gridops

gridops is an open-source and customizable grid operations model for three major U.S. interconnections (Western, Eastern, ERCOT). Production cost modeling in gridops utilizes linear DC Optimal Power Flow (DC-OPF) approximation to optimize hourly unit commitment (UC) and economic dispatch (ED) processes over a user-configurable rolling horizon.

Purpose

gridops was created to:

• Address weather and water dynamics and associated vulnerabilities in U.S. bulk power systems,
• Link with a wide range of hydrometeorological datasets and energy system/machine learning models to explore future uncertainty in grid operations,
• Analyze current and future grid stress and reliability trends under a wide range of weather conditions, generation mixes, and demand growth scenarios.
• Create simple and user-tailored representations of the U.S. interconnections to find a balance between model accuracy and run time to answer research questions of interest.

Installation

pip install gridops

Quickstart

import gridops as go

# Instantiate the model
model = go.Model(
    region="west",              # U.S. Western Interconnection
    problem="linear",           # Linear programming (economic dispatch only)
    solver_name="appsi_highs",  # Using open-source HiGHS solver
    solver_params=None,         # Optional – using default solver parameters
)

# Run for 7 days with a 24-hour rolling horizon
model.run(
    config_file="example_config_file.yml",  # Path to the model configuration file
    n_days=7,                               # Running for 7 days
    horizon_hours=24,                       # Optional – using default 24-hour horizon
)

Configuration file

gridops needs a YAML file that maps each input CSV and NPY files to their file path and each output Parquet to its destination. Check out below for a minimal example (see gridops/data/example_config_file.yml for a full template):

generator_parameters_file: Input/data_genparams.csv
line_parameters_file: Input/line_param.csv
# … all other input entries …
flow_file: Output/flow.parquet
mwh_file: Output/mwh.parquet
# … all other output entries …
restart_file_directory: Restart/

Supported solvers

Users can pass any desired solver parameters via the solver_params:

model = go.Model(
    solver_name="appsi_highs",
    solver_params={"presolve": "on", "solver": "ipm", "time_limit": 3600},
    ...
)

License

BSD-2-Clause – See LICENSE.

Disclaimer

See DISCLAIMER.

Reference

If you use gridops in an academic context, please cite the following article.

Akdemir, K. Z., Oikonomou, K., Kern, J. D., Voisin, N., Ssembatya, H., & Qian, J. (2024). An open-source framework for balancing computational speed and fidelity in production cost models. Environmental Research: Energy, 1(1), 015003. https://doi.org/10.1088/2753-3751/ad1751