Abstract
Within the H2020 project TRUST-PV [1], the TRUST-PV Risk Matrix has been created with the aim to standardize and classify maintenance tickets during operation and maintenance (O&M) of PV plants. The categorization of O&M tickets is based on the affected component, subcomponent, and event/failure in the field and by assigning accordingly a failure ID. Once an individual ticket is standardized, the cost priority number (CPN) methodology, initially developed in the previous H2020 project Solar Bankability [2] and further refined in TRUST-PV, is used for the calculation of the economic impact of maintenance tickets and the consequent prioritization of the ticket. The CPN, given in €/kWp, estimates the cost of an event by considering downtime cost due to performance loss and fixing cost due to repair. Hereby, downtime costs are calculated based on a method created by Lindig et al. [3], which estimates the power loss with an optimized machine learning model. For an accurate assessment of the fixing cost of an event or failure, this work continues with the evaluation of individual maintenance tickets of O&M providers in order to get information about the proposed action to take. Only by knowing the action to take, or solution, to a specific issue in the field, the fixing cost due to repair can be properly determined. The final CPN is the sum of downtime costs and fixing costs, and acts as a metric for the developed approach which will be integrated into an automated decision support system (DSS) platform [4] for field technicians to optimize the operational phase of PV plants. Thus, for purposes of financial impact calculation of fixing technical risks, the core of this work is to extend the TRUST-PV Risk Matrix by a Solution Matrix. This matrix suggests one or several actions to take, so-called solutions, for all events in the risk matrix. Then, for all solutions, the cost contributions are defined based on the analysis of real maintenance tickets, and integrated with literature research, market analysis and data provided by O&M companies. The first step is to extract the time to fix the failure from maintenance tickets, , define the number of personnel required, , and their hourly tariff, . Multiplying these parameters gives us the labour costs, . The detection costs, instead, are usually defined per kW nominal capacity, since detection methods are usually applied to the whole plant or at least a bigger part of the plant. Thus, the Solution Matrix will just include a suggestion for a detection or inspection method. A further component of the matrix are the repair costs, , including all material and replacement costs necessary for the repair of the issue. Finally, transport costs, , are listed in the solution matrix, even if they only apply if replacement components or materials are needed and no spare part management is in place. All cost parameters mentioned above are country, plant, contract dependent and need to be adjusted and updated continuously.