Abstract:
Photovoltaic (PV) grid connected systems are increasing all over the world. By the end of 2000, thetotal installed power of such application is about 473 MWp. These systems provide electricity at peakloads. As output of the systems depend on radiation level and electrical loads, connecting more PV gridconnected systems to the distribution networks raise concerns on electrical quality and safety due toislanding effects.In 1998 the Energy Policy and Planning Office (EPPO) of Thailand initiated a national preprogram onrooftop grid connected systems for domestic households, government offices and schools, and Royalinitiatedproject sites. All the systems were installed by the Electricity Generating Authority ofThailand (EGAT). No monitoring of the systems has been undertaken. The research of this thesis is onthe on-site measurement (between June and July 2001) and analysis of the system at the Satree SamutPrakarn School under the national program. The system consists of a 4.2 kWp X-Si Siemens array, a3.5 kW Leonics grid-connected inverter (ModelG304). The system is connected to the distributionnetwork of the Metropolitan Electricity Authority (MEA).Parameters under monitoring are radiation level, inverter outputs, harmonics (voltage, current andpower) at the point of common coupling-PCC between the inverter, the load and the distributiontransformer.As the monitoring took place on the system in actual operation, no controlled experiments werepossible, especially variation in radiation levels. Consequently, a laboratory-based system was set up inthe Department of Electrical Engineering of the University. A separately-excited 10 kW DC generator,160 V and 63 A, replaced a PV array. The DC output of the PV array is represented by the generatoroutput.On the actual system, three conclusions can be made. First, the voltage at PCC depends on radiationlevel. Moreover, the radiation level and the rate of radiation change due to clouds affect the voltage.As radiation changes slowly, the variation in the voltage at PCC is not large and is observed to bewithin the limit of the standards set by the electricity utilities.Secondly, a unity power factor is observed at the radiation above 200 W/m2. But the power factordecreases rapidly at low radiation. Even though the observed value is small but connecting thousandsof distributed systems to the network and the effects of larger reactive power flowing into the networkshould be further investigated.Thirdly, significant harmonics observed are odd harmonics, except the second and fourth harmonics.At low radiation range (200-400 W/m2) and high radiation range (400-800 W/m2), harmonicscharacteristics are rather similar. Below 200 W/m2 the total harmonic distortion current is larger than7.5 %, and above the standards whereas the total harmonic distortion voltage is rather dependent on theradiation but within the acceptable standards.The laboratory system is used to study the voltage flickering, the power factor and harmonics at thePCC due to changes in inverter inputs, represented by variation in the DC output of the generator. Threepoints are noted. First, the line voltage largely determines the PCC voltage. This depends on the rate ofchange in inverter inputs, or radiation change. Secondly, power factor is not related to the rate ofchange. Near the cut-off level of the inverter, around 0.10 kW, the power factor is about 0.53 andoutside the acceptable value set by the standards. The corresponding reactive power supplied by thesystem to the line is about 220 VAR. Thirdly, most odd harmonics correlate with the rate of change ininverter input, or radiation change. The total harmonic distortion current and voltage nature are similarto those observed with the actual PV system.