Design process optimisation of solar photovoltaic systems

The design processes for solar photovoltaic (PV) systems is improved to achieve higher reliability and reduced levelised cost of energy (LCOE) throughout this thesis. The design processes currently used in the development of PV systems are reviewed. This review process included embedding the author in a project to deliver four rooftop PV systems which totalled a megawatt of installed generating capacity, which at the time represented very significant system sizes. The processes used in this are analysed to identify improvement potential. Shortcomings are identified in three main areas: safety assurance, design process integration and financial optimisation. Better design process integration is required because data is not readily exchanged between the industry standard software tools. There is also a lack of clarity about how to optimise design decisions with respect to factors such as shading and cable size. Financial optimisation is identified as a challenge because current software tools facilitate optimising for maximum output or minimum cost, but do not readily optimise for minimum levelised cost of energy which is the primary objective in striving for grid parity. To achieve improved design process integration and financial optimisation, a new modelling framework with the working title SolaSIM is conceived to accurately model the performance of solar photovoltaic systems. This framework is developed for grid connected systems operating in the UK climate, but it could readily be adapted for other climates with appropriate weather data. This software development was conducted using an overarching systems engineering approach from design and architecture through to verification and validation. Within this SolaSIM framework, the impact of shading on array and inverter efficiency is identified as a significant area of uncertainty. A novel method for the calculation of shaded irradiance on each cell of an array with high computational efficiency is presented. The shading sub-model is validated against outdoor measurements with a modelling accuracy within one percent. Final verification of the over-arching SolaSIM framework found that it satisfied the requirements which were identified and actioned. The author installed the new CREST outdoor measurement system version 4 (COMS4). COMS4 is a calibrated system which measures 26 PV devices simultaneously. Validation of SolaSIM models against COMS4 found the modelling error to be within the 4% accuracy target except two sub-systems which had electronic faults. The model is validated against PV systems and found to be within the specified limits.