PV MODULE PEAK POWER: COMPARISON BETWEEN MANUFACTURER DATA AND LABORATORY RESULTS M. Alonso-Abella, J.P. Silva, F. Fabero, N. Vela and F. Chenlo Ciemat PV Laboratory Avda. Complutense, 22 – 28040 Madrid, Spain. Phone: +34913466492; Fax: +34913466037; E-mail: miguel.alonso@ciemat.es ABSTRACT: The Ciemat PV Laboratory usually performs I-V curve measurements of PV modules in order to determine their power and other related characteristic electrical parameters. This paper describes a procedure to perform the peak power control of large provisions of PV modules for big PV power plants. The proposed procedure includes the following steps: initial comparison of measurement results between manufacturer and laboratory considering a sample of at least 10 PV modules in order to check and calibrate power and verify measurement repeatability; sample selection from a given manufacturer batch; sample testing; modification of the manufacturer flash report data for the batch according to the testing results obtained for the sample in the laboratory; application of the conditions provided in the contract of sale signed between the PV module manufacturer and the owner of the PV plant according to the values given by the manufacturer and corrected by the laboratory results. In this work data corresponding to 6,333 PV modules, 140 models from 52 manufacturers, tested by the Ciemat PV laboratory during the last 18 months, are included. Keywords: PV Module, nominal power, Quality Control 1 INTRODUCTION Spanish grid-connected PV market has increased from 100 MWp to almost 1.7 GWp during 2008, mainly as a result of the current feed-in tariff in force (0.44 c€/kWh according RD 661/2007 until September 2008). The sales price of PV modules is related to their peak power, i.e., the power delivered at Standard Test Conditions (STC). PV promoters and PV module buyers are aware that there could be some differences between the PV manufacturer data (rated values and flash-report) and the peak power measured by external PV laboratories. Peak power declared by manufacturers usually lies in ±3% to ±5% from a central nominal value but sometimes the real peak power of a single PV module is lower than the declared power. In order to guarantee the installed PV peak power, paying only for the “real” Wp, and also to accurately estimate yearly energy productions, quality controls are usually requested by PV modules buyers. Among other requirements, it becomes usual nowadays to randomly select a sample of PV modules from a production batch to be tested in an independent external laboratory. Different pass/fail criteria can be defined as a result of manufacturer and purchaser negotiations if a single tested PV module of the sample is out of specifications, for example batch rejection, financial claim or an additional supply of PV modules. Ciemat PV Laboratory has got a testing infrastructure for the I-V curve determination at STC [1] and performs peak power quality controls of PV module production batches as support of the PV Spanish industry and business [2]. I-V curve measurements can be done either in indoor conditions in a Class AAA [3] 10 ms pulsed solar simulator or in outdoor conditions. This activity is supported by other activities in the Laboratory, for example irradiance calibrations, international traceability through participation in round robins and intercomparisons, spectral mismatch calculations and in- house electrical calibrations. 2 QUALITY CONTROL PROCEDURE To ensure the energy production and the quality of large PV grid connected plants one of the main issues is to control the supplies of PV modules by measuring their real peak power at STC. It is also interesting to know some other basic electrical parameters (Isc, Voc, Im, Vm and FF) in order to reduce mismatch losses in the plant. Manufacturers deliver the nominal peak power of PV modules with a given tolerance. Some manufacturers, even for the same cell technology, divide their PV module production in different models depending on the final power obtained for the manufactured modules. Historically if the power of a module was included inside the allowed tolerance then the module was accepted as belonging to a given model. For the acceptance of a supply it is a common practice to classify PV modules in batches (typically a batch is equal to the number of modules that fit in an international transport container). Due to the current evolution of the PV market it is additionally required nowadays that the mean power of a whole batch has to be equal or higher to the nominal power of the model of PV module considered. Every batch has to be characterized for its acceptance, rejection or penalty. Typically a representative sample from 1% to 5% of the total number of PV modules in each batch (population), minimum 12 to 16 modules per batch, is randomly selected for laboratory testing. For the acceptance of a PV module large supply, two main points should be considered: 1. The nominal power provided by the manufacturer in the flash report for each PV module should fit within the allowed power tolerance. The mean power of the batch should be equal or higher than the nominal power of the model considered. 2. Manufacturer procedures to measure the electrical characteristics of PV modules must be appropriate. A key point in the process of acceptance of batches that simplifies the whole process and minimise further problems is to compare, prior to start the process itself, measurements performed by the manufacturer and by the