Any analysis conducted in terms of performance?

[DiegoLucadeTena]

I am mostly interested in its performance compared to other tools like pandapower, PowSyBl etc. for contingency analysis and sensitivity analysis. Thank you.

[petersalemink95]

Hi @DiegoLucadeTena,

Thanks for your question!
We have conducted benchmarks against PandaPower and OpenDSS.
You can find the benchmark script (and results of a certain run) in our power-grid-model-benchmark repository: https://github.com/PowerGridModel/power-grid-model-benchmark/blob/main/Power%20Grid%20Model%20Benchmark.ipynb

You can run the script yourself if you want to, so you can compare performance on your own machine.

If you have any other questions, please let me know!

[mgovers]

@DiegoLucadeTena There was a publication on this in https://ieeexplore.ieee.org/document/10267233 .

[DiegoLucadeTena]

Impressive! Can you please indicate the hardware used and the number of nodes in all the tests? Thank you!

[mgovers]

The info on the number of nodes is $10 * 100 + 1 = 1001$ in
Cell #5 in the PGM Benchmark script (https://github.com/PowerGridModel/power-grid-model-benchmark/blob/main/Power%20Grid%20Model%20Benchmark.ipynb)

The original benchmark as published in https://ieeexplore.ieee.org/document/10267233 contains the following line on the hardware that was used:

The tests are performed using an Intel i7-8850H processor and 40 GB RAM, single-threaded in Linux (Windows Subsystem for Linux).

[TonyXiang8787]

Hi @DiegoLucadeTena,

It is strongly recommended that you run the benchmark script in your own hardware to see the actual performance. You can also adjust the size of the network and scenarios at the beginning of the notebook.