Quantifying Pipeline Leak-Detection Capability Under Uncertainty

Pipeline leak detection should not be described by a single sensitivity figure or a generic claim of “real-time detection.” Practical capability depends on the physical leak signature, measurement uncertainty, inventory estimation, operating state, sensor coverage, detection method, alarm behaviour and the evidence used to demonstrate performance.

This white paper presents a series-specific Leak Detection Capability Map for describing and evaluating pipeline leak-detection capability under defined operating and measurement conditions.

The framework organizes six connected review areas:

physical leak signature — including product, pressure, release mode, location, environment and consequence;

instrumentation evidence — including flow, pressure, temperature, density, timing, calibration and measurement uncertainty;

detection method — including computational pipeline monitoring, statistical, acoustic, fibre-optic, vapour, liquid and other sensing approaches;

operating-mode discrimination — including steady-state, transient, startup, shutdown, line packing and other abnormal or changing conditions;

alarm and confirmation behaviour — including alarm logic, persistence, false-positive behaviour, controller interpretation and confirmation workflow; and

performance verification — including commissioning tests, controlled releases, simulation, historical-event analysis, functional testing, drills and continuing KPI review.

The analysis draws on publicly available sources including API RP 1175, Pipeline Leak Detection — Program Management, 2nd edition; PRCI research on computational pipeline monitoring, leak-detection thresholds and retrofit sensing; PRCI material on measurement uncertainty; and PHMSA control-room-management guidance.

A central proposition is that a leak-detection claim is defensible only when the claimed capability is tied to the relevant leak signature, operating mode, instrumentation uncertainty, detection logic, alarm behaviour, response time and performance evidence under representative conditions.

The paper emphasizes that leak detection is a programme rather than a product specification. Algorithm sensitivity alone does not determine operational effectiveness. Instrument reliability, maintenance, procedures, operator response, alarm rationalization, system testing and continuous improvement all contribute to actual detection capability.

Measurement and inventory uncertainty are treated as fundamental limits on observability. Perfect balance between incoming and outgoing mass, volume or energy is not achievable in real systems. Meter bias, sensor accuracy, density estimation, timing, temperature, pressure and linepack uncertainty can therefore establish a practical floor below which small leaks become difficult to distinguish from normal measurement and operational noise.

Operating condition also matters. A detector that performs well during steady-state operation may behave differently during pump starts, valve movements, batch transitions, line packing, slack-line conditions or planned maintenance. Detection capability should therefore be characterized by operating mode rather than assumed to be constant.

The paper also addresses false alarms and operator workload. High nominal sensitivity can be undermined by nuisance alarms, weak confirmation workflows or declining operator trust. Performance assessment should consequently include false-positive behaviour, alarm persistence, confirmation requirements and control-room workload, not only detection threshold.

Complementary detection technologies are examined from the perspective of diversity of detection physics and blind spots. Combining technologies can improve resilience when their strengths and limitations are genuinely different; using multiple technologies with the same failure modes provides less benefit.

Finally, the paper emphasizes the need to test, measure and continuously improve leak-detection performance. Controlled releases, simulations, hydraulic transient replay, historical-event analysis, functional testing and operator-response drills can all contribute evidence, depending on the system and operating context. Material changes to instrumentation, operating conditions or detection logic should trigger review, retesting or retuning.

The Leak Detection Capability Map is a practitioner synthesis, not a new leak-detection technology or standard. It does not replace API, PRCI, PHMSA, regulatory, operator or project-specific requirements, and it does not establish universal minimum detection thresholds, response times or numerical acceptance criteria.

The scope is deliberately limited to publicly discussable sensing, monitoring, analytical assurance, alarm interpretation and human engineering decision support. Shutdown, isolation and other consequential physical-control actions remain within separately engineered operating and safety systems.

This paper forms No. 26 of the From Project to Live Asset — Independent Practitioner Research Series, Phase III, extending the wider steel-pipe lifecycle research programme from corrosion reassessment into evidence-bounded pipeline leak-detection capability.

Author-written abstract, reproduced from the authoritative Zenodo DOI record.