Assessing the design options for the optimization of a 2MWp ground mounted PV system in Malta

Abstract

The objective of large scale PV installation in space constrained countries has gradually shifted from production maximisation to that of reaching an optimised economic performance. A main reason for this has been the decline in the price of PV modules. This has resulted in systems with lower tilt angles, increased mutual shading and overloading of inverters by design in the continuous quest of balancing reduced yields with development costs over the entire lifespan of the project. This approach has shifted the traditional view of evaluating a PV installation from a Euro/watt approach to the use of current optimisation metrics of Levelised Cost Of Electricity (LCOE) or Internal Rate of Return (IRR). This is the concept of value based design. In fact, optimisation can be simply described as a balancing act, evaluating tradeoffs to assess which combination gives the best economic performance. By its very nature, optimisation is an iterative process. Large installations provide room for design flexibility where the additional returns justify the energy and cost of the optimisation work. Optimisation work is very much site specific. What has been proven as the best approach for other countries does not necessary apply to Malta. This dissertation evaluates a number of design parameters which, together with production modelling within the context of the Maltese solar climate, aims at identifying an optimum design for the building of a 2MWp ground mounted PV farm. Among the aspects considered, are layout optimisation with issues of tilt and cross shading and inverter architecture whether string or central, including the controversial dc-toac ratio. Central inverter architecture is found to provide a better economic performance over string inverter architecture. The dominant factor in this respect is that no allowance has to be made for the complete replacement of a central inverter within the project life as is the case with string inverters. Yield optimisation is reached with a dc-to-ac ratio of 1.20, however, economic optimisation with respect to string inverter architecture is achieved with a much higher ratio of 1.38. The economic metrics have shown a high sensitivity to the method of vi land acquisition for the project. In the case of an outright purchase, best economic performance is reached with those plant parameters which minimise land usage, that is, low tilt angle and high limit angle. In the case of a land lease agreement, maximising energy production takes the leading role over land usage minimisation resulting in a plant utilising a higher tilt angle and a lower limit angle when compared to the previous scenario. This research highlights the importance that plant optimisation studies should be undertaken prior to any large scale PV development, as plant parameters which guarantee the best return on investment are not static but vary according to the costs and revenue structures of the project at the time of development.