APPLICATION How accurate is a commercial monitoring system for photovoltaic plant? Lorenzo Fanni 1 * , Mattia Giussani 2 , Matteo Marzoli 2{ and Miglena Nikolaeva-Dimitrova 1 1 Institute for Renewable Energy, EURAC Research, Viale Druso 1, 39100 Bolzano, Italy 2 Ricerca sul Sistema Energetico RSE S.p.A, Via Rubattino 54, 20314 Milano, Italy ABSTRACT According to uncertainty calculations, the values recorded by means of commercial monitoring systems are expected to be less accurate than those recorded by a system optimized for the measurement of electrical parametersthe so-called dedicated system (DS). This study aims to verify if a larger expected uncertainty for commercial system (CS) actually turns into a larger spread of the measurements around the true value. In the Airport Bolzano Dolomiti plant, CS and DS are installed on the same photovoltaic arrays. The comparison performed considers the detailed uncertainty budget for the two systems using three performance indicatorsenergy, yield and performance ratio. Results show that the uncertainty level of the CS is much larger; for example, on performance ratio, it is about four times larger with respect to the optimized one (respectively Æ16% and Æ4%). Three sources mainly contribute to the uncertainty: measurements of irradiance, current and voltage. The measured values of the electrical parameter are compared in order to verify if the results of the budget calculations turn into a real difference. Results show that the CS is accurate in measuring current and voltage, respectively, ~2% and ~5% of difference from the DS, but not for irradiancehere, the difference is higher than 10%. In particular, the irradiance measured by the CS is systematically smaller; therefore, the performance ratio calculated through the CS is always overestimated and often larger than 100%. Copyright © 2012 John Wiley & Sons, Ltd. KEYWORDS uncertainty calculation; monitoring system; performance; outdoor test; photovoltaic systems *Correspondence Lorenzo Fanni, Institute for Renewable Energy, EURAC Research, Viale Druso 1, 39100 Bolzano, Italy. E-mail: lorenzo.fanni@libero.it Present address: IMT EPFL PVLab, Rue Breguet 2, 2000 Neuchâtel, Switzerland. { Present address: SUPSI ISAAC, 6952 Canobbio, Switzerland. Received 2 December 2011; Revised 8 October 2012; Accepted 22 October 2012 1. INTRODUCTION Nowadays, the number of large photovoltaic (PV) plants is increasing signicantly, and the task of their proper monitoring and maintenance becomes a critical issue. Normally, monitoring activities of the plant performance are carried out through the monitoring system integrated into the array inverter, whereas only few plants are equipped with a dedicated monitoring system. Because of its importance for economical purposes (e.g., feed-in tariff remuneration), the energy injected into the grid is always accurately measured through a Class 1 electricity metering device (1% of accuracy error on produced energy [1]), but other parameters related to the intrinsic features of the system are not measured with the same accuracy. This fact poses problem for fair comparison between different PV plants. There are various standards that outline recommendations and guidelines for the monitoring system. The international standard IEC 61724 Photovoltaic system performance mon- itoringGuidelines for measurement, data exchange and analysis[2] denes the maximum allowed uncertainties for the measuring sensors. According to this standard, irradiance accuracy should be smaller than 5%, voltage and current accuracy smaller than 1% and alternate cur- rent (AC) power accuracy smaller than 2%. The standard gives also suggestions regarding the sampling interval. The international standard IEC 60904-1 Measurement of currentvoltage characteristic[3] deals in particular with the measurement of the IV curve, dening more PROGRESS IN PHOTOVOLTAICS: RESEARCH AND APPLICATIONS Prog. Photovolt: Res. Appl. 2014; 22:910922 Published online 20 December 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/pip.2328 Copyright © 2012 John Wiley & Sons, Ltd. 910