5.9% of quartz samples since 2000 never yielded a measurement

Of the 167,619 quartz samples in MSHA's open-data Quartz Samples dataset, collected between January 3, 2000 and June 23, 2026 across 3,431 distinct coal and metal/nonmetal mines, 9,825 (5.9%) carry a void code and never produced a usable result. These are samples MSHA's own quality-control process set aside before a quartz value was recorded.

The void reasons name where the monitoring record loses data. Insufficient weight gain on the filter accounts for 3,564 voids, and MMU not in producing status for 3,105. After those, the counts fall off: unacceptable timeframe (921), quartz laboratory void (490), control filter invalid (349), invalid sample type (339), contaminated (258), and invalid certification number (202). Some of these point to field or lab failures, while others are procedural.

Once a sample clears QC, the record is near-complete. Among the non-voided samples, only 272 of 167,619 (0.16%) are missing both a quartz percentage and a quartz microgram value, so the gap sits almost entirely in the voided fraction rather than in silent holes among passing samples.

A few caveats belong with these numbers. A void count is not an exposure-risk claim, and we cannot say what any voided sample would have measured had it been usable. MMU not in producing status is at least partly administrative, meaning the sample was drawn when the mining unit was not producing rather than because the lab work failed, so grouping it with weight-gain or contamination voids would overstate any reading of it as a single failure category. And because we have no denominator of worker-hours, none of these counts is a per-worker exposure rate. What the 5.9% documents is narrow: the share of samples that carry a void code, no more.

The full sample record, void codes included, is available per mine and operator through our index.