A discrepancy is not a root cause. It is a difference between two values, two systems, or two expectations that needs to be explained. The explanation may sit in sampling, analysis, calculation, communication, or comparison method.
The first task is to define exactly which values disagree and under what conditions.
Define the comparison properly
Before troubleshooting, confirm that the two values are being compared on the same basis. Time alignment, pressure basis, normalisation method, averaging period, and reference conditions all matter.
Many apparent discrepancies are comparison errors rather than measurement errors.
Trace the chain from source to result
Work from raw inputs through each processing stage to the final output. For energy values, this often means checking gas composition, calculation basis, flow input, communication mapping, and totalisation logic in sequence.
A structured trace is more reliable than jumping straight to the final number.
Use independent evidence where possible
Archived reports, trends, diagnostic logs, field checks, and independent tests all help establish whether the discrepancy is persistent, intermittent, real, or apparent. Good evidence reduces the risk of correcting the wrong problem.
Confidence grows when several independent clues align.
Aim for explanation, not only correction
It is possible to make two values match without understanding why they differed. That is not good enough in important measurement systems. The investigation should end with a clear technical explanation and a defensible corrective action.
A solved discrepancy should also become a learning record.