Firstly, thank you for your reply to my first post :http://forum.pvsyst.com/viewtopic.php?f=22&t=1397&p=0&e=0&sid=d750c9a4e98f51edeac50844eb3d05a1#p3934 My second and more practical question may appear simple, but I am highly interested in the accuracy of data. Our SCADA systems are currently connected to analogue thermopile pyranometers. Our daily generation reports consist of a 15 minute time interval and since we do not have dataloggers, what we see are instantaneous values obtained throughout the day. We want to accurately convert W/sq.m to kWh/sq.m everyday to calculate PR. My questions are: 1. Do we directly multiply the instantaneous values with 15 minutes and add up the values for the day? 2. Is this time sufficient or do we choose to then log 1 minute or (any other suggested) intervals? 3. Do we then integrate and find the area under the curve, if so what mathematics are involved? What are the limits of the integral? 4. What back-end calculations does PVsyst do in this regard? 5. How do we reduce the error in our observed values? Thanks again for your replies in advance.
