Data and code for traceable reuse of historical HAZOP–LOPA records in a tert-butyl acetate process
Description
This dataset supports the study of traceable reuse of historical HAZOP–LOPA records for equipment change review in a tert-butyl acetate process. It contains de-identified transcriptions of 75 HAZOP and 28 LOPA records from ten historical tables covering five process areas. Original Chinese source fields and record identifiers are retained, together with English analysis views, calculation inputs and table exports. The Python programs reproduce arithmetic products, risk-code comparisons, factor-link tests, network perturbations, registered equipment-change queries, and a search of safety-valve labels across all 28 LOPA records. They also check the information requirements for six review tasks and reproduce eight calculation states from three published studies. A single command generates CSV results, a JSON report and an HTML summary. Blank input templates and a separate adapter are provided for users to map records from their own sites. The worked adapter example uses the existing TBA records. Users supply the equipment identities, applicable historical scales and supporting engineering evidence for their sites. The participating company has authorized public dissemination of these de-identified data. Data and explanatory documentation are licensed under CC BY 4.0; software is licensed under MIT. FILE_INDEX.csv describes all 103 files in the archive and their applicable licenses.
Files
Steps to reproduce
1. Download Source_Data.zip and extract all files, preserving the directory structure. 2. Use Python 3.9 or later. No additional Python packages are required. 3. Open a terminal in the extracted folder and run: python run_all.py On Windows, you can instead double-click Run_Reproduction.cmd. 4. Open RESULTS.html in the newly generated results directory. CSV output tables and a detailed JSON report are stored in the same directory. The program checks file integrity, repeats the calculations and searches, evaluates the declared information requirements, and runs the software tests. 5. Consult OUTPUT_MAP.md to locate the outputs corresponding to the paper. DATA_GUIDE.md explains the data fields and sources. SITE_REUSE_GUIDE.md describes the separate adapter and blank templates for applying the method to explicitly mapped records from another site. The release was tested on Windows with Python 3.12.14 and 3.14.4. The programs use the Python standard library and do not require network access.
Institutions
- Guangdong University of Petrochemical TechnologyGuangdong, Maoming