Minimum sample size and SEG characterisation adequacy per AIHA IHSS and NIOSH methods
| Risk Band | 95th %ile / OEL | Min n recommended | Rationale |
|---|---|---|---|
| Band 1 Very highly controlled |
< 1% | 1–3 Qualitative only |
Exposure so low that sampling mainly confirms absence; quantitative characterisation is not cost-effective. |
| Band 2 Well controlled |
1–10% | 3–6 | Sufficient to confirm low exposure; not designed for precise characterisation of the distribution. |
| Band 3 Action level |
10–50% | 6–10 | Adequate characterisation for trend monitoring; formal UCL calculations are feasible. |
| Band 4 Near OEL |
50–100% | ≥ 10 | Formal characterisation required; sufficient samples needed to resolve whether the 95th percentile exceeds the OEL. |
| Band 5 Exceeds OEL |
> 100% | ≥ 6 Immediate control; sampling to verify |
Controls must be implemented immediately. Further sampling confirms effectiveness of controls, not initial characterisation. |
| Work pattern | Typical GSD range | Example |
|---|---|---|
| Single substance, steady process, homogeneous SEG | 1.3–1.8 | Constant solvent cleaning operation |
| Mixed tasks, some variability | 1.8–2.5 | General manufacturing |
| Intermittent tasks with high peaks | 2.5–3.5 | Spray painting, welding |
| Highly variable or complex SEG | 3.5–5.0 | Mining, construction, agriculture |
| Zone | Decision criterion | Action |
|---|---|---|
| Accept | UCL on 95th percentile < OEL | Conclude compliant. Document and periodically re-survey to verify conditions have not changed. |
| Reject | LCL on 95th percentile > OEL | Conclude non-compliant. Implement controls immediately. Re-survey after controls are in place. |
| Indeterminate | Confidence interval straddles OEL | Collect more data, apply the precautionary principle, or implement interim controls and re-survey. |