Table input and output for PyPSA

Build and run a PyPSA network from an Excel table; suitable for users with little programming experience

This repository holds an interface for PyPSA that allows to build a network from a table input (xlsx or csv) file. It is designed to be easy to use for users with little programming experience. It allows to build a network from a table input file and to optimize a PyPSA network with a single command without requiring detailed knowledge of the PyPSA syntax. Its limitations lie in the complexity of the network that can be built. For more complex networks, we recommend to use PyPSA-Eur or PyPSA-Earth.

The table input file

The following information is read in from the input file (see test/test_case.xlsx for an example):

• Case Data holds general information about the setup like
  • input/output paths
  • time range
  • solver, logging level, numeric scale factor
  • units

• Component Data holds information about the components in the network and their attributes like
  • component type and name
  • bus(ses) connected to this component
  • capital and marginal costs
  • time series file names
  • other attributes like efficiency, max_hours, etc

Technology data

Costs and other technology inputs can be read in from the PyPSA technology database or manually specified in the input file.

• To read values from the database, adjust the costs_path to the cost assumptions to be used (default is 2020 costs) and choose the name of the technology exactly as in the database. In the cells that should be read from the database, enter "db". If this attribute does not exist in the database, this will cause an error. To multiply the database value by a factor, enter "factor*db_attr", e.g. "3*db_capital_cost". Empty cells default to PyPSA defaults.
• To manually define values, just enter the value in the input file. Use technology names that are different than those in the database to avoid confusion.
• Combinations of both are possible as well.

(See test/test_case_db_values.xlsx for an example)

Installation

Clone this repository

with --recursive (this clones PyPSA as a submodule )

git clone https://github.com/carnegie/table_pypsa --recursive

which creates a directory table_pypsa, cd into that director with

cd table_pypsa

Install dependencies in the environment

When you're running for the first time, create a new environment based on the env.yaml file. We recommend using conda for easy package management.

If you're running for the first time, create the environment with

conda env create -f env.yaml

Then activate the environment with

conda activate table_pypsa_env

every time you want this code.

Install a solver

For using Gurobi, the installation is already included in the environment but you still need to obtain and activate a license. Free licenses for academics are available.

For other solvers, see the installation instructions in the PyPSA documentation.

Run PyPSA

To run PyPSA, you need to have a case input file and data input files.

pyPSA is run with the command

python run_pypsa.py -f <input_file>

where <input_file> is the case input file, e.g.

python run_pypsa.py -f test/test_case.xlsx

Note: If you run into a problem executing the code, try running within a 'bash' shell.

Create a new project based on table_pypsa

To add table_pypsa as a submodule in a new repository, do

git submodule add -b main https://github.com/carnegie/table_pypsa

Make sure to update the submodule regularly by doing

git submodule update --remote --recursive

To run PyPSA in that repository, then run

python table_pypsa/run_pypsa.py -f <path_to_your_case_file>