A few years ago I worked for a wind turbine company (Admittingly a crappy one) and I was one of two software engineers,
My first job was to analyse all of the data about 300 turbines were sending back home, and build a dashboard.
Well a few weeks later after checking all of the data, it turns out that the entire fleet of turbines actually consumed slightly more energy than they produced.
This was because of many manufacturing problems meant there were several frequencies that the turbine couldn’t spin at because it would generate a resonance and shatter or break something. So the system was frequently applying the break to stop the turbine destroying itself.
After several days of head scratching and double-checking and triple-checking the data, I went to the mechanical engineers and they finally sheepishly admitted they knew these things were net energy consumers rather than producers.
This doesn’t even take into account the energy budget of cost of production and distribution.
This isn’t the case for the big turbines you see out in the field, I’m really just reenforcing the parent comments point that total energy needs to be considered.