PROJECT TITLE: Imaging-Based Diagnostics of Photovoltaic Arrays
Collaborators: DTU Photonics (DTU), Sky-watch (SW), Kenergy (KE), SiCON (SC) and Skive Municipality (SM).
Funding: Innovation Fund Denmark.
Photovoltaic systems (PV) are now a major electricity source in many countries, reaching 227 GWp installed cumulative capacity worldwide in 2015. As this market continues to grow PV system reliability and production predictability becomes even more significant, as any failures and downtime, especially in large solar plants leads to production loss and significant cost.
More than 2% all PV panels fail after 11-12 years of operation, well before their warranty expires. Considering the cost of the PV panels is more than 40% of the total cost of a PV plant, and that multi-MW scale PV plants are common nowadays, fast and accurate detection and localization of failures in PV panels has become increasingly critical.
Aim: This PhD project aims to develop diagnostic and failure identification methods for PV panels primarily based on electroluminescence (EL) and infrared (IR), enabling improved utilization of PV assets.
Project specific goals are:
- Research and development of the necessary EL and IR imaging techniques for PV panels applicable on drones during daytime and/or night-time with focus on modulating the electric current through the panels, aiming at improved failure detection potential.
- In-house/Experimental testing and correlation of different PV panel failures and degradation of PV panels with the acquired EL and IR image patterns.
- Developing diagnostic algorithms based on the different faults patterns acquired from EL and IR images for automatic fault identification and severity assessment.
- Modification of the developed fault detection algorithms with the adaptation of real practical scenario and environmental sustainability aspects.
- Integration and field testing of the EL/PL detection on the drone inspection system with the image analysis, PV panel localization, automatic fault detection.
Publications in journals and conference papers may be found at VBN.