BEHAVIOUR MODELING OF A PHOTOVOLTAIC GENERATOR BASED ON MEASUREMENTS Salim Bouchakour a , Ahmed Tahour b , Kamel Abdeladim a , Houari Sayah c , Amar Hadj Arab a , Farida Cherfa a , Karim Kerkouche a and Bilal Taghezouit a a Centre de Développement des Energies Renouvelables BP 62 Route de l'Observatoire Bouzaréah, 16340, Algiers, Algeria Tel.: +213 21 901 446; Fax: +213 21 901 560 b Masacara University, Route of Mamounia, 29000, Mascara, Algeria c Electrical Department, Djillali Liabes University, 22000, Sidi Bel Abbes, Algeria s.bouchakour@cder.dz ABSTRACT: This paper describes the development of behavioral model of photovoltaic (PV) array generation. The PV array was modeled using an analytical model, which takes into consideration the electrical characteristics provided by the manufacturer data sheet, the temperature and the irradiance level. The needed electrical parameters are re-evaluated in real operational conditions, for better accuracy by carrying outdoor measurements on PV module (PVM). The assessment of the results was done, both for the static and the dynamic cases, in order to provide an experimental validation of the PVM model and the PV array performance model. The simulation results have been achieved using MATLAB/Simulink environment. Results has shown good agreement with experimental data, whether for the I–V characteristics and for dynamic evolution. Keywords: PV array; Inverter; PV system; solar radiation; modeling; simulation. 1 INTRODUCTION The amount of photovoltaic (PV) systems in the distribution network is expected to grow. Increasing penetration levels of distributed grid-connected PV systems may affect the management and structure of distribution network. For these, the important issue of the supervision of these PV array generation. The actual models to describe solar panel performance are more related to physics, electronics and semiconductors than to power systems. Some of the models require several parameters such as the temperature coefficients, photon current, open circuit voltage, series/shunt resistance of the device, etc. Also some of the required parameters in those models are not available by the manufacturer data sheets so it is required to find the information in other sources. At the same time, these models can be impractical and too complex for common tasks in power systems such as power flow, harmonic analysis, sensitivity analysis, load matching for maximum power transferred from the source to the load, etc [1], [2]. To solve these problems and to maximize the use of information provided by field tests, this paper we study behaviour modeling of photovoltaic panel based on the electrical characteristics under standard test conditions (STC) and I-V Curves. This model will be more beneficial and practical for future different works on power systems analysis. 2 GRID-CONNECTED PHOTOVOLTAIC SYSTEM DESCRIPTION The grid-connected PV system plan, currently in service, was achieved in cooperation with the Spanish Agency for International Development Cooperation (AECID). The installation is located at “Centre de Développement des Energies Renouvelables” (CDER) in Bouzaréah, Algiers (latitude 36.8°N, longitude 3°E and 345m of altitude). It started operating on June 2004, the installation operate without storage system. The electricity produced by the PV array feeds our laboratory loads, meanwhile in case of good weather conditions the extra PV generation is injected into the grid, otherwise, the backup is assured by the grid. Fig. 1 shows a diagram of the CDER grid-connected PV system. The electrical energy was calculated through several energy meters; monophasic energy meter calculate the output power for each PV array, three phases energy meter is a bidirectional meter that calculate the energy imported and exported to the grid. The grid-connected PV system includes 90 modules covering a total area of 76m² with an installed capacity of 9,45kWp. The PV generator was designed in three equal PV arrays, of 30 modules for rated power around 3.15kWp; each one was built interconnecting 15 modules in series and 2 in parallel. Three monophasic inverters, of 2.5kWp nominal power, were used to inject PV generation to the grid. The specification of PV module (ISOFOTON I -106/12) and inverter (FORNIUS IG 30) are summarized in tables I and II. Table I: PV module ISOFOTON specifications at STC Symbol Parameter Value Units NOCT Nominal Operating Cell Temperature 47 °C V Nominal voltage 12 V P max Max power 106 ±5% Wp I sc_ref Short-circuit current 6,54 A V oc_ref Open-circuit voltage 21,6 V I mpp_ref MPP current 6,1 A V mpp_ref MPP voltage 17,4 V Table II: Inverter fronius specification at rated conditions Symbol Parameter Value Units V mpp MPP-voltage range 150 - 400 V P n ac-nominal power 2500 W Cos φ Power factor 1 f Frequency range 49,8 - 50,2 Hz V Grid Grid voltage range 195 – 253 V η Efficiency 92,7 - 94,3 % 28th European Photovoltaic Solar Energy Conference and Exhibition 4091