How Crude Oil Volume Reconciliation Works (And Why It Keeps Taking So Long)

Every month, gathering operators across the Permian, Bakken, and Eagle Ford face the same ordeal: reconciling crude oil volumes across dozens of receipt and delivery points, hunting down variances in spreadsheets, and trying to close the books before the settlement deadline. For a facility running six LACT units, this process can consume three to five days of manual work — every single month.

The frustrating part? Most of that time isn't spent on actual analysis. It's spent collecting data from different systems, re-keying numbers, and chasing down discrepancies that an automated system would have flagged on day one. This post walks through exactly how crude oil volume reconciliation works at a gathering facility — step by step — so you can see where the bottlenecks are and what it looks like when software handles the heavy lifting.

What Is Crude Oil Volume Reconciliation?

Volume reconciliation is the process of comparing every barrel that enters a gathering system against every barrel that leaves it. The goal is simple: the numbers should balance. Oil received from producers at LACT receipt points should equal oil delivered to the pipeline, refinery, or truck loading facility at delivery points — minus a small, contractually agreed pipeline loss allowance (PLA).

When the numbers don't balance, you have a variance. Variances aren't just accounting problems — they represent real money. A 0.5% unresolved variance on a 50,000-barrel-per-day facility is 250 barrels. At $70 per barrel, that's $17,500 every day that someone is either overpaying or underpaying.

Reconciliation exists to find those variances, diagnose their causes, and resolve them before settlements go out. It protects both the gathering operator and the producers they serve.

The Five Stages of Volume Reconciliation

Whether you're doing reconciliation manually or with software, the process follows the same logical stages. Here's how it works at a mid-size gathering facility with six LACT receipt bays and two delivery meters.

Stage 1: Data Ingestion

Everything starts with getting the data into one place. For each LACT unit, you need:

• Meter volumes: Gross observed volume (GOV), gross standard volume (GSV), and net standard volume (NSV) from each flow computer.
• Quality data: BS&W readings, API gravity from lab analysis, and temperature/pressure at time of measurement.
• Meter factors: The current proving-derived meter factor for each LACT.
• Batch records: Start/stop times, opening/closing totalizer readings, and any alarm events during the batch.

In a manual operation, this data comes from multiple sources: SCADA exports, TransLog or Microload electronic files from flow computers, handwritten field tickets, and lab reports emailed as PDFs. Someone has to collect all of it, verify the formats, and key it into a master spreadsheet.

With reconciliation software, data ingestion is automated. Flow computer data feeds directly into the system via SCADA integration or file imports. Lab results are entered once. Field tickets are digitized. The system normalizes everything into a single data model — no re-keying, no format conversion, no waiting for someone to email a spreadsheet.

Stage 2: Validation

Raw data is never ready for reconciliation as-is. Before you can compare volumes, you need to validate that the data is complete, consistent, and correct. Validation catches problems like:

• Missing data: A LACT unit that didn't report for 6 hours — did it shut down, or did the SCADA feed drop?
• Duplicate records: The same batch appearing twice because a flow computer retransmitted after a communication error.
• Out-of-range readings: A BS&W reading of 15% when the well typically runs at 1.5% — is the probe fouled, or did the well start producing excess water?
• Meter factor anomalies: A meter factor that jumped 0.003 between provings — normal drift, or a mechanical problem? (The significance of a meter factor change depends on meter type and operating conditions — for a Coriolis meter, 0.003 would warrant immediate investigation, while for some PD meters in variable crude, it may be within normal drift.)
• Temperature correction errors: Volumes reported at observed temperature rather than standard 60°F.

In a spreadsheet, validation means manually scanning every row, cross-referencing against expected ranges, and flagging anything that looks off. For six LACT units generating hundreds of batch records per month, this is tedious and error-prone. It's the stage where mistakes creep in quietly — a duplicate batch that inflates receipt volumes by 800 barrels, an uncorrected temperature reading that shifts a settlement by thousands of dollars.

Automated validation runs configurable rules against every record as it arrives. Duplicates are flagged instantly. Out-of-range values trigger alerts. Missing data gaps are identified in real time instead of discovered during month-end close. The system doesn't get tired at 4 PM on the last day of the month.

Stage 3: Volume Balancing

This is the core of reconciliation: comparing total receipts against total deliveries and identifying the variance. The basic formula is straightforward:

Let's walk through a real example. Your facility has six LACT receipt bays and two delivery meters. In a given month:

In this example, 242,600 barrels came in, 241,450 went out, and 849 barrels are within the contracted PLA. That leaves a net variance of 301 barrels — about 0.12% of throughput. That's within normal tolerance for most gathering contracts, but it still represents over $21,000 at $70/bbl. Whether that warrants investigation depends on the contract terms and the operator's risk tolerance.

The point of this stage is to surface variances quickly. In a manual process, building this table requires pulling data from six different LACT flow computers and two delivery meters, ensuring all volumes are corrected to the same standard conditions, and checking that the time periods align. That alone can take a full day. Software generates this view automatically, updated in real time as data flows in.

Stage 4: Variance Investigation

When variances exceed the acceptable threshold, you need to find the cause. Common sources of reconciliation variances include:

• Meter drift: A receipt meter reading 0.2% high because it hasn't been proved recently. This inflates receipts relative to deliveries.
• Temperature correction mismatches: One meter reporting at 60°F standard while another reports at observed temperature. Even a 5°F difference can swing volumes by 0.3–0.5%.
• BS&W discrepancies: The inline probe reads 1.0% but the lab sample shows 2.5%. Which one do you use for the net volume deduction? The answer differs by contract.
• Timing mismatches: A batch that closed at 11:58 PM on the 31st at the receipt LACT but wasn't recorded at the delivery meter until 12:03 AM on the 1st. Those barrels show up in different reporting periods.
• Data entry errors: A transposed number in a manually entered field ticket. LACT 4 delivered 29,870 barrels but someone keyed 28,970 — a 900-barrel error hiding in plain sight.
• Actual losses: A small leak in a gathering line, or vapor losses from a tank battery. These are real volume losses, not measurement errors.

Manual investigation means pulling the raw data for each suspect meter, recalculating volumes by hand, checking proving records, calling field operators to verify readings, and comparing against historical patterns. For a single variance, this can take hours.

With software, investigation starts from a drill-down: click on the variance, see which meters are contributing, compare against historical baselines, and view the underlying batch records with all corrections applied. The system can automatically flag the most likely cause based on patterns — "LACT 3's meter factor changed 0.004 mid-month; re-prove recommended" — so the analyst knows where to focus.

Stage 5: Approval and Settlement

Once variances are investigated and either resolved or documented, the reconciliation needs sign-off before settlements can be generated. This is where the balance report becomes the official record.

The approval process typically involves:

• Measurement technician review: Confirms that all meter data is accurate, meter factors are current, and any corrections have been applied properly.
• Accounting review: Verifies that volumes tie to the settlement statements being generated, contract terms (PLA percentages, pricing tiers) are applied correctly, and any variance adjustments are documented.
• Management approval: Final sign-off that the reconciliation is complete and settlements can be issued.

In a spreadsheet-based process, "approval" often means emailing a workbook around and hoping everyone reviews the right version. There's no audit trail showing who changed what, when, or why. When a producer disputes a settlement six months later, reconstructing the reconciliation from email chains and spreadsheet versions is a nightmare.

In a software-based workflow, approval is structured: each reviewer sees the same data, comments and adjustments are tracked, and the final approved reconciliation is locked with a full audit trail. When disputes arise, the evidence is one click away.

Why Manual Reconciliation Takes So Long

Looking at those five stages, the bottlenecks are obvious:

• Data collection alone takes 1–2 days. Gathering data from six LACT flow computers, two delivery meters, lab reports, and field tickets — and getting it all into one spreadsheet in a consistent format.
• Validation is done by eyeball. Scanning hundreds of rows for anomalies, duplicates, and missing data. The human eye misses things, especially under deadline pressure.
• Volume balancing is formula-heavy. Every LACT needs temperature corrections (API MPMS Ch. 11), BS&W deductions, and meter factor applications. One wrong cell reference in the spreadsheet cascades through the entire reconciliation.
• Variance investigation is reactive. You don't discover the problems until month-end, by which point the evidence (real-time sensor data, operator observations) may be stale.
• Approval is informal. No structured workflow, no audit trail, no way to prove what was reviewed and by whom.

Add it all up and a facility with six LACT units typically spends 3–5 days on month-end reconciliation. Multiply that by 12 months, and you're looking at 36–60 person-days per year consumed by a process that could be running continuously in the background.

What Changes With Automated Reconciliation

The five stages don't go away with software — the physics and accounting don't change. What changes is when and how the work happens:

• Data ingestion is continuous. Flow computer data feeds into the system daily or in real time. No month-end data collection scramble.
• Validation runs automatically. Every record is checked against configurable rules as it arrives. Duplicates, gaps, and anomalies are flagged immediately — not discovered on day 28.
• Volume balancing updates daily. The balance report is always current. By the time month-end arrives, the reconciliation is already 95% done.
• Variance detection is proactive. A 500-barrel variance that appears on day 5 gets investigated on day 5, when field data is fresh and operators remember what happened.
• Approval is structured. Digital review workflows with comments, adjustments, and full audit trail.

The result: month-end close goes from 3–5 days to a few hours of review and sign-off. The reconciliation work hasn't disappeared — it's been redistributed from a painful monthly sprint into a continuous, automated background process.

See It in Action

COYOTE Measurement's reconciliation engine automates all five stages of crude oil volume reconciliation. It ingests data from LACT flow computers (including TransLog and Microload formats), validates every record against configurable rules, generates real-time balance reports, and flags variances as they occur — not at month-end.

If you're spending days on spreadsheet-based reconciliation every month, schedule a demo to see how it works with your data.