Electricity Partial Equilibrium Model (PEM)

RTI

Updated by Shane Weisberg, October 2021

Run Instructions:

Using the terminal, navigate to the "Shell" directory and run the shell script run_pe.sh Add the flag -o to this command to run the original, unmodified optimization model If bash is not installed on your system (windows users), you can download gitbash from https://git-scm.com/downloads and configure VS Code to run bash in its terminal

Contents:

GAMS Files

• model.gms: This is the main script that will run the optimization problem. It contains the definition of the objective function and all constraints. It calls the other essential .gms files described below.
• data.gms: This is the script responsible for loading and preparing data. In the original version of the model, the user can set a carbon tax and other parameters in this script.
• sets.gms: This is the script responsible for defining all sets used by the optimization problem.
• findings.gms: This script will produce reports and outputs.

Shell Scripts

The directory "Shell" contains the following two bash scripts:

• run_pe.sh - this script should be used to initiate model runs
• csv_to_gdx_batch.sh - this script is able to convert csv data input files to gdx files and must be run once any time the input data is updated

Data Files

The directory "Data" contains the following:

• In the "updated_data" directory there are csv files containing supporting data for the model.
  • PE_capacity.csv, PE_pf.csv, PE_dele.csv, PE_fomcost.csv, PE_vomcost.csv, PE_heatrate.csv, PE_count.csv contain data that has been updated using the MEEDE dataset
  • PE_capcost_2020.csv, PE_dadj_2020.csv, PE_pbio_2020.csv, PE_biosupply_2020.csv, PE_peak_2020.csv, PE_pele_2020.csv contain original unedited data truncated to only contain data pertaining to the year 2020.

Output Files

The directories "Model/Output" and Model/Results" are set up to receive outputs from the model. "Model/Results" contains results in csv format on generation, capacity, and costs. These files can be read and processed by the included R scripts to allow for easy comparisons of model runs.