Understanding and mitigating the negative impact of natural organic matter on filtration processes

Abstract

The overall objective of this project was to develop and understand effective coagulation processes for optimizing filtration performance during the treatment of high-TOC concentration raw waters. Specific objectives were to determine the impact of coagulation pH, ζ-potential, and alum, addition model (single vs. two-stage) on floe formation and filtration performance. Another goal was to evaluate a photometric dispersion analyzer (PDA) as an indicator of overall process performance. Two-stage coagulation processes were proposed as a means to mitigate restabilization of floe by natural, organic matter, particularly for high-TOC concentration water. Bench and pilot-scale experiments compared floe development, filtration performance, and ζ-potential distributions of post-rapid mix and settled water particles for single and two-stage coagulation processes over the pH range of 6.0 to 7.4. Overall, in terms of filtration performance during the treatment of high-TOC concentration, waters, single-stage high-pH and two-stage coagulation processes were most effective. The use of relatively high coagulation pH conditions (e.g. 7.5) along with alum doses optimized by ζ-potential optimized, particle removal in the settling process aid significantly increased the operating range, in terms of ζ-potential, for optimum filtration performance. Two-stage coagulation processes also improved filtration performance, evidenced primarily by increased filter run lengths prior to particle breakthrough. Two-stage coagulant addition consistently increased turbidity removal during sedimentation, oftentimes dramatically when compared to single-stage processes using the same alum dose. This result correlated with the significant increase in floe formation rates and floe size consistently observed for the two-stage processes, particularly in high-TOC concentration waters. The fact that two-stage processes did not impact floe formation in low-TOC conditions supported the proposed mechanistic advantage of this process, i.e. mitigation of NOM-induced floe restabilization by allowing NOM-Al(OH)3(S) reactions to occur primarily in the first coagulation stage. The photometric dispersion analyzer (PDA) appears to be a useful alternative to jar tests for determining optimum coagulant doses and quickly evaluating different coagulation scenarios. Overall, processes that the PDA indicated were optimum, in terms of the kinetics of floe development, were also optimum in terms of filtration performance.