bonsamurais / bentso Goto Github PK

Sometimes people have a data cache.

When the final optimization model from @cdgaete is checked in, need to remove/move notebooks, delete temp files, etc.

• Generate pumped hydro generation mixes
• Filter by time interval (e.g. mix just for daytime)
• Include trade
• Conceptual model for markets, consumption, generation, and trade
• Conversion from ENTSO-E labels to B ontology

follows from 0.1 release described in #3

@bkuczenski @cdgaete @ntropy-esa @jakobsarthur

We need a strategy on how to account for the electricity used to charge pumped hydro (we don't have any other storage in the API, at least as far as I can tell).

The problem is that pumped hydro is a generator, but they don't list it as a load. We need to keep it in the generation mix to have energetic balances.

The API does have filling of hydro storage plants (search for "Aggregated Filling Rate of Water Reservoirs and Hydro Storage Plants") - but this is only weekly totals.

ecoinvent and electricitymap.org just use the average generating mix - which itself already includes pumped hydro (gasp!). I also asked the open energy modelling folk, but didn't find anything helpful.

So, the best I can some up with is to charge with the weighted average mix when pumped hydro is not being used, with weights coming from the inverse of the electricity price - the assumption being that lower prices are more likely to times when pumped hydro would be charged. But maybe you have a better idea? You can see the available data in the example notebook.

Needs to include the following:

• Store ENTSO API key as environment variable
• CI for linux and windows
• CI for test coverage
• Auto-building docs
• Register package on PyPI
• Cache data retrieved from ENTSO in default directory
• Use caches from other directories/network mounts
• Simple interface to specify data to retrieve

Need to save results locally to avoid too many requests. Should integrate with CLI to delete old data, show how much is used, etc.; see ocelot repo for examples.