Ceramic wastewater plants that sail through a single-metric acceptance check often surface a different kind of problem six months after handover: rising sludge disposal costs, dosing drift that operators manage manually each shift, and reuse water that meets turbidity on a clean-feed day but not on a high-load one. The damage is not catastrophic immediately—it accumulates through compounding operational adjustments that were never logged, never challenged during commissioning, and never compared against what the system was contracted to deliver. The judgment that prevents this is straightforward: acceptance must close on the full water loop, not on any individual output measured in isolation. What follows gives you the criteria to distinguish a defensible acceptance record from one that is likely to be disputed when performance falls short.
Test under normal ceramic wastewater feed conditions
Running acceptance on diluted, synthetic, or pre-settled feed is the most common shortcut that invalidates the commissioning record. Ceramic wastewater from grinding, polishing, or glazing lines carries a variable combination of fine suspended solids, colloidal clay particles, and fluctuating pH from process chemistry—none of which behave the same way as a prepared test slurry. If the system is tested under tidier conditions, the floc formation, settling velocity, and chemical demand observed during testing will not reflect what the plant actually imposes.
The practical consequence is that acceptance thresholds calibrated on clean feed can mask two instabilities that only appear under real load: poor floc formation at low or variable pH, and recirculating fines that progressively blind the downstream filter press. By the time either problem is visible in production, the commissioning window has closed and the contractual baseline has been set against test conditions that no longer apply.
Define what “normal feed” means for your site before the acceptance period starts. This is not a standardised input—it reflects your process lines, seasonal variation in raw materials, and the shift-by-shift range of suspended solids the system will actually receive. The design basis should already contain the feed envelope the supplier used for sizing; acceptance testing should operate within that envelope, not below it.
Measure reuse-water turbidity TSS and pH together
Measuring turbidity alone gives you a real-time optical proxy, but it can pass while total suspended solids remain above the design limit—particularly when fine colloidal particles scatter light at levels that look acceptable but carry enough mass to accumulate in spray nozzles or fouling-sensitive reuse equipment. TSS and pH belong in the same measurement run, logged simultaneously, because the three parameters are causally linked: pH affects coagulation efficiency, coagulation efficiency drives TSS removal, and TSS affects turbidity. Separating them into sequential checks misses the coupling.
Turbidity measurement during acceptance can reference the optical method framework in ISO 7027-1:2016, but that standard addresses measurement technique, not pass/fail values. The numeric thresholds that matter for your site are those written into the quoted design basis. If the design basis specifies a reuse-water turbidity limit and a TSS limit independently, both need to be met simultaneously, not on separate test runs.
The practical failure pattern here is pH drift during extended cycling. Ceramic process water often becomes more acidic under sustained load as soluble species accumulate. If pH slides below the coagulation optimum without triggering a dose adjustment, TSS removal degrades before turbidity visibly worsens—meaning a turbidity-only check can record a pass at the moment TSS has already exceeded the design limit. A combined measurement protocol, logged at defined intervals throughout each test cycle, is the minimum needed to detect that coupling before acceptance closes.
Track sludge volume cake handling and filtrate return
Sludge output is the most frequently deferred acceptance criterion and the one most likely to generate a post-handover cost dispute. The volume of sludge produced per unit of feed processed, the solids content of the pressed cake, and the volume and quality of filtrate returned to the front of the system all affect the economic and operational balance of the plant in ways that water quality metrics alone do not reveal.
ISO 5667-13:2011 provides a useful reference for collecting representative sludge samples at defined points in the process rather than from the most convenient access point, but it does not set performance targets. Those targets belong in the design basis: expected cake solids percentage, maximum sludge volume per shift, and the filtrate quality assumed when calculating the recycle load on the sedimentation stage. If filtrate from the Filtre-presse à membrane is returned to the sedimentation inlet without being characterised during acceptance, the commissioning record cannot confirm that the recycled load matches the design assumption.
The trade-off that matters here is between test speed and downstream visibility. Skipping sludge volume tracking and filtrate return measurement shortens the acceptance period by a day or two, but it defers the cost of discovering that cake moisture is above design, that the filter requires more press cycles per shift than specified, or that filtrate fines are progressively loading the Tour de sédimentation verticale beyond its sizing basis. Each of these conditions is diagnosable during acceptance and difficult to attribute contractually once the plant is in production. Track sludge volume and filtrate return as primary acceptance parameters, with the same logging rigour applied to reuse-water quality. Further detail on filtrate behaviour during dewatering cycles is covered in Gestion du filtrat dans les opérations de filtre-presse.
Record chemical dose and operator adjustments
Every manual adjustment made during acceptance testing is a data point about whether the system is self-stabilising or intervention-dependent. If coagulant and flocculant dose rates are adjusted multiple times per shift by the operator to maintain water quality, those adjustments need to appear in the acceptance log with a timestamp, a reason, and the parameter that triggered them. A commissioning record that shows clean reuse water but contains no dosing data is not evidence of a stable system—it is an incomplete record.
The PAC and PAM dose rates observed during acceptance should be compared against the design basis consumption figures. Significant divergence in either direction has a different implication: higher-than-specified dose rates may indicate that the feed is harder to treat than modelled, or that the sedimentation stage is underperforming; lower rates may mean the feed is atypically clean, which will not represent normal conditions. Either way, the data needs to be present in the record, not averaged out or omitted because the water quality output looked acceptable. An intelligent dosing system that logs dose rates automatically reduces the risk of gaps in this record and gives the acceptance team traceability without relying on operator-written notes.
What this section of the acceptance record ultimately demonstrates is whether the system’s chemical consumption is predictable, proportionate to feed variation, and within the operating envelope the supplier quoted. That predictability has direct procurement implications: annual reagent costs, storage sizing, and resupply frequency are all based on the design-basis dose rate. If actual dose rates during acceptance differ materially, those figures should be revised before handover rather than inherited as an unresolved discrepancy.
Check stable operation across repeated cycles
A single successful run demonstrates capability under one set of conditions. Repeated-cycle testing is what demonstrates that the system recovers reliably after each sludge draw-down, filter press cycle, or dosing adjustment—and that performance does not degrade progressively across the test period. The number of cycles required is not universal; it follows from the design basis and the cycle frequency the plant expects to operate.
What repeated cycling reveals that a single run does not: filter blinding that begins after the second or third press cycle, sedimentation zone recovery time after a high-load event, and the lag between a pH shift in the feed and the automatic dose response. These are not edge-case failure modes—they are normal operating conditions compressed into the acceptance window. If any cycle produces results that fall outside the design basis, the correct response is root-cause review before the next cycle, not continuation of the test with a plan to address the deviation post-handover.
Cycle-to-cycle variance should be assessed against the design basis, not against the best run recorded. If the first cycle performs well and subsequent cycles show declining cake solids or rising reuse-water TSS, the trend matters more than the initial result. A flat or improving trend across repeated cycles, with dosing remaining within design range and no unplanned operator interventions, is the practical evidence that the system is ready for handover.
Compare results with the quoted design basis
The design basis—whether embedded in the project specification, the supplier’s technical offer, or the agreed URS equivalent—is the only contractually meaningful reference point for acceptance. Comparing results against general industry benchmarks or against what similar ceramic plants achieve elsewhere does not confirm that the delivered system meets what was purchased for this site at this capacity.
The comparison should be structured rather than impressionistic. Reuse-water turbidity, TSS, and pH results should be laid against the design limits for each. Sludge volume per shift should be checked against the sizing assumption. Chemical dose rates should be compared with the quoted consumption figures. Filtrate quality and return volume should be verified against the recycle load the system was designed to handle. Where the EPA Water Reuse Guidelines informed the design basis for reuse-water quality targets, they provide useful background context for those limits—but they do not replace the specific design values the supplier committed to at contract stage.
Any result that sits at the boundary of the design basis deserves scrutiny before acceptance is recorded. A turbidity reading that consistently lands at 95% of the design limit suggests the system has less operational headroom than expected and may not hold the limit under a higher-load day. That observation belongs in the acceptance record, with the supplier’s response documented, rather than being treated as a pass because the threshold was not breached.
Close acceptance only when the loop is repeatable
Acceptance has a practical gateway function: it is the last point at which discrepancies are the supplier’s problem rather than the operator’s. Closing acceptance before the full loop—water quality, sludge output, dosing stability, filtrate return—has been demonstrated as repeatable transfers that responsibility prematurely.
Repeatability in this context does not require statistical precision across dozens of cycles. It requires that water quality, sludge output, chemical consumption, and operator intervention load are internally consistent across the test period, aligned with the design basis, and not dependent on continuous manual correction to remain within acceptable bounds. If the commissioning log shows repeated manual adjustments without a documented cause and response pattern, it is difficult to defend the claim that the loop is inherently stable rather than propped up by operator attention. That distinction matters in any post-handover performance dispute, and it matters in regulatory review if reuse-water quality becomes a compliance issue.
Before sign-off, the acceptance record should allow any qualified reviewer to answer four questions from the documentation alone: what was the feed condition during each test run, what did the system produce in terms of reuse water and sludge, how much chemical was consumed and how was the dose adjusted, and did each cycle perform consistently with the previous ones. If any of those questions requires the commissioning engineer’s memory rather than a written record, the acceptance package is not complete.
The real risk in ceramic wastewater acceptance is not a single missed measurement—it is the cumulative effect of treating sludge output, filtrate return, and operator adjustments as secondary evidence rather than primary acceptance criteria. A stripped-down test that closes quickly will often pass on the day and drift in production, generating cost and attribution disputes that the commissioning record cannot resolve.
Before closing acceptance on any ceramic wastewater recycling system, confirm that the test period included real feed conditions, that all parameters were logged simultaneously across repeated cycles, and that the results have been compared explicitly with the figures the supplier committed to at contract stage. If the design basis consumption figures, sludge volume estimates, or reuse-water limits were not specified with enough precision to make that comparison, clarifying them before acceptance is more productive than discovering the gap after handover.
Questions fréquemment posées
Q: What if our production line isn’t running at full capacity during the acceptance window — can we simulate normal feed?
A: No, acceptance testing should be postponed until the system can process real, full-strength ceramic wastewater. Simulated or diluted feed masks floc formation and settling instabilities that only appear under actual load, and any thresholds calibrated on synthetic feed will not hold once production starts, invalidating the commissioning baseline.
Q: After closing acceptance, what ongoing checks are needed to prevent the system from drifting outside the design basis?
A: Establish a monthly mini-cycle where reuse-water turbidity, TSS, pH, chemical dose rates, and sludge cake moisture are logged simultaneously and compared against the acceptance reference figures. This creates an early-warning trend before a cost or compliance issue accumulates.
Q: Do these acceptance criteria still apply if the recycled water is only used for floor washing or non-critical rinsing?
A: The water-quality targets for turbidity and TSS can be relaxed to match the actual reuse application, but sludge volume and chemical consumption must still be tracked. A plant that ignores sludge output during acceptance risks undiscovered disposal cost escalation regardless of reuse-water destination.
Q: Is a manual operator log sufficient for recording dosing adjustments during acceptance, or is automated logging required?
A: A manual log can be sufficient if it captures every adjustment with a timestamp and the specific trigger, but automated logging from an intelligent dosing system removes the risk of missing entries and brings the traceability needed if a post-handover dispute arises. The deciding factor is completeness, not the tool.
Q: The full acceptance protocol would delay our production start — is the extended testing really worth the downtime?
A: The production impact of thorough acceptance testing is almost always smaller than the cost of resolving sludge handling, dosing drift, or water quality failures after handover. A truncated test shifts discovery risk into operations, where fixing problems is more expensive and contractually harder to attribute.
