Chemicals decarbonization model

This repository contains the chemicals decarbonization model, as created by SYSTEMIQ/University of Cambridge/University of Tokyo funded by the Mitsibishu Chemicals Corporation (MCC). This README describes how to set up and use the model. A more detailed description about how the model works can be found in docs/.

Windows

Installing Python

The simplest way is to install the necessary version of Python directly:

• Python 3.9
• When installing, select the option to "add to PATH"
• If installing Python this way, you can use any terminal to run the commands below (e.g. Powershell, or git-bash).

Alternatively, if you are already using Conda, you can use it to get the correct version of Python.

• If using Conda, all the commands below should be in the "miniconda prompt" or "anaconda prompt"
• Navigate to Pathways-Chemical-Industry folder
• Create a Conda environment with the correct version of Python:

  conda create -n Pathways-Chemical-Industry python=3.9

  Don't "activate" the environment now.

Either way, we then need to install virtualenv, which manages all the dependencies of the code to get the correct versions installed. The simplest way is to run pip install virtualenv.

• Navigate to Pathways-Chemical-Industry folder
• Then runvirtualenv Pathways-Chemical-Industry

Setting up and running the model

If using Command Prompt

1. Open a terminal / git-bash
2. Navigate to Pathways-Chemical-Industry folder and activate the environment If using Conda conda activate Pathways-Chemical-Industry If using virtualenv Pathways-Chemical-Industry
3. Install project requirements using pip install -r requirements.txt
4. Create a data/ directory, and add the input data sheet (Master Template - python copy.xslx) into it.
5. Make a copy of the config_template.py file and rename it to config.py. Make changes only to config.py.
6. Run the model using python -m main
7. Outputs are added in the output/ directory

If using PyCharm

1. Navigate to chemicals-decarbonization folder and activate the environment
2. Install project requirements using pip install -r requirements.txt
3. Create a data/ directory, and add the input data sheet (Master Template - python copy.xslx) into it.
4. Make a copy of the config_template.py file and rename it to config.py. Make changes only to config.py.
5. Run the model using python -m main
6. Outputs are added in the output/ directory

Running the model on codespaces

Github offers a way to run software on their infrastructure: Codespaces. Setting up a codespace can be done directly from the repository, and can be accessed through the browser (or VSCode). The benefit of this is that you can use a bigger machine than you own (more cores / memory), to reduce workload on your own hardware. Be careful, this is not free!

The setup on codespaces is exactly the same as on your local machine.

Configuration options

Now, you can run the model with different configurations, by changing these values in config.py:

• RUN_PARALLEL runs the model on multiple cores at the same time, speeding up computation by the number of cores you have
• MODEL_SCOPE allows running the model worldwide, or only for Japan
• CHEMICALS defines the chemicals to run the model for
• run_config allows running parts of the model individually
• PATHWAYS define the pathways that you run the model for, and SENSITIVITIES the sensitivities. It will run all combinations; if you choose 2 pathways and 2 sensitivities, this results in 4 model runs.

There are more configuration options, a complete explanation is in config.py.

Contacts

Technical questions: [email protected]

Subject matter questions: [email protected], [email protected], [email protected] [email protected]