Addressing PV plant performance test uncertainties: Case studies with data analysis

This paper describes methods commonly used for PV plant performance tests. The performance target is calculated for a 30 MWac PV plant currently operating in Austin, Texas. Actual plant data that was collected during performance test in February 2012 is presented and analyzed against targets set by different methods. Uncertainties emerging from the selection of the transposition algorithm (Hay vs. Perez) are studied. The paper also presents uncertainties emerging from acceptance or rejection of certain irradiance data sets, changing test duration, and implementing data exclusions as per standard ASTM E2848-13 are presented.