When several wells produce into a common gathering system before anything is separated, the operator still has to work out how much each well contributed, but there is no separate meter on each well's oil, gas, and water. Full well stream allocation solves this for the case where the stream being measured is the unseparated, multiphase mixture straight off the wells. It divides a single measured total stream back to the individual wells using the wells' periodic well tests as the basis for their relative contributions. This guide explains what full well stream allocation is, how the wet stream and the well tests combine to split production back to each well, and how it differs from allocation done after separation.
Full Well Stream Allocation in one line: Full well stream allocation is the method of dividing a single, unseparated multiphase stream, measured before any separation, back to the individual wells that produced it. Because no well has its own dedicated phase metering, each well's share is estimated from its periodic well test, which characterises what that well produces, and those tested rates are used as the ratios to apportion the total measured stream. It is the pre-separation case of back-allocation, applied to the full wet stream rather than to separated oil, gas, and water.
The defining feature of full well stream allocation is that the measurement point sees the whole wet stream, oil, gas, and water together, before it has been separated into phases. Several wells flow into a common line, and a meter, or a measurement of the combined stream, captures the total the wells collectively delivered over a period. What it does not capture is which well delivered what, because the wells are commingled upstream of the measurement and there is no per-well metering of the individual phases. The allocation exists to answer that missing question.
This is why the phrase full well stream matters. The stream being allocated is the complete, unprocessed production as it comes off the wells, not the clean, separated oil or gas that emerges downstream of a separator. Allocating this raw multiphase mixture is a harder problem than allocating a separated single-phase stream, because the total itself is a blend of phases in proportions that vary from well to well, and the goal is to hand each well back its share of that blend consistent with what the well actually produces.
Because the individual wells are not separately metered, the allocation cannot be a direct measurement and is instead an estimate anchored to a measured total. The one thing known with the most confidence is the combined stream, and the allocation's job is to distribute that known total among the contributing wells in the right proportions. Everything therefore hinges on how those proportions are established, which is where the well test comes in.
The proportions come from well tests. Periodically each well is routed to a test facility and measured on its own for a period, which characterises what that well produces, its rate and the mix of phases in its stream. Between tests, the well is assumed to keep producing in roughly the proportions its last test showed. The test rates for all the wells feeding the common stream become the ratios by which the measured total is divided: a well whose test showed it producing a larger share of the group's total is allocated a proportionally larger share of the measured stream.
Mechanically, the allocation compares each well's tested contribution against the sum of all the wells' tested contributions to form a fraction for each well, and then applies those fractions to the actual measured total for the period. If the tests together imply the wells should have produced a certain combined amount, and the real measured stream came in different, the fractions still govern how the real total is split, so the measured stream is honoured as the true quantity and the tests only decide the shares. This is the essence of back-allocation: measure the aggregate accurately, and use the tests to distribute it.
The quality of full well stream allocation therefore rests directly on the well tests, and its main sources of error follow from them. Tests are snapshots taken infrequently, while wells change between tests as they decline, water out, or shift in gas ratio, so a well's true share can drift away from its last test before the next one corrects it. This is why test frequency and test quality matter so much here, and why a well suspected of changing significantly is a candidate for an earlier test. The measured total keeps the aggregate honest, but only fresh, representative tests keep the split among wells honest.
Full well stream allocation is data-hungry in a very practical way: it needs the measured total stream for every period and it needs the well tests kept current and correctly associated with the right wells, and pulling those together across a field is exactly the kind of task field data systems and SCADA support. A cloud SCADA such as Merobix can gather the combined stream measurement and the operational data around it, and hold the well-test results that set the allocation fractions, so the inputs the allocation depends on live in one place rather than scattered across sites and spreadsheets.
Visibility into the state of the well tests is where monitoring earns its keep for this method, because the allocation is only as good as the tests behind it. Surfacing when each well was last tested, and flagging wells whose tests are stale, helps an operator keep the test programme current so the fractions stay representative. Trending the measured total and the operational signals can also hint when a well has changed materially since its last test, for instance if the character of the combined stream shifts, which is the cue to schedule an earlier test before the allocation drifts too far.
It is worth being clear about the division of labour. The allocation arithmetic, dividing the measured total by the tested fractions, is a calculation performed in an allocation or production accounting system, not something the SCADA does in place of it. What cloud monitoring contributes is reliable, centralised, and current inputs: an accurate measured stream, well tests that are visible and up to date, and early warning when a well's behaviour suggests its share has changed. For fields with many wells commingled into unseparated streams and lightly staffed sites, keeping those inputs trustworthy in one accessible system is what makes full well stream allocation defensible period after period.
Full well stream allocation divides the complete unseparated multiphase stream, measured before any separation, back to the wells, whereas allocation after separation works with clean, single-phase oil, gas, or water that has already been split out by a separator. Allocating the raw wet stream is harder because the total is a blend of phases in proportions that vary by well. Both use well tests to set the shares, but the full well stream case applies them to the commingled multiphase total rather than to separated phases.
Because the individual wells are commingled before the measurement point and are not separately metered, so their relative contributions cannot be measured directly and must be estimated. Periodic well tests characterise what each well produces, and the tested rates become the ratios used to split the measured total stream among the wells. The measured total is honoured as the true quantity, and the tests only decide each well's share of it, which is the defining logic of back-allocation.
The main source of error is the well tests going stale. Tests are infrequent snapshots, but wells change between them as they decline, water out, or shift in gas ratio, so a well's true share can drift away from its last test before the next test corrects it. Keeping test frequency and quality high, and testing a well earlier when it is suspected of changing significantly, is what keeps the split among wells accurate, since the measured total only guarantees the aggregate is right.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.