So kilowatts per hour is a valid measure in this case.
I hadn't considered this, but of course it is obvious in hind site.
Lol you could probably differentiate one more time and figure out the plants ability to change the rate of changing the power output. If it wasn't a technical limitation it would still be true because of the latency between sensing a higher demand and issuing the new output.
So I guess kw/hour/hour is a unit too.
I think that becomes important when considering start-up conditions. For instance, when you turn on an incandescent lamp, the current changes as the filliment warms up. A * t^-2
And picking up on the thermal conductivity post of a few days ago, at aluminium smelters, power outages are a problem because as the metal cools, it takes more energy (current) to reheat - another kw/time measure.
The short answer is probably blame the boffins - they have a measurement for everything.