Using Reliability Modeling and Accelerated Life Testing to Formulate a Cost Model of Photovoltaic Systems with Different Architectures

As the number of photovoltaic (PV) systems grows annually and ages, the importance of long term reliability to the levelized cost of electricity (LCOE) has become increasingly clear. Likewise innovative PV system architectures being considered present opportunities for differentiation and growth into new markets, but still need to have a model for system reliability in order to justify their LCOE. Other benefits of reliability modeling include setting reliability goals for individual failure modes, determining optimum replacement and maintenance strategies, as well as prioritizing which failure modes require the most consideration in terms of research and testing. Physics of failure and materials properties are key inputs to the outcome of these models. This paper presents the results of reliability models based on accelerated testing, field data and theoretical calculations that were undertaken in order to determine the cost model of reliability of PV systems of different system architectures. Three examples for PV modules are presented here (bypass diodes, solder joints and microinverters) and are used to show how understanding of the reliability of systems based upon materials properties can be used to formulate models for lifetime and costs.