1. Overview for the NEG Resource Adequacy Project

This repository contains the unit commitment and economic dispatch (UCED) model for the Northeast Power Grid (NEG) of China for the study "Resource Adequacy Under Institutional Constraints and the Low-Carbon Energy Transition in China". The model is written using Julia and uses Gurobi as the optimization solver. Results are visualized using R and Python. The model solves weekly optimization problems for the 12 weeks spanning August to October.

This repository is mainly structured into two components:

1. 2021 retrospective analysis (uced-neg-2021): Focused on historical data and scenarios. In the base case, the derated rate of coal units is set at 35%, while in the sensitivity analysis, it is assumed to be 40%.
2. 2030 forward-looking analysis (uced-neg-2030): Focused on future projections and scenarios.

In addition, figures_for_paper contains the figures displayed in the JEPO paper.

Contact Ming (m2wei at ucsd dot edu) if you run into any issues.

---

2. System Requirements

Hardware

• Operating System: macOS or Windows.
• Minimum Requirements:
  • 4 GB RAM
  • 1 GB disk space
  • Intel Core i5 or equivalent processor.

Software

• Julia: Version 1.8.3
• Gurobi: Version 10.0.1 (requires a valid license)
• Python: Version 3.9 (for visualization and post-processing)
• R: Version 4.3.1 (for visualization and post-processing)

Julia Dependencies

• JuMP=1.16.0
• DataFrames=1.3.6
• CSV=0.10.11
• Missings=1.1.0

Python Dependencies

• pandas=1.5.3
• numpy=1.23.5
• gurobipy=10.0.2
• matplotlib=3.7.1
• geopandas=0.12.3
• plotly=5.14.0

R Dependencies

• dplyr=1.1.4
• readr=2.1.4
• sf=1.0.16
• ggplot2=3.4.3
• ggpubr=0.6.0
• ggmap=3.0.2
• viridis=0.6.4
• hrbrthemes=0.8.0
• tidyr=1.3.0
• here=1.0.1
• cowplot=1.1.1
• gridExtra=2.3
• RColorBrewer=1.1.3

---

3. Installation Guide

1. Install Julia dependencies:

   using Pkg
   Pkg.add(["JuMP", "DataFrames", "CSV", "Missings", "Gurobi"])

2. Install Python dependencies:

   pip install pandas numpy gurobipy matplotlib geopandas plotly

3. Ensure you have a valid Gurobi license. Academic licenses can be obtained here.

---

4. Running the Model

Local Execution

1. Run the model:

   julia Run.jl

2. Outputs will be saved in the Batch directory, organized by scenario and week.

Cloud Execution

1. Create a shell script adapted to your server environment.

2. Submit the job to the cluster, e.g.:

   sbatch Run_cluster.sh

3. Retrieve results using tools like FileZilla.

---

5. Data Description

• Fuels_data: Fuel cost and availability data.
• Generators_data: Information on generators, including:
  • Resource type (e.g., solar, wind, coal).
  • Capacity, ramping, and heat rate.
  • Fuel requirements and costs, etc.
• Generators_variability: Hourly variability of each generator.
• Heat_time: Heating and non-heating periods for each province.
• Load_data: Hourly demand data for four zones in the Northeast China Grid and two zones in the North China Grid.
• Network_forward/Network_reverse: Transmission network setup in both directions.
• Transmission_MLT: Hourly interprovincial/interregional transmission amount stipulated by Medium to Long-term (MLT) contracts.
• Operating_reserve: Reserve requirements for loads and VRE.
• other_inputs: Initial and final states of storage; reservoir's minimum level.

---

6. Script Description

• Run.jl: Main script to execute the model.
• Paths.jl: Defines input/output directories.
• ReadFiles.jl: Reads input data files.
• SetCreation.jl: Creates data sets for indexing.
• EDUCModel.jl: Core UCED model function.
• RecordCSV.jl: Saves main optimization outputs to CSV.
• ProcessDispatch.jl: Processes and records hourly dispatch results.
• Plot_figureX.py: Visualizes results and plots figure for the JEPO paper (X represents the figure number).