RTO Performance Verification: What to Prepare Before a Buyer Acceptance Test

An RTO performance verification is not simply a day when someone records outlet data. It is a controlled demonstration that the installed system can operate in an agreed process case and produce the evidence defined in the purchase or commissioning documents.

The difficult work usually happens before the test: deciding which production sources run, defining stable conditions, checking instruments and sample locations, agreeing how missing data will be handled, and assigning each acceptance row to an owner. If those decisions remain implicit, a technically sound test can still produce an argument about whether the result is representative.

This guide is an engineering preparation framework for buyers, operations, EHS, commissioning, and suppliers. It does not replace the contract, approved design basis, site procedures, or the competent test organization responsible for sampling and interpretation.

1. Define the approved test case before anyone takes a reading

Write the test case in plain language and attach the current process and RTO documents. At minimum, identify the process sources connected to the system, the intended production state, the permitted operating mode, and the duration or number of valid runs required by the agreed test plan.

Record the conditions that can materially change the result:

• Which exhaust sources are operating, and are they expected to run simultaneously?
• What production rate, material, solvent or coating recipe, and shift condition represent the approved case?
• What airflow, temperature, pressure, VOC composition, and moisture range is expected at the RTO inlet?
• Which bypasses, recirculation paths, dilution air, purge sequences, or startup states are excluded?
• What happens if the process changes or a reading leaves the agreed range?

A test case is useful only when operations can recognize it on the day. Use tags, historian trends, batch records, and operator sign-off to show that the stated case actually occurred.

2. Turn the acceptance requirement into measurable rows

Convert the contract or commissioning requirement into an acceptance matrix. Do not leave a phrase such as “meets performance” without a measurement definition, owner, and evidence source.

For each row, state whether it is a guaranteed value, a design reference, an observation, or a condition for a valid test. That distinction keeps an observation from being mistaken for a contractual conclusion.

3. Check instruments, sampling points, and data quality

Before the test, make an instrument list with tag, range, unit, calibration status, and data source. Confirm that the range is suitable for the expected operating window; a sensor can be in calibration and still be poorly suited to the value being measured.

Sampling preparation deserves the same discipline. Verify that ports are accessible, safe, identified, and located where the agreed method expects them. Confirm probe clearance, conditioning requirements, utilities, and any blank or leak checks required by the test organization. Do not infer a valid result from an easy-to-reach port that does not represent the intended stream.

Synchronize clocks across the PLC, historian, portable instruments, laboratory labels, and operator logs. A five-minute mismatch can make a process change appear unrelated to a sample. Define the master time reference and record any correction.

Use a short pre-test data-quality check:

1. Compare the control-system value with an independent portable or reference reading where practical.
2. Confirm units and reference conditions, especially for flow, temperature, and concentration.
3. Check for flat-lined, clipped, implausibly fast, or intermittently missing signals.
4. Record calibration certificates or verification results in the test file.

4. Prepare the RTO and process sources for a stable run

The equipment should be mechanically and operationally ready before the first valid run starts. Complete open punch-list items that could affect the test, review active alarms, and confirm the approved valve sequencing, burner state, fan configuration, and safety permissives. A performance test is not the right time to discover that a temporary override is still active.

The process side is equally important. Confirm source routing, damper positions, production recipe, material feed, and expected solvent or VOC loading. If a source is unavailable, document the substitution and obtain the test lead’s decision on whether the run remains valid.

Define a stabilization rule that operators can follow. It may use a sustained period of acceptable temperature, flow, pressure, source status, and alarm-free operation. The exact values should come from the approved design basis or test plan, not from a generic blog threshold.

Keep an event log with exact timestamps for startup, stabilization, sample start/stop, recipe or flow changes, alarms, manual interventions, and shutdown. A clean event log often explains more than a single average.

5. Coordinate emissions sampling and chain of custody

When external sampling or laboratory analysis is involved, align the field team, plant, and supplier before the test date. Confirm who supplies equipment, who observes the method, who labels samples, and who receives the final data package.

The sampling plan should state the analytes, locations, number of runs, duration, field blanks or duplicates where required, sample containers, preservation, transport, and reporting units. The plan must also explain what happens when a sample is invalid or a process event interrupts a run.

Use a chain-of-custody record that connects sample ID to location, start/stop time, operating case, sampler, preservation, and laboratory receipt. Keep copies with the test report. These records support traceability; they do not by themselves prove compliance with a particular jurisdiction’s rule.

6. Separate test observations from pass/fail conclusions

During the test, capture facts first: the operating state, instrument values, sample IDs, alarms, changes, and deviations. The technical team can then compare those facts with the agreed acceptance matrix.

Use a deviation log with four fields:

• what happened and when;
• which acceptance row or test assumption it affects;
• whether the run is valid, invalid, or requires a technical decision;
• who approved the disposition and what follow-up evidence is required.

Avoid editing data to make a trend look cleaner. If a value is corrected, preserve the original, the reason, the method, and the person who approved the correction. If a result is outside the agreed band, report it and explain the operating context rather than labeling it a failure without review.

7. Close the evidence loop after the test

The acceptance package should be usable by someone who was not in the control room. Assemble the signed test plan, operating-case evidence, calibration and method records, raw data, calculations, sample and laboratory records, event/deviation log, photos of relevant setup, and the final acceptance matrix.

Hold a short close-out with operations, maintenance, engineering, EHS, procurement, and the supplier. Identify open actions such as repeat testing, instrument repair, documentation correction, or a design-basis update. Assign an owner and due date for each action. Link the final record to the RTO equipment tags and future maintenance or spare-parts reviews.

FAQ

Is an RTO performance verification the same as a regulatory compliance test?

Not necessarily. A buyer acceptance or commissioning test may use contract-defined conditions and methods, while a regulatory test follows the applicable permit or rule. Confirm the governing requirement and competent test organization for the site.

What if production cannot reach the planned operating case?

Document the actual source status, recipe, flow, loading, and duration. Ask the responsible test lead and contract owners whether the run is representative, needs qualification, or must be repeated. Do not silently substitute a different case.

How much data should be retained?

Retain the complete package defined by the contract, site document control, and test method: raw data, calculations, calibration evidence, sample records, event logs, and approved deviations. A summary without traceable source data is difficult to review.

Can the PLC historian alone prove performance?

Historian data is valuable for operating context, but some acceptance rows require independent verification, calibrated instruments, or laboratory analysis. Match each conclusion to the evidence type stated in the acceptance matrix.

Can SERNO help prepare an acceptance matrix?

SERNO can help organize the process-source data, RTO operating points, measurement questions, and evidence owners for a site-specific technical discussion. Final acceptance decisions remain with the contract parties and qualified site personnel.

Conclusion

RTO performance verification is strongest when the test case, measurement method, equipment state, and evidence package are agreed before the first sample. Define measurable acceptance rows, check instruments and sample points, stabilize the process, log every event, and separate observations from conclusions. That preparation gives buyers a reviewable record and gives plant teams a clear path for any repeat or corrective action.

For an upcoming RTO acceptance test, SERNO can review the available design basis, source data, operating scenarios, and draft acceptance matrix for a focused engineering discussion. Use the actual installation documents and qualified site procedures for final operating, safety, and compliance decisions.