Here is a small bit of open science:
- true record of three-phase grid voltage at CentraleSupélec (Rennes campus)
- Python code to analyze the harmonic content (by least squares fitting, not FFT)
Published open source & open data: https://github.com/pierre-haessig/grid-voltage
Open question: how specific to Rennes campus is the grid voltage shape?
More grid voltage records data (by RTE)
That being said, after a tiny data survey, I found that RTE made available last year a dataset of 12053 records of grid records (specifically grid fault events), i.e. a much richer database! https://github.com/rte-france/digital-fault-recording-database
Context: grid connected inverter
This was created in the context of another investigation: the harmonic current perturbation in a grid connected inverter. And in the end: the harmonic current looks indeed proportional to the harmonic voltage, so the grid imperfection (albeit perfectly within compliance standards) is the likely cause (along with the simplistic inverter control which make no effort to reject harmonics)...
