Turbidity Excursion After a Reservoir Turnover
A surface-water plant draws from a stratified reservoir. In late October, cooling surface water mixes with deeper water; the intake is at 8 m in a 22 m reservoir. Raw turbidity climbs from 3 to 18 NTU over 36 hours, raw manganese from 0.02 to 0.31 mg/L, and raw pH drops from 7.6 to 7.1. The plant runs alum at 28 mg/L, polymer at 0.4 mg/L, coagulation pH 6.8, settled turbidity 2.1 NTU, and filtered turbidity 0.09 NTU. Finished manganese is 0.04 mg/L against a 0.05 mg/L secondary standard. The operator has 8 hours of clearwell at current demand and cannot reduce production because the transmission main has no alternate supply. Jar tests show that at pH 6.8, manganese removal is negligible and settled turbidity rises to 2.9 NTU at the current alum dose.
1. What is the most defensible short-term treatment adjustment, and why is raising the alum dose alone not sufficient?
Read the worked answer
The raw-water shift is a chemical change, not just a solids change: manganese and lower pH mean the coagulation window has moved. Raising alum alone adds acidity and sulfate and will not oxidize soluble Mn(II), so settled turbidity may improve while manganese passes through. The defensible move is to (1) confirm the Mn is soluble (filtered sample, not total), (2) add an oxidant ahead of coagulation - permanganate at a dose just below the point where pink color appears in the applied water, or chlorine if DBP precursor and ammonia levels allow - and (3) re-optimize coagulant pH with the jar test rather than blindly increasing alum. Because clearwell is only 8 hours and there is no alternate supply, the operator should stage the change, monitor filtered Mn and turbidity at least hourly, and have a contingency (blending, reduced rate, or public notice) ready if filtered Mn exceeds 0.05 mg/L.
2. The plant has no alternate supply. What monitoring and decision points should be set before the change is made?
Read the worked answer
Set explicit trigger points before acting, not after. Reasonable ones: filtered turbidity action level at 0.10 NTU (individual filter) and 0.30 NTU (combined); filtered manganese at 0.05 mg/L; permanganate residual checked by visual pink in a settled sample; and chlorine demand re-checked because the oxidant competes with the coagulant. Because clearwell is 8 hours, any upset reaches the distribution within one shift, so monitoring frequency should match that lag - hourly filtered turbidity and Mn, and a distribution Mn/turbidity check at the clearwell outlet. The decision tree: if filtered Mn exceeds 0.05 mg/L after two hours of the oxidant change, escalate to reduced production rate or blending; if turbidity exceeds 0.30 NTU combined, stop the change and revert. Document the raw-water trend so the next turnover is anticipated rather than reacted to.