Treating for Enterococci

About the author:

Paul Wood, P.E., is a senior associate at Lockwood, Andrews & Newnam Inc. (LAN), a national planning, engineering and program management firm. He can be reached at [email protected]. 

The city of Corpus Christi Greenwood Wastewater Treatment Plant (WWTP) uses ultraviolet (UV) light for disinfecting its effluent discharge. Originally permitted for  E. coli, the parameter was changed when Texas aligned its discharge parameters to the U.S. EPA discharge criteria. Facilities with ocean or brackish water discharges now use enterococci as the indicator organism for effluent disinfection status.

A new UV disinfection system was required at the Greenwood WWTP due to repeated flooding of the existing UV facility and difficulty in consistently meeting the enterococci limits where it had previously consistently met its previous  E. coli discharge limits.

In July 2015, the city  contracted Lockwood, Andrews & Newnam Inc. (LAN), a national planning, engineering and program management firm, to prepare  a preliminary engineering report detailing the differences in treating for E. coli versus enterococci, along with recommendations to ensure that a future UV installation would meet limits. These included higher applied dose requirements to inactivate enterococci and filtration prior to UV disinfection to ensure nominal particle size of suspended solids passing through the UV system so that UV exposure could be ensured. LAN subsequently was selected to prepare the final engineering documents for system construction.

UV Disinfection Standards

It is known that enterococci generally are more robust and survive against stressors better than E. coli. Additionally, they persist longer and can be transported further in some water environments.

The Recommended Standards For Wastewater Facilities 2014 edition (the Ten State Standards) states: 

”Ultraviolet (UV) disinfection process design, operating data, and experience are developed, but design standards are not well established. Expected performance of the UV disinfection units for the full operating range of flow rates shall be based upon experience at similar full-scale installations or thoroughly documented prototype testing with the particular wastewater. Critical parameters for UV disinfection units are dependent upon manufacturers’ design, lamp selection, tube materials, ballasts, configuration, control systems, and associated appurtenances. Proposals on this disinfection process will be reviewed on a case-by-case basis at the discretion of the reviewing authority under the provisions of Paragraph 53.2.”

In fact, UV design standards are well established. In 1986, the U.S. EPA published UV design guidelines in its Design Manual for Disinfection. In this manual, an estimated 125 wastewater treatment plants used ultraviolet light for disinfection at that time. TrojanUV, which designs and manufactures pressurized and open-channel UV disinfection systems, claims that currently it has more than 10,000 installations worldwide. In 2015, the Water Environment Federation published “Ultraviolet Disinfection for Wastewater, Low-Dose Application Guidance for Secondary and Tertiary Discharges.”

When specifying UV disinfection units, engineers normally require UV manufacturers to provide bioassay validation reports. These reports provide the information needed to determine the size of a specific UV system to provide a required UV dose. A default dosage of 30 milliwatt-seconds per square centimeter (mW-s/cm²)—or millijoules per square centimeter (mJ/cm²), which has become a more common measurement—typically is used for achieving E. coli effluent standards of 126 colony forming units per 100 milliliters (cfu/100 ml). 

Interestingly, this default value finds its origin in the Ten State Standards. Its value assumes that water transmittance is greater than 65% for a light source with 254 nanometers wavelength, and biochemical oxygen demand (BOD) is less than 30 milligrams per liter (mg/L). This value has been used for decades but does not take into account the actual levels of E. coli present that may actually need to be inactivated (i.e. the log reduction required). The actual concentrations seen by the UV system depend upon the wastewater treatment process being used, whether the effluent is simply secondary effluent or if it has received tertiary treatment or if it has been further processed through some sort of filtration. The actual needed dosage may be less than a default 30 mJ/cm². This can only be determined by performing collimated beam tests on the actual wastewater being treated.

Collimated Beam Test

Knowing that enterococci is more robust than E. coli, LAN consulted vendors and journal articles in the UV industry. No direct comparisons between the default dosage for E. coli and enterococci could be found. A “rule of thumb” was that 1.5 times the dosage used for E. coli would be needed to get the enterococci inactivation desired. 

This equated to 45 mJ/cm² to achieve 35 cfu/100 ml of enterococci. One vendor recommended 50 mJ/cm² to achieve 35 cfu/100 ml enterococci and require the effluent to have particle sizes no greater than 10 µ to prevent hideout. To better quantify the actual dose required at the facility, the project team decided that a collimated beam test should be undertaken. Two samples were taken, properly preserved, and sent for analysis. One sample was unfiltered, the second sample was filtered through a clean, unused, cloth media filter material.

The results indicated that the transmittance through the wastewater was in line with the transmittance that was to be specified: 65%. The actual results were 66.4% for the filtered sample and 64.2% for the unfiltered sample. Total suspended solids were 5 mg/L and 9 mg/L, respectively.

For the samples taken, 87.8% of the filtered sample particles were 10 µ or less and 83.4% of the unfiltered particles were 10 µ or less. The filtered sample was subjected to a five-point collimated analysis, and the unfiltered sample was subjected to a six-point collimated beam analysis. 

Results

The collimated beam test demonstrated that a UV dose of 10.43 mJ/cm² was sufficient to provide a 2.9 log inactivation of enterococci on the filtered sample which had an influent count/100 ml of 948 and that a dose of 9.81 mJ/cm² was sufficient to provide a 1.798 log reduction of enterococci on the unfiltered sample with an influent count of 1,577.

The question then becomes what dose is really required to obtain the needed log reduction? Metcalf and Eddy (fourth edition) gives table 2-25 for typical microorganism concentrations in raw wastewater, including enterococci between 104 to 105 cfu/100 ml. Similarly, it provides table 12-2 for removal percentages by different treatment processes including activated sludge at 90 to 98%. In the worst case, 90% removal of 105 leaves 104 cfu/100 ml. To get to 35 cfu/100 ml (assuming an initial count of 50,000 cfu/100 ml), a 3.155-log inactivation is needed. 

Metcalf and Eddy also gives table 12-22, which shows total coliform counts at various points in a wastewater treatment process. No similar study for enterococci could be readily found. 

The raw influent concentration of total coliform is three to four orders of magnitude greater than that for enterococci. Filtered and nitrified effluent counts for total coliforms are in the range of 104 to 106 cfu/100 ml. Assuming reductions across the upstream processes are similar for both total coliforms and enterococci, three orders of magnitude (three log) reduction are expected for enterococci by the biological and filtration processes or 50 to 5,000 cfu/100 ml (conservatively, assuming rounding) for enterococci. Therefore, typical ranges of 0.15 and 2.15 cfu/100 ml are needed for inactivation. For complete inactivation of 50,000 cfu/100 ml, a 3.699 log inactivation is needed.

Conclusion

There was data for one point, with a high count of 1,577 cfu/100 ml. Was this a typical (or more importantly, worst case) value? Unfortunately, the log reduction is dependent on the initial concentration, and initial counts were on the low side of the “typical” range. 

With only one data point, the answer was unknown. The project team did not want to specify a system that would only provide a dose of 10.43 mJ/cm² based on initial discussions with the manufacturer and other sources. A reasonably conservative approach would be to provide a system that could provide 45 mJ/cm². 

For the Greenwood WWTP UV system, this dose of 45 mJ/cm² could be provided with three modules in series. A safer approach would be to design the UV channels to accept a fourth module to allow it to achieve greater than the 50 mJ/cm² dose. 

The city of Corpus Christi ultimately decided that if the UV channel was being built for four modules, then four modules should be installed. This likely is an extremely conservative approach, but until more installations with enterococci are reported on, the project team considered it to be a safe option. The system can be paced, and ongoing influent and effluent bacteria counts can be used to better tune the dosage. This project is expected to begin construction in 2020.