The Town of Essex in southwestern Ontario upgraded its facultative lagoon treatment systems with an alternative wastewater treatment plant — a continuous-fill, intermittent-discharge sequencing batch reactor (CFID-SBR), targeting even higher effluent quality.

The new treatment plant is composed of three CFID-SBR reactors with a total average capacity of 1.2 mgd (flow rate estimated for the next 20 years). Provision has been made to expand plant capacity by 33 percent by building a fourth reactor to treat an ultimate design flow of 1.6 mgd.

The plant of Essex Sewage Works was commissioned in early January 2006. Since then, the CFID-SBR system has consistently met its effluent requirements. Six months after a successful startup of the first two reactors, process-performance testing began. The warm-period testing started in the last week of July and continued through October, followed by cold-period testing from November through February.

Having passed the two testing periods with two reactors in operation, the testing continued at higher flow due to inflow and infiltration (I&I) through March and April 2007 with only one reactor in operation. The system demonstrated that it could efficiently handle a constantly changing flow with high-rate I&I.

Premier Tech Aqua designed the CFID-SBR process and provided major process equipment. Stantec Consulting of Windsor, Ont., designed the plant and provided all engineering for construction. The plant is operated by Ontario Clean Water Agency.

 

SBR design approach

The main treatment objectives for the CFID-SBR were to reduce BOD5, TSS, ammonia and total phosphorus. Chemically assisted phosphorus removal ensures consistent compliance with the total phosphorus requirement.

Biological process simulation using BioWin (EnviroSim Associates) as a platform was part of the process design. The simulation considered the cycle time and the particular hydraulic management of the CFID-SBR and followed the variation of the concentrations of pollutants being treated.

The scenarios that were checked showed the CFID-SBR response first to the effect of low temperature (46 degrees F) compared to the design temperature (59 degrees F). The effluent quality for ammonia and BOD remained below the limits.

The second effect assumed 50 percent higher flow per reactor, lasting one week, as would occur if one of the three reactors were suddenly put out of service for one week. All the parameters were kept as design conditions at normal average flow. The simulation results confirmed that the CFID-SBR system as designed would meet the targeted effluent criteria.

 

Field data and results

Daily precipitation data for the Windsor area (from Environment Canada record) showed a perfect match between precipitation spikes and daily flow peaks throughout the test period (Figure 1). For March and April 2007, the one-reactor configuration faced a continuous, extremely high hydraulic load, most likely due to I&I.

The average design flow per reactor was exceeded by more than 70 percent, and the maximum treated daily volume was on average 4.6 times the design average daily flow. The hourly peak flow reached 40 gps, 8.5 times the average design flow per reactor.

The influent pumps were equipped with variable-frequency drives. As the influent intensity progressively increased and decreased, the CFID-SBR sequence did not simply jump between two modes (normal and storm). Instead, it continuously and smoothly adjusted every five minutes to the variation of inflow. Treatment steps were automatically and progressively shortened or extended depending on influent variations. Consequently, the system was able to optimize the process equipment at all times.

The SBR mixed liquor temperature was significantly affected by influent temperature, and both the SBR and influent temperatures were affected by ambient temperature.

Average monthly ambient temperatures for March and April were 40 and 47 degrees F. Because the SBR was roofless and exposed to the weather, influent and ambient temperatures during those months contributed to cooling down the SBR, which tends to hinder treatment performance. Although January and February were the coldest months, the impact on the SBR was less visible because of low precipitation, thus low I&I.

With excessive incoming flow, the time for the treatment cycle is reduced considerably, meaning the Static-Fill step is skipped on a regular basis and the Fill-Settle step is reduced progressively. This may well have contributed greatly to gradually diminishing the effectiveness of the biological selector effect, thus raising the sludge volume index number.

Analyses of effluent TSS removal (Figure 2) for March and April confirmed the suspicion about settling deterioration under stress conditions. Total phosphorus in the effluent displayed the same trend as TSS despite alum dosing. Chemical sludge made up about 8 percent of total sludge, and that may have contributed to maintaining relatively good settling sludge.

Organic matter concentrations expressed as BOD (Figure 3) showed that the plant had no difficulty demonstrating excellent performance. Even with only one reactor in operation, BOD was consistently below 5 mg/l.

Full nitrification was observed. Even at low temperatures, total ammonia in the effluent remained well below the target concentration of 1 mg/l (Figure 4). For the last three months of testing with one reactor in operation, total ammonia did not exceed on average 0.6 mg/l. This means the treatment cycle as engineered provided enough aeration time to complete the nitrification reaction. Another important aspect is that filling with wastewater during settling and decanting did not affect effluent quality, thanks to the baffle wall.

Nitrate (NO3) and nitrite (NO2) concentrations in the influent were monitored regularly. In general, total influent concentration (NO2 + NO3) on average was below 0.25 mg/l, and thus the impact of these constituents on the total balance was not significant.

The percentage removal of total nitrogen during the warm period and with both reactors in operation was 70 percent on average. In the one-reactor configuration, removal dropped to 60 percent. The average BOD loading for March and April was about 440 pounds of BOD per day. The synthesis of new biomass and nitrogen assimilation could not have been responsible for more than 25 percent removal. Therefore, the excess nitrogen removal can only be attributed to the denitrification process.

 

Conclusion

Process-performance testing started in July 2006 and continued through April 2007. The CFID-SBR system performance was excellent and consistently in compliance with the effluent quality criteria suggested for design.

Analyses of BOD5, TSS, ammonia and phosphorus concentrations in the effluent were consistently very low and almost never exceeded the discharge limits. Full nitrification was observed. Inherently, the CFID-SBR system provided partial denitrification despite a lack of mechanical mixing.

The CFID-SBR, with its added flexibility, allowed the treatment process to provide more than satisfactory results in particularly challenging conditions where only one reactor was in operation with a high I&I rate.

 

About the author

K. Khier Chibani, M.Sc., is an SBR process designer with Premier Tech Aqua. He can be reached at khik@premiertech.com.