What’s new

Next release

SSP 3.0 development

By topic:

Non-model changes

  • the --datafile option for the the material-ix build command was removed since it is not used anymore.

  • a material-ix calibrate command has been added that can be used to run the MACRO calibration step in isolation.

  • values from model parameter data that have a year_act and year_vtg columns greater than the technology lifetime are dropped since they are inflating the model size unnecessarily.

  • a module model.material.share_constraints for commodity share constraints was created, but is not fully used in the build yet.

Todo

consider differentiating the constraints by SSP.

Demand

  • the demand commodity level of methanol has been changed from final_material to demand.

  • an additional demand projection setting for aluminum, steel and cement has been introduced called “highest” which SSP5 scenarios are mapped to

  • the 2025 demands are fixed to the projected SSP2 values for each SSP.

  • Aluminum base year demand has been updated based on IAI MFA output for 2020

  • Cement base year demand has been updated based on GlobBulk Consulting data for 2020

  • Steel base year demand has been updated based on worldsteel association data for 2020

Post processing

  • the new industry reporter has been implemented model.material.report, with mapping files in data/material/reporting.

Model development

General

  • the power sector module has been deactivated for the SSP builds.

  • coal_i technology is share constrained in each region based on 2020 IEA statistics.

  • the low temperature share constraint for other industry is updated to reflect explicit modelling of heavy industry.

Todo

compile low temperature literature and set up references.

  • the .tools.cost module has been updated to run in “gdp” mode when called by the materials build.

  • the .tools.cost module is called twice if --update_costs option is True in material-ix build command to be able to get the correct cost projections for the non-MESSAGEix-Materials industry technologies (e.g coal_i, sp_el_I etc.)

  • the following utility functions were added/updated in utils.py. - for mapping country names with ISO 3166-1 alpha-3 codes with pycountry.

    Todo

    integrate in util.pycountry.

    • adding a R12 column with region values mapped based on a iso column of the same dataframe.

    • the MACRO calibration excel input file updater is now also updating the demand_ref tab based on the given scenario demand parameter values.

    • the MACRO calibration excel input file updater can be run in “extrapolation” mode or just take values from the scenario directly.

Aluminum

  • 2025 demand SSP2 fixing as mentioned under General changes.

  • the aluminum build reads SSP differentiated input data by reading from data/materials/aluminum/<SSPX>/aluminum_techno_economic.xlsx.

  • the historical capacity of smelters is calibrated until 2020 using the genisim dataset.

  • the historical activity of smelters is calibrated until 2020 using British Geological Survey data (Soderberg activity is calibrated assuming capacity shares computed from genisim dataset).

  • the historical activity of smelters is calibrated until 2020 using British Geological Survey data.

  • the historical activity of alumina refining is calibrated until 2020 using British Geological Survey data.

  • a trade model for alumina is introduced to be able to calibrate refining and smelting activity.

  • the process heat fuel consumption is calibrated using IAI data.

  • the regional alumina refining process heat intensities are calibrated using IAI data.

Ammonia

  • 2025 demand SSP2 fixing as mentioned under General changes.

  • a unit issue in the emission factor of biomass_NH3_ccs technology was fixed.

Cement

  • 2025 demand SSP2 fixing as mentioned under General changes.

  • the cement build reads SSP differentiated input data by reading from data/materials/cement/<SSPX>/Global_cement_MESSAGE.xlsx.

  • selected regional CEMENT values in residual_industry_2019.csv were updated since they seemed wrong when comparing the data to the IEA cement report.

Todo

TODO: paste link to IEA cement report here.

  • the heat input for clinker CCS addons was lowered a lot.

Todo

insert US NREL publication reference here.

Other industry

  • 2025 demand SSP2 fixing as mentioned under General changes.

  • the demands for the other industry are generated based on IEA historical data of the non-explicitly modelled sectors.

Note

the new demand model still sits in a private repository.

  • the sector furnaces were missing non CO2 emission factors of their original MESSAGE counterpart (e.g coal_i). Thus, they were copied from the originals and scaled with the input coefficients.

  • the furnaces are now writing into the IndThermDemLink relation, that is required for the MESSAGE-GAINS linkage.

Methanol

  • the model structure was slightly updated to simplify and correct the carbon emission balance accounting.

    • the negative emission coefficients that represent the carbon stored in long lived products, were moved from meth_t_d technology to a new technology meth_ind_fs.

    • the carbon balance of MTO_petro was not correct. The process emissions and the input/output parameters were updated based on new literature.

    Todo

    compile literature list and summarize in a paragraph.

Petrochemicals

  • together with the changes in Methanol the carbon accounting was improved:

    • the negative emission coefficients that represent the carbon stored in long lived products, were moved from steam_cracker_petro technology to the production_HVC technology.

    • since carbon capture in plastics was moved upstream, ethanol_to_ethylene_petro technology needed a positive emission factor to represent the combusted chemicals part produced with ethylene feedstock from ethanol.

Steel

  • the steel build reads SSP differentiated input data by reading from data/materials/steel/<SSPX>/Global_steel_MESSAGE.xlsx.

  • the maximum recycling relation was updated.

  • the minimum recycling relation was changed for SSP1.

  • the cokeoven_steel technology got its own bottom up CO2 emission accounting coefficient, to separate that from industrial emissions and move it to transformation emissions.

  • the cost changes done in the last PR (insert steel hydrogen PR link) were also transferred to the tools.cost module and custom reduction rates were assigned.