This repository contains the data and code to reproduce the engineering-economic optimization (EEO) modeling results from the study "Simulating institutional heterogeneity in sustainability science". Optimization code is written using JuMP for Julia and data processing and figures are generated using Python.

File organization

data

It contains two data folders to run EEO model.

1. powergenome is the main input data which provides major data resources such as load and generators in western USA.
2. extra contains additional data. It also includes data_settings.yml, which is a setting file to run PowerGenome package in order to get input data provided in powergenome. PostProcessing.jl is the Julia code which we use to edit/adjust the raw data provided by PowerGenome run.

figs

It contains figs.ipynb notebook to generate Figure 3 in the main text and Figure S.2.1 in the supplementary file. It reads data from powergenome and extra folders, as well as model results from different folders. Running this notebook does not require any virtual environment. base environment can be used to successfully generate the figures. Users should install odfpy package to read results of ABM model.

model

It includes Julia code to run EEO model and replicate the corresponding results in the paper. batch contains two .sh files, which are bash files to submit jobs in TORQUE clusters. Inside batch, m2050_tbaseline_pregionalces_rbaseline_linear.sh is used to replicate results with heterogeneity. m2050_tbaseline_pregionalces_rregionalized_linear.sh is used to replicate reference results (e.g., without institution or barriers).

results

eeo folder contains EEO modeling results with and without heterogeneity (e.g., reference). abm and iam contain their modeling results in appropriate formats, solely to generate figures in the paper.

Reproducing EEO results

There exist two ways of reproducing the results. One is to run the model in a TORQUE environment, and the other is to run on a local machine.

To run the model in TORQUE

1. Make sure that Julia and Gurobi are installed.
2. Put folders data and model in the same directory.

To replicate heterogeneity results,

3. Uncomment Line 116 in ExpansionModel.jl
4. Submit m2050_tbaseline_pregionalces_rbaseline_linear.sh in model/batch.

To replicate reference results,

3. Comment Line 116 in ExpansionModel.jl
4. Submit m2050_tbaseline_pregionalces_rregionalized_linear.sh in model/batch.

To run the model in local machines

1. Make sure that Julia and Gurobi are installed.
2. Put folders data and model in the same directory.

To replicate heterogeneity results,

3. Uncomment Line 116 in ExpansionModel.jl
4. Change runname variable in Line 4 of Run.jl to m2050_tbaseline_pregionalces_rbaseline_linear
5. Create an empty folder in model/batch with name Results_m2050_tbaseline_pregionalces_rbaseline_linear
6. Create an empty folder in model/batch/Results_m2050_tbaseline_pregionalces_rbaseline_linear with name Dispatch.
7. Run Run.jl file.

To replicate reference results,

3. Comment Line 116 in ExpansionModel.jl
4. Change runname variable in Line 4 of Run.jl to m2050_tbaseline_pregionalces_rregionalized_linear
5. Create an empty folder in model/batch with name Results_m2050_tbaseline_pregionalces_rregionalized_linear
6. Create an empty folder in model/batch/Results_m2050_tbaseline_pregionalces_rregionalized_linear with name Dispatch.
7. Run Run.jl file.