Flare Data Gap
You are auditing a municipal solid waste landfill that operates a gas flare. The permit requires at least 98% destruction efficiency and continuous monitoring with valid data for at least 95% of operating hours each quarter. For Q2, the continuous monitoring system recorded 91% valid hours: 6% missing due to analyzer failures and 3% invalid due to calibration drift. A stack test conducted during Q2 reported 97.2% destruction efficiency. The operator says the stack test is biased by high moisture and that the flares were visually clean. The calibration log for the analyzer is missing for a 14-day period that overlaps two exceedance alarms. The landfill environmental specialist states that the data gaps are normal and that the stack test is the only reliable evidence. Your audit objective is to determine conformity with the permit monitoring and performance requirements.
1. Assess the sufficiency and reliability of the available evidence and determine whether a nonconformity exists.
Read the worked answer
Evidence assessment. The permit has two separate requirements: performance (98% destruction) and monitoring data availability (95% valid hours). The 91% valid hours is a monitoring nonconformity on its face. The stack test at 97.2% is below 98%, but its reliability is contested because of moisture bias and because the analyzer calibration log is missing for 14 days overlapping alarms. The operator's visual observation is not a substitute for continuous monitoring.\n\nConclusion: there is sufficient evidence of at least a monitoring nonconformity and a credible indication of a performance exceedance. The auditor should not dismiss the stack test solely on the operator's claim, nor treat it as conclusive without the test protocol and calibration data. The finding should distinguish confirmed data availability failure from probable destruction efficiency failure pending further evidence. This is a judgement call that requires documenting conflicting evidence rather than averaging it away.\n\nJudgement point: the auditor's role is to weigh evidence, not to resolve technical disputes beyond competence; a specialist may be needed.
2. What additional evidence should be gathered, and how should conflicting evidence be handled in the audit report?
Read the worked answer
Further evidence: the continuous monitoring QA/QC plan, raw analyzer output, calibration gas certificates, maintenance work orders, alarm logs for the 14-day gap, stack test protocol and moisture correction calculations, and interviews with the monitoring technician. I would also request the certification of the stack test lab and the chain of custody for samples.\n\nResolving conflict: use triangulation. If the analyzer was out of calibration, its data may be invalid, but the stack test still needs a valid protocol. If the stack test is invalid, the permit may still be violated by the data availability failure. The auditor should report the facts and the limitation, not choose the most convenient source. A follow-up test or third-party review may be needed. The report should state what is confirmed, what is unresolved, and what evidence would close the gap.\n\nJudgement point: conflicting evidence is common in compliance audits. The report should not overstate certainty; it should explain the basis for each conclusion.
3. The operator claims the data gaps are normal and the stack test is biased. How would you test those claims?
Read the worked answer
Testing the operator's claims. For the data gaps, compare Q2 against prior quarters and the permit's 95% requirement. Review analyzer failure history, maintenance response times, and whether the same failure repeated. For the stack test bias claim, review the test protocol, moisture measurement, sample conditioning, and whether the test was conducted under representative operating conditions. Check whether the lab corrected for moisture as required and whether the analyzer calibration was valid during the test.\n\nCorroboration: interview the monitoring technician and the stack test contractor separately, review alarm response records, and inspect the analyzer maintenance log. If the operator's claim is supported, the auditor should still report the data availability failure and the missing calibration records. If the bias claim is unsupported, the stack test becomes stronger evidence of a performance issue.\n\nJudgement point: an unsupported explanation does not neutralize evidence. The auditor documents the claim, tests it, and reports the result.