2.11 Process Monitoring, Trend Interpretation & Plant Optimization
Key Takeaways
- Trend direction and rate of change carry more diagnostic information than any single instantaneous reading, which is why operators evaluate charts and graphs rather than spot values.
- A settled-water turbidity above about 2 NTU on a conventional plant points at coagulation or sedimentation, not at the filters, and fixing the filters will not correct it.
- Partnership for Safe Water optimization goals set individual filter effluent at or below 0.10 NTU in 95 percent of readings with a maximum of 0.30 NTU, well below the regulatory 0.3 and 1.0 NTU limits.
- A slow steady drift usually means a consumable is changing such as media, membrane, or chemical strength, while a step change points at an equipment state change such as a valve, pump, or setpoint.
- Chemical and energy optimization must always be validated against turbidity, disinfection CT, and residual compliance before a reduced dose is accepted as the new normal.
Process Monitoring, Trend Interpretation & Plant Optimization
Two of the highest-value job tasks in the ABC Water Treatment outline are "identify trends and abnormal operation in plant processes by interpreting data from gauges, meters, charts, and graphs" and "make appropriate changes in plant processes to optimize performance and efficiency." Both are application-level tasks - the exam gives you data and asks what it means.
1. The Daily Monitoring Set
A conventional surface water plant should have, at minimum, the following data recorded every shift or continuously:
| Location | Parameter | Why it matters |
|---|---|---|
| Raw water | Turbidity, temperature, pH, alkalinity, TOC or UV254, color | Sets coagulant type and dose; temperature drives reaction rate |
| Rapid mix | Chemical feed rates, mixer amps | Confirms dose actually delivered |
| Flocculation | Paddle speed, visual floc size | Floc quality is the earliest predictor of settled turbidity |
| Settled water | Turbidity, pH | The single best measure of whether coagulation is working |
| Filter influent | Turbidity, filter aid dose | Loading rate on the filters |
| Individual filter effluent | Turbidity (continuous), headloss, flow rate, run time | Pathogen barrier performance |
| Combined filter effluent | Turbidity | Compliance |
| Clearwell | Chlorine residual, pH, temperature, level | CT compliance |
| Entry point to distribution | Residual, fluoride, orthophosphate | Distribution water quality |
| Plant-wide | Flow in and out, chemical usage, power | Efficiency and mass balance |
2. Unit-Process Benchmarks
These are the numbers that let you say which unit is failing, rather than that the plant is failing. They come from the composite correction program methodology used in comprehensive performance evaluations.
| Point | Good performance | Interpretation if exceeded |
|---|---|---|
| Settled water turbidity | Below 2 NTU (below 1 NTU is excellent) at raw turbidity under 100 NTU | Coagulation chemistry or sedimentation hydraulics |
| Filter influent turbidity | Below 2 NTU | Filters are being asked to do the clarifier's job; runs will shorten |
| Individual filter effluent, steady state | Below 0.10 NTU | Media, ripening, or floc strength |
| IFE peak after backwash | Below 0.30 NTU and recovering within 15 minutes | Ripening or backwash sequence |
| Filter run length | 40 to 100 hours | Short runs mean high loading or weak floc |
| Unit filter run volume (UFRV) | Above 5,000 gal/sq ft per run | Below 3,000 signals a chemistry problem |
| Backwash water | Under 4 percent of production | Above suggests over-washing or short runs |
Unit filter run volume is a useful compact metric: UFRV = filtration rate (gpm/sq ft) x run time (min). A filter running at 3.0 gpm/sq ft for 48 hours yields 3.0 x 2,880 = 8,640 gal/sq ft - healthy.
3. Reading a Trend
The mistake novices make is to react to a single reading. The exam rewards reading the shape of the trace.
| Trend shape | Most likely meaning | First checks |
|---|---|---|
| Step change (abrupt, then flat at a new level) | Something switched state | Valve position, pump start/stop, setpoint change, chemical tank changeover, analyzer recalibration |
| Slow monotonic drift over days or weeks | A consumable is changing | Chemical strength decaying, media fouling, membrane fouling, probe drift, seasonal temperature |
| Diurnal cycle | Demand-driven | Flow rate changing loading rates and detention times through the day |
| Sawtooth | Cyclic process | Backwash cycles, batch chemical makeup, pump alternation |
| Increasing noise with unchanged mean | Instrument or sample line problem | Air in the sample line, fouled optics, failing probe, low sample flow |
| Flat line at a plausible value | Frozen signal | Clogged sample line, dead analyzer, held SCADA value - the most dangerous failure because it looks normal |
Verify before you act. A turbidity spike on one instrument with no corroborating change in headloss, flow, or chemical feed is an instrument problem until a grab sample proves otherwise. Conversely, a credible excursion supported by two independent signals demands action immediately, not after further study.
4. Troubleshooting Logic
Settled water turbidity climbing
- Did raw water change? Turbidity, temperature, pH, alkalinity, or color. A cold-water event alone can double the required coagulant dose because hydrolysis and floc formation slow with temperature.
- Is the dose actually being delivered? Check day tank drawdown, not the pump dial. Confirm chemical strength, especially with a new delivery.
- Is the pH in the coagulation window? Alum is effective roughly 5.8 to 7.5; ferric across 4 to 11. Coagulant hydrolysis consumes alkalinity, and low-alkalinity water can be driven out of the window by the dose itself.
- Run a jar test. Bracket the current dose above and below, and test a pH adjustment at the best dose.
- Check hydraulics. Short-circuiting from wind, density currents from temperature differences, or accumulated sludge reducing effective basin volume.
Chlorine residual falling at constant dose
- Rising demand: raw water organics, ammonia, iron, manganese, sulfide, or nitrite.
- Falling feed: gas ejector fouling, hypochlorite decay, pump losing prime.
- Rising flow: dose is per unit volume; if control is manual and flow doubles, the dose halves.
- Nitrification in a chloraminated distribution system - accompanied by rising nitrite and falling total chlorine in storage.
Filter runs suddenly shortening
- Higher filter influent turbidity - a clarifier problem, not a filter problem.
- Overdosed polymer or coagulant creating a surface mat.
- Air binding - dissolved gas coming out of solution in the bed, common when cold saturated water warms; shows as high headloss with low turbidity.
- Media loss or mudballs reducing effective bed volume.
5. Optimization Targets Beyond Compliance
The Partnership for Safe Water optimization goals set performance well below regulatory limits, because regulatory limits are the failure threshold, not the operating target:
- Settled water turbidity at or below 1.0 NTU (raw under 10 NTU) or at or below 2.0 NTU (raw above 10 NTU), in 95 percent of readings.
- Individual filter effluent at or below 0.10 NTU in 95 percent of readings, excluding the 15-minute post-backwash period.
- Maximum IFE at or below 0.30 NTU.
- No IFE spike above 0.30 NTU after backwash.
Compare: the regulation permits 0.3 NTU in 95 percent of CFE readings with a 1.0 NTU ceiling. The optimization goal is roughly a factor of three tighter, and it is set at the individual filter rather than the combined effluent - because a combined reading averages away one bad filter.
6. Energy and Chemical Optimization
The ABC outline lists "optimize the use of energy and chemicals" as an administrative task, and it is one of the few places an operator directly moves the utility's budget.
Chemical levers
- Enhanced coagulation at the true optimum, verified by jar test each season, not by a dose that has not changed in five years.
- Coagulant selection. Polyaluminum chloride consumes far less alkalinity than alum and performs better in cold water, often eliminating a supplemental base feed.
- Moving the chlorine application point downstream to cut both dose and DBP formation.
- Right-sizing filter aid. Too much polymer shortens runs and increases backwash water.
Energy levers
- Pump efficiency testing and trimming or replacing impellers that have drifted off the best efficiency point.
- Variable frequency drives where flow varies - power varies with the cube of speed, so a 20 percent speed reduction cuts power by about half.
- Off-peak pumping into storage where the utility's rate structure has demand charges.
- Backwash water recovery, which recovers both the water and the energy already spent treating it.
The discipline that makes optimization safe: change one variable at a time, hold it long enough to see the full effect through the process detention time, and validate against turbidity, CT compliance, and distribution residual before adopting the new setpoint. A chemical saving that costs a log of Giardia inactivation is not a saving.
A conventional plant reports combined filter effluent turbidity of 0.12 NTU, but settled water turbidity has climbed from 1.4 NTU to 5.8 NTU over three days while filter run lengths dropped from 60 hours to 22 hours. Where should the operator focus?
A dissolved oxygen analyzer trend shows a perfectly flat line at 6.2 mg/L for 36 hours while plant flow, temperature, and blower output have all varied substantially. What should the operator suspect first?
How do the Partnership for Safe Water optimization goals for individual filter effluent turbidity compare with the regulatory turbidity requirements under the Surface Water Treatment Rules?