Low Impact Development Technology: Design Methods and Case Studies

Biofiltration devices are a potentially effective option for the treatment and disposal of stormwater runoff from urban areas. However, the performance of these systems, and other infiltration devices, are affected by factors such as the texture, structure, and degree of compaction of the media during their construction. This paper presents the results of laboratory and field-scale studies conducted to provide insight on media characteristics of a poorly operating biofilter facility located in Tuscaloosa, AL. Double ring infiltrometer tests and soil compaction measurements were conducted along a large biofilter to determine the in-situ characteristics of the media. Infiltration measurements were also made during rain events for comparison to the results from the lab tests and the double-ring infiltration measurements. The effects of different compaction levels on the infiltration rates through the soil media were also examined during laboratory column tests, along with benefits associated with adding sand to the media mixture. The results indicated that soil compaction has dramatic effects on the infiltration rates; therefore, care needs to be taken during stormwater treatment facilities’ construction to minimize detrimental compaction effects. Adding sand to a media having large fractions of clay-sized particles helps minimize the detrimental effects of compaction on the infiltration rates also.

A newly constructed bioretention cell is like any other natural system; it can be expected to change over time as its ecosystem develops and establishes itself. Among these changes is the development of an enriched organic layer, or topsoil, at the soil surface, which is produced as plants grow and decay, and colonizing invertebrates and bacteria incorporate the resulting material into the soil. This enriched organic layer has properties significant to engineers, including increased porosity, increased cation exchange capacity, and increased bacterial activity. A field study was conducted to characterize the development of bioretention soils over time. A chronosequence consisting of 10 bioretention cells of ages ranging from one to ten years was assembled. Parameters measured included: horizon depth, soil texture, soil organic matter, earthworm abundance and diversity, root biomass, and macroinvertebrate abundance and diversity. This study revealed a great deal of variation between sites in terms of their design and development, but also uncovered general trends that seem to apply universally. All sites were found to exhibit a characteristic soil profile in which soil organic matter and macro-biological activity decrease exponentially with depth. Earthworms were found at all sites, and their burrowing activity may play a key role in preventing clogging by solids deposition. The study reveals that biological activity is ubiquitous in bioretention soils, despite their location in isolated and degraded settings, such as parking lot islands. These findings may have significance for engineers, as they are likely to impact a bioretention cell’s hydraulic and pollutant removal performance. Further research is needed to quantify these effects, and to explore the relationship between the emerging bioretention ecosystem and its encompassing urban ecosystem.

In this paper we demonstrate that in the upper soil column, vegetation type and cultivation practices can have as much of an influence on soil moisture retention and infiltration as soil texture class. First, we show that better a priori estimates of infiltration for natural and naturalized areas can be derived using soil data from natural landscapes rather than from cultivated areas. Our infiltration estimates, derived from German forest soil water retention data, compare significantly better to measured data from natural/naturalized sites than estimates derived from widely used USDA models. We also applied the DRAINMOD water balance model to compare drainage and runoff characteristics of the same soil for suburban lawn and adjacent mixed pine and hardwood forest on a site in Raleigh, North Carolina. For DRAINMOD inputs, lawn soil moisture and saturated hydraulic conductivity were estimated using the USDA model while the forest soil properties were estimated using the Teepe adjustments. Long-term continuous model results show that for the same soil, without accounting for any interception losses, the average annual runoff was 0.14 in. and 9.1 in. for the forest and lawn areas, respectively.

Little is known about the ultimate fate of petroleum hydrocarbons in bioretention areas or the factors that influence their fate, including vegetation choice. In this work, laboratory-scale bioretention cells were constructed inside sealed glass columns and spiked with ¹⁴C-naphthalene to permit an accurate accounting of naphthalene fate. Three columns were operated for approximately 5 months: an unplanted control column, a column planted with Blue Joint Grass, and a column planted with purple prairie clover (a legume). Naphthalene volatilization, leaching, biodegradation (mineralization), sorption, and plant uptake were determined. Adsorption to soil was the dominant naphthalene removal mechanism within the columns, although mineralization and vegetative uptake also were important. Contaminant volatilization was negligible and leaching of the contaminant was minor after some initial washout. Enrichment of the naphthalene degrader community (p<0.05) in the columns was measured using biodegradation batch experiments. The vegetated columns experienced enhanced enrichment compared to the unplanted columns (p<0.05). This research suggests that vegetation not only provides enhanced aesthetic appeal to bioretention cells but also measurable pollution control benefits.

Current bioretention designs are not very effective at nutrient retention for a variety of reasons. Negatively charged organic amendments compete for the few phosphorus (P) binding sites available in the sand matrix. The rapid rate of flow through the systems does not permit enough time for effective nitrogen (N) retention to occur. To improve bioretention designs for nutrient retention, it is necessary to understand the processes involved in nutrient transformations. Given this understanding, it is then possible to develop approaches that improve nutrient retentions. Our mesocosm studies have shown that advances in bioretention system design substantially improve retention of N and P from runoff. Using advanced media, our research over three years has confirmed our initial findings on P retention trends. After accumulating 190 g-m^-2 P (the equivalent of nearly five decades of stormwater runoff), the advanced media had a median TP discharge concentration of 0.077 mg/l. The median Ortho-phosphate discharge concentration was 0.015 mg/l. N retention by the outlet regulated treatment was 66% in even quite large (~6 month ARI) events, and 62% for oxides of N. The corresponding free discharge treatment was significantly less effective, at 27% and 19% retention, respectively. A year later, total N retention in small storm events exceeded 70%, while nitrate retention exceeded 90%. These results confirm how advanced media combined with outlet controls substantially raise the bar for treating runoff nutrients. This is become ever more important with the adoption of strict numeric criteria in TMDLs in affected waterbodies throughout the US.

The International Stormwater BMP Database (BMP Database) is a long-term project that began in 1994 through the vision of members active in the Urban Water Resources Research Council of the American Society of Civil Engineers (ASCE) and leadership of the U.S. Environmental Protection Agency (EPA). Funded for many years by EPA, the project is now supported by a coalition of partners including the Water Environment Research Foundation, Federal Highway Administration, American Public Works Association, as well as ongoing support from ASCE’s Environmental and Water Resources Institute. The project is a long-term, multifaceted effort that includes guidance on BMP monitoring (Geosyntec and WWE 2009a), standardized database reporting information, and recommended performance analysis protocols.

This paper covers the experience and lessons learned from the Ballard Roadside Raingarden, Phase 1 Project. This project involved installing bioretention facilities along eight blocks of city right-of-way for combined sewer overflow reduction goals in 2010. This paper details the design and construction experience and where Seattle Public Utilities (SPU) made missteps that resulted in the removal of 40% and retrofit of 50% of the constructed rain gardens.

The Selby Bay community in Edgewater Maryland in Anne Arundel County was evaluated for opportunities to apply Low Impact Development (LID) Stormwater Management practices and Environmental Site Design (ESD) techniques in an effort to reduce discharges and gain water quality improvements. The community has very narrow right of ways (40 ft), many existing utilities, flat slopes and a high water table. The methods described in this paper can be applied in many coastal Bay communities to improve drainage and water quality. A detailed hydrologic and hydraulic analysis of the study area was conducted with an innovative twist. In accordance with the provisions of the Maryland Stormwater Management Manual, the TR-55 model was used to calculate the flows using the Runoff Curve (CN) number method. However, the modified CN approach was used to model the benefits of providing storage through use of swales. In general, this method uses a reduced CN and an increased time of concentration (Tc) to account for the flow attenuation provided by the swales. Ultimately, the use of bioswales in the upper portions of the drainage area provides the following benefits: water quality treatment through infiltration and filtration processes, detention storage on-site both in the swales and in the biosoils that will be installed or enhanced under the swales, increased time of concentration by lagging the initial rainfall, reduced runoff curve numbers reflecting the infiltration and on-site storage, and a reduction of the downstream peak discharges allowing the use of smaller less expensive pipe systems to convey the runoff.

They were the best of alleys, they were the worst of alleys… the following paper tells the story of the City of Richmond’s Green Alley Program pilot projects. Design details, challenges, and lessons learned are included to help inform and improve the practice of permeable alley retrofits.

This study investigated the mass and concentration of nitrate in the runoff from black EPDM membrane (control) roofs, modular green roof units, and built-inplace (BIP) green roof models. Results for mass and concentration were similar. Results indicated that the mean nitrate mass in the runoff from the green roof systems was statistically higher than from the control roofs and that the mass from the planted BIP systems was statistically greater than from the unplanted systems. The mean nitrate mass in the runoff of the BIP models decreased as depth increased, requiring at least 15 cm to meet a 10 ppm limit. The mean mass from the 10-cm planted BIP models was statistically less than from the 10-cm modular units, apparently as a result of the drainage layer in the BIP system. The 5-cm planted BIP system was statistically similar to the modular system, even though the modular system provides better plant performance. Despite continued fertilization, the nitrate mass in the planted BIP models decreased over time, likely as a result of the degradation of the organic fraction of the growth media.

There are two primary approaches to green roof construction, those being termed tray systems and built-up systems. In general, tray systems use a series of plastic trays placed upon the roof like “tiles” while built up systems are comprised of rolls or large tiles of geotextiles which served to function as a drainage layer, root barrier and water retention layer. This paper provides a narrative comparison of the two different types of approaches based on common green roof design elements. These elements include water retention, weight, installation, etc.

The benefits of controlling stormwater runoff volume that results from land development activities have been well documented and are generally accepted by contemporary stormwater management practitioners. Although infiltration practices have been used for many years to mitigate the impacts associated with increased stormwater runoff, the benefits of more passive and non-structural approaches have only recently been recognized. Unfortunately, methods to quantify and assess those benefits have been limited, ranging from relatively simple empirical methods based on percentage of impervious cover to highly complex deterministic models which are beyond the needs of site-level analysis. In addition, the benefits from these so-called “green infrastructure” practices are generally associated with reductions in the annual runoff volume. Traditional stormwater management has relied on event-based methods to evaluate stormwater impacts and verify regulatory compliance. The Delaware Sediment & Stormwater Program has developed a methodology based on the Natural Resource Conservation Service’s Runoff Curve Number (RCN) to estimate the annual runoff from developing lands and runoff reduction benefits associated with Green Technology Best Management Practices (GTBMPs) used to comply with State stormwater regulations. A spreadsheet compliance tool has also been developed based on the methodology.

Low impact development (LID) design should fully consider local hydrological, climate, and soil characteristics. With the increasing popularity of LID and recognition of its contribution to sustainable development, many municipalities are developing or have developed LID design guide or manual. This paper discusses the processes, experiences and lessons learned from developing a LID-BMPs Design Guide for the City of Edmonton. This project had a series of workshops and roundtables to engage stakeholders to gain support and solicit feedbacks on the Design Guide. A modelling analysis was conducted to compare the LID-oriented neighborhood design with conventional neighborhood design regarding hydrologic performance, water quality improvement, life cycle cost, and benefits.

This paper describes a hydrologic and hydraulic (H&H) modeling technique to quantify the benefits of stormwater Low Impact Development (LID) controls in terms of reduction of Combined Sewer Overflow (CSO) events, volume, and peak overflow rate. This paper can help sustainable developers estimate the number and size of LID controls needed to capture a certain amount of CSO discharge in a typical year. A Pennsylvania case study is presented to illustrate H&H modeling results.

A numerical model has been developed to define the saturated seepage zone under Low Impact Development (LID) devices, such as an infiltration trench, rain garden, bioretention pond, and underground infiltration chamber. The numerical model is established on Microsoft Excel spreadsheets and executed with computer programs as embedded functions in the spreadsheets. The programs were written using Visual Basic. The model can show the graphical results of the seepage zone. The boundary element method is applied in the model. To determine the free surface of the saturated seepage zone, a few numerical iterations will be required in the model. The seepage flow rate in steady state can be computed with this model.

Transient storage plays an important role in pollutant assimilation in urban streams by allowing additional time for physical, chemical and biological transformation and uptake. While riparian forests provide many ecological benefits, few studies have examined the impact of riparian vegetation on the seasonal variability of transient storage. We investigated transient storage in four reaches of Dead Run, Baltimore, Maryland and four reaches of Valley Creek, Chester County, Pennsylvania. All reaches drained urban areas with impervious surface cover ranging from approximately 13%–28% in Valley Creek to 40%–60% in Dead Run. In each watershed, two reaches had a riparian forest which provided complete canopy cover over the stream channel and two reaches had an open canopy over at least 23% of the reach. Transient storage in both watersheds varied widely, accounting for as much as 25% of median travel time in Dead Run and tributary reaches of Valley Creek and as little as 0.01% of median travel time in the Valley Creek main stem reaches. In closed canopy reaches, there was no statistically significant seasonal trend in the median travel time attributable to transient storage (Fmed). However, in open canopy reaches, the time of year explained approximately 59% of the variation in Fmed, increasing from spring to summer. We conclude that the absence of canopy cover will induce a seasonal pattern in which transient storage is minimized in spring and maximized during the summer and early autumn growing season.

This paper summarizes the system engineering, installation, and start-up of an innovative stormwater management system using a “Gravity Drain Field” subsequent to a 2008 concept paper on the subject by Lolcama and Gauffreau in ASCE Geotechnical Special Publication 178. The stormwater management concept resulted from a collaboration of hydrogeologists and geotechnical engineers as engineered drainage for disposal of large stormwater volumes into epikarstic bedrock. The concept provides compliance with recent NPDES regulations in karst terrain by utilizing a series of low pressure, Class V injection wells (gravity drains). The gravity drain engineered recharge system replicates the movement of stormwater into the deeper epikarstic flow conduits that was occurring at a site prior to development, with the exception of isolating the stormwater from the soil mantle to prevent sinkhole formations, and flow capacity management. The concept has been developed into a working stormwater management system, and 19 Gravity Drains have been installed at a karst site located in the greater Philadelphia area. At this site, the method has the potential to recharge a volume of water greater than that of a 2-year storm event in a fully developed condition, which equates to 29.9 acre ft of stormwater, or 9.7 million gallons during each storm event. This paper discusses the selection of the gravity drain locations; the gravity drain and monitoring system layout; the hydrogeologic testing of the drains and well field array; the injection well and the drain field engineering specifications and installation; the automated remote hydrogeologic monitoring of the drain field including water table and subsurface temperature; and the stormwater pre-treatment and conveyance to the drain field.

Onondaga County, New York, in partnership with the City of Syracuse, has a substantial financial commitment to using green procedures to address its longstanding CSO issues. Its court ordered program shifts from building centralized industrial wastewater treatment structures and massive pipe storage systems to using decentralized green infrastructure (GI) approaches to hold, infiltrate and clean polluted stormwater runoff and reduce CSOs. Specifically, this order stemmed from an extended official negotiation process among the County, New York State Department of Environmental Conservation (DEC), Atlantic States Legal Foundation, Inc. (ASLF), and various stakeholders and community advocates, and it requires Onondaga County to apply green infrastructure approaches, as a complement to gray projects, to reduce CSO volume and meet water quality standards. This has put this community in the forefront nationwide for addressing stormwater and CSOs.

Mature trees contribute significantly to urban stormwater management and provide many other benefits. This presentation will explain the different processes by which trees provide stormwater management, as well as the magnitude of stormwater benefits possible with trees, supported by the latest research and case studies. A growing awareness of the multiple benefits of large trees has many cities developing initiatives to plant large numbers of trees. New York City, for example, aims to plant one million trees over the next decade. But without paying attention to how these trees are planted, the trees will never grow large enough to produce anywhere near the level of ecological services they are ultimately capable of providing or meet the expectations of those proposing the tree planting programs. Studies have found that trees surrounded by pavement in urban downtown centers only live for an average of 13 years, a mere fraction of their much longer lifespan under natural conditions. The most significant problem urban trees face is the inadequate volume of soil useable for root growth. Research has shown that trees need approximately 2 cubic feet of soil volume for every 1 square foot of canopy area. Most urban trees, confined to a 4 x 4 x 4 in. tree pit hole, have less than 1/10th the rooting volume they need to grow large. Additionally, the hunt for super-trees that can tolerate the stress of urban environments has resulted in low species diversity, rendering urban forests very susceptible to catastrophic losses such as those from Dutch Elm Disease and Emerald Ash Borer. This presentation will provide holistic policy, design, and management recommendations for how to create healthy, resilient urban forests that will allow trees to grow large enough to provide significant stormwater volume, quality and rate benefits.

With the issuance by the State of Maryland of its 2010 stormwater permit, Montgomery County (Montgomery) was required to retrofit 20% of its older, untreated or poorly treated impervious surfaces by 2015. In January 2012, the County released its final strategy for meeting this retrofit goal. The strategy includes using Environmental Site Design (ESD), or green infrastructure, for 18% of its retrofit obligations, with the bulk of the remainder to be achieved through stormwater pond retrofits. Unit costs for some innovative green retrofits are lower than others, and different mixes of green practices can be applied to different land cover categories. Using alternative mixes of these innovative green practices and independent local cost data, a back-of-the-envelope analysis indicates that it may be possible – and affordable – to apply green stormwater retrofit practices to more than half of the Anacostia Watershed’s targeted 1421 impervious acres in Montgomery County. A new unit cost metric, dollars per Acre-inch of runoff reduced, is introduced. Examination of an alternative green retrofit scenario for Montgomery’s Anacostia watershed area suggests that this approach merits further in-depth consideration, both for Montgomery County and for other stormwater permittees facing similar imperviousness restoration mandates.

A variety of trail and channel stabilization and runoff reduction techniques were implemented within Philadelphia’s Wissahickon Valley Park in Philadelphia. In addition to “conventional” bioinfiltration systems (not discussed here), the methods employed and described in some detail below include (1) the replacement of actively enlarging gullies below stormwater outfalls with stable “stepped infiltration swales”; (2) the obliteration and hydrologic restoration of steep, improperly routed trails; and (3) the installation of trailside microtopographic basins, “contour soakage trenches”, designed to accept and infiltrate runoff from active trails and unpaved roadways. These best practices have wide application within semi-natural urban nature parks generally, many of which are seriously impaired hydrologically because of land use history, intensive human use today, and ongoing urban development along their periphery.

This paper presents the process used to identify water quality issues in an impaired watershed and the design and construction of stormwater basin retrofits to address the impairment issue. A 2008 Watershed plan for Northfield Brook in Connecticut identified several water quality issues which lead to the impairment of the water quality in Northfield Brook Lake. The lake, which is under the jurisdiction of the Army Corps of Engineers, is closed many times during the year for recreational swimming uses due to excessive Escherichia coli (E.coli) concentrations. The E.coli concentrations routinely exceed the US EPA standard of the geometric mean of 5 samples over a 30 day period being less than 125 CFU/100 ml. While a direct source of the E.coli was not able to be determined, non-point source runoff from a medium density residential subdivision in Thomaston, CT was considered to be a potential source. The subdivision is served by two detention basins which are non-functioning at the current time. Due to the lack of detention, the downstream channel has experienced significant gully erosion with the result of over 1,000 cubic yards of sediment being deposited into Northfield Brook Lake. The Northwest Conservation District (NCD) applied for and received Section 319 funding from the CT Department of Environmental Protection (DEP) to address this impairment issue by hiring a consultant to design retrofits for the failing Highwood Estates stormwater basins. The goal of the retrofits was to increase the removal of coarse and fine sediments, on which E.coli is commonly attached, and to reduce peak rates of runoff.