Required C T Value for 5-log Virus Inactivation at Full Scale

Transcription

1 E18 Required C T Value for 5-log Virus Inactivation at Full Scale MICHAEL J. ADELMAN, 1 MICHAEL PHELPS, 2 ROBERT T. HADACEK, 1 OLIVER R. SLOSSER, 1 SIMON CALVET, 1 JOAN OPPENHEIMER, 1 AND JAMES H. BORCHARDT 1 1 MWH Global, Pasadena, Calif. 2 Camrosa Water District, Camarillo, Calif. The required free chlorine C T (concentration times time) value for 5-log virus inactivation was measured at a full-scale water recycling plant through a suite of tracer, chlorine decay, and virusseeding tests on a baffled-channel chlorine contactor. This plant produces fully nitrified, low-turbidity filtered effluent. MS2 bacteriophage virus was seeded into the contactor at 10 6 pfu/ml, and inactivation of this virus, along with native coliform, were measured. A consistent MS2 inactivation rate >0.28 log-l/mg-min was observed for three different flows, consistent with other benchand pilot-scale studies. At C T values >22 mg-min/l, the facility complied with California recycled water disinfection requirements by achieving >5.6-log inactivation of MS2 and reducing 10 4 cfu/100 ml influent total coliform to <2.2 cfu/100 ml in the effluent. These results showed how complete nitrification and effective suspendedsolids removal allowed a full-scale plant to realize favorable disinfection kinetics with free chlorine. Keywords: C T value, disinfection, free chlorine, inactivation kinetics, MS2 bacteriophage, water reuse There is interest in determining the C T value (the mathematical product of concentration and time) required for adequate chlorine disinfection of recycled water at full scale. As the need for water recycling around the world increases in response to growing demand, limited supplies, and climatic uncertainty, the questions of which disinfectant to use and what C T value to target are important to both existing facilities seeking to increase capacity and new facilities for which a rational design basis must be selected. Beyond the chlorination step itself, the effectiveness of disinfection also depends on the design and operation of the upstream processes at these facilities to produce effluent that is readily disinfected. Existing guidelines for recycled water disinfection are often quite stringent. For example, water quality requirements in California (CDPH 2014) stipulate a chlorine C T value of 450 mg-min/l to disinfect tertiary (i.e., filtered) recycled water, which is based on the kinetics of chloramine disinfection (Dryden et al. 1979). Recognizing that different disinfection processes inactivate pathogens at different rates, recycled water guidelines often allow for alternative disinfectants. For example, a.2 of the California regulations (CDPH 2014) also allows an exception from the C T requirement of 450 mg-min/l if the disinfection process is validated by demonstrating 5-log inactivation of a surrogate virus. For any disinfection process, an appropriate surrogate virus should be more resistant than the pathogenic human viruses to the proposed disinfectant so that the measured reduction of the surrogate will be conservative. MS2 bacteriophage is commonly used in disinfection studies, and it has been demonstrated as a conservative surrogate for the specific cases of chlorination (Tree et al. 1997) and disinfection of wastewater effluents for reuse applications (Sigmon et al. 2015). A variety of constituents in wastewater affect chlorination by reacting with chlorine (e.g., ammonia) or shielding microorganisms (e.g., suspended solids), and removal of these constituents can improve disinfection efficiency (Tchobanoglous et al. 2003). For water recycling plants that produce fully nitrified, low-turbidity effluent with low chlorine demand, it is possible to use free chlorine and achieve kinetics for pathogen inactivation that are much better than those realized with chloramination (Hirani et al. 2014). C T values lower than the 450 mg-min/l California regulation would allow higher flows at existing plants and smaller footprints for the disinfection infrastructure at new plants. Selecting a C T value appropriate to the wastewater quality also prevents overchlorination and reduces the formation of disinfection by-products that may affect potential reuse. In the current study, the required free chlorine C T value for 5-log virus inactivation was measured at a full-scale water recycling facility. The goal of this research was to re-rate the chlorine contactor at the plant to operate at lower C T values, and this was achieved through a suite of tracer tests, chlorine decay tests, and virus-seeding tests. The disinfection process at this plant initially operated at the default minimum C T of 450 mg-min/l, even though the plant reliably achieves nitrification and produces low-turbidity filtered effluent. Other plants in California have

2 E19 been approved for operation at lower C T values. For example, at the San Jose Creek East Water Reclamation Plant in Los Angeles County, a lower C T value was approved on the basis of pilot testing (Huitric et al. 2013). The novel methodological contribution of the current study was to carry out the testing in support of re-rating the disinfection process at full scale. MATERIALS AND METHODS Treatment plant description. This study took place at the Camrosa Water Reclamation Facility (CWRF) in Camarillo, Calif., which is owned and operated by the Camrosa Water District. The liquidphase treatment train of this plant (Figure 1) consists of bar screening, biological treatment in an oxidation ditch, secondary clarification, media (sand) filtration, and chlorination. Because virtually all effluent from this plant is reclaimed for crop irrigation, it is essential that the facility comply with recycled water treatment standards. Before the current study, the plant could meet these standards at flow rates only up to 1.5 mgd, at which point the chlorine contact basins would provide insufficient C T under the default regulatory minimum. This limitation sparked interest in studying the disinfection process for the purposes of re-rating this portion of the plant. The CWRF is operated with a mean cell residence time of 20 to 25 days to achieve nitrification, and the ammonia concentration of the effluent typically is below 1.0 mg/l ammonia as nitrogen (NH 3 -N). The mixed liquor suspended solids concentration is maintained at between 3,500 and 4,000 mg/l, and the dissolved oxygen in the oxidation ditch is maintained between 0.2 and 0.5 mg/l. During normal operation, chlorine is dosed into the clarifiers (to control algal growth and meet part of the chlorine demand), as well as into the contact basin, directly downstream of the tertiary filters. The chlorine dose is approximately 10 mg/l in each location. Turbidity and free chlorine in the disinfected filtered effluent are monitored by the CWRF control system. Continuous monitoring allows plant operators to verify that a free chlorine residual of at least 1 2 mg/l can be maintained during the disinfection process. Variable frequency drives on the oxidation ditch aerators provide operational flexibility to achieve fully nitrified secondary effluent by fine-tuning the aeration achieved in the oxidation ditch. Chlorine contactor description. The disinfection process at the CWRF consists of two parallel, baffled contact basins rated for a total flow of 1.5 mgd. These contactors, which predate the rest of the CWRF, limit the overall rating of the plant; the other treatment processes at the CWRF can treat up to 3.25 mgd. Chlorine is dosed in the form of sodium hypochlorite (12.5%) at the inlet box, where it is blended into the filtered effluent by an electric mixer. At the inlet box, flow is split into the two parallel basins. In the contactor, water flows over and under a series of wooden baffles and over an outlet weir and into a common outlet box. The dimensions of the contactor are shown in Figure 2. The contact basins are 10 ft wide and 110 ft long between the inlet box and the outlet box. When the plant is operating at the current rated capacity, the water depth in the basins is 7.7 ft, which gives a volume of 8,470 ft 3 (63,360 gal). The theoretical retention time is 122 min for each basin at the current rated capacity of 0.75 mgd (519 gpm) per basin. The baffles divide the reactor into 20 baffle spaces, each of which has a volume of ft 3 (3,168 gal) and a theoretical contact time of 6.1 min at the current rated capacity. The expectation that the existing contactor would be effective at higher flow rates was attributable to a high expected baffling factor, based on the favorable geometry to prevent short-circuiting and a length-to-width ratio of 11:1. Long, narrow geometry and significant baffling in a reactor typically favor advection over dispersion and allow the volume to be more fully used (Tchobanoglous et al. 2003). The specific case of a long channel with closely spaced vertical baffles allows kinetic processes to be modeled on the basis of FIGURE 1 Process flow schematic of the CWRF, with chlorination points and disinfection analyzers Influent flow splitter Clarifiers NaOCI Ammonia turbidity NaOCI Free chlorine turbidity Treated effluent Oxidation ditches Tertiary filters Chlorine contactors Headworks Return activated sludge Waste activated sludge CWRF Camrosa Water Reclamation Facility, NaOCl sodium hypochlorite

3 E20 contact time (Weber-Shirk & Lion 2010). Residence times of at least 90% of the theoretical value have been reported in other studies of well-baffled contactors (Rodgers & Butler 2010). Test flow rates and flow control. Testing was conducted at four different flow rates. The low flow rate corresponded to the current rated peak capacity (1.5 mgd), the medium flow rate corresponded to the proposed average daily capacity (2.25 mgd), and the high flow rate corresponded to the proposed peak hourly capacity of the chlorine contactor (3.24 mgd). A medium-high flow provided an additional test point within the range of flow rates that the basin was likely to experience (2.74 mgd) and was used only in the tracer study. Table 1 presents the test flows for the study and the corresponding theoretical contact times at the end of the contact basins, as well as the sample points used at each flow rate. The plant flows listed in Table 1 were the total flows when both contact basins were in use. For this study, only one contact basin (south) was used at the test flow rates shown in the table. The south basin was selected for testing on the basis of a preliminary slug test that showed that its residence time was slightly lower than that of the north basin with both in operation. Running the tests at one-half of the proposed plant flows on the portion of the contactor that is less effectively baffled was conservative, because the contact time achieved by the two basins in parallel was expected to be greater than the contact time of the south basin alone. To achieve consistent flow rates, a temporary pumping configuration was used (Figure 3). A temporary gate was inserted at the entrance to the second channel (north) to isolate this channel from the inlet box. The flow rate for each test was controlled by a diesel-powered temporary pump, which transferred water from the inlet of the north channel to the inlet of the south channel. The pump discharge hose was pointed toward the channel bottom to facilitate mixing of the first baffle space. A magnetic flow FIGURE 2 Plan-view diagram of the existing contactor 110 ft N High baffle Submerged baffle Inlet box 10 ft Outlet weir NaOCl dosing NaOCl sodium hypochlorite TABLE 1 Test flow rates, theoretical contact time, and sample points Test Scenario Plant Flow (Two Basins) mgd (gpm) Test Flow (One Basin) gpm Theoretical Contact Time min Sample Points a Low flow 1.5 (1,042) , 3, 5, 7, 12 Medium flow 2.25 (1,562) , 5, 7, 9, 12 Medium-high flow 2.74 (1,903) , 6, 8, 10, 12 High flow 3.24 (2,260) 1, , 7, 9, 11, 12 a The location of sample points is shown in Figure 3.

4 E21 meter 1 on the inlet piping connected to the pump measured the flow rate for the test. Varying the pump motor speed provided coarse flow control, and an adjustable-position butterfly valve was used to fine-tune the flow rate. The total plant flow rate always remained above the test flow rate for the duration of each test to ensure an adequate supply of water to the pump. Sample points are shown in Figure 3. At each flow rate, sample points were selected (Table 1) at theoretical residence times of approximately 10, 20, 30, and 40 min. Samples were also taken from the last baffle space during every test. For each flow rate, the same sample points were used for each phase of testing. Tracer study and chlorine decay testing. A set of tracer tests was conducted to measure the actual modal contact time at each sample point. The tracer tests consisted of a slug addition of fluoride at the first baffle space of the reactor along with monitoring conducted at points downstream to measure the reactor response. The slug of fluoride was added at the discharge point of the temporary pump for effective mixing. Fluoride ion selective electrodes 2 (ISEs) were placed at the end of the contact basin and four intermediate locations (Table 1). The probes continuously monitored the fluoride concentration, and the data were electronically logged. The chlorine decay study was conducted on the filtered effluent at three different test flows. On the basis of the average range of chlorine doses at the CWRF, the chlorine decay tests were performed at doses of 6, 8, and 10 mg/l into the filtered effluent at each test flow rate. Under normal operation, chlorine is also dosed into the secondary clarifiers at a constant 10 mg/l to prevent algae growth on the weir. Although the chlorine added at this location does not produce a free residual, it does satisfy a portion of the chlorine demand. Therefore, the experiments for the chlorine decay and virus seeding were conducted with this chlorine addition at the clarifiers. During this testing, 1-L samples were taken at five points along the contactor (Table 1) and analyzed for free chlorine as well as temperature, ph, nitrite, nitrate, ammonia, and ultraviolet (UV) absorbance. General weather conditions were also recorded. Virus-seeding study. The final phase of testing was a virusseeding study in which MS2 bacteriophage was injected into the contactor influent as a surrogate virus. This virus is one of the surrogate viruses identified in section a.2 of the California recycled water regulations (CDPH 2014). To quantify the rate and extent of disinfection, both MS2 virus and coliform bacteria were monitored along the contactor during the virusseeding test. Seed stock solution with an MS2 bacteriophage concentration of approximately pfu/ml was obtained from a commercial laboratory 3 for the virus-seeding studies. The MS2 solution was refrigerated on receipt at the CWRF. For all tests, the seed stock solution was diluted 10 times to obtain an MS2 bacteriophage concentration of approximately pfu/ml. A 32-gal plastic bin was used to store and mix the diluted seed stock solution. The stock was metered into the contactor with a peristaltic pump to achieve an initial concentration of pfu/ ml. The virus was injected near the temporary pump outlet in the first baffle space to facilitate mixing. The stock tank was sampled at the beginning and end of each test to verify that it maintained the target virus concentration. Before each test, 100-mL samples of tertiary effluent were collected to evaluate the background concentration of coliform and MS2 bacteriophage. During the test, chlorine was injected into the contactor at a dose selected to give 1 2 mg/l free residual at the end of the contactor, based on the results of the chlorine decay study. During the course of testing, four samples were taken at each of five sample points (Table 1), at least 20 min past the measured modal contact time for that location. One 1-L sample was analyzed immediately on site for free chlorine, ph, temperature, and UV absorbance; triplicate 100-mL sample bottles were FIGURE 3 Temporary pump configuration and sample point locations North contact basin N Meter South contact basin (inlet closed, basin used for testing)

5 E22 TABLE 2 Summary of analytical methods Parameter Analytical Method Minimum Sample Volume ml Preservatives Time of Analysis (for the Current Study) Fluoride Method 9214 a (USEPA 1996) 100 NA Continuous (sensor) BOD Method 5210B (Standard Methods 2012) 300 NA Within one day of collection Alkalinity Method 3020B (Standard Methods 2012) 200 Refrigeration Within one day of collection NH 3 -N Method (USEPA 1993a) 100 H 2 SO 4 Immediately after collection NO 3 -N/NO 2 -N Method (USEPA 1993b) 200 H 2 SO 4 Immediately after collection UV 254 NA NA NA Immediately after collection Free/total chlorine Method (USEPA 1978) NA NA Immediately after collection Turbidity Method 2130B (Standard Methods 2012) NA NA Continuous (sensor) ph Method 4500-H + (Standard Methods 2012) NA NA Immediately after collection Temperature NA NA NA Immediately after collection TOC Method 5310C (Standard Methods 2012) 100 H 2 SO 4, refrigeration Off site, within 28 days Total coliform bacteria Method 9222B (Standard Methods 2012) 100 Sodium thiosulfate, refrigeration Off site, within 24 hours MS2 bacteriophage Method 1602 (USEPA 2001) 100 Sodium thiosulfate, refrigeration Off site, within 24 hours BOD biochemical oxygen demand, H 2 SO 4 sulfuric acid, NA not applicable, NH 3 -N ammonia-nitrogen, NO 2 -N nitrite-nitrogen, NO 3-N nitrate-nitrogen, TOC total organic carbon, USEPA US Environmental Protection Agency, UV 254 ultraviolet absorbance at 254 nm a Modified method to allow for continuous in situ monitoring preserved with sodium thiosulfate pellets, cooled in ice, and sent off site for microbial analysis. At the end of the test, the chlorine pump was turned off with the MS2 injection still on, and triplicate 100-mL samples were taken from the first baffle space. This allowed for an accurate measurement of the initial concentration unaffected by chlorine. One full virus-seeding test was run at each flow rate, and two additional tests were started but not completed because of problems with the setup (loss of chlorine feed or insufficient MS2 dose). Some data from these latter tests were used before the test was stopped. Each full virus-seeding test generated a total of 34 sample bottles for MS2 and coliform analysis. The 34 samples comprised duplicate samples from the MS2 stock tank at the beginning of the test and duplicate samples at the end of the test, for a total of four samples; triplicate samples of the tertiary effluent for background concentrations; triplicate samples from the first baffle space (both with and without chlorine feed) as well as from the first three intermediate sample points, for a total of 15 samples; and two sets of triplicate samples from the last intermediate sample point and the last baffle space, for a total of 12 samples (these double sets of triplicate samples were generated to confirm effective disinfection in the water leaving the contactor). Analytical methods. Quantification of fluoride during the tracer testing used a modified version of Method 9214 (USEPA 1996), as previously demonstrated for full-scale tracer testing (Loux 2011). This required calibration of the ISE with a series of standard solutions in the expected range of the measurement (e.g., 0 to 60 mg/l) each day that testing was performed. In the case of the current study, fluoride standard solutions were made up in a matrix of tertiary wastewater from the plant so that a calibration curve could be developed for each probe, specific to the conditions of the CWRF and in the absence of an ionic strength adjustment buffer. Table 2 summarizes the analytical methods used in this study. Some analyses were conducted continuously (with sensors). Some analyses were performed by off-site laboratories, and the remainder were performed at the CWRF laboratory, either immediately or within one day of collection. RESULTS AND DISCUSSION Measured contact time and free chlorine residual. During tracer testing, the modal contact time (corresponding to the peak tracer concentration) was observed to be very close to the theoretical value at the end of the contactor for all flow rates, which indicates effective baffling (Table 3). In the chlorine decay testing, free chlorine was monitored along the contactor at chlorine doses of 6, 8, and 10 mg/l. This helped to characterize the chlorine demand and decay characteristics TABLE 3 Test Summary of tracer test results Flow gpm Theoretical Contact Time min Modal Contact Time min Low flow Medium flow Medium-high flow High flow 1,

6 E23 TABLE 4 Conditions during each virus-seeding test Filtered Water Average Conditions C T Value Test NH 3 -N mg/l Turbidity ntu UV 254 ph Temperature C Free Cl 2 Residual mg/l Final C T mg-min/l Low flow Medium flow High flow C T concentration times time, Cl 2 chlorine, NH 3 -N ammonia-nitrogen, UV 254 ultraviolet absorbance at 254 nm within the contactor under different flow conditions. The measured chlorine demand/decay ranged from 3.8 to 8.5 mg/l, leaving a final free residual from 0.7 to 5.3 mg/l. This test showed that an adequate free chlorine residual could be maintained in the contactor while dosing between 6 and 10 mg/l of chlorine. This was made possible by effective nitrification and suspended solids removal by the CWRF. With low total organic carbon (2 4 mg/l), biochemical oxygen demand (1 3 mg/l), ammonia (<0.04 mg/l), UV absorbance ( ), and nitrite (<0.01 mg/l) in the filtered water, the variability in chlorine decay was largely the result of weather conditions and residence time. Increased sunshine led to FIGURE 4 MS2 pfu/ml, Total Coliform cfu/100 ml, and C T mg-min/l Virus and coliform results for high-flow test MS2 (dosed into reactor) Coliform (in situ) C T value Distance Along Contactor ft C T concentration times time increased chlorine loss attributable to UV oxidation; longer residence time led to increased decay because of an increased time for kinetically limited reactions between the chlorine and dissolved organics to progress. Results for virus-seeding study. During the virus-seeding tests, the chlorine contactor was monitored to quantify conditions that affect the extent of disinfection. Table 4 presents the average temperature and ph along the reactor during each virus-seeding test as well as the free chlorine residual and C T value at the end of the contactor. The low ammonia and turbidity values indicated effective nitrification and solids removal during each test. For each sampling point, the geometric mean was calculated from the triplicate sampling data for MS2 bacteriophage (dosed into the reactor) and the native total coliform. The geometric mean gives a count of the numerical concentration of MS2 and coliform, which allows log inactivation to be calculated as shown in Eq 1: pc* = log N T N0 where pc* is the log inactivation at a given point, N T is the concentration at the sample point (pfu/ml or cfu/100 ml), and N 0 is the initial concentration at the first baffle space (pfu/ml or cfu/100 ml). In addition, the C T value at each sample point (in units of mg-min/l) was calculated from the free chlorine concentration measured during the test and the observed modal contact time from the tracer study. Figure 4 shows an example plot of the virus and coliform results for the high-flow test. The value on the bar graph is the geometric mean of the triplicate samples at each point, and the error bars show the standard error. Table 5 summarizes the observed extent of inactivation during the virus-seeding studies at each flow rate. Following chlorine addition in each test, all points at the end of the contactor and most intermediate points were below the detection limit, including all 17 data points for MS2 and 16 of 22 data points for coliform. Values in Table 5 and subsequent tables are reported accordingly. The contactor demonstrated compliance with the California requirements for disinfected tertiary recycled water. Log inactivation. The contactor achieved greater than 5.6- log inactivation of MS2 bacteriophage virus at each flow rate, compared with the required 5-log inactivation. Coliform count. The final effluent contained <1.7 cfu/100 ml at each flow rate, which is less than 2.2 MPN/100 ml (1)

7 E24 TABLE 5 Test Extent of MS2 and coliform inactivation at the end of the contactor Date MS2 Inactivation log Coliform Inactivation log Final Coliform cfu/100 ml Low flow 8/4/2014 >5.66 >2.78 <1.1 Medium flow 7/23/2014 >5.73 >2.89 <1.7 High flow 9/2/2014 >6.14 >3.98 <1.2 California standard for recycled water to be used for ediblecrop irrigation under b (CDPH 2014). Disinfection kinetics. For both MS2 and coliform, a normalized kinetic parameter K was calculated between the first and second sample point for each test. In disinfection processes, conventional engineering practice approximates inactivation of microorganisms as a first-order process based on Chick s law and assumes that the rate of disinfection is linearly proportional to free chlorine concentration (Tchobanoglous et al. 2003). Both of these assumptions are implicit in the C T concept. Therefore, the kinetic parameter was normalized to chlorine concentration using Eq 2: TABLE 6 Observed kinetic parameters for inactivation MS2 Virus Total Coliform K = pc* 1 2 C2 T 2 (2) FIGURE 5 MS2 Inactivation log Test Log inactivation Rate (K) log-l/mg-min MS2 inactivation versus C T for all tests Analytical values Measurements below the detection limit C T = 22 mg-min/l Log inactivation 5-log inactivation C T Value mg-min/l C T concentration times time Rate (K) log-l/mg-min Low flow >5.56 > Medium flow >5.92 > High flow >6.14 > K normalized kinetic parameter Each point represents the geometric mean of triplicate samples; points above C T = 100 mg-min/l represent double triplicate samples. where K is the normalized kinetic parameter (log-l/mg-min), pc * 1 2 is the observed log inactivation from the first to the second sample point (log), and C 2 T 2 is the actual C T value at the second sample point (mg-min/l). Table 6 shows the observed inactivation rate of MS2 and coliform for each test. Coliform levels were detectable at the second sample point, and a relatively consistent inactivation rate was observed for coliform in all experiments. For MS2, the fact that the concentration was below the detection limit by the second sample point means that only the lower limit of this kinetic parameter can be reported (as shown in the table). This kinetic parameter was comparable in its order of magnitude to values calculated from the literature, including pilotscale studies (Hirani et al. 2014, Mansell et al. 2008) and bench-scale MS2 inactivation data (Coulliette et al. 2010). The disinfection process used in the current study demonstrated the favorable kinetics of virus inactivation that are possible with free residual chlorine in low-turbidity ( 1 ntu) nitrified water, compared with the much slower kinetics of combined chlorine disinfection that provide the basis for the California recycled water regulations. In addition, the observation that coliform bacteria were inactivated at a slower rate than MS2 was also consistent with expectations for free chlorine disinfection of wastewater effluent (Mansell et al. 2008). Minimum required C T value. Data presented in Table 5 show that a 5-log reduction in MS2 concentration was reached for the three tested flows at the end of the basin. However, it should be noted that this 5-log reduction occurred very early in the disinfection process and therefore at lower C T values than the reductions at the end of the contactor. The minimum required C T must consider both MS2 inactivation and coliform concentration to meet the California regulatory requirements for disinfected tertiary recycled water. To determine the minimum required C T, the data from the virusseeding tests were plotted as MS2 inactivation versus C T (Figure 5) and total coliform count versus C T (Figure 6), including results from all intermediate sample points along the basin. Each point on these figures represents the geometric mean of triplicate samples, and points above C T = 100 mg-min/l represent double triplicate samples. Hollow points on these

8 E25 FIGURE 6 Coliform Count cfu/100 ml C T = 22 mg-min/l C T concentration times time Total coliform versus C T for all tests C T Value mg-min/l 2.2 cfu/100ml Each point represents the geometric mean of triplicate samples; points above C T = 100 mg-min/l represent double triplicate samples. TABLE 7 C T Value mg-min/l Analytical values Measurements below the detection limit Data points near proposed minimum C T value MS2 Virus pc* log pc* log Total Coliform Count cfu/100 ml 18.2 ND >2.3 < > > > > > C T concentration times time, ND no data at a given point, pc* log inactivation at a given point TABLE 8 Matrix Reported C T values for 5-log MS2 inactivation Experiment C T for 5-log Inactivation Reference MBR effluent Bench and pilot tests 30 Hirani et al MBR effluent Pilot test with simulated fiber breakage 30 Mansell et al Well water Bench-scale jar tests 18 Coulliette et al Tertiary effluent Pilot tests, including episodes of high turbidity C T concentration times time, MBR membrane bioreactor 3 Huitric et al figures indicate analytical values, and filled points indicate measurements that were below detection limit. Table 7 shows the data points plotted in Figures 5 and 6 that had C T values around 20 mg-min/l. On the basis of the data in Table 7, a free chlorine C T value of at least 22 mg-min/l is required; this C T value is plotted in Figures 5 and 6. Above this CT value, full compliance with California recycled water disinfection requirements, including the MS2 inactivation (>5 log) and total coliform count (<2.2 MPN/100 ml), was demonstrated at all three flow rates. This minimum C T value is consistent with what would be calculated from the kinetic parameters in Table 6. The C T value for 5-log virus inactivation in this study is also similar to values found by other studies with dispersed virus, as shown in Table 8. Given the difficulty of measuring very low C T values at full scale, it is possible that the minimum required C T may be even lower than reported here under the conditions at the CWRF. This would certainly be consistent with the observation of Huitric and colleagues (2013), who found that sufficient disinfection could be achieved by a free chlorine C T value of only 3 mg-min/l. It is also useful to compare these results with the familiar C T values used for chlorine clearwells in potable water disinfection. Water treatment plants must meet the Surface Water Treatment Rule requirements of 4-log inactivation of viruses and 3-log inactivation of Giardia cysts (USEPA 2014), and they will often achieve 2-log removal of Giardia in the filtration process. Under the conditions in the current study (ph 8, temperature 25 C, and free chlorine residual 1 mg/l), the standard disinfection tables predict a C T value of 2.0 mg-min/l for 4-log virus inactivation and a C T of 19 mg-min/l for 1-log inactivation of Giardia cysts (USEPA 1991). The lower C T value for virus inactivation is consistent with the very rapid kinetics observed by Huitric and colleagues (2013), and the governing C T around 20 mg-min/l represents a similar minimum operating point as was found in this study. CONCLUSIONS The required free chlorine C T value for 5-log inactivation of MS2 bacteriophage was measured in a full-scale water reclamation facility that produces fully nitrified, low-turbidity filtered effluent. A consistent MS2 inactivation rate greater than 0.28 log-l/mg-min was observed for three different flow rates, which is consistent with rates observed in other bench- and pilot-scale studies. At free chlorine C T values of at least 22 mg-min/l, the facility demonstrated full compliance with California recycled water disinfection requirements for MS2 virus and total coliform. These results demonstrated how complete nitrification during secondary treatment and effective solids removal during tertiary treatment allow a full-scale treatment plant to take advantage of favorable disinfection kinetics in clear, ammonia-free water. In such a facility, operating at lower C T values allows for a smaller footprint and lower capital costs to complete the disinfection process. The results of the current study can be used (along with appropriate factors of safety) to inform the design or operating targets for free chlorine disinfection of recycled water, provided that the tertiary water is of sufficient quality.

9 E26 ACKNOWLEDGMENT The authors thank the many individuals who collaborated on this study, including Melanie Holmer, Nathan Griffin, Mauricio Gonzalez, Patrick Stahl, June Lovel, and Mary Portillo at MWH Global; and Keith Kohr, Kevin Wahl, Graham Moland, Robert Barone, and Terry Curson from the Camrosa Water District in Camarillo, Calif. The authors especially thank Zakir Hirani (now with ConocoPhillips) for his work on the test protocol, and Richard Lin (now with Metropolitan Water District of Southern California) for reviewing the manuscript. The Camrosa Water District provided financial support for this work. ENDNOTES 1 WaterMaster, ABB, Stäfa, Switzerland 2 IntelliCAL TM, Hach, Loveland, Colo. 3 BioVir Laboratories, Benicia, Calif. ABOUT THE AUTHORS Michael J. Adelman (to whom correspondence may be addressed) is an environmental engineer with MWH Global, 300 N. Lake Ave., Ste. 400, Pasadena, CA USA; adelmanmj88@gmail.com. He has a BS degree from Lafayette College in Easton, Pa., and an MS degree from Cornell University in Ithaca, N.Y. He is interested in the intersection of theory and practice; his process background includes filtration, sedimentation, disinfection, membrane treatment, ion exchange, bioremediation, and municipal waste composting. Michael Phelps is water quality manager for the Camrosa Water District, Camarillo, Calif. Robert T. Hadacek is an environmental engineer, Oliver R. Slosser is a civil engineer, and Simon Calvet is an environmental engineer with MWH Global in Pasadena. Joan Oppenheimer is principal environmental scientist and James H. Borchardt is water treatment tech director at MWH Global in Pasadena. PEER REVIEW Date of submission: 05/01/2015 Date of acceptance: 07/27/2015 REFERENCES CDPH (California Department of Public Health), Disinfected Tertiary Recycled Water. Division 4: Environmental Health, Regulations Related to Recycled Water. 22 CCR , Sacramento, Calif. Coulliette, A.D.; Peterson, L.A.; Mosberg, J.A.W.; & Rose, J.B., Evaluation of a New Disinfection Approach: Efficacy of Chlorine and Bromine Halogenated Contact Disinfection for Reduction of Viruses and Microcystin Toxin. American Journal of Tropical Medicine and Hygiene, 82:2:279. Dryden, F.D.; Chen, C.L.; & Selna, M.W., Virus Removal in Advanced Wastewater Treatment Systems. Journal of the Water Pollution Control Federation, Hirani, Z.M.; Bukhari, Z.; Oppenheimer, J.; Jjemba, P.; LeChevallier, M.W.; & Jacangelo, J.G., Impact of MBR Cleaning and Breaching on Passage of Selected Microorganisms and Subsequent Inactivation by Free Chlorine. Water Research, 57: Huitric, S.-J.; Munakata, N.; Tang, C.-C.; Kuo, J.; Ackman, P.; Friess, P.; Souza, K.; & Barnard, R., Determining Free Chlorine Residual CT Values to Meet California Title 22 Five-Log Virus Inactivation Requirement. Proc. Water Environmental Federation Technical Exhibition and Conference (WEFTEC) 2013, Chicago. Loux, B., New Technology for Contact Time Certification in Water Reuse Permits. Proc. WateReuse California 2011, Dana Point, Calif. Mansell, B.; Huitric, S.-J.; Munakata, N.; Kuo, J.; Tang, C.-C.; Ackman, P.; Friess, P.; & Selna, M., Disinfection of Membrane Bioreactor Permeate Using Free Chlorine: Virus Inactivation and Disinfection Byproducts Formation. Proc. WEFTEC 2008, Chicago. Tchobanoglous, G.; Burton, F.L.; & Stensel, H.D., Wastewater Engineering: Treatment and Reuse. McGraw-Hill, New York. Rodgers, M. & Butler, T., Effective Volume Factor of 0.9 for the Garrett A. Morgan Water Treatment Plant Reservoir. Proc. AWWA Ohio Section 2010 Annual Conf., Columbus, Ohio. Sigmon, C.; Shin, G.-A.; Mieorg, J.; & Linden, K.G., Establishing Surrogate Virus Relationships for Ozone Disinfection of Wastewater. Environmental Engineering Science, 32:6:451. Standard Methods for the Examination of Water and Wastewater, 2012 (22nd ed.). APHA, AWWA, & WEF, Washington. Tree, J.A.; Adams, M.R.; & Lees, D.N., Virus Inactivation During Disinfection of Wastewater by Chlorination and UV Irradiation and the Efficacy of F+ Bacteriophage as a Viral Indicator. Water Science and Technology, 35:11 12:227. USEPA (US Environmental Protection Agency), Drinking Water Contaminants. water.epa.gov/drink/contaminants/ (accessed Apr. 23, 2015). USEPA, Method 1602: Male-Specific (F+) and Somatic Coliphage in Water by Single Agar Layer (SAL) Procedure. Office of Water, Washington. USEPA, Method Potentiometric Determination of Fluoride in Aqueous Samples With Ion-Selective Electrode. Office of Water, Washington. USEPA, 1993a. Method Rev Determination of Ammonia Nitrogen by Semi-Automated Colorimetry. Environmental Monitoring Systems Laboratory, Office of Research and Development, Cincinnati. USEPA, 1993b. Method Rev Determination of Nitrate-Nitrite Nitrogen by Automated Colorimetry. Environmental Monitoring Systems Laboratory, Office of Research and Development, Cincinnati. USEPA, Guidance Manual for Compliance With the Filtration and Disinfection Requirements for Public Water Systems Using Surface Water Sources. Office of Drinking Water, Washington. USEPA, Method Chlorine, Total Residual (Spectrophotometric, DPD). Approved for National Pollutant Discharge Elimination System. Office of Water, Washington. Weber-Shirk, M.L. & Lion, L.W., Flocculation Model and Collision Potential for Reactors With Flows Characterized by High Peclet Numbers. Water Research, 44:18:5180.