JAWRA article - Rain Gardens

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

ASLESON, NESTINOEN, GULLIVER, HOZALSKI, AND NIEI3ER
computed coefficient of variation (CV) values was used
to determine the appropriate distribution for the mea-
sured Ksat values. Possible outliers were not removed
due to the highly heterogeneous soil expected in vege-
tative landscapes and the uncertainty involved with
removal of such data points.
The median or mean Ksat value from the infiltrom-
eter results; the surface area, A (m and volume, V
(m of the water in the rain garden at the time of
the test were used to predict the drain time, t, using
the following equation:
V
Drain Time = (2)
Ks.tA
Equation (2) will provide a rough estimate of the
drain time, subject to the following assumptions:
(1) The mean piezometric head over the soil, repre-
sented by VIA, is appropriate to estimate drain
time, and
(2) The soil is saturated at the conclusion of filling,
when drain time measurements commenced.
Synthetic Drawdown Tests
The flow rate needed to fill a rain garden with water
for the simulated runoff assessment was calculated to
determine if an adequate supply of water was avail-
able. The parameters necessary to estimate this flow
rate include the surface area of the basin, the esti-
mated or measured infiltration rate, and the storage
capacity of the basin. After filling the rain garden with
water to the highest level possible, the water level vs.
time was recorded using a staff gauge and a stop
watch. Measurements also were collected every second
and averaged over 10 s intervals with an ultrasonic
sensor (MassaSonic, M -5000) mounted to a postset at
the lowest point in the basin. The time required to fill
each rain garden was —30 min. Measurements of
water level began as soon as the inflow was turned off.
The reported drain times are based on the actual vol-
ume of water used during the synthetic drawdown test
and not the maximum capacity of each site.
RESULTS AND DISCUSSION
Visual Inspection
All of the rain gardens contained various species of
native perennial vegetation. Four of the sites con-
tained new plantings and were considered to be in
good health based on their early stage in develop-
ment. Four rain gardens suffered from an obvious
lack of infiltration observed during a visual inspec-
tion, including ponded water, the presence of hydric
soils and wetland plants, and a lack of plant growth
on compacted soil. These rain gardens failed the
Level 1 inspection and will require rehabilitation.
Vegetation health at the Cottage Grove (7) site was
rated as poor based on the presence of failing trees.
Nevertheless, the prairie grasses and perennial
plants appeared to be established and growing well.
With the exception of the four sites determined to be
nonfunctional and the Cottage Grove (7) site, the
other sites bad well - established vegetation that
appeared to be healthy.
The results of the inspection of the soils at each
rain garden are provided in Table 2. Rain gardens
typically have mulch covering the soil surface for
moisture and weed control. The topsoil of several rain
gardens consisted of a sandy loam soil. Examining
the soil profile, as described in the methods section
for visual inspection, of the rain gardens is important
for the detection of restrictive soil layers that may be
present due to improper construction or have formed
over time due to prolonged saturation. Soil profiles
can give especially useful information as it is possible
that a restrictive layer contributes to the desired
drain time not being met. At the UM — St. Paul (6)
campus rain garden, for example, a lower permeabil-
ity silt loam soil layer was found to underlay the
sandy loam topsoil, which was later found to be an
adjustment made during construction. Two other rain
garden sites [Thompson Lake (11) and UM — Duluth
(12)] had an underlying native soil of finer texture
(silt loam and clay) than the overlying topsoil (loamy
sand and sandy loam). These two sites were designed
with underdrains to compensate for the restrictive
layer. The soil profile at the Cottage Grove (7) site
consisted of forty inches of sand overlying gravel. The
poor retention of water and nutrients in the sandy
soil were a potential cause of the failing plants
observed during inspection of the vegetation.
Two sets of samples were taken for the UM —
Duluth (12) rain garden as it contained two different
types of soil. The mean bulk densities of the Cottage
Grove (7) rain garden (Table 2) and the coarse sand
overlying the underdrain system at the UM — Duluth
(12) rain garden were 1.57 and 1.53 g /cm respec-
tively, which are typical values for sand. There was
several bulk densities of — 1.0 -1.2 g /cm which is
indicative of loamy soils and was expected based on
the texture of the topsoils in the rain gardens. Sandy
soils with a relatively low volume of pores may have
a bulk density of —1.6 g /cm whereas aggregated
loams and clay soils fall below 1.2 g /cm (Hillel,
JAWRA 1024 JOURNAL OF THE AMERICAN WATER RESOURCES ASSOCIATION