Quality controls are useful because they can reveal a problem. Their failure should trigger a defined investigation, even when the associated production samples look plausible. For an exploration owner, the immediate decision is which results can be used, which must be held and what evidence is required before interpretation resumes.
Know which question each control asks
Standards, blanks and duplicates examine different aspects of the sampling and analytical chain. The programme should specify their purpose, placement and acceptance criteria before results arrive. A laboratory certificate alone does not show that the overall programme met those criteria. JORC's checklist expressly addresses assay quality controls and whether accuracy and precision have been established. The owner's technical team should be able to explain the controls in ordinary language: what a result might reveal, what it cannot reveal and which person has authority to release the batch.
Investigate the chain rather than choose a convenient cause
A hypothetical blank returning an unexpected value could prompt checks of preparation, identification, handling and analysis. The appropriate investigation depends on the circumstances. Automatically blaming contamination or dismissing a result as clerical would bypass the evidence. Preserve the original submissions, certificates and communications while checking possible causes. The scope should extend to neighbouring samples where relevant, with a documented reason for that boundary. An investigation becomes credible when it can explain why the accepted resolution addresses the suspected mechanism rather than merely produces more reassuring numbers.
Control the replacement record
Reanalysis and replacement results need clear relationships to the original records. The database should retain both versions with their status and the reason for a change. Quietly overwriting the original certificate makes later verification difficult and can leave different teams using different datasets. CIM's guidance on verification and database protocols supports this discipline. A model release should identify the dataset actually used and any held results. Technical staff need a reliable way to prevent provisional values from reaching interpretation files simply because an automated export treated every laboratory number as final.
Budget for the time required to resolve exceptions
Assay turnaround forecasts should include the possibility of quality investigations. Investors should be wary of a timetable that allows no time between receiving results and publishing an interpretation. Owners can set a communication rule that describes the hold and its consequences without speculating about the eventual outcome. The benefit is practical: a controlled delay may prevent expensive follow-up drilling based on a questionable pattern. A quality system earns its place when failed controls can change the work programme, the database and the publication date, even under pressure to announce progress.
