08-07-2019, 05:34 PM

(08-07-2019, 04:18 PM)Antti-L Wrote:Abi Afthab Wrote:Thanks a lot for detailed reply. How did you get this maximum and minimum load of this process in each season? Is this calculated by (Var_FIn/(8760*year fraction)) of the process in each time slice? Here in my case, the electricity input is the Var_FIn. I am little confused about this term ''maximum load of the process'". I am very sorry for asking this again.If the capacity is defined by the output (as it was for the electrolyser), the load of the process is the power level of the output flow. The power level of the output flow in each timeslice can be obtained by dividing the corresponding flow (VAR_FLO.L(...)) by the year fraction G_YRFR (and optionally converting it to the capacity unit). For example, for power plants the load at any given time is regularly calculated in this way, usually converted to MW. The maximum load in each season is obtained by taking the maximum (highest) value of these levels over the timeslices in the season, and the minimum is obtained by taking the minimum (smallest) value.

Thanks a lot Anti.

So if I model part load efficiencies and the model is adjusting the output flows in such a way that there are no part load efficiencies in any time slice that means there should be a corresponding investment in storage also, right? What I mean is if I model without part load efficiencies, I should have less investment in storage?

Although, the logic you explained is sensible, I feel it makes more sense to model this with MILP so that this can be done in discrete unit sizes. If this was done with MILP, the model would have done more investment in the electrolyser instead of storage, right?

But I have seen the documentation mentioning that this extension of TIMES and MILP results won't vary much if there are large number of units. In my case, the 2050 electrolyser capacity requirement constitutes almost 2500+ hydrogen refuelling stations. What do you think about it? I am trying to find out a point justifying this type of modelling.

Regards

Abi Afthab