Optimization with storage respecting time dependancy

Q&A
1

I am just starting off with energy modelling in general and oemof in particular. So, please apologize any possible triviality of my topic!

The optimization problem I would like to approach is the following: take a (simplified) model with source, sink and storage:

                      Bus
                       |
 Src ----------------->| 
 (time dependant)      |
                       |------------------> Sink (fix)
                       |                (time dependant)
                       |
                       | 
                       |--------->
                       |          Storage
                       |<---------

The idea is that the storage will buffer the energy to compensate for differences between sink and source (at a given point in time).

The optimization should now result in a nominal value for the source, keeping its time dependant structure. Of course, in this simple case it should match the nominal value of the sink (with normalized time dependant weights).

What is the best way to approach this optimization?

One (naive) idea is to optimize the model with an additional excess source and/or sink, recalculate the nominal value of the source, recreate and reoptimize the model until the excess entities are consistent with zero?!

But most probably there is a ‘omoef’-intrisic solution for the problem? Any comments are appreciated!

2

@andreasfurtjes, welcome to the oemof community.

You can use the investment option if you want to optimise the capacity of your components. See these examples especially v1 might meet your needs.

Assuming that the electricity demand (Sink) is fix you can find the cheapest combination of e.g. pv and storage or wind, pv and storage or the optimal size of a storage for a given source and so on.

3

Thank you, Uwe, for your fast reply. I will investigate the example program!