Pipeline pressure monitoring and leak detection for fuel pipelines

Pipeline pressure monitoring is the continuous reading of pressure transmitters at several points along a pipeline, shown to operators with history and alarms. A pressure drop that does not fit the operating pattern, such as a pump change or a valve move, can point to a leak. Locating it and judging its size needs extra methods, for example line balance or the analysis of pressure waves.

This guide explains how a monitoring system is built, how to read a pressure trend, which leak detection methods exist and where the limits lie. Newcomers get a plain explanation first in each section, and engineers get the design detail after it. The examples come from fuel pipelines between terminals and booster stations.

A pressure transmitter turns pressure into an electrical signal, usually 4-20 mA, and the rest of the system carries that signal to a screen. The pieces are the same on a short pump line and on a transmission pipeline, only the distances change.

Citia's CATALYST PressMon follows this chain. It links pressure transmitters on a pipeline to a PLC or HMI, an industrial network and a dashboard. Operators see each measuring point live, review history when they investigate a fault and get alarms when pressure leaves its range. Alarms and interlocks follow the facility's control design. For the signal side, see how 4-20 mA and HART instruments send data.

Typical alarm levels are low-low, low, high and high-high. Many sites add a rate-of-change alarm, since a fast fall is often more telling than a slow one, and a comparison between neighbouring points on the same pipe. Levels come from the line's design pressure and operating envelope, so they are set per pipeline.

Transmitter quality decides data quality. Mount each transmitter with an isolation valve and a vent so it can be checked without stopping the line, and keep its calibration record. Choose a range that fits the design pressure with margin for surge: a range that is too wide loses resolution, and one that is too narrow clips the peaks.

Pressure on a working pipeline is never flat. It rises when a pump starts, falls when a valve opens downstream and changes with the product in the line. A leak adds one more change, and the skill is telling it apart from the others.

Comparing neighbouring points is powerful. If a drop appears at the transmitter just downstream of a pump and not further along, the cause is probably the pump or its valve. If it travels along the line, something is happening in the pipe. Event history from the control system, such as pump starts and valve commands, is what allows this reading.

Operators also need context on the same screen. Show pump status, valve positions and flow next to the pressure trend, so an anomaly can be judged without opening four other displays.

The weak spots are also well known. Small leaks sit inside the noise. Product changeovers and temperature changes move pressure without any leak. A transmitter that has drifted can look like a slow leak, which is why calibration records matter. For that reason, pressure monitoring is usually the first layer of leak detection and seldom the last.

Computational pipeline monitoring (CPM) is the name for software that watches pipeline data and helps the controller spot hydraulic anomalies. API RP 1130 covers it for liquid pipelines. API RP 1175 looks at the whole leak detection program, including control room procedures, training and the testing of equipment and alarms.

Monitoring is also a question of organisation. Decide who watches the screen, who is called when an alarm fires and how long a pressure alarm may stay open before it is escalated. Written procedures and regular drills turn a technical alarm into a response that actually happens.

No single method is best everywhere. Sites combine methods so that each covers the gap of another. The design question is not which one to buy, but what leak size, detection time and location accuracy the operator needs, and what the pipeline's instruments can support.

Real-time transient modelling goes a step further. A hydraulic model of the line runs next to the real pipe, and the system compares the model with the measurements. It can see smaller leaks than simple trend watching, but it needs a calibrated model, good instruments and time to tune.

Citia's PressMon communication restoration at Balongan is a field example of why the link matters. The work covered replacing fibre-optic converters, panel assembly, installation, restoring the connection and function tests, with checks of SCADA and web displays and alarm functions. A PressMon reference at TBBM Plumpang shows the web-based view with log history and alarms for abnormal conditions.

At Pertamina Priok, pressure and temperature transmitters and gas detectors feed the control system. When a value leaves its range, it raises an alarm and sends a trip command to the shutdown valve. That is monitoring connected to action.

The same data helps outside emergencies. A pressure history that shows how the line behaves at each pump change gives maintenance a baseline for pump wear, partly blocked strainers and valve problems. Alarm levels also need a review after commissioning: read the first months of alarm history, remove alarms that fire without any action and adjust levels, following the alarm management practice in ISA-18.2.