Filter Run Ends Early: Turbidity Creep and Backwash Decisions
A conventional surface-water plant (4 MGD design, 2.8 MGD current) has four dual-media filters. Filter 3 has been running 62 hours. Its effluent turbidity rose from 0.08 NTU to 0.31 NTU over the last 6 hours while the other three hold at 0.10–0.14 NTU. Headloss on Filter 3 is 6.1 ft versus 4.8 ft on the others; the terminal headloss setpoint is 8 ft. Raw turbidity jumped from 4 to 11 NTU after a rain, and the coagulant dose was raised 20% this morning. Filter 3 also shows a slightly higher finished-water particle count. The state's combined filter effluent limit is 0.3 NTU (95th percentile) with a maximum of 1 NTU, and individual filter effluent must be monitored.
1. Do you backwash Filter 3 now, and what is your reasoning against the headloss and turbidity triggers?
Read the worked answer
Backwash now. The decision is not driven by headloss - 6.1 ft is below the 8 ft terminal setpoint - but by the individual-filter effluent turbidity trend and the particle count.
Reasoning:
- Effluent turbidity more than tripled (0.08 → 0.31 NTU) in 6 hours while the other filters stayed stable. That pattern points to a developing problem inside Filter 3 (media disturbance, channeling, or a breakthrough from the higher coagulant load), not normal ripening.
- 0.31 NTU is at/above the 0.3 NTU combined limit and is an individual-filter excursion. Even if the combined effluent is still under 0.3, an individual filter trending up is a leading indicator of a compliance problem and a pathogen-barrier concern.
- Particle counts corroborate the turbidity; two independent signals agreeing is stronger than one.
- The 20% coagulant increase after the rain means the filter is seeing a different floc; a filter that was stable at the old dose may be overloaded or under-ripened at the new one.
Do not wait for 8 ft headloss. Headloss is a capacity trigger; turbidity is a water-quality trigger, and water quality governs.
2. What do you check and record during and after the backwash, and what would make you take Filter 3 out of service?
Read the worked answer
During the backwash, verify and log:
- Backwash rate and duration against the design (e.g., target gpm/sf and bed expansion); confirm you actually achieve the expansion needed to clean the bed.
- Backwash water turbidity - a spike that clears is normal; a muddy finish means incomplete cleaning.
- Air scour (if equipped) operation and sequence.
- Media condition - look for media loss to the troughs, mudballs, or a disturbed surface after the wash.
After return to service, watch the filter-to-waste/ripening period: turbidity should fall quickly to the baseline (≤0.10 NTU). A slow or erratic recovery, a repeat rise within a few hours, or a rising particle count means the problem is not just dirty media.
Take Filter 3 out of service if:
- Effluent turbidity cannot be held below the individual limit after a backwash and a short ripening period;
- You see media loss or a damaged underdrain/air grid;
- Headloss rises abnormally fast after a clean backwash (suggests surface blinding or media problems);
- The filter is the source of a combined-effluent exceedance.
Also review the coagulant change: confirm jar tests or streaming-current setpoint match the new dose, because a filter problem that started with a dose change is often a chemistry problem, not a media problem.