Welcome to GridOverture
GridOverture is an open-source Python library that unifies power system solvers behind a single, clean API. Instead of learning the quirks of each solver individually, you write one script and run it against any backend.
Load. Solve. Compare.
import gridoverture as go
net = go.load("ieee39")
result = go.run_load_flow(net, solver="veragrid")
comparison = go.compare(net, solvers=["andes", "veragrid", "pypsa"])
Who is GridOverture for?
Researchers
Benchmark solvers side-by-side, generate ML-ready datasets, and reproduce cascade simulations calibrated against PowerGraph (ETH Zurich, NeurIPS 2024).
Engineers
Run power flow and OPF studies across multiple backends without rewriting your pipeline. Switch solvers with a single parameter.
Students
Learn power systems with built-in IEEE test networks and an approachable Python API. No boilerplate, no configuration files.
What problems does it solve?
- Solver lock-in: Each power system tool has its own data format, API conventions, and output structure. GridOverture abstracts these differences behind a unified interface.
- Reproducibility: Compare results across solvers deterministically. The cross-solver comparison engine highlights discrepancies in bus voltages, line flows, and losses.
- Dataset generation: Produce ML-ready datasets from realistic cascade simulations for GNN and other graph-based models.
- Fragmented tooling: Instead of maintaining separate scripts for ANDES, PyPSA, pandapower, and others, maintain one.
Key Features
7 Solver Backends
andes, veragrid, pypsa, pandapower, dc, pypsa-dc, pandapower-dc
8 Test Networks
ieee14, ieee24, ieee39, ieee57, ieee118, ieee300, uk_29bus, wecc240
Cascade Simulator
Calibrated against PowerGraph (ETH Zurich, NeurIPS 2024)
Cross-Solver Comparison
Side-by-side voltage, flow, and loss comparisons
OPF Support
PyPSA LOPF and pandapower AC OPF
Apache 2.0 License
Free for commercial and academic use