STP Performance Testing & Acceptance: The Guarantee-Test Guide

An STP can look finished — tanks full, blowers humming, treated water running clear — and still not be doing its job. The only way to know whether a plant actually delivers the water quality it was sold on is to test it, under load, against numbers written into the contract. That test is the moment the plant stops being a promise and becomes an asset. Get it wrong — or skip it — and the buyer inherits a plant that may quietly breach discharge norms for years.

This guide is the acceptance playbook: what to sample, where and how, at what flow, and what happens when the results don't pass.

A performance guarantee test proves one thing and one thing only: that at full design flow, on real sewage, the plant drives BOD, COD and TSS below the guaranteed outlet values — repeatably, not once by luck.

What "performance" actually means here

Performance is defined in the contract, not by opinion. Every STP supply-and-install contract should carry a Performance Guarantee (PG) clause listing the outlet values the plant must achieve. For a domestic STP in India these are pegged to CPCB / State Pollution Control Board discharge and reuse norms, and typically read something like:

The exact figures depend on the reuse or discharge route — water going to toilet flushing or cooling towers may carry tighter turbidity limits than water bound for landscape irrigation. If you are unclear on why these parameters matter or how the raw numbers are set, the primer on what an STP is explains BOD, COD and TSS from first principles.

Test at design flow — not at whatever's arriving

The single most common way a guarantee test is quietly fudged is by running it at a partial load. A newly commissioned plant in a half-occupied building might see only 30–40% of its rated flow. At that gentle loading almost any plant will produce clean water — the biology is barely stressed. That result proves nothing about behaviour at full occupancy.

A valid test must run the plant at or near its rated design flow — the KLD figure the plant was sized for. If real inflow is short, you engineer the load:

• Recirculate treated or raw water to top the equalisation tank up to design flow.
• Feed with tankered sewage or a made-up organic load if occupancy is genuinely low.
• Run the plant at that flow continuously for the test window, not in a brief spurt.

Confirm the flow itself is real: read the inlet flow meter, cross-check against the equalisation tank draw-down and pump run-hours, and log it. If you are unsure the plant was even sized correctly for the building, the STP Capacity Calculator and the sizing guide let you sanity-check the rated KLD before you certify against it.

Sampling: where, how and how often

Bad sampling produces meaningless numbers. Two sample points matter:

• Inlet — raw sewage after the bar screen and grease trap but before biological treatment, so you capture true incoming strength.
• Outlet — the final treated water after filtration and disinfection, at the point it actually leaves for reuse or discharge.

Sampling inlet and outlet together lets you compute percentage removal, which is often a more honest test than the outlet number alone — a plant fed unusually weak sewage can hit outlet targets while removing very little.

Grab vs composite. A single grab sample is a snapshot and swings with the daily load cycle. The defensible method is a 24-hour composite — samples drawn every hour (or flow-proportioned) and pooled — so the result reflects a full day's diurnal peaks and troughs. For a guarantee test, composite sampling over the test period is the norm; grabs are for spot-checks only.

Then handle the samples properly, or the lab result is worthless:

• Use clean, parameter-appropriate bottles; refrigerate at ~4 °C.
• Respect holding times — BOD in particular degrades fast and should reach the lab within hours.
• Maintain a written chain of custody from bottle to bench.
• Test at a NABL-accredited laboratory; the report from an accredited lab is what regulators and arbitrators will accept.

The guarantee-test protocol

A defensible acceptance test is a structured event, agreed in writing before anyone opens a valve. A workable protocol:

1. Precondition the plant. The biomass must be mature. A plant needs weeks of stable operation for the microbial culture to fully establish — jump the trial-run and commissioning stages and you are testing an unripe plant that will fail on biology, not on design.

2. Fix the test window. Typically three consecutive days at design flow, so a pass has to be repeatable, not a one-day fluke.

3. Hold flow at rated KLD throughout, logged hourly.

4. Draw 24-hour composites at inlet and outlet each day.

5. Test at a NABL lab for the full PG parameter list.

6. Pass criterion: every outlet parameter must sit at or below its guaranteed value on all test days — not on average, on every day.

7. Witness and document. Client, consultant and contractor jointly witness sampling and countersign the log. Attach flow records, lab reports and photographs to the acceptance certificate.

Only when all three days clear does the plant earn its Performance Guarantee Certificate, which should feed directly into the handover checklist. Keep the whole file — it is your evidence for the Consent to Operate and any future dispute.

What to do when it fails

A failed test is a diagnosis prompt, not a catastrophe. Read which parameter failed, because each points somewhere specific:

• High BOD/COD — the biology isn't coping. Suspect insufficient aeration (check blower output and DO in the aeration tank), an immature or washed-out biomass, hydraulic overloading, or short-circuiting through the tank.
• High TSS — solids are escaping separation. Suspect a poorly settling sludge (bulking), an overloaded or short-circuited clarifier, or, in a membrane plant, fouled or breached UF membranes — see the MBR and MBBR guides for technology-specific causes.
• High coliform — disinfection is under-dosing. Check the chlorination contact time and residual, or the UV lamp intensity and sleeve fouling.
• pH excursions — chemical dosing or an industrial slug entering a domestic plant.

The remedy path runs cheapest-first: operational tuning (raise DO, adjust sludge wasting, correct dosing, extend contact time) before process fixes before invoking commercial remedies. Contractually, the PG clause should let you re-test after correction, and should tie release of the retention/performance bond to a clean pass. Never let go of the retention money until the certificate is signed.

The bottom line

Performance testing is the hinge between construction and ownership. Done properly — full design flow, composite inlet and outlet samples, a NABL lab, three repeatable days, joint witnessing — it gives the buyer a plant proven to meet its guarantees and gives the operator a baseline to run against for the next decade. Done lazily, at half-load with a single grab sample, it certifies nothing and hides the failures that surface the day the building fills up.

From here, tie the result back into the broader Sewage Treatment Plants guide library, and once the plant is proven, quantify what that clean water is worth with the Water Reuse Savings Calculator — the number that turns a compliance obligation into a recovered asset.