Skip to main content
Dryad

Technology- and facility-level energy, cost, and environmental performance in U.S. chemicals, cement, iron and steel, food, and non-manufacturing industries

Data files

Nov 04, 2025 version files 6.27 MB

Click names to download individual files

Abstract

This U.S. industrial facilities and technology dataset is a technology- and facility-level collection of technological, cost, energy, and emissions attributes for six manufacturing and three non-manufacturing U.S. industries. The dataset is a JSON array organized by industry. Each industry entry (except for mining and agriculture) contains four sections: author list, assumptions, emerging technologies, and existing facilities. The non-manufacturing industry inventories 18 solutions across agriculture, mining and construction sectors and 6 categories, documenting qualitative benefits and quantitative energy/emissions reduction potentials with low/average/high estimates. The dataset integrates data exclusively from publicly available data sources including EPA's Greenhouse Gas Reporting Program, U.S. Geological Survey, industry reports, peer-reviewed research to provide a unified resource for energy systems modeling and analysis.

The assumptions section standardizes units and conversions and provides fuel and feedstock prices for operating expenditure (OPEX) calculations, Chemical Engineering Plant Cost Index (CEPCI) values for capital cost (CAPEX) harmonization, and price/inflation indices to align values to a common base year. Emerging technologies with minimal to no market share in U.S. commercial-scale production facilities are characterized across six industries: ammonia (10 processes including autothermal reforming, renewable hydrogen-based, biomass gasification, methane pyrolysis, and ATR with CCS); cement (26 processes including conventional wet/dry kiln variants, low-SCM dry kiln with preheater+precalciner, full CCS, and full electrification options); ethanol (8 processes including dry mill and wet mill variants, electrified process heat via heat pumps, and dry mill BAT with CCS); ethylene and propylene (21 processes including electrified steam cracking with electricity-cost bands, ethanol-to-ethylene, MTO, and NGL-to-olefins); iron and steel (10 processes including H2DRI–EAF, NGDRI–EAF with CCS, and molten oxide electrolysis, with varied scrap utilization scenarios); and food (9 cross-cutting process-heat decarbonization options such as hot-water and steam heat pumps, electric boilers, RNG/biogas boilers, and solar thermal steam).

The existing facilities section covers: ammonia (36 facilities across 20 states; SMR 97.2%, coal gasification 2.8%); cement (97 facilities across 35 states; conventional dry kiln 90.7%, wet kiln 9.3%); ethanol (201 facilities; dry mill 95.5%, wet mill 4.5%); ethylene/propylene (35 facilities across 6 states; steam cracking 100%); iron and steel (102 facilities across 31 states; EAF 86.4%, BF–BOF 7.8%, DRI 2.9%, hybrid BF–BOF/EAF 0.9%); and food (production and energy intensity by state and five subsectors: animal slaughtering, dairy, fruit and vegetable, grain and oilseed milling, and sugar). 

The dataset offers a broad range of use cases through its standardized JSON structure and comprehensive documentation, potentially offering interoperability with common analytical tools. Primary uses envisioned for this dataset include energy systems optimization modeling, multisectoral and integrated assessment modeling of the industrial sector or the broader economy (but with higher fidelity of technology characterization), technology assessment comparing conventional and emerging production routes, spatially resolved production capacity planning analysis, and economic analysis of technology deployment costs. The dataset's facility-level granularity enables bottom-up modeling approaches while maintaining compatibility with top-down sectoral analyses. Technical features enhancing reusability include standardized coordinate systems (WGS84) for GIS integration, consistent economic units (2018 USD) for temporal comparisons, and modular data structure supporting selective extraction.