GR4SP Suite: Additional Data and Simulation Results

This entry is for the GR4SP Suite, a model and simulation tool used to analyse the Victorian electricity system's history and potential future transition pathways. Most of these files were used or generated with the Jupyter Notebook scripts shared in the author's GitHub repository (https://github.com/gr4sp/simulationEngine).

 This version adds the validation material for the simulation-mode assessment of the model, together with the business-as-usual run it is assessed against. It does not repeat files from earlier versions; see "Relation to earlier versions" below.

Files in this version

- gr4sp_SOBOLhypopast2021-Mar-03_includ_wholesale_month.tar.gz (972 MB) — the validation ensemble. 105,000 GR4SP runs over 24 uncertainties, generated with an EMA Workbench Saltelli sample of N = 2100. It is a hypothetical-past design: it varies only parameters that could plausibly have differed over the historical period and omits the forward-looking parameters of the sensitivity-analysis ensemble, making the two designs distinct and not interchangeable. Unlike the earlier sampling archives, it also carries the monthly wholesale price outcome, which is what allows the wholesale series to be validated at monthly resolution. The archive holds plain CSV outcome matrices, one per outcome, so it can be read without the EMA Workbench. Two rows, 41357 and 41379, are empty and are dropped in analysis.

- VICSimDataYearSummary_bau19982051.csv (0.6 MB): the pinned business-as-usual run, annual, 1998 to 2050, 922 columns. This is the single deterministic run the validation statistics compare against the historical record for wholesale price, renewable share, retail tariffs and greenhouse gas emissions.

- VICSimDataYear_seed_1621753227017-BAU-1902-2051.csv (1.0 MB): a business-as-usual run reported annually from 1902 to 2050, 1015 columns, named by its random seed. It extends back through the model's historical mapping period, before the 1998 start of the summary file, and is the baseline for several of the analysis notebooks.

- eet_hypoPast.csv (1.4 kB): the Elementary Effects Test (Morris) screening carried out over the historical period. It lists 34 uncertain model inputs, and for each of four indicators (greenhouse gas emissions, primary spot wholesale price, tariffs, renewable energy) gives the input's rank by mu* (columns ending -M) and by sigma (columns ending -S); a blank means the input did not rank for that indicator. It is included as context for the choice of parameters varied in the validation ensemble, which it informed but did not determine: 22 of the 34 inputs rank on at least one indicator, 20 of those are among the ensemble's 24, and the ensemble both adds four inputs that rank nowhere here and omits two that do.

Relation to earlier versions

 Zenodo versions do not inherit each other's files, and this deposit's files are spread across two of them. Version 2.0, record 8320754 (https://zenodo.org/records/8320754), holds the fourteen scenario sampling archives (BAU, LCT, JT, ST and the Sobol sensitivity ensemble), the YAML input settings, the Energy Vulnerability and Sectoral Network Analysis data, and the policy-mix results. This version holds the four files listed above and nothing else.