Journal of Sustainable Water in the Built Environment

Abstract

Combined sewer overflow (CSO) is one of the major water pollution problems faced by many municipalities that are on combined sewer systems. Green infrastructures (GIs) can help mitigate urban stormwater problems including CSOs. A field-based study over 3 years investigated the effectiveness of rain gardens and planter boxes as GIs for stormwater control in residential urban areas. The objectives of this study were to (1) investigate the impact of field-scale GIs on stormwater volume reduction in combined sewers at paired sites in an urban area, and (2) develop a methodology, including the application of a series of statistical methods, for analyzing field-based data and addressing the variations of field-collected data. The paired sites were both located in the City of St. Louis, Missouri. The test site had 12 rain gardens and six planter boxes, whereas the control site had no GIs. The stormwater was separated from the combined flows measured in sewers using the antecedent subsequent dry-days estimation method and were then normalized by drainage area and rainfall amount. The statistical methods deployed for this study included the Anderson Darling normality test, the Mann-Whitney U-test, and the Monte Carlo randomization test. This study revealed that, based on the normalized stormwater volume in the sewers from all of acceptable rainfall events, the means were ${0.409\pm 0.356\mathrm{m}}^3/{\mathrm{m}}^2\text{-}\mathrm{m}$ ($0.255\pm 0.222\mathrm{gal.}/\mathrm{sqft}\text{-}\mathrm{in.}$) at the test site and ${0.704\pm 0.979\mathrm{m}}^3/{\mathrm{m}}^2\text{-}\mathrm{m}$ ($0.439\pm 0.610\mathrm{gal.}/\mathrm{sqft}\text{-}\mathrm{in.}$) at the control site, revealing a 42% reduction of stormwater; the medians were ${0.321\mathrm{m}}^3/{\mathrm{m}}^2\text{-}\mathrm{m}$ ($0.200\mathrm{gal.}/\mathrm{sqft}\text{-}\mathrm{in.}$) at the test site and ${0.428\mathrm{m}}^3/{\mathrm{m}}^2\text{-}\mathrm{m}$ ($0.267\mathrm{gal.}/\mathrm{sqft}\text{-}\mathrm{in.}$) at the control site, revealing a 25% reduction of stormwater. For small rainfall events, the reduction was higher, at 62%. The methodology developed in this study can be used for other field-based studies. The findings provide needed information for the application of GIs to manage stormwater to help reduce CSOs.

Practical Applications

Green infrastructure provides numerous benefits for managing stormwater. However, field-scale studies that use continuously operating sewer systems to investigate their effect on stormwater in combined sewers are limited due to many hard-to-control factors, such as complex hydraulic conditions and flow variations in sewers and limitations in flow monitoring devices. This study developed a methodology to analyze flow data in combined sewers to quantify the effect of green infrastructure in reducing flow and, thus, combined sewer overflows. Stormwater volumes at a test site and a control site were separated from the measured combined flows using the antecedent subsequent dry-days estimation method and were normalized by the drainage area and rainfall amounts of the site. The results showed that residential rain gardens and planter boxes are a tool to help meet requirements to reduce combined sewer overflows. They reduced runoff within the sewer by 25% for the median of the measurements over the 3-year study and reduced runoff from small rainfall events by 62%.

Green stormwater infrastructure (GSI) can provide multiple benefits in addition to stormwater management. However, there is a need to improve GSI siting to ensure these benefits are realized. We present a planning algorithm that hones in on ‘sweet spots’ of GSI implementation that are hydrologically optimal, feasible, and provide more equitable access to the benefits of GSI. We apply this approach in Lancaster, a city in Pennsylvania, US, with multiple stormwater-related challenges. To identify sweet spots, we first leveraged available spatial data to derive maps of five key criteria, including hydrology, vegetation, property ownership, sewer system type, and social vulnerability. We then normalized each layer and combined them using two different weighting schemes, including an ‘Even Weights’ and a ‘People’s Choice’ scenario based on a choice experiment embedded in a community survey. The survey indicated a preference for prioritizing the hydrology and sewer system criteria. Sweet spots for GSI implementation under each scenario were mapped based on the 90th percentile of the final combined key criteria layers. Comparisons between the two weighting schemes indicated a 73% overlap in sweet spot locations. We also found a small percentage (16%) of existing GSI in Lancaster overlapped with the sweet spots, indicating an opportunity to target future GSI implementation in the remaining sweet spots. Despite being demonstrated in a specific city, this relatively simple approach leveraging widely available spatial data can be applied and customized elsewhere and help improve future GSI siting methods.

Traditional (piped) drainage systems are designed to drain quickly. It is therefore reasonable to design a system assuming that it will be empty at the start of the rainfall event. However, the restoration of full capacity in sustainable drainage systems (SuDS) or low impact development is dependent on slower processes (e.g., evapotranspiration, infiltration, or the consumption of harvested rainwater). Current design guidance often does not advise on reasonable assumptions to make regarding SuDS retention capacity. In addition, SuDS have the capacity to control both runoff volumes and flow rates during both routine and extreme storm events. This presents two further interlinked challenges: first, to identify relevant metrics to define SuDS performance; and, second, to define appropriate performance criteria for system design. Using rainwater harvesting as an example, it is argued that continuous simulation supports the calculation of a full range of performance metrics, properly accounting for retention, and empowering users to set design targets that are appropriate for a desired level of protection. Six independent metrics are considered to characterize performance in response to both routine and extreme rainfall events, and a scatter pie plot is introduced as a clear visual indicator of system performance across multiple targets. While current UK guidance for SuDS prioritizes flood risk mitigation and aims to provide protection up to the 1 in 100 year event, it is argued that such stringent expectations may be acting as a deterrent to SuDS uptake, particularly at the domestic scale. Here, lower design thresholds (for household SuDS) in the region of the 1 in 2 year event and 95% of annual runoff are recommended.