Coagulation Collapse After a Storm: Same Alum Dose, New Raw Water
A 3.0 MGD conventional surface water plant treats a 900-acre reservoir. On Monday the raw water is 6 NTU, 88 mg/L alkalinity as CaCO3, pH 7.4, and the plant runs 26 mg/L alum with settled turbidity of 1.1 NTU. Overnight a storm delivers 2.6 inches of rain in 14 hours. Tuesday raw water reads 78 NTU, 41 mg/L alkalinity, pH 6.6, and color up to 60 CU. The operator keeps the alum feed at 26 mg/L. Settled turbidity climbs to 7.4 NTU, filter runs fall from 58 hours to 9 hours, and combined filter effluent rises to 0.29 NTU. Finished pH drops to 6.7 and the clearwell free chlorine residual falls from 1.1 to 0.6 mg/L at an unchanged chlorine feed rate. The plant has caustic soda, an anionic polymer feed, and unused filter-to-waste piping. The clearwell provides about 100 minutes of contact time at design flow, and the distribution system has old unlined iron mains and leaded service lines in two neighborhoods.
1. Explain why the unchanged alum dose no longer produces acceptable settling, and describe the adjustments you would make within the first four hours.
Read the worked answer
What changed. The 26 mg/L dose was tuned for 6-10 NTU water. The storm raised particle load roughly tenfold and color to 60 CU, so coagulant demand rose sharply while the dose stayed flat - the plant is now underdosing. Alkalinity fell from 88 to 41 mg/L and raw pH to 6.6; alum also consumes roughly 0.45 mg/L of alkalinity per mg/L of product, so the reaction drives settled pH toward the lower edge of the alum solubility range. The result is light, slow-settling floc, which is exactly what 7.4 NTU settled water shows. The filters are not the problem yet; they are being overloaded.
First four hours. (1) Run jar tests on the new raw water at 26, 34, 42, and 50 mg/L alum plus polymer, reading settled turbidity at 10, 20, and 30 minutes, and pick the dose at the knee of the curve. (2) Feed caustic to hold settled pH around 6.8-7.2; low pH weakens floc and threatens corrosion control downstream. (3) Start the anionic polymer as a floc aid rather than simply overfeeding alum. (4) Slow the filters, extend backwash, and use the filter-to-waste piping. (5) Recheck alkalinity and pH every two hours and log every change.
2. Combined filter effluent has reached 0.29 NTU. Is the plant in violation, and what should you do about filtration right now?
Read the worked answer
The combined filter effluent limit is 0.3 NTU in at least 95 percent of monthly measurements and never above 1.0 NTU. A single 0.29 NTU value is technically compliant but has essentially no margin, and it is part of a trend, not an isolated reading: settled water is 7.4 NTU and runs collapsed to 9 hours. That combination means breakthrough is likely on the next run.
Individual filter effluent is a separate question. Sustained individual filter values near or above the state assessment triggers (commonly 0.5 NTU and 1.0 NTU over consecutive 15-minute readings) can require a filter assessment even when the combined number looks acceptable, so review each filter's effluent and turbidity profile.
Actions: place the worst-performing filter on filter-to-waste; verify backwash rate, duration, bed expansion, and mudball condition; check for coagulant carryover and post-precipitation; hold plant flow down while raw water settles; move to continuous turbidity monitoring with an alarm at 0.25 NTU; and document the event and corrective actions in case an assessment is later required. Treating 0.29 NTU as proof that filtration is working is the mistake.
3. The clearwell residual fell to 0.6 mg/L as pH dropped. Explain the cause and the risk, and give the sequence of corrective actions.
Read the worked answer
The easy diagnosis is the wrong one. Lower pH shifts the chlorine equilibrium toward HOCl, the stronger disinfectant, so a drop from 7.4 to 6.7 would not by itself depress residual. The residual fell because demand rose: 60 CU of color means far more natural organic matter, and unsettled floc plus aluminum carryover consume chlorine before it ever reaches the clearwell.
Sequence. (1) Fix coagulation and settling first; pouring more chlorine into a dirty clearwell only makes THMs and HAAs. (2) After turbidity and color improve, run a dose step test and set a clearwell outlet target near 1.5-2.0 mg/L total chlorine to hold 0.5-1.0 mg/L through distribution and never below the state minimum (commonly 0.2 mg/L). (3) Recompute CT after the pH and temperature change. Lower pH actually reduces required CT for free chlorine, so this is a check, not a crisis, but the 100-minute clearwell must be verified against 4-log virus inactivation at peak flow. (4) Track the DBP consequence: sample TOC now and schedule TTHM/HAA5 per the Stage 2 schedule, and confirm TOC removal credit is not lost if coagulation pH shifts. Residual failures in old iron and leaded areas also raise corrosion and nitrification questions later.