Consumption-band compatibility and tariff precision in industrial heat electrification screening

# Data download and analysis replication package Article: Consumption-band compatibility and tariff precision in industrial heat electrification screening. This package contains only data acquisition and quantitative analysis. It contains seven frozen public Eurostat snapshots, two narrow benchmark extracts downloaded on 9 September, and the public ITAC workbook archive downloaded on 9 September. The IEA source specification and provenance are supplied in data_external; report PDFs are cited, not redistributed. The public ITAC workbook's ASSESS and Terms sheets provide bill quantities, costs and field definitions. Source hashes are included. No credentials, restricted data or interactive software are required for quantitative reproduction. The national reporting-method screen is explicit in protocol/05_metadata_screen.json; source locators and decisions are in reports/NATIONAL_METADATA_AUDIT.csv. Data providers are not responsible for the transformations or conclusions. ## Reproduce with one command Use Python 3.11 or later. Install requirements.txt first. requirements-tested.txt records the exact numerical package versions used in the original run; logs/environment.json records Python and operating-system details. From the unzipped directory run: ``` python scripts/run_all.py ``` The command verifies cached hashes, rebuilds panels and all quantitative tables and figures. The normal run reads the included snapshots without contacting Eurostat. No GPU is used. To create a new data vintage, explicitly run `python scripts/01_download.py --refresh`. This is outside frozen-data reproduction and may change results. Review all downstream outputs after any refresh. `python scripts/08_download_benchmark.py` optionally re-acquires the narrow published-benchmark extracts; frozen reproduction uses the included, hashed snapshots instead. ## Result map | Result | Generated source under outputs/tables unless noted | Main script | | ------------------------------------------------------------ | ---------------------------------------------------------------------------------------------- | ------------------------------------------------ | | Headline band-gap results | metadata_sensitivity.csv and summary.csv | 02_analyse.py and 05_metadata_sensitivity.py | | Fiscal contributions | metadata_sensitivity.csv | 05_metadata_sensitivity.py | | Weighted dispersion | metadata_sensitivity.csv and dispersion_summary.csv | 04_composition.py and 05_metadata_sensitivity.py | | Institutional comparison | reports/INSTITUTIONAL_VALIDATION.md and NATIONAL_METADATA_AUDIT.csv | Documented source review | | Heat allowances | heat_summary.csv | 06_heat_model.py | | Band-choice consequences | band_choice_summary.csv | 07_band_choice.py | | Published benchmark portability | published_benchmark_summary.csv | 09_benchmark_robustness.py | | Tariff robustness | tariff_robustness_summary.csv | 09_benchmark_robustness.py | | Paired plant bill diagnostic | itac_benchmark_summary.csv | 10_itac_extract.py and 11_itac_analysis.py | | Screened decomposition figure | outputs/figures/05_screened_decomposition.\\* | 05_metadata_sensitivity.py | | Heat allowance figure | outputs/figures/06_heat_allowance.\\*; heat_curve.csv | 06_heat_model.py | | Band choice figure | outputs/figures/07_band_choice.\\*; band_choice_country.csv | 07_band_choice.py | | Coverage, country results, historical and diagnostic files | see individual CSV files under outputs/tables | 02_analyse.py and 03_diagnostics.py | | Annual component support and consumption-weighted dispersion | annual_component_summary.csv and dispersion_summary.csv | 02_analyse.py and 04_composition.py | | Threshold and influence results | band_choice_thresholds.csv, band_choice_hurdle_intervals.csv and band_choice_leave_one_out.csv | 07_band_choice.py | | Plant-bill sample flow and charge components | itac_sample_flow.csv, itac_benchmark_yearly.csv and itac_charge_components.csv | 10_itac_extract.py and 11_itac_analysis.py | All country results are machine-readable rather than only embedded in plots. Figures correspond to source stems 01_price_gaps, 02_exact_decomposition, 03_country_heterogeneity and 04_volume_weighted_dispersion for the original panel, and 05–07 for the screened, heat and band-choice extensions. ## Interpretation notes The metadata, heat and robustness extensions were selected after the original results; their dates and scope are in protocol/. Scenario counts are not independent observations. Country sampling intervals in the price analysis do not account for every common shock. The IB/I2 comparator is illustrative. The additional IE/I3 pair is documented against an IEA comparison, but transporting it to this heat service does not reproduce or invalidate that report's project model. Tariff-deviation bounds are stipulated, not measured distributions. The same energy service, efficiencies, country, tax definition and time weights are retained in each comparison. The ITAC exercise uses 2,073 complete manufacturing assessment records and 2,008 on national common benchmark support. These are selected US records, not a representative plant population or European tariff data. Actual historical bill unit expenditures are not marginal contract offers. US dispersion frequencies do not calibrate the European price bounds. The plant extension is specified in protocol/10_plant_bill_validation.json. Original source queries, retrieval times and source hashes are in data_raw/provenance.json and data_external/provenance.json. Source attribution and reuse information are in reports/DATA_SOURCES.md.