Solar farm performance monitoring

The customer's solar plant had a number of failures causing production losses which led to missing monthly and annual targets.  Over a period of time, this amounted to a considerable cost to the company. Additionally, a lack of accurate planning of replacement parts inventory resulted in a high cost through excess safety stock driven by reactive maintenance processes.  The existing SCADA/PLC systems used to identify, detect and report the causes were very manual and time consuming, often providing notifications too late to avoid a loss of power event. It was also possible for IGBT & Inverter failures to go undetected for a period of time, thus impacting solar farm production targets for many days.  The customer wanted to automate the failure identification process and predict failures ahead of time to reduce the impact of failures on production and reduce spare parts inventory through better planning.

Arundo carried out an initial pilot to demonstrate that critical failures such as inverter failures and IGBT switch failures could be detected ahead of time. Arundo also demonstrated the notifications were consistently in enough time to avoid loss of power events.  The models also showed that significant efficiencies can be gained through advance notification of faults through correct classification of potential failures to improve the replacement parts inventory planning.