Late-October Turbidity Spike: Raising Alum Without a Jar Test
Riverton Municipal Water Works (pop. 9,400) treats 3.5 MGD in a conventional coagulation, flocculation, sedimentation, and dual-media filtration plant drawing from a 620-acre surface reservoir. In the last week of October, reservoir turnover plus 1.8 inches of rain pushed raw turbidity from 4 NTU to a steady 35–42 NTU. Current raw water: pH 7.7, alkalinity 48 mg/L as CaCO3, temperature falling from 68°F to 55°F, TOC 5.2 mg/L. The plant feeds alum at 26 mg/L with no polymer and no pH adjustment. Settled turbidity has climbed from 1.5 to 8 NTU, combined filter effluent is running 0.25–0.45 NTU, filter runs have collapsed from 55 hours to 16 hours, and Filter 3 logged 1.2 NTU for one four-hour reading yesterday. The shift operator wants to raise the alum dose to 45 mg/L and change nothing else.
1. Evaluate the operator's plan to raise alum from 26 to 45 mg/L. Show the alkalinity and pH implications with numbers, and explain what must be verified before the feed rate changes.
Read the worked answer
Alkalinity math. Alum consumes roughly 0.45–0.5 mg/L of alkalinity (as CaCO3) per mg/L applied. At 26 mg/L that is about 12 mg/L consumed, leaving ~36 mg/L and a pH near 7.4–7.5. At 45 mg/L the demand is about 20–22 mg/L, leaving ~26 mg/L and dropping pH by roughly 0.7–1.0 unit, into the low 6s. That is not automatically bad - alum flocculation for turbidity often works well between about 5.5 and 7.0 - but without lime feed you cannot hold a target pH, and you are moving blind.
Why the plan is incomplete.
- Higher dose raises sludge production, shortens runs further, and increases the load on already-struggling filters.
- Overdosing can restabilize particles (charge reversal), so more alum can make settled turbidity worse.
- The raw water changed in three ways at once: turbidity up 10x, TOC 5.2 mg/L (coagulant demand), and temperature down 13°F (slower floc formation, higher viscosity). The fix may be polymer, more mixing/flocculation energy, or lower loading - not just more alum.
- Filter runs of 16 hours with 0.25–0.45 NTU effluent say the filters are receiving solids that sedimentation should have removed.
Before changing feed: run a jar test (never dose a 3.5 MGD plant from a hunch), verify the alkalinity/pH target, confirm rapid-mix and flocculator performance in cold water, and check whether sedimentation overflow rate is the real bottleneck.
2. Filter 3 recorded 1.2 NTU for a four-hour period. What are your immediate actions, and what are the compliance consequences?
Read the worked answer
First, verify. Check the turbidimeter calibration and the sample line, confirm the reading is real, and pull a grab sample. A false 1.2 NTU reading triggers the same paperwork as a real one, so eliminate the easy explanations first.
Then act. With a conventional plant, combined filter effluent must stay at or below 0.3 NTU in at least 95% of monthly measurements and must never exceed 1 NTU. A single required measurement above 1 NTU is an exceedance. Therefore:
- Take Filter 3 out of service or place it on filter-to-waste if piping permits; do not send that water to the clearwell.
- Check the other three filters immediately and increase effluent monitoring frequency to at least every two hours while the raw water is unsettled.
- Confirm the clearwell turbidity; it was 0.14 NTU, which is good news and tells you the excursion was short.
- Consider reducing plant flow to lower filter loading rate until coagulation is under control.
Compliance. Document the reading, the corrective action, and the times. Notify the Kentucky Division of Water as required by your permit. A turbidity MCL violation is normally a Tier 2 public notice, due within 30 days, though you should confirm the tier with the state. Sustained filter performance failures can also trigger a filter assessment or comprehensive performance evaluation, so the record you write today matters later.
3. Design the jar test program: doses, pH points, what to measure, and what result would make you add polymer.
Read the worked answer
Run 1 - dose sweep at raw pH. Six jars at 15, 25, 35, 45, and 55 mg/L plus a control, same rapid mix (about 30 seconds at high G), same flocculation time and speed, same 25–30 minute settling. Because the water is colder, match the flocculation time you actually have in the plant rather than an ideal one.
Run 2 - pH sweep at the best dose. Four or five jars at pH 6.0, 6.5, 7.0, and 7.5, adjusted with lime or acid, to find where the floc settles and filters best. This tells you whether to add lime feed.
Measure: settled turbidity at 10, 20, and 30 minutes; pH and alkalinity before and after; filtered turbidity through a 0.45-micron membrane (this predicts filter performance better than settled turbidity); floc size and settling rate by observation; and residual aluminum if you have the capability.
Decision rule. Choose the dose and pH giving the lowest filtered turbidity with acceptable settled turbidity - not simply the clearest supernatant. If settled turbidity is good (under 2 NTU) but filtered turbidity stays above 0.1–0.15 NTU, or if pin floc carries over and runs stay short, add a polymer (0.5–1.0 mg/L as filter aid or as a flocculant aid) and test it as a final jar in each series. Also verify the polymer is ANSI/NSF Standard 60 certified and confirm the maximum use rate.