AI-assisted HAZOP

We develop AI methods that support Hazard and Operability (HAZOP) studies. Machine-readable P&IDs and large language models help HAZOP teams prepare studies faster and check them more systematically, while the engineers stay responsible for the results.

Challenge

HAZOP studies are a cornerstone of process safety. A multidisciplinary team systematically walks through the P&IDs of a plant, node by node, and discusses how deviations from the design intent can occur, what their consequences are, and which safeguards are in place. HAZOP studies are time-consuming and depend on the experience of the team. Most of the information they need (P&IDs, design documents, previous HAZOP reports) is available only as documents that computers cannot interpret, which makes it hard to reuse knowledge and to check results for consistency.

Our approach

We combine three building blocks:

  • Machine-readable plant information. Legacy P&IDs are digitized and converted into Smart P&IDs based on the DEXPI standard, which can be represented as knowledge graphs (see digitization of engineering diagrams and pyDEXPI).
  • LLM interaction with P&IDs. With ChatP&ID, large language model agents query the P&ID knowledge graph through retrieval tools, which grounds their answers in the actual plant topology.
  • Generative AI for HAZOP worksheets. Generative models propose nodes, deviations, causes, consequences, and safeguards based on the engineering documents, previous HAZOP reports, and simulations. The HAZOP team reviews, edits, and decides.

The aim is to make HAZOP studies more efficient and more consistent, and to support engineers rather than replace them. AI support for HAZOP is also one of the planned features of the Digitization Companion (DigiCo).

Partners

We work on AI-assisted HAZOP together with the European Process Safety Centre (EPSC), an industry association for process safety, and its Operations Director, Dr. Tijs Koerts.

Team

Key publications

  • Alimin, A. A., & Schweidtmann, A. M. (2026). GraphRAG for engineering diagrams: ChatP&ID enables LLM interaction with P&IDs. AIChE Journal, 72, e70540. doi:10.1002/aic.70540
  • Schweidtmann, A. M. (2024). Generative artificial intelligence in chemical engineering. Nature Chemical Engineering, 1, 193. doi:10.1038/s44286-024-00041-5 · Details
  • Goldstein, D. P., Schulze Balhorn, L., Alimin, A. A., & Schweidtmann, A. M. (2025). pyDEXPI: A Python framework for piping and instrumentation diagrams (P&IDs) using the DEXPI information model. Systems and Control Transactions (Proceedings of ESCAPE 35). doi:10.69997/sct.139043
← All projects