Stormwater runoff may contain high levels of pollutants and is regulated by the Federal National Pollution Discharge Elimination System (NPDES). Stormwater bioretention (BR) systems are often used to satisfy these regulations. BR systems collect accumulated runoff that leaches into groundwater. A greenhouse study evaluated nutrient and metal removal among plant species that are typically found growing in BR systems. A field demonstration study assessed citric acid enhanced metal bioaccumulation potential under typical BR system conditions. The greenhouse experiment examined pollutant retention, and bioaccumulation potential for six plant species undergoing three hydraulic and pollutant loads. Results verified there was 98% recovery of total phosphorous over the study period. Biomass increased with higher hydraulic and pollutant loads for all species. Phragmites australis, Carex praegracilis, and Carex microptera took up significantly more total phosphorous and nitrogen mass into shoots than Typha latifolia, Scirpus valid us, and Scirpus acutus. This study also found that 89% of applied metals were removed within the top 27 em of soil in all treatments. Similar results were found regarding copper, lead, and zinc concentrations and bioaccumulation. Carex praegracilis, and Carex microptera exhibited higher metal distribution in plant tissue and exfiltrate, and lower distribution in the soil media than the other species. This indicated species differences in biological and chemical processes taking place within the simulated BRsystems. The field experiment investigated citric acid enhanced metal bioaccumulation potential among three different plant species under representative BR conditions. Citric acid significantly increased metal concentrations in the soil pore water for the planted treatments, but this did not result in increased metal uptake into plant tissue. However, notable differences were found among species, where Carex microptera accumulated more AI, Cr, Cu, and Fe in the above ground tissue than Helianthus maximiliani and Typha /atifolia (except for Cu in Helianthus). These results provide greater insight into the biological and chemical process that affect transport, uptake and translocation of nutrients and metals, and confirm the importance of species selection in BR systems to optimize nutrient and metal retention and recovery from stormwater runoff to minimize subsequent groundwater pollutant loading.

The introduction of pavement, buildings, and other impervious surfaces to urban landscapes greatly influences the quantity and quality of urban stormwater runoff. In this study, we designed and implemented modern stormwater monitoring technologies to establish a “smart” stormwater sensor network within the Northwest Field Canal (NWFC), an urban water conveyance located in Logan, Utah, USA. This network was designed to collect flow and water quality data at high frequencies and simultaneously at multiple locations. The observatory’s innovative method of inter-site communication and changing sampling frequencies during storm events was able to capture short duration events at the upstream and downstream ends of the NWFC and at multiple outfalls in the canal simultaneously without human intervention. We then investigated statistical regression models between turbidity and TSS so as to predict TSS at high frequencies. Finally, the addition of the high-frequency discharge data in the calibration procedure for a stormwater simulation model developed using the Environmental Protection Agency’s Stormwater Management Model did little to improve model performance at the downstream end of the canal, but did provide important insight into the overall contribution of discharge from individual stormwater outfalls to the NWFC. The results of this study inform water professionals on how to build and operate automated monitoring systems and how to create high-frequency estimates of TSS and TP loads in urban water systems.

The combined effect of land-use alterations and introduced diffuse anthropogenic pollutants to the earth's surface in suburban/urban zones often sparks a decrease in stormwater quality in the area, and contributes to nonpoint source pollution in receiving waters. The ponds at the Utah Botanical Center (UBC) located in Kaysville, UT, regularly experience algal blooms, which in turn cause low dissolved oxygen levels in the waters, indicating high concentrations of inflowing pollutants. The goal of this thesis paper was to describe the findings from the water quality monitoring implemented at both the inlet and outlet of the UBC ponds in order to assess pollutant loading to the ponds. A survey was mailed to the homeowners in the drainage area with the intention of gaining a baseline understanding of residents' perceptions of stormwater issues, and their lawn care practices that might influence stormwater quality. Results from the weekly monitoring found that the TN, TP and TSS levels were all below respective medians reported for urban areas around the United States. Baseflow separation calculations revealed that 47% of inflow was due to precipitation falling onto the watershed and therefore 53% of inflow was a product of non-stormflow. With only 47% of the inflow coming from local runoff, potential effectiveness of educational efforts was considered minimal. Survey results reported that 86% of respondents had never received educational materials regarding stormwater. Second, fertilizer is used by 92.3% of respondents and in most cases, homeowners perform more than one application per year. Of the respondents, 98.1% of them believed that individual residents had an impact (positive or negative) on the quality of water resources in the area. No significant association was found between the education component of the survey and whether participants undertook certain stormwater-related behaviors.

Water quality continues to be threatened by human development activities such as stormwater runoff from urbanization. This study addresses the question of how stormwater Best Management Practice (BMP) system design choices affect pollutant removal efficiency, through the examination of 12 case study sites (across five states) that use three common BMP system design types (detention, retention, and wetland channel). Water quality information was obtained from the International Stormwater Database and site composite grab samples. Development conditions were inventoried by orthophotos and assessed using Geographic Information System (GIS) and AutoCAD data. This study relates Total Suspended Solid (TSS) load removal efficiency with a series of form-based design variables, which landscape architects can control through their scope of services. System design characteristics of age, size, form, and material selection, along with site-specific precipitation regimens and extent-of-development, were compared with use of TSS removal efficiency in order to determine possible design relationships. Primary removal efficiency methods of individual design types were then evaluated to prioritize conclusions for practical applications. Mixed findings from this study report that one material-based and two form-based design variables hold significant correlation with the TSS removal efficiency. However, site-contextual variables, especially the Natural Resource Conservation Service (NRCS) hydrologic weighted curve value, play an important role in BMP performance and may supersede possible design impacts. The study also reveals that for some BMP sites, a deviation from original design intent may jeopardize system removal efficiency.

abstract: Imagine you live in a place without any storm water or wastewater systems! Wastewater and storm water systems are two of the most crucial systems for urban infrastructure. Water resources have become more limited and expensive in arid and semi-arid regions. According to the fourth World Water Development Report, over 80% of global wastewater is released into the environment without adequate treatment. Wastewater collection and treatment systems in the Kingdom of Saudi Arabia (KSA) covers about 49% of urban areas; about 25% of treated wastewater is used for landscape and crop irrigation (Ministry of Environment Water and Agriculture [MEWA], 2017). According to Guizani (2016), during each event of flooding, there are fatalities. In 2009, the most deadly flood occurred in Jeddah, KSA within more than 160 lives lost. As a consequence, KSA has set a goal to provide 100% sewage collection and treatment services to every city with a population above 5000 by 2025, where all treated wastewater will be used. This research explores several optimization models of planning and designing collection systems, such as regional wastewater and stormwater systems, in order to understand and overcome major performance-related disadvantages and high capital costs. The first model (M-1) was developed for planning regional wastewater system, considering minimum costs of location, type, and size sewer network and wastewater treatment plants (WWTPs). The second model (M-2) was developed for designing a regional wastewater system, considering minimum hydraulic design costs, such as pump stations, commercial diameters, excavation costs, and WWTPs. Both models were applied to the Jizan region, KSA. The third model (M-3) was developed to solve layout and pipe design for storm water systems simultaneously. This model was applied to four different case scenarios, using two approaches for commercial diameters. The fourth model (M-4) was developed to solve the optimum pipe design of a storm sewer system for given layouts. However, M-4 was applied to a storm sewer network published in the literature. M-1, M-2, and M-3 were developed in the general algebraic modeling system (GAMS) program, which was formulated as a mixed integer nonlinear programming (MINLP) solver, while M-4 was formulated as a nonlinear programming (NLP) procedure. / Dissertation/Thesis / Doctoral Dissertation Civil, Environmental and Sustainable Engineering 2019

Nutrients are introduced into the waters of the Commonwealth of Pennsylvania through various methods, mainly through point and non-point sources. Non-point sources would include agricultural means however, there are often other non-point sources that are often overlooked. Water runoff from stormwater is an important contributor to pollution entering various watersheds. Data was collected at 4 locations (the initial point of reference, residential, urban, and industrial) on Mill Run and 2 locations on Woodcock Creek, through surface water collection methods in dry and wet weather. Nitrate, Nitrite, Total Phosphorus, and Total Nitrogen were among the data collected. Total Phosphorus had an increasing trend in concentration as the study progressed. Nitrate, Nitrite and Total Nitrogen showed no increasing trends. Location 3 (Urban Zone) produced the largest increase in Total Phosphorus. Weather conditions played a role in higher Nitrate concentrations while warm water produced higher Nitrite concentrations

Monitoring of a regional stormwater management facility, located on the Virginia Tech campus in Blacksburg VA, was conducted in order to assess its efficacy in reducing nonpoint source pollutant losses downstream. The facility design includes both an upper water quality (wet) pond and a lower 100-yr-event quantity (dry) pond. These on-stream ponds capture both baseflow and storm runoff from the southern portion of the Virginia Tech campus and surrounding lands, and release the water back to the unnamed stream shortly above its conjunction with Stroubles Creek, a tributary of the New River. Monitoring sites for flow measurement, water quality sampling, and biotic assessments (habitat evaluation and rapid bioassessment of benthic macroinvertebrates) were located above and below each of the ponds. Both grab samples and automated samples were collected at these stations. Between 1997 and 1999, water quality grab samples included 35 baseflow samples and 22 stormflow samples. The grab samples were analyzed for concentrations of total suspended solids (TSS), metals, bacteria, and nutrients as well as temperature, pH, dissolved oxygen, conductivity, total organic carbon (TOC), and chemical oxygen demand (COD). Automated flow-weighted sampling was initiated in February of 1999 and results are reported through the end of October 1999. Thirty-three storms in 1999 were monitored for flow and various water quality parameters (TSS, TOC, COD, and nutrients). Pollutant loads and pollutant removal estimates were calculated with regard to the wet pond, dry pond, and the combined facility. Two types of pollutant removal efficiencies were calculated: (1) the EMC efficiency, based on pollutant concentrations from individual storms; and (2) the SOL efficiency, based on pollutant loads, to estimate long-term performance over the study period. Benthic macroinvertebrate sampling and habitat assessment were performed in both 1997 and 1999. In addition, a preliminary investigation of pond characteristics was conducted, including measurements of water quality and composition, sediment deposition and composition, and residence time. As a system, the stormwater management facility appears to have minimum impact on improving the downstream water quality. Pollutant concentrations and loads both appear to increase downstream of the facility as compared to upstream, during both storm event and baseflow periods. Monitoring results of the benthic assemblages showed evidence of moderate to high impairment at all sampling locations, and habitat assessments showed evidence of high sedimentation levels within the stream, even after installation of the stormwater management facility. Total suspended solids (TSS) concentration removal efficiency was 10% for the combined wet pond and dry pond system, much lower than the 80 to 90% TSS removal expected for properly functioning stormwater management facilities (Hartigan, 1989). There is some evidence of sedimentation within the ponds because of a slight reduction in sediment-bound constituent export, but the dissolved nutrient constituents had either very low and most often negative (indicating pollutant export) removal efficiencies. Concentrations of metals measured in the stream often exceeded their respective acute and chronic water quality criteria at all sampling locations. Pollutant removal efficiencies measured in the wet pond are atypical of those reported in the literature (Schueler, 1993). Insufficient residence time (two days compared to the optimal two weeks), and wet pond embankment failure are likely the principal causes of the wet pond's inadequate performance and thus, the inadequate performance of the overall facility. TSS removal efficiencies were low in the wet pond (19% for concentrations and 33% for loads) compared to the 80 to 90% expected for similar ponds. Nevertheless, the wet pond reduced the concentrations of several pollutants typically associated with TSS and not likely to be associated with the fill material for the wet pond embankment. Zinc concentrations in sediment cores were highest near the pond inlet, where the majority of sedimentation occurs. During storm events, the following results were noted. Copper and zinc concentrations in 1998 were lower at the pond outlet as compared to the pond inlet, and TOC concentrations and loads were also reduced by the wet pond (13% for concentrations and 12% for loads). However, sedimentation is also expected to remove phosphorusl, and wet pond phosphorus loads were only reduced by 10% and 3% for orthophosphorus and total phosphorus, respectively. Because the wet pond is undersized with respect to the watershed it serves (surface area less than 1% of the watershed area (0.87 ha), as compared to the 3% ratio often recommended for optimal pollutant removal (Athanas, 1988)), higher removal efficiencies were found during baseflow periods. The greatest reductions in baseflow concentrations were for ammonia (67%), nitrate (57%), total nitrogen (54%), and COD (45%). However, the residence time of two days appears to be insufficient to reduce fecal coliform concentrations in the stream, and over 40% of the fecal coliform samples collected exceeded the water quality standard for contact recreation (DEQ-WQS, 1997). Furthermore, the wet pond did not appear to reduce TSS or TOC during baseflow periods. Export of TSS (-29% EMC efficiency) and TOC (-44% EMC efficiency) from the wet pond during baseflow periods is likely due to the wet pond embankment failure as well as pond eutrophication. Eutrophication processes are favored by the water temperature increase as flow passes through the shallow wet pond. The wet pond increased downstream temperatures by approximately 8°C above inflow temperatures during the summer, and to levels above 21°C which cannot be tolerated by sensitive coldwater species (Schueler, 1987). The dry pond did not remove dissolved nutrient constituents or other pollutants during baseflow periods, but there is some evidence of sedimentation within the dry pond during storm events. During storm events, the dry pond was effective in removing TSS, with a concentration removal efficiency of 69% (EMC efficiency) and loading removal efficiency of 43% (SOL Efficiency). Removal of TKN and total phosphorus (36% and 37% respectively for concentrations) within the dry pond is further evidence of sedimentation within the dry pond. The wet pond embankment was built in 1997, and monitoring occurred during a potential stabilization period when evidence of water quality benefits are slow to appear, especially with respect to downstream habitat and aquatic communities. Some benefits which could have been observed more immediately may have been negated or masked by the progressive erosion of the wet pond embankment as a result of a design flaw. Further complicating the results is the appearance; based on observations of extended drawdown time and results from a water budget analysis in the wet pond (where inflow substantially exceeds inflow); that groundwater interacts with the pond in a complicated fashion, possibly including both recharge and discharge. To fully understand the impact of the stormwater management facility on the water quantity and quality within this tributary of Stroubles Creek, monitoring efforts should continue after the wet pond embankment is repaired and is fully operational. If biotic community improvement is desired, the stabilization period could be defined by the time necessary to flush out accumulated sediment within the channel. Monitoring efforts should also expand to include the investigation of the groundwater regime and water level fluctuations within the wet pond. Further measurements of pollutant removal processes and influences upon those processes within the wet pond should also be considered. Last, the influence of the stormwater management facility on downstream flow regimes should be investigated to assess the adequacy of its performance with regard to flow control and prevention of stream channel degradation. / Master of Science

No description available.

No description available.

Due to the large amount of impervious surface cover, urban areas are at high risk for flooding and, in cities with combined sewer systems, subject to sewer overflow during heavy storm events. The Pennsylvania Department of Transportation (PennDot) is currently reconstructing and expanding parts of Interstate 95 (I-95) through the city of Philadelphia. Due to both federal and local laws, PennDOT must account for the stormwater runoff and minimize outflow to the sewer system. To do so, PennDOT has plans to construct a series of stormwater management practices (SMPs) adjacent to I-95 to control the volumes of highway runoff. In partnership with Villanova University, Temple University has been tasked with monitoring these SMPs, known as bioswales, to provide insight and guidance as the project moves forward and to ensure mistakes aren’t reproduced in future construction. This research is contributing to the overall project goals by testing the application of geophysical monitoring to one of the bioswales known as SMP A. Unlike commonly used point measurements, geophysical surveys are non-invasive and provide extensive spatial coverage. Specifically, this research involves the use of electrical resistivity tomography (ERT), in which a series of cable-connected electrodes are placed in the ground and measure electric potential differences when an electric current is applied. Once processed, the results are a contoured subsurface image of the distribution of electrical resistivity (the inverse of electrical conductivity). If multiple surveys are taken over time, the data can be differenced, known as time lapse inversion, to quantify changes in electrical resistivity. ERT is a favorable for these SMPs as survey results are sensitive to changes in soil moisture and fluid conductivity, which are essential parameters when tracking infiltration and road salt influx at these SMPs. Additionally, the ERT data can be converted to soil moisture values using Archie’s law, which is important for determining soil moisture at points where no sensors are currently placed. We built and installed three ERT survey lines connected to an on-site monitoring station in April 2019 and collected quasi-daily measurements until monitoring seized in November 2021. One way to test SMPs is through a simulated runoff test, in which an SMP is flooded with water from an external source and the SMP’s response is recorded. During September 2020, Villanova University performed an SRT at SMP A, while we performed ERT surveys before, during, and after the SRT to track the infiltration and dry-out cycle. Knowing how long the soil at an SMP takes to recover to pre-storm soil moisture levels is essential in understanding an SMP’s performance and functionality. We were successfully able to capture the wet-up associated with the SRT and the corresponding dry-out period with the ERT data, which showed around a 20% decrease in resistivity when soil sensors indicated saturation. This resistivity change began to decrease and finally reached pre-SRT levels (0 – 5% change) after 68 hours, leading to our estimate of a three day recovery time for SMP A. Interestingly, inflow/outflow measurements at SMP A showed that only 24% of the input water exited the SMP via the overflow drain, meaning the rest of the water remained in the SMP. This discrepancy was solved with our ERT data, which showed that the decrease in resistivity, and therefore increase in soil moisture, was seen at depths beyond the 0.60 m layer of amended fill the SMP contained. Overall, the water was infiltrating past this layer and into the urban soil below. Initially it was thought that the native urban soil would impede infiltration, hence SMP A was designed around this assumption. However, our geophysical results indicate that the native urban soil underlying the SMP has an infiltration rate of 10 cm/hr and is contributing to the overall function of the SMP. This was unknown as previous monitoring was focused on the layer of amended fill material, not the underlying native soil. The relationship between electrical resistivity and soil moisture, fluid conductivity, and porosity is known as Archie’s law, who derived an empirical formula that allows electrical resistivity data to be converted to soil moisture values. However, this equation requires quantifying two parameters, m (also known as the cementation factor) and n, the saturation exponent. Researchers commonly use pre-published values for m and n, or establish site-specific values by fitting Archie’s law to a set of soil moisture and conductivity data. However, as soil is heterogeneous, one set of m and n values may not be accurate across an entire site, especially with the presence of hysteresis, where one soil moisture value can correspond to multiple conductivity values depending on whether the soil is experiencing imbibition or drainage. Additionally, m and n can change over time as soil fabric changes, as well as soil conductivity changes due to the influx of road salt during winter months. In December 2019, we finished installing 16 TEROS12 soil sensors at SMP A, which recorded soil volumetric water content (VWC) and bulk electrical conductivity (bulk EC) every five minutes for nearly two years. These sensors were at six different locations within SMP A at depths of either 0.10 m, 0.30 m, or 0.60 m. We selected 13 storm events and fit Archie’s law to the soil VWC and bulk EC data to get values for m and n. While we were able to find m and n for all events, including events that exhibited hysteresis in soil VWC and bulk EC, each sensor had a different pair of m and n values. This discrepancy was surprising, given that the soil at SMP is a homogeneous, sandy-loam fill with no more than 10% clay. However, even sensors at the same depth show statistically significant differences. We also found that m and n were changing over time, notably m was increasing over time, possibly due to porosity changes. This result indicates that multiple sensors are needed to accurately calculate m and n, even at sites with relatively homogeneous soil. Most notably, the reason why we had success in fitting Archie’s law for every sensor was due to our accounting for changes in porewater conductivity. Most researchers assume a constant value for porewater (fluid) conductivity in Archie’s law. However, we found that not accounting for porewater conductivity changes lead to severe misestimation of soil VWC, even getting physically impossible values (VWC > 1.0 m3/m3) in some cases. Therefore, accounting for changes in porewater conductivity is essential when using Archie’s law. Road salt transport in SMPs is a concern, especially in Philadelphia, which is subject to winter storms and freezing conditions. In some PennDOT SMPs, the presence of road salt in the soil during leaf-out has been suspected to be the cause of stunted plant growth and pre-mature plant mortality. Vegetation is an important aspect of the SMPs, as they provide evapotranspiration pathways, aesthetics, and soil erosion control. Thus, vegetation impairment affects SMP functionality, and plants often need to be replaced, increasing maintenance costs. To track and assess the spatial distribution of road salt, we performed ERT surveys along three lines, with two lines in the topographically lower portion of the SMP, or flood zone, and the other line on the elevated bank parallel to the other lines. All three of these lines had vegetation. In total, we collected 900 ERT surveys from October 2020 to September 2021, sufficiently covering the winter months and growing season. During February 2021, the soil sensors indicated significant increases in conductivity, with sensors ranging from 5.0 – 20.0 mS/cm, compared to pre-winter values of 0.1 – 0.6 mS/cm. The winter ERT surveys show the formation of a shallow conductive (< 10 Ω) layer in the top 0.25 m of soil, and an overall decrease in resistivity of up to 70%. This change decreased over the spring and summer months, indicating that dilute runoff was flushing the salt through the soil column. However, flood-zone ERT data still showed a 20% decrease in resistivity in June when compared to pre-winter data, indicating that road lingered in the soil during the spring and summer months. In May, we began taking bimonthly measurements of plant height, width, and leaf chlorophyll content (SPAD) on plants along the ERT lines, then in July took leaf tissue, root tissue, and root-zone soil samples and analyzed them for sodium content. We found that the plants along Lines 2 and 3 (flood-zone) had statistically significant stunted growth when compared to the plants along the elevated bank, as well as elevated sodium levels (> 400 mg/kg) in root tissue. No detectable sodium was found in leaf tissue samples. The stunted growth and elevated root sodium in the flood-zone plants indicate that early spring storms are not enough to flush out the road salt, and therefore artificial flooding may be required before leaf-out to ensure plant survival. We also suggest planting salt-tolerant plant species in areas of SMPs prone to flooding, such as the topographically lower portions. ERT can also be used to guide the placement of these plant species, as ERT can delineate areas of higher conductivity. / Geoscience