Challenges of CSO Disinfection by Medium Pressure UV Light

Objectives:
This presentation will cover the operation, monitoring, and maintenance challenges from implementing medium pressure ultraviolet (MP UV) light for disinfection of filter effluent and primary effluent during wet weather conditions of a Combined Sewer Overflow (CSO) Wastewater Treatment Plant (WWTP). The use of MP UV light to disinfect filter effluent and primary effluent is a unique application of this technology given the challenges to ensure adequate performance; only a few facilities worldwide have implemented this treatment approach. This presentation offers insightful information relevant for wet weather treatment process technologies and strategies.
Introduction: The City of Richmond Long Term Control Plan to reduce CSOs identified that increasing the excess Combined Sewer Overflow wet weather treatment at the WWTP was the most cost-effective method to improve the James River bacteriological water quality. For this purpose, implementation of a CSO Wet Weather UV disinfection facility was designed. These improvements allow the Plant to increase its flow treatment capacity from 75 MGD to 140 MGD during wet weather conditions (Figure 1). The construction of these facility improvements has been finalized, a certificate to operate was issued in April 2021, and currently the system uses medium pressure UV light to disinfect filter effluent and primary effluent during wet weather conditions. Originally, the UV Building at the Richmond WWTP had three UV Reactors installed in UV Channels No. 1, 2 and 3 that are designed to disinfect filter effluent (Figure 2). During normal operation of the new wet weather UV system, primary effluent passes through three additional reactors in the UV Building (reactors 4, 5 and 6 in Figure 2). The UV system automatically turns on and off UV reactors based on the flow (dry weather or wet weather flow) and the required UV dose. High E. coli counts were detected early in September 2021 in the UV reactors that disinfect filter effluent. There was concern that the high E. coli counts would exceed the discharge limit in the City's NPDES permit (126 N/100 mL, monthly geometric mean). Plant staff have checked the operation of the disinfection system and found it to be operating normally. The lamp hours for all filter effluent reactors were at about 6,000 hours, which is ~ 50% of their expected useful life. An investigation was developed to determine the possible causes of the high E. coli counts. The Plant is monitoring the performance of the UV system through microbial inactivation analysis to corroborate compliance with permit requirements. This presentation will focus the discussion on lessons learned from the operational, monitoring, and maintenance challenges during normal operation of the medium pressure UV reactors to disinfect filter effluent and primary effluent. The lessons learned are transferable and usable at other facilities and situations involving UV disinfection. Methods: An outside laboratory analyzed duplicate effluent samples for comparison with the City's laboratory E. coli counts to confirm this was not a cross-contamination issue. A visual inspection of the sleeves and UV lamps for the 3 filter effluent UV reactors was performed to identify if fouling on the sleeves is interfering with the delivery of UV radiation for water disinfection. The City suspects that a deposition on the sleeves is iron from the ferric chloride feed for phosphorous removal. The City collected samples from the sleeves build-up and sent it to an external lab for iron concentration analysis. The UV dose calculation and UV intensity monitoring system were also inspected. Key parameters driving disinfection performance of this system are Total Suspended Solids (TSS) and microbial content. UV transmittance (%UVT) determines the UV intensities needed to provide the required UV dose (mJ/cm2) to achieve the target microbial inactivation.
Preliminary Results: The City began to observe an increase in E. coli counts on September 2nd, 2021. Usually counts are in the single digits (<10 N/100 mL). On September 2nd the counts increased to double digits and continued increasing to triple digits until September 18th. On September 18th the City decided to set the UV reactors in operation (reactors 1 and 3) to 100% power. This increased the UV dose to 120 mJ/cm2; usually the target UV dose is set to 30 mJ/cm2. The reactors remained with the 100% power setting until September 21st at this time the target UV dose setting was adjusted to 60 mJ/cm2 and by September 28th, the City adjusted the UV dose back to the usual 30 mJ/cm2. The E. coli counts went back to the single digits after these adjustments. Figure 3 shows the E. coli counts from Outfall 101 during September. Note that the vertical axis is in logarithmic scale. The dates that do not show a bar had an E. coli count of 1 N/100 mL. Figures 4 and 5 show the UV Transmittance (UVT) readings during September 3rd to September 17th. The UV reactors for the filter effluent are designed for a minimum UVT of 65%. Typically, the UVT is in the 70-75% range. For some of the days that the Plant recorded increased E. coli counts the UVT ranged from ~50-60%. However, a good correlation was not observed between the days that the E. coli counts increased and when the UVT decreased. The data suggests that the decreased UVT was picked up by the UVT analyzers and the dose pacing control of the UV system adjusted the lamp power and number of banks/lamps in service to provide the set UV dose (~ 30 mJ/cm2).
Anticipated Conclusions: The preliminary results show that the UV system requires a significant maintenance effort given increased lamp fouling on both filter effluent and primary effluent UV reactors. Banks 1B and 2B from UV reactors 1 and 2, respectively, where the two banks in operation during the days that the plant recorded E. coli counts in the triple digits, September 11th to 18th. These banks showed heavy fouling on the quartz sleeves and stainless-steel components. It is likely that the fouling on the quartz sleeves was affecting the delivery of UV radiation for water disinfection. A UV intensity sensor is installed on each Bank in each UV reactor. The UV Intensity sensors are intended to monitor the status of the sleeves; if the UV intensity decreases, as detected by the UV intensity sensor, it is an indication that the sleeves might need servicing. The UV intensity sensor is a limited monitoring tool, since the sensor is only looking at one lamp as a representation of the whole bank, but it provides relevant information on the status of the quartz sleeves. It is important to regularly provide maintenance/cleaning to this sensor since it can also get build-up/fouling similar to the sleeves.