High photovoltaic penetration analysis : case study of Gozo high and low voltage networks

Abstract

This dissertation studies the impact of high photovoltaic penetration on Gozo Low and High Voltage Networks by developing different models based on SMART meter data. The main aim was to develop a practical rule for the maximum allowable level of distributed (household) photovoltaics that could be sustained on a low voltage network without causing any overvoltage issues. This study makes use of the Instantaneous Renewable Penetration Metric - the ratio of renewable power to the system load at any given instant - to determine the impact of photovoltaics on a network. A Structured Query Language model was developed for four Gozo substations known to have overvoltage issues. Generation/Consumption generic profiles were derived. This model showed that high photovoltaic penetration issues were expected during the spring months between 11:00 – 15:00. More precise models were designed for two of the above Gozo networks using renowned power systems software DIgSILENT. Geographic Information System data, physical network component characteristics and the generic profiles derived earlier were used as data sources. Again these models showed that high photovoltaic penetration issues were expected around midday for the spring months. The results of numerous load flow studies (a method used to study existing power systems or future network expansions) for different photovoltaic penetration scenarios showed that a combination of measures were needed to bring all node voltages on all feeders within the limits defined by European Standard EN50160. However it was noted that substation/feeder active power losses had a more influential and limiting effect than overvoltage rise issues. The analysis showed that feeder active power losses: (a) Increased rapidly (exponentially) when the instantaneous Photovoltaic Penetration Ratio went beyond 250%. (b) Remained comparable to those of a ‘No Photovoltaic Scenario’ when photovoltaic penetration level was kept around the 200%. (c) Decreased when compared to those of a ‘No Photovoltaic Scenario’ for penetration levels around the 150% mark or less. The 200% Instantaneous Photovoltaic Penetration limit seemed to be an optimum maximum level to adopt for low voltage networks. The impact of having all Gozo domestic substations reaching this limit was studied for the critical month of May. The results showed that the high voltage network was starting to experience reverse power flow under such conditions indicating this limit should not be exceeded. The impact of installing central battery storage within substations was studied using another software package (HOMER). Batteries have a positive effect by reducing the Maximum Instantaneous Photovoltaic Penetration Level during the mid-day period. The 200% Maximum Instantaneous Photovoltaic Limit is a practical rule Enemalta could adopt to ensure that substation/feeder losses do not exceed those losses experienced when no Photovoltaics were present. Such a rule can be applied for those substations of a ‘domestic’ nature (high household loads) throughout Malta and Gozo. The main scope of the thesis was thus achieved.