AI_H2

Short description – Converting gas networks to hydrogen–natural gas blends raises the requirements for leak monitoring. Hydrogen escapes more readily than methane, even through small leaks, which makes it well suited as an indicator gas for early detection. In an interdisciplinary junior research group at WHZ, ten early-career researchers from four faculties are developing a systemic approach to energy-autonomous, AI-supported leak detection: from sensor technology to decision support for network operators.

Project Duration:
Nov 1, 2026 to Oct 31, 2028

Project Management:
Prof. Dr. Robert Täschner (WHZ)

Participating WHZ professorships:
Prof. Dr. Robert Täschner, Prof. Dr. Peter Hartmann, Prof. Dr. Ulrich Walther, Prof. Dr. Mirko Bodach, Prof. Dr. Martin Dannemann, Prof. Dr. Christoph Laroque, Prof. Dr. Matthias Richter, Prof. Dr. Daniel Kretz

The EU Methane Regulation 2024/1787 obliges network operators to monitor and repair leaks. At the same time, blending hydrogen into existing natural gas networks increases the risk of leaks and the associated safety risk. The project therefore aims to deliver a concrete systemic approach to leak detection in hydrogen/natural gas networks, built on three pillars:

  1. Hydrogen detection: Development of two complementary sensor concepts: a chip-based MOEMS sensor with palladium-coated photodiodes, and a fibre-optic sensor for distributed, spatially resolved measurement.
  2. System design: Investigation of leakage behaviour through simulation (FEM, CFD) and on a scaled pipeline test rig, together with the development of an energy-efficient electronics platform for autonomous sensor nodes.
  3. Data analysis and system integration: Statistical and AI-based methods for real-time detection of critical states, prediction of maintenance needs, and decision support at the level of the overall network.

Proof of function will be provided using demonstrators and the WHZ pipeline test rig.

 

The Industry Analytics research group and Professor Laroque contribute to work package 3, "Data Analysis and System Integration", with a focus on developing a decision support system (DSS) for the overall hydrogen/natural gas network. The DSS brings together the results of state detection, ageing models and AI-based predictive maintenance in a unified environment. This environment visualises the state of the entire network and uses analytical interaction patterns to support the assessment of risk situations down to individual measuring points. Information from existing operational information systems is also integrated to improve the basis for preventive maintenance planning and rapid intervention. Finally, the system is implemented as a prototype and validated using real and simulated data.

 

Prof. Dr. Christoph Laroque
Professor of Business Analytics 
christoph.laroque[at]whz.de

Adrian Rössl
Research Assistant 
adrian.roessl.l41[at]whz.de