Introduction
Most instrumentation on a facility exists to control the process or protect it. Metering is different: it exists to settle money. The fiscal meter at the export point is where the operator's production becomes a number that a buyer pays for and a government taxes, and that number has to be defensible to an auditor, a partner, and a regulator years after the flow passed through. A few tenths of a percent of bias on a large export stream is a meaningful sum every single day — which is why metering attracts a level of accuracy, redundancy, and paperwork that the rest of the instrument scope never sees.
On a facility that gathers production from more than one field or more than one owner — increasingly common as marginal fields tie back to shared early production facilities — there is a second metering problem on top of the first: allocation. The total measured at export has to be shared back to the contributing wells and owners fairly, and that sharing is its own measurement discipline with its own accuracy expectations.
This post covers the distinction between custody-transfer and allocation metering, the meter technologies and where each fits, how the uncertainty budget is built, and how allocation works when several parties share one facility.
Custody Transfer vs Allocation — Why the Distinction Matters
The two jobs look similar — both measure flow — but they answer different questions and carry different accuracy obligations:
- Custody-transfer (fiscal) metering measures hydrocarbons changing ownership or crossing a tax boundary. It is the legal cash register: highly accurate, regulated, and subject to defined standards and approvals. Typical targets are around ±0.25–0.5% for liquids and ±1.0% for gas, with the regime (for example the UK measurement guidelines, plus API MPMS, AGA, and ISO 5167) prescribing how it is built, proved, and reported.
- Allocation metering divides a commingled total back to its sources. It does not cross an ownership boundary by itself; it apportions the fiscal total among wells or owners. A lower absolute accuracy is acceptable because errors largely net out across the allocation — but it must be consistent and unbiased, because a systematic skew always favours one party over another.
The practical consequence: you spend the accuracy budget at the custody-transfer point and you spend the fairness budget on the allocation scheme. Confusing the two — over-engineering an allocation meter, or under-specifying a fiscal one — wastes money in one place and creates exposure in the other.
Meter Technologies and Where Each Fits
There is no universal meter; each technology trades accuracy, rangeability, pressure loss, and cost differently:
| Technology | Best for | Notes |
|---|---|---|
| Orifice plate (AGA-3 / ISO 5167) | Gas, well-understood service | Cheap, robust, fully standardised — but square-root response limits rangeability to ~3:1, permanent pressure loss, and accuracy depends on the installation (straight lengths, plate condition) |
| Ultrasonic (multipath) | Gas and liquid export, wide turndown | No moving parts, very low pressure loss, wide rangeability — but needs a known flow profile (straight lengths or flow conditioner) and is sensitive to fouling and wet gas |
| Coriolis | Liquid custody transfer, dosing | Measures mass directly plus density — very accurate and turndown-rich, independent of flow profile — but size and cost limit it on large lines, and gas voids upset it |
| Turbine | Clean liquid custody transfer | Excellent repeatability when proved regularly via a prover — but moving bearings wear, and it needs clean fluid and proving |
The choice follows the service: a large gas export line is usually orifice or multipath ultrasonic; a liquid custody point is often Coriolis or proved turbine; an allocation point can use a less demanding device because the consistency, not the absolute accuracy, is what matters. Whatever the primary device, custody streams are built with redundancy — duty/standby runs, a master meter or prover for in-situ proving, and a flow computer doing the standards-based correction.
The Uncertainty Budget
A fiscal measurement is never just the primary device. The reported quantity comes from a chain, and the overall uncertainty is the individual contributions combined in quadrature (root-sum-square), per ISO 5168:
- The primary element (orifice discharge coefficient, ultrasonic path geometry, Coriolis calibration).
- The secondary instruments — differential-pressure, static-pressure, and temperature transmitters — each with its own uncertainty.
- The fluid properties — density for liquids, composition and compressibility (z-factor) for gas, brought in from an on-line densitometer, a gas chromatograph, or a sampling regime.
- The flow computer algorithms — standard-volume conversion, and for gas the energy calculation (calorific value per ISO 6976, compressibility per AGA-8). Gas is increasingly sold on energy, not volume, so the calorific value is part of the cash measurement.
- Proving / calibration — how recently and how well the meter was proved against a reference.
Because the contributions combine in quadrature, the largest single term dominates the result — so there is no point buying a 0.1% transmitter to pair with a 1% composition measurement. The budget tells you where to spend: tighten the dominant contributor, not the one that is already small. A meter that is mechanically perfect but fed a stale gas composition is not an accurate fiscal measurement.
Allocation on a Shared Facility
When several fields or owners feed one facility, the fiscal total at export is real money that must be split. The mechanics:
- Each source is characterised by an allocation meter or by well tests (periodic measurement of each well's rate and composition), giving each contributor a theoretical production figure.
- The fiscal export total rarely equals the sum of the theoretical inputs — there are measurement errors, line pack, and process losses. The difference (the imbalance or mismatch) has to be distributed back, usually pro-rata to each party's theoretical share, so that the allocated totals sum exactly to the fiscal measurement.
- Composition matters as much as volume: owners are allocated their own oil, gas, and water, so the allocation has to track quality, not just total flow — especially where streams of different value commingle.
The fairness principle is that no party should be systematically advantaged by the scheme. Smaller fields suffer disproportionately from allocation uncertainty — a fixed measurement error is a larger fraction of a small contributor's share — which is why a marginal field tying into a shared facility cares intensely about the allocation methodology written into the agreement. The metering engineer and the commercial agreement are tied together here: the scheme has to be technically sound and contractually agreed before first oil.
Worked Example — Gas Export Uncertainty
Scenario: a gas export stream is measured for custody transfer. The question is whether the chain meets a ±1.0% target on energy.
The contributions (illustrative):
- Multipath ultrasonic meter: ±0.5%
- Static pressure transmitter: ±0.2%
- Temperature transmitter: ±0.15%
- Gas composition / calorific value (ISO 6976 from a chromatograph): ±0.7%
- Compressibility (AGA-8): ±0.2%
Combined in quadrature:
√(0.5² + 0.2² + 0.15² + 0.7² + 0.2²) = √(0.25 + 0.04 + 0.0225 + 0.49 + 0.04) ≈ √0.8425 ≈ ±0.92%
The chain meets the ±1.0% target — but only just, and the composition term (±0.7%) dominates. If the gas composition drifts and the chromatograph is not maintained, that term blows out and the whole measurement fails its target, no matter how good the ultrasonic meter is. The budget says plainly where the money and attention belong: keep the composition measurement honest. Buying a more accurate flow meter while neglecting the GC would improve the smallest contributor and leave the result essentially unchanged.
Common Pitfalls
- Confusing custody-transfer and allocation requirements. Over-specifying an allocation meter wastes money; under-specifying a fiscal one creates real exposure. Match the accuracy to the job.
- Optimising the flow meter and ignoring composition. On gas energy measurement the calorific-value term often dominates the uncertainty budget. A great meter fed a stale composition is not an accurate measurement.
- Measuring volume when the contract pays for energy. Gas sales are increasingly on energy. If the flow computer is not doing the ISO 6976 / AGA-8 calculation correctly, the cash number is wrong even if the volume is right.
- Forgetting rangeability. An orifice plate sized for design rate is inaccurate at turndown (square-root response, ~3:1 range). If the stream varies widely, choose a wide-turndown technology or stage the meters.
- Neglecting proving and calibration. Fiscal accuracy decays without regular proving against a reference. A meter is only as accurate as its last proving — design the prover or master meter in.
- Leaving allocation methodology to the engineers alone. The allocation scheme is a commercial agreement as much as a measurement. Agree it contractually — especially for small fields tying into a shared facility — before first oil, not after a dispute.
Conclusion
Metering is the one instrumentation discipline where the deliverable is money, not control. The fiscal meter is the cash register that turns production into a number a buyer pays and a regulator taxes; the allocation scheme decides how that number is shared among the wells and owners that produced it. Both demand a rigour the rest of the instrument scope never sees.
Match the technology to the service, build the uncertainty budget honestly and spend the accuracy where the dominant term lives — usually composition on gas, not the flow meter itself — and prove the meters on a schedule. Where a facility is shared, treat the allocation methodology as a commercial document agreed before first oil, because a systematic bias is not a rounding error, it is mis-paying a partner every day the field produces.
