A new procedure is presented to analyse the performance of grid-tied PV facilities. It needs limited amounts of data that are easily sourced and is based on knowledge of the analysed system and its mode of operation. The procedure is applied, in a case study, to compare real PV production at two 100 kWp grid-connected PV installations. Located in the same geographical region, the installation of these two facilities followed the same construction criteria - PV panels, panel support system and wiring - and the facilities were exposed to the same atmospheric temperature and solar radiation. They differ with regard to their inverter technology: one facility uses an inverter with an integrated transformer system and the other uses a transformerless inverter. The results show that the transformerless inverter system performed better than the isolated system by a factor of 1.2%, which, in economic terms, represents more than 2000 is an element of/year. (C) 2013 Elsevier Ltd. All rights reserved.