|
Description
|
This dataset accompanies the paper "Replacing natural gas in Norwegian methanol production: A national-scale MILP optimization of a waste-to-methanol supply chain". It contains all input data, processing code, optimization results, and supporting documentation needed to reproduce the study in full. The study develops a deterministic mixed-integer linear programming (MILP) model to evaluate whether domestic Norwegian waste and biomass resources can support a national upstream supply chain for methanol production, as a long-term substitute for natural gas. The model covers 356 mainland Norwegian municipalities and five industrial hub nodes as potential conversion sites, 110 approved coastal ports as aggregation and export points, and six feedstock-specific thermochemical conversion modules: hydrothermal liquefaction of food waste and sewage sludge, fast pyrolysis of woody waste and forest harvest residues, and thermal cracking of household and industrial plastic waste. For each module, bio-oil yield, capital cost, operating cost, electricity demand, external heat demand, and emission intensity parameters are derived from the peer-reviewed techno-economic literature using triangular fuzzy numbers and defuzzified into single planning values before optimisation. The model determines which conversion plants should be activated, how much feedstock should be processed at each site, and how the resulting waste- and biogenic-derived oil (WBD-oil) should be routed to coastal ports, under four cost- and emission-constrained scenarios. Each scenario is solved using a lexicographic two-pass procedure: the first pass maximises annual WBD-oil delivered to port, and the second pass minimises total annual system cost while preserving the maximum output achieved in the first pass. Transport distances between supply nodes and ports are computed using the Open Source Routing Machine (OSRM) with real Norwegian road-network data, capturing the routing constraints imposed by fjord and mountain geography. The deposit is organised into four folders. The input data folder contains all raw feedstock statistics from Statistics Norway (SSB), municipality boundary and port registry geospatial files, and four parameter workbooks documenting the techno-economic, yield, emission, and minimum capacity assumptions with full literature traceability. The pipeline folder contains six sequentially numbered Python scripts covering node construction, port table assembly, distance matrix computation, cost and emission parameter generation, MILP optimisation, and results analysis, each accompanied by a detailed code manual. The results folder contains 31 publication-quality figures and 10 statistics tables generated by the analysis script, consolidated into a single Excel workbook. The data descriptor folder contains a variable dictionary covering every column in every deposit file, provided in both Word narrative and Excel lookup table formats. An interactive Leaflet.js map visualising the optimised supply chain across all four scenarios is included at the deposit root. (2026-03-19)
|
