Storm Event at a Low-Alkalinity Conventional Plant
Your 4.2-mgd conventional plant (6-mgd design) treats a low-alkalinity river with alum coagulation, sedimentation, and six dual-media filters at 4 gpm/sf. Baseline raw water is 12 NTU, pH 7.6, alkalinity 85 mg/L, 8 °C, and alum runs 32 mg/L with a settled-water target under 2 NTU. Two days of rain change everything: raw turbidity 65 NTU, pH 7.1, alkalinity 42 mg/L, water temperature 9 °C. Settled water is now 4.5 NTU, filter runs have fallen from 55 hours to 20 hours, and combined filter effluent is 0.24 NTU while one filter reads 0.45 NTU. Clearwell free chlorine is 0.7 mg/L, finished pH is 7.0, and the superintendent wants to 'just double the alum.'
1. Before changing any chemical feed, what information do you need, and what is your first process adjustment?
Read the worked answer
**Do not double the alum blindly.
1. Run the alkalinity budget. Alum consumes roughly 0.45–0.5 mg/L of alkalinity per mg/L of dose. At 32 mg/L you consume about 15 mg/L of the 42 mg/L available, leaving ~27 mg/L. Doubling to 64 mg/L consumes ~30 mg/L, leaving ~12 mg/L and driving post-alum pH toward 5.8–6.2. Below roughly pH 6.0–6.5, aluminum solubility rises, floc gets light and poor-settling, and you send aggressive water to the filters and the distribution system. The plant would get worse, not better.
2. Jar test before you touch the feed. Run a dose series (roughly 20–70 mg/L alum) and a parallel series with a cationic polymer as a filter aid; measure settled turbidity, settled pH, floc size and settle rate, and filtered turbidity - not settled turbidity alone. Turbidity removal after a storm is often limited by particle character, not coagulant dose.
3. Likely first move: a moderate alum increase (e.g., 32 → 42–48 mg/L), 0.05–0.1 mg/L polymer as a filter aid, and caustic soda or soda ash to hold post-sedimentation pH at 6.8–7.2 so the coagulant keeps working and you do not ship corrosive water. Confirm the change with a second jar test and with settled-water turbidity at the plant.
4. Also check: raw water manganese and organics (a storm can spike both), whether a PAC feed is warranted, and whether the plant should reduce production to lower filter loading while the floc is fragile.
2. Combined filter effluent is 0.24 NTU and one filter is at 0.45 NTU. Is this a compliance problem, and what do you do?
Read the worked answer
Not a violation today, but it is a warning.
- Combined filter effluent is 0.24 NTU - under the 0.3 NTU level that must be met in at least 95% of monthly measurements, and far under the 1 NTU absolute ceiling.
- The individual filter at 0.45 NTU is under the 1.0 NTU individual-filter limit (95% of monthly measurements). Confirm the exact triggers in your state's rule, since the self-assessment thresholds use consecutive 15-minute readings and get tighter for a filter that has already exceeded once.
What actually matters is the trend. You are one upset away from a trigger on a unit that was probably running 0.10–0.15 NTU a week ago. Investigate that filter specifically:
- Verify the data - calibrate/compare the turbidimeter, check the sample line for air bubbles or biofilm, and pull the 15-minute data rather than a grab value.
- Filter profile - turbidity versus depth through the bed to see whether the problem is at the surface (floc carryover/blinding) or deeper (media disturbance).
- Backwash quality - at 9 °C, bed expansion for a given backwash rate is higher than in summer; verify rate, duration, air scour, and that you are not washing media over the troughs. Look for mudballs, cracks, and pinholing after the wash.
- Upstream - 4.5 NTU settled water means floc is reaching the filters; a filter cannot polish its way out of that.
Interim measures: put that unit on filter-to-waste if you have it, lower the plant rate, or take the unit out of service for a backwash and inspection. Do not wait for the number to reach 1.0 NTU.
3. How should you manage filter run length and filtration rate for the next several days, and what are the tradeoffs?
Read the worked answer
Shorten runs deliberately rather than waiting for breakthrough.
At 65 NTU raw and 4.5 NTU settled, solids loading is 3–4 times normal. The dangerous failure mode is not terminal headloss - it is turbidity and particle breakthrough while the filter still has headloss capacity. Particles pass before the effluent turbidimeter climbs, which is why you cannot manage this on turbidity alone.
Set conservative triggers: backwash at 15–20% above clean-bed headloss, or when individual filter effluent rises about 0.1 NTU above that unit's clean-bed baseline - whichever comes first. Expect runs of 12–20 hours.
Tradeoffs to manage:
- Washwater. Runs of 20 hours instead of 55 roughly triple backwash frequency, which means roughly 2–3 times the washwater volume. Verify the clearwell and washwater tank can supply it without cutting into contact time (CT), finished-water storage, or fire protection. If the recovery basin cannot keep up, you cannot simply backwash more.
- Rate. Dropping the filtration rate from 4 gpm/sf to about 3 gpm/sf improves particle capture substantially. If demand allows, reduce plant output and draw from storage.
- Filter-to-waste on the worst unit costs you finished water but protects the combined effluent.
- Documentation. Log the shortened runs, the backwash justification, and any filter-to-waste periods. If an individual-filter trigger is hit, you will need that record for the self-assessment report to the primacy agency.