Challenge
A solar farm operator was seeing a reduction in yield across their 15 solar farm assets along with an increase in operations and maintenance spend.
Given the remote nature of the assets the client wanted to better leverage existing data to optimize operations but they struggled with where to start
Approach
We performed a rapid 4 week diagnosis using over 3 years of historical production data from across the solar farm assets and identified multiple areas of impact:
- Automated shading detection - analysis of hourly power production allowed identification of modules suffering from shading effects cause by overgrown vegetation.
- Identifying trackers needing calibration - analysis of tracker and power production data allowed continuous identification of tracker modules needing recalibrating
- Optimizing cleaning frequency - analysis of historical weather and pollen data together with power production enabled optimal cleaning frequency per site to be defined
- Detecting Inverter failure - continuous monitoring enabled identification of at-risk inverters based on failure signal modelling
We then worked with the operations & maintenance team to design and implement digital tools to capture the identified value at stake. This involved changing ways of working and re-thinking workflows, such as weekly maintenance team planning
Impact
Implementing solutions to optimize the maintenance and operations activity allowed the client to boost yield at underperforming sites by 3-5%. This was made possible by improving prioritization of existing maintenance resources without any incremental OPEX.