solph speed up

[simnh]

Pyomo tends to be quite slow with lots of memory usage for large LPs. This has been discussed by other frameworks like calliope, PyPSA or urbs as well.

There are currently different options and developments:

1. Moving code to Julia (which is significantly faster)
2. Using PyPy which is supported by Pyomo >= 5.6
3. Using a lower-level API of Pyomo like this one here: https://github.com/Pyomo/pyomo/search?q=linear_constraint which should be faster.

I assume the potential seed up is highest at 1 and lowest at 3.

Somebody interested in working on this?

[p-snft]

This might become obsolete if #586 is successful. I will keep the issue open until then. However, this is nothing for the short term? I set it to "nice to have".