State-Level Green Hydrogen Modeling in Mexico: A Spatially Explicit Techno-Economic Framework for Alkaline, PEM, and SOEC Electrolyzers

Description

This repository contains the processed state-level input tables, model outputs, and Python scripts used in Scenario-based mapping of state-level green hydrogen potential in Mexico: Policy-relevant insights from a multi-technology framework (Energy and Climate Change, Vol. 7, Dec 2026, 100240; DOI: 10.1016/j.egycc.2026.100240). Research hypothesis. Mexico’s green-hydrogen potential and competitiveness are not determined by solar/wind resource alone; instead, feasible production and cost outcomes emerge from the joint interaction of renewable capacity factors, suitable land, water constraints, and grid accessibility, and these interactions differ by electrolysis technology (AWE, PEM, SOEC) and over time (2025–2060). What the data shows / notable findings. The dataset provides a screening-scale, comparable state-by-state ranking of (i) feasible H₂ production potential (after constraints) and (ii) plant-gate LCOH trajectories. Results show strong spatial heterogeneity: high-resource states can still be limited by grid distance or water caps, while some states achieve lower costs due to more favorable electricity price assumptions and capacity factors. Deterministic sensitivity indicates that electricity price (or LCOE) and electrolyzer efficiency typically dominate LCOH outcomes, with CAPEX and fixed O&M having smaller but still meaningful effects. What the data is and how it was gathered. Public geospatial layers were harmonized to Mexico’s 32 state polygons and aggregated using zonal statistics / class aggregation to create input CSVs: solar (GHI), wind (100 m; onshore baseline with offshore diagnostics for coastal states), land suitability (reclassified land-cover classes summed to suitable hectares), water context/caps (state-level volumes; 9 L/kg H₂), and grid accessibility (distance-to-grid metrics plus proximity shares). These inputs drive technology scripts that compute baseline H₂ outputs, apply grid and water derating, project to 2060 using technology efficiency pathways, and pass projections to an LCOH module that computes plant-gate costs and a ±10% univariate sensitivity panel. How to interpret/use. Outputs are screening-level (strategic planning), not site-specific project feasibility. They do not model dispatch, congestion, permitting, or full downstream transport/storage/conversion unless optional adders are provided. Use the data to identify priority corridors, compare technologies consistently, and test policy levers (power price, interconnection readiness, efficiency/utilization, and water sourcing). Steps to reproduce. Install Python 3.10+ and numpy, pandas, geopandas, shapely, pyproj, rasterio, matplotlib (optional: rasterstats, fiona, rtree). Update BASE_DIR/DATA_DIR/OUT_DIR paths. Optionally regenerate state input CSVs from rasters/shapefiles, then run: (1) AWE/PEM/SOEC baselines → (2) projections to 2060 → (3) LCOH → (4) sensitivity and plotting scripts.

Steps to reproduce

Install tools Install Python 3.10+ Create an environment and install: numpy pandas geopandas shapely pyproj rasterio matplotlib Optional (only if your preprocessing scripts need them): rasterstats fiona rtree Unzip and keep folder structure Keep input folders together (solar rasters, wind, land, water, electricity, and the model folders). Set paths in scripts In each script, edit BASE_DIR / DATA_DIR / OUT_DIR to your local folders. If running preprocessing/maps, also point to mexican-states.shp. (Optional) Rebuild state-level input CSVs from rasters Run the preprocessing scripts to generate/update: MEXICOSOLAR.csv MEXICOWIND.csv (onshore + offshore diagnostic for coastal states) MEXICOLAND.csv MEXICOWATER.csv and MEXICO_water_availability.csv MEXICOGRIDELECTRICITY.csv (If these CSVs are already provided and you don’t need to regenerate them, skip this step.) Run technology baseline models Run ALKALINEMEXMODEL.py → ALKALINE_H2_OUTPUT_REALISTIC.csv Run PEMMODELMEX.py → PEM_H2_OUTPUT_REALISTIC.csv Run SOECMEXMODEL.py → SOEC_H2_OUTPUT_REALISTIC.csv Run projections to 2060 Run ALKALINEMEXPROJECT2060.py → ALKALINE_H2_PROJECTION_CLEAN.csv, ALKALINE_H2_NATIONAL.csv Run PEMMEXPROJ2060.py → PEM_H2_PROJECTION_CLEAN.csv, PEM_H2_NATIONAL.csv Run SOECMEXPROJECT2060.py → SOEC_H2_PROJECTION_CLEAN.csv, SOEC_H2_NATIONAL.csv Run LCOH model Run LCOHMEX.py (reads the three *_H2_PROJECTION_CLEAN.csv) → LCOH_PER_STATE_BY_TECH.csv Optional: run DISTANCEBYSTATE.py first if you’re enabling delivered-cost adders. Run sensitivity + figures Run MEXICOSENSIVILITY.py → SENSITIVITY_RESULTS_ALL_YEARS.csv

Institutions

• Chongqing University

  Chongqing, Chongqing

Categories

Related Links

Licence