Audit Planning, Site Inspection, and Sampling Procedures
Audit planning begins with scope: which buildings, which assemblies, how many visits, and what evidence will demonstrate conformity. Review the specification, approved submittals, installer certifications, and prior reports before mobilizing. Define audit criteria explicitly - code edition, project specification section, manufacturer requirements - so every finding ties to a requirement rather than an opinion.
Visits generally follow construction sequence: pre-installation verification of materials and mock-ups, in-progress inspection during application and detailing, and pre-concealment verification. Sampling must be defensible. Random selection documented by a repeatable method (grid, elevation, unit number) beats convenience sampling. Judgmental sampling is acceptable for high-risk details such as parapets, roof-wall junctions, and window perimeters, but should be reported as judgmental, not presented as representative.
Trap: declaring a project conforming because accessible areas looked good; define the sample before the visit and report results against it. Another trap: letting the contractor steer the sample, or quietly shrinking it because the crew was rushed.
Record environmental conditions, material lot or batch, and instrument calibration status. Photograph with location identifiers. Handle test specimens per the method and keep traceability intact.
Worked example: to be 90% confident of finding at least one nonconformity when the true defect rate is 10%, solve (1 − 0.1)ⁿ ≤ 0.10, giving n ≈ 22. Inspecting only 10 locations leaves roughly a 35% chance of missing a 10% defect rate, which is rarely defensible in a report.