Filter Effluent Climbing After a Runoff Event
A 12 MGD conventional surface-water plant treats a river source through coagulation, flocculation, sedimentation, and four dual-media filters. At 06:00 raw turbidity is 4 NTU and alum is fed at 28 mg/L. By 14:00 an upstream thunderstorm pushes raw turbidity to 31 NTU, raw color from 15 to 60 color units, and raw pH from 7.6 down to 7.0. The operator raises the alum dose to 42 mg/L and holds the polymer feed unchanged.
At 16:00 settled-water turbidity is 2.8 NTU (normally 0.8), Filter 3 effluent has climbed from 0.08 to 0.38 NTU over 90 minutes, its headloss is 5.1 ft of the 8.0 ft available, run time is 41 hours, and the other three filters read 0.09 to 0.14 NTU. Plant flow is 9.5 MGD. The 1.5 MG clearwell is 68 percent full, and projected demand for the next eight hours is 12.5 MGD. The state primacy agency enforces a combined filter effluent limit of 0.3 NTU in at least 95 percent of monthly samples, with an individual filter limit of 1.0 NTU at all times.
1. Filter 3 has not reached terminal headloss, but its effluent turbidity is rising. What do you do first, and how do you decide between backwashing and taking the filter out of service?
Read the worked answer
Verify the number before you act on it. Check the turbidimeter against a grab sample and its calibration, and check the sample line for air bubbles or a partial plug. A rise from 0.08 to 0.38 NTU after a raw-water upset is plausible, but so is a fouled flow cell.
Interpret the reading. 0.38 NTU is not a violation by itself: individual filters must stay under 1.0 NTU at all times and under 0.5 NTU 15 minutes after backwash, while the 0.3 NTU limit applies to combined filter effluent in 95 percent of monthly samples. The trend is the problem, and it is happening while the other three filters sit at 0.09 to 0.14 NTU.
Headloss tells you the failure mode. At 5.1 ft of 8.0 ft available after only 41 hours, the filter is not loaded with solids. Rising effluent at low headloss points to material passing through rather than storage capacity running out: coagulant floc shearing through the bed after the coagulation upset, or a media defect such as a boil, crack, mudball, or disturbed underdrain.
Action. Take Filter 3 out of service, backwash, and return it on filter-to-waste if you have it. If post-backwash effluent does not come back below roughly 0.10 NTU and hold there, keep the filter out and inspect the media surface and probe the bed. Record the values and notify the shift supervisor.
2. Forty minutes after Filter 3 returns to service at 0.07 NTU, Filter 1 has risen from 0.11 to 0.29 NTU. What does that pattern tell you, and what do you do across the plant?
Read the worked answer
Two filters failing within an hour of each other is not two filter problems. It points upstream: settled-water turbidity of 2.8 NTU against a normal 0.8 means the filters are being asked to remove solids the sedimentation basin should have captured.
Why did coagulation fail? The raw water changed in three ways at once: turbidity 4 to 31 NTU, color 15 to 60, pH 7.6 to 7.0. Alum was raised from 28 to 42 mg/L by feel, with no jar test and no change to polymer. Higher color and lower pH generally raise coagulant demand sharply, but a dose increase without pH correction can also move you off the optimum coagulation pH zone. A polymer feed failure would produce exactly this picture too, so verify the feed pump and day tank before assuming it is chemistry.
Actions: run jar tests on the current raw water across a range of alum doses and pH values; adjust alum and polymer so settled turbidity comes back under about 1 NTU; trim pH with caustic if the jar tests call for it; reduce plant flow to lower the filter loading rate in gpm/sf while the chemistry is unstable; monitor each filter's effluent hourly instead of every four hours; backwash on turbidity rise rather than waiting for terminal headloss. Track the combined filter effluent running average against the 95 percent rule and keep the state informed if you approach it.
3. The clearwell is 68 percent full and demand is projected at 12.5 MGD for the next eight hours. How does that constrain your decision to cut plant flow, and what is your plan?
Read the worked answer
Do the arithmetic first. 68 percent of 1.5 MG is 1.02 MG in storage. At 9.5 MGD production against 12.5 MGD demand, the system draws down about 3 MGD, or 125,000 gallons per hour, so 1.02 MG lasts roughly eight hours. That is exactly the length of the projected peak, which means storage buys you time to fix coagulation but not enough time to idle filters or run a long, deep reduction in production.
Plan: use the clearwell as a buffer; cut plant rate modestly (toward 8.5 to 9 MGD) to lower filter loading while you re-optimize the dose with jar tests; keep Filter 3 out until its effluent is stable; call in the second operator; monitor clearwell level hourly and set a trigger level, say 40 percent, at which you must restore full production or bring in purchased water. Tell the distribution supervisor so pressure management can be coordinated, and have the low-pressure and precautionary boil notice procedures ready in case pressure drops below 20 psi.
Document the turbidity values, the dose changes, the jar test results, and the decision points. If the combined filter effluent exceeds the limit, that record is what shows the agency you responded rather than ignored it.