Filter Run Collapse After a Storm Event
A 3.5-mgd conventional surface water plant treats a reservoir source with aluminum sulfate coagulation, flocculation, sedimentation, and four dual-media filters of 350 ft² each. Free chlorine is applied ahead of a 250,000-gallon baffled clearwell. On Tuesday a storm pushed raw turbidity from 4 to 45 NTU, raw alkalinity down to 28 mg/L, raw pH 7.6, and water temperature 8 °C. The operator raised the alum dose from 18 to 26 mg/L. Settled turbidity now runs 1.8 NTU against a 1.0 NTU internal goal. Filter runs have fallen from 40 hours to 16 hours, terminal headloss of 6.5 ft arrives in about 13 hours, and filter effluent turbidity climbs from 0.08 to 0.28 NTU late in each run. One filter read 0.9 NTU for two consecutive 15-minute readings two hours after backwash. Clearwell effluent is 1.2 mg/L free chlorine at pH 7.4, and plant flow is being held at 3.5 mgd to keep the clearwell above 12 ft.
1. What is happening in the filters, and what would you change in the next 24 hours? Include at least one calculation.
Read the worked answer
Check the hydraulics before blaming the media. 3.5 mgd is 2,430 gpm spread over 1,400 ft² of filter area, or 1.74 gpm/ft² - comfortably under a typical 4 gpm/ft² design rate. The filters are not overloaded.
The problem is upstream. Settled water at 1.8 NTU is roughly twice the design solids load arriving as fragile, low-density floc that penetrates the bed instead of forming a cake at the surface. That single fact explains both symptoms: short runs, and headloss and turbidity arriving at the same time instead of headloss coming first.
Next 24 hours:
- Run jar tests across a dose/pH/polymer matrix instead of simply adding more alum. At 28 mg/L raw alkalinity, each mg/L of alum consumes about 0.45 mg/L as CaCO3, so 26 mg/L burns roughly 12 mg/L of alkalinity and pulls pH down; confirm the coagulated pH is actually in the optimum band and consider a filter-aid polymer.
- Verify backwash: 20–25% bed expansion, air scour working, no mudballs or media loss.
- The 0.9 NTU after backwash is classic filter ripening. Filter-to-waste until IFE drops below 0.1 NTU.
- Recalibrate all online turbidimeters against grab samples.
2. Your state requires 3-log Giardia inactivation, and the CT table value at pH 7.4 and 8 °C is 129 mg·min/L. With a baffle factor of 0.5, is the plant complying at 3.5 mgd? What are your options?
Read the worked answer
Do the arithmetic, do not assume.
- Detention time = 250,000 gal ÷ 2,430 gpm = 103 min
- T10 = 103 × 0.5 = 51 min
- Actual CT = 1.2 mg/L × 51 min = 61 mg·min/L
- Required CT = 129 mg·min/L
The plant is at about 47% of required CT. That is a treatment technique failure, not a recordkeeping problem.
Options, cheapest and safest first:
- Improve the clearwell. Better inlet/outlet baffling to lift the T10 factor from 0.5 toward 0.7 gives 72 mg·min/L with no chemical change.
- Cut flow through the clearwell during high-turbidity events, which raises detention time directly.
- Raise the free chlorine residual. At these conditions you need roughly 2.5 mg/L to close the gap - but note that storm-driven TOC plus a higher residual means more THM/HAA formation.
- Add a second application point downstream of the filters so you build CT without pushing chlorine through the filters.
Recalculate at peak hourly flow, not at the 3.5 mgd you are holding today. At 4.5 mgd, CT falls to roughly 48 mg·min/L.
3. Two hours later the same filter reads 1.2 NTU on two consecutive 15-minute individual filter effluent samples. What do you do, and what does it trigger?
Read the worked answer
Take that filter out of service now. Do not wait for a third reading. Send it to waste or recycle, then inspect media depth and loss, bed condition, backwash history, valve position, and the turbidimeter itself against a grab sample. A 1.2 NTU IFE shortly after a clean backwash usually means breakthrough through a disturbed or channeled bed, not a meter error.
Regulatory side: individual filter effluent above 1.0 NTU on two consecutive 15-minute samples is an exceedance that requires follow-up - typically a filter profile and a self-assessment, with escalating requirements if it repeats. Separately, combined filter effluent for conventional filtration must be at or below 0.3 NTU in at least 95% of monthly readings and must never exceed 1 NTU. A CFE above 1 NTU is a treatment technique violation with public notice consequences.
Notify the state drinking water program the same day, document the corrective actions and times in the operational log, and compute the CFE 95th percentile before month end. Confirm the exact follow-up trigger and deadline in your state's rule - the 15-minute averaging and the assessment tiers are where people get caught out.