Dear,
I am working on a simulation of a stand-alone pv-hydrogen-battery system. I am using the mastercontroller, "Unit 36", to control the power flow between the solar, battery, electrolyzer, fuel cell and load. Unfortunately when running the simulation for a whole year with timestep of 1 hour and the successive solver and differential equation algorithm of Heun or Euler it gives me convergence issues and stops the simulation. Ofcourse I can set the limit for the errors so high but then the simulation finishes with really strange results. For example even the solar energy coming in is totally changed throughout the year. If I make the timestep smaller, such as 0.01 then it works kinda oke but still the solar energy curve is totally different then it should be when nothing would be connected to it. See the two graphs added of the solar energy curve throughout the year.....
Does anybody has a clue on how to tackle this convergence issue? Do I need to make the stepsize even smaller? But then it takes so long, that when I want to optimize my system with TRNOpt it takes probably about 3 days per simulation which is ridiculous by my understanding.
Can someone please help me out or point me towards a direction on where to find a solution?
Thanks in advance!
Kind regards,
Michel