Final Report
Urban Stormwater Fine Sediment Filtration Using Granular Perlite
(P037)
Principal Investigator: Russell Wigart, County of El Dorado
December 2011
Table of Contents
Abstract........................................................................................... 2
Introduction.................................................................................... 3
Background .................................................................................... 3
Goals, Objectives, and Hypotheses .............................................. 4
Approach, Methodology and Location of Research................... 4
Issues and Constraints................................................................... 8
Data Collection ............................................................................. 10
Meteorology .................................................................................................................. 10
Flow and Volume .......................................................................................................... 11
Water Quality................................................................................................................ 12
Concentration Based Load ....................................................................................... 12
Modeled Load ........................................................................................................... 15
Project Findings ........................................................................... 17
Cost Analysis................................................................................................................. 17
TMDL Crediting............................................................................................................ 18
Turbidity and TSS ......................................................................................................... 19
Conclusion .................................................................................... 20
Hypotheses Validation .................................................................................................. 21
References ..................................................................................... 22
Appendix A ................................................................................... 23
Appendix B ................................................................................... 25
Abstract
The infiltration of stormwater is not always practical as a treatment Best Management
Practice (BMP) in the Lake Tahoe Basin, so an effective and economical treatment
alternative is needed. The operation of manned sedimentation or filtration systems for
stormwater is not currently technically or economically feasible. Granular Perlite has
been tested and utilized as an unmanned stormwater filter media, however, the benefits
and costs relative to improving Lake Tahoe clarity are not well understood. In order to
further the understanding of the benefits and costs of a perlite filter media as a treatment
BMP in the Lake Tahoe Basin, the County of El Dorado conducted full scale testing of
this BMP in an existing urban stormwater system that currently discharges directly to
Trout Creek. This paper presents data collected from field tests of perlite filtration from a
drop inlet in an urban setting. Actual event based data collected for an entire Water Year
indicates that perlite has potential as a water quality BMP. The measured data suggests
that the concentration of <16 micron particles can be reduced by as much as 75% on an
event basis with an annual load reduction for this study indicating a 43% to 34%
reduction respectively for Total Suspended Solids (TSS) and <16 micron TSS. Although
the perlite filter data suggests it to be effective at fine particle removal, the study did
unveil some operational concerns relative to maintenance and flooding that depending on
the site may render this type of filtration not practical for an urban subdivision setting in
the Lake Tahoe Basin.
2
Introduction
Disturbance within the Lake Tahoe Basin increased rapidly in the late 1950s’ including
an extensive road network located mainly in the lower reaches of the watershed (Reuter,
2003). The most dominant pollutant of concern for Lake Tahoe clarity is sediment
particles less than 16-microns (Swift, 2005). Roberts (2007) estimated that 72% of the
less than 20-micron sediment load to Lake Tahoe originated from the urban upland
source category establishing the importance of treating urban stormwater. In order to
improve the clarity of Lake Tahoe, the cost effective removal of fine sediment from
urban runoff is a fundamental problem since the benefits of erosion control measures that
target total suspended sediment may be less than anticipated as earlier hypothesized by
Jassby (1999).
A highly effective and economical means of reducing the transport of fine sediment from
the urban environment to Lake Tahoe involves infiltration of runoff into subsurface soils.
However, the infiltration of stormwater is not always practical due to surface water
proximity, high groundwater, or the absence of available land for the construction of
infiltration systems at stormwater outlets. In order to meet the Total Maximum Daily
Load (TMDL) goals as identified by Roberts (2007), an effective and economical
treatment alternative to infiltration must be identified.
Background
Sedimentation of fine sediment by gravity or filtration of stormwater are alternatives to
infiltration for the removal of sediment from urban runoff. However, in order to achieve
high removal efficiencies of fine sediment from highway runoff using sedimentation,
some type of particle coagulation and flocculation methods are required (Kang, 2007).
Furthermore, the California Department of Transportation investigated the feasibility of
treating stormwater in the Lake Tahoe Basin by sedimentation and fine sand filtration.
They found that without chemical dosing, none of the sedimentation and fine sand
filtration systems tested consistently met surface water discharge standards that included
a maximum turbidity of 20 NTUs (Cal Trans, 2003). Treatment of stormwater using
manned stoichiometric particle destabilization systems is difficult and undesirable (Kang,
2007) and for these reasons a more effective filtration alternative for treatment is required
for locations where infiltration is not feasible.
Granular perlite media is proposed as a filtration media because of its use in proprietary
treatment systems for the removal of fine sediment from stormwater. Aqua Filter
stormwater filter has demonstrated between 65% and 80% reduction in total suspended
solids (TSS) including sediment less than 20-micron (Mailloux, 2006). In 2008, the
County of El Dorado Department of Transportation (DOT) conducted a series of smallscale filtration tests to evaluate the effectiveness of perlite for reducing turbidity in
simulated stormwater. The effectiveness was measured to range between a 40% and 90%
reduction in turbidity. These small-scale tests, existing Aqua Filter literature, the inert
composition of perlite, and the low cost of the media have demonstrated the potential for
perlite to reduce fine sediment, turbidity and TSS in stormwater.
3
For this study a granular perlite filter media system was installed in existing drainage
inlets within a functioning stormwater collection system that discharges directly to Trout
Creek. The filter was maintained, operated, and evaluated for a period of 12 consecutive
months.
Goals, Objectives, and Hypotheses
The goal of this research is to advance understanding in the Lake Tahoe Basin relative to
media filtration options available for the treatment of fine sediment in urban runoff by
evaluating perlite filtration media. This research quantified the water quality benefits,
fine sediment treatment, measured load reduction and operational concerns relative to
treatment. The hypothesis of this research is that filtering urban stormwater with granular
perlite will be:
•
effective in significantly reducing fine sediment from urban runoff
•
practical for installation and operation in a drainage inlet within an extensive
storm drain system
•
economic relative to other filtration alternatives for the treatment of fine sediment
Approach, Methodology and Location of Research
This research included the direct installation, operation, and maintenance of three perlite
granular filters in three different existing drainage inlets within an existing stormwater
system. The filters were operated and maintained for a period of 12 consecutive months.
This approach allowed for the measurement and evaluation of the filter during a complete
hydrological season based on average stormwater conditions. The research was
conducted in the Montgomery Estates subdivision which is bounded within the Cold
Creek and Trout Creek watersheds (Figure 1). Montgomery Estates was selected due to
the presence of an existing stormwater conveyance system, lack of treatment of
stormwater in the subdivision, stormwater outfall connectivity directly to Trout Creek,
location within half of a mile to the Sierra House Elementary School meteorological
station (Sierra House), and the proximity to the residences of the principal investigator.
The typical spacing between drainage inlets in the Montgomery Estates subdivision is
approximately 300 feet. With drainage inlets on each side of the 24-foot road, the
impervious surface contributing flow for each drainage inlet is approximately 3,600 ft2.
Three existing concrete drainage inlets were modified using technology developed by
Aqua Shield called the Aqua-Guardian (Detail 1).
4
Figure 1 Montgomery Estates Project Location
Unfiltered stormwater samples were collected manually as flow entered the drainage inlet
and filtered stormwater was collected at the filter outlet. Precipitation was collected
using measurements from the existing Sierra House meteorological station maintained by
the County of El Dorado. The turbidity of the unfiltered and filtered stormwater was
measured in the field using a Hach 2100 portable turbidimeter. Water quality samples
were collected and analyzed for TSS and Particle Size.
The three perlite filtration systems were installed in the Montgomery Estates subdivision
to evaluate the efficiency of fine particle removal and operational effectiveness (Figure
5). The perlite systems were custom fabricated to be installed in an 18” x 24” drop inlet
(Figure 2).
The design of this filter system incorporates existing technology developed by Aqua
Shield called the Aqua-Guardian (Detail 1).
5
Detail 1
Utilizing a proprietary device made sense from both the fabrication and design
perspective as it limited the unknown variables which would be inherent in a totally new
design. In this regard, if proven effective and efficient at fine particle treatment, the
retrofit of many drop inlets can be made at several locations without having individual
agencies researching, fabricating and testing varying systems.
The completed system can be placed into the inlet and is maintained by removing and
checking the perlite pillow sacks located at the bottom of the system. Several high flow
bypasses are incorporated into the system; however clogging still occurred due to pine
needle and organic material accumulation within the Perlite system.
6
Figure 2
7
Issues and Constraints
It was found that the filter system is susceptible to clogging at both the drop inlet and
within the filter mechanism. Debris and detritus are transported during events, clogging
the system, thereby not allowing water to free flow through the media or the drop inlet
(Figure 3). This was evident during almost all of the storm events analyzed for this
study. For the purpose of this study the filter was cleaned prior to predicted precipitation
events to ensure that sampling could be done. Though this step was taken, it did not
solve the problem of the filter clogging prior to sample collection along with the resultant
flooding. As a result the associated sampling interval for the perlite system was not at as
high a resolution as intended.
Figure 3
Though the system was effectively monitored, the preliminary results have indicated that
the filter and DI plug fairly easily with debris and material transported from the road.
This is discouraging in that it creates both flooding and liability issues. The liability
issues alone may render drop inlet filtration in an urban setting infeasible and impractical.
Site (P2) in the project area had to be decommissioned due to flooding that occurred after
a single event (Figure 4). The water from a rain storm filled the drop inlet with detritus,
debris and sediment thereby impeding the overflow bypass of the perlite system and
causing the road to flood. Emergency measures were then taken to clear the drop inlet.
8
Figure 4
9
Data Collection
Meteorology
The Sierra House Weather Station was maintained as part of this study with data
collected on a 15 minute interval for Temperature, Humidity, Dew Point, Wind Speed,
Wind Direction, Rain and Barometric Pressure. The weather station contained a heated
rain gauge to record snow water equivalents. A Summary of Precipitation is included in
Table 1 (below).
Table 1
WY 2011
Total
Precipitation
Precitation as
Snow
(<32degree)
Precitation as
Rain
(>32degree)
Peak rain
Intensity (in/hr)
34.7
14.6
20.1
0.92
All data from the weather station is available from the DOT upon request.
10
Flow and Volume
Load estimates were calculated based on both measured flow data and utilizing rain
intensity data from within the project area. The DOT used survey data from within the
watershed for determining the impervious area tributary to the drop inlets under
investigation. Utilizing the precipitation intensity data and the impervious surface area,
the corresponding flow was calculated for rain events using the following equation.
Q=I×A
Q = Flow (cfs)
I = Rainfall Intensity (in/hr)
A = Contributing Impervious Area (acres)
Q (cfs ) = I (in
hr
) × hr
sec
2
× A(acre) × ft
acre
× ft
in
Storm generated volumes were calculated by multiplying the discharge times the interval
of data collection. For example:
Volume( ft 3 ) = Q ( ft
3
) × Interval ( s × m)
s
m
Or
Volume = Q × ∆t
Table 2 below is the storm generated volume for WY 2011 for all sites in the study area.
Table 2
Site
P1
P2
P3
11
Annual Runoff
Volume (cf)
29,980
77,450
14,990
Water Quality
The data below is reported for all sites within the study area for the duration of the study.
Site P1 is located on Plateau Circle (West), site P2 is Plateau Circle (East) and P3 is
located on Lupine Trail all within the Montgomery Estates Subdivision in El Dorado
County (Figure 4 above). The analysis indicates a potential for fairly good removal
efficiencies for all systems during the events monitored, assuming no maintenance issues.
Load Estimates
Concentration Based Load
The concentration of pollutants was measured with the reduction of concentration per
event and Water Year calculated based on measured results. The concentration based
effectiveness was calculated regardless of whether the filter was plugged or in bypass
mode. Nearly every time County of El Dorado staff went to measure the filter it was
plugged and not operating properly. As a result, the filter was cleaned during events and
allowed to run for 15 minutes before a sample was collected for water quality
measurements. The following are the concentration based results for the monitoring and
the reductions associated with the perlite media filtration. The filtration efficiency
ranged from -19% to 86% effective for TSS and -14% to 75% effective for <16 micron
TSS. Depending on the event size, rain intensity, and the number of antecedent dry
days, the effectiveness of the filter varied. For Water Year 2011 the overall concentration
efficiency of the filters was 43% and 34% respectively for TSS and <16 micron TSS
(Table 3). Table 4 contains the concentration based efficiency results for individual
events.
12
Table 3
Date
Time
Site
10/04/10
1740
10/04/10
1740
10/05/10
950
10/05/10
950
10/05/10
1035
10/05/10
1035
10/05/10
1120
10/05/10
1120
10/05/10
1720
10/05/10
1720
10/05/10
1905
10/05/10
1905
NTU
TSS
P1-1
P1-2
P2-1
P2-2
P3-1
P3-2
P2-2-1
P2-2-2
P3-2-1
P3-2-2
P3-3-1
P3-3-2
47.2
39.7
48.3
31.1
12.9
11.0
421.0
90.8
21.2
20.7
9.2
10.8
590.0
83.0
27.0
23.0
26.0
11.0
160.0
59.0
18.0
15.0
6.0
18.0
mg / l
<100 um
<63 um
<20
<16
323.6
236.5
98.1
63.6
51.5
24.4
23.1
21.2
14.4
20.9
18.6
10.8
22.2
18.6
10.3
10.0
8.8
5.0
149.0
139.7
102.0
57.1
55.2
45.3
16.8
15.4
9.4
13.0
11.8
7.4
5.7
5.3
3.6
16.5
14.1
7.5
<10
<1
83.2
20.8
12.8
9.4
9.0
4.3
92.4
42.4
8.1
6.5
3.2
6.4
55.8
14.4
9.6
6.7
6.5
3.0
71.0
34.9
5.7
4.6
2.4
4.5
3.3
0.7
1.1
0.5
0.4
0.2
7.5
5.9
0.2
0.1
0.1
0.2
10/24/10
10/24/10
10/24/10
10/24/10
1430
1430
1730
1730
P3-1-1
P3-1-2
P3-2-1
P3-2-2
6.9
8.0
5.9
4.9
11.0
9.0
25.0
19.0
8.5
8.7
15.3
13.5
6.7
7.5
11.5
10.1
3.5
4.1
5.5
4.6
3.1
3.7
4.8
3.8
2.1
2.5
3.3
2.7
0.1
0.1
0.1
0.1
12/02/10
12/02/10
12/02/10
12/02/10
1600
1600
1620
1620
P3 in
P3 out
P3 in-2
P3 out-2
47.0
20.7
73.6
34.9
110.0
35.0
130.0
56.0
81.5
32.2
114.7
56.0
65.0
29.5
100.3
56.0
32.2
17.3
58.3
56.0
27.0
14.4
50.0
56.0
17.7
9.2
33.5
4.8
0.8
0.4
2.2
1.2
12/07/10
12/07/10
1450
1450
P3 in-1
P3 out-1
6.0
4.0
15.0
3.0
14.1
3.0
13.0
3.0
7.9
3.0
6.7
2.7
4.5
1.6
0.3
0.1
12/14/11
12/14/11
12/14/11
12/14/11
12/14/11
12/14/11
12/14/11
12/14/11
910
910
935
935
1000
1000
1030
1030
P3-1-1
P3-1-2
P1-1-1
P1-1-2
P1-2-1
P1-2-2
P3-2-1
P3-2-2
19.5
17.2
43.5
32.6
37.6
48.0
13.1
11.0
18.0
16.0
61.0
40.0
64.0
60.0
12.0
8.0
16.8
15.0
56.2
37.3
45.9
50.4
11.3
8.0
15.6
13.9
51.8
33.8
38.4
42.7
10.2
7.6
11.8
10.5
29.3
20.0
21.0
21.8
7.1
5.6
10.8
9.6
24.5
17.2
18.2
18.5
6.5
5.2
8.4
7.5
16.0
11.9
12.9
12.4
5.1
4.2
0.6
0.5
1.1
0.8
0.9
0.8
0.5
0.4
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
03/28/11
1430
1430
1800
1800
1650
1650
1630
1630
1700
1700
P3-in-1
P3-out-1
P3-in-2
P3-out-2
P1-in-2
P1-out-2
P1-in-1
P1-out-1
P3-in-1
P3-out-1
7.3
8.3
3.5
4.6
21.1
29.4
411.0
289.0
10.2
4.0
11.0
4.0
1.0
10.0
41.0
54.0
440.0
380.0
4.0
5.0
11.0
4.0
0.0
10.0
41.0
54.0
426.2
365.9
4.0
5.0
11.0
4.0
0.0
10.0
41.0
54.0
414.7
327.9
4.0
5.0
11.0
4.0
0.0
10.0
41.0
54.0
349.3
190.1
4.0
5.0
11.0
4.0
0.0
10.0
41.0
54.0
322.3
166.4
4.0
5.0
11.0
4.0
0.0
10.0
41.0
54.0
245.6
117.9
3.9
4.3
5.2
2.4
0.0
7.1
23.5
29.7
15.2
7.0
0.4
0.4
05/09/11
05/09/11
930
930
P1-1
P1-2
20.0
23.0
17.0
23.0
17.0
22.0
16.9
20.0
15.4
14.6
15.0
13.5
13.4
10.9
1.4
1.0
05/25/11
05/25/11
05/25/11
05/25/11
1600
1600
1615
1615
P1-1-In
P1-2-out
P3-1-in
P3-1-out
50.3
60.5
31.4
39.2
36.0
43.0
61.0
52.0
34.5
41.4
52.5
46.5
31.7
37.6
41.0
36.4
24.5
28.1
22.6
19.8
22.6
25.8
20.0
17.4
17.6
20.0
14.7
12.6
1.2
1.5
0.8
0.8
06/28/11
06/28/11
06/28/11
06/28/11
06/28/11
06/28/11
06/29/11
06/29/11
06/29/11
06/29/11
2230
2230
2245
2245
2300
2300
550
550
750
750
P1-1-1
P1-1-2
P1-2-1
P1-2-2
P1-3-1
P1-3-2
P1-4-1
P1-4-2
P1-5-1
P1-5-2
Avg EMC inflow
Avg EMC outflow
EMC (% Reduction)
38.7
58.6
88.9
78.6
64.3
45.4
18.0
16.6
82.6
31.5
59.3
38.4
35%
67.0
60.0
170.0
140.0
110.0
63.0
62.0
20.0
160.0
99.0
87.6
50.3
43%
59.9
54.3
150.9
126.3
102.2
59.4
58.6
18.7
135.9
91.1
71.4
46.6
35%
48.0
43.9
127.3
101.5
88.5
51.9
52.8
16.6
96.1
81.1
61.5
41.2
33%
30.7
26.5
72.4
52.4
57.1
32.9
37.5
11.4
61.3
58.3
40.8
26.8
34%
27.9
23.5
62.8
45.3
50.6
29.0
34.3
10.5
58.6
53.3
36.8
24.2
34%
20.8
16.7
42.8
30.5
35.3
20.3
26.6
8.2
49.8
41.1
27.7
17.0
39%
1.2
0.9
2.6
1.8
1.9
1.1
1.7
0.5
3.9
2.7
2.8
2.5
12%
13
Table 4
Date
Time
Site
10/4/2010
1740 P1
10/5/2010
950 P2
10/5/2010
1035 P3
10/5/2010
1120 P2-2
10/5/2010
1720 P3-2
10/5/2010
1905 P3-3
Overall Event Efficiency
P1 Efficiency
P2 Efficiency
P3 Efficiency
Date
Time
Site
10/24/10
1430 P3-1-2
10/24/10
1730 P3-2-1
P3 Event Efficiency
Date
Time
Site
12/02/10
1600 P3
12/02/10
1620 P3
P3 Event Efficiency
Date
Time
Site
12/07/10
1450 P3
P3 Event Efficiency
Date
Time
Site
12/14/11
910 P3-1-1
12/14/11
935 P1-1-1
12/14/11
1000 P1-2-1
12/14/11
1030 P3-2-1
Event Efficiency
P1 Efficiency
P3 Efficiency
Date
Time
Site
03/28/11
1430 P3-1
03/28/11
1800 P3-2
03/28/11
1630 P1-1
03/28/11
1650 P1-2
03/28/11
1700 P3-1
Event Efficiency
P1 Efficiency
P3 Efficiency
Date
Time
Site
05/09/11
930 P1
P1 Event Efficiency
Date
Time
Site
05/25/11
1600 P1-1
05/25/11
1615 P3-1
Event Efficiency
P1 Efficiency
P3 Efficiency
Date
Time
Site
06/28/11
2230 P1-1
06/28/11
2245 P1-2
06/28/11
2300 P1-3
06/29/11
550 P1-4
06/29/11
750 P1-5
P1 Event Efficiency
14
NTU
16%
36%
15%
78%
2%
29%
16%
57%
9%
NTU
-16%
18%
1%
NTU
56%
53%
54%
NTU
34%
34%
NTU
12%
25%
-28%
16%
6%
-1%
14%
NTU
-14%
TSS
<100 um
86%
80%
15%
10%
58%
55%
63%
62%
17%
22%
48%
86%
39%
37%
46%
80%
36%
39%
TSS
<100 um
18%
-3%
24%
12%
21%
4%
TSS
<100 um
68%
60%
57%
51%
63%
56%
TSS
<100 um
80%
79%
80%
79%
TSS
<100 um
11%
11%
34%
34%
6%
-10%
33%
29%
21%
20%
22%
16%
12%
20%
TSS
<100 um
64%
64%
% Reduction
<63 um
<20
78%
75%
12%
25%
53%
51%
60%
56%
23%
21%
<16
75%
27%
52%
54%
21%
<10
74%
30%
53%
51%
18%
<1
80%
53%
60%
21%
25%
46%
75%
40%
36%
46%
75%
41%
36%
45%
74%
40%
36%
48%
80%
37%
43%
% Reduction
<63 um
<20
-11%
-18%
12%
17%
<16
-20%
20%
<10
-21%
18%
<1
-16%
15%
0%
0%
-1%
0%
% Reduction
<63 um
<20
55%
46%
44%
4%
<16
47%
-12%
<10
48%
86%
<1
54%
44%
25%
17%
67%
49%
% Reduction
<63 um
<20
77%
62%
<16
60%
<10
64%
<1
68%
60%
64%
68%
<16
11%
30%
-2%
20%
<10
11%
26%
4%
18%
<1
15%
24%
11%
12%
15%
14%
16%
15%
14%
15%
15%
15%
14%
15%
17%
13%
% Reduction
<63 um
<20
64%
64%
<16
64%
<10
64%
<1
55%
45%
78%
36%
38%
1%
49%
77%
62%
% Reduction
<63 um
<20
11%
11%
35%
32%
-11%
-4%
25%
21%
15%
12%
18%
30%
-39%
61%
14%
-32%
-25%
14%
-32%
-25%
21%
-32%
-25%
46%
-32%
-25%
48%
-32%
-25%
52%
-32%
-10%
54%
-26%
-9%
9%
-5%
24%
5%
-9%
19%
5%
-9%
19%
7%
-5%
19%
13%
7%
19%
14%
8%
19%
18%
10%
27%
18%
14%
23%
% Reduction
<63 um
<20
-18%
6%
<16
10%
<10
18%
<1
29%
10%
18%
29%
<16
-14%
13%
<10
-13%
14%
<1
-19%
4%
-1%
-15%
13%
-1%
-14%
13%
0%
-13%
14%
-7%
-19%
4%
% Reduction
<63 um
<20
9%
14%
20%
28%
41%
42%
69%
70%
16%
5%
<16
16%
28%
43%
70%
9%
<10
20%
29%
42%
69%
18%
<1
21%
32%
44%
70%
31%
33%
35%
40%
NTU
-15%
-15%
NTU
-20%
-25%
-23%
-20%
-25%
NTU
-51%
12%
29%
8%
62%
12%
TSS
<100 um
-35%
-29%
-35%
-29%
TSS
<100 um
-19%
-20%
15%
11%
-2%
-19%
15%
-4%
-20%
11%
TSS
<100 um
10%
9%
18%
16%
43%
42%
68%
68%
38%
33%
35%
34%
-18%
6%
% Reduction
<63 um
<20
-18%
-15%
11%
13%
-4%
-18%
11%
31%
32%
Modeled Load
Using the rain intensity data along with the perlite filter efficiency results, the modeled
flows were used to calculate potential load reductions in a non-bypass / non-clogged
state. These load based values are estimated for all monitored storm events. This is an
example of the types of load reductions that could be expected if the filter system were to
operate 100% effectively without the need for bypass or clogging. Though this was not
the case for this study, the results suggest this may be possible with higher frequency
maintenance of the filter and pre-treatment of flows with the removal of coarse sediment
and detritus.
P1
P1
10/5/2010
Rain
12/14/2010
Rain on Snow
3/28/2011
Snowmelt
5/9/2011
Snowmelt
5/25/2011
Snowmelt
6/28/2011
Rain
15
Pre Filter Load (lb)
Post Filter Load (lb)
mg/l
NTU TSS (mg/l) <100 um <63 um <20
<16 <10 <1
5.32
66.52
36.49
26.67 11.06 9.38 6.30 0.38
4.48
9.36
7.17
5.81 2.75 2.35 1.63 0.08
Load Reduction
0.85
57.16
29.32
20.85
8.31
7.03
4.67 0.30
Pre Filter Load (lb)
Post Filter Load (lb)
2.69
3.02
3.57
4.38
3.30
3.40
3.02
2.86
1.74
1.51
1.47
1.29
0.98 0.07
0.89 0.06
Load Reduction (lb)
-0.34
-0.81
-0.10
0.16
0.23
0.18
0.09 0.01
Pre Filter Load (lb)
Post Filter Load (lb)
2.50
1.84
2.78
2.51
2.70
2.42
2.63
2.21
2.25
1.41
2.10
1.27
1.65 0.22
0.99 0.21
Load Reduction (lb)
0.66
0.27
0.27
0.43
0.84
0.83
0.66 0.01
Pre Filter Load (lb)
Post Filter Load (lb)
0.13
0.15
0.11
0.15
0.11
0.15
0.11
0.13
0.10
0.10
0.10
0.09
0.09 0.01
0.07 0.01
Load Reduction (lb)
-0.02
-0.04
-0.03
-0.02
0.01
0.01
0.02 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.34
0.40
0.24
0.29
0.23
0.28
0.21
0.25
0.16
0.19
0.15
0.17
0.12 0.01
0.13 0.01
Load Reduction (lb)
-0.07
-0.05
-0.05
-0.04
Pre Filter Load (lb)
Post Filter Load (lb)
2.46
1.94
4.79
3.21
4.27
2.94
3.47
2.48
2.18
1.53
1.97
1.36
1.48 0.09
0.98 0.06
Load Reduction
0.52
1.57
1.33
0.99
0.65
0.61
0.49 0.04
Total Modeled Load
Reduction (lb) All
Monitored Events
1.60
58.11
30.74
22.37 10.01
8.63
5.92 0.36
-0.02 -0.02 -0.02 0.00
P2
P2
10/5/2010
Rain
Pre Filter Load (lb)
Post Filter Load (lb)
Load Reduction
mg/l
NTU TSS (mg/l) <100 um <63 um <20 <16 <10 <1
26.46
10.54
9.70
9.07 6.56 5.93 4.54 0.49
6.87
4.62
4.40
4.17 3.16 2.92 2.35 0.36
19.58
5.92
5.31
4.90 3.40 3.01 2.19 0.12
P3
P3
10/5/2010
Rain
10/24/2010
Rain
12/2/2010
Rain on Snow
12/7/2010
Snowmelt
12/14/2010
Rain on Snow
3/28/2011
Snowmelt
5/25/2011
Snowmelt
16
Pre Filter Load (lb)
Post Filter Load (lb)
mg/l
NTU TSS (mg/l) <100 um <63 um <20 <16 <10 <1
1.63
1.88
1.68
1.48 0.87 0.76 0.55 0.03
1.60
1.65
1.48
1.30 0.75 0.65 0.46 0.02
Load Reduction
0.03
0.23
0.19
0.17 0.13 0.12 0.09 0.01
Pre Filter Load (lb)
Post Filter Load (lb)
0.54
0.54
1.52
1.19
1.00
0.94
0.77 0.38 0.33 0.23 0.01
0.74 0.37 0.32 0.22 0.01
Load Reduction
0.00
0.34
0.06
0.03 0.01 0.01 0.01 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.24
0.11
0.49
0.18
0.40
0.18
0.34 0.18 0.16 0.10 0.01
0.17 0.15 0.14 0.03 0.00
Load Reduction (lb)
0.13
0.30
0.22
0.16 0.03 0.01 0.08 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.04
0.03
0.10
0.02
0.09
0.02
0.09 0.05 0.04 0.03 0.00
0.02 0.02 0.02 0.01 0.00
Load Reduction (lb)
0.01
0.08
0.07
0.07 0.03 0.03 0.02 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.58
0.50
0.53
0.42
0.50
0.41
0.46 0.33 0.30 0.24 0.02
0.38 0.28 0.26 0.21 0.02
Load Reduction (lb)
0.08
0.11
0.09
0.08 0.05 0.04 0.03 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.13
0.12
0.12
0.10
0.12
0.09
0.11 0.08 0.07 0.06 0.00
0.09 0.07 0.06 0.05 0.00
Load Reduction (lb)
0.02
0.02
0.02
0.02 0.01 0.01 0.01 0.00
Pre Filter Load (lb)
Post Filter Load (lb)
0.21
0.26
0.41
0.35
0.35
0.31
0.27 0.15 0.13 0.10 0.01
0.24 0.13 0.12 0.08 0.01
Load Reduction (lb)
-0.05
0.06
0.04
0.03 0.02 0.02 0.01 0.00
Total Modeled Load
Reduction (lb) All
Monitored Events
0.22
1.14
0.70
0.55 0.29 0.24 0.25 0.02
Project Findings
The results of this study indicate that this type of filtration system can be effective, but
will be maintenance intensive and therefore, cost prohibitive. Though the system may
have better success in a parking lot or commercial area, it is unlikely that this type of
treatment will ever be implemented in a municipal setting in the County of El Dorado,
with its’ nearly 250 miles of roads including 153 miles of primary and secondary roads.
In addition to this system being both time consuming and expensive to maintain, it also
creates potential liability issues with flooding and plugging of storm drain infrastructure.
Cost Analysis
Even in the best of operating conditions, this system costs approximately $1500 per filter
unit and requires maintenance every event. The total cost for filter replacement including
labor is approximately $50 per storm or approximately $2000 / year. The first year cost
for the filters is approximately $2500 with a subsequent cost estimated at $2000 a year
with materials and labor.
The County of El Dorado estimated the filter was in operation approximately 10% of the
time as clogging was observed for every event monitored, requiring maintenance. This
was difficult to quantify as no pressure transducer was located in the filtration device
making it difficult to know exactly when it failed. The Cost benefit ratio is calculated
below for a 100% treatment scenario. The County of El Dorado anticipates that
approximately 1 hour of maintenance is needed per location and per storm event totaling
a maintenance labor of approximately 40 hours / year. The cost of maintenance is
calculated with a fully burdened rate of a $50 / hr for a maintenance worker annually.
This cost does not include equipment or materials. From this exercise it was determined
the unit cost of sediment removal to be at best case scenario between $29-$57 / lb of TSS
removed. The <16 micron portion of TSS was calculated between $85-$170 / lb. These
estimates are based on the system operating correctly every storm and assuming all
maintenance takes place to ensure operational performance during an event.
Annual Unit Cost (based on modeled loads)
TSS
(mg/l) <100 um
NTU
P1
P3
Efficiency All
Events
Avg EMC (mg/l)
35%
59
43%
88
Avg Load (lbs)
Avg Load (lbs)
39
20
70
35
Annual
Maintenance
Cost
$ / lb sediment
$ / lb sediment
17
$
$
$
2,000
51
102
$
$
29
57
$
$
<63 um
<20
<16
<10
<1
35%
71
33%
62
34%
41
34%
37
39%
28
12%
3
46
23
38
19
26
13
24
12
20
10
43
86
$
$
53
105
$
$
77
153
$
$
85
170
$
$
99
199
1
0
$
$
3,158
6,317
TMDL Crediting
In the Lake Tahoe Basin, a TMDL credit equals 1.0 × 1016 or 200 pounds fine sediment
particles with a diameter smaller than 16 micron (Lake Clarity Crediting Handbook
2011). For the best case scenario in attempting to understand credits as a result of this
treatment and potential incorporation into the Tahoe Pollutant Load Reduction Model
(PLRM), the County of El Dorado used the modeled flow data and calculated pollutant
removal rates with corresponding efficiencies to estimate load reductions. The range in
load annually based on actual monitoring data and modeled flow data determined that the
annual <16 micron load reduction can range from 12-24 pounds / filtration unit / year.
This equates to .06 - .12 credits annually per filtration unit when operated optimally with
intensive maintenance.
18
Turbidity and TSS
For the duration of the study it was found that a fairly good relationship between
Turbidity (NTU) and TSS exists. The R² was .91 and .92 respectively for TSS and <16
micron TSS.
Total Suspended Solids (TSS) vs. Turbidity (NTU)
TSS vs. NTU
450
400
R2 = 0.9085
350
300
TSS
250
200
150
100
50
0
0
50
100
150
200
250
300
350
400
450
NTU
<16 micron Total Suspended Solids (TSS) vs. Turbidity (NTU)
<16 Micron TSS vs. NTU
450
400
350
300
TSS
250
200
R2 = 0.9238
150
100
50
0
0
50
100
150
200
250
NTU
19
300
350
400
450
Conclusion
The results of this study indicate that there is a potential benefit to the use of perlite as a
filtration treatment media with large maintenance constraints. The reduction found in a
field setting resulted in a concentration reduction of 43% and 34% respectively for TSS
and <16 micron TSS. These were based on measured values. The total modeled load
reduction for all sites based on measured EMC values was 65 lbs and 11.9 lbs
respectively for TSS and <16 micron TSS, however the estimated actual load reduction
after plugging and clogging was significantly less.
The system while when operating effectively during events is efficient at fine particle
removal, the maintenance required for individual storms is time intensive and costly.
Issues also exist with liability as a result of clogging and potential flooding. For
example, during one event the flooding of a street caused flow to overtop the curb in the
middle of the night exposing underground power lines outside the Right of Way.
Another event required maintenance by a homeowner and DOT staff to prevent water
from entering the homeowner’s house. These systems if not maintained every event are
subject to clogging and could do more harm than good from a liability standpoint. Users
of these types of systems should consider the organic loading associated from the area as
pulverized organic debris appeared to be the primary cause of clogging and malfunction.
Once this occurred, not only did the system not work properly, but eventually clogged the
high flow bypass causing inundation of the drop inlet and flooding of the street. Slime
was frequently witnessed also, plugging the inlet screen. The source of the slime is
unknown at this time.
This type of treatment may be a benefit to areas with limited options for treatment such as
a parking lot or small drainage. With reasonable maintenance these systems can be
effective for fine particle removal.
Overall, the County of El Dorado’s experience utilizing this system was discouraging.
During winter months in freezing conditions it was observed that the filter would clog
with snow and be rendered ineffective. Maintaining the perlite filtration system in
freezing conditions was even more time intensive and difficult. The drop inlet system
appears impractical in the urban subdivision within Montgomery Estates, however the
concept still has validity for use within sediment traps where clogging does not cause
backwatering or flood conditions.
The County of El Dorado staff believes that infiltration should still be the primary
treatment of choice where feasible. Removing impervious cover and replacing with
infiltration will yield a net volume reduction and large treatment effectiveness and
pollutant load reduction. Infiltration treatment is easier to maintain, has a lower unit cost
and large treatment effectiveness as the volume of water is eliminated or reduced. With
perlite filtration, the best case scenario is a concentration reduction. Using infiltration;
the best case scenario is a concentration reduction and a pollutant load volume reduction
from entering a receiving water body.
20
Hypotheses Validation
Were the hypotheses as developed for this research validated by collected data?
The results of this study were used to test three hypotheses.
Research Hypotheses Tested:
•
Will filtering urban stormwater with granular perlite be effective in significantly
reducing fine sediment from urban runoff
o Hypothesis Validated - The data suggest that fine sediment load in
stormwater can be significantly reduced by perlite filtration for the
conditions tested.
•
Will filtering urban stormwater with granular perlite be practical for installation
and operation in a drainage inlet within an extensive storm drain system
o Hypothesis Not Validated - This system requires a high level of
maintenance and can be a liability concern during storm events, therefore
for the conditions tested was not practical for this type of installation.
•
Will filtering urban stormwater with granular perlite be economic relative to
other filtration alternatives for the treatment of fine sediment
o Hypothesis Validated with further investigation needed – Although
the system requires elevated levels of maintenance, the fine sediment
capture capacity is reasonable relative to cost. The system if maintained
properly and consistently can yield load reductions, but the utility of this
system must be evaluated individually on a site by site basis.
21
References
Caltrans (2007). Caltrans Lake Tahoe Storm Water Small-Scale Pilot Treatment Project
Phase II Report. December, 2003.
Jassby, Alan D. et al. (1999). Origins and scale dependence of temporal variability in the
transparency of Lake Tahoe, California-Nevada. Limnology and Oceanography 11 (2):
282-294
Kang, Joo H. et al. (2007). Particle destabilization in highway runoff to optimize
pollutant removal. Journal of Environmental Engineering. 133 (4): 426-434.
Lahontan Water Quality Control Board and Nevada Division of Environmental
Protection. 2011. Lake Clarity Crediting Program Handbook: for Lake Tahoe TMDL
Implementation v1.0. Prepared by Environmental Incentives, LLC. South Lake Tahoe,
CA.
Mailloux, James T. (2006). Verification testing of the gravity-flow aqua-filter filtration
cartridge with Sil-Co-Sil 106.
Reuter, J. E. et al., (2003). An integrated watershed approach to studying ecosystem
health at Lake Tahoe, CA-NV. Managing for Healthy Ecosystems, 1283-1298.
Roberts, David M. and Reuter, John E. Draft Lake Tahoe Total Maximum Daily Load
Technical Report, California and Nevada, September 2007.
Swift, Theodore J. et al. (2005). Water clarity modeling in Lake Tahoe: Linking
suspended matter characteristics to secchi depth. Aquatic Science 68: 1-15.
22
Appendix A
Selected Photos
October 4, 2010 at 5:30 am…
With a rain intensity of .25 inches / hr the filter system completely clogged and the street flooded.
Emergency measures taken to remove the filter and allow drainage. Resulted in damage to slope on
backside of curb and exposing electric power line.
October 5, 2010 at 11:40 am…
Damage to slope on backside of curb and exposure of three 600 volt power lines. Stormwater generated
Erosion did not make it to surface waters.
23
May 25, 2011
Material Accumulation from a single storm event rendering the filter inoperable
May 27, 2011 at 1:30pm…
Typical clogging after 1 rain event. The system was maintained just prior to the event.
24
Appendix B
Rain Event Data Example
The storm precipitation total for the October 3-5, 2010 event was 2.1 inches with
maximum rain intensity of .23 inches in 15 minutes (Below). A total of 12 samples were
collected for the duration of the storm.
Rain Event 10-4-10
0.25
Precipitation
Samples
(in/15min)
0.2
0.15
0.1
0.05
0
Rain
Date
10/04/10
10/04/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/05/10
10/2/10
12:00
10/2/10
20:24
Time
1740
1740
950
950
1035
1035
1120
1120
1720
1720
1905
1905
Site
P1-1
P1-2
P2-1
P2-2
P3-1
P3-2
P2-2-1
P2-2-2
P3-2-1
P3-2-2
P3-3-1
P3-3-2
Date
Time
Site
10/04/10
1740 P1-1
10/04/10
1740 P1-2
10/05/10
950 P2-1
10/05/10
950 P2-2
10/05/10
1035 P3-1
10/05/10
1035 P3-2
10/05/10
1120 P2-2-1
10/05/10
1120 P2-2-2
10/05/10
1720 P3-2-1
10/05/10
1720 P3-2-2
10/05/10
1905 P3-3-1
10/05/10
1905 P3-3-2
Date
Time
Site
10/4/2010
1740 P1
10/5/2010
950 P2
10/5/2010
1035 P3
10/5/2010
1120 P2-2
10/5/2010
1720 P3-2
10/5/2010
1905 P3-3
Overall Event Efficiency
P1 Efficiency
P2 Efficiency
P3 Efficiency
25
10/3/10
4:48
10/3/10
13:12
NTU TSS
47.2
39.7
48.3
31.1
12.9
11
421
90.8
21.2
20.7
9.17
10.8
NTU
47.2
39.7
48.3
31.1
12.9
11
421
90.8
21.2
20.7
9.17
10.8
NTU
16%
36%
15%
78%
2%
29%
16%
57%
9%
10/3/10
21:36
590
83
27
23
26
11
160
59
18
15
6
18
10/4/10
6:00
10/4/10
14:24
10/4/10
22:48
10/5/10
7:12
10/5/10
15:36
10/6/10
0:00
Percent of TSS
<100 um <63 um <20
<16
<10
<1
55%
40%
17%
14%
9%
1%
77%
62%
29%
25%
17%
1%
86%
78%
53%
48%
36%
4%
91%
81%
47%
41%
29%
2%
85%
72%
40%
35%
25%
2%
91%
80%
46%
39%
28%
2%
93%
87%
64%
58%
44%
5%
97%
94%
77%
72%
59%
10%
93%
86%
52%
45%
32%
1%
87%
79%
49%
43%
31%
1%
96%
88%
59%
53%
41%
2%
91%
78%
42%
36%
25%
1%
mg / l
TSS (mg/l) <100 um
<63 um
<20
<16
<10
<1
590
323.62
236.51
98.08
83.16
55.84
83
63.60
51.54
24.40
20.81
14.44
27
23.13
21.18
14.43
12.83
9.60
23
20.86
18.63
10.77
9.36
6.75
26
22.15
18.61
10.32
8.97
6.47
11
9.99
8.75
5.01
4.33
3.03
160
148.99
139.67
101.99
92.35
70.97
59
57.12
55.25
45.29
42.40
34.91
18
16.77
15.42
9.39
8.14
5.69
15
13.02
11.82
7.37
6.46
4.64
6
5.75
5.25
3.55
3.20
2.44
18
16.47
14.10
7.51
6.44
4.49
TSS
<100 um
86%
80%
15%
10%
58%
55%
63%
62%
17%
22%
48%
86%
39%
37%
46%
80%
36%
39%
% Reduction
<63 um
<20
78%
75%
12%
25%
53%
51%
60%
56%
23%
21%
45%
78%
36%
38%
46%
75%
40%
36%
3.34
0.67
1.09
0.51
0.43
0.17
7.54
5.94
0.19
0.14
0.12
0.16
<16
75%
27%
52%
54%
21%
<10
74%
30%
53%
51%
18%
<1
80%
53%
60%
21%
25%
46%
75%
41%
36%
45%
74%
40%
36%
48%
80%
37%
43%