I'm a quality and compliance manager at an industrial operator. I review every piece of treatment equipment before it goes into our plants — roughly 40 vendor quotes and 12 installed units a year. In 2025, I rejected 31% of first-time deliveries because the nameplate didn't match the spec sheet.

Below is the checklist I now run on every wastewater treatment equipment sourcing project. Seven steps, built for industrial and mining plants running roughly 50 to 500 m³/day. If you're a package plant under 20 m³/day or a municipal MBR, some of these compress — I'll flag where near the end.

Step 1: Lock the effluent numbers before you open a catalog

Before you call a single wastewater treatment equipment supplier, write down your discharge limits. Not "we need to meet EPA standards." Your actual permit numbers, in mg/L, with sampling frequency and averaging period.

The US federal floor under 40 CFR Part 133 is 30 mg/L BOD₅, 30 mg/L TSS, and pH 6–9 as monthly averages for secondary treatment. Mining permits are usually tighter — suspended solids at 15 mg/L or less, and metals limits (copper, zinc, lead) that vary by state. If your permit says 10 mg/L ammonia as N, that is a different plant than one with no ammonia limit at all.

Why first: every sizing calculation, every material choice, every control scheme downstream depends on this. I've watched buyers spec a 200 m³/day lamella clarifier for a job that turned out to be 80 m³/day. The unit was fine. The order was wrong.

Step 2: Screen the manufacturer's quality system, not the brochure

Anyone can print "ISO certified" on a website. Ask for the certificate number and the scope. ISO 9001:2015 covers quality management; ISO 14001:2015 covers environmental management. A supplier with ISO 9001 but no ISO 14001 is not automatically a red flag — plenty of good fabricators are small enough to run without it.

What matters more: does the certificate scope actually cover the product line you're buying? I've seen suppliers with ISO 9001 for tanks who quoted inline mixers that were subcontracted out and never actually covered. The certificate was real. The coverage wasn't.

And ask for the date of their last surveillance audit. Certificates expire. (Note to self: add this to the standard RFQ template — we still email it case by case.)

Step 3: Match the material grade to the actual chemistry

"Stainless steel" is not a spec. 304 and 316L are different metals with different corrosion behavior. If your wastewater has chlorides above roughly 200 mg/L, 304 will pit. Sometimes within a year. 316L, with 2–3% molybdenum, handles chlorides better — but not indefinitely, and not at elevated temperatures.

When I compared two clarifier tanks side by side after 14 months in service — same duty, one in 304, one in 316L — I finally understood why the original spec mattered so much. The 304 unit had pitting along the weld heat-affected zone. The 316L was clean. The cost difference at purchase was $4,300. The 304 tank cost us $19,000 in repair and downtime.

Check the mill certificate against ASTM A240 for the plate and the consumables spec for the welds. Most failures I've seen are at welds, not in the parent metal.

Step 4: Demand the mass balance and hydraulic profile in writing

This is the step most buyers skip. What I mean is that most quotes give you a capacity in m³/day and a footprint, and stop there — no influent load assumption, no head-loss calculation, and no way to check whether the unit will actually meet your permit number once the water hits the outlet.

Ask for a mass balance and a hydraulic profile showing the water line from inlet to outlet. If the supplier can't produce this in two weeks for a standard unit, they are assembling from a template, not designing to your job.

Bonus check: does the mass balance close? If influent BOD is 800 mg/L and effluent is claimed at 20 mg/L, the removal calculation should be traceable. If it isn't, ask where the mass went. (Surprise, surprise — sometimes the answer is "we assumed.")

Step 5: Re-check the electricals against your local code

Equipment sourced from overseas — or even from a different state — often ships to the electrical code of the manufacturer's home market. NEMA enclosures are not identical to IP ratings, and the "UL" sticker on a component doesn't always mean the assembled panel is UL-listed.

Walk your electrician through the panel schematic before the PO goes out. Confirm voltage, phase, frequency, enclosure rating, and whether the panel needs a UL 508A label. I've seen a $90,000 packaged treatment unit arrive with 60 Hz motors — fine for the supplier, not fine for a 50 Hz site. The retrofit cost more than the original freight.

Step 6: Write a witness-test protocol, not a "FAT certificate"

Everyone tells you to demand a factory acceptance test. They warned me to read the test procedure, not just the certificate. I didn't listen once and signed off on a "FAT certificate" that turned out to be a one-page checklist the supplier wrote themselves. The unit failed commissioning six weeks later because the pumps were undersized for a duty point the FAT never actually held. That lesson cost us a $27,000 retrofit and a two-week delay.

What to write instead: a protocol that names every test, the acceptance criterion in engineering units, and who signs off. Run the pumps at the design duty point and measure head, flow, and power draw. Hydrotest the vessels per ASME Section VIII — typically 1.3× MAWP for the hydrotest. Document results in the same units as your spec sheet, not "within tolerance."

If you can't be there, hire a third-party inspector. The cost is small compared to the retrofit.

Step 7: Tie payment milestones to commissioning, not shipment

This is the quiet one. Most payment terms are 30% deposit, 60% on shipment, 10% on commissioning — or something close. That structure creates an incentive to ship, not to commission. The invoice is paid before the unit proves it can meet the spec.

What works better for us: 30% deposit, 30% on FAT, 30% on arrival and mechanical completion, 10% on final commissioning and performance test. We held that last 10% on a 2024 project for 45 days past arrival while the supplier fixed a clarifier weir level. If the last 10% had been released on shipment, we'd still be chasing them.

We didn't have a formal milestone structure until the third time a commissioning delay stretched past 30 days on a project where the supplier had already been paid in full. Cost us when they quietly moved on to the next job. Now every PO includes the milestone table, and every supplier signs it before the deposit leaves.

Notes and common errors

Three things I see regularly, beyond the seven steps.

Taking "industry standard" as a specification. It isn't one. If the supplier won't put the number in writing, the number doesn't exist.

Skipping the change-control step. If you specify the unit, then approve a "small" change by email, then approve another — by the time the unit arrives, the nameplate spec and the PO spec are different documents. Someone has to reconcile them. That someone is you.

Ordering the same unit twice without revisiting the chemistry. A clarifier that worked on one site does not always transfer. Chloride, temperature, and TDS change with the process, not with the model number.

One honest limitation. This checklist is built for industrial wastewater treatment equipment sourcing at 50–500 m³/day. If you're a municipal plant under a design-build contract, or a package plant under 20 m³/day, some steps compress — the mass balance in Step 4, for instance, is often delivered as a standard submittal by the engineering firm rather than by the supplier. In that case, check the submittal, not the catalog. For an indirect discharge to a municipal sewer with loose local limits, Step 1 reduces to a chemistry check and you can spend the saved effort on Step 6. It also helps less if you're buying refurbished or second-hand equipment — most of these checks assume a documented design basis, which used units rarely come with.

None of these steps is clever. They're the boring ones. But in four years of reviewing these deliveries, 31% of first shipments got rejected — and the failure was almost always in the boring steps, not the technical ones.