NEPM (Neural Engine Pipeline Monitoring)

Pipeline leak detection

The transportation of fluids via pipelines is increasing worldwide – with good reason. Over long distances, pipelines are among the safest, most environmentally efficient and economical transport systems. Nevertheless, leaks represent a potential safety, environmental and economic risk.

Leaks can occur for many reasons, including corrosion, material fatigue, construction damage, operational errors, ground movement, earthquakes or illegal product tapping. To minimise these risks, pipeline operators rely on advanced leak detection systems that provide fast, reliable and actionable information.

Selecting a suitable system

Selecting an appropriate leak detection system is a complex task. The system must be adapted to the specific pipeline, product and operating conditions, while also complying with international standards and regulations such as API RP 1130, API 1175, TRFL or CSA Z662 Annex E.

Modern pipelines increasingly require solutions that are robust against changing operating conditions, product changes and transient effects – and that can turn large volumes of sensor data into reliable decisions.

From RTTM to NEPM

Evolution of internal pipeline monitoring

The RTTM (Real‑Time Transient Model) has long been recognised as a leading technology for internal pipeline leak detection. RTTM mathematically models pipeline behaviour and compares real-time measurement data from the physical pipeline with a simulated or “virtual” pipeline.

With E‑RTTM (Extended RTTM), KROHNE expanded this principle by adding leak signature analysis and pattern-based evaluation, significantly improving sensitivity and reliability.

NEPM represents the next evolutionary step. It builds on the proven foundations of RTTM and E‑RTTM and extends them into a neural, multi‑method monitoring framework.

Multi‑method leak detection with NEPM

An NEPM-based system creates a dynamic digital representation of the pipeline using:

• real-time measurements from flow, pressure and temperature sensors

• advanced hydraulic modelling of the pipeline

• pattern recognition and signature databases

• AI-driven optimisation and adaptive learning algorithms

Measurement data is taken at the inlet and outlet of the pipeline and at intermediate locations such as pump stations, valve stations or block valves. Based on this data, NEPM continuously calculates flow, pressure, temperature and density profiles along the entire pipeline.

These calculated values are permanently compared with measured values from flowmeters and pressure transmitters. When a deviation is detected, the neural engine analyses the event using multiple methods in parallel. It determines whether the deviation is caused by normal operational effects, instrumentation issues, transient conditions – or by a real leak, whether gradual or sudden.

Handling changing and transient operating conditions

A key strength of NEPM is its ability to reliably monitor pipelines under dynamic and challenging operating conditions. NEPM works with transient states and continuously adapts to changing conditions such as:

• start-up and shutdown sequences

• batch or product changes

• pump or valve operations

• sensor or communication failures

The combination of real-time hydraulic modelling, pattern recognition and AI-based optimisation provides exceptional robustness. By continuously learning from operational data, NEPM improves detection performance over time and offers a high degree of protection against false alarms.

NEPM in PipePatrol NEO

NEPM is the core monitoring technology of PipePatrol NEO, KROHNE’s latest generation pipeline monitoring platform. It is suitable for liquid and gas pipelines, including liquefied gases and supercritical products.

PipePatrol NEO with NEPM supports compliance with API RP 1130, API 1175, TRFL and CSA Z662 Annex E, while providing operators with enhanced transparency, faster response times and greater confidence in pipeline integrity.