get_starting_costs is not respected in calculate_generation_cost

[maurerle]

The get_starting_costs should be used in energy_generation_costs.

It looks like this is an oversight in the framework.
This should reflect the price calculated from get_operation_time and get_starting_costs

• ramping constraints working
• startup cost respected (cold_start in dmas,
• downtime_warm_start used

To me, it does not make sense to have a warm_start at all, as we do not have data for that?

Pypsa also models only the following:

    marginal_cost=20,
    min_down_time=2,
    start_up_cost=5_000,
    up_time_before
    min_up_time=3,

So this looks like a larger refactoring.

@nick-harder do you know how the values for warm_start_cost were estimated?

[gugrimm]

Can you provide more details? Which units? Which strategies?
I would argue that it does make sense to have warm_start and associated costs in a highly RES system (e.g. electrolyzers) and in other systems also for powerplants.

[mthede]

On the starting cost estimation: it's based on Kumar et al. (2012) .

I saw that there are two different cost values logged. "total_cost" and "energy_generation_cost". The starting costs seem to be incorporated in the total costs for flexable and learning strategies?