What Is an Oilfield Test Separator: Well Testing, Sizing and Instrumentation

What Is an Oilfield Test Separator: Well Testing, Sizing and Instrumentation

Answering the core question: What is an oilfield test separator? An oilfield test separator is a portable, skid-mounted pressure vessel used to measure the individual oil, gas and water production rates of a single well during a production test. It is functionally a small production separator, typically 0.3 to 2 m in diameter, but it is engineered for mobility and measurement rather than for continuous duty: the vessel is mounted on a skid or a trailer with its own piping, valves, meters, control panel and often a burner or a tank, so it can be moved between wells and connected to a wellhead or a test manifold. A test runs for 4-24 hours and produces the gas-oil ratio, the water cut, the API gravity and the absolute rates, with an accuracy of plus or minus 2-5% when correctly operated. That data drives allocation, reservoir management and workover decisions.

1. What a Well Test Actually Measures and Why

The purpose is not merely to separate the fluid but to produce defensible numbers, and four quantities are obtained:

• Oil, Gas and Water Rates: The three phase rates are the primary output. Gas is metered on the vapour outlet, usually with an orifice meter for its simplicity and its wide acceptance, or with a turbine, a Coriolis or an ultrasonic meter where a wider turndown or a direct mass measurement is required; a 10:1 to 100:1 turndown is needed because a single skid may test both a high-rate and a marginal well. Oil and water are metered on their respective outlets with positive displacement, turbine or Coriolis meters, with Coriolis increasingly preferred because it measures mass and density directly and therefore gives the water cut and the API gravity from the same instrument. Rates are normally reported at standard conditions, so the separator pressure and temperature are recorded continuously and a shrinkage factor is applied to correct the oil volume from separator to stock tank conditions.
• Gas-Oil Ratio and Water Cut: The gas-oil ratio, expressed in standard cubic feet per barrel, and the water cut, as a percentage of the total liquid, are the two derived numbers that matter most for reservoir management. GOR typically ranges from 50 to 5,000 scf/bbl, and a rising GOR over successive tests signals gas coning or breakthrough, which is one of the classic triggers for a workover or a rate reduction. Water cut ranges from 0 to 95%, and a rising trend maps water encroachment, channeling or a failed cement job. Both are more useful as a trend than as a single value, which is why tests are repeated on a schedule, typically monthly or quarterly for producing wells, and why the test conditions, including the choke size and the flowing pressure, must be recorded so that successive tests are comparable.
• Fluid Sampling and PVT Analysis: A well test is usually the only practical opportunity to obtain a representative fluid sample, and the separator is where it is taken. Oil and gas samples are collected at the separator under pressure in special sampling cylinders, and recombined in the laboratory in the measured producing gas-oil ratio to reconstruct the reservoir fluid. From that, a pressure-volume-temperature analysis gives the bubble point, the formation volume factor, the solution gas ratio and the viscosity, all of which are required for reservoir simulation and for material balance calculations. Getting a valid sample requires the well to be flowed steadily at a rate below the critical rate long enough to stabilise, which is why a test is often preceded by a clean-up period of several hours to a day.
• Well Performance and Deliverability: Beyond simple rates, a properly run test establishes the relationship between the flowing pressure and the production rate, from which the productivity index and, for gas wells, the absolute open flow potential are derived. A multipoint test flows the well at three or four increasing choke sizes, holding each until the pressure and the rates stabilise, and plots the rate against the pressure drawdown. The result tells the engineer how much rate is available for a given drawdown, whether there is formation damage indicated by an unexpectedly low productivity index, and whether the well is being produced above its critical rate, which causes coning and sand production. This is the information that justifies or rules out a stimulation, a workover or a rate change.

2. Equipment Design and Test Procedure

Three design features and one discipline separate a good test from a worthless one:

• Vessel Sizing for a Portable Unit: A test separator is sized by the same Souders-Brown gas capacity as a production vessel, but with two additional constraints. It must handle the full range of rates expected across all the wells it will serve, since one skid is moved from well to well, which pushes the design towards a wider turndown and sometimes towards interchangeable internals or a second smaller skid for marginal wells. And it must be transportable, which limits the diameter and the length to road-legal dimensions and therefore limits the gas capacity; where a high-rate well exceeds the capacity, the test is run at reduced rate through a smaller choke, or two skids are operated in parallel. Liquid retention is shorter than in a production separator, typically 1-5 minutes, because the objective is measurement rather than maximum separation efficiency.
• Instrumentation and the Metering Package: The metering package is what makes the skid a test separator rather than just a small separator. It includes the gas meter with its pressure and temperature transmitters and, for an orifice plate, a differential pressure transmitter with a range that matches the expected flow; the liquid meters with their own temperature and density measurement; a level controller on each liquid outlet that must hold steady, since a hunting level controller produces oscillating rates and ruins the test; a pressure controller on the gas outlet that holds the separator pressure steady, typically 2-5 bar above the flowline; and a data acquisition system that logs everything at intervals of a minute or less so that stability can be demonstrated rather than assumed. All meters are calibrated on a defined schedule and the calibration certificates are part of the test record.
• Test Procedure and Stabilisation: The discipline is what produces a valid number. The well is flowed to clean up and to remove drilling or workover fluid, which may take several hours to a day. It is then flowed at the test rate through a fixed choke, and the rates, pressures and temperatures are monitored until they are stable, typically defined as less than 2-5% variation over a period of one to four hours depending on the standard applied. Only then does the formal test period begin, running for 4-24 hours depending on the well and the purpose. The choke size, the wellhead pressure and temperature, the separator pressure and temperature and the tank or meter readings are all recorded. A test stopped early, or started before stabilisation, produces numbers that look reasonable but are wrong, and since these numbers drive allocation and investment decisions, that error is expensive.
• Safety, Layout and Environmental Controls: Well testing is one of the higher-risk routine operations in production, and the skid must be designed for it. The vessel is built to ASME VIII Division 1 with a relief device sized for the blocked-outlet and fire cases and routed to a flare or a burner boom, since a test separator is a temporary installation that often has no permanent flare. The skid is earthed and bonded, laid out with the wellhead upwind and with an escape route, and operated with gas detection where hydrogen sulphide is possible. Liquid is routed to a test tank whose volume is metered independently as a cross-check on the liquid meters, or to the production manifold if the test is an allocation test. Produced water and any hydrocarbon are contained rather than discharged, and the whole operation is covered by a written procedure and a permit.

Oilfield Test Separator Configurations Comparison Matrix

Frequently Asked Questions (FAQ)

Q: What is the difference between a test separator and a production separator?

A: A production separator runs continuously, is sized for the full field rate and for the whole production life, and is optimised for separation efficiency and for long residence times of 3-30 minutes. A test separator runs intermittently, is sized for a single well, is optimised for measurement accuracy and mobility, and accepts a shorter residence time of 1-5 minutes because its job is to measure rather than to treat. In practice the test separator is skid or trailer mounted with its own meters, control panel, data acquisition and often a tank or a burner, so it can be moved between wells, whereas a production separator is a permanent installation. Many facilities have both: a production separator handling the commingled stream and a test separator on a test manifold that can isolate and measure any individual well.

Q: How long should a well test run?

A: Long enough to demonstrate stability, and that is the criterion rather than a fixed clock time. Typical practice is a clean-up period of several hours to a day, followed by a stabilisation period during which the pressures and rates are watched until they vary by less than about 2-5% over one to four hours, and then a formal test period of 4-24 hours. Simple allocation tests on stable wells may be as short as 4-8 hours, while an initial test on a new well, a test for reservoir study or a deliverability test on a gas well may run 24-72 hours or longer. The key discipline is that the formal test period must not start until stability is demonstrated and recorded, because a test started early produces numbers that look plausible but are systematically wrong, and since those numbers drive allocation between partners they are difficult to correct later.

Q: Why is separator pressure and temperature recorded during a test?

A: Because all reported rates are at standard conditions, and the conversion depends on the actual conditions in the vessel. Gas volume depends on pressure and temperature through the gas law and the compressibility factor, so the measured actual flow must be corrected to standard cubic feet or standard cubic metres. Oil volume depends on temperature for thermal expansion and, more significantly, on pressure because the oil leaving a separator contains dissolved gas that flashes as the pressure falls on the way to the stock tank: the oil therefore shrinks between the separator and the tank, typically by 1-20% depending on the pressure and the crude, and a shrinkage factor or a formation volume factor is applied to correct for it. Without accurate separator pressure and temperature records, the reported rates carry an error that is often larger than the meter error itself.

Q: Can a multiphase flow meter replace a test separator?

A: Increasingly yes for allocation and surveillance, but not entirely for formal well testing. A multiphase meter measures the gas, oil and water fractions and velocities without separating the stream, giving continuous rate data from a compact installation and eliminating the need to divert a well to a test separator. Typical accuracy is plus or minus 5-10% on each phase, against 2-5% for a well-run test separator, and the accuracy degrades at high gas volume fractions and at high water cut. Multiphase meters are therefore excellent for trending and for allocation where the accuracy is acceptable to the partners, and they are widely used offshore and in remote fields where a test separator is impractical. They do not, however, provide a representative fluid sample for PVT analysis, so a separator test is still required periodically for that purpose.