Manual Reconciliation Leaves Imbalances Unresolved for Too Long
The reconciliation workflow is known: pull tank battery gauge tickets and truck tickets from Quorum, export OSIsoft PI flow meter readings by lease and gathering segment, load both into Excel, identify discrepancies, trace the source, and generate correction requests for SAP posting. At 0.8% average imbalance generating $1.2M annually in unrecovered revenue, the question isn't whether the reconciliation matters — it's whether a monthly manual process in Excel is the right way to run it.
Weekly Imbalance Detection, Source-Level Tracing, Human-Approved Corrections
A measurement reconciliation agent reads ticket data from Quorum and cross-references it against OSIsoft PI flow meter readings by lease and gathering segment on a weekly cadence. When volume imbalances are detected, the agent traces the likely source discrepancy — meter drift, tank gauge variance, truck ticket mismatch — and generates an imbalance resolution report. The measurement director reviews the findings and approves allocation corrections before they are posted in SAP. The weekly cycle replaces a monthly manual process.
Targeting Recovery of $800K+ in Annual Imbalance Revenue
This is a direct revenue recovery story. The agent is designed to target recovery of $800,000 or more of the $1.2M annual imbalance by catching discrepancies weekly rather than monthly, and by automating the identification of the source — which is the step that currently takes most of the manual time. Manual reconciliation effort drops 70–90%. The agent is live and processing measurement data in approximately 4 weeks. For a gathering operator with multiple leases and segments, the cumulative recovery across a full year makes the economics straightforward.
How does the agent determine the source of a volume discrepancy?
The agent cross-references Quorum ticket records against OSIsoft PI readings at the lease and segment level, flagging where the divergence occurs in the flow path. It categorizes discrepancies by likely source type — meter variance, gauge ticket gap, truck ticket mismatch — so the measurement team knows where to investigate first.
Are allocation corrections applied automatically?
No. The agent generates a resolution report and queues the proposed correction, but the measurement director reviews and approves before anything is posted in SAP. Human sign-off is required for every allocation correction.