Get documents about "Bioretention in an Urban Context by Michael J. Coyle,
Document Sample
Contact: Mike Coyle, 215.569.2900, x3393, mcoyle@klingstubbins.com
This article was featured in the August 2007 issue of Construction Specifier.
Bioretention in an Urban Context
by Michael J. Coyle, PE, and Kevin Selger, RLA, ASLA, LEED® AP
Increasingly, development projects are required to provide stormwater best management practices (BMPs) to
improve water quality.1 Historically, the measures used in urban projects have involved underground devices (e.g.
sand filters or subterranean storage chambers) due to a lack of usable surface space.2
However, when land area is available (or mandated, as in the case of jurisdictions requiring open-space or
parkland), surface level BMPs can be employed, such as bioretention areas. This strategy involves a soil and plant-
based stormwater filtering process to rid water of potentially toxic content. A critical component of low-impact
development (LID) philosophies, it uses green space to keep stormwater on-site, improving its quality with natural
means, and minimizing runoff by maximizing infiltration.
Bioretention areas have porous soil covered with a thin layer of mulch. Various plants (e.g. grasses, shrubs,
trees, or a combination thereof, site permitting) encourage evapotranspiration, maintain soil porosity, encourage
biological activity, and promote uptake of some pollutants. Water runoff from impervious, hardscape areas (e.g.
sidewalks, roofs, and parking lots) is directed into the bioretention area where it then infiltrates through the
A R CH I T E C T U R E plant/mulch/soil environment.
ENGINEERING
INTERIORS
This can mean excellent removal of total suspended solids and phosphorus, while also allowing—to a lesser
PLANNING
extent—for the removal of nitrogen and nitrates.3
Cambridge
L a s Ve g a s
Philadelphia
Raleigh
San Francisco
Wa s h i n g t o n , D C
By using a case study of a recent construction project, this article examines the suitability (and effi cacy) of large-
scale bioretention systems within an urban setting.
Case study in point
A Wilmington, Delaware, high-density residential redevelopment (including a 22-story apartment tower, 63
townhouses, parking, and green space along the Christina River) incorporates bioretention systems. The existing
site had been completely paved and once functioned as a fuel storage facility.
Designated as a brownfield, the remedial action was to cap the
existing soil in place with a geotextile barrier overlain with imported,
clean soil. This soil cap (at least 0.6 m [2 ft] thick across the site)
meant limited growing depth, which proved to be a challenging
aspect in implementing on-site bioretention above, while minimizing
infiltration below.
Bioretention and sand filters are accepted by the state’s Department
of Natural Resources and Environmental Control (DNREC) as
standalone BMPs to meet Delaware’s target of 80-percent removal
of total suspended solids (TSS), a measurement formerly known
as ‘non-fi lterable residue’ (NFR). For this project, bioretention was
selected as the primary BMP because of the availability of land
and because it was felt the method offered a more cost-effective
solution than sand filters, which were still used in areas without
sufficient surface area.
This Wilmington, Delaware, high-density residential Since the site was designated as a brownfield, trees and shrubs
redevelopment was once a fuel storage facility. Transformed
could not be used due to fears over root systems penetrating the
from a brownfi eld into a residential complex, it incorporates
subtle bioretention systems. contaminated earth. This limited the use of plants to mainly wet-
tolerant perennials and grasses. Overland flows, underground pipe
systems, and up-well structures feed water runoff into 16 bioretention areas, which are underdrained with a
geotextile-wrapped gravel layer and perforated pipe system. These areas allow for 152.4 mm (6 in.) of ponding
depth (up to the 10-year storm), before the water fl ows into a raised, grated inlet basin with an outflow pipe
designed to handle larger rainfall events (i.e. up to the 100-year storm).
Regulations, requirements, and criteria
With any development project, there are typically requirements to control or manage the rate and volume of
runoff leaving a site, along with improving water quality. In the case of the Wilmington project, fl ood control
or detention were not requirements because there was already a net decrease in impervious coverage—before
construction, the site had been completely asphalt-paved.
While the state mandated permanent water quality measures to achieve 80-percent removal effi ciency for
suspended solids, there were no specifi c requirements to remove other pollutants. In Delaware, the mandated
design criteria for water quality includes managing the rate and volume of flow from the 50.8-mm (2-in.) Natural
Resources Conservation Service’s (NRCS’s) Type II rainfall event and up to a maximum of 25.4 mm (1 in.) of runoff.
DNREC also provided general design guidance for bioretention area planning. While working with the state
during the course of this project, adjustments were made to design criteria to account for constraints from the
site’s existing conditions. Figure 1 summarizes the design guidelines.
Bioretention design
An important design aspect involved figuring out how to dig without having heavy equipment further compact
the soil under the bioretention areas. The width of the areas was designed so a bucket excavator (i.e. track-hoe)
could access the middle of the excavation from outside the bioretention zone. The excavators had an approximate
6.1-m (20-ft) reach, which provided a maximum of 12.2 m (40 ft) of width for an area.
The biofiltration media (i.e. soil) was mechanically mixed to provide a homogenous blend to promote even infi
ltration and permeability throughout. This mixture was placed in 0.3-m (1-ft) lifts and was partially spread with
the excavator bucket and hand-raked to minimize compaction. The project site contains 16 separate bioretention
areas ranging in size from 6.3 to 244 m2 (68 to 2627 sf) of soil media surface area. Dimensions were determined
by dividing the maximum rainfall volume of 25.4 mm (1-in.) over the contributing drainage area by the load of
0.8 m (2.75 ft) per design guidance from DNREC’s Division of Soil and Water. The maximum 25.4-mm runoff
volume was used because all drainage areas had a runoff coeffi cient of at least 0.50. The bioretention areas
received runoff from paved surfaces, building roofs, and landscape areas. The post-developed site is 63-percent
impervious, which equates to about 2 ha (5.04 acres) of impervious surfaces.
As mentioned, stormwater inflow occurs through both underground piping and overland flow. One of the
bioretention areas has a unique inflow system in which the underground piping discharges below the bottom
of the zone into a sumped catch basin structure. Stormwater fills the catch basin and discharges through its
top grate into the bioretention area. This design became a necessity due to the site’s flatness, in addition to
brownfield-related constraints, which drove the infl ow piping below the basin.
Throughout Delaware, rooftop drainage must be treated for water quality, but further municipal regulations
meant this drainage needed to be sent to underground piping instead of on-grade. The catch basin acts as a
plunge pool to dissipate large flows and collect sediment. This unique design required special attention to the
hydraulics of the conveyance system due to the high tail-water conditions.
The bioretention facilities are sized to ensure routing occurs. This prevents the volume of the water quality
storm (i.e. 25.4 mm [1 in.] of runoff in Delaware) from exceeding the maximum storage depth of 152.4 mm (6
in.), allowing filtration through the soil. The depth was determined by routing the storms with an hourly media
infi ltration rate of 23.4 mm (0.92 in.). This is the minimum required to drain the 0.8-m (2.75-ft) load within
Delaware’s 36-hour draw down time.4
Although the actual infiltration will most likely be higher, achieving a rate of 23.4 mm/hr conservatively deals with
the possibility of clogging of the media due to silting and lack of maintenance. Runoff from larger storms will
pond deeper than 152.4 mm (6 in.) and fl ows into the top grate of the outflow structure and into the main storm
system draining to the Christina River. Figure 2 illustrates a typical section of a bioretention area.
Filtering soil media
The bioretention facilities
are 0.6-m (2-ft) thick soil
media filters consisting of
a mixture of equal parts
sand, tripleshredded aged
hardwood mulch, and
sphagnum peat moss. This is
the DNREC-recommended soil
mixture under its 2005 Green
Technology: The Delaware
Urban Runoff Management
Approach—Standards, Specifi
ations, and Details for Green
Technology BMPs to Minimize
This Wilmington, Delaware, high-density residential redevelopment was once a fuel storage facility. Transformed
Stormwater Impacts from Land from a brownfi eld into a residential complex, it incorporates subtle bioretention systems.
Development. The sand must
also be silica-based, conforming to ASTM International C 33, Standard Specifi cation for Concrete Aggregates.
Additionally, a 76.2-mm (3-in.) mulch layer is specified at the surface to increase filtering capability and retain
moisture within the soil media.5
Underdrain system
Beneath the soil is a 101.6-mm (4-in.)
diameter, perforated polyvinyl chloride
(PVC) pipe underdrain within a 152.4-
mm (6-in.) thick layer of No. 57 stone
(double-washed), surrounded with woven
geotextile on the top and sides. The state
permitted a reduction in the stone layer
thickness from the required 228.6 mm (9
in.) to accommodate the limited depth of
clean fill within the brownfield cap and to
stay above the spring high groundwater
elevation.
Water runoff from impervious, hardscape areas (such as this parking lot) can be directed into the
bioretention area where it then infi ltrates through the plant/mulch/soil environment for removal of
potentially harmful constituents.
Plant materials
Along with the soil, plants are integral for filtering and absorption. Another aspect to the selection and planting
design for the Wilmington project involved using the bioretention areas as landscape features. In several cases,
they were located in park-like open spaces next to the riverwalk, apartment tower, and townhouse complexes. As
such, they had to have an aesthetic look and be integrated into the overall landscape design.
Most of the plants are native to the region, with some considered ‘facultative’ (i.e. wetland or wet-tolerant).
Since native species are already suitable for regional conditions, they required little maintenance inputs (e.g.
watering, fertilization, pesticides). Plants were selected for tolerance to periodic inundation and ponding, as well
as resistance to drought. Another component in the selection process was finding species that could endure salt,
as the region experiences snow and ice in the winter and de-icing chlorides are frequently used on pavements. An
ongoing issue for landscape near paved areas concerns concentrations of salt in the runoff, which can affect plant
survivability and the bioretention system’s ability to filter and absorb suspended solids and pollutants.5
DNREC guidelines state mature plant growth size should occupy no more than 50 percent of the bioretention
surface area. The reasoning for this involves:
• providing open, mulched areas for increased fi ltering;
• ensuring plantings do not grow so dense they decrease surface permeability; and
• allowing maintenance access for mulch replacement.
Given the selected plant’s growth habits, the 50-percent maximum plant coverage requirement works out to
about 21 plants per 9.29 m2 (100 sf). The plan calls for locating the individual species in groupings of fi ve to
10 per group to add an aesthetic design aspect. However, in the authors’ experience, this 50-percent maximum
plant coverage could be problematic because it likely increases the potential for weed establishment and impacts
the landscape aesthetics, both of which will make bioretention strategies less desirable to clients, designers, and
the public.
Weeds (e.g. Canadian Thistle) seed themselves into open areas and out-compete the installed plantings. Once
weeds become established they are difficult to manage, requiring intensive manual removal/maintenance or
chemical control (the latter would be a path of last resort, given the goal of reducing pollution). In the authors’
opinion, 100-percent plant coverage mitigates these issues without decreasing surface permeability. The plants
and their roots in the soil become part of the matrix that filters pollutants.
Water quality analysis
Two of the 16 bioretention areas were
modeled with the Delaware Urban Runoff
Management Model (DURMM) program for
the pre- and post-development conditions.
This estimates pollutant loads based on event
mean concentrations (EMCs) associated with
different types of pervious and impervious
land covers. DURMM uses the area and EMC
of individual land covers to develop an area-
weighted average EMC for the pre- and post-
development drainage areas. The parameters
of the bioretention area are then entered
into the model to determine the adequacy of
soil depth and surface area in removing the
incoming pollutant load.
Illustrated in Figures 3 and 4, the results of the
DURMM for bioretention areas 14 and 16 showed 100-percent pollutant removal for TSS, phosphorus, and nitrate.
(These two areas in particular were chosen for accessibility of inf luent and eff luent sampling points.) Unlike other
filtering BMPs, pollutant removal efficiency is not dependent on design parameters. Since the required media
depth provides as much removal as possible, there are no design-dependent adjustments to performance.6
A field-testing program of the two bioretention areas was performed using the laboratory services of Analytics
Corp., to analyze the stormwater influent (pre-treatment) and effluent (post-treatment) for pollutant loading and
removal rates for suspended solids, phosphorous, and nitrogen. The sampling results are summarized in Figure
5.
Lessons learned
During both the Wilmington project’s design and construction phases, multiple issues were uncovered and solved.
The authors highly recommend developing a working relationship with the municipal/state review agency and the
reviewing professional assigned to the project. This became extremely important with the project’s unique design
challenges, especially with the issues revolving around the brownfield.
It is also important to ensure construction notes and specifi cations align with regulatory requirements and
guidelines. The specifi cations should provide for the contractor submitting shop drawings, samples, and product
data for review and approval by the designer prior to purchase and installation. This can assist in the process of
ensuring proper installation through construction administration/observation. In Delaware, a certified construction
reviewer is required to observe the installation of the stormwater quality measures.
Erosion controls must be maintained throughout the project, especially with online bioretention systems. If
possible, the installation of soil mix should be scheduled toward the end of the project or after most of the site is
stabilized. Plant materials and container/plug sizes should be locally available. If large quantities or special species
are required, it is recommended a contract growing program be established early so plants are ready when
bioretention areas are installed.
The Wilmington project has shown large-scale bioretention can be incorporated into a dense urban redevelopment.
It also illustrates the importance of having a strong relationship with the reviewing agency to work through
design issues. Without this rapport, the stormwater treatment strategy may have been discounted early in the
design process. Bioretention can also become a feature in the landscape and not relegated to the edges of
parking lots and roads at the back of a site, while improving runoff quality and decreasing development’s impact
on the surrounding environment.
Notes
1
This article is adapted from the authors’ presentation at the 2006 StormCon conference held in Denver, Colorado.
They wish to thank three KlingStubbins colleagues—Joseph Cuilla, PE (principal and chief civil engineer), Donald
Gallagher, RLA, LEED AP (senior landscape architect), and Robert Maloney, RLA (project landscape architect).
2
For more on one such option, see “Breaking the Surface—Commercial stormwater management from the
ground up” by Fred Dotson in the November 2006 issue of The Construction Specifier.
3
See “Optimizing Bio-retention Design to Improve De-nitrification in Commercial Site Runoff,” by W. Hunt et
al (North Carolina State University’s Department of Biological and Agricultural Engineering). Visit www.bae.
ncsu.edu/cont_ed/main/handouts/ASCEbio-ret.pdf. See also Green Technology: The Delaware Urban Runoff
Management Approach—A Technical Manual for Designing Non-structural BMPs to Minimize Stormwater
Impacts from Land Development by William C. Lucas (Integrated Land Management and DNREC’s Division of Soil
and Water Conservation, 2004).
4
See the aforementioned DURMM: The Delaware Urban Runoff Management Approach manual.
5
For more on the dangers of de-icing, see “Architectural Metal Corrosion—The de-icing salt threat,” by Catherine
Houska, CSI, in the December 2006 issue of The Construction Specifier.
6
In the authors’ experience, Delaware’s recommended soil mixture is not commonly used outside the state. An
alternative soil mixture contains an actual soil component (screened topsoil or existing in-situ soils), sand, and
an organic material (e.g. compost, peat moss, or leaf mold). The alternative mixture comprises 50 to 60 percent
soil, 30 to 40 percent sand, and 10 percent organic material. Replacing most of the organic component (i.e. peat
and mulch) with a soil component is desirable to sustain and encourage plant growth, while also decreasing
settlement from decomposition of the larger quantities of organics in the DNREC mixture.
Authors
Michael J. Coyle, PE, has more than 15 years of experience as a civil engineer in the environmental and land
development fields and currently works out of the Philadelphia, Pennsylvania, offices of KlingStubbins. He is
an associate member of the American Society of Civil Engineers (ASCE). Coyle can be contacted via e-mail at
mcoyle@klingstubbins.com. Kevin M. Selger, RLA, ASLA, has more than eight years of experience as a landscape
architect and a site planner for a wide range of projects across the United States and overseas. Also employed by
KlingStubbins, he is a member of the American Society of Landscape Architects (ASLA) and is a licensed/registered
landscape architect in several states. Selger can be contacted via e-mail at kselger@klingstubbins.com.
KlingStubbins provides professional services in all major disciplines within the realm of architecture, engineering,
interiors, planning, and landscape architecture. The firm consists of more than 550 professionals in its Cambridge,
MA; Las Vegas, NV; Philadelphia, PA; Raleigh, NC; San Francisco, CA; and Washington, DC offices. Its areas of
market focus and specialization include Corporate/Commercial, Government, Health Care, Higher Education,
Hospitality/Entertainment, Institutional/Civic, Mission Critical, and Research and Development.
KlingStubbins can be found online at www.klingstubbins.com.
###
Related docs
