In order to calculate 25 year approx yield, I have been running a batch simulation where I increase the module quality loss by 0.5% per year and then run the simulation 25 times. This way only the modules are "degrading." I'd like to get some feedback on what you think about this approach.