In the past days I completed the implementation of the improved system integration modeling in WITCH, according to the guidelines that I had described in the April 22 post. Fortunately, the number of infeasibilities and computational problems has been acceptably limited, which allowed me not to waste too much time in annoying debugging operations. The only small, critical point concerns depreciation, that I had to leave pending. This aspect is in fact quite awkward because it substantially affects the modeling of the entire energy system, and it will require a more in-depth and dedicated work. After all, it does not directly regard system integration, so it can be temporarily “frozen” with no consequences on the completeness of this work.

As I mentioned in my previous posts, I will describe in detail the new formulation and the relevant results in the presentation that I will give at FEEM next month.

Having completed this task, I was able to dedicate myself to other two activities, which will be my focus in the remaining weeks of the first phase here in Berkeley. The first one is the dynamics of decarbonization, which I have already talked about in the introductory presentation and in my previous post. The second one is an activity that I had not mentioned yet. It regards the exploration of future pathways of solar PV learning in Integrated Assessment Models, i.e. the impacts on PV investment costs and penetration deriving from different assumptions about the learning rate, which is the core parameter linking capacity deployment with cost reduction. I will give more details about this work in a dedicated post in the next weeks.

Both these activities will be discussed along with the main system integration task in my next June’s presentation, naturally.