Home // IARIA Congress 2025, The 2025 IARIA Annual Congress on Frontiers in Science, Technology, Services, and Applications // View article

Early Response Prediction for H2 Sensors

Authors:
Raduan Sarif
Carlo Tiebe
Christian Herglotz

Keywords: H2 Safety; H2 Leakage Detection; H2 Sensor Data Analysis; H2 Sensor Response Predictions; First-Order (FO) Model.

Abstract:
Green hydrogen (H2) is essential for the global transition to clean energy; it will significantly reduce emissions from heavy industry and the long-distance transport system. H2 can be used as fuel in fuel cells, storing surplus renewable energy, and as a feedstock in industrial processes. However, H2 faces significant safety challenges during storage and transportation. Accidents due to H2 leakage and explosions raise serious concerns due to its high flammability, rapid diffusion in air, and extremely low ignition energy. To mitigate risks associated with H2 leakages, reliable and automated H2 safety systems are essential for emergency repairs or shutdown. An early response from H2 sensors is crucial for early warning in accidents. The earlier response time of H2 sensors is often constrained by their sensor principle, which is heavily influenced by the sensor material’s properties. This study explores methods for earlier sensor response through predictive algorithms. Specifically, we investigate transient response predictions using a First-Order (FO) model and propose improvements through the First-Order with early response and the First-Order with adapted early response model. Both models can predict the stable value of the H2 sensor response from a small time window, which is 70.89% and 83.72% earlier, respectively, than the time required for the sensor hardware to reach it physically. The model’s performance is evaluated by calculating the fitting error with a 2 % threshold. Our current research lays the groundwork for future advancements in real-time sensor response predictions for hydrogen leakage.

Pages: 131 to 137

Copyright: Copyright (c) IARIA, 2025

Publication date: July 6, 2025

Published in: conference

ISBN: 978-1-68558-284-5

Location: Venice, Italy

Dates: from July 6, 2025 to July 10, 2025