A custody-transfer meter is the point where money changes hands, so a quiet drift in its accuracy can go unnoticed while it steadily mismeasures every barrel or cubic foot that passes. Check metering guards against that blind spot by installing a second, independent meter alongside the custody meter and comparing the two - if they agree, both are trusted, and if they diverge, something has changed. This guide defines a check meter, explains how the two meters are compared to catch drift or resolve a dispute, what deviation is considered acceptable, and how dual-meter data in SCADA supports confidence in the measurement.
Check Metering in one line: A check meter is a second meter installed in series with a custody-transfer meter to independently verify the volume it measures. The two meters see the same flow, so their readings should agree closely, and comparing them continuously catches a meter that has drifted, failed, or is being questioned in a measurement dispute. As long as the two stay within an agreed deviation, the custody measurement is trusted; when they diverge, the check meter provides an independent reference for investigating which meter is wrong.
The fundamental problem a check meter addresses is that a single custody meter has no witness. It measures the flow, that number is billed, and there is no independent way to know whether it is still accurate between provings other than trusting it - so a gradual drift, a partial blockage, a failing sensor, or a subtle installation problem can quietly bias every measurement until the next verification. On a high-value stream, even a small percentage error accumulates into a large discrepancy over time, and it may not be discovered until a proving or an audit long after the money has moved.
A check meter gives that single measurement a continuous, independent second opinion. Placed in series so it sees exactly the same flow, it should read the same volume, and the running comparison of the two turns a silent drift into a visible divergence the moment it grows beyond normal scatter. This converts measurement integrity from a periodic check into a continuous one - instead of trusting the custody meter until the next proving, operators see day to day whether it still agrees with an independent reference.
Check meters are most common where the stakes justify the cost of a second measurement: high-value custody points, streams shared between parties who each want assurance, and situations where a measurement dispute has occurred or is anticipated. The check meter is often a different technology or at least an independently maintained instrument, so that a systematic problem affecting one is unlikely to affect the other identically, preserving the independence that makes the comparison meaningful.
In use, the two meters' readings are compared over a defined period - the volumes they each register over an hour, a day, or a batch - and the difference is expressed as a deviation. Because no two meters are perfect and each carries its own small uncertainty, they will never agree exactly, so the comparison is judged against an acceptable band rather than expecting a zero difference. As long as the deviation stays within that agreed band, both meters are considered healthy and the custody meter's reading is accepted.
The band that counts as acceptable is set by agreement between the parties and by the uncertainty of the meters involved, and it is chosen so that normal instrument scatter falls inside it while a genuine problem falls outside. A deviation that grows beyond the band, or that trends steadily in one direction rather than scattering randomly around zero, is the signal that one meter has changed - a steadily widening gap points to drift, while a sudden step points to a discrete failure or a change in conditions. The pattern of the deviation, not just its size, helps diagnose what has happened.
When a divergence appears, the check meter is what makes the investigation possible. Because there are now two independent measurements, the operator can prove or examine each meter to determine which one moved, rather than facing a single reading with no reference to test it against. This is also what makes check metering valuable in a dispute: if two parties disagree over a custody volume, the independent check measurement provides a second data point to reconcile against, and a documented history of the two meters agreeing supports confidence that the custody figure was sound.
Check metering only delivers its value if the comparison is actually watched, and that is where continuous SCADA monitoring turns two separate readings into an ongoing verification. Bringing both the custody and the check meter into one system lets the deviation between them be computed and trended automatically, so instead of comparing volumes manually after the fact, the operator sees the running difference in real time and its history over days and weeks. A comparison that has been steady and suddenly starts to widen is visible immediately rather than at the next audit.
A cloud SCADA such as Merobix reads both meters and can present their volumes side by side along with the calculated deviation, and it can raise an alarm when that deviation moves outside the agreed band. That automated watch is what makes check metering a continuous safeguard rather than a periodic manual exercise - the operator is told the moment the two meters begin to disagree, which is precisely when a drifting or failing meter can be caught before it has biased a large volume. The logged deviation history also becomes the evidence trail that supports the measurement in an audit or dispute.
For an operator responsible for measurement across many custody points, this dual-meter visibility scales the assurance without scaling the manual effort. Rather than periodically reconciling meter readings by hand, the operator relies on SCADA to watch every check-metered point continuously and to flag only the ones where the two meters have diverged. The documented, time-stamped record of the two meters agreeing day after day is what underpins confidence in the custody figures, and the alarm on divergence is what ensures a problem is investigated while it is still small.
A check meter is a second meter permanently installed alongside the custody meter that continuously compares readings against the same flow, so it catches drift as it happens between provings. Meter proving is a periodic verification in which the custody meter is checked against a reference standard such as a prover to establish a correction factor at that moment. They are complementary: proving establishes accuracy at intervals, while a check meter provides continuous assurance in between and an independent reference if a dispute arises.
No two meters agree exactly because each carries its own small uncertainty, so the comparison is judged against an agreed acceptable band rather than expecting zero difference. The band is set by the parties and by the meters' uncertainties, chosen so that normal instrument scatter falls inside it while a genuine problem falls outside. A deviation that grows beyond the band, or that trends steadily in one direction instead of scattering around zero, is the signal that one meter has changed and needs investigation.
In a dispute over a custody volume, a single custody meter offers only one reading with nothing independent to test it against. A check meter provides a second, independent measurement of the same flow, so the parties have another data point to reconcile against, and each meter can be proved to determine which one is accurate. A documented history of the two meters agreeing also supports confidence that the custody figure was sound, which is why check metering is common where disputes have occurred or are anticipated.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.