When to Trust a Statistical Drilling Benchmark and When Not To
Drilling benchmarks are often expressed as averages, medians, percentiles, or “best offset” values. But a benchmark can be statistically stable and still be engineeringly irrelevant. This article explains how sample size, independence, formation mix, BHA and rig context, metric definition, weighting, measurement consistency, and selection bias affect the reliability of drilling benchmarks. It also shows why raw sensor sample count should not be confused with the number of independent drilling observations and why distributions are often more useful than a single target value.
Read article →