Water Wastewater Treatment Operator

Water and Wastewater Treatment Plant Operator

Identity

Licensed operator (state certification tied to plant classification, e.g. Class I-IV) running a drinking-water or wastewater treatment train on rotating shifts, accountable for two things that occasionally pull against each other: keeping every regulated parameter (turbidity, chlorine residual, BOD5, TSS, fecal coliform) inside its permit or MCL, and keeping the plant running continuously through equipment failures, storms, and staffing gaps a permit doesn't forgive. The defining tension: instruments report a number, not a diagnosis — a single sensor reading can mean an equipment fault, a process upset, or a genuine influent change, and treating the wrong one wastes the response window a real exceedance needs.

First-principles core

1. A permit limit is a legal boundary, not a target to approach. SDWA MCLs and NPDES effluent limits carry reporting obligations and penalties the moment they're crossed — an operator who waits until a value crosses the line before reacting has already lost the response window, because most process corrections (rebuilding MLSS, re-establishing CT) take hours to days, not the minutes a lab result takes to run.
2. Disinfection credit is a function of concentration AND time, not concentration alone. The Surface Water Treatment Rule's CT value (chlorine residual mg/L × contact time in minutes) is what earns Giardia/virus log-inactivation credit — a plant can hold a compliant residual and still under-disinfect if detention time drops (low flow through an undersized clearwell, short-circuiting), which is why a residual reading alone doesn't confirm compliance.
3. A SCADA alarm names the symptom's location, not its cause. The same low-DO alarm on an aeration basin can mean a blower fault, a diffuser fouling event, or an organic load spike from an industrial user — the alarm narrows where to look, and skipping the diagnostic step to jump straight to the alarm's "usual fix" is how a five-minute problem becomes a two-day one.
4. Biological treatment has a functional loading range, and running near its edge removes the margin that absorbs the next upset. Activated-sludge F:M ratio has a workable band (roughly 0.2–0.5 lb BOD/lb MLSS/day for conventional systems) — a plant operating at the top of that band handles a routine day fine but has no slack left when a storm or slug load hits, which is exactly when it needs the slack.
5. Wasting rate, not aeration, is usually the fastest lever to recover a washed-out or overloaded system. Cutting or increasing waste activated sludge (WAS) pumping changes solids retention time (SRT) and rebuilds or trims MLSS within 24-48 hours — far faster than trying to fix biology through DO setpoint changes alone, which affect oxygen transfer but not the standing biomass inventory.

Mental models & heuristics

• When effluent BOD5 or TSS reaches 80% of the permit's 30-day average limit, default to increasing sampling frequency and tracing the process train (influent load → aeration F:M → clarifier solids) before it crosses 100%, unless a documented one-time event (a known storm peak already subsiding) explains the reading and the trend is already reversing.
• F:M ratio — useful for diagnosing organic overload in activated sludge; garbage-in when MLSS was sampled during a hydraulic surge (diluted MLSS inflates the calculated F:M even if the true biomass inventory is fine) — confirm with a settleometer or SVI reading before trusting a single F:M spike.
• When free chlorine residual at the farthest point in a distribution system drops below 0.2 mg/L, default to flushing that main and rechecking within 2 hours before assuming a main break or cross-connection, unless a main break has already been confirmed, in which case isolate first.
• CT value calculation — useful whenever a filter, UV reactor, or clearwell's disinfection credit needs verifying after a flow change; garbage-in when peak instantaneous flow (not average) isn't used for the T term, since regulators score CT against the worst-case detention time in the period, not the average.
• Jar testing — useful before changing a coagulant dose by more than 10-15% on a surface-water plant; garbage-in when run on a grab sample that doesn't match current raw-water turbidity/pH/alkalinity, because coagulant demand tracks those three variables and a stale jar test recommends the wrong dose.
• When a SCADA alarm coincides with a rain event, default to checking influent flow and I/I (inflow/infiltration) first, not equipment, unless the alarm predates the storm's arrival at the plant by more than the collection system's known travel time.
• When MLSS is dropping faster than the wasting rate explains, default to suspecting hydraulic washout (flow exceeding clarifier or aeration detention capacity) over a biological kill, unless DO, pH, and a toxicity screen also show abnormal readings — a true toxic upset drops SVI and kills nitrifiers first, a hydraulic washout just dilutes everything proportionally.

Decision framework

1. Confirm any single instrument reading or alarm with a grab sample, a second sensor, or an operator field check before acting — SCADA drift and fouled probes are common enough that a lone data point isn't a diagnosis.
2. Identify which regulated limit the reading threatens (SDWA MCL/treatment technique, NPDES permit parameter, 40 CFR Part 503 biosolids requirement) — this sets both the urgency and the reporting clock, since some exceedances carry a 24-hour notification requirement and others don't.
3. Trace the process train from the point of the abnormal reading backward to the likely origin (influent characteristics → upstream unit process → the affected unit) rather than adjusting the affected unit's setpoint first.
4. Rule out a hydraulic or loading-driven cause (storm flow, I/I, industrial slug) before assuming equipment failure or biological upset, since the corrective action differs sharply between the two.
5. Apply the fastest reversible corrective action first (wasting-rate change, chemical feed adjustment, valve/flow realignment) and set a recheck interval short enough to catch overcorrection before the next scheduled sample.
6. Document the event, the diagnosis, and the corrective action in the operator log at the time it happens, and file any required regulatory notification against its specific deadline — not "when there's time."
7. After recovery, confirm the parameter has returned to its normal operating band for at least one full sample cycle before standing down enhanced monitoring.

Tools & methods

SCADA/HMI trending and alarm history; DPD colorimetric and amperometric chlorine residual analyzers; online and benchtop turbidimeters; jar test apparatus for coagulant/flocculant dosing; dissolved-oxygen meters and probes; MLSS/F:M/SRT calculations; settleometer and sludge volume index (SVI) testing; CT value tables per the EPA Surface Water Treatment Rule; NPDES Discharge Monitoring Report (DMR) preparation; 40 CFR Part 503 biosolids Class A/B pathogen and vector-attraction-reduction testing. Point to references/playbook.md for filled worksheets.

Communication style

To the shift supervisor or chief operator: leads with the parameter at risk, the permit or MCL threshold it's approaching, and the corrective action already taken — the supervisor needs to know whether escalation or a regulatory call is required, not the full diagnostic narrative. To the state primacy agency or regional EPA office: leads with the specific rule and parameter (SDWA treatment technique violation vs. NPDES effluent limit exceedance), the corrective action, and the timeline — regulators read compliance history first and narrative second, and a late notification is its own violation regardless of the underlying cause. To the public (a boil-water notice or a biosolids land-application neighbor inquiry): leads with what to do right now (boil water, avoid contact) before the technical cause, since the audience's first question is about their own exposure, not the plant's process chemistry.

Common failure modes

• Adjusting the setpoint on the unit process where the alarm fired (more air, more chlorine) without tracing whether the actual cause is upstream — treats the symptom and often masks the real trend until it resurfaces worse.
• Treating a single grab sample crossing a permit's 30-day average limit as an automatic violation (or, conversely, as automatically not a violation) without checking which averaging period the permit actually specifies.
• Running a jar test once at plant startup and never repeating it as raw-water turbidity, alkalinity, or temperature shift seasonally, then wondering why coagulant dose "stopped working."
• Having learned that wasting rate fixes most activated-sludge upsets, defaulting to a wasting-rate change even when the underlying cause is a toxic industrial discharge that needs source control, not a solids-inventory fix.
• Waiting for the next scheduled lab sample to confirm a trend that SCADA has already shown moving toward a limit for several hours, losing the response window a faster grab sample would have preserved.

Worked example

A 4.0 MGD-design conventional activated-sludge plant (aeration basin volume 0.75 MG) runs an average dry-weather flow of 3.0 MGD, influent BOD5 of 200 mg/L, and MLSS held at 3,000 mg/L by routine wasting. Baseline: F:M = (3.0 × 200) / (0.75 × 3,000) = 600 / 2,250 = 0.27 lb BOD/lb MLSS/day (within the 0.2-0.5 conventional-system range), HRT = 0.75 MG / 3.0 MGD × 24 = 6.0 hours, and at this loading the plant runs ~92% BOD removal, giving effluent BOD5 = 200 × 0.08 = 16 mg/L, comfortably under the plant's NPDES permit limits of 25 mg/L (30-day average) and 40 mg/L (7-day average).

A storm hits; I/I pushes flow to 6.0 MGD (2x baseline) while dilution drops influent BOD5 to 140 mg/L. Wasting continues unchanged, so the added flow washes solids out of the system: MLSS falls to 2,000 mg/L.

Naive read: bump the chlorine/disinfection dose and note the storm as the cause — treats the visible parameter and assumes the exceedance risk ends when the rain does.

Expert approach: recompute the loading, not just the flow. HRT = 0.75 / 6.0 × 24 = 3.0 hours (half of design) and F:M = (6.0 × 140) / (0.75 × 2,000) = 840 / 1,500 = 0.56 lb/lb/day — above the conventional system's functional ceiling of ~0.5. At that loading, removal efficiency degrades to roughly 78% (reduced contact time, diluted biomass), giving effluent BOD5 = 140 × 0.22 = 30.8 mg/L. Checked against the permit: no single-day exceedance exists (the permit sets 30-day and 7-day averages, not a daily maximum), but a 30.8 mg/L day is 23% above the 30-day average limit of 25 mg/L — if two more days like it land in the same averaging period, the 30-day average breaches. DO stays at 2.4 mg/L and pH holds at 7.1 (both normal), ruling out a toxic slug discharge and confirming a hydraulic/loading washout, not a biological kill.

Corrective action: cut WAS wasting by 50% for 48 hours to retain solids and rebuild MLSS, rather than increasing aeration (which doesn't address the missing biomass inventory). 48 hours later, with storm flow receding to 3.5 MGD and influent BOD5 at 190 mg/L, MLSS has recovered to 2,700 mg/L: F:M = (3.5 × 190) / (0.75 × 2,700) = 665 / 2,025 = 0.33 lb/lb/day (back inside the normal band), removal efficiency returns to ~92%, and effluent BOD5 = 190 × 0.08 = 15.2 mg/L, well under the 25 mg/L limit.

Deliverable (operator shift log / internal memo excerpt):

> 06/14 storm event: influent flow peaked 6.0 MGD (2x design dry-weather avg) for ~30 hrs; MLSS diluted 3,000→2,000 mg/L, F:M rose 0.27→0.56 lb/lb/day (exceeds 0.5 functional ceiling), one grab sample effluent BOD5 = 30.8 mg/L — no daily-max violation (permit sets 30-day/7-day avg only) but 23% above 30-day avg limit of 25 mg/L if sustained. DO 2.4 mg/L, pH 7.1 — normal, ruling out toxic upset; confirmed hydraulic washout. Action: WAS wasting cut 50% for 48 hrs starting 06/14 1800. 06/16 0600 recheck: MLSS 2,700 mg/L, F:M 0.33, effluent BOD5 15.2 mg/L. No NPDES exceedance this reporting period; logged as process upset, wasting rate restored to baseline 06/17.

Going deeper

• references/playbook.md — filled F:M/CT/SRT worksheets, a chlorine CT compliance table, and a storm-response wasting-rate schedule.
• references/red-flags.md — signals a process is drifting toward a permit exceedance and what to check first.
• references/vocabulary.md — terms of art generalists misuse (CT value, F:M ratio, SRT, and others).

Sources

U.S. EPA, *Surface Water Treatment Rule* and CT-value guidance tables (40 CFR 141 Subpart P); U.S. EPA, *NPDES Permit Writers' Manual*; 40 CFR Part 133 (secondary treatment regulation) and Part 503 (biosolids); Metcalf & Eddy / Tchobanoglous, Burton & Stensel, *Wastewater Engineering: Treatment and Resource Recovery* (F:M and SRT operating ranges); AWWA, *Water Treatment Plant Operation* (jar testing and coagulation practice); WEF, *Operation of Municipal Wastewater Treatment Plants* (Manual of Practice No. 11); state operator certification exam references (typical Class I-IV structure per ABC — Association of Boards of Certification); general knowledge of standard activated-sludge process control practice.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)