samples
from evening, night and weekend storms were collected the following day. The samples were
redistributed into laboratory bottles, labeled, logged, preserved, and refrigerated until the
analyses were performed at CRWR.
4.2.5 Numerical Analysis
Concentration Reduction
A concentration reduction was calculated for each constituent by finding the average
concentration of the constituent observed for the highway runoff and the median discharge
and applying the following formula:
()
%100?
?
=
r
sr
C
CC
R
where
R = concentration removal efficiency, %
C
r
 = average concentration observed in runoff from highway (mg/L, CFU, or
NTU)
C
s
 = average concentration observed in discharge from vegetated buffer strip
(mg/L, CFU, or NTU)
The average concentrations were calculated in a process involving several steps.  An
event mean concentration (EMC) for the constituent was calculated for each storm.  The
EMC is an average concentration for a storm calculated using concentrations from several
discrete samples which are weighted according to the amount of flow that was passing the
collection point around the time each sample was taken.  Appendix B includes sample
concentrations and associated flow volumes used for weighting the samples.
The flow associated with each sample was determined using Flowlink software and
was dependent on the sampling schedule for the site.  Normally, the flow associated with
each sample was the volume of runoff that passed the sampling tube from the time halfway
between the previous sample and the current sample to the time halfway between the current
61
sample and the subsequent sample.  If samples 3, 4, and 5 of a storm were taken at 6 A.M., 7
A.M., and 8 A.M., then sample 4 would be associated with the volume of flow passing the
flow meter bubbler tube between 6:30 and 7:30 A.M.  The time interval before and after the
first and last samples was normally equal to standardize these calculations.
An average of all flow-weighted averages for each storm was used to calculate the
final concentrations (listed in Table 4.5).  The average is the preferred estimator for the mean
of a lognormally distributed data with coefficient of variation less than 1.2 (Gilbert, 1987).
The storm concentration data for the sites are lognormally distributed, and the coefficient of
variation for the majority of the flow-weighted averages of constituents was less than 1.2.
The average was used all constituents for simplicity.  Summaries of flow-weighted averages
for all storms and the average concentration calculations for each site are presented in
Appendix C.
Any concentration that was below the detection limit of the analytical procedure was
assumed to be equal to the detection limit for the purpose of this evaluation.  This approach
resulted in conservative (lower) removal efficiencies.  The majority of concentrations below
the detection limit were observed for samples from the filter strips.  Hence, assuming the
detection limit was likely to increase the average concentrations in the discharge of the filter
strip to a greater extent than in the highway runoff, and as a result, the calculated removal
efficiencies will be smaller and more conservative.
Load Reduction
The observed reductions in concentrations demonstrate the ability of a vegetated
buffer strip to remove constituents via sedimentation, filtration, dilution, biological activity,
and other physical and chemical mechanisms.  However, additional removal of constituents
occurs as the runoff infiltrates through the soil.  The reduction in total load includes the
effects of infiltration and represents the total reduction in the mass of constituents that occurs
in the filter strip.
  An annual pollutant load is the mass of a particular constituent that is discharged
through an outfall over a one-year period.  Calculating a reduction in the constituent load
62
requires some interpretation.  In this study, the calculation of load reduction is directed at
establishing the difference between the constituent load assuming the highway runoff were
conveyed directly to a storm sewer without treatment and the load from the highway runoff
after treatment by the filter strip.
Reduction in pollutant load was calculated as a percent of total load for each site
using the following formula:
()
%100?
?
=
H
FH
L
LL
R
where
R = reduction in pollutant load from the highway as a result of treatment by the
vegetated buffer strip, %
L
H
 = annual pollutant load to receiving waters if the runoff from the highway was
not treated by the filter, kg/yr
L
F
 = annual pollutant load to receiving waters from the vegetated buffer strip
drainage area with runoff from the highway being treated by the filter, kg/yr
Annual pollutant loads (L
H
 and L
F
) were calculated using an adaptation of the ?simple
method? (EPA, 1992).  The simple method was converted for metric units.  The simple
method used in this study is defined by the following equation:
()( )( )[]()()( )00001.0ACRvCFPL =
where
L = annual pollutant load at the outfall of the drainage area (kg/yr)
P = average annual precipitation in Austin, Texas (82.6 cm/yr)
CF = correction factor that adjusts for small storms where no runoff occurs (0.9)
Rv = runoff coefficient of the drainage area concerned (m
3
 runoff/m
3
 rainfall)
C = average concentration of the pollutant (mg/L)
A = drainage area (m
2
)
The number 0.00001 is a conversion factor used to obtain correct units.  Additional notes
63
concerning the origin of drainage area and runoff coefficient values are given below.
Drainage Area
The load after treatment by the vegetated buffer strip (L
F
) was calculated based on a
drainage area, A, that was assumed to be the entire drainage area of the outfall for the
vegetated buffer strip.  The load assuming the vegetated buffer strip was not treating the
highway runoff (L
H
) was calculated assuming the drainage area A was the area of the
highway pavement.
Runoff Coefficient
A runoff coefficient is the fraction of volume of rainfall that produces runoff in a
drainage area.  In other words, the runoff coefficient is the fraction of rainfall from an area
that does not infiltrate into the soil.  The coefficients used to calculate L
F
, the constituent
loads after treatment by the filter strip, were calculated using flow data measured at the two
filter strip collection drains and rainfall data collected at each site.  The volume of rainfall
was calculated by multiplying rainfall depth for each storm by the catchment area.  Runoff
volume was calculated using Flowlink software with the collected flow data.  Plotting rainfall
and runoff volumes for all storms results in a linear trendline.  The slope of this graph is the
runoff coefficient.  The runoff coefficients used to calculate the loads without treatment by
the filter strip (L
H
) was 0.95.
4.2.6 Grab Samples
In addition to the continuous monitoring at the two filter sites, grab samples were
taken along the length of the vegetated buffer strip at U.S. 183 during 5 rain events.  The
objective of these grab samples was to determine where the treatment was occurring, i.e.
down the length of the median or along the side slopes of the median.
Grab samples were collected at points 240, 180, 120, 60, and 0 meters upstream of
the drop inlet along the center of the median at the U.S. 183 site.  The samples were collected
while standing on the northbound side of the centerline of the median since only the
64
southbound lanes of 183 drain into the filter.  Samples were collected starting at the upstream
end of the median in order to find changes in concentration for the runoff as it traveled down
the median.
4.3 Field Results
4.3.1 Runoff Coefficients
The runoff coefficient for each site was calculated using the data plotted in Figure 4.4
and Figure 4.5.  The calculated runoff coefficient for the Walnut Creek site is 0.30.  This
value agrees well with runoff coefficients for other sites in Austin with comparable
percentages of impervious cover (Barrett, 1997).  The runoff coefficient for the filter strip at
U.S. 183 was initially calculated to be 0.66 (Figure 4.5); however, a value of approximately
0.40 is normal for a drainage area that is 52% paved, such as the U.S. 183 filter strip drainage
area.  The higher-than-expected runoff coefficient was suspected to be caused by runoff
entering the drainage area from unanticipated sources.
65
y = 0.2986x - 104.44
R
2
 = 0.8002
0
200
400
600
800
1000
1200
1400
1600
1800
0 500 1000 1500 2000 2500 3000 3500 4000 4500
Rainfall volume, m
3
Runoff volume, m
3
Figure 4.4.  Runoff coefficient of the filter strip drainage area at Walnut Creek.
y = 0.6621x - 1.9194
R
2
 = 0.8808
0
100
200
300
400
500
600
700
800
900
1000
0 200 400 600 800 1000 1200 1400 1600
Rainfall volume, m
3
Runoff volume, m
3
Figure 4.5.  Initial calculation of runoff coefficient of filter strip drainage area at U.S.
183.
An inspection of the site proved this to be the case; erosion at the upstream drain at
66
the U.S. 183 site caused a large amount of flow to bypass the drain and flow into the
catchment area of the U.S. 183 filter strip.  It was thus impossible to define the area that
should be used for rainfall volume calculations at the U.S. 183 site.  Therefore, a runoff
coefficient and area for the filter strip drainage area at the U.S. 183 site were assumed.  The
area used was 13,000 m
2
.  This is the area of highway and median that would have drained to
the filter strip drop inlet if the upstream drain erosion had not occurred.
The runoff coefficient was calculated using results from a recent study which
developed a relationship between runoff coefficient and impervious cover based upon
monitoring of multiple storm events at each of 18 sites in the Austin area (Barrett, 1997).
The study used the following second-order equation to describe the relationship:
Rv = 0.3428(IC)
2
 + 0.5677(IC) + 0.0125
where
Rv = runoff coefficient, m
3
 runoff/m
3
 rainfall
IC = fraction of impervious cover for the site.
According to this equation, the runoff coefficient for a site with 52% impervious cover is
expected to be 0.40.  This value was used for calculation of L
F
.  In summary, the pollutant
load calculations for the U.S. 183 filter are the best possible estimate of what the loads would
be if the filter were not receiving unintended runoff from other drainage areas.
4.3.2 Concentration and Loading Reductions
The average concentrations and percent concentration reduction observed at both
field sites are given in Table 4.5.  Table 4.6 includes the pollutant loads and loading
reductions observed at both sites.
67
Table 4.5.  Reductions in concentrations observed at two vegetated buffer strips.
U.S. 183 Walnut Creek
Road Mean Swale Mean Reduction Road Mean Swale Mean Reduction
Constituent mg/L mg/L % mg/L mg/L %
TSS 157 21 87 190 29 85
Turbidity** 55 17 69 70 16 78
Fecal Col* 96000 280000 -192 NA 240000 NA
Fecal Strep* 23000 40000 -74 7100 41000 -477
COD 94 37 61 109 41 63
TOC 33.9 16.7 51 41.3 19.5 53
Nitrate 0.91 0.46 50 1.27 0.97 23
TKN 2.17 1.46 33 2.61 1.45 44
Total P 0.55 0.31 44 0.24 0.16 34
Zinc 0.347 0.032 91 0.129 0.032 75
Lead 0.138 0.082 41 0.093 0.077 17
Iron 3.33 0.69 79 2.04 0.51 75
* units are CFU/100mL, ** units are NTU
68
Table 4.6.  Constituent loadings with and without treatment by the vegetated buffer
strip.
U.S. 183 Walnut Creek
Untreated Treated Load Untreated Treated Load
Load, L
H
Load, L
F
Reduction Load, L
H
Load, L
F
Reduction
Constituent kg/yr kg/yr % kg/yr kg/yr %
TSS 748 79 89 5320 671 87
Turbidity** 265 66 75 1980 367 81
Fecal Col* 4600 11000 -136 NA 56000 NA
Fecal Strep* 1100 1500 -41 2000 9600 -380
COD 450 144 68 3060 952 69
TOC 162 65 60 1160 455 61
Nitrate 4.3 1.8 59 36 23 36
TKN 10.3 5.63 46 73 34 54
Total P 2.65 1.20 55 6.73 3.70 45
Zinc 1.66 0.124 93 3.62 0.75 79
Lead 0.661 0.317 52 2.61 1.79 31
Iron 15.9 2.66 83 57 11.8 79
* 10
9
 CFU/yr,  ** NTU*L/yr
Discussion of Concentration and Loading Reductions
In general, the monitoring results demonstrate good to excellent (often greater than
75%) removal rates for suspended solids and metals, good removal of organic compounds
(60-70%), moderate removal rates for nutrients (25-60%), and negative removal of bacteria.
In addition, though the highway runoff and the filter strip discharge concentrations often
differ between the two sites, the removal rates for all constituents between sites are
69
remarkably similar.
The constituent loading removal rates observed at the two filter strips are
considerably higher than those found in previous studies (Young et al, 1996; Yu and
Benelmouffok, 1988).  This observation is not true for all constituents and all studies.  The
Young et al (1996) report refers to a filter strip study with comparable TSS, phosphorus, and
lead removals (70, 40, and 25 percent, respectively) to this study, but removal efficiencies
reported for zinc and nitrate/nitrite (40 and 10 percent, respectively) were lower than those
found for the Austin, Texas filter strips.  Yu and Benelmouffok (1988) report lower removal
efficiencies for sediments, nutrients, and metals than the removals seen in this study.  The
reason for the higher removal efficiencies observed in the Austin, Texas study is difficult to
identify with certainty.  One possible reason is that the filter strips in other studies treated
runoff from a larger drainage area than the filter strips in this study, which treated runoff only
from a 3-lane highway.  The Yu and Benelmouffok filter drained an 18-acre area near a
highway and shopping center complex.  The larger drainage area could have resulted in
higher runoff velocities and water depths, thereby reducing the effectiveness of the filter
strip.  The difference in drainage areas might explain why filter strips may be ?unreliable in
urban settings? (Schueler et al, 1992), but more appropriate for treating runoff from areas
with relatively small drainage areas, such as highways, as demonstrated by the results of this
study.  Highways provide a relatively small catchment area for a filter strips that lie along
their entire length.  Water depths and velocities are normally low and filter strips can act
effectively in such a configuration.
The results of this study indicate that filter strips of relatively short lengths, 7 to 9 m,
can be effective for removing a variety of constituents in highway runoff.  The consistency
seen in removal efficiencies between the two sites further confirms the removal efficiencies,
and indicates that similar removal efficiencies could be expected for filter strips with similar
characteristics to those studied here.  This observation is particularly promising since
medians that already are present along highways in Austin and in other areas may be of
comparable size, geometry and other aspects to those monitored in this study.  Thus, the
inclusion of an effective BMP in the design of a highway is straightforward.  The highway
70
runoff can be allowed to drain as sheet flow down the sides of a grassy median or shoulder
area.  This design could be implemented for highways already built by the removal of curbs
so that runoff flows into the median to the storm drains along the median for runoff
collection.
The pollutant removal capabilities of filter strips treating highway runoff are
comparable to those of structural controls, such as sand filters.  A comparison of removal
efficiencies for the monitored filter strips and several sand filters is given below in Table 4.7.
In the Highwood and BCSM sand filters, sedimentation and filtration occur in one basin; the
Seton Pond facility has separate detention and sand filtration basins.  The removal
efficiencies for the sand filters reflects pollutant removal only for the runoff that was
captured by the facility and does not reflect reduction in removal efficiency caused by bypass
of runoff during large storms.  All three sand filters are located in the Austin, Texas area; the
Seton Pond results are from a monitoring study performed in conjunction with this study.
The filter strip removal efficiencies are comparable to sand filter removal efficiencies for all
constituents.
Table 4.7  Comparison of filter strip performance with three sand filtration systems.
Sand Filters
(% mass reduction)
Vegetated Buffer Strips
(% mass reduction)
Constituent Highwood BCSM Seton Pond U.S. 183 Walnut Creek
TSS 86 75 98 89 87
COD 29 40 88 68 69
TOC 43 38 62 60 61
Nitrate -18 -42 64 59 36
TKN 40 60 65 46 54
Zinc 40 74 94 93 79
Iron 57 65 95 83 79
71
The Federal Highway Administration (FHWA) makes two recommendations that are
refuted to some extent by the results of this research.  First, the FHWA recommends that the
slopes of filter strips used to treat runoff be less than 5 percent to prevent gullies which can
disrupt sheet flow.  The average slopes of the filters monitored in this study, however, are 9
and 12 percent at the Walnut Creek and U.S. 183 sites, respectively.  No gullies were
witnessed along the median sides at either site.  It may be that the short filter length and
relatively small catchment area (3 highway lanes plus shoulders) for the filter strips
prevented the formation of gullies.  Differences in rainfall intensity or antecedent dry periods
between the FHWA study and the Austin study may also explain why no gullies were
witnessed at the Austin filter strips.  Second, the FHWA cites the results of a study which
suggest use of filter strips only for roadways with a maximum of 2 lanes and average daily
traffic of 30,000 (Young et al, 1996).  Both filters strips studied in Austin, Texas were 3-lane
(each direction) highways and had daily traffic of 47,000 (Walnut Creek) and 111,000 (U.S.
183); nevertheless, the filter strips were effective at removing contaminants in runoff.
Results indicate that filter strips are effective for 3-lane (each direction) highways at average
daily traffic counts greater than 50,000.
Removal efficiencies for copper were not calculated because copper concentrations in
a large majority of the samples were less than the detection limit, 0.006 mg/L.  These data
indicate that copper in the runoff coming from highways in Austin, Texas is minimal.
The calculated removal efficiencies for lead are considerably lower than removal
efficiencies for iron or zinc, or for suspended solids.  It is difficult to explain these data.
Lead is one of the least soluble metals in urban runoff (Wiginton et al, 1986; Barrett et al,
1995b), and as a result one would expect lead would have a strong association with
particulate matter in runoff.  This would make lead easily removed by such processes as
sedimentation and filtration in the vegetated buffer strips.  The lower removal efficiencies
observed for lead are thus contrary to expectations.  The data reported by other research
shows lead to be removed equally or better by vegetated BMPs over other metals
(Municipality of Metropolitan Seattle, 1992).  Other results are similar to the data observed
in this study (Young et al, 1996).  Occasional problems with the analytical equipment used
72
for lead analyses compromised the reliability of the lead concentrations detected for some
samples.
4.3.3 Grab Sample Results
The grassy medians monitored for this project were initially thought to be acting as
grassy swales, that is, treatment was thought to occur as the runoff traveled in deep flow
along the center of the median.  However, the medians responded more like vegetated buffer
strips, which treat runoff as the sheet flow travels over a broad vegetated slope.  The
treatment occurred along the sides of the median and in not the center.
The results of the grab samples are summarized in Figure 4.6.  These data show the
change in concentration of TSS along the length of the median.  Total suspended solids was
used as an indicator constituent for determining the removal pattern.  The data reveal that a
small reduction in concentration occurs down the length of median; however, this removal
accounts for only a small part of the over 80% reduction in total TSS concentration.  The
average TSS concentration observed from the road at this site is 128 mg/L, therefore the
majority of removal of TSS must be occurring along the side of the median.  Therefore, the
median acts as a vegetated buffer strip, not a grassy swale.
This observation indicates that the length of the median has only a small effect on
pollutant removal.  A longitudinally long (i.e., long in the direction perpendicular to flow)
filter strip is not required to achieve removal of constituents.  Thus, a median that filters sheet
flow from a very short length of road, but is similar in other respects to those monitored in
this study would be expected to have comparable removal capabilities.  Other factors, such as
the length and slope of the sides of the median and density and type of vegetative cover, may
have a greater effect than the median?s longitudinal length on the efficiency of filter strips
along highways.
73
0
10
20
30
40
50
60
70
80
0 50 100 150 200 250
Distance from drop inlet, m
TSS concentration, mg/L
2/7/97
2/12/97 AM
2/12/97 PM
3/25/97
5/9/97
Figure 4.6. TSS concentrations along the center of the median for 5 storm events.
4.3.4 Other Monitoring Results
During the monitoring phase of this study, two important observations were noted
regarding filter strips; both observations demonstrate the need for filter strip maintenance.
Significant channel erosion occurred at the bottom of the Walnut Creek median.  In February
1997, seven washouts were noted along the 1055 m of median.  All were in the center of the
median and ranged from 0.15 to 0.91 m in width, 0.15 to 0.45 m in depth, and 4.5 m to 28 m
in length.  The washout areas were primarily bedrock with some sediment, and devoid of
vegetation (Figure 4.7).  Such washouts diminish the effectiveness of filter strips by
contributing sediments to receiving waters and reducing any treatment that may occur along
the length of the median.  In addition, the washouts can present aesthetic problems and
maintenance problems, such as during the mowing of gullies.  No erosion was noted at the
74
U.S. 183 site.  The longitudinal slope of the Walnut Creek median (along the median
centerline) averages 1.7%, while the average longitudinal slope at U.S. 183 is only 0.7%.
Higher velocities are associated with steeper slopes; this may explain why erosion occurred
at the Walnut Creek median.  Future design of filter strips should consider measures to
prevent erosion.  The use of additional drop inlets along the median may alleviate the erosion
along the Walnut Creek median.
Figure 4.7.  Erosion at the Walnut Creek vegetated buffer strip.
The second observation regarding the filter strips in the field is the presence of a
sediment ?lip? that formed along parts of the edge where the pavement meets the grassy
median at the U.S. 183 site.  This lip, which formed from the settling of sediment at the
75
pavement/median interface, grew until highway runoff was prevented from entering the
median and was instead diverted to a curb and gutter system.  The runoff thus traveled
toward receiving waters untreated.  This problem has been noted for grassed swales by other
researchers as well (Schueler et al, 1992).  This type of lip can likely be avoided during
construction by ensuring that the level of the soil near the pavement edge is lower than the
pavement.  Periodic maintenance can remove sediments from along the highway/median
interface.
4.4 Effects of Metals on Vegetated Areas
4.4.1 Concerns Regarding Metals Deposition on Vegetated Areas
Metals in highway runoff are removed by sedimentation, filtration, infiltration into
soil, and possibly other mechanisms in vegetated buffer strips, thereby protecting receiving
waters from the toxic effects of metals.  These metals, however, accumulate in various forms
in the filter strip itself.  The fate and effect of these accumulated toxic metals on the
environment is a natural concern.  The objective of this portion of the study is to make a
broad assessment of the risk to human health and the environment posed by metal deposition
from highway runoff in vegetated buffer strips.
A simple mass balance of metals entering and leaving the vegetated buffer strip
indicates that metals are accumulating in the strip.  The metal loads presented in Section 4.3
can be used for such a mass balance.  For example, at the U.S. 183 site, approximately 1.44
kg of zinc per year enters the filter strip from highway runoff.  However, only 0.07 kg/yr of
zinc exits the filter strip.  The difference, or 1.37 kg per year, is deposited over the area of the
filter strip.  The removal of metals from the filter strip by wind and infiltration is assumed to
be negligible.
The fate of metals after deposition, and the metals concerns with regard to protecting
human health and the environment, should be understood before addressing any assessment
76
of risk.  Once removed from highway runoff, possible fate of trace metals within vegetated
buffer strips include the following:
1. Residence in an insoluble form, i.e., attached to particulate matter, in the soil
matrix;
2. Uptake of soluble metals by plants;
3. Uptake by animals who consume plants with accumulated metals;
4. Leaching of soluble metals from the soil into groundwater;
5. Removal from the filter strip to receiving waters by runoff from subsequent storm
events;
6. Some evaporation of the metals is possible, as documented in recent studies
(Carpi and Lindberg, 1997); and
7. Removal from the filter strip by wind action on particulates containing metals.
The primary concerns for trace metals applied to vegetated areas are the following:
1. Phytotoxicity, or toxicity to plants that uptake metals;
2. Toxicity to animals that eat plants with high metal concentrations;
3. Contamination of groundwater resources that are sources of drinking water or
provide habitats for plant and animal species.
4.4.2 Use of Part 503 Regulations to Assess Environmental Risk
Assessment of the risk to human health and the environment from the accumulation
of metals in the roadside environment has not been reported in any detail.  A recent
regulation developed by the U.S. Environmental Protection Agency may be used to assist in
such an assessment.  This regulation, The Standards for the Use or Disposal of Sewage
Sludge, or Title 40 of the Code of Federal Regulations (CFR), Part 503, provides
comprehensive requirements for the management of biosolids generated during the process
77
of treating municipal wastewater.   This regulation was passed in 1993 in compliance with
requirements of the Clean Water Act of 1987.  Of particular interest to this study is that the
regulations provide annual and cumulative limits for the application of metals on cropland.
4.4.3 Justification of Use of 503 Regulations for Stormwater
The 503 Regulations for biosolids disposal were based upon an estimate of the
environmental risk of biosolids application on cropland.  Nonetheless, a meaningful
comparison is possible between rates of deposition allowed by the regulations and rates of
deposition found on the filter strips in this study.  The notable differences in the situation for
which the 503 Regulations were developed and their use for this study include the following:
? Land use.  The biosolids regulations were intended for regulating land used to
grow crops for human and animal consumption.  Metals that are absorbed by
crops are harvested and removed from the area.  Vegetated BMPs, normally do
not have this mechanism for removal of metals from the site unless mowing
clippings are collected and removed from the area.
? Nature of applied material.  The biosolids regulations pertain to application of
biosolids effluents from municipal wastewater treatment plants.  This analysis
investigates the risk associated with highway runoff.
The similarities between the situation for which the 503 Regulations were developed and
treatment of highway runoff by a vegetated buffer strip include the following:
? The environmental risks involved in metals deposition from highway runoff on
filter strips are the same as those present when applying biosolids to cropland:
phytotoxicity, toxicity to animals eating plants, and groundwater contamination.
? Both the application of biosolids on cropland and the treatment of highway runoff
over a filter strip involve the spreading of a substance that is primarily water with
some solids, including metals, over land.
? The land uses in question both contain significant vegetation.
The 503 regulations provide a starting point for an assessment of risk.  A more
78
accurate risk assessment requires an extensive study specifically regarding environmental
concerns of pollutant deposition on grassy areas from highway runoff.
4.4.4 Metals Limitations Placed by the 503 Regulations
The metals limitations that are part of the 503 Regulations include annual and
cumulative limits for 10 metals.  The annual loading limits are the maximum amount of
metal, in kilograms of metal per hectare per year, that may safely be applied to cropland; the
cumulative loading limits are the cumulative amount of metal, in kilograms per hectare, that
may be safely applied to cropland over time.  The 503 Regulations require that biosolids
application must cease if either of these limits are exceeded.
An annual metals loading rate at each site was calculated, and the calculated rate was
compared to the limits provided by the 503 Regulations.  This comparison provided
information regarding the current presence of risk.  Second, the time in years until the
cumulative loading rate limitations were exceeded was calculated.  This time is the site life
for each site based upon metals limitations.
Annual metals loading rates for each metal were calculated by the following formula:
F
FH
A
LL
R
?
=
where
R = annual metal loading rate for one metal over the vegetated buffer strip,
kg/ha/yr
L
H
 = annual metal load generated by the portion of the highway that drains onto
the vegetated buffer strip, kg/yr
L
F
 = annual metal load that exits the vegetated buffer strip, kg/yr
A
F
 = area of the vegetated buffer strip
The annual metal loads from the highway and buffer strip, L
H
 and L
F
, were previously
79
presented in Table 4.6 (page 68).  The site life calculation used the following formula:
R
Limit
SL
cum
=
where
SL = site life of the vegetated buffer strip based on metals limitations, yr
Limit
cum
 = cumulative metal loading limitation from the 503 regulations, kg/ha
R = annual metal loading rate for one metal over the vegetated buffer strip,
kg/ha/yr
4.4.5 Metals Risk Analysis Results and Discussion
The calculated annual metals deposition rate for each site for two metals is presented
in Table 4.8, along with the 503 Regulations limits for comparison.  Calculated site lives
based upon metals limitations for the two metals are presented in Table 4.9.
Table 4.8.  Annual metals loading rates, in comparison to the 503 Regulations.
503 Regulations Limit* U.S. 183 Filter Strip Walnut Creek Filter Strip
Metal kg/ha/yr kg/ha/yr kg/ha/yr
Zinc 140 4.9 9.2
Lead 15 1.2 0.25
* For metals in biosolids applied to cropland
80
Table 4.9.  Site lives based upon metal deposition limitations.
U.S. 183 Filter Strip Walnut Creek Filter Strip
Metal years years
Zinc 570 304
Lead 244 1202
The metals loading rates at the two sites for lead and zinc are lower than the annual
metals loading limits prescribed by the 503 Regulations.  Indeed, the metal loading rate on
the filter strips was less than one tenth of the rate limits for application of metals in biosolids
to cropland.  Therefore, metal deposition from highway runoff on roadside grassy areas may
not pose any risk to human health and the environment.  This conclusion is reinforced by
other considerations.  The conservative nature of the 503 Regulations when applied to BMPs
and the minimal effects of highway runoff on groundwater shown by previous research
further support this claim.
The site lives for each site based upon both metals accumulation in the filter strip was
over 200 years.  Therefore, no adverse effects are likely to occur as a result of metals
accumulation in the strips for at least 200 years.
This analysis was performed for only two metals in highway runoff.  Copper was
found at concentrations below detection limits in highway runoff in this study, and iron is not
regulated by the 503 Regulations.  Other metals, however, could be investigated.  Cadmium,
in particular, has a low annual loading limit (1.90 kg/ha/yr) in the 503 Regulations, and is
found in highway runoff, though in low concentrations (Barrett et al, 1995b).  Nickel and
chromium also are detected in low concentrations in highway runoff and are regulated by the
503 Regulations.
81
4.5 Summary of Field Study Results
Vegetated buffer strips can effectively remove many constituents in highway runoff.
The percent removal of mass of constituents in runoff within the filter strips was above 85%
for total suspended solids; 68%-93% for turbidity, chemical oxygen demand, zinc, and iron;
36%-61% for total organic carbon, nitrate, total Kjeldahl nitrogen, total phosphorus, and
lead; and negative removal of bacteria.  These data indicate that relatively short (7-9 m) filter
strips with moderate slopes (9-12%) can treat highway runoff efficiently.  Filter strips that
traverse highways treat a relatively small drainage area.  This set of conditions may be the
reason that the evaluated filter strips were effective, while in the past filter strips have been
reported to be unreliable for treating runoff in developed areas.
The removal efficiencies observed at both sites, despite differences in vegetation,
traffic density, median side slope and longitudinal (centerline) slope, are similar.  Therefore,
other filter strips, even with some varying characteristics, are likely to treat highway runoff
with similar effectiveness.  The observed data indicate that treatment of highway runoff
occurred along the sides of the median, and not along the center of the median.  Hence, an
effective best management practice for treating highway runoff is accomplished by allowing
runoff from the highway pavement to pass as sheet flow down a smooth, vegetated area of at
least 8 meters in length and slope less than 9 to 12%.
The rate of zinc and lead deposition from highway runoff on the filter strips is less
than one tenth the maximum deposition rate allowed by the 503 Regulations, which limit
application rates of metals in biosolids to cropland.  Any threats to human health and the
environment from metals deposition from highway runoff on vegetated areas are small.
Accumulation of metals in the monitored filter strips could continue for over 200 years
without risk.
82
Chapter 5 Conclusions and Recommendations
5.1 Channel Swale Conclusions
The conclusions from experiments on the channel swale are the following:
? Removal of TSS, COD, total phosphorus, TKN, zinc, and iron was highly correlated with
swale length.  No trend was observed for nitrate.
? Most of the reduction in the concentration of constituents in runoff occurred in the first
20 meters of the swale.   Little improvement in water quality was observed during the last
20 meters.  Swales longer than 20 m may not be cost effective.
? The removal efficiency for constituents of particulate nature, such as suspended solids,
organic material, and metals with the exception of zinc, decreased with increased water
depth.  No relationship between water depth and removal efficiency was observed for
nitrate and TKN.  It is uncertain whether decreasing water depth, decreased velocity, or
both were responsible for increases in removal efficiency for particulate constituents.
Increasing water depth and velocity of runoff in a swale will impede the swale?s
performance for most constituents.
? The removal efficiency of the grassed swale changed between dormant and growing
season for only one constituent.  Total suspended solids were removed best in the
growing season, during which there is a combination of new grass and remaining
dormant grass resulting in high grass blade densities.
? Dormant Buffalo grass did not decay until the subsequent growing season.  Grassed
swales can still be effective at removing contaminants during the dormant season.
? Percolation of runoff through layers of soil and gravel into the underdrain reduced
concentrations of all constituents except nitrate.
? The removal efficiencies for the grassed swale in the channel were similar to grassed
swales of other studies (Municipality of Metropolitan Seattle, 1992; Schueler et al, 1992)
and similar swales can be expected to have comparable removal efficiencies.
83
5.2 Field Study Conclusions
The conclusions of the field study are the following:
? Vegetated channels designed solely for stormwater conveyance can be as effective as
sand filters for reducing the concentrations and loads of constituents in highway runoff.
The percent reduction in pollutant mass transported to receiving waters was above 85%
for total suspended solids; 68%-93% for turbidity, chemical oxygen demand, zinc, and
iron; and 36%-61% for total organic carbon, nitrate, total Kjeldahl nitrogen, total
phosphorus, and lead.
? Simple, V-shaped highway medians or shoulder areas with length of at least 8 meters, full
vegetative cover, and slopes less than 9 to 12 percent provide protection to receiving
waters against constituents in highway runoff.  Consequently,  many highways in the
state which have vegetated channels are already employing an effective best management
practice.
? The removal efficiencies for the two filter strips were similar, despite significant
differences in vegetation, traffic density, median side slope, and longitudinal (median
centerline) slope.  Other comparable filters may have similar removal efficiencies.
? The removal efficiencies for the two filter strips are comparable to removal efficiencies
for sedimentation and filtration controls.
? Grab samples confirmed that the removal of constituents occurred down the sides of the
median, and not down its longitudinal length.  A longitudinally long median is not
required for effective removal of constituents from highway runoff.
? The slopes and lengths recommended in this report are appropriate for highways, but may
not be sufficient for other situations.  The small drainage areas provided by highways
may explain why the filter strips were effective.
? The deposition rates of lead and zinc on the filter strips were less than one tenth the
allowable rate for metals application on cropland.  Threats to human health and the
environment from metals deposition from highway runoff on vegetated areas are
84
minimal.
? Vegetated buffer strips and grassed swales can be used as a pretreatment alternative for
structural runoff controls, such as sand filters, which can clog from sediment loads.
5.3  Recommendations
The recommendations of this study are the following:
1. Include vegetated buffer strips or grassed swales in the design of new highways or
renovation of old highways.  Vegetated BMPs are especially beneficial in
environmentally sensitive watersheds or recharge zones; in addition, they could be used
when regulations require enhancement of highway runoff water quality.  However, use
vegetated BMPs only when sufficient space is available and geometry and climate allow
for appropriate slopes and sufficient vegetative cover.  Effective vegetated buffer strips
can be included in highway design at low cost and with little obstruction to other
highway design objectives.
2. Avoid curb-and-gutter systems for removal of runoff from new highways and roadways.
Instead, allow the runoff to exit the pavement as sheet flow into grassy medians or
shoulder areas.  It is recommended that sheet flow be maintained.
3. Filter strips should have a maximum slope of 9 to 12 percent and a minimum length of 8
meters.
4. Include effective erosion control techniques in highway median design.  A storm drain
system with drop inlets can be used in conjunction with vegetated channels to minimize
erosion and maintain shallow water depths in the swales.
5. Swale length, water depth, and velocity have a significant impact on the removal
efficiency of grassed swales.  Consider these factors in the design of grassed swales.  One
study combined the effects of these factors by recommending a 9 minute minimum
hydraulic detention time for runoff in a grassed swale (Municipality of Metropolitan
Seattle, 1992).  Ignore the effect of season on swale efficiency for design considerations.
85
BIBLIOGRAPHY
Barrett, M.E., 1997, Water Quality and Quantity Inputs for the Urban Creeks Future Needs
Assessment, Draft Report to the Drainage Utility Department, City of Austin, Texas, Center
for Research in Water Resources, University of Texas, Austin, TX.
Barrett, M.E., Malina, J.F. Jr., Charbeneau, R.J., and Ward, G.H., 1995a, Characterization of
Highway Runoff in the Austin, Texas Area, CRWR 263, Center for Research in Water
Resources, Bureau of Engineering Research, Austin, TX.
Barrett, M.E., Zuber, R.D., Collins, E.R. III, Malina, J.F Jr., Charbeneau, R.J., and Ward,
G.H., 1995b, A Review and Evaluation of Literature Pertaining to the Quantity and Control
of Pollution from Highway Runoff and Construction, Second Edition, CRWR 239, Center for
Research in Water Resources, Bureau of Engineering Research, Austin, TX.
Carpi, A., and Lindberg, S.E., 1997, ?Sunlight-Mediated Emission of Elemental Mercury
from Soil Amended with Municipal Sewage Sludge?, Environmental Science & Technology,
Volume 31, No. 7, pp. 2085-2091.
Dorman, M.E., Hartigan, J., Johnson, F., and Maestri, B., 1988, Retention, Detention and
Overland Flow for Pollutant Removal from Highway Stormwater Runoff: Interim Guidelines
for Management Measures, FHWA/RD-87/056, Versar Inc., Springfield, VA.
Gilbert, R.O., 1987, Statistical Methods for Environmental Pollution Monitoring, Van
Nostrand Reinhold Company Inc., New York, NY.
Gish, C.D., and Christensen, R.E., 1973, ?Cadmium, Nickel, Lead and Zinc in Earthworms
from Roadside Soil,?, Environmental Science and Technology, Vol. 7, No. 11, pp. 1060-
1062.
Glick, R.H., Wolfe, M.S., and Thurow, T.S., 1993, Effectiveness of Native Species Buffer
Zones for Nonstructural Treatment of Urban Runoff, Technical Report No. 156, Texas Water
Resources Institute, Texas A&M University.
Howie, B., and Waller, B.G., 1986, Chemical Effects of Highway Runoff on the Surficial
Aquifer, Boward County, Florida, USGS Water Resources Investigations Report 86-4200.
Lagerwerff, J.V., and Specht, A.W., 1970, ?Contamination of Roadside Soil and Vegetation
with Cadmium, Nickel, Lead, and Zinc,? Environmental Science and Technology, Vol. 4,
No. 7, pp. 583-586.
Maryland Department of the Environment, Sediment and Stormwater Administration, 1985,
86
Inspector?s Guidelines Manual for Stormwater Management Infiltration Practices, Baltimore,
MD.
Municipality of Metropolitan Seattle, 1992, Biofiltration Swale Performance,
Recommendations, and Design Considerations, Publication 657, Water Pollution Control
Department, Seattle, WA.
Schueler, T.R., 1987, Controlling Urban Runoff: A Practical Manual for Planning and
Designing Urban BMPs, Department of Environmental Programs, Metropolitan Washington
Council of Governments, Washington, DC.
Schueler, T.R., Kumble, P.A., Heraty, M.A., 1992, A Current Assessment of Urban Best
Management Practices, Techniques for Reducing Non-Point Source Pollution in the Coastal
Zone, Anacostia Restoration Team, Department of Environmental Programs, Metropolitan
Washington Council of Governments, Washington, D.C.
Seattle Engineering Department, 1993, Dayton Avenue Swale Biofiltration Study, Drainage
and Wastewater Utility, Seattle, WA.
U.S. Environmental Protection Agency, 1992, Guidance Manual for the Preparation of Part 2
of the NPDES Permit Applications for Discharges From Municipal Separate Storm Sewer
Systems, EPA 833-B-92-002.
U.S. Environmental Protection Agency, 1993, Urban Runoff Pollution Prevention and
Control Planning, EPA/625/R-93/004, Office of Research and Development, Cincinnati, OH.
Waller , B.G., Klein, H., and Lefkoff, L.J., 1984, Attenuation of Stormwater Contaminants
from Highway Runoff Within Unsaturated Limestone, Dade County, Florida, USGS Water
Resources Investigations Report 84-4083.
Wiginton, P.J., Jr., Randall, C.W., and Grizzard, T.J., 1986, ?Accumulation of Selected Trace
Metals in Soils of Urban Runoff Swale Drains,? Water Resources Bulletin, Vol. 22, No. 1,
pp. 73-79.
Young, G.K., Stein, S., Cole, P., Kammer, T., Graziano, F., and Bank, F., 1996, Evaluation
and Management of Highway Runoff Water Quality, FHWA-PD-96-032, Federal Highway
Administration, Washington, D.C.
Yousef, Y.A., Wanielista, M.P., and Harper, H.H., 1985a, ?Removal of Highway
Contaminants by Roadside Swales,? Transportation Research Record, 1017, pp. 62-68.
Yu, S.L., and Benelmouffok, D.E., 1988, ?Field Testing of Selected Urban BMP?s,? Critical
Water Issues and Computer Applications, Proceedings of the 15
th
 Annual Water Resources
Conference, ASCE, New York, NY, pp. 309-312.
87
Yu, S.L., Kaighn, R.J. Jr., Liao, S.L., and O?Flaherty, C.E., 1995, The Control of Pollution in
Highway Runoff through Biofiltration Volume I: Executive Summary, VTRC 95-R28,
Virginia Transportation Research Council, Charlottesville, VA.
88
APPENDIX A
Individual Sampling Results from Channel Experiments
89
40-meter  L ab S wal e R aw Data
TSS Turbidity COD TOC Nitrate TKN TP Zinc Lead Iron Copper
Sample mg/L NT U mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
F1-0 340 268 48 16.9 0.04 1.107 0.4 0 0.1 0 0
F1-1 240 204 25 6.3 0.06 0.866 0.23 0.1 0.3 6 0
F1-2 200 228 37 6.3 0.09 1.278 0.26 0.1 0.3 6.3 0
F1-3 258 272 31 8.7 0.11 1.422 0.29 0.1 0.4 6.1 0
F1-4 258 252 51 13.4 0.17 1.757 0.43 0.1 0.4 6.6 0
F1-41 312 276 34 6.3 0.12 1.4 0.31 0.1 0.3 5.6 0
F1-42 218 220 29 9.9 0.11 0.866 0.24 0.1 0.3 4.9 0
F1-43 186 236 30 6.3 0.15 1.217 0.26 0.1 0.4 5.9 0
F1-5 152 148 37 6.3 0.15 1.051 0.16 0 0.3 3.3 0
F2-0 594 316 47 32.5 0.13 1.734 0.41 0.251 0.326 11.611 0.021
F2-1 320 292 35 22.1 0.14 1.597 0.34 0.164 0.197 7.819 <.006
F2-21 300 296 41 24.7 0.2 1.649 0.44 0.325 0.181 6.504 <.006
F2-22 226 296 37 22.1 0.17 1.55 0.33 0.184 0.235 7.528 <.006
F2-23 242 292 32 19.5 0.17 1.22 0.31 0.142 0.181 6.624 <.006
F2-24 128 204 28 16.1 0.13 1.224 0.25 0.059 0.125 3.57 <.006
F2-3 292 26 18.5 0.2 1.372 0.3 0.118 0.152 5.905 <.006
F2-4 262 284 31 18.3 0.16 1.194 0.29 0.112 0.19 5.844 <.006
F2-5 160 180 26 14.1 0.19 0.937 0.19 0.031 0.086 3.057 <.006
F3-01 440 240 69 15 0.19 6.344 2.38 0.2 0.4 5.3 <.05
F3-02 624 260 39 18.9 0.19 1.427 0.28 0.2 0.3 6.1 <.05
F3-03 474 230 37 20.7 0.21 1.45 0.31 0.4 0.5 9.9 <.05
F3-04 678 230 31 24.7 0.19 1.15 0.26 0.205 0.138 4.906 0.012
F3-1 300 210 24 15.1 0.21 1.32 0.23 0.14 0.17 2.99 <0.006
F3-2 230 210 23 13.3 0.19 0.967 0.21 0.121 0.125 2.62 <0.006
F3-3 208 210 24 13.2 0.22 0.778 0.18 0.109 0.09 2.434 <0.006
F3-4 194 200 26 13.4 0.21 1.205 0.2 0.096 0.184 2.148 <0.006
F3-5 104 150 21 13.3 0.21 0.923 0.13 0.011 0.046 1.218 <0.006
F4-0A 423 65 35 26.3 <DL 1.349 0.22 0.108 0.176 2.833 <0.006
F4-1A 250 66 27 25.1 <DL 1.123 0.22 0.015 0.165 1.75 <0.006
F4-2A 201 63 21 30.9 <DL 1.233 0.15 0.032 0.06 1.274 <0.006
F4-3A 129 56 18 22.6 <DL 0.896 0.14 0.041 0.073 1.107 <0.006
F4-4A 80 55 13 21.3 <DL 0.891 0.15 0.046 0.065 0.93 <0.006
F4-5A 47 34 12 18.9 <DL 0.797 0.08 <0.002 0.063 1.817 <0.006
F4-0B 184 67 34 33 <DL 1.269 0.23 0.15 0.136 3.168 <0.006
F4-1B 111 66 24 26.3 <DL 1.039 0.17 0.024 <0.042 1.681 <0.006
F4-2B 94 60 21 20.9 <DL 0.551 0.22 0.039 0.086 4.582 <0.006
F4-3B 80 54 21 24.2 <DL 3.723 0.36 0.029 0.088 0.979 <0.006
F4-4B 73 40 16 19.7 <DL 0.303 0.15 0.041 0.042 0.802 <0.006
F4-5B 46 32 14 19.3 <DL 0.929 0.1 0.006 0.066 0.478 <0.006
F5-01 207 83 46 5.7 0.14 1.778 0.18 0.162 0.228 4.642 <0.006
F5-02 174 88 39 5.7 0.15 1.877 0.2 0.155 0.137 3.446 <0.006
F5-03 191 88 39 9 0.15 2.025 0.2 0.145 0.15 3.835 <0.006
F5-04 238 86 42 9 0.15 1.646 0.18 0.111 0.181 2.525 <0.006
F5-05 201 85 39 5.7 0.15 1.795 0.19 0.153 0.295 3.644 <0.006
F5-06 206 85 25 16.3 0.15 1.717 0.19 0.247 0.205 4.175 <0.006
F5-1 144 83 43 9.6 0.18 1.203 0.13 0.09 0.089 2.366 <0.006
F5-2 107 79 30 9.6 0.12 1.075 0.11 0.053 0.103 1.597 <0.006
F5-3 85 73 27 6.2 0.11 1.052 0.1 0.033 0.067 1.4 <0.006
F5-4 86 71 27 6.2 0.12 1.013 0.1 0.038 0.048 0.886 <0.006
F5-5 28 39 20 2.9 0.12 0.784 0.02 <0.002 0.127 0.758 <0.006
90
TSS Turbidity COD TOC Ni trate TKN TP Zinc Lead Iron Copper
Sampl e mg/L NT U mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
F6-0A 300 200 47 40.2 0.17 2.025 0.28 0.26 NA 3.491 0.034
F6-1A 159 164 38 14.3 0.19 1.525 0.2 0.15 NA 1.758 0.018
F6-2A 108 128 22 16.1 0.19 1.468 0.13 0.107 NA 1.395 0.007
F6-3A 66 120 27 11.5 0.21 1.058 0.11 0.069 NA 0.886 0.008
F6-4A 90 104 16 11.5 0.19 1.398 0.23 0.034 NA 0.893 0.009
F6-5A 58 104 7 11.7 0.19 1.447 0.1 0.059 NA 0.866 0.01
F6-0B 290 176 93 40.4 0.23 2.153 0.38 0.26 NA 3.447 0.03
F6-1B 142 152 16 21.2 0.23 1.519 0.15 0.139 NA 1.976 0.014
F6-2B 100 136 20 16.5 0.25 1.116 0.34 0.108 NA 1.789 0.012
F6-3B 76 120 27 14.1 0.23 1.35 0.11 0.057 NA 1.022 0.008
F6-4B 60 100 14 13.9 0.23 1.334 0.13 0.038 NA 0.745 <0.006
F6-5B 60 100 26 13.9 0.25 1.032 0.16 0.045 NA 0.856 0.01
F7-0A 320 160 63 50.2 0.12 2.267 0.46 0.178 NA 2.949 0.026
F7-1A 104 100 64 22.3 0.1 1.375 0.25 0.06 NA 1.204 0.012
F7-2A 75 80 18 22.1 0.07 1.755 0.38 0.016 NA 0.951 0.007
F7-0B 286 168 90 48.1 0.11 2.006 0.62 0.221 NA 3.501 0.041
F7-1B 124 98 29 47.3 0.11 1.482 0.23 0.163 NA 2.567 0.016
F7-2B 72 82 25 22.4 0.1 1.504 0.25 0.089 NA 1.785 0.014
F7-0C 323 168 75 30.7 0.12 2.287 0.49 0.372 NA 5.626 0.053
F7-1C 134 100 33 22.7 0.14 1.881 0.23 0.139 NA 2.347 0.017
F7-2C 70 79 27 25.3 0.12 1.987 0.22 0.071 NA 1.612 0.012
F8-0A 384 80 30 NA 0.16 1.494 0.22 0.26 NA NA NA
F8-1A 76 77 23 NA <0.1 1.466 0.14 0.21 NA NA NA
F8-2A 92 65 19 NA <0.1 1.266 0.12 0.19 NA NA NA
F8-3A 68 62 22 NA <0.1 1.168 0.11 0.21 NA NA NA
F8-4A 58.5 41 24 NA <0.1 1.284 0.12 0.11 NA NA NA
F8-5A 36 44 15 NA 0.12 1.235 0.09 0.1 NA NA NA
F8-0B 275 86 32 NA 0.13 1.41 0.21 0.26 NA NA NA
F8-1B 129 79 32 NA 0.13 1.506 0.15 0.15 NA NA NA
F8-2A 86 69 25 NA 0.24 1.125 0.12 0.12 NA NA NA
F8-3B 71 49 28 NA 0.35 1.13 0.12 0.1 NA NA NA
F8-4B 60 59 25 NA 0.33 0.879 0.19 0.09 NA NA NA
F8-5B 37 45 16 NA 0.24 1.107 0.09 0.06 NA NA NA
F9-0A 420 160 36 NA 0.16 1.976 0.23 0.34 NA NA NA
F9-0B 344 180 33 NA 0.14 1.567 0.26 0.33 NA NA NA
F9-0C 302 180 28 NA 0.14 1.992 0.26 0.31 NA NA NA
F9-1A 218 150 24 NA 0.14 1.358 0.2 0.22 NA NA NA
F9-1B 118 150 28 NA 0.17 1.168 0.18 0.18 NA NA NA
F9-1C 128 150 26 NA 0.17 1.23 0.17 0.19 NA NA NA
F10-0A 226 160 44 NA 0.11 1.513 0.29 0.23 NA NA NA
F10-0B 218 170 41 NA 0.14 1.433 0.24 0.26 NA NA NA
F10-1A 140 150 30 NA <0.1 1.452 0.15 NA NA NA
F10-1B 150 160 36 NA 0.14 1.387 0.22 0.16 NA NA NA
F10-2A 110 150 34 NA 0.12 1.21 0.13 NA NA NA
F10-2B 96 140 29 NA 0.14 1.121 0.18 0.4 NA NA NA
F10-3A 110 150 31 NA 0.14 1.225 0.16 0.243 NA NA NA
F10-3B 112 150 32 NA 0.14 1.125 0.17 0.14 NA NA NA
F10-4A 116 120 33 NA 0.11 1.039 0.14 NA NA NA
F10-4B 102 160 31 NA 0.15 1.225 0.16 NA NA NA
F10-5A 58 150 23 NA 0.12 1.135 0.12 0.09 NA NA NA
F10-5B 60 120 29 NA 0.15 1.098 0.12 0.17 NA NA NA
F11-0B 271 200 NA NA 0.15 1.591 0.27 0.38 NA NA NA
F11-1B 130 170 27 NA 0.13 1.22 0.15 0.19 NA NA NA
F11-2B 79 150 25 NA <0.1 1.326 0.11 0.135 NA NA NA
F11-5B 54 120 24 NA 0.18 1.215 0.1 0.169 NA NA NA
F11-0A 280 190 NA NA 0.14 1.925 0.35 NA NA NA
F11-1A 127 170 30 NA 0.13 1.353 0.15 0.28 NA NA NA
F11-2A 71 150 26 NA 0.13 1.593 0.17 NA NA NA
F11-5A 55 120 24 NA 0.14 0.851 0.13 0.11 NA NA NA
91
APPENDIX B
Flow Data and Sample Concentrations
For Four Field Monitoring Locations
92
Highway runoff at U.S. 183 site
Sample Date/Time Flow Vol
Cum
Flow
TSS
Conc
Turbidity
Conc
Fecal Col
Conc
Fecal Str
Conc
E. coli
Conc
COD
Conc
TOC
Conc
Nitrate
Conc
TKN
Conc
Total P
Conc
Zinc
Conc
Lead
Conc
Iron
Conc
Copper
Conc
No. Collected L L mg/L NTU CFU/100ml CFU/100ml CFU/100ml mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
Storm 12
1 5/27/96 7:53 8090 8090 882 228 CG 670000 77000 431 212.0 4.40 NA 1.77 1.447 0.394 14.516 <0.006
2 5/27/96 8:07 86980 95070 86 37 6000 13400 290 28 24.4 0.84 NA 0.38 0.250 0.165 2.76 <0.006
3 5/27/96 8:22 22240 117310 14 50 7300 3700 1500 11 12.9 1.50 NA 0.22 0.049 0.099 0.890 <0.006
Storm 13
1 5/30/96 1:54 2140 2140 30 12 CG 21000 CG 93 39.7 1.70 NA 0.62 0.013 0.111 0.655 <0.006
2 5/30/96 2:09 3070 5210 52 35 435000 10000 CG 13 12.8 0.85 NA 0.38 0.002 0.117 2.160 <0.006
3 5/30/96 2:24 3450 8660 14 17 CG 25000 CG 0 12.8 0.85 NA 0.33 0.002 0.094 0.572 <0.006
4 5/30/96 2:54 126320 134980 8 11 129000 12000 16000 24 13.3 6.60 NA 0.35 0.002 0.068 0.398 <0.006
5 5/30/96 3:24 40680 175660 0 6 226000 29000 14000 35 16.3 1.75 NA 0.38 0.002 0.106 0.108 <0.006
6 5/30/96 4:24 590 176250 8 5 CG 30000 50000 43 16.8 5.80 NA 0.40 0.002 0.104 0.176 <0.006
Storm 15
1 6/22/96 11:38 4800 4800 328 47 NA NA NA 612 82.7 3.95 7.224 0.74 0.595 0.198 6.710 0.015
2 6/22/96 11:52 1890 6690 52 35 NA NA NA 66 34.3 1.60 2.617 0.26 0.114 0.195 1.889 <0.002
3 6/22/96 12:07 100 6790 46 47 NA NA NA NA 14.4 NA NA NA NA NA NA NA
4 6/22/96 12:37 10 6800 NA NA NA NA NA NA 11.8 NA NA NA NA NA NA NA
Storm 16
1 6/25/96 10:35 5450 5450 0 50 CG 14700 3000 179 50.4 7.30 3.465 0.70 0.851 0.264 8.296 0.007
2 6/25/96 10:50 7650 13100 64 15 138000 4000 700 16 15.9 7.00 1.109 0.22 0.058 0.165 1.440 <0.006
3 6/25/96 11:05 700 13800 52 27 12000 400 0 30 15.9 1.80 1.198 0.20 0.022 0.127 1.129 <0.006
4 6/25/96 11:35 3590 17390 252 13 4000 200 0 30 14 1.28 1.477 0.18 0.006 0.191 0.612 <0.006
5 6/25/96 12:05 1090 18480 84 40 5200 6400 90 45 21 4.95 0.936 0.22 0.143 0.201 1.388 <0.006
Storm 19
1 8/22/96 10:03 5710 5710 32 46 NA NA NA 209 37.6 2.7 3.089 0.52 0.289 0.169 4.457 0.006
93
2 8/22/96 10:17 230 5940 9 26 NA NA NA 38 9 1.55 0.61 0.2 0.038 0.092 0.840 <0.006
Storm 20
1 8/23/96 17:07 3590 3590 42 39 NA NA NA 124 41.9 1.6 1.786 0.31 0.122 0.085 1.609 <0.006
2 8/23/96 17:21 1710 5300 19 15 NA NA NA 36 9.1 0.54 0.846 0.2 0.002 0.042 0.422 <0.006
3 8/23/96 17:36 1400 6700 7 14 NA NA NA 35 6.9 0.47 0.416 0.2 0.002 0.042 0.402 <0.006
4 8/23/96 18:06 2470 9170 7 13 NA NA NA 37 11.2 0.77 1.259 0.18 0.002 0.042 0.787 <0.006
5 8/23/96 18:36 3240 12410 4 15 NA NA NA 28 2.6 0.54 1.38 0.15 0.002 0.042 0.227 <0.006
6 8/23/96 19:36 2850 15260 10 17 NA NA NA 20 NA 0.44 0.796 0.15 0.002 0.042 0.451 <0.006
Storm 21
1 8/29/96 12:09 3480 3480 22 42 510000 4000 2100 128 43.8 1.15 0.325 0.39 0.154 0.077 2.680 <0.006
2 8/29/96 12:23 200 3680 14 19 130000 5700 2800 34 7.5 0.67 1.299 0.15 0.002 0.042 0.560 <0.006
Storm 22
1 9/18/96 15:31 9250 9250 360 120 CG 12700 NA 163 61.1 2.50 3.329 0.52 0.321 0.153 5.555 <0.006
2 9/18/96 15:45 790 10040 36 34 CG 2300 NA 30 15 1.80 1.462 0.23 0.019 0.073 2.245 <0.006
3 9/18/96 16:00 20880 30920 47 64 CG 7000 NA 99 33.5 2.20 1.827 0.35 0.046 0.074 1.638 <0.006
4 9/18/96 16:30 1380 32300 23 24 CG 3400 NA 19 3.7 1.60 0.853 0.17 0.019 0.042 1.318 <0.006
5 9/18/96 17:00 20 32320 10 23 500000 4300 NA 24 8.1 1.75 0.745 0.2 0.016 0.042 0.603 <0.006
Storm 23
1 10/17/96 16:42 17300 17300 65 40 <200000 26000 NA 125 47.9 1.15 3.147 1.12 1.100 0.3 10.3 0.1
2 10/17/96 16:56 1100 18400 44 25 230000 <20000 NA 32 7.2 1.10 1.180 0.26 0.100 0.050 1.1 <0.05
Storm 24
5 10/27/96 14:00 5410 5410 334 116 20000 59000 NA NA 60.1 0.46 6.192 2.24 1.200 0.3 8.5 0.1
6 10/27/96 14:02 910 6320 184 40 4800 20100 NA 88 55.6 0.55 1.862 0.65 0.500 0.1 5.0 <0.05
94
Filter strip discharge at U.S. 183 site
TSS Turbidity Fecal Col Fecal Str E. coli COD TOC Nitrate TKN Total P Zinc Lead Iron Copper
Sample Date/Time Flow
Vol
Cum
Flow
Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc
No. Collected L L mg/L NTU CFU/100ml CFU/100ml CFU/100ml mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
Storm 9
1 3/27/96 14:36 5130 5130 5 12 NA 28500 NA 51 22.3 0.49 0.28 0.07 0.024 0.042 0.212 <0.006
2 3/27/96 15:05 9720 14850 5 19 NA NA NA 57 22.3 0.81 0.26 0.18 0.019 0.042 0.347 <0.006
3 3/27/96 15:35 9920 24770 5 26 6136 NA NA 55 21.0 0.86 0.26 0.24 0.014 0.042 0.365 <0.006
4 3/27/96 16:05 8770 33540 4 21 NA NA NA 40 17.9 0.78 0.26 0.23 0.002 0.042 0.262 <0.006
5 3/27/96 17:05 16830 50370 2 23 NA NA NA 32 32.4 0.55 0.26 0.21 0.006 0.042 0.262 <0.006
6 3/27/96 18:05 10020 60390 3 21 NA NA NA 30 16.8 0.41 0.26 0.2 0.012 0.042 0.295 <0.006
Storm 10
1 4/5/96 17:19 1360 1360 40 NA NA NA NA 61 28.9 NA NA 0.38 0.016 0.042 0.320 <0.006
2 4/5/96 17:48 13540 14900 24 NA NA NA NA 26 18.3 NA NA 0.26 0.007 0.042 0.331 <0.006
3 4/5/96 18:18 14670 29570 16 NA NA NA NA 18 16.6 NA NA 0.23 0.002 0.042 0.238 <0.006
4 4/5/96 18:48 15760 45330 16 NA NA NA NA 5 13.1 NA NA 0.15 0.002 0.042 0.232 <0.006
5 4/5/96 19:48 29020 74350 16 NA NA NA NA 0 11.3 NA NA 0.16 0.002 0.042 0.430 <0.006
Storm 11
1 4/22/96 12:43 1520 1520 17 10 NA NA NA 82 25.5 0.98 NA 0.46 0.002 0.089 0.253 <0.006
2 4/22/96 13:12 13760 15280 6 7 NA NA NA 67 21.2 0.80 NA 0.36 0.002 0.122 0.223 <0.006
3 4/22/96 13:42 10700 25980 3 9 NA NA NA 64 24.9 0.82 NA 0.33 0.002 0.095 0.992 <0.006
4 4/22/96 14:12 11900 37880 5 5 NA NA NA 58 25.5 0.80 NA 0.30 0.002 0.067 0.211 <0.006
5 4/22/96 15:12 4240 42120 5 9 NA NA NA 60 20.4 0.73 NA 0.30 0.002 0.097 0.101 <0.006
Storm 12
1 5/27/96 8:01 33120 33120 188 40 310000 4400 36000 60 66.4 1.10 NA 1.40 0.090 0.208 3.672 <0.006
2 5/27/96 8:30 68510 101630 24 23 440000 4700 610000 79 26.2 0.37 NA 0.78 0.002 0.086 0.990 <0.006
3 5/27/96 9:00 24790 126420 6 25 510000 4900 630000 85 24.4 0.40 NA 0.85 0.002 0.117 0.332 <0.006
4 5/27/96 9:30 12450 138870 14 20 158000 590 63000 83 26.1 0.37 NA 0.89 0.003 0.086 0.214 <0.006
5 5/27/96 10:30 10010 148880 12 14 159000 560 24000 83 31.2 0.40 NA 0.88 0.002 0.094 0.192 <0.006
95
Storm 13
1 5/30/96 2:37 29610 29610 200 64 36000 34000 21000 131 39.6 2.70 NA 0.52 0.364 0.168 3.461 <0.006
2 5/30/96 3:07 169080 198690 36 38 <20000 4800 13000 45 16.6 6.20 NA 0.23 0.070 0.079 1.147 <0.006
3 5/30/96 3:37 37940 236630 0 25 4500 2900 3500 22 8.1 1.00 NA 0.16 0.027 0.076 0.562 <0.006
4 5/30/96 4:07 18200 254830 130 34 1800 3600 6000 7 11.7 0.46 NA 0.55 0.313 0.206 3.104 <0.006
Storm 15
1 6/22/96 16:38 2110 2110 32 15 NA NA NA 45 19.0 0.76 2.55 0.54 0.002 0.042 0.356 <0.006
2 6/22/96 17:08 33270 35380 52 30 NA NA NA 46 22.9 3.85 2.22 0.49 0.002 0.117 1.548 <0.002
3 6/22/96 17:37 17850 53230 24 20 NA NA NA 45 20.7 0.83 2.04 0.42 0.002 0.161 0.557 <0.006
4 6/22/96 18:07 9100 62330 16 10 NA NA NA 43 17.8 NA 0.74 0.41 0.002 0.060 0.258 <0.006
Storm 16
1 6/25/96 11:00 13430 13430 0 6.4 990000 55000 140000 60 27.4 NA 2.62 0.39 0.013 0.180 0.580 <0.006
2 6/25/96 11:28 33340 46770 52 14 560000 24000 21000 15 18.9 5.40 1.76 0.27 0.002 0.148 0.626 <0.006
3 6/25/96 11:59 35520 82290 52 10 8900 25000 500 24 15.9 4.40 1.48 0.21 0.002 0.145 0.481 <0.006
4 6/25/96 12:28 22190 104480 48 7 360000 12400 11000 29 17.8 1.00 1.32 0.19 0.002 0.123 0.436 <0.006
5 6/25/96 13:28 11470 115950 40 9 260000 8800 1400 26 16.3 1.90 2.15 0.18 0.002 0.149 0.404 <0.006
Storm 18
1 8/11/96 14:45 67360 67360 116 31 CG 32000 560000 77 16.4 0.78 15.1 0.51 0.002 0.164 1.001 <0.006
2 8/11/96 15:15 46490 113850 20 17 1590000 7300 300000 42 13.1 0.36 1.76 0.35 0.002 0.141 0.856 <0.006
3 8/11/96 15:45 23960 137810 8 6.6 1890000 7000 460000 66 14.1 1 1.65 0.37 0.002 0.165 0.311 <0.006
4 8/11/96 16:15 19540 157350 48 4.7 1410000 5000 510000 60 17.4 0.5 1.81 0.37 0.002 0.131 0.174 <0.006
5 8/11/96 17:15 10060 167410 4 3.2 770000 5100 420000 70 20.7 0.44 1.83 0.38 0.002 0.145 0.110 <0.006
6 8/11/96 18:15 2990 170400 8 2.5 420000 8800 50000 79 23 0.26 3.11 0.48 0.002 0.144 0.148 <0.006
Storm 19
1 8/22/96 19:14 9810 9810 8 5 NA NA NA 86 25.9 0.27 2.42 0.53 0.002 0.124 0.152 <0.006
2 8/22/96 19:43 12150 21960 1 10 NA NA NA 67 19.9 0.4 1.6 0.31 0.002 0.090 0.273 <0.006
3 8/22/96 20:13 7100 29060 2 9 NA NA NA 58 15.9 0.3 1.79 0.26 0.002 0.093 0.201 <0.006
4 8/22/96 20:43 6350 35410 1 7 NA NA NA 55 15.9 0.21 1.42 0.25 0.002 0.101 0.198 <0.006
Storm 20
1 8/23/96 17:51 8490 8490 5 4 NA NA NA 76 29 0.32 2.9 0.4 0.002 0.043 0.120 <0.006
2 8/23/96 18:20 31780 40270 7 14 NA NA NA 61 20.2 0.33 1.21 0.27 0.002 0.042 0.285 <0.006
96
3 8/23/96 18:50 27500 67770 3 11 NA NA NA 54 13.9 0.23 1.04 0.2 0.002 0.042 0.351 <0.006
4 8/23/96 19:20 34580 102350 8 7 NA NA NA 44 13.7 0.16 1.15 0.19 0.002 0.064 0.276 <0.006
5 8/23/96 20:20 31650 134000 4 4 NA NA NA 33 11.7 0.1 0.89 0.16 0.002 0.093 0.176 <0.006
6 8/23/96 21:20 15470 149470 2 3 NA NA NA 36 11.7 0.1 1.05 0.15 0.002 0.042 0.068 <0.006
Storm 21
1 8/29/96 15:16 68130 68130 90 33 240000 91000 26000 43 14 1 1 0.44 0.003 0.068 2.064 <0.006
2 8/29/96 15:45 39410 107540 62 57 98000 74000 10000 26 9.5 2.1 1 0.27 0.002 0.069 3.725 <0.006
3 8/29/96 16:15 20380 127920 16 19 55000 44000 8000 22 7.5 0.79 1 0.22 0.002 0.042 0.904 <0.006
4 8/29/96 16:45 15570 143490 0 11 43000 34000 12000 19 7.5 1.15 NA 0.15 0.002 0.042 0.425 <0.006
5 8/29/96 17:45 8200 151690 5 6.5 2300 30000 10000 23 9.6 3.35 1.58 0.15 0.002 0.042 0.513 <0.006
Storm 22
1 9/18/96 16:11 22870 22870 5 5 NA NA NA 27 5.6 1.5 NA NA 0.002 0.042 0.411 <0.006
2 9/18/96 16:40 35580 58450 10 9 NA NA NA 23 7.8 1.3 0.89 0.38 0.002 0.090 0.598 <0.006
3 9/18/96 17:10 20810 79260 7 12 NA NA NA 24 10 1.2 1.14 0.28 0.002 0.042 0.439 <0.006
4 9/18/96 17:40 16140 95400 3 7 NA NA NA 21 10 1.25 0.77 0.26 0.002 0.042 0.222 <0.006
5 9/18/96 18:40 7690 103090 1 5 NA NA NA 21 8.4 1.3 0.57 0.26 0.002 0.042 0.675 <0.006
Storm 25
1 11/7/96 2:02 33830 33830 19 14 NA NA NA 57 35.0 0.30 1.996 0.99 0.002 0.1 0.3 <0.0
2 11/7/96 2:31 48830 82660 34 16 NA NA NA 27 17.9 0.20 1.179 0.39 0.050 0.042 0.5 <0.0
3 11/7/96 3:01 28150 110810 18 13 NA NA NA 23 14.1 0.22 0.320 0.35 0.1 0.042 0.4 <0.0
4 11/7/96 3:31 32430 143240 3 13 NA NA NA 21 14.1 0.17 0.804 0.33 0.002 0.042 0.218 <0.006
5 11/7/96 4:31 45110 188350 3 6.4 NA NA NA 24 16.6 0.15 0.122 0.31 0.002 0.042 0.2 <0.0
6 11/7/96 5:31 34040 222390 3 6.5 NA NA NA 21 18.7 0.21 0.194 0.26 0.002 0.042 0.222 <0.006
Storm 28
1 12/15/96 5:06 30430 30430 23.00 13 NA NA NA 38 20.0 0.74 1.702 NA 0.002 0.067 0.330 <0.006
2 12/15/96 5:35 73040 103470 13.00 14 NA NA NA 18 13.8 0.45 1.015 NA 0.132 0.042 0.510 <0.006
3 12/15/96 6:05 66530 170000 9.00 11 NA NA NA 5 13.5 0.21 0.508 NA 0.002 0.049 0.278 <0.006
4 12/15/96 6:35 85060 255060 4.00 10 NA NA NA 0.0 13.5 0.26 0.522 NA 0.002 0.042 0.320 <0.006
5 12/15/96 7:35 114750 369810 3.00 10 NA NA NA 13 15.4 0.15 0.457 NA 0.002 0.042 0.213 <0.006
6 12/15/96 8:35 105530 475340 4.00 12 NA NA NA 3 13.5 0.11 0.406 NA 0.002 0.042 0.337 <0.006
97
Storm 26
1 11/24/96 5:22 11870 11870 8.00 22 NA NA NA 58 19.3 0.81 2.130 0.54 0.002 0.080 0.398 <0.006
2 11/24/96 5:51 25510 37380 7.00 14 NA NA NA 40 8.8 0.42 1.475 0.31 0.075 0.042 0.334 <0.006
3 11/24/96 6:21 37160 74540 0 17 NA NA NA 30 10.0 0.19 1.165 0.22 0.002 0.042 0.490 <0.006
4 11/24/96 6:51 46660 121200 12.00 24 NA NA NA 25 6.3 0.19 0.985 0.22 0.111 0.042 0.292 <0.006
5 11/24/96 7:51 97870 219070 5.00 19 NA NA NA 21 4.4 0.13 0.664 0.17 0.002 0.042 0.751 <0.006
6 11/24/96 8:51 75910 294980 7.00 18 NA NA NA 14 5.0 0.11 0.537 0.15 0.002 0.046 0.532 <0.006
Storm 29
1 2/7/97 6:56 14120 14120 16 26 9000 220000 NA 77 24.7 1.9 2.88 0.79 0.025 0.096 0.78 0.012
2 2/7/97 7:25 25750 39870 5 25 2400 300000 NA 44 18.1 1.6 1.56 0.31 0.012 0.049 0.523 0.012
3 2/7/97 7:55 21300 61170 6 25 2000 37800 NA 43 17.3 1.4 0.92 0.22 0.022 0.116 0.619 0.011
4 2/7/97 8:25 22210 83380 10 25 2000 38000 NA 33 15.5 1.15 0.88 0.21 0.118 0.093 0.794 0.02
5 2/7/97 9:25 39370 122750 4 21 2500 2500 NA 37 16.8 0.76 0.97 0.19 0.032 0.042 0.639 0.011
6 2/7/97 10:25 38670 161420 5 25 2000 24000 NA 34 16.2 0.69 1.33 0.21 0.053 0.093 0.799 0.011
Storm 30
1 2/12/97 6:08 20180 20180 17 23 NA NA NA 44 20.8 1.05 1.43 0.38 0.025 0.455 0.008
2 2/12/97 6:37 40080 60260 7 25 NA NA NA 25 22.5 1.4 0.85 0.16 0.042 1.054 0.011
3 2/12/97 7:07 41290 101550 1 18 NA NA NA 12 14.2 0.96 0.72 0.14 0.025 0.529 0.01
4 2/12/97 7:37 64780 166330 3 20 NA NA NA 8 12.6 0.65 1.48 0.13 0.02 0.621 0.008
5 2/12/97 8:37 60960 227290 6 23 NA NA NA 13 10.9 0.49 0.79 0.11 0.028 0.857 0.015
6 2/12/97 9:37 30650 257940 6 21 NA NA NA 12 10.9 0.4 0.7 0.09 0.025 0.706 0.008
Storm 31
1 3/11/97 11:54 11660 11660 34 21 NA NA NA 64 17.9 0.35 1.72 0.46 0.06
2 3/11/97 12:24 21520 33180 30 39 NA NA NA 47 25.5 0.22 1.56 0.26 0.12
3 3/11/97 12:54 21520 54700 16 33 NA NA NA 35 15.6 0.24 1.02 0.23 0.06
4 3/11/97 13:24 30360 85060 14 27 NA NA NA 27 15.2 0.11 1.2 0.23 0.04
5 3/11/97 14:24 25470 110530 9 22 NA NA NA 24 12.9 0.1 1.04 0.2 0.26
6 3/11/97 15:24 11560 122090 6 19 NA NA NA 25 12.9 0.1 0.77 0.18 0.05
Storm 32
1 3/25/97 11:41 25280 25280 14 19 5800 40000 NA 44 21.3 0.61 2.12 NA* 0.09
2 3/25/97 12:10 37170 62450 NA 19 3200 25000 NA 34 17.3 0.47 1.25 NA 0.07
98
3 3/25/97 12:40 24250 86700 NA 17 930 26600 NA 26 15.5 0.47 0.89 NA 0.05
4 3/25/97 13:10 22580 109280 1.5 16 500 43000 NA 30 15.5 0.32 1.48 NA 0.06
5 3/25/97 14:10 21720 131000 3 15 480 13500 NA 26 16 0.27 1.01 NA 0.06
6 3/25/97 15:10 44860 175860 4.5 16 240 23200 NA 22 14.8 0.31 1.26 NA 0.07
Storm 33
1 4/2/97 18:31 9940 9940 14 16 116000 114000 NA 48 16.8 1.06 1.46 0.46 0.05
2 4/2/97 19:01 11710 21650 7 20 65000 110000 NA 46 20.3 0.94 1.33 0.28 0.04
3 4/2/97 19:31 9660 31310 8.5 20 22000 105000 NA 41 18 1.06 0.97 0.27 0.04
4 4/2/97 20:01 10030 41340 6 17 20000 126000 NA 38 16.8 0.85 0.97 0.21 0.08
5 4/2/97 21:01 13230 54570 2 15 10000 98000 NA 35 18 0.21 0.85 0.17 0.08
6 4/2/97 22:01 10810 65380 1 12 5900 60000 NA 32 14.5 0.1 0.86 0.3 0.03
Storm 34
1 4/25/97 10:51 27260 27260 14.5 3.5 NA NA NA 41 23.8 0.91 2.21 0.5 0.09
2 4/25/97 11:20 44640 71900 5.5 4.9 NA NA NA 15 13.2 0.38 0.99 0.23 0.07
3 4/25/97 11:50 48150 120050 3 2.9 NA NA NA 14 11.5 0.34 1.1 0.17 0.04
4 4/25/97 12:20 92120 212170 2 4 NA NA NA 16 8.6 0.32 1.1 0.19 0.1
5 4/25/97 13:20 151340 363510 4 3.9 NA NA NA 11 7 0.25 0.86 0.12 0.07
6 4/25/97 14:20 63060 426570 2.5 2.4 NA NA NA 17 7.3 0.15 0.85 0.15 0.05
Storm 35
1 5/9/97 6:54 68120 68120 47 7.4 NA 77000 NA 72 NA 0.77 2.4 NA 0.06
2 5/9/97 7:24 117780 185900 40 7.5 NA 21900 NA 34 NA 0.35 1.49 NA 0.05
3 5/9/97 7:54 69090 254990 12.5 6.4 NA 17700 NA 36 NA 0.3 1.13 NA 0.05
4 5/9/97 8:24 70960 325950 5.5 5.7 NA 17100 NA 41 NA 0.36 1.22 NA 0.03
5 5/9/97 9:24 71230 397180 1 5 NA 15500 NA 44 NA 0.39 1.22 NA 0.07
6 5/9/97 10:24 64870 462050 4.5 4.6 NA 13400 NA 35 NA 0.41 0.86 NA 0.06
Storm 36
1 5/27/97 16:03 115150 115150 28 13 31000 55000 NA 32 13.8 0.67 1.34 0.27 0.06
2 5/27/97 16:33 74090 189240 NA* 10 8900 39000 NA 18 6.2 0.42 1.25 0.12 0.08
3 5/27/97 17:03 42570 231810 NA* 10 12100 28000 NA 18 9.6 0.36 1.27 0.15 0.04
4 5/27/97 17:33 37460 269270 4 11 13000 21000 NA 18 11.9 0.3 0.58 0.16 0.1
5 5/27/97 18:33 23520 292790 1 11 20000 36000 NA 29 16.6 0.3 0.5 0.17 0.1
99
6 5/27/97 19:34 8540 301330 NA* 10 5500 14700 NA 24 15.6 0.26 0.5 0.12 0.14
100
Highway runoff at Walnut Creek site
TSS Turbidity Fecal Col Fecal Str E. coli COD TOC Nitrate TKN Total P Zinc Lead Iron Copper
Sample Date/Time Flow
Vol
Cum
Flow
Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc
No. Collected L L mg/L NTU CFU/100ml CFU/100ml CFU/100ml mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
Storm 6
1 2/29/96 10:23 40 40 522 192 NA NA NA 232 NA 4.400 3.15 0.58 0.388 0.116 7.489 0.023
2 2/29/96 10:53 1080 1120 430 164 NA NA NA 153 NA 1.330 2.06 0.54 0.272 0.097 6.301 <0.006
3 2/29/96 11:23 260 1380 328 140 NA NA NA 116 NA 1.320 1.80 0.46 0.244 0.057 4.989 0.007
4 2/29/96 12:23 1550 2930 228 108 NA NA NA 118 NA 0.520 1.10 0.33 0.209 0.042 3.196 <0.006
5 2/29/96 13:23 2050 4980 199 62 NA NA NA 80 NA 0.350 0.84 0.24 0.110 0.042 3.473 <0.006
6 2/29/96 14:23 400 5380 125 58 NA NA NA 65 NA 0.300 0.85 0.17 0.088 0.042 2.077 <0.006
Storm 8
1 3/26/96 1:41 20 20 560 54 NA NA NA 324 102.9 3.400 4.36 0.43 0.187 0.042 2.281 <0.006
2 3/26/96 2:10 70 90 456 26 NA NA NA 291 123.1 5.900 NA 0.38 0.180 0.042 1.611 <0.006
Storm 10
1 4/5/96 11:29 0 0 340 NA NA NA NA 226 89.9 NA NA 0.31 0.160 0.042 2.331 <0.006
2 4/5/96 11:59 270 270 460 NA NA NA NA 217 81.7 NA NA 0.33 0.184 0.042 2.781 0.007
3 4/5/96 12:29 40 310 688 NA NA NA NA 209 75.1 NA NA 0.010 0.149 0.042 2.282 0.007
4 4/5/96 13:29 0 310 392 NA NA NA NA 219 68.5 NA NA 0.010 0.129 0.042 2.002 <0.006
5 4/5/96 14:29 410 720 296 NA NA NA NA 122 45.2 NA NA 0.010 0.137 0.042 2.335 0.007
6 4/5/96 15:29 540 1260 660 NA NA NA NA 133 50.6 NA NA 0.010 0.193 0.042 3.233 <0.006
7 4/5/96 17:03 1240 2500 664 NA NA NA NA 99 45.4 NA NA 0.010 0.149 0.042 2.505 <0.006
8 4/5/96 18:03 2110 4610 308 NA NA NA NA 52 23.3 NA NA 0.010 0.113 0.042 1.613 <0.006
9 4/5/96 19:03 260 4870 152 NA NA NA NA 53 19.8 NA NA 0.010 NA NA NA NA
12 4/5/96 22:03 1220 6090 400 NA NA NA NA 49 21.7 NA NA 0.010 NA NA NA NA
Storm 11
1 4/28/96 23:06 320 320 706.00 124 NA NA NA 353 90 5.890 NA 0.12 0.394 0.277 4.916 <0.006
2 4/28/96 23:36 1070 1390 344.00 45 NA NA NA 71 24.1 4.600 NA 0.29 0.056 0.072 1.100 <0.006
3 4/29/96 0:06 270 1660 180.00 55 NA NA NA 87 31.6 1.430 NA 0.28 0.082 0.097 2.362 <0.006
101
4 4/29/96 1:06 30 1690 175.00 55 NA NA NA 100 43.6 0.820 NA 0.26 0.097 0.085 1.600 <0.006
Storm 12
1 5/27/96 7:15 180 180 554 192 NA NA NA 515 178.4 2.200 NA 0.91 0.666 0.37 12.17 0.063
2 5/27/96 7:45 1560 1740 376 168 NA NA NA 169 52.5 1.100 NA 0.35 0.132 0.129 3.353 <0.006
3 5/27/96 8:15 5620 7360 32 26 NA NA NA 16 19.5 0.650 NA 0.08 0.013 0.119 1.243 <0.006
4 5/27/96 9:15 410 7770 0 40 NA NA NA 30 12.9 0.650 NA 0.07 0.021 0.056 0.443 <0.006
Storm 14
1 6/4/96 3:22 900 900 82 31 CG 8000 CG 116 53.6 2.31 NA 0.58 0.335 0.394 6.816 0.031
2 6/4/96 3:52 5200 6100 16 7 12000 2600 11000 28 3.8 0.49 NA 0.05 0.002 0.042 0.509 <0.006
3 6/4/96 4:22 1160 7260 8 5 1000 2800 45000 21 4.4 0.49 NA 0.04 0.002 0.048 0.238 <0.006
4 6/4/96 5:22 80 7340 0 4 1600 2700 18000 20 6.2 0.85 NA 0.04 0.002 0.014 0.194 <0.006
5 6/4/96 6:22 0 7340 0 4 145000 9300 27000 30 9.9 0.98 NA 0.05 0.002 0.042 0.22 <0.006
6 6/4/96 7:22 0 7340 0 4 163000 8400 53000 34 9.6 1.26 NA 0.05 0.002 0.063 0.212 <0.006
Storm 15
1 6/22/96 11:37 330 330 216 42 NA NA NA 191 58.1 5.00 2.419 0.17 0.215 0.085 2.654 <0.006
2 6/22/96 12:06 600 930 42 33 NA NA NA 164 51.7 4.60 3.629 0.26 0.053 0.135 0.908 <0.006
3 6/22/96 12:36 0 930 50 45 NA NA NA 141 55.6 1.70 6.149 0.26 0.047 0.149 0.754 <0.006
Storm 16
1 6/25/96 10:59 320 320 168 30 NA NA NA 120 33.2 5.000 3.049 0.35 0.186 0.27 5.062 <0.006
2 6/25/96 11:28 740 1060 76 24 NA NA NA 76 24.4 3.400 1.953 0.32 0.036 0.173 1.256 <0.006
3 6/25/96 11:58 370 1430 64 25 NA NA NA 71 24.6 4.800 1.738 0.15 0.06 0.118 1.405 <0.006
4 6/25/96 12:58 10 1440 60 30 NA NA NA 95 38.3 4.650 NA 0.14 0.04 0.216 1.426 <0.006
Storm 19
1 8/22/96 9:58 1000 1000 NA 47.000 NA NA NA 80 26.3 1.550 1.864 0.410 0.053 0.135 0.908 <0.006
2 8/22/96 10:27 10 1010 11.000 29.000 NA NA NA 310 53.9 2.350 3.139 0.280 0.047 0.149 0.754 <0.006
3 8/22/96 10:57 0 1010 11.000 31.000 NA NA NA 297 54.6 6.600 2.648 0.300 0.174 0.147 3.837 0.012
5 8/22/96 12:57 70 1080 29.000 38.000 NA NA NA 332 46.3 5.000 3.383 0.310 0.03 0.055 1.09 <0.006
7 8/22/96 18:08 1020 2100 28.000 52.000 12000 1600 900 327 36.1 0.440 1.233 0.590 0.011 0.073 0.784 <0.006
8 8/22/96 18:37 60 2160 11.000 19.000 7000 6800 4800 58 17.0 1.900 1.003 0.170 0.031 0.069 1.245 <0.006
9 8/22/96 19:07 40 2200 12.000 33.000 200000 5300 20000 70 19.0 1.450 1.375 0.170 0.254 0.216 5.787 <0.006
10 8/22/96 20:07 10 2210 14.000 26.000 250000 9500 5000 88 22.6 1.400 1.583 0.210 0.063 0.084 0.95 <0.006
102
11 8/22/96 21:07 0 2210 15.000 26.000 240000 9700 12000 107 30.3 1.500 1.868 0.240 0.002 0.043 0.58 <0.006
12 8/22/96 22:07 0 2210 14.000 24.000 180000 7800 4500 73 23.0 3.800 1.469 0.200 0.014 0.123 0.66 <0.006
Storm 20
1 8/23/96 17:02 880 880 80.000 24.000 NA NA NA 74 41.9 1.000 1.227 0.340 0.104 0.053 2.792 <0.006
2 8/23/96 17:30 1710 2590 11.000 14.000 NA NA NA 22 20.2 0.300 6.496 0.230 0.005 0.042 1.146 <0.006
3 8/23/96 18:00 960 3550 8.000 7.000 NA NA NA 12 7.1 0.320 0.276 0.120 0.002 0.042 0.276 <0.006
4 8/23/96 19:00 560 4110 2.000 10.000 NA NA NA 30 9.3 0.370 0.878 0.120 0.002 0.042 0.359 <0.006
5 8/23/96 20:00 250 4360 8.000 9.000 NA NA NA 20 7.1 0.200 1.146 0.110 0.002 0.042 0.175 <0.006
6 8/23/96 21:00 0 4360 5.000 9.000 NA NA NA 35 11.4 0.710 0.923 0.140 0.002 0.042 0.025 <0.006
Storm 21
1 8/29/96 12:17 780 780 16.000 24.000 NA NA NA 54 33.0 3.500 2.043 0.140 0.012 0.053 0.895 <0.006
2 8/29/96 12:46 0 780 15.000 27.000 NA NA NA 78 22.3 2.300 1.897 0.100 0.028 0.104 0.85 <0.006
3 8/29/96 13:16 0 780 15.000 3.600 NA NA NA 74 24.2 3.350 2.673 0.110 0.002 0.068 0.767 <0.006
4 8/29/96 14:16 0 780 15.000 19.000 NA NA NA 59 15.9 1.300 1.770 0.090 0.002 0.06 0.65 <0.006
5 8/29/96 15:16 2670 3450 12.000 19.000 NA NA NA 52 13.7 4.900 1.390 0.090 0.029 0.092 0.634 <0.006
6 8/29/96 16:16 220 3670 14.000 19.000 NA NA NA 48 13.8 1.400 1.150 0.080 0.018 0.072 1.974 <0.006
Storm 23
1 10/17/96
16:32
4890 4890 297 200 <2000000 <20000 NA 155 55.3 1.30 5.482 0.19 0.4 0.2 5.8 <0.05
2 10/17/96
17:01
510 5400 78 52 2200000 <20000 NA 53 10.0 1.09 1.635 0.17 0.050 0.1 1.3 <0.0
3 10/17/96
17:31
0 5400 27 27 <200000 20000 NA 65 14.3 1.20 1.809 0.14 0.050 0.1 1.0 <0.0
4 10/17/96
18:31
0 5400 32 31 <2000000 <20000 NA 36 11.0 0.88 1.336 0.11 0.050 0.050 1.0 <0.0
Storm 24
1 10/27/96 5:31 740 740 24 25 <20000 3400 NA 57 29.7 0.73 NA 0.14 0.050 0.042 0.8 <0.0
2 10/27/96 6:00 290 1030 16 25 <2000 2600 NA 99 40.5 0.93 1.531 0.13 0.050 0.042 0.3 <0.0
3 10/27/96 6:30 590 1620 21 43 <2000 5300 NA 104 36.2 0.75 2.993 0.14 0.050 0.042 0.6 <0.0
4 10/27/96 7:30 170 1790 16 22 <2000 2800 NA 94 49.2 0.93 1.960 0.14 0.050 0.042 0.042 <0.0
5 10/27/96 8:30 0 1790 17 26 <2000 2400 NA 122 49.2 1.05 1.845 0.15 0.050 0.042 0.3 <0.0
103
6 10/27/96 9:30 0 1790 14 25 <2000 2000 NA 127 38.3 0.95 NA 0.14 0.050 0.042 0.3 <0.0
Storm 25
1 11/7/96 1:22 3150 3150 120 54 0 4400 NA 115 61.7 1.90 2.370 0.34 0.036 0.079 1.061 <0.006
2 11/7/96 1:51 3570 6720 20 27 <2000 2000 NA 30 12.5 0.45 0.776 0.17 0.002 0.074 0.481 <0.006
3 11/7/96 2:21 2090 8810 3 19 0 2000 NA 8 10.0 0.22 0.952 0.09 0.002 0.088 0.165 <0.006
4 11/7/96 3:21 1800 10610 4 16 <2000 6100 NA 38 21.2 1.40 0.159 0.12 0.002 0.076 0.204 <0.006
5 11/7/96 4:21 3640 14250 13 19 0 2200 NA 33 18.3 0.59 0.217 0.15 0.009 0.042 0.628 <0.006
6 11/7/96 5:21 210 14460 16 17 0 2100 NA 14 15.4 1.00 0.460 0.12 0.002 0.045 0.340 <0.006
Storm 27
1 12/4/96 11:40 190 190 27.00 77 <2000 5100 NA 117 35.4 2.10 4.519 0.17 0.232 0.042 4.841 <0.006
2 12/4/96 12:09 1380 1570 166.00 63 <2000 5100 NA 127 37.4 2.40 3.339 0.27 0.645 0.198 12.09 0.038
3 12/4/96 12:39 2590 4160 132.00 59 <20000 2700 NA 53 33.1 0.70 2.075 0.15 0.373 0.080 6.802 0.015
4 12/4/96 13:39 2080 6240 52.00 55 <2000 3100 NA 52 27.1 0.40 1.248 0.10 0.091 0.114 3.704 <0.006
5 12/4/96 14:39 980 7220 44.00 52 <2000 3700 NA 60 27.1 0.96 1.597 0.12 0.033 0.072 1.789 <0.006
6 12/4/96 15:39 580 7800 69.00 72 <2000 12100 NA 56 33.1 1.05 1.709 0.18 0.108 0.067 4.312 <0.006
Storm 28
1 12/15/96 1:24 510 510 43.00 27 NA NA NA 117 58.5 4.10 3.418 NA 0.043 0.057 1.180 <0.006
2 12/15/96 1:53 510 1020 23.00 25 NA NA NA 121 45.9 4.10 3.094 NA 0.110 0.087 0.787 <0.006
3 12/15/96 2:23 230 1250 21.00 20 NA NA NA 140 51.4 >10 3.544 NA 0.047 0.042 0.873 <0.006
4 12/15/96 3:23 230 1480 15.00 24 NA NA NA 135 53.2 >10 2.854 NA 0.027 0.042 0.583 <0.006
5 12/15/96 4:23 4840 6320 343.00 45 NA NA NA 139 67.4 2.10 2.413 NA 0.268 0.080 4.692 <0.006
6 12/15/96 5:23 10600 16920 202.00 47 NA NA NA 19 28.0 0.53 1.309 NA 0.005 0.057 1.643 <0.006
Storm 26
1 11/24/96 3:53 1240 1240 184.00 44 0 11900 NA 86 23.4 1.52 1.464 0.21 0.100 0.124 2.214 <0.006
2 11/24/96 4:21 200 1440 12.00 34 0 5900 NA 49 15.7 1.85 1.484 0.13 0.010 0.165 0.669 <0.006
3 11/24/96 4:51 8110 9550 41.00 37 <2000 8000 NA 52 17.7 1.95 0.840 0.10 0.002 0.042 0.652 <0.006
4 11/24/96 5:51 5760 15310 33.00 23 0 2000 NA 13 6.4 0.37 1.182 0.06 0.004 0.042 0.991 <0.006
5 11/24/96 6:51 3670 18980 11.00 20 <2000 1800 NA 12 6.4 0.50 0.460 0.04 0.002 0.139 0.193 <0.006
6 11/24/96 7:51 6720 25700 90.00 23 0 2000 NA 17 8.3 0.34 0.501 0.05 0.002 0.042 0.371 <0.006
Storm 29
1 2/6/97 11:44 480 480 336 290 NA NA NA 326 91.1 8.4 7.4 0.53 0.043 0.116 0.559 0.012
104
2 2/6/97 12:13 380 860 124 150 NA NA NA 268 73.6 8 6.78 0.32 0.106 0.183 2.213 0.039
3 2/6/97 12:43 100 960 81 150 NA NA NA 287 73.6 8.4 7.15 0.32 0.14 0.157 2.429 0.05
4 2/6/97 13:43 0 960 121 160 NA NA NA 260 88.9 10 7.4 0.31 0.138 0.142 2.558 0.05
5 2/6/97 14:43 0 960 152 180 NA NA NA 235 90.3 10 7.01 0.35 0.121 0.183 2.481 0.043
6 2/6/97 15:43 0 960 117 190 NA NA NA 324 83.8 10 6.72 0.34 0.165 0.229 3.387 0.059
Storm 30
1 2/12/97 4:14 640 640 440 86 <2000 80000 NA 94 44.0 4.1 3.86 0.14 0.086 1.939 0.015
2 2/12/97 4:43 1160 1800 74 74 <2000 2000 NA 41 30.4 2.8 2.44 0.14 0.078 2.943 0.021
3 2/12/97 5:13 3760 5560 118 61 <2000 2000 NA 44 22.5 1.1 2.26 0.16 0.089 1.773 0.014
4 2/12/97 6:13 3780 9340 65 63 <2000 2000 NA 44 12.6 0.88 1.11 0.09 0.076 2.219 0.011
5 2/12/97 7:13 3720 13060 216 59 <2000 1600 NA 35 19.2 0.6 1.38 0.22 0.125 2.169 0.029
6 2/12/97 8:13 1910 14970 182 55 <2000 1400 NA 37 17.6 0.8 1.45 0.12 0.086 2.223 0.038
Storm 31
1 3/10/97 1:51 5930 5930 1124 272 NA NA NA 213 111.0 NA 2.93 0.66 0.584
2 3/10/97 2:20 4770 10700 46 29 NA NA NA 35 6.2 NA 1.1 0.22 0.002
3 3/10/97 2:50 1310 12010 8 21 NA NA NA 17 6.2 NA 0.75 0.06 0.09
4 3/10/97 3:50 700 12710 12 25 NA NA NA 33 13.4 NA 0.61 0.12 0.172
5 3/10/97 4:50 410 13120 8 17 NA NA NA 24 8.0 NA 0.88 0.05 0.072
6 3/10/97 5:50 220 13340 NA 20 NA NA NA 22 8.0 NA 1.28 0.06 0.002
Storm 32
1 3/25/97 9:40 1240 1240 680 136 NA NA NA 139 144.5 1.99 2.25 0.97 0.36
2 3/25/97 10:09 1390 2630 158 260 NA NA NA 46 70.9 1.23 3.04 0.33 0.1
3 3/25/97 10:39 470 3100 62 53 NA NA NA 91 30.8 1.37 2.75 0.25 0.13
4 3/26/97 11:39 5850 8950 318 66 NA NA NA 54 54.3 0.7 1.71 0.33 0.08
5 3/26/97 12:39 1290 10240 38 36 NA NA NA 66 20.6 0.82 1.49 0.13 0.24
6 3/26/97 13:39 1600 11840 22 34 NA NA NA 73 30.8 1.19 1.77 0.18 0.08
Storm 33
1 4/2/97 17:15 8990 8990 480 290 NA NA NA 304 122.5 2.3 5.51 0.84 0.51
2 4/2/97 17:44 1900 10890 184 170 NA NA NA 149 44.6 0.77 2.77 0.61 0.21
3 4/2/97 18:14 930 11820 40 100 NA NA NA 87 20.3 1.02 2.2 0.29 0.12
4 4/2/97 19:14 2300 14120 135 56 NA NA NA 125 31.9 0.73 2.19 0.46 0.2
105
5 4/2/97 20:14 1600 15720 26 29 NA NA NA 70 21.6 0.74 2.06 0.24 0.12
6 4/2/97 21:14 1570 17290 34 28 NA NA NA 71 18.1 0.71 1.7 0.25 0.1
Storm 34
1 4/25/97 2:48 1080 1080 77 37 15000 20500 NA 187 69.5 5 4.3 0.34 0.29
2 4/25/97 3:17 1290 2370 15 20 35000 16500 NA 73 32.9 4.35 2.76 0.16 0.17
3 4/25/97 3:47 1240 3610 13 18 26000 6600 NA 37 19.0 2.61 1.9 0.08 0.09
4 4/25/97 4:47 1240 4850 12 28 28000 25000 NA 80 31.0 2.7 2.46 0.13 0.27
5 4/25/97 5:47 1400 6250 198 150 50000 62000 NA 130 45.2 3.96 3.21 0.18 0.12
6 4/25/97 6:47 1510 7760 307 150 19000 45000 NA 162 NA 2.01 3.43 0.59 0.4
Storm 36
1 5/27/97 15:49 3690 3690 556 114 NA NA NA 150 80.2 1.17 4.11 80.2 0.23
2 5/27/97 16:19 3880 7570 114 30 NA NA NA 20 11.9 0.29 0.84 0.01 0.09
3 5/27/97 16:49 0 7570 60 30 NA NA NA 48 27.1 0.76 1.46 0.08 0.09
106
Filter strip discharge at Walnut Creek site
TSS Turbidity Fecal Col Fecal Str E. coli COD TOC Nitrate TKN Total P Zinc Lead Iron Copper
Sample Date/Time
Flow
Vol
Cum
Flow
Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc Conc
No. Collected L L mg/L NTU CFU/100ml CFU/100ml CFU/100ml mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L
Storm 5
1 12/17/95 7:25 2500 2500 147 32 40000 26500 NA 71 NA 1.110 0.650 0.35 0.080 0.085 2.353 <0.006
2 12/17/95 7:40 71100 73600 120 28 40500 22500 NA 56 NA 0.650 0.58 0.23 0.056 0.042 1.756 0.007
3 12/18/95 7:55 31800 105400 9 5.8 12500 22000 NA 35 NA 0.390 0.45 0.23 0.002 0.042 1.197 <0.006
4 12/19/95 8:25 28560 133960 2 4.5 18500 30000 NA 35 NA 0.240 0.43 0.27 0.002 0.042 0.180 <0.006
5 12/20/95 9:25 23900 157860 4 4 NA NA NA 41 NA 0.420 0.40 0.11 0.002 0.042 0.177 <0.006
Storm 6
3 2/29/96 11:57 3760 3760 28 48 180000 20000 100000 98 NA 1.780 1.10 0.03 0.083 0.042 1.107 <0.006
4 2/29/96 12:27 10900 14660 115 45 TNTC 31000 20000 69 NA 1.460 1.19 0.03 0.062 0.042 1.607 <0.006
5 2/29/96 13:27 35940 50600 53 35 TNTC 29000 210000 53 NA 1.000 0.80 0.03 0.014 0.042 0.632 <0.006
6 2/29/96 14:27 29420 80020 28 21 NA NA NA 47 NA 0.930 0.90 0.06 0.011 0.042 0.604 <0.006
Storm 15
1 6/22/96 16:56 3100 3100 42 3 NA NA NA 58 17.7 1.89 1.299 0.17 0.002 0.093 0.335 <0.006
2 6/22/96 17:10 5780 8880 22 19 NA NA NA 84 32.2 1.89 3.977 0.24 NA NA NA NA
3 6/22/96 17:40 3470 12350 24 10 NA NA NA 90 37.7 4.6 4.048 0.16 0.004 0.075 0.145 <0.006
4 6/22/96 18:10 3450 15800 14 16.000 NA NA NA 77 27.1 7.4 0.904 0.150 0.002 0.135 0.181 <0.006
5 6/22/96 19:10 1710 17510 14 20 NA NA NA 72 31.8 NA 1.934 0.13 NA NA NA NA
6 6/22/96 20:10 480 17990 NA NA NA NA NA 72 38.6 NA 3.584 0.20 NA NA NA NA
Storm 16
1 6/25/96 11:38 5320 5320 NA 5 NA NA NA 58 24.9 3.7 2.371 0.26 0.002 0.142 0.183 <0.006
2 6/25/96 11:52 6690 12010 48 4.7 NA NA NA 74 28.1 1.8 3.24 0.29 0.002 0.165 0.206 <0.006
3 6/25/96 12:22 6060 18070 36 4.7 NA NA NA 70 26.6 2.8 1.104 0.17 0.002 0.159 0.191 <0.006
4 6/25/96 12:52 4830 22900 36 5 NA NA NA 58 24.7 1.9 1.526 0.12 0.002 0.152 0.191 <0.006
Storm 18
1 8/11/96 14:52 79590 79590 80 4.7 CG 2600 430000 35 20.400 0.980 3.696 0.160 0.002 0.121 0.138 <0.006
107
2 8/11/96 15:07 52110 131700 108 9.8 CG 10400 1000000 110 34.800 1.150 3.647 0.450 0.015 0.135 0.671 <0.006
3 8/11/96 15:37 50350 182050 192 3.5 CG 5700 480000 63 16.800 0.700 1.898 0.270 0.002 0.128 0.343 <0.006
4 8/11/96 16:07 25790 207840 8 2.7 1450000 5200 280000 77 27.800 0.590 2.181 0.390 0.002 0.173 0.338 <0.006
5 8/11/96 17:07 12850 220690 12 3 1160000 6000 400000 93 29.600 0.700 2.673 0.420 0.002 0.098 0.099 <0.006
6 8/11/96 18:07 10580 231270 4 1.9 620000 2500 63000 67 39.200 0.700 1.930 0.390 0.002 0.127 0.344 <0.006
Storm 19
2 8/22/96 18:20 10500 10500 7 8.000 380000 7000 9000 97 20.800 4.600 3.039 0.250 0.002 0.126 0.38 <0.006
3 8/22/96 18:34 9720 20220 8 10.000 120000 25000 190000 60 20.600 4.600 1.577 0.170 0.002 0.098 0.188 <0.006
4 8/22/96 19:04 7700 27920 0 5.000 20000 15000 20000 52 16.600 4.200 1.393 0.190 0.002 0.11 0.544 0.068
5 8/22/96 19:34 12820 40740 6 5.000 180000 15000 24000 52 15.000 1.300 2.175 0.210 0.002 0.076 0.405 <0.006
Storm 20
2 8/23/96 17:16 5360 5360 7 3.500 NA NA NA 24 12.100 1.100 0.609 0.090 0.041 0.069 0.143 <0.006
3 8/23/96 17:30 42390 47750 6 4.700 NA NA NA 48 18.400 0.680 0.591 0.200 0.002 0.042 0.717 <0.006
4 8/23/96 18:00 48920 96670 3 5.700 NA NA NA 33 13.600 0.330 1.218 0.190 0.002 0.042 0.007 <0.006
5 8/23/96 18:30 41780 138450 3 4.500 NA NA NA 18 13.700 0.200 1.107 0.180 0.002 0.042 0.018 <0.006
6 8/23/96 19:30 47610 186060 3 3.500 NA NA NA 38 14.100 0.150 0.781 0.190 0.002 0.061 0.007 <0.006
Storm 21
1 8/29/96 12:26 6870 6870 21 13.000 NA NA NA 24 17.200 2.000 3.696 0.11 0.002 0.134 0.067 <0.006
2 8/29/96 12:40 9170 16040 14 4.500 NA NA NA 45 17.200 2.500 3.647 0.12 0.002 0.124 0.103 <0.006
3 8/29/96 13:10 5950 21990 1 21.000 NA NA NA 47 18.400 4.400 1.898 0.12 0.002 0.07 0.109 <0.006
4 8/29/96 13:40 7510 29500 6 6.300 NA NA NA 48 18.400 1.400 2.181 0.13 0.002 0.073 0.9 <0.006
5 8/29/96 14:40 8810 38310 8 7.200 NA NA NA 39 12.700 4.200 2.673 0.14 0.002 0.141 0.148 <0.006
6 8/29/96 15:40 229040 267350 2 4.000 NA NA NA 40 14.600 0.510 1.930 0.13 0.002 0.044 0.049 <0.006
Storm 27
1 12/4/96 12:59 7610 7610 5.00 3.3 2000 2000 NA 9 9.4 0.15 0.482 0.15 0.060 0.042 0.214 <0.006
2 12/4/96 13:14 9110 16720 19.00 31 22000 22700 NA 40 22.4 1.40 1.244 0.17 0.021 0.061 1.514 <0.006
3 12/4/96 13:44 9680 26400 25.00 40 48000 28000 NA 37 23.7 1.30 1.113 0.15 0.021 0.114 1.593 <0.006
4 12/4/96 14:14 9740 36140 13.00 35 2000 16000 NA 30 27.4 1.40 1.149 0.12 0.002 0.120 0.887 <0.006
5 12/4/96 15:14 9030 45170 13.00 31 4400 12400 NA 19 21.4 0.81 0.893 0.11 0.005 0.247 1.071 <0.006
6 12/4/96 16:14 9500 54670 21.00 25 2000 11700 NA 27 18.4 0.93 0.843 0.11 0.009 0.072 0.838 <0.006
Storm 28
108
1 12/15/96 4:55 10410 10410 11.00 15 NA NA NA 11 20.1 0.80 0.677 NA 0.002 0.042 0.156 <0.006
2 12/15/96 5:10 32570 42980 17.00 20 NA NA NA 64 39.0 2.20 1.595 NA 0.002 0.087 0.470 <0.006
3 12/15/96 5:40 227380 270360 6.00 12 NA NA NA 41 21.5 1.00 1.353 NA 0.002 0.042 0.294 <0.006
4 12/15/96 6:10 251380 521740 11.00 12 NA NA NA 9 13.8 0.24 0.763 NA 0.002 0.042 0.426 <0.006
5 12/15/96 7:10 154060 675800 7.00 13 NA NA NA 16 15.4 0.21 0.582 NA 0.002 0.042 0.382 <0.006
6 12/15/96 8:10 159220 835020 6.00 9 NA NA NA 4 17.5 0.17 0.650 NA 0.005 0.137 0.473 <0.006
Storm 26
2 11/24/96 5:52 18430 18430 10.00 23 570000 460000 NA 50 23.2 2.10 0.500 0.29 0.002 0.042 0.202 <0.006
3 11/24/96 6:22 33560 51990 11.00 16 2000 300000 NA 40 11.9 1.10 0.970 0.25 0.002 0.099 0.169 <0.006
4 11/24/96 6:52 37270 89260 5.00 14 250000 65000 NA 29 10.1 0.39 0.744 0.18 0.012 0.061 0.235 <0.006
5 11/24/96 7:52 101270 190530 36.00 12 2000 50000 NA 18 10.1 0.29 0.720 0.16 0.002 0.061 0.552 <0.006
6 11/24/96 8:52 95450 285980 6.00 14 210000 76000 NA 18 10.1 0.24 0.574 0.19 0.002 0.051 0.269 <0.006
Storm 29
1 2/7/97 6:21 12250 12250 30 12 NA NA NA 17 4 0.78 0.889 0.1 0.273 0.247 5.982 0.054
2 2/7/97 6:36 27410 39660 24 26 NA NA NA 73 12.8 1.9 1.927 0.16 0.053 0.11 1.197 0.014
3 2/7/97 7:06 23260 62920 139 59 NA NA NA 79 25.6 1.6 1.591 0.25 0.067 0.087 1.286 0.015
4 2/7/97 7:36 16040 78960 37 38 NA NA NA 42 16.5 1.16 1.278 0.16 0.04 0.113 1.012 0.009
5 2/7/97 8:36 30080 109040 14 28 NA NA NA 39 12.3 1.1 1.908 0.16 0.007 0.042 0.011 0.002
6 2/7/97 9:36 45570 154610 12 24 NA NA NA 41 12.5 1 1.599 0.14 0.046 0.042 0.572 0.015
Storm 30
1 2/12/97 5:36 6890 6890 16 25 <2000 13200 NA 25 19 3.5 0.411 0.08
2 2/12/97 5:50 23260 30150 18 26 <2000 13400 NA 38 17.3 3.3 1.464 0.08 0.036 0.357 0.008
3 2/12/97 6:20 26330 56480 30 26 <2000 15500 NA 36 15.8 1.85 1.139 0.11 0.107 0.488 0.024
4 2/12/97 6:50 41820 98300 11 20 <2000 190000 NA 19 12.5 1.2 1.225 0.09 0.029 0.458 0.007
5 2/12/97 7:50 85090 183390 9 16 <2000 15700 NA 16 14.2 1.15 0.678 0.14 0.032 <0.04
2
0.48 0.009
6 2/12/97 8:50 43660 227050 13 22 2600 14300 NA 19 14.2 0.73 1.081 0.08 0.055 0.168 0.6 0.019
Storm 31
1 3/10/97 2:02 29730 29730 19 27 NA NA NA 14 13.4 NA 0.879 0.08 0.05
2 3/10/97 2:17 139310 169040 80 32 NA NA NA 38 43.1 NA 1.883 0.21 0.07
3 3/10/97 2:47 60010 229050 78 30 NA NA NA 21 14.7 NA 1.121 0.09 0.08
109
4 3/10/97 3:17 35060 264110 16 15 NA NA NA 17 10.2 NA 0.741 0.09 0.03
5 3/10/97 4:17 14450 278560 6 11 NA NA NA 24 12.5 NA 0.988 0.09 0.09
6 3/10/97 5:17 7950 286510 4 10 NA NA NA 19 16.1 NA 1.23 0.11 0.09
Storm 32
1 3/25/97 11:42 5900 5900 9.5 5.3 NA NA NA 17 18 1.26 1.199 0.09 0.12
2 3/25/97 11:56 29440 35340 6 21 NA NA NA 46 24.2 1.59 1.907 0.1 0.15
3 3/25/97 12:26 28110 63450 12.5 22 NA NA NA 40 20 0.78 0.663 0.22 0.09
4 3/25/97 12:56 21420 84870 4 17 NA NA NA 31 14 0.47 0.786 0.16 0.24
5 3/25/97 13:56 11420 96290 2.5 16 NA NA NA 29 15.8 0.39 0.901 0.19 0.05
6 3/25/97 14:56 50970 147260 1 15 NA NA NA 33 17.5 0.61 0.786 0.16 0.09
Storm 34
1 4/25/97 8:41 4050 4050 34 6.5 10900 25000 NA 49 36.9 2.14 1.24 0.2 0.13
2 4/25/97 8:55 8050 12100 11 7 9300 40000 NA 42 24.5 2.07 1.482 0.14 0.28
3 4/25/97 9:25 6680 18780 20 19 19200 45000 NA 51 19.7 1.59 1.5 0.18 0.2
4 4/25/97 9:55 16660 35440 6 16 10500 52000 NA 47 19.9 1.45 1.659 0.2 0.08
5 4/25/97 10:55 88360 123800 7 13 8200 40000 NA 48 19.7 1.34 1.255 0.12 0.1
6 4/25/97 11:55 117430 241230 16.5 14 6800 26000 NA 35 17.4 0.99 1.25 0.18 0.18
110
APPENDIX C
EMCs and Final Concentration Averages
for Four Field Sites
111
St
orm
 
EM
C s
u
m
m
ary
 f
o
r roa
d
,
 1
8
3
 @ 
M
o
p
a
c
 
s
i
te 
(C)
1
9 s
t
orm
s
T
S
S
T
u
r
bidity
F
e
c
a
l
 C
o
l
F
e
c
a
l
 S
t
r
E
.
 co
li
C
O
D
T
O
C
N
i
tr
a
t
e
T
K
N
T
o
t
a
l P
Zin
c
L
e
a
d
I
r
o
n
Flo
w
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
Storm
Date
L
m
g
/L
N
T
U
C
FU
/1
00
m
l
C
F
U
/
1
0
0
m
l
C
FU
/1
00
m
l
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
St
orm
 12
5
/
27
/9
6
1
17
31
0
1
27
5
3
6
2
6
5
5
6
8
4
2
5
81
0
5
3
3
5.2
1
.21
N
A
0
.45
0
.29
4
0
.
16
8
3
.21
6
S
t
o
r
m
 1
3
5/
3
0
/
9
6
1
7
6
2
5
0
7
10
15
77
26
16
31
3
1
5
6
3
4
27
14
.
3
5.
2
1
N
A
0.
3
6
0.
0
0
2
0
.
0
79
0.
3
6
8
St
orm
 15
6
/
22
/9
6
6
80
0
2
47
4
4
N
A
N
A
N
A
4
5
8
6
8.1
3
.29
5
.92
2
0
.
60
0
.
45
9
0
.19
7
5
.
34
8
S
t
o
r
m
 1
6
6/
2
5
/
9
6
1
8
4
8
0
11
7
2
7
8
3
2
0
2
64
23
N
A
69
26
.
0
5.
6
6
1.
8
6
8
0
.
3
5
0
.
2
85
0.
2
0
0
3
.
2
86
St
orm
 19
8
/
22
/9
6
5
94
0
3
1
4
5
N
A
N
A
N
A
2
02
3
6
.5
2
.
66
2
.
99
3
0
.51
0
.27
9
0
.
16
6
4
.31
7
S
t
o
r
m
 2
0
8/
2
3
/
9
6
1
5
2
6
0
17
21
N
A
N
A
N
A
52
17
.
1
0.
8
0
1.
1
9
9
0
.
2
0
0
.
0
30
0.
0
5
2
0
.
7
23
St
orm
 21
8
/
29
/9
6
3
68
0
2
2
4
1
4
89
34
8
4
09
2
2
13
8
1
23
4
1
.8
1
.
12
0
.
37
8
0
.38
0
.14
6
0
.
07
5
2
.56
5
St
orm
 22
9
/
18
/9
6
3
23
20
1
3
5
7
8
N
A
8
36
1
N
A
1
12
3
9
.7
2
.
25
2
.
20
6
0
.39
0
.12
3
0
.
09
5
2
.76
0
S
t
o
r
m
 2
3
10
/
1
7
/
9
6
18
40
0
6
4
3
9
N
A
N
A
N
A
1
1
9
45
.
5
1.
1
5
3.
0
2
9
1
.
0
7
1
.
0
40
0.
2
8
5
9
.
7
50
S
t
o
r
m
 2
4
10
/
2
7
/
9
6
63
20
31
2
1
0
5
17
81
1
5
3
3
9
9
N
A
N
A
59
.
5
0.
4
7
5.
5
6
9
2
.
0
1
1
.
0
99
0.
2
7
1
7
.
9
96
S
t
o
r
m
 2
5
11
/
7
/
9
6
3
0
6
0
0
81
22
23
55
42
44
N
A
11
15
.
4
0.
5
3
0.
6
1
6
0
.
1
6
0
.
1
26
0.
0
4
2
1
.
5
16
St
orm
 26
1
1
/2
4/
96
2
5
6
6
0
4
0
2
4
2
00
0
5
13
7
N
A
3
8
5
.5
0
.
41
0
.
59
0
0
.43
0
.02
2
0
.
08
2
0
.37
0
S
t
o
r
m
 2
8
12
/
1
5
/
9
6
55
18
0
9
8
2
6
N
A
N
A
N
A
2
0
1
8
.
7
0
.
5
5
0
.
8
85
N
A
0.
0
9
3
0
.
0
88
1.
8
3
0
St
orm
 30
2
/
12
/9
7
1
01
10
1
3
3
6
6
N
A
N
A
N
A
6
7
2
5.0
0
.46
1
.34
6
0
.
27
0
.
23
N
A
2
.
55
8
St
orm
 32
3
/
25
/9
7
2
04
90
3
2
8
1
05
N
A
N
A
N
A
8
1
3
5
.5
0
.
43
2
.
05
5
0
.58
0
.44
N
A
N
A
St
orm
 33
4
/
2/
97
1
2
3
6
0
5
22
2
0
6
1
83
63
3
7
0
3
5
N
A
1
22
5
1
.9
1
.
63
0
.
30
8
0
.69
0
.69
N
A
N
A
St
orm
 34
4
/
25
/9
7
3
83
0
1
46
6
0
N
A
N
A
N
A
3
8
4
0
.3
2
.
47
2
.
45
5
0
.56
0
.35
N
A
N
A
S
t
o
r
m
 3
5
5/
9
/
9
7
15
24
00
38
9
3
6
N
A
8
3
0
6
N
A
1
9
N
A
0
.
9
1
3
.
2
49
0.
6
8
0.
4
8
N
A
N
A
St
orm
 36
5
/
27
/9
7
8
73
89
1
5
9
4
8
8
66
04
5
7
0
0
7
N
A
8
6
3
4.3
0
.94
2
.17
6
0
.
30
0
.
41
N
A
N
A
Str
a
ight A
v
e
r
a
g
e
15
7
5
5
9
5
9
6
4
23
37
8
7
8
6
1
9
4
3
3
.
9
0
.
9
1
2
.
1
67
0.
5
5
0.
3
4
7
0
.
1
38
3.
3
2
9
Coe
f
f 
of Va
ria
n
ce
0.
9
0
0.
8
1
1.
6
3
0.
9
8
0.
8
9
1.
0
9
0.
4
9
0.
9
3
0.
7
6
0.
7
6
0.
9
0
0.
5
9
0.
8
3
112
S
t
o
r
m
 
E
M
C
 sum
m
ar
y
 for
 sw
al
e,
 18
3
 
@
 Mo
pa
c si
t
e
 
(
D
)
2
3 st
o
r
m
s
TSS
Tur
bi
di
ty
Fe
ca
l
 Col
F
e
c
a
l
 St
r
E
.
 c
o
l
i
COD
T
OC
Ni
t
r
a
t
e
T
KN
Tota
l
 P
Z
i
n
c
L
e
a
d
I
r
o
n
Fl
ow
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
St
o
r
m
D
at
e
L
m
g
/
L
N
T
U
C
FU
/
1
00m
l
C
FU
/
1
00m
l
C
FU
/
1
00m
l
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
St
o
r
m
 9
3/
27/
96
60
39
0
4
21
N
A
N
A
N
A
42
23
.
3
0.
65
0.
262
0.
20
0.
011
0.
042
0.
294
St
o
r
m
 10
4/
5/
9
6
74
35
0
1
8
N
A
N
A
N
A
N
A
1
0
1
4
.
3
N
A
N
A
0
.
1
9
0
.
003
0.
042
0.
330
St
o
r
m
 11
4/
22/
96
42
12
0
5
7
N
A
N
A
N
A
6
4
2
3
.
4
0
.
8
0
N
A
0
.
3
3
0
.
002
0.
096
0.
404
St
o
r
m
 12
5/
27/
96
14
88
80
56
26
38
02
61
40
45
40
04
95
76
35
.
2
0.
54
N
A
0.
95
0.
022
0.
119
1.
359
St
o
r
m
 13
5/
30/
96
25
48
30
56
39
14
80
4
7
8
2
4
1
2
0
1
5
49
17
.
7
4.
61
N
A
0.
28
0.
115
0.
098
1.
469
St
o
r
m
 15
6/
22/
96
62
33
0
3
8
2
4
N
A
N
A
N
A
4
5
2
1
.
4
2
.
7
1
1
.
966
0.
46
0.
002
0.
119
1.
036
St
o
r
m
 16
6/
25/
96
11
59
50
50
10
37
30
30
24
17
3
2
4
6
5
1
27
18
.
5
3.
71
1.
728
0.
24
0.
003
0.
146
0.
518
St
o
r
m
 18
8/
11/
96
17
04
00
58
19
15
11
616
16
65
5
4
5
2
0
5
6
6
4
1
5
.
7
0
.
6
4
7
.
066
0.
42
0.
002
0.
153
0.
702
St
o
r
m
 19
8/
22/
96
35
41
0
3
8
N
A
N
A
N
A
6
8
2
0
.
0
0
.
3
1
1
.
832
0.
35
0.
002
0.
102
0.
212
St
o
r
m
 20
8/
23/
96
14
94
70
5
8
N
A
N
A
N
A
48
15
.
4
0.
20
1.
176
0.
21
0.
002
0.
058
0.
240
St
o
r
m
 21
8/
29/
96
15
16
90
59
34
14
51
82
71
12
1
1
7
1
2
3
32
11
.
1
1.
40
1.
003
0.
32
0.
002
0.
061
2.
088
St
o
r
m
 22
9/
18/
96
10
30
90
7
8
N
A
N
A
N
A
24
8.
1
1
.
3
2
0
.
901
0.
32
0.
002
0.
059
0.
471
St
o
r
m
 25
11
/
7
/
9
6
2
2
2
3
9
0
1
4
1
1
N
A
N
A
N
A
2
9
1
9
.
3
0
.
2
0
0
.
775
0.
43
0.
025
0.
051
0.
312
St
o
r
m
 28
12
/
1
5
/
96
47
53
40
7
1
1
N
A
N
A
N
A
1
0
1
4
.
4
0
.
2
5
0
.
630
N
A
0.
022
0.
045
0.
322
St
o
r
m
 26
11
/
2
4
/
96
29
49
80
6
1
9
N
A
N
A
N
A
2
4
6
.
5
0.
19
0.
874
0.
21
0.
026
0.
045
0.
539
St
o
r
m
 29
2/
7/
9
7
16
14
20
7
2
4
2
7
9
8
8
3
6
7
6
N
A
41
17
.
4
1.
12
1.
299
0.
27
0.
044
0.
077
0.
690
St
o
r
m
 30
2/
12/
97
25
79
40
5
2
2
N
A
N
A
N
A
1
6
1
4
.
4
0
.
7
8
1
.
001
0.
15
0.
027
N
A
0.
726
St
o
r
m
 31
3/
11/
97
12
20
90
17
28
N
A
N
A
N
A
35
16
.
6
0.
17
1.
206
0.
25
0.
11
N
A
N
A
St
o
r
m
 32
3/
25/
97
17
58
60
6
1
7
1
8
2
3
2
7
8
0
9
N
A
30
16
.
6
0.
41
1.
326
N
A
0.
07
N
A
N
A
St
o
r
m
 33
4/
2/
9
7
65
38
0
6
17
38
59
6
1
0
1
6
2
9
N
A
4
0
1
7
.
5
0
.
6
8
1
.
066
0.
28
0.
05
N
A
N
A
St
o
r
m
 34
4/
25/
97
42
65
70
4
4
N
A
N
A
N
A
16
9.
6
0
.
3
2
1
.
039
0.
18
0.
07
N
A
N
A
St
o
r
m
 35
5/
9/
9
7
46
20
50
21
6
N
A
2
6
4
7
8
N
A
43
N
A
0.
42
1.
397
N
A
0.
05
N
A
N
A
St
o
r
m
 36
5/
27/
97
30
13
30
19
11
19
07
7
4
0
4
0
0
N
A
24
11
.
4
0.
48
1.
122
0.
19
0.
07
N
A
N
A
St
rai
g
h
t
 A
v
erag
e
21
17
27
63
54
40
38
1
1
8
1
2
6
8
3
7
1
6
.
7
0
.
4
6
1
.
456
0.
31
0.
032
0.
082
0.
689
C
o
e
f
f
 o
f
 V
a
ri
an
ce
1.
009
0.
555
1.
766
0.
831
1.
238
0.
495
0.
365
1.
187
0.
974
0.
558
1.
086
0.
459
0.
753
113
S
t
or
m
 E
M
C s
u
m
m
a
r
y
 f
o
r
 r
oad,
 
W
a
lnut
 
Cr
eek
 
s
i
t
e
 
(
s
ite W
CR)
24 s
t
o
r
m
s
T
S
S
T
u
r
b
i
d
i
t
y
F
ecal Co
l
F
ecal S
t
r
E
.
 col
i
CO
D
T
O
C
Ni
t
rat
e
T
K
N
T
o
t
al
 
P
Z
in
c
L
ead
I
r
on
Fl
ow
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
E
M
C
EM
C
Sto
r
m
Da
te
L
m
g
/
L
NT
U
C
F
U
/1
00m
l
C
F
U
/1
00m
l
C
F
U
/1
00m
l
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
m
g/
L
Sto
r
m
 6
2/
29/96
5380
257
100
NA
NA
NA
107
NA
0.
67
1.
224
0.
33
0.
178
0.
054
3.
960
Sto
r
m
 8
3/
26/96
90
479
32
NA
NA
NA
298
118.
6
5
.
3
4
4
.
360
0.
39
0.
182
0.
042
1.
760
S
t
o
rm
 10
4/
5/96
6090
432
NA
NA
NA
NA
81
34.
2
N
A
N
A
0
.
0
2
0
.
139
0.
042
2.
181
S
t
o
rm
 11
4/
28/96
1690
383
62
NA
NA
NA
127
38.
1
4
.
2
7
N
A
0
.
2
6
0
.
125
0.
115
2.
033
S
t
o
rm
 12
5/
27/96
7770
111
59
NA
NA
NA
59
29.
5
0
.
7
8
N
A
0
.
1
5
0
.
052
0.
123
1.
878
S
t
o
rm
 14
6/
4/96
7340
23
10
9889
3295
17211
38
10.
0
0
.
7
2
N
A
0
.
1
1
0
.
043
0.
086
1.
236
S
t
o
rm
 15
6/
22/96
930
104
36
NA
NA
NA
174
54.
0
4
.
7
4
3
.
200
0.
23
0.
110
0.
117
1.
528
S
t
o
rm
 16
6/
25/96
1440
93
26
NA
NA
NA
85
26.
5
4
.
1
2
2
.
143
0.
28
0.
076
0.
181
2.
141
S
t
o
rm
 19
8/
22/96
2210
26
48
20496
2077
1819
202
31.
2
1
.
1
6
1
.
593
0.
48
0.
036
0.
103
0.
953
S
t
o
rm
 20
8/
23/96
4360
23
14
NA
NA
NA
31
19.
5
0
.
4
5
3
.
035
0.
21
0.
024
0.
044
1.
130
S
t
o
rm
 21
8/
29/96
3670
13
20
NA
NA
NA
52
17.
8
4
.
3
9
1
.
514
0.
10
0.
025
0.
083
0.
770
S
t
o
rm
 23
10/
17/96
5400
276
186
NA
NA
NA
145
51.
0
1
.
2
8
5
.
119
0.
19
0.
367
0.
191
5.
375
S
t
o
rm
 24
10/
27/96
1790
21
31
NA
3840
NA
83
35.
4
0
.
7
9
2
.
422
0.
14
0.
050
0.
042
0.
581
S
t
o
rm
 25
11/
7/96
14460
36
28
NA
3085
NA
47
25.
4
0
.
8
9
0
.
927
0.
18
0.
011
0.
069
0.
562
S
t
o
rm
 27
12/
4/96
7800
98
59
NA
4114
NA
68
31.
6
1
.
0
1
2
.
050
0.
16
0.
280
0.
107
6.
050
S
t
o
rm
 28
12/
15/96
16920
227
44
NA
NA
NA
63
41.
4
1
.
2
1
1
.
794
NA
0.
085
0.
064
2.
451
S
t
o
rm
 26
11/
24/96
25700
54
28
NA
4373
NA
30
11.
4
0
.
9
5
0
.
809
0.
07
0.
007
0.
061
0.
664
S
t
o
rm
 29
2/
6/97
960
226
220
NA
NA
NA
299
82.
4
N
A
7
.
129
0.
43
0.
078
0.
147
1.
409
S
t
o
rm
 30
2/
12/97
14970
147
62
NA
5159
NA
43
20.
1
1
.
1
4
1
.
729
0.
15
0.
093
0.
093
2.
139
S
t
o
rm
 31
3/
10/97
13340
526
136
NA
NA
NA
112
53.
2
N
A
1
.
849
0.
39
0.
280
NA
NA
S
t
o
rm
 32
3/
25/97
11840
256
88
NA
NA
NA
67
57.
9
1
.
0
0
1
.
948
0.
35
0.
131
NA
NA
S
t
o
rm
 33
4/
2/97
17290
295
188
NA
NA
NA
209
77.
6
1
.
5
7
3
.
925
0.
63
0.
341
NA
NA
S
t
o
rm
 34
4/
25/97
7760
113
72
29253
30588
NA
112
38.
8
3
.
3
7
3
.
002
0.
25
0.
226
NA
NA
S
t
o
rm
 36
5/
27/97
7570
329
71
NA
NA
NA
83
45.
2
0
.
7
2
2
.
434
0.
01
0.
158
NA
NA
S
t
raig
h
t
 A
v
erag
e
190
70
19879
7066
9515
109
41.
3
1
.
2
7
2
.
610
0.
24
0.
129
0.
093
2.
042
Co
ef
f
 
of
 V
arian
ce
0.
83
0.
83
0.
49
1.
35
1.
14
0.
71
0.
61
0.
85
0.
59
0.
64
0.
81
0.
49
0.
75
114
St
orm
 EMC s
u
m
m
ary
 
for s
w
a
l
e
,
 W
a
l
n
u
t
 Creek
 s
i
t
e
 (
s
i
t
e
 
W
C
S)
22
 
st
orm
s
TSS
Turb
i
d
i
ty
Fe
cal
 
Col
F
e
cal
 
Str
E.
 
col
i
COD
TOC
Ni
trate
T
KN
Total
 P
Z
i
n
c
L
e
a
d
I
ro
n
Fl
ow
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
EMC
Sto
r
m
Da
t
e
L
m
g
/L
NTU
C
F
U
/1
00m
l
C
F
U
/1
00m
l
C
F
U
/1
00m
l
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
m
g
/L
WC
-
3
4/
29
/9
4
6
6
284
62
NA
NA
NA
NA
35
NA
0.
49
NA
0.
22
0.
04
1
N
A
0
.
2
4
9
WC
-
5
4/
30
/9
4
3
7
536
15
NA
NA
NA
NA
40
NA
0.
45
NA
0.
12
0.
02
4
N
A
0
.
1
3
8
WC
-
6
5/
2/
94
41
068
19
NA
NA
NA
NA
47
NA
0.
28
NA
0.
10
0.
02
0
N
A
0
.
4
7
4
WC
-
9
5/
28
/9
4
3
1
004
10
NA
NA
NA
NA
30
NA
0.
87
NA
0.
10
NA
NA
NA
WC
-
2
0
10
/1
8/
94
15
486
1
5
5
N
A
1
1
600
0
8
0
000
NA
38
NA
0.
20
NA
0.
09
0.
03
1
0
.
0
0
7
1.
08
7
W
C
-33
5/
8/
95
19
733
0
2
0
N
A
N
A
N
A
N
A
2
1
N
A
N
A
N
A
N
A
N
A
N
A
N
A
Sto
r
m
 5
12
/2
0/
95
15
786
0
5
9
1
6
2
9
154
24
055
NA
46
NA
0.
50
0.
50
1
0
.
2
2
0
.
0
2
8
0.
04
3
1
.
1
2
9
Sto
r
m
 6
2/
29
/9
6
8
0
020
51
32
NA
28
762
NA
55
NA
1.
07
0.
90
4
0
.
0
4
0
.
0
2
3
0.
04
2
0
.
7
7
7
Sto
r
m
 1
5
6/
4/
96
17
990
24
14
NA
NA
NA
63
29
.9
3.
69
2.
73
5
0
.
1
8
0
.
0
0
3
0.
10
1
0
.
2
1
6
Sto
r
m
 1
6
6/
7/
96
22
900
41
5
N
A
N
A
N
A
6
6
2
6
.
2
2
.
5
3
2
.
1
1
1
0.
22
0.
00
2
0
.
1
5
5
0.
19
4
Sto
r
m
 1
8
8/
11
/9
6
2
3
127
0
9
5
5
11
958
78
55
07
53
413
5
6
7
2
5
.
1
0
.
8
9
2
.
9
8
7
0.
30
0.
00
5
0
.
1
3
0
0.
33
2
Sto
r
m
 1
9
8/
22
/9
6
4
0
740
6
7
18
699
1
1
5
324
58
983
66
18
.1
3.
49
2.
10
7
0
.
2
1
0
.
0
0
2
0.
10
1
0
.
3
7
3
Sto
r
m
 2
0
8/
23
/9
6
1
8
606
0
4
5
N
A
N
A
N
A
3
4
1
4
.
8
0
.
3
6
0
.
9
2
1
0.
19
0.
00
3
0
.
0
4
8
0.
17
5
Sto
r
m
 2
1
8/
29
/9
6
2
6
735
0
3
5
N
A
N
A
N
A
4
0
1
4
.
9
0
.
8
5
2
.
0
6
5
0.
13
0.
00
2
0
.
0
5
4
0.
08
0
Sto
r
m
 2
7
12
/4
/9
6
5
4
670
16
28
13
874
15
951
NA
28
20
.9
1.
03
0.
97
0
0
.
1
3
0
.
0
1
8
0.
11
1
1
.
0
4
5
Sto
r
m
 2
8
12
/1
5/
96
83
502
0
8
12
NA
NA
NA
20
18
.0
0.
51
0.
90
0
N
A
0
.
0
0
3
0.
06
2
0
.
3
8
9
Sto
r
m
 2
6
11
/2
4/
96
28
598
0
1
7
1
4
1
4
034
8
1
1
639
3
N
A
2
4
1
1
.
2
0
.
5
0
0
.
6
9
0
0.
19
0.
00
3
0
.
0
6
1
0.
34
9
Sto
r
m
 2
9
2/
7/
97
15
461
0
3
8
3
1
N
A
N
A
N
A
5
0
1
4
.
2
1
.
2
7
1
.
6
2
7
0.
16
0.
06
0
0
.
0
8
4
1.
15
5
Sto
r
m
 3
0
2/
12
/9
7
2
2
705
0
1
4
2
0
N
A
4
7
200
NA
22
14
.5
1.
45
0.
98
2
0
.
1
1
0
.
0
4
4
NA
0.
47
3
Sto
r
m
 3
1
3/
10
/9
7
2
8
651
0
6
0
2
7
N
A
N
A
N
A
2
8
2
7
.
8
N
A
1
.
4
1
6
0.
15
0.
06
7
N
A
N
A
Sto
r
m
 3
2
3/
25
/9
7
1
4
726
0
5
18
NA
NA
NA
36
18
.7
0.
83
1.
01
2
0
.
1
6
0
.
1
2
2
NA
NA
Sto
r
m
 3
4
4/
25
/9
7
2
4
123
0
1
3
1
4
8
0
6
4
3
3
900
NA
42
19
.0
1.
22
1.
29
5
0
.
1
6
0
.
1
4
7
NA
NA
Stra
i
g
h
t A
vera
ge
29
16
24
147
3
4
0
788
29
655
9
4
1
1
9
.
5
0
.
9
7
1
.
4
5
1
0.
16
0.
03
2
0
.
0
7
7
0.
50
8
Co
eff
 of Var
i
a
n
ce
0.
87
0.
61
1.
77
0.
88
1.
13
0.
37
0.
29
0.
88
0.
51
0.
38
1.
25
0.
54
0.
74
115