top of page

Automatic Crude Oil Sampling for Custody Transfer: What API MPMS 8.2 and ISO 3171 Expect

Writer: Tech Inc
Tech Inc
Nov 7
4 min read

In a custody transfer, the meter measures the volume and the sample decides how much of that volume is water and sediment. Both numbers go on the invoice. An error of 0.1% in the water content of a 500,000-barrel cargo is 500 barrels of oil that someone pays for and nobody receives. This guide is for terminal, pipeline and metering engineers who need to specify an automatic crude oil sampling system, or understand why an existing one is being challenged.

The standards in brief

  • API MPMS Chapter 8.2, published jointly as ASTM D4177, is the standard practice for automatic sampling of petroleum and petroleum products.

  • ISO 3171 covers automatic pipeline sampling of petroleum liquids and is widely used outside North America.

  • API MPMS Chapter 8.1 (ASTM D4057) covers manual sampling, which remains the fallback when an automatic system is out of service.

  • API MPMS Chapter 8.3 (ASTM D5854) covers the mixing and handling of liquid samples once they are collected.

The documents differ in detail, but they agree on the principle: a representative sample needs every part of the system to work, from the pipe to the laboratory. Always check requirements against the current editions named in your contract.

Where automatic samplers are used

Automatic sampling is expected wherever crude changes hands by meter: pipeline custody transfer points, LACT units at production sites, marine terminals loading or discharging tankers, tank-to-pipeline transfers and export lines from production facilities. In each case the sample result is applied to the full metered volume, so the cost of a biased sample scales with throughput.

Five elements of a representative sample

  1. A well-mixed stream at the sample point. Water and sediment must be dispersed across the pipe, either by natural turbulence or by a static or power mixer.

  2. A correctly designed and located probe. The probe opening faces upstream, sits in the well-mixed zone near the centre of the pipe, and is downstream of the mixing source.

  3. Flow-proportional grabs. Each grab is a fixed, repeatable volume, and grabs are taken in proportion to flow, so every part of the batch is represented by its share.

  4. Sealed receivers. The collected sample is held without losing light ends or water, and without contamination from the previous batch.

  5. Proper mixing before the laboratory draws its sub-sample. Water settles in a receiver within minutes, so the receiver must be mixed, and its homogeneity confirmed, before analysis.

A weakness in any one of these cannot be compensated by the others. A perfect sampler on a stratified stream still gives a wrong answer.

Mixing and the sample point

Crude and water separate quickly at low velocity, so the standards include methods for assessing whether the stream is adequately mixed at the probe location, based on pipe size, flow rate, fluid properties and upstream fittings. Where natural mixing is not enough across the full flow range, a static mixer or a power mixing system is added. Our article on stratification and power mixing covers the options in detail.

Tech Inc. sample receivers and laboratory mixing unit with company logo

Flow-proportional sampling and receivers

The sampler controller takes a flow signal from the meter, usually pulses from the flow computer, and triggers a grab each time a set volume has passed. Grab volume and pacing are chosen so the receiver fills to its target level by the end of the batch, without overfilling if the batch runs long. Too few grabs and a short-lived water slug can be missed; too many and the receiver fills early. The controller should recalculate pacing if the nominated batch volume changes.

Receivers are typically fixed or portable stainless steel vessels with a mixing system; for volatile crudes, constant-pressure receivers keep light ends in solution. When a batch ends, the receiver is mixed, the sample is drawn into laboratory containers, and the receiver is emptied and flushed before the next batch. A sample receiver and laboratory mixing unit lets the laboratory do this to a repeatable procedure rather than by hand-shaking.

Proving the system: the water-injection test

A sampling system should be proven, not assumed. The standards describe an acceptance test in which a known volume of water is injected upstream of the sampler during a controlled transfer. The water content measured in the collected sample, less the baseline water in the crude, is compared with the injected amount. If the result falls within the tolerance the standard sets, the system as installed, meaning mixing, probe, sampler, receiver and laboratory handling together, is shown to be representative. Confirm the test procedure and acceptance criteria in the current edition before planning the test.

Why sampling results get disputed

  • Time-based rather than flow-proportional grabs, so low-rate periods are over-represented.

  • A probe in a poorly mixed location, or one moved during a later modification.

  • Receivers not mixed thoroughly before sub-sampling, or not cleaned between batches.

  • Missed grabs that nobody noticed because the controller had no fill monitoring.

  • Laboratory handling that does not follow the sample-handling standard.

What the controller must do

  • Pace grabs from meter pulses or a 4–20 mA flow signal, with a configurable volume per grab.

  • Count grabs and compare expected with actual receiver fill, alarming on missed grabs or low fill.

  • Monitor receiver level or weight continuously, so a leaking or overfilling receiver is caught during the batch.

  • Switch receivers at batch change, and record the batch start, end, volume and grab count.

  • Interlock with the mixer and the fast-loop pump, so grabs are only taken when the stream is mixed and flowing.

  • Report to the flow computer, DCS or SCADA, and keep a batch history for audits.

Specify the system, not just the sampler

We design in-line and fast-loop automatic sampling systems, the automatic sampler controller and receivers in Vaughan, Canada and build them in Chennai, India, with NACE MR0175/ISO 15156 materials for sour crude and EN 10204 3.1 material certificates. Each system is factory acceptance tested, witnessed by the client or a third-party inspector. If you are choosing a layout, read fast-loop versus in-line automatic samplers. Browse our automatic sampling systems and the full oil and gas systems range. Send your line size, flow range, crude properties and batch sizes to mail@techincresearch.com or call +91 739 749 8656 and we will return a P&ID, GA drawing and priced offer.

Comments


bottom of page