CICEET Progress Report for the period 9/01/05 Through 3/05/06

Project Title: Transport and Capture of Pathogens from Urban Stormwater Runoff Using Bioretention
Principal Investigator(s): Allen P. Davis, University of Maryland
Additional Investigator(s): Eric A. Seagren, University of Maryland; Jeffrey S. Karns, USDA
Project Start Date: September 1st, 2005

Figures

Figure 1

Figure 2

Figure 3

Figure 4

Tables

Table 1

Table 2

Tasks to meet objectives
Only Task 1 was initiated during the first six months of this project.

Progress on Tasks
The preparation of materials has been completed, and the column has been set up. A bromide tracer experiment and one column transport experiment using a derivative of Escherichia coli O157:H7 have been completed and analyzed. The experimental methods and data analysis are briefly described in the following paragraphs.

Bacterial Growth
A derivative of E. coli O157:H7, stain B6914, was used as the test indicator organism. The original inoculum of strain B6914 was provided by J. S. Karns (Environmental Microbial Safety Laboratory, USDA). To prepare the feed solution, a colony of strain B6914 from an agar plate was picked and used to inoculate in 5 mL of LB broth containing ampicillin (Ap) at 100 µg/mL in a sterile centrifuge tube which was incubated at 37°C with shaking (120 rpm) overnight. Then, a 1 mL sample from this liquid culture was aseptically transferred into 50 mL LB broth (Ap 100 µg/mL) in a sterile 250 mL flask with a cotton plug. This culture was incubated at 37°C with shaking (120 rpm), and then the cells was harvested during the late exponential phase (OD_600nm ª 1.0). To harvest the cells, the culture broth was centrifuged (4,500 rpm) for 10 min, the cell pellet was washed with phosphate buffered saline (PBS, pH=7.5) and then centrifuged again. After pouring off the supernatant, the cells were suspended in PBS as a stock solution.

Column Setup and Bromide Tracer Experiment
Column experiments were performed in a glass chromatography column (Kontes, Vineland, NJ) with a 20 µm porosity polyethylene bed support. The inner diameter of this column is 2.5 cm and the height is 23 cm. Experiments to date have been performed using conventional bioretention media (sandy loam soil/sand/mulch mixture). The sandy loam soil was collected from Tantallon, Maryland, the sand was obtained from a local home supple store, and the mulch was obtained from the College Park Department of Public Works. The sand, soil and mulch were sieved through a 1.18 mm sieve, and then mixed at a volume ratio of 5:3:2. This ratio is typical for use in the bioretention facilities. Subsequently, the media mixture was dry-packed into the column to a depth of about 21.7 cm. Prior to initiating the tracer studies, the column was flushed to reach saturation using synthetic runoff. The content of the synthetic runoff was based on information on urban stormwater runoff chemistry. The characteristics of this water are presented in Table 1, Davis et al., 2001).

For the bromide tracer experiment, sodium bromide was dissolved in the synthetic runoff to obtain a bromide concentration of 200 mg/L. Then the influent was continuously pumped into the column from the top for 6 hrs. During this time, the flow rate was maintained at approximately 20 mL/hr (approach velocity = 4 cm/hr). This represents a common storm (return period < 1 year) concentrated by a factor of 20 from the drainage area to the treatment facility. The effluent samples were collected every quarter- or half-hour from the bottom of the column. A schematic diagram of the column experimental set up is shown in Figure 1. The Bromide concentration of influent and effluent samples was measured using a Cole-Parmer 27502-04, 05 bromide electrode (Cole-Parmer Instrument Company, Vernon Hills, IL). The detection limit for the bromide probe is 0.01 mg/L in the presence of other ions.

Pathogen Transport Experiment
The E. coli O157:H7 stain B6914 in PBS stock solution was diluted in the synthetic runoff to a final concentration of 10⁵~10⁶ CFU/mL. The runoff was then continuously pumped into the column from the top at the flow rate of 20 mL/hr for 6 hrs. The effluent samples were collected every half hour or one hour from the bottom of the column, and influent samples were collected at the start and the end of the experiment. Strain B6914 concentrations in influent and effluent samples were determined using the method of heterotrophic spread-plate counts adapted from Standard Method 9215 (APHA et al. 1995).

Filtration Modeling
Bacterial transport can be described using a one-dimensional advection-dispersion equation (van Genuchten and Alves, 1982; Harvey and Garabedian, 1991).

(1)

In equation (1), R is the retardation factor, c is the concentration of bacteria suspended in aqueous solution (ML^-3), D is the hydrodynamic dispersion coefficient (L²T^-1), x is the distance (measured from the entrance of column), n is the pore water velocity (LT^-1), kc is deposition coefficient (T^-1), and Rd is the rate at which bacteria are detached from the collector (ML^-3T^-1). The hydrodynamic dispersion coefficient D, pore water velocity n, and porosity q were determined using the bromide conservative tracer test data, as described below. The bacteria deposition coefficient can be determined based on colloid filtration theory (Tien and Payatakes, 1979), as follows:

(2)

where dc is the diameter of the collector. For conventional bioretention media with low uniformity (d₆₀/d₁₀ = 4.9), d₁₀ was selected as d_c (Martin et al., 1996). The sticking coefficient a and the single collector collision efficient h0 can be calculated using the adjusted correlation equations developed by Tufenkji and Elimelech (2004). For this initial data analysis of the pathogen transport experiments, detachment is neglected (R_d = 0). The analytical solution to the one-dimensional advection-dispersion equation (1) was provided by van Genuchten and Alves (1982).

The bromide breakthrough curve also can be described by the one-dimensional advection-dispersiion equation (R = 1, k_c = 0, and R_d = 0). A FORTRAN program “trafit1d” was used to fit Equation (1) to the tracer data (Seagren, 1994). The best-fit parameters (hydrodynamic dispersion D, pore velocity n and porosity q) were obtained by minimizing the sum of the squares of absolute residuals between the normalized experimental bromide tracer data and the normalized flux-averaged concentration calculated using the one-dimensional non-reactive solute transport model of Parker and van Genuchten (1984).

Difficulties A few experimental difficulties have been encountered in the experiments performed to date. Most importantly, strain B6914 concentration in the synthetic runoff was stable for approximately three hours, but then greatly decreased over time, interfering with long-term column studies. This may be due to low nutrient levels or the toxicity of oil in the synthetic runoff. This is not expected to be a problem if new synthetic runoff suspending fresh strain B6914 is made and applied to the column every three hours during column experiments.

Project Objectives for Next Reporting Period

Objectives
Comparing conventional bioretention media and engineered bioretention media for the capture and destruction of pathogens using column studies, and setting up the “long-term” sustainability studies.

Tasks to Meet Objectives
The goal for the next reporting period is to complete proposed Task 1 and Task 2, evaluating conventional bioretention media and engineered bioretention media (iron oxide-coated sand) for the capture and destruction of pathogens using column studies. Based on these results, a reengineered design for maximizing pathogen attenuation will be subsequently developed in Task 3.

Also during this time, Task 4, “long-term” sustainability studies, will be initiated using the same general experimental design as in Task 1.

Work Plan for Next Reporting Period
One Ph.D. student is working full time on this project. An undergraduate student will be hired and will begin assisting with sampling and routine laboratory work. The laboratory column experiments will become “routine”. Once the experimental design is finalized, some new columns will be ordered and set up.

Anticipated Success in Meeting Project Objectives
At this point, the project is somewhat behind the originally schedule, but it is still early. We are optimistic that some of the lost time will be made up in the near future. In a few months, we will have a better feel for the chances of keeping on schedule.

Overall Project Timeline Update
A revised project timeline is presented below. The dark bars represent the original timeline. The light bars are the revised timeline.

Preliminary Data
Modeling the bromide breakthrough curve with the one-dimensional advection-dispersion model resulted in a good fit of the observed data (See Figure 2). The best-fit parameters (hydrodynamic dispersion D, pore velocity n and porosity q) are presented in Table 2.

The effect of two different stain B6914 concentrations in synthetic runoff on the capture by conventional bioretention media has been tested. In the two bacterial transport column experiments, the stain B6914 concentrations suspended in the synthetic runoff were 2.0_10⁵ CFU/mL and 3.6_10⁶ CFU/mL, respectively. For the experiment with the input stain B6914 concentration of 2.0_10⁵ CFU/mL, the maximum effluent cell concentration was reached at 2.5 hrs (or 1.36 pore volumes). Most of the input cells were retained by media (C/C_o of 0.075) during the six-hour column experiment (See Figure 3). However, for the input stain B6914concentration of 3.6_10⁶ CFU/mL, the maximum effluent cell concentration greatly increased (C/C_o of 0.88, See Figure 4). For both experiments, cell breakthrough was slightly retarded relative to the bromide tracer experiment (Average retardation coefficient R=1.12). Modeling the bromide breakthrough curve with the one-dimensional advection-dispersion model resulted in a generally good fit of the experimental data (See Figure 3 & Figure 4).

From the results obtained so far, it is obvious that the removal efficiency of E. coli by conventional bioretention media decreases with increasing input concentration of E. coli O157:H7 stain B6914 in the synthetic runoff. Thus, the conventional bioretention media possesses a much higher removal capacity to remove E. coli O157:H7 stain B6914 from synthetic stormwater runoff when the lower influent concentration was applied (Co = 2.0_10⁵ CFU/mL).

Dissemination
Publications: None
Workshops: None
Conferences: None
Manuals, Protocols: None
Outreach Activities: None
Contact with End Users: None
Patent, Copyright, Invention Disclosure Activity: None

Expenditures
Expenditures are in the range expected for the work accomplished to date.

End User Advisor Feedback
Name: Neil Weinstein
Organization: The Low Impact Development Center, Inc.
Location: 5010 Sunnyside Avenue
Suite 200
Beltsville, Maryland 20705
Phone number: (301) 982-5559
E-mail: nweinstein@lowimpactdevelopment.org

Research Response
1.) This research can be used to address impairments to water bodies from non-point source wet and dry weather flows. This includes loads from developed and non-developed lands. The results of the technology can be used for structural and non-structural BMP applications.

2.) Not applicable

3.) Challenges:
a. Variability in soils conditions from bench testing to field (e.g.composition of soil matrix, moisture content, temperature, wetting and drying, bulk density)
b. Composition of mulch and other amendments
c. Variability in field soil conditions
d. Variability in source and load of pollutants and duration of exposure
e. Effects of soil chemistry from maintenance (e.g. fertilization, pesticides, mulching,)
f. Effects of other pollutants (e.g. oils, grease, salts)
g. Effects on plants
h. Change in effectiveness over time

4.) These questions do not appear to be addressed in the first phase, which appears to be just setting up the basic research approach.

5.) This is very valuable research, due to the significant number of estuaries and water supply reservoirs that are being listed due to this family of contaminants. A low cost and effective approach is critical and this technology has tremendous potential.

References
APHA, AWWA, and WEF. 1995. Standard Methods for the Examination of Water and Wastewater, American Public Health Association, American Water Works Association, and Water Environment Federation, Washington, D.C.

Davis, A.P., Shokouhian, M., Sharma, H. and Minami, C. 2001. Laboratory Study of Bioretention for Urban Storm Water Management. Water Environ. Res., 73, 5-14.

Harvey, R.W., Garabedian, S.P. 1991. Use of colloid filtration theory in modeling movement of bacterial through a contaminated sandy aquifer. Environ. Sci. Technol. 25, 178-185.

Martin, M.J., Logan, B.E., Johnson, W.P., Jewett, D.G., Arnold, R.G., Member, ASCE. 1996. Scaling bacterial filtration rates in different sized porous media. J. Environ. Eng. 122, 407-415.

Parker, J.C., van Genuchen, M.Th. 1984. Determining transport parameters from laboratory and field tracer experiments. Bulletin 84-3. Virgia Agric. Exp. Stn. Blackburg, VA.

Rajagopalan, R., Tien, C. 1976. Trajectory analysis of deep-bed filtration with the sphere-in-a-cell porous media model. AIChE J. 22, 523-533.

Seagren, E.A. 1994. Quantitative evaluation of flushing and biodegredation for enhancing in situ dissolution of nonaqueous phase liquids. Ph.D dissertation, Univ. of Illinois, Urbana Champaign.

Tufenkji, N., Elimelech, M. 2004. Correlation equation for predicting single-collector efficiency in physicochemical filtration in satured porous media. Environ. Sci. Technol. 38, 529-536.

van Genuchten, M. Th., Alves, W.J. 1982. Analytical solutions of the one-dimensional convective-dispersive solute transport equation. United States Department of Agriculture (USDA), Agricultural Research Service, Technical Bulletin Number 1661.