A simulation-based dataset for anomaly detection in hydrogen blend transport networks.

Area of Science:

• Energy Systems Engineering
• Industrial Process Monitoring
• Data Science for Infrastructure

Background:

• Hydrogen transport via pipelines is crucial for sustainable energy but requires robust safety monitoring.
• Existing multivariate datasets for hydrogen transport networks are scarce, impeding data-driven monitoring method development.
• Safety concerns include flammability risks from leaks, compressor failures, and component response delays.

Purpose of the Study:

• To address the limited availability of multivariate data for hydrogen transport networks.
• To provide a synthetic dataset for developing and evaluating data-driven monitoring and anomaly detection algorithms.
• To support the advancement of digital twins for hydrogen transport infrastructures.

Main Methods:

• Simulation of a representative industrial hydrogen pipeline segment using MATLAB Simscape.
• Generation of time-series data from distributed virtual sensors.
• Inclusion of both normal operating conditions and various anomalous scenarios (leaks, compressor failures, delayed responses).

Main Results:

• A comprehensive synthetic dataset capturing transient and steady-state dynamics of hydrogen transport.
• Data covers normal operations and critical failure scenarios.
• The dataset is suitable for testing algorithms for monitoring, anomaly detection, and digital twins.

Conclusions:

• The presented synthetic dataset effectively fills a critical data gap in hydrogen transport research.
• This resource facilitates the development and validation of advanced monitoring techniques for industrial hydrogen pipelines.
• Enables enhanced safety, reliability, and efficiency in hydrogen energy infrastructure.