Objectives

The overall objective of this project is to improve an existing cloud-based product designed to optimise the O&M strategies and asset management of PV power plants. In particular, the proposed project is predominantly progressing further the field of PV operational data quality and automatic failure and performance loss diagnosis and prognosis. This will be achieved by the online (i.e. real-time) identification and classification of failures (open- and short-circuit, bypass diode and inverter failures, potential induced degradation, etc.) and trend based performance losses (soiling, degradation, etc.) in PV power plants. As such, improvements in the PV plant availability, performance ratio (PR therefore, energy yield), O&M costs and hence, LCOE will be achieved.

The scientific and technological objectives described by the main challenges of the project and the associated key performance indicators (KPIs) are to:

1. Formulate data quality algorithms for inspecting and treating missing and erroneous data achieving data availability > 95%.


2. Develop and verify data-driven and statistical algorithms for real-time failure and performance loss diagnosis and prognosis achieving failure and performance loss detection and classification accuracies > 98.5% and > 92% (under clear-sky conditions), respectively.


3. Combine the developed algorithms in order to achieve PV plant availability > 96% and a 4% relative increase in PR (and energy yield). The PR improvement is also aligned with the associated target given by the SET-Plan; i.e. annual PR ≥ 82% (for small commercial plants) and 87% for larger plants installed after 2020.


4. Implement and incorporate the developed algorithms into a product (TRL6) for optimum PV asset management taking into account financial aspects, which will enable optimum hardware replacement/maintenance, cleaning schedules and methods, etc., achieving a relative reduction of up to 10% in O&M costs.


5. Perform an economic analysis and evaluate the impact on LCOE by benchmarking the up to 6.5% relative reduction in LCOE (assuming average O&M costs of 2%, CAPEX 1 €/W, average energy yield 1300 kWh/kWp and linear degradation rate of 0.8%/year) due to the combined improvements in performance ratio (O3) and O&M costs (O4).

The commercial challenge of this project is to enhance and improve the TRL of the cloud-based product of Alectris by incorporating data quality algorithms, an optimised methodology for the estimation of soiling and appropriate algorithms for automatic and uninterrupted monitoring enabling predictive and preventive maintenance of PV power plants. As such, the early identification and classification of failures and performance loss mechanisms will be achieved and the PV plant availability, intervention, response and resolution times will be improved. Therefore, actions will be able to be taken by the corresponding asset owners or operators/contractors in order to safeguard the PV performance and minimise the investment risks and LCOE.