ENHANCED PRIMARY TREATMENT: FULL-SCALE PILOT ANSWERS MANY QUESTIONS

The King County Wastewater Treatment Division initiated an investigation in 2001 of the potential use of ferric chloride addition for enhanced primary treatment at the South Treatment Plant in Renton, Washington. The South Treatment Plant treats an annual average flow of 70 million gallons per day (mgd). The objective of the study was to determine if the benefits of enhanced solids removals using ferric chloride (FeCl3) for advanced primary treatment (APT) in the primary clarifiers, and resultant energy savings in the secondary and solids processes, outweighed the cost of purchasing ferric chloride and flocculant aid polymer.The project began in March 2001 with jar testing of various doses of ferric chloride and polymer. The target primary clarifier removal rates were 50 percent biochemical oxygen demand (BOD) and 80 percent total suspended solids (TSS). These removal rates were seen during jar testing at doses of 30 mg/L ferric chloride and higher (the stoichiometric demand based on influent phosphorus concentration was approximately 35 mg/L). However, the cost of the chemicals at a dose of 30 mg/L would likely make enhanced primary treatment infeasible. The study initially targeted a dose of 15 to 20 mg/L ferric chloride, with a dose of 0.2 mg/L anionic polymer to see if some enhancement in the removal rates would provide a benefit to the plant operations budget. Two trial periods were planned: one under partial flow conditions (1 to 4 primary clarifiers) and one under full-flow conditions (8 primary clarifiers).The first trial (2 weeks long) performed in July 2001 (using one primary clarifier at a ferric chloride dose of 15 to 20 mg/L) did not show removal rates near the target levels in the primary clarifiers. However, a change in the influent wastewater characteristics negatively affected plant performance overall, including the ferric chloride trial. The plant influent was monitored over the next few months. Once it returned to historically “normal” conditions, another trial would be scheduled. The second trial (Round 2) was run in winter 2001 at the same doses using four clarifiers. The results were near the target levels for BOD and TSS removal. The conclusion of the second trial was to wait until summer 2002 and run a trial. In addition, a series of settling tests indicated that the target removals could be achieved with less chemical addition, but longer settling time.A trial was run for 6 weeks during July and August 2002. Target doses were 15 to 20 mg/L ferric chloride, with a dose of 0.2 mg/L anionic polymer. TSS removals were not as good as those seen in Round 2 (December 2001), but the BOD removals did meet the target of 50 percent. However, the anticipated decrease in secondary energy use was not observed. Increased cake dewaterability (+1 to 1.5 percent) was measured 30 days after adding ferric chloride (digester solids retention time is 20 to 30 days). This was notable because the cake percent solids had historically dropped over the summer, so the actual increase may have been greater than 1.5 percent. As in the summer 2001 study, several unexplainable performance-related issues at the plant may also have affected the study.The conclusion of the summer 2002 trial was that other factors again seemed to be affecting plant performance. Nonetheless, it was suspected that a higher dose of chemical was needed to achieve the desired performance results. To confirm this hypothesis, a trial was run in October 2002 at a dose of 30 to 35 mg/L, with a polymer dose of 0.2 mg/L. The results were positive. During the October trial, four additional primary clarifiers were put in service (for eight total) to increase the detention time in the primaries. The additional detention time helped improve performance in the primary clarifiers. Primary clarifier TSS removal increased and secondary energy use dropped. However, it was still difficult to know if the variations in the process performance were due to the use of ferric chloride.One final trial was conducted in December 2002 (Round 8) to answer the remaining unanswered question—whether wet weather conditions promote better enhanced primary treatment performance. Unfortunately, the trial was short (3 days) and each day a different chemical dose was added. The findings were inconclusive.In summary, the results derived from the partial flow testing were positive and showed potential annual cost savings if ferric chloride were used for APT at the South Treatment Plant. However, additional data gathered during the subsequent full-flow testing program carried out during 2002 failed to provide the TSS removal rate necessary to develop a cost-effective program. In addition, dosing ferric chloride in the lower range of 10 to 15 mg/L, while more cost-effective, did not provide adequate enhancement to the secondary and solids processes to completely offset the higher chemical costs.Seasonal (winter) use of ferric chloride may provide high removal rates at a lower dose. Not enough data were collected in the Round 8 test to decisively conclude this. However, a definite benefit of the trial was the ability to increase the hydraulic loading rate to the belt filter presses. This will provide operational flexibility later this year when the belt filter presses are replaced with centrifuges.