Improve industrial electricity demand calculation #63

I copied this from timtroendle/possibility-for-electricity-autarky#1, opened by @brynpickering .

Industry electricity demand is currently calculated as follows:

European quantity of manufactured goods per sub-class -> plant quantity of manufactured goods based on its relative contribution to sub-class emissions -> plant electricity consumption based on its energy input per unit weight of manufactured good. This is done for the three main emitting industries: ferrous metals, chloralkali, and Aluminium.

The annual energy consumption for industry, broken down into industry subclasses and by input energy source, is published by Eurostat for all European countries.

I would propose to keep the relative quantity of manufactured goods to be based on annual emissions, but equate that to national statistics. Annual emissions scraped from the EU-ETS register (e.g. this plant) can then be used to have plant contributions be specific to the model year. So the workflow would look as follows:

National industry energy consumption per sub-class -> plant energy consumption based on its relative contribution to sub-class emissions. This would be done for all industry sub-classes.

From a first pass of the EU-ETS data, I have >5000 sites mapped, which gives a good coverage of industry in all participating countries. Non-participating countries (Switzerland, Western Balkans) could just use the national-level data from Eurostat, although Switzerland has its registry that I'm looking into how one could access...