JAWRA article - Rain Gardens

• Project Type: General OWC
• Title: Performance Assessment on Rain Gardens
• Date: 12/15/2009
• Water Conservation - Doc Type: Reports

ASLESON, NESTINGEN, GULLIVER, HOZALSKI, AND N IEEER
University of Minnesota -St. Paul Campu
Inlet
Inlet
•
O
• ° O
i
i
• O o •
• O
f�
Hydraulic Conductivity
[cmis]
•
0.0000006 - 0.0006
0.0006- 0.0012
O
0.0012 - 0.0024
0.0024 - 0.0048
•
0.0048 - 0.0096
•
0.0096 - 0.016
FIGURE 7. Map Showing the Range of Ksat Values
Measured Using the MPD Infiltrometer at Various Locations
Within the UM — St. Paul (6) Rain Garden.
1994; Tsegaye and Hill, 1998; Jang and Liu, 2004;
Regalado and Munoz - Carpena, 2004).
Given the distributions of Ksat, shown in Table 3
and Figure 8, we can now determine the number of
measurements required to accurately estimate the
true mean of the Ksat. This will provide guidance on
selecting the appropriate number of measurements to
conduct other experiments or Level 2 assessments.
Equation (3) can be used to compute the estimated
number of measurements (1) required to be within
specified range of the mean (C.L — µ), where z is the
tabulated za/2 value for the desired confidence level of
estimation (Klute, 1986).
N _ ( SD x z Z
/ I 3
CI — µ
The number of measurements (1) that would be
necessary to obtain a mean Ksat value within selected
levels of tolerance (i.e., maximum acceptable differ-
ence between the true and computed mean values)
was calculated for each rain garden assuming a 95%
confidence interval. The results of these calculations
along with the actual number of measurements made
(1) for each rain garden are shown in Table 4.
Synthetic Drawdown Tests
Three of the sites were evaluated using synthetic
drawdown testing (i.e., Level 3). Selection of these
sites was based on the size of the rain garden and
availability of a water supply. The rain gardens
selected for synthetic drawdown testing were Cottage
Grove (7), RWMWD #5 (5), and the UM — St. Paul (6)
campus. A fire hydrant was used to fill both the
RWMWD #5 (5) and UM — St. Paul (6) rain gardens.
The Cottage Grove (7) rain garden required the use
of a water truck because there was no fire hydrant
nearby. The combination of limited volume of water,
relatively low delivery flow and high infiltration rate
of the soil only allowed the site to be filled to 28% of
its estimated maximum capacity. The RWMWD #5
(5) rain garden was filled to 72% of the maximum
capacity to prevent overflowing the basin. The UM —
St. Paul (6) rain garden was the only site filled to
capacity during the synthetic drawdown test. This
site had an overflow weir connected to a second rain
garden. The rain garden was filled until water began
to overflow into the next rain garden and measure-
ments began when overflow ceased.
The measured drainage times of the three rain
gardens are plotted in Figure 9. The Cottage Grove
(7), RWMWD #5 (5), and UM — St. Paul (6) rain gar-
dens drained in 0.14, 3.13, and 2.14 h respectively.
All three rain gardens drained well before the desired
maximum 48 -b. drainage period, which indicated that
the rain gardens were infiltrating water.
The infiltration rate tests were also used to esti-
mate the drain time based on the same initial volume
used for each synthetic drawdown test. The appropri-
ate Ksat value to estimate drainage time, if the piezo-
metric head difference is assumed to be equal and
the low Ksat values are spaced uniformly throughout
the basin, may be determined by conceptually placing
two media of differing Ksat next to each other with
one piezometric head applied. An application of
Darcy's law to the flow through these two media
results in an overall flow that is dependent upon the
arithmetic mean of the two Ksat values. The drain
time is thus proportional to the inverse of the arith-
metic mean of the Ksat values. The drain times com-
puted using the arithmetic mean Ksat values for all
three rain gardens, however, were lower than the cor-
responding synthetic drawdown test drain times (Fig-
ure 9). It is believed that the location of the higher
Ksat values around the perimeter of the basins influ-
enced the spatially averaged, arithmetic mean to a
JAWRA 1028 JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION