Data and code for a new sizing framework for photovoltaic, wind, battery, and hydrogen green buildings across three climate zones
Description
This dataset supports the manuscript entitled “A new sizing framework for a photovoltaic/wind/battery/hydrogen green building system incorporating Wasserstein robustness, future climate, and hourly validation: Case studies of Shenzhen, Golden, and Edinburgh”. The dataset contains processed climate, building-load, cost, technology, and siting inputs; 63 complete historical and mid-century future-climate hourly scenarios for Shenzhen, Golden, and Edinburgh; two-stage finite-support Wasserstein distributionally robust capacity-optimization code; hourly redispatch and feasibility-restoration code; and the resulting capacity, economic, reliability, and sensitivity outputs for a standardized 20,438.66 m2 office/campus building. The archive contains source data for Figs. 1–36 and Tables 1–30, supplementary figures, final hourly evaluation records, strict onsite feasibility results, future-climate audits, re-optimized sensitivity cases, solver traces, data-acquisition scripts, public-source provenance records, checksums, and claim-level verification files. The uncertainty support combines ERA5 historical years from 2011 to 2025 with CMIP6 pseudo-years from ACCESS-CM2, MRI-ESM2-0, and UKESM1-0-LL under SSP2-4.5 and SSP5-8.5. Third-party inputs remain subject to their original reuse terms. Where redistribution is not permitted, the archive provides source URLs, query parameters, retrieval scripts, access metadata, and provenance records. The dataset supports inspection and reproduction of the reported cross-climate system-sizing and hourly-validation results.
Files
Steps to reproduce
1. Download and extract Renewable_Energy_Mendeley_Data_Upload_v1.1_FINAL.zip. 2. Read README_MENDELEY_DATA.md for the package structure, dataset provenance, scenario coverage, and integrity information. 3. Change the working directory to analysis_2026/submission_2026. 4. Run the analysis scripts in the following order: python build_submission_inputs.py python run_wdro_extensive.py --planning-timestep-hours 12 python repair_hourly_designs.py python evaluate_strict_onsite.py python evaluate_historical_design_future.py python build_manuscript_extensions.py python build_publication_tables.py python make_publication_figures.py 5. The processed scenarios are located in data/hourly_scenarios. Capacity, economic, reliability, sensitivity, and solver results are provided under results. Final manuscript figures are provided under figures/article_main, and manuscript table source data are provided under results/template_tables. 6. The Wasserstein distributionally robust optimization stages are computationally intensive. Archived result tables, hourly evaluation records, numerical residual checks, and solver traces are supplied so that the reported results can be inspected without rerunning every optimization. 7. File integrity can be checked using CHECKSUMS_SHA256.txt. FILE_MANIFEST.csv lists all archived files, and PACKAGE_VALIDATION.json records the automated package validation results.