[X] Annual Electricity Demand Analysis: big differences, about 4.7 TWh.
[ ] Annual Electricity Demand Adressing Mismatch I: to improve how pypsa-earth estimates electricity demand, move away from the 60%-40% (industry, GDP) used in the literature to one taking into account electrification rates.
[ ] Annual Electricity Demand Adressing Mismatch II: use data provided by Kumbuso. This includes hourly demand data for Zambia for one year, and a snapshot of the state of the system to match regional distribution. Do we need more?
[ ] Comparing load profiles between the data sources we have and the demand from PyPSA is essential, the missmatch, in terms of total TWh, can be adjusted based on our own scenarios
[ ] Write section in validation report.
How is demand data obtained in pypsa-earth?
PyPSA-Earth uses a package (GEGIS I think) that estimates annual demand data for the years 2030, 2040, 2050, and 2100 at every node from the baseline years 2011, 2013 and 2018 (in particular, 2013 seems to be the core year). In a given region, it might use data from that region and other regions with similar economic/development characteristics.
Given these limitations, the proposed approach for demand validation is the following:
[ ] Past data: Try to get demand data for 2011, 2013, and 2018 that pypsa-earth uses to generate future demand data. 2011, 2013, and 2018 are the baseline years pypsa-earth uses. Ask Max for how to get it from pypsa-earth. Compare it to the annual generation @KumbusoBird provided for the years 2011 to 2021.
[ ] Future data: Look at the results for 2030, 2040, 2050, and 2100 pypsa-earth provides as demand estimations, and compare with external sources.
After all, demand analysis should be done with several different scenarios, to try to analyse the most likely future outcomes. Hence, if we feed in customized data or analyse several different scenarios generated by pypsa-earth, we should obtain "robust" results w.r.t. different future demand outcomes. Hence, we can understand this validation step as a first step towards the creation of several demand scenarios.
How is demand data obtained in pypsa-earth? PyPSA-Earth uses a package (GEGIS I think) that estimates annual demand data for the years 2030, 2040, 2050, and 2100 at every node from the baseline years 2011, 2013 and 2018 (in particular, 2013 seems to be the core year). In a given region, it might use data from that region and other regions with similar economic/development characteristics.
Given these limitations, the proposed approach for demand validation is the following:
[ ] Past data: Try to get demand data for 2011, 2013, and 2018 that pypsa-earth uses to generate future demand data. 2011, 2013, and 2018 are the baseline years pypsa-earth uses. Ask Max for how to get it from pypsa-earth. Compare it to the annual generation @KumbusoBird provided for the years 2011 to 2021.
[ ] Future data: Look at the results for 2030, 2040, 2050, and 2100 pypsa-earth provides as demand estimations, and compare with external sources.
After all, demand analysis should be done with several different scenarios, to try to analyse the most likely future outcomes. Hence, if we feed in customized data or analyse several different scenarios generated by pypsa-earth, we should obtain "robust" results w.r.t. different future demand outcomes. Hence, we can understand this validation step as a first step towards the creation of several demand scenarios.