An online custody chromatograph runs continuously and feeds the energy measurement that settles gas transactions, so its accuracy has real money riding on it. But an online analyzer can drift, lose calibration, or develop a fault, and there needs to be an independent way to confirm it is telling the truth. That check is a cross-verification: a spot sample of the gas is captured in a cylinder and sent to a reference laboratory, and the lab's composition is compared against what the online GC reported at the same time. This guide describes how to verify an online custody GC by pulling a spot cylinder sample to a reference lab, how the two composition sets are reconciled, the tolerance that triggers an investigation, and how a host stores paired online-versus-lab records that can resolve custody-transfer disputes.
GC Cross-Verification in one line: Cross-verifying a custody gas chromatograph means capturing a representative spot sample of the flowing gas in a sample cylinder, sending it to a reference laboratory for an independent composition analysis, and comparing the lab result against what the online GC reported for the same gas. The two composition sets and the properties derived from them are reconciled, and if they agree within a defined tolerance the online analyzer is confirmed, while a difference beyond the tolerance triggers an investigation into the analyzer's calibration or the sampling. Paired online-versus-lab records provide the documented, independent evidence that resolves custody-transfer disputes over gas quality.
The verification starts with capturing gas from the same stream the online GC measures, into a sample cylinder, in a way that faithfully represents the flowing composition. This is more delicate than it sounds, because a poorly taken sample can misrepresent the gas, particularly at the heavy end, if the sampling drops liquids out or fractionates the gas. The sample is drawn following a proper spot-sampling procedure so the cylinder ends up holding gas of the same composition that was actually flowing, and it is labelled with the time and point so it can later be matched to the online analysis from that moment.
The cylinder is then sent to a reference laboratory that runs its own analysis on well-maintained, carefully calibrated equipment, producing an independent composition for the sampled gas. The point of using an external or reference lab is precisely that it is independent of the field analyzer: its result does not share the field GC's calibration, its columns, or its potential faults, so if the two agree, that agreement is meaningful, and if they disagree, the disagreement points to something in one of them rather than a shared error. The lab result becomes the reference against which the online GC is judged.
For the comparison to be fair, the lab result has to be matched to what the online GC reported for the same gas, which is why the sample time and point matter. The online analysis taken at or nearest the sampling moment is the one to compare, so that both numbers describe the same gas rather than gas that has since changed. Keeping the sampling record and the corresponding online analysis paired from the start is what makes the eventual reconciliation clean rather than an argument about which readings to compare.
Reconciliation compares the online and lab compositions component by component and, importantly, compares the properties derived from them, especially the heating value and relative density, because those are what actually drive the settlement. Small differences are expected: two analyses of the same gas by different instruments will never be identical, and both the online GC and the lab carry their own uncertainties. The question is not whether they match exactly but whether they agree closely enough that the online analyzer can be trusted for custody, which is where a tolerance comes in.
A tolerance is set, typically on the derived heating value and on key components, defining how far the online GC may differ from the lab reference before the difference is considered significant. Inside that tolerance the online analyzer is confirmed and continues in service; outside it, the difference triggers an investigation. That investigation looks at both sides: whether the online GC has drifted or lost calibration, whether its carrier or columns are degrading, and equally whether the spot sample was taken or handled poorly, because a bad sample can make a good analyzer look wrong. The tolerance is the trip point that separates normal instrument-to-instrument scatter from a genuine problem worth chasing.
It matters that a tolerance exceedance does not automatically condemn the online GC, because the sample is just as capable of being the culprit. A dropped-liquid sample or a mishandled cylinder can shift the lab's heavy-end result and create an apparent disagreement that has nothing to do with the analyzer. A sound investigation therefore treats the exceedance as evidence that something is off in the online GC, the sampling, or the lab, and works to identify which, rather than assuming the field instrument is at fault. Getting this right is what keeps the verification from either missing real drift or falsely flagging a healthy analyzer.
The lasting value of a cross-verification is not just the pass-or-fail at the time but the documented record it creates, and that record is most useful when the online analysis and its paired lab result are stored together with their times, the point, and the reconciliation outcome. A history of these paired records builds a track record of the online GC's agreement with independent references over time, which shows whether the analyzer is consistently accurate or gradually drifting, and it provides the evidence base for any question about the measurement's validity during a given period.
This documentation is exactly what custody-transfer disputes turn on. When two parties disagree about the quality of gas delivered over some period, the argument is settled by evidence, and a stored series of online analyses reconciled against independent lab references is strong evidence about what the gas actually was and whether the metering analyzer was accurate at the time. Being able to produce the paired records for the period in question, showing the online GC agreed with the lab within tolerance, is often what resolves a dispute in the operator's favor, or conversely reveals a real problem that has to be corrected retroactively.
A cloud SCADA platform such as Merobix is well placed to hold these paired records because it already stores the online analyses with their timestamps, and it can attach the corresponding lab result and reconciliation outcome to each verification event. It can trend the online-versus-lab difference over successive verifications so a slow drift becomes visible before it crosses tolerance, alarm when a verification exceeds tolerance, and flag when the next scheduled verification is due so the online GC is never left unchecked for too long. Keeping the online analyses, the lab references, and the reconciliation history in one place turns the periodic cross-check into a continuous, auditable record that both protects the operator and settles disputes with documented, independent evidence.
An online custody GC feeds energy measurement that settles real transactions, but it can drift, lose calibration, or develop a fault, and there needs to be an independent way to confirm it is accurate. A spot sample analyzed by a reference lab does not share the online GC's calibration, columns, or potential faults, so its agreement with the online analysis is meaningful confirmation and its disagreement is a genuine signal. The cross-check provides the independent evidence that the metering analyzer can be trusted for custody.
A tolerance is set on the derived heating value and key components defining how far the online GC may differ from the lab reference before the difference is considered significant, because two analyses of the same gas will never be identical. Inside the tolerance the analyzer is confirmed; outside it, the difference triggers an investigation. That investigation examines both the online GC, for drift or calibration loss, and the sampling and lab, because a poorly taken sample can make a healthy analyzer appear wrong.
When two parties disagree about the quality of gas delivered over a period, the dispute is settled by evidence, and a stored series of online analyses reconciled against independent lab references shows what the gas actually was and whether the metering analyzer was accurate at the time. Being able to produce paired records showing the online GC agreed with the lab within tolerance often resolves the dispute, while a documented exceedance reveals a real problem to correct. Storing these records with their times and outcomes makes the history auditable.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.