System integration, storage, and grid in WITCH – Additional work needed
Talking about Deliverable D1.1, I had mentioned the work needed to align the WITCH version developed for the MERCURY project by Marco, Simone and me with the official version of the code (that we normally call “master”). Simone took care of this activity, which seemed to have been completed.
However, in the last weeks some technical issues arose. The problems are not related to the quality of the new modules and their integration in the master version of the model – Simone did a great job – but to the running time, that in the new model version has become much higher than before. In general, this is somehow expectable, given the higher complexity of the optimization process with the inclusion of curtailment, the new flexibility and capacity constraints depending on the renewable share, the new technology differentiation for storage, etc. but in the last modeling meetings the extent of the phenomenon has been assessed to be too large. Therefore in the next weeks I will have to put my hands on the code in order to simplify the formulation (in terms of pure coding, i.e. without sacrificing the new modeling features that we added in our work) and thus accelerate the running process.
I say “I” because unfortunately I will no longer be able to count on Simone, who in the meantime completed his internship at FEEM. Too bad, also because over the past months he had become more expert than me concerning all the small and often fundamental technical details of the code, that I will have to get acquainted with again! Also for this reason, this activity is likely to take some time (probably weeks). This is quite bad, but after all I would really like the new model version of system integration, storage, and grid to be fully included in the WITCH master, so this activity is an absolute priority.
So, let’s work and forget about the shifted agenda for the moment!