The Potency Discrepancy on a Live Batch
A cultivator harvests a 40 kg wet batch of a high-THC cultivar. The pre-harvest COA from Lab A reports 28.4% total THC and passes pesticides and microbials. After drying and curing, you submit a sample to Lab B for the required final testing before sale. Lab B reports 21.1% total THC, a passing microbial panel, but detects piperonyl butoxide at 0.08 ppm - below the state action limit of 0.1 ppm but above your internal 0.05 ppm spec. The batch is 8 kg dried flower and represents a $60,000 wholesale order due in 5 days. Lab A's sample was taken 10 days before harvest; Lab B's sample was taken from the top colas only. Your SOP requires a representative sample across the batch.
1. How do you interpret the discrepancy and the pesticide result, and what do you do before release?
Read the worked answer
Interpretation:
- Potency gap (28.4% → 21.1%) is plausible from drying/curing moisture loss and from sampling differences (top colas vs. representative sample), but it is large enough to require investigation. Moisture content and sampling method must be reconciled.
- Piperonyl butoxide at 0.08 ppm is below the state limit but above your internal spec. That is a yellow flag: it suggests possible prior pesticide use or contamination and may indicate other unlisted actives. A single below-limit result does not guarantee the batch is clean.
Actions before release:
- Do not release. Place the batch on hold in the inventory system.
- Review the sampling SOP: Lab B's top-cola sample was not representative. Retake a composite sample per SOP (multiple points, homogenized) and resubmit to Lab B or a third lab.
- Request Lab A's raw data and confirm sample chain of custody and moisture correction basis.
- Ask the lab to screen for a broader pesticide panel, including actives that co-occur with piperonyl butoxide (a common synergist).
- If the resample confirms below-limit but above internal spec, escalate to QA management for a documented disposition decision; consider whether the internal spec should trigger rejection.
Reasoning: below-limit is not the same as clean; the sampling flaw undermines confidence in both results.
2. What quality-system changes would prevent this situation from recurring?
Read the worked answer
Sampling: Mandate representative composite sampling (e.g., 8–12 cores across the batch, including mid and lower canopy) with documented moisture content. Train samplers and audit compliance.
Lab management: Use a primary and confirmatory lab with written chain-of-custody. Require labs to report raw data and method limits. Establish a policy for discordant results (retest triggers, third-lab arbitration).
Specifications: Distinguish state action limits from internal specs and define escalation when a result falls between them. Document the rationale for each internal spec.
Pre-harvest controls: The pre-harvest COA is a screening tool, not a release test. Track pesticide use records and applicator logs; investigate any synergist detection as a potential indicator of undisclosed actives.
Batch records: Require moisture, sampling map, and lab COAs to be linked to each batch ID before QA release.
Reasoning: the failure was process (sampling) and governance (spec escalation), not just chemistry.