The Hidden Rows in the NOx Spreadsheet
You are auditing a food processor's environmental management system. Boiler nitrogen-oxide monitoring is a permit condition, with a limit of 120 mg/m3. The environmental coordinator says the data are clean. While reviewing the spreadsheet, you find hidden rows for 14 days when readings exceeded 150 mg/m3. The boiler operator says they zeroed the sensor because it drifts after washdown. The maintenance log has no calibration entries for two of six analyzers. The coordinator's manager says the hidden rows are just formatting. A junior technician privately tells you they were told to delete bad data. The audit is halfway through, and the coordinator asks to see your notes.
1. Evaluate the sufficiency and reliability of the evidence you have so far.
Read the worked answer
You have three evidence types: document (the spreadsheet and hidden rows), interview (the operator, manager, and technician), and observation (the maintenance log gap). None is sufficient alone, but together they are enough to pursue a finding. The hidden rows are objective evidence that records were altered or concealed; the manager's formatting explanation is implausible given 14 exceedance days. The operator's statement explains a possible cause but also shows an uncontrolled adjustment practice. The missing calibration entries undermine the reliability of all analyzer data for the affected period.
The technician's statement is corroborative but should be treated carefully because it is oral and may expose the technician to retaliation. Do not conclude falsification yet; conclude that data reliability and record integrity are not demonstrated. That is enough for a nonconformity under 9.1.1 and potentially 9.1.2 if compliance evaluation used the altered data.
2. What corroboration and evidence-preservation steps would you take next?
Read the worked answer
Preserve the original electronic file before it can change: save a copy with date, time, and hash, and request the audit trail, version history, and file metadata. Ask for the raw analyzer output and the permit reports submitted for the 14 days. Obtain calibration certificates, maintenance work orders, and the analyzer QA/QC procedure. Interview the operator and coordinator separately, using open questions, and compare their accounts with the records. Check whether the zeroing practice was authorized and whether it was recorded.
Sample around washdown days and compare with any reference-method results. Follow the data trail from analyzer to spreadsheet to regulatory report. Do not confront staff aggressively or promise confidentiality you cannot provide. If evidence suggests deliberate alteration, stop and consult your audit programme manager and legal adviser before expanding the investigation.
3. How would you classify and document the finding at this stage?
Read the worked answer
Document a nonconformity for failure to ensure monitoring data are reliable and controlled, citing 9.1.1 and the permit monitoring requirement. State the facts: hidden rows showing 14 exceedance days, missing calibration for two analyzers, and an undocumented sensor-zeroing practice. State that the reliability of reported data is not demonstrated. Avoid using fraud or falsification as a conclusion unless you have direct evidence; describe the record-integrity issue instead.
Classification depends on impact: if the data were used in compliance reports and the exceedances were not evaluated, a major nonconformity may be justified because the monitoring process cannot be trusted. If it is isolated and corrected, a minor may be appropriate. Record the evidence source, date, and persons interviewed, and keep the original files. Because the coordinator asked to see your notes, explain that working notes remain audit records and are not negotiable; provide the finding through the normal reporting process.