A 0.1% error sounds negligible until it's applied to volumes measured in millions of barrels or billions of cubic feet per year. At that scale, small systematic biases — not random noise, which tends to average out, but consistent over- or under-registration — accumulate into figures that show up directly on a balance sheet.
The math is straightforward: take annual throughput, multiply by the bias (not the full uncertainty band, since random components partially cancel over time), and multiply by unit price. For a mid-size custody transfer point, that arithmetic routinely produces six- or seven-figure annual exposure from an error that would pass a casual inspection of the meter's spec sheet.
Where this gets more subtle is in multiphase and allocation metering, where the 'error' isn't a single meter's calibration drift but a combination of phase-split uncertainty — water cut, gas volume fraction — propagating through an allocation formula shared across multiple wells or working interests. A small, consistent bias in phase fraction can quietly redistribute value between parties long before it's large enough to trigger a dispute.
This is why custody transfer measurement is one of the few places where investing in a tighter, better-documented uncertainty budget pays for itself directly rather than as a compliance cost. Knowing the real expanded uncertainty — and being able to show where it comes from — turns a number counterparties have to trust into one they can independently verify.