Separating reasoning from serialization eases engineering design tasks
Separating Engineering Reasoning from DEXPI Serialization in LLM-Based Greenfield Surface-Process Design: A Three-Case Study for Underground Gas Storage
Software EngineeringArtificial Intelligence
Summary
Designing underground gas storage involves creating detailed engineering descriptions and computer-readable files that follow strict standards. This paper shows that splitting the design process into two parts—first making a simple engineering outline, then converting it into the full standard format—can reduce errors related to the format itself. The authors found that while this separation makes it easier to spot where problems happen, it does not completely fix mistakes made in the engineering design. They tested this idea on three example cases involving different gas storage scenarios.
What this means in practice
- •For process engineers: Simplify creating and validating engineering designs for underground gas storage by separating design reasoning from strict data serialization.
- •For software developers in process industry: Build design tools that isolate serialization errors by generating simpler intermediate representations before full standard-compliant outputs.
Authors
Qingchuan Zhu, Shuyue Tong, Pengju Ren
Abstract
Large language models can produce engineering descriptions and structured process representations, but standards-level serialization can substantially increase the generation burden. This diagnostic study examines whether separating engineering reasoning from Data Exchange in the Process Industry (DEXPI) serialization changes where representation and engineering failures occur in constrained greenfield surface-process design for underground gas storage. We compare Direct DEXPI generation with generation of a lightweight Engineering Intermediate Representation (IR) on three cases: single-pressure injection, withdrawal and export, and dual-pressure injection. All six conditions use one fixed model snapshot, qwen3.8-max-0902, with one completed hosted generation per condition. Direct outputs are XSD-valid in 2 of 3 cases, while all 3 Engineering IR outputs are structurally valid under a minimal validator. Direct prompt inputs contain approximately 121.8k-121.9k tokens, compared with 617-699 tokens for the Engineering IR prompts. Engineering feasibility does not uniformly favor the IR: one IR output is rejected for an explicit cooling-state contradiction. The cases also reveal two distinct Direct DEXPI failure modes: engineering inconsistency and standards-level serialization failure. The observed comparison shows that, in these evaluated method bundles, deferring DEXPI serialization substantially reduces representation burden and helps isolate serialization failure, but reducing representation burden alone does not eliminate engineering inconsistencies.