Grantee Research Project Results
Final Report: Evaluating Biofiltration in Small Urban Areas: Chico, California Case Study
EPA Grant Number: SV836930Title: Evaluating Biofiltration in Small Urban Areas: Chico, California Case Study
Investigators:
Institution:
EPA Project Officer:
Phase: II
Project Period: October 1, 2016 through September 30, 2018 (Extended to September 30, 2019)
Project Amount: $74,971
RFA: P3 Awards: A National Student Design Competition for Sustainability Focusing on People, Prosperity and the Planet - Phase 2 (2016) Recipients Lists
Research Category: P3 Awards , P3 Challenge Area - Safe and Sustainable Water Resources , Sustainable and Healthy Communities
Objective:
Impervious surfaces directly alter the hydrologic cycle, as a greater percentage of rainfall becomes stormwater runoff instead of infiltrating the ground surface. This increases the quantity and magnitude of stormwater runoff, which directly impacts geomorphic and ecological processes in urban watersheds with higher risk of flooding and scour potential in creeks and rivers (Hawley and Bledsoe, 2011). In addition to sediment, storm runoff picks up pollutants such as heavy metals, petroleum products (oil and grease), pesticides, and fertilizers, diminishing downstream water quality. As a result, stormwater discharges originating from Municipal Separate Storm Sewer Systems (MS4s) are regulated as point sources of pollution by the 1987 Amendments of the Clean Water Act (CASQA, 2009). While permitting for MS4s serving large urban centers (population greater than 100,000) has been in place since the 1990s, MS4s in smaller municipalities have only been included in the permitting process recently and are in need of cost-effective solutions to manage storm runoff (CASQA, 2009).
Biofiltration systems are a type of structural stormwater best management practice (BMP) used to reduce the quantity and improve the quality of urban storm runoff in a cost-effective manner. These vegetated filtration systems are designed to slow down velocity, reduce volume, and retain pollutants of stormwater runoff using a combination of plants and filter media such as sand, gravel, compost, and soil (Guo, 2009; Lim et al., 2015). They enhance natural processes such as sedimentation, filtration, sorption, plant and microbial uptake, and biodegradation (EPA, 1999). Biofilters are increasingly popular green infrastructure (GI) management solutions to urban stormwater runoff, due to their flexibility in terms of size, location, configuration, and appearance (Guo, 2009). For example, biofilters can be grass filter strips along roadsides intercepting sheet runoff, bioswales capturing concentrated runoff in outlet drains in broad open channels, or even be incorporated as low-footprint street-trees. Despite the rapid implementation of biofilters (Dietz and Clausen, 2006), pollutant removal performance is not often assessed after construction and long-term monitoring generally does not exist. Without monitoring data, it is difficult to determine the effectiveness of biofiltration systems and how to improve their design.
The overall goals of this research project were to a) improve water quality in small urban cities and campuses and b) increase awareness about urban green water infrastructure by identifying the practicality of using biofiltration in small urban areas, which requires identifying preferred configurations, evaluating performance of filter media to remove pollutants, performing a cost-benefit analysis, and disseminating results to the local community using K-12 outreach and meetings with local officials. This project aims to remove pollutants from runoff that could end up in streams and rivers; it therefore aligns with the mission of Section 104 of the Clean Water Act focused on reducing water pollution. This research embodies the principles of sustainability and seeks solutions that protect the environment and strengthen urban communities.
The project objectives were to fill in a knowledge gap in biofiltration systems in small urban areas by 1) identifying key design parameters to achieve biofilter resiliency and 2) carrying out detailed long-term performance monitoring of existing biofilters. The innovative aspect of this project lies in its focus on testing pollutant removal efficiency in biofilters because the monitoring of current systems is mainly limited to hydrologic performance. Phase II research was conducted both in the laboratory and in the field. Long-term (~18-month) laboratory experiments extended investigations from Phase I in order to identify optimal plant-media combinations and design configurations to achieve resilient biofilters. In-depth monitoring of biofiltration systems on college campuses in the Chico area (CSU Chico and Butte College) assessed long-term hydrologic and pollutant removal performance of existing biofilters. In addition, a life cycle analysis (LCA) of biofiltration designs was conducted. Increased public awareness about urban green water infrastructure was achieved through regular education outreach events in the local community and numerous presentations at scientific conferences.
Summary/Accomplishments (Outputs/Outcomes):
The overall project goals, to 1) identify key design parameters to achieve biofilter resiliency and 2) to carry out detailed long-term performance monitoring of existing biofilters, were achieved. Goal 1 was accomplished by conducting column experiments both in the laboratory and in the greenhouse. For Goal 2, 17 storm events were sampled between Oct. 2016 and Sept. 2019 with increasing frequency and detail over time. More stormwater samples were collected than expected (127 samples), with the sampling of three storms in 2016/17 (16 samples), six storms in 2017/18 (60 individual samples), and eight storms in 2018/19 (51 samples). During specific storms, samples were collected throughout the storm to capture the variability of stormwater composition and biofiltration performance. Students and PI Matiasek disseminated Phase I and Phase II results at more conferences than planned with 40 oral and poster presentations at 20 international, national, or regional professional meetings including the American Geophysical Union 2019 and 2016 Fall Meetings, the Association for the Sciences of Limnology and Oceanography 2019 Aquatic Sciences Meeting, the 2019 Annual Meeting of the Association for Environmental Studies and Sciences, and the 2018 Summer Meeting of the Association for the Sciences of Limnology and Oceanography (see complete list in section 7).
Outreach activities were implemented at the CSU Chico Hands-On Lab (Nov. 2017), at three local K-12 schools (2018), during the CIELO Program Summer Institute for K-2 teachers focused on sustainability in STEM (Jun. 2018 and Jun. 2019, CSU Chico), and at the Hearthstone School in Oroville, CA (2018-19). PI Matiasek met with architects, engineers, and CSU Chico Facilities Management & Services during the new Science building design phase to consult on stormwater management (Jun. 15, 2017). The EPA P3 student team conducted soil grain size analysis with construction site soil to assess the feasibility for stormwater bioretention basins (Oct. 2018). A test plot was built in Jan. 2019 to further investigate the performance of the local campus soil. This project resulted in convincing landscape architects to use local soil as the main substrate for the stormwater GI features of the new Science building instead of an imported soil mix. In addition, small design modifications were approved to enable flow measurements (weirs) and soil water sampling (addition of perforated drain pipe) in two bioswales and three flow-through planters. The P3 project findings directly contributed to these design amendments. Project findings were also presented to CSU Chico campus representatives for the new Campus Master Plan. These presentations replaced the meetings with the City of Chico representatives (who did not agree to meet) and the CSU Chico Sustainability Committee (only reactivated in 2018-19) expected in the original proposal.
The performance of existing biofiltration systems and the composition of urban storm runoff in Chico, California were assessed with 127 samples collected during 17 storm events between Oct. 2016 and Sept. 2019 (Goal 2). Storm runoff in parking lots, streets, and building downspouts had relatively stable pH values throughout the rainy season (7.0 ±0.6, mean ±standard deviation). In contrast, electrical conductivity (EC, indicator of salinity) and turbidity measurements displayed large variability (EC = 8 – 270 μS/cm, mean 48 ±49 μS/cm, turbidity = 4 – 583 NTU, mean 56 ±86 NTU). Similarly, concentrations of nutrients and total petroleum hydrocarbons (TPH) varied greatly in urban runoff. Nitrate concentrations ranged from non-detectable (ND) to 60.9 mg/L (mean: 5.1 ±8.6 mg/L), ammonium concentrations ranged from 0.04 to 21.2 mg/L (mean: 1.4 ±2.5 mg/L), dissolved organic carbon (DOC) concentrations ranged from 0.4 to 1543 mg/L, and TPH concentrations ranged from ND to 8.3 mg/L. E.Coli, indicating fecal contamination, were only detected in 1 of 13 samples. While there is no environmental screening level (ESL) for nutrients, increased nutrient concentrations can lead to algal blooms and eutrophication in downstream waters. Large DOC concentrations are problematic for drinking water purposes since DOC is a precursor for carcinogenic disinfection byproducts. TPH concentrations all exceeded their ESL (100 μg/L) when detected (15 of 35 samples). Heavy metal concentrations also greatly varied in storm runoff, ranging from ND to 1,690 μg/L (copper during the first storm of the 2017-18 season). Zinc and copper were the most abundant metals in storm runoff, followed by lead, chromium, and nickel. Cadmium concentrations were low but exceeded their ESL when detected (in 6 of 16 samples). These results indicate that urban storm runoff in Chico, CA is highly variable and can contain high concentrations of nutrients, heavy metals, and oil and grease. The high spatial and temporal variability in these constituents of concern represents a challenge for stormwater management purposes, making the modeling and prediction of storm runoff composition difficult.
Three local biofiltration systems were monitored throughout the project period (Goal 2). The CSU Chico rain garden displayed poor water treatment, with no change in water quality between its inlet and outlet. This lack of water treatment was attributed to the system design. In this rain garden, rooftop runoff is collected in an underground cistern until the cistern reaches its capacity. Only then, a pump empties the cistern into a constructed “streambed” made of large pebbles, which overflows into a storm drain, giving water a limited time (~ 5 min) to infiltrate. Performance would likely be enhanced by decreasing the pump flowrate to increase water residence time and infiltration. This recommendation was made to CSU Chico Facilities Management & Services and taken into account during the design phase of the new Science building. A small bioswale receiving runoff from a City of Chico parking lot (“City Lot 5”) was added to the list of field sites in 2018 due to its proximity to CSU Chico campus. The City Lot 5 bioswale performance was monitored during three storm events (1/24/18, 4/6/18, and 10/3/18). During the 1/24/18 and 10/3/18 storms, turbidity and nutrient concentrations remained unchanged between inlet and outlet locations, while metal concentrations decreased slightly (5-35%) on 1/24/18 and substantially (25-47%) on 10/3/18. During the 4/6/18 storm, the bioswale did not improve storm runoff quality, with no distinguishable trend in water quality parameters throughout the duration of the storm. These results suggested that the small bioswale only treated storm runoff to a limited extent, likely due to short water residence times.
A vegetated bioswale receiving runoff from a 5-acre parking lot at the local community college (Butte College) was monitored the most extensively during the project period, an effort that regularly involved Butte College students. Eleven storms were sampled, three of which sampled throughout the event. From 2016 to 2018, grab samples were collected and flows were estimated using a highly erroneous float method. In collaboration with Butte College, CSU Chico Civil Engineering and CSU Chico Concrete Industry Management faculty, flow-measuring devices (flumes and weirs) were designed and installed in May 2018. Pressure transducers were deployed at each inlet and outlet of the bioswale in Oct. 2018 to record storm flows on a continuous basis (1-min intervals) and estimate pollutant loads for specific storm events. The detailed monitoring of five storms in 2018/19 greatly expanded our understanding of the system performance. With a 900-ft length, this bioswale displayed larger hydrologic residence times (30 min to 4 h) than the City Lot 5 bioswale (~ 15 min) and decreased storm runoff volumes by 38% to 78%. This system displayed a greater ability to remove TPH and heavy metals than nutrients and fecal coliforms. TPH were completely removed from parking lot runoff, with ND concentrations at the outlet. The bioswale removed on average 79% of metal mass inputs (loads) during the five storms intensively monitored in 2018/19, while nutrient loads were only decreased by 42% on average (range: -208% to 99% removal). Fecal contamination was detected at the outlet of the bioswale, with greater E.Coli counts than in parking lot runoff at the bioswale inlets (data from three storms).
Monitoring of the bioswale throughout the 2017-18 rainy season pointed to a potential seasonality in its water treatment performance. While nitrate, phosphate, and DOC concentrations decreased between the inlets and the outlet locations during the first storms of the season (fall), the bioswale acted as a source of these nutrients during the late season (spring) storms, with increases in nitrate (26%), DOC (210%), and phosphate (410%) concentrations from inlets to the outlet on 4/6/18. This seasonal pattern was also observed for the metals chromium (136% increase), copper (13% increase), and nickel (116% increase). The more detailed monitoring of flows during the 2018-19 rainy season improved our understanding of the bioswale performance and suggested that soil moisture conditions could be a stronger control on water treatment than seasonal variability. During the first major storm event of the rainy season (Nov. 21-23, 2018), the bioswale first acted as a sink for metals and nutrients during the rising limb of the storm hydrograph, with load reductions ranging from 36% (phosphate) to 94% (zinc). Towards the end of the same storm (falling limb of the hydrograph), as soils were saturated, the bioswale became a source of nutrients with load increases ranging from 30% (phosphate) to 208% (ammonium). In contrast, the bioswale acted as a sink of both nutrients and metals during the rising limb of a spring storm (March 2019). These results highlighted the role of antecedent soil moisture conditions; the bioswale can remove efficiently pollutants from runoff with initially dry soils, while it can become a source of nutrients under saturated soil conditions. Vegetated bioswales are known to efficiently remove metals but tend to increase nutrient and coliform concentrations because they also serve as habitat for plants and animals. We documented the phytoaccumulation of chromium, copper, lead, nickel, and zinc in the bioswale over the course of a rainy season (Oct. 2018 – May 2019), with preferential accumulation in plant roots compared to shoots. This finding confirmed that metals were retained within the bioswale and was consistent with greenhouse column tests (described below). Overall, the study of the Butte College bioswale contributed new knowledge on stormwater treatment in vegetated biofiltration systems by highlighting a potential seasonal variability in performance, the importance of antecedent soil moisture conditions, and the role of plants in metal uptake, which is relevant information for the management of these systems.
In order to identify key design parameters contributing to biofilter performance and resiliency (Goal 1), column tests were conducted in the laboratory and subsequently in the greenhouse. During year 1 (2016-17), triplicate PVC columns (12” tall) were built in the laboratory according to Phase I findings, which identified the best performing filtration mix as a 2:1 soil-sand mixture with local soil (Vina loam). The biofilter columns consisted of 2 cm of pebbles (for drainage), 1 cm of fine sand (for particles retention), 22 cm of filtration mix (2:1 soil-sand), 5 cm of mulch (bark) and 3-4 Santa Barbara sedges (Carex barbarae). Synthetic stormwater of known metal, nutrient, and sediment concentrations was applied to test columns on 3/2/17 and 4/1/17 and its composition after biofiltration was used to assess pollutant removal. Test variables included five native, drought-resistant plants (deer grass – Muhlenbergia rigens, spreading rush – Juncus patens, Santa Barbara sedge – Carex barbarae, California rose – Rosa Californica, blue sage – Salvia clevelandii), four media additives (zeolite, activated carbon, compost, rice hull biochar), and biofilter depth (24”). Native plants were investigated to assess their role in heavy metal removal in biofilters and to determine the fate of metals in plants. Dissolved copper, lead, nickel, and zinc were quantitatively removed (77-100%) from the applied stormwater. Over time, metal removal from stormwater was maintained or improved. To increase nutrient retention in biofilters, four media additives were added to the filtration mix of triplicate test columns (15% by volume); zeolite, activated carbon, compost, and rice hull biochar. Rice hull biochar was acquired locally from another EPA P3 Phase I project at Butte College (project SU836790). Over time, biofilters with media additives removed more nitrate (max. 33% by zeolite) than controls, which leached nitrate. Ammonium was efficiently removed by biochar and zeolite (60 – 96%), while the removal performance of activated carbon (40%) was lower than controls. Phosphate removal varied greatly among media additives: activated carbon and zeolite (83 – 94% removal) outperformed controls (72%), while biochar and compost leached additional phosphate (20 – 22%). All biofilters leached DOC, with biofilters containing compost releasing up to 17 times more DOC than was applied. Overall, compost additives were less efficient at removing nutrients in biofilter columns than zeolite and biochar additives.
During year 2 (2017-18), column tests were moved from the laboratory to a temperature-controlled greenhouse to ensure optimal plant growth with ample supply of natural light. New columns were built similarly to the previous year. Test variables included the same five native, drought-resistant plants as in year 1 and five media additives. Native plants were investigated to assess their role in heavy metal removal in biofilters and to determine the fate of metals in plants. Biofilters effectively retained 99% of the six applied metals (Cd, Cr, Cu, Ni, Pb, Zn) across five applications between Oct 2017 and March 2018. During this six-month experiment, plants preferentially accumulated metals in their roots relative to their shoots. Five media additives were investigated for their potential to improve nutrient retention in biofilters (15% by volume) in a long-term (18 months) test: rice hull biochar, walnut shell biochar, logging residuals biochar, water treatment residuals, and zerovalent iron. Of note, the rice hull biochar was acquired locally from another EPA P3 Phase II project at Butte College (project SV839370). Six applications of synthetic stormwater were performed between June 2018 and Oct 2019. Ammonium was quantitatively (>91%) removed by all additives, which generally performed similarly to control columns (no additive). After an initial leaching phase observed during the first synthetic runoff application, additives enhanced nitrate removal (14-98%) compared to controls (2-44% removal), with highest removal in biofilters amended with zerovalent iron. Nitrate removal was highly variable between synthetic runoff applications, suggesting other potential controlling factors such as soil moisture or seasonal variability in plant uptake. Water treatment residuals and zerovalent iron were the only additives to maintain a consistent phosphate removal (23-100%) over time compared to controls and biochar additives that leached additional phosphate (7-632% increase). These results indicated that amending the filtration mix of biofilters with byproducts of industrial (zerovalent iron and water treatment residuals) and agricultural (biochars) activities generally improved nutrient removal from stormwater runoff. Zerovalent iron and water treatment residuals were especially promising with high nitrate and phosphate removals.
This project produced the anticipated outputs specified in the original proposal, with the exception of the preparation of a guidance document for small cities. Similar resources are publicly available (e.g., Guidance for Stormwater and Dry Weather Runoff Capture at Schools, Office of Water Programs at CSU Sacramento, Dec 2018, Stormwater Quality Design Manual, Cities of the Sacramento Region, July 2018), therefore the EPA P3 student team chose to focus on its field monitoring efforts and increase its number of presentations at scientific conferences.
Conclusions:
This project characterized the typical composition of urban storm runoff and evaluated the performance of three biofiltration systems in the small city of Chico in northern California (Goal 2). The project findings highlighted key biofilter design criteria for optimal water treatment. Vegetation (with associated soil and microorganisms) and greater hydrologic residence times enhanced heavy metal and total petroleum hydrocarbons removal, while additional management strategies appeared necessary to efficiently remove nutrients and fecal coliforms from vegetated biofiltration systems. The water treatment performance of the Butte College bioswale was likely controlled by antecedent soil moisture conditions and displayed seasonal variability, two factors necessitating more investigation. We also evaluated the role of native plants and media additives in a greenhouse setting to identify key design parameters contributing to biofilter performance and resiliency (Goal 1). Metal phytoaccumulation was documented in all plants tested and trace metals were found to accumulate in the root zone rather than in above-ground plant tissues. Amending biofilters with byproducts of industrial (zerovalent iron and water treatment residuals) and agricultural (biochars) activities generally improved nutrient removal from stormwater runoff, with greatest water treatment from zerovalent iron and water treatment residuals.
To investigate the feasibility of media additives in bioswales, a Life Cycle Analysis (LCA) compared the current Butte College bioswale with hypothetical bioswales amended with the additives tested in greenhouse tests (15% by volume, all other building specs maintained identical). Each system was considered for a 20-year lifetime in the LCA software SimaPro with a treatment unit of one m3 of treated storm runoff. Amendments decreased the eutrophication potential of bioswales by ~20% compared to the current system, with the hypothetical bioswale amended with water treatment residuals leading to the lowest eutrophication potential (8.0 vs 10.3 g N equivalents per m3 of treated runoff). In other impact categories, such as global warming impact and stratospheric ozone depletion, the bioswale amended with zerovalent iron had the greatest impact, exceeding the impact of the current system, while the bioswale amended with water treatment residuals had the lowest impact. These LCA results indicate that the additives considered would further reduce the risk of eutrophication in urban creeks if amended to vegetated bioswales, and that water treatment residuals were the most promising alternative due to their lowest global warming impact and stratospheric ozone depletion.
The design and amendment recommendations of this study consist in preferentially using vegetation over rocks, lengthening flow paths to increase residence times, and amending systems with waste products from other activities, such as water treatment residuals, zerovalent iron, and biochars. These design recommendations were demonstrated to be technically effective and are economically feasible as they consist in slight design changes to biofiltration systems or involve repurposing waste byproducts. This study demonstrated that these easy-to-implement design modifications have quantifiable benefits to people, prosperity, and the planet by improving urban storm runoff water quality and decreasing the risks of flooding and eutrophication in urban creeks. These findings are relevant to the EPA’s mission to protect human health and the environment by providing a critical assessment of design criteria of biofiltration systems to improve the water quality of urban storm runoff.
References:
CASQA (2009). Introduction to hydromodification. White paper. California Stormwater Quality Association. www.casqa.org
Dietz, M.E., & Clausen, J.C. (2006). Saturation to improve pollutant retention in a rain garden. Environmental Science and Technology 40 (4), 1335–1340.
EPA (1999). Preliminary Data Summary of Urban Storm Water Best Management Practices. US Environmental Protection Agency. EPA-821-R-99-012. https://www3.epa.gov/npdes/pubs/usw_a.pdf
Guo, J. C. (2009). Preservation of Watershed Regime for Low-Impact Development through Detention. Journal of Hydrologic Engineering, 15(1), 15-19
Hawley, R. J., & Bledsoe, B. P. (2011). How do flow peaks and durations change in suburbanizing semi-arid watersheds? A southern California case study. Journal of Hydrology, 405(1), 69-82.
Lim, H.S., Lim, W., Hu, J.Y., Zieglar, A., Ong, S.L. (2015). Comparison of filter media materials for heavy metal removal from urban stormwater runoff using biofiltration systems. Journal of environmental management 147: 24-33.
Journal Articles:
No journal articles submitted with this report: View all 23 publications for this projectSupplemental Keywords:
Storm water management; urban water planning; low impact development; best management practices; green infrastructure; water treatment technology; pollutant removal; bioretention; water cycle; cost benefit analysis; environmental education.Relevant Websites:
Photo shoot with student researchers and Jason Halley Exit
Environmental Chemistry Laboratory at CSU Chico Exit
Progress and Final Reports:
Original AbstractP3 Phase I:
Evaluating Biofiltration in Small Urban Areas: Chico, California Case Study | Final ReportThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.