Field and Laboratory Evaluation of Bioretention Bacterial Removal

Field and Laboratory Evaluation of Bioretention Bacterial Removal from Urban Stormwater Runoff in South Texas Ahmed Mahmoud, Augusto Sanchez, Javier Guerrero and Kim D. Jones Civil Engineering Department University of Texas Rio Grande Valley & Environmental Engineering Department Texas A&M University-Kingsville 1

* * Storm-Water Runoff as Non-Point Source Pollution * Bioretention research field and lab results in STC State of the science of bacteria removal in bioretention systems 2

* Urban runoff as a Non-point source pollution carries sediment, nutrients, heavy metals and pathogens into surface waters. * Stormwater runoff was among the top three sources of pollution in lakes, ponds, reservoirs, and estuaries (USEPA 2000). * http: //www. starkenvironmental. com/b-0 -stormwater. html Based on their design, Green Infrastructure or Low Impact Development (bioretention, permeable pavement, wetlands and rain gardens) practices can use a combination of treatment mechanisms to improve stormwater runoff. 3

Wetland (Weslaco, LRGV TX) Bioretention (Mc. Allen, LRGV TX) Bio-swale, (Brownsville, LRGV TX) Green Roof (San Juan, LRGV TX) Rain Harvesting system (Weslaco, TX) Permeable Pavement (La Feria, LRGV TX)

Arroyo Colorado watershed (http: //arroyocolorado. org) 5

* Controlling surface water and associated runoff pathogens are considered a growing challenge due to the variety of organisms and strains that can exist (e. g. Clostridium and Salmonella) http: //www. doh. wa. gov/Communityand. Environ ment/Drinking. Water/Contaminants/Coliform Positive relation between E. coli and sickness http: //water. usgs. gov/edu/bacteria. html 6

* Bioretention is an infiltration practice through porous media; that uses a biologically active filtration bed to remove contaminants. * One of the most commonly used LID practices. * Significant reduction of runoff volume provided by the bioretention cells with water quality improvement by substantially reducing the various pollutants. https: //www. hydrologystudio. com/help/bioretention-ponds. htm 7

Processes Sedimentation Filtration System Components Mulch Adsorption Coarse Sand Absorption Pore Space Cation Exchange Capacity Polar / Non-polar Sorption Microbial Action (aerobic / anaerobic) decomposition / nitrification / denitrification Plant Uptake Cycling Nutrients / Carbon / Metals Biomass Retention (Microbes / Plant) Evaporation / Volatilization Surface Area Complex Organics Microbes Biofilm Plants “Ecological Structure” Source: Bioretention “Rain Garden” Technical Seminar San Francisco Regional Water Quality Control Board Santa Clara Valley Urban Runoff Pollution Prevention Program 8

* In comparison among stormwater infiltration performance studies, bioretention has been estimated to achieve even greater bacterial removal rates than sand filtration systems * Because bioretention media promotes biological activity BMP Proposed Fecal Coliform Removal Ability Dry extended detention basin Medium Wet detention basin (wet pond) Medium Stormwater wetlands Medium Sand filter High Bioretention High Grassed swale Low (NCDENR – Stormwater Best Management Practices Manual, 2007) 9

* The core bioretention media is local material of 30 inch depth finely well graded sand washed from silt and clay, predominate grain size range (0. 81. 2 mm), commercially known as washed sand number 20. Layer of 3 inch gravel 30 inch Washed Sand #20 7 to 9 inch crushed stone STC bioretention cell cross section (Source: Perez Consulting Engineers, Civil Drawing 2011) 10

* South Texas College (STC) in Mc. Allen, Texas has a parking lot area is 32, 812 square-feet parking interspersed with 250 ft linear bioretention cell and the rest of the parking lot is traditional asphalt pavement. * The field scale monitoring of constructed LID structures was conducted evaluate each site’s effectiveness in reduction of storm water bacteria and nutrients during storm events. East end of the STC parking lot study section 2 1 1 - Bioretention Cell (BRC) drainage area 2 - Traditional Asphalt Pavement (TAP) drainage area 11

Tubing connected to 6 in weir Solar Panel Rain Gauge (ISCO 674) Signature Flow meter Autosampler 12

13

* 45 storm events were monitored for flow between March 2016 till April 2017 ranging from 0. 07 to 1. 95 inch total was 20. 73 inch. * Bioretention cell total volumetric runoff was less than traditional asphalt pavement by 82%. * Comparing by the size of the storm event, bioretention has shown capacity to eliminate entirely the runoff for certain small rain events, while runoff was observed in medium and large events under certain rainfall depth, intensity and antecedent dry period conditions. 14

Runoff hydrographs for 6 hours (November 8, 2016), rain event totaling 16. 5 mm (0. 65 inch) 15

Runoff hydrographs for 10 hours (March 3, 2017), rain event totaling 38. 3 mm (1. 51 inches) 16

Total runoff volume (m 3) for the bioretention cell and antecedent dry period (days) with different rainfall depth (mm) for small storm events (1. 8 to 7. 9 mm) 17

Total runoff volume (m 3) for the bioretention cell and antecedent dry period (days) with different rainfall depth (mm) for medium and large storm events (8. 1 to 49. 5 mm) 18

Number of samples * TSS BOD TKN TP E. coli 21 19 21 21 21 To determine of the effect of ADPs on the water quality results, two samples were collected on the same day after having two discreet storm events in 09/26/16, first event occurred in the morning and the other event occurred in the afternoon with 8 hours difference. * Bioretention cell effluent were significantly better quality than traditional asphalt pavement runoff, showing bioretention cell’s removal potential for various pollutants associated with the surface runoff Total Removal Efficiency % TSS BOD TKN TP E. coli 96 51 65 83 49 19

20

A comparison between total Kjeldahl nitrogen (TKN) load (mg/m 2) for the Traditional Asphalt Pavement (TAP) and Bioretention Cell (BRC) at different storm events for 21 collected samples. 21

1000000 Log E. coli (MPN/m 2) 100000 1000 10 1 BRC A comparison between E. coli concentration (mg/L) for the Traditional Asphalt Pavement (TAP) and Bioretention Cell (BRC) at different storm events for 21 collected samples. Total concentration reduction 49% TAP A comparison between E. coli load reduction (mg/m 2) for the Traditional Asphalt Pavement (TAP) and Bioretention Cell (BRC) at different storm events for 21 collected samples. Total load reduction 96% 22

* In general, the longevity for bacteria is greater in moist soil than in dry soil * Eight samples of the bacterial effluent concentration of the bioretention were higher than traditional pavement * Five of the eight samples were preceded by storm events a day before collecting it, indicating that antecedent dry period has a negative impact on bacterial removal. 23

A comparison between E. coli load (MPN/m 2) for the Traditional Asphalt Pavement (TAP) and Bioretention Cell (BRC) at different storm events for 21 collected 24 samples.

* Straining due to filtration and attachment has been suggested as two major mechanisms for bacterial capture in porous media. * Attachment has been suggested to be more significant, due to the small cell size of bacteria when compared with media-particle diameter. * Several factors can affect bacterial attachment on the particle surface of porous media including the texture and the size of the porous media, the presence of organic matter and biofilms, temperature, flow rate and electrostatic charge. * Due to the complexity of bioretention biological processes, the laboratory scale has merit in the simplification of a complex full scale system 25

• Bench-scale biofilter column length was 21 cm and 5 cm diameter • The stormwater inflow rate was chosen based on the average rainfall intensity data that have been collected by the rain gauge at the STC site • Samples from the effluent were analyzed with an IDEXX Quanti-Tray system to determine E. coli concentrations. • Calculate the mass removal efficiency per mass of sand. 26

https: //www. idexx. com/water/products/colilert. html 27

Parameter Value Media Bulk density 1. 6 gm/cm 3 (630 gm) Flow rate 50 m. L/hr Inflow Volume 10 L Pumping duration 200 hours (8. 3 days) Inflow Concentration 1011. 2 MPN/m. L Bacteria load reduction 1933 MPN/gm per mass of sand/day 28

Experimental setup for the laboratory column studies. The Erlenmeyer flask contained the influent (tracer or bacterial solution), which was pumped at constant flow rate through the column that contain 100 gm of washed sand #20. An outflow samples were collected by fraction collector to compare the normalized outflow with respect to time. 29

Advection-dispersion equation for the conservative tracer Analytical solution to this equation by Ogata and Banks (1961) Equilibrium adsorption for microbes 30

Sand Column Breakthrough Point Bromide 70 min 150 min Bacteria (70% lower than C 0) 31

Bioretention cell with STX River Rock Bioretention cell with STX Recycled concrete 32

* The bioretention cell at STC showed significant stormwater flow reduction (82%) and substantially reduced the concentration of various pollutants including TSS, BOD 5, TKN, TP and E. coli with average removal at 96%, 51%, 65%, 83% and 49%; respectively. * Antecedent dry periods (ADPs) appeared to impact the bioretention hydrological performance for medium and large storm events. Likewise, ADPs seem to affect the reduction of total nitrogen and indicator bacteria, with more impact on bacteria and less on total nitrogen. * Lab experimental results showed that bacteria transport through bioretention porous media could be modeled using a onedimensional advection-dispersion-attachment (A-D-A 1 D) model 33

Acknowledgments The authors wish to thank the Texas Commission on Environmental Quality, Non-Point Source Team for their support and funding for this work under TCEQ Contract 581 -14 -30049, and the project team including Tim Cawthon of TCEQ, Rene Gonzalez of STC engineering, Martin Knecht of STC, Taufiqul Alam from Environmental Engineering at TAMUK, Yeasir Rahman now at USF, and GLO Coastal Management Program for funding La Esquina Subdivision Project. 34

35

* The media characterization results of the sand sample showed that the media used in the STC sand bioretention cell is well graded sand Parameter * Particle size (mm) Sand Lab results (d 10=0. 1), (d 30=0. 18) and (d 60=0. 28) Hydraulic Conductivity 5. 24 in/hr. Porosity 0. 354 Moisture content 14. 15% Inorganic content 99. 4% Organic Content 0. 6% The media characterization results of STX river rock and STX recycled concrete Parameter Dominant particle size STX River Rock STX Recycled Concrete 1 inch 0. 5 inch Hydraulic Conductivity Porosity 0. 43 cm/sec 0. 45 0. 03 cm/sec 0. 49 36

37

Bioretention Traditional Pavement % Reduction Kg/ha. year TSS (mg/m 2) 317 80027 99. 6 1108. 5 BOD (mg/m 2) 177 1425 87. 5 17. 4 TKN (mg/m 2) 36 304 88 3. 9 TP (mg/m 2) 8. 9 109. 6 92 1. 47 E. coli (MPN/m 2) 1. 5 E 06 2. 1 E 07 75* - * Geometric mean value 38

39

40

Drainage Area Sources (sq ft) Vegetation 37428 Road 5992 Undeveloped Access Road Total Drainage Area 2178 45598 41

Area of Los Fresnos bioretention 300 (sq ft) Drainage Area 16675 (sq. ft) Area of the Biofilter Column 0. 59 (sq. ft) Drainage Area equivalent to the biofilter column 32 (sq. ft) Average Total Rainfall intensity (inch/hour) 0. 103 Flow rate for the column (L/hr) 1. 56 42

Experimental Plan Volume (L) Flow Rate (L/hr) Time (hours) First Event 10 1. 5 6. 6 Second Event 17 1. 5 11. 3 Third Event 30 1. 5 20 43

The bacterial removal efficiency percentages for river rock and recycled concrete at different inflow volumes after biofilm formation. Recycled Concrete River Rock 10 L 17 L 30 L Number of samples 10 10 8 10 10 10 Minimum 28 45 6 -1 -5 -21 Maximum 85 85 51 62 68 64 Average 50 70 29 31 45 22 Median 49 76 28 27 50 21 44