american-public-power-association-419672-unsplash-1024x682

WHAT IS PREDICTIVE MAINTENANCE?

Predictive Maintenance strategy is emerging nowadays as an essential field to keep high technical and economic performances of solar PV plants over time by scheduling maintenance through statistical tools.

WHY IS PREDICTIVE MAINTENANCE NEEDED?

Similar to medical care, prevention has proven to be much more cost effective than reactive approaches. Analyses of panel quality can reveal early signs of problems such as solar cell potential induced degradation or PID, which can lead to power losses of up to 50% and may be difficult to spot at very early stages within traditional operations and maintenance schedules.

Costs for the common risks of solar panels versus the cost of preventive methods can be displayed in the graph below. Green colour stands for the range of predictive costs and the red colour stands for the distribution of failures cost such as potentially induced degradation, internal corrosion and snail trail contamination. It is clearly seen that normally the risks outweigh the cost of predictive maintenance while the worst-case scenarios exceed it greatly.

High customization costs, the need of collecting a great number of physical variables and of a stable Internet connection on field generally limit solar panel effectiveness, especially for farms in remote places with unreliable communication infrastructures. The lack of a predictive component in the maintenance strategy is also a hindrance to minimize downtime costs.

HOW TO PREDICT MAINTENANCE?

In order to enhance and improve the system performance, preventive maintenance of solar power plants is done by implementing Operation & Maintenance (O&M) activities using predictive analytics and supervisory control and data acquisition (SCADA). With the help of the internet (cloud) based EA-PSM software, every function needed can be performed under one roof.

Since keeping the implementation costs and model complexity on a low level is important, EA-PSM uses emerging statistical methods based on Data Mining which are an effective approach both for fault prediction and maintenance scheduling.

CONCLUSION

Predictive maintenance enables PV system operators to move from a traditional reactive maintenance activity towards a proactive maintenance strategy, improving decision-making process thanks to a complete information on the incoming failure before the faults occur and reduce OPEX. The small cost of predictive maintenance can prevent expenses 10 times its size.

Subscribe to our Newsletter

Receive the latest news of Digital Twins and IT solutions for energy efficiency

Paieška

Search