What Is Pipeline Batching?

Back to Blog

How Batches Move Through the Line

In a batched (fungible or segregated) products pipeline, one product is pumped in behind another in a planned sequence. There is usually no mechanical separator between them; the liquids stay largely apart because turbulent pipe flow mixes them only in a limited zone at the boundary. That mixing zone, the interface, grows slowly with distance as diffusion and the pipe's velocity profile blend the two products.

The scheduling of the sequence matters. Shippers order the batches so that adjacent products are compatible - for example placing a lower-grade gasoline next to a higher grade rather than diesel next to jet fuel - which limits the value lost at the interface. Some operators inject a small separating batch or use a batching pig between highly incompatible products.

The Interface, Transmix, and Batch Tracking

The commingled material at the boundary is called interface or transmix. Because it is a blend of two products, it fails the specification of both, so it is cut out at the delivery point and either downgraded into a compatible batch, reprocessed, or sent to a transmix processing plant that redistills it. Managing interface volume - keeping it small and cutting it accurately - directly affects the economics of a products line.

Operators track each batch's location using the known pumped volumes, flow measurement, line fill, and density or other quality sensors that detect the interface passing a point. As a batch nears its destination, operators divert (cut) it into the correct tank at the right moment. In oil and gas, batching is standard on refined-products and NGL pipelines, and increasingly relevant where crude of different qualities shares a line. The interface-detection instruments and batch positions are monitored through the pipeline control system.

The Instruments That Find the Interface

Interface detection depends on some measurable property stepping from one product's band to the next. Density is the workhorse: an online densitometer at pump stations and delivery points trends the passing stream, and the interface announces itself as a ramp between two stable density bands. Where adjacent grades sit too close in density to discriminate, operators add other instruments - sonic velocity meters, optical or color-based detectors, or analyzers keyed to a property that genuinely differs between the two products. Which instrument fits is a product-pair and site-specific decision.

Detection works alongside dead reckoning. Batch tracking projects each boundary down the line from metered volumes and line fill, producing an expected arrival window at every station; the analyzer then confirms the actual edge. The deviation between expected and detected arrival is itself diagnostic - persistent earliness or lateness points at meter drift, line fill error, or slack line. The same meters and line-balance calculations feed the pipeline leak detection system, so batch tracking accuracy and leak detection quality tend to rise and fall together.

A Worked Example: Cutting a Batch at a Terminal

Here is the shape of a delivery cut with symbolic products A and B, A ahead of B in the line:

1. Batch tracking projects the A/B interface into an expected arrival window at the delivery meter.
2. The delivery densitometer trends steady inside product A's density band.
3. At the first sustained deviation from A's band, delivery swings from A's tank to the transmix tank.
4. The reading ramps through the mixing zone and settles into B's band.
5. When the reading holds inside B's band per the site's cut criteria, delivery is cut to B's tank.
6. The volume recorded between the two cuts is booked as transmix.

Where exactly the two cuts sit inside that density transition is a site- and product-specific decision. Cut wide and you downgrade good product into transmix; cut narrow and you risk contaminating a tank that must meet specification. The cut criteria come from the product specifications and site procedures, and the recorded cut volumes flow into custody transfer monitoring and each shipper's batch accounting.

Fungible vs Segregated Service

Products pipelines run in two commercial modes. In fungible service, shippers tender product meeting a common specification and receive equivalent product at the destination - not the same molecules. That lets the operator consolidate many small nominations into full batches, and it is how most refined-products systems run. In segregated service, a shipper's specific batch travels intact end to end, which branded, additized, or specialty grades require. Segregated service consumes more line capacity because batches cannot be pooled, and minimum batch sizes in either mode are set by the pipeline's tariff, not by physics.

Scheduling then becomes a repeating cycle of grades arranged so every adjacency is acceptable. Jet fuel gets the most protective treatment because aviation specifications are the least forgiving: sequences are built so interface material is always cut away from the jet batch, with downgrades flowing toward more tolerant products. A published cycle also gives shippers predictable windows to nominate into, which is why a batched products line feels more like a scheduled railway than an open pipe.

What Complicates Batching in Real Operations

Interface growth is not constant. The mixing zone lengthens with distance travelled and grows faster at low velocity, where flow is less turbulent and the velocity profile stretches the boundary. Shutdowns add a second mechanism: in a line with elevation changes, the denser product can migrate downhill through the lighter one while the line sits idle, smearing an interface that was tight when pumping stopped. Restart planning after an extended shutdown accounts for this, per the operator's procedures.

Midpoint activity complicates the map further. Injections and deliveries part-way along the line split batches, change volumes, and create new interfaces, so the tracking model has to represent every junction, not just the two ends. Slack line at high elevation points corrupts the volume arithmetic that dead reckoning relies on. None of this breaks batching - it is routine on large systems - but it explains why batch tracking is a serious software discipline rather than a spreadsheet.

Frequently Asked Questions