Iberdrola’s automatic replenishment algorithm (ARA) analyses data from sensors and smart meters which has helped improve reliability, safety and cost of network operations. The average outage duration per customer was 44 minutes in 2018 compared with an average of 1.14 hours in the rest of Spain. Iberdrola also become 1.8 times faster at network fault recovery in 2018 due to its AI-based ARA tool, automatically detecting and isolating the smallest area of the network to speed up recovery, and reducing operating costs by 12%.
In Spain, thanks to the STAR Project (Remote Network Management and Automation System), with an investment of €2 billion, 10.8 million smart meters have been installed and provide remote management, supervision and automation capabilities to 9.000 transformation centers. As a result, an accurate and flexible source of information can eliminate the need for onsite meter reading or manual bill validation, as well as wasteful paperwork.
SAIDI (called TIEPI in Spain). See here the improvements from last years, being 44 minutes in 2018, and the average in Spain is 1,143 hours in 2018, see the information from ministry of industry and energy in Spain: https://sedeaplicaciones.minetur.gob.es/eee/indiceCalidad/total.aspx, so Iberdrola is 25% better than the rest of the sector in Spain. For clarification, a DSO with a SAIDI of more than 60minutes means that reliability is still high (>99%).
Digitalisation in Renewable Asset Management to improve efficiency
Iberdrola has digitalised its entire renewables fleet and is applying AI to increase the accuracy of forecasting and improve wind plant performance, reducing component failure and increasing reliability. A single dispatch centre centralises operations and enables cloud-based predictive maintenance in nine European countries, with one each in the US, UK and Brazil. It uses a cloud-based Advanced System of Predictive Analysis which detects faults in the main components of wind assets in their early stages and suggests corrective actions.
The MeteoFlow system uses meteorological simulations to optimise renewables facilities. These meteorological simulations generate about 2 Terabytes per day of three-dimensional grid data using machine learning techniques. AI is also used to detect anomalies in PV solar plants, to forecast hydroelectric power production and to predict natural disasters.
Read more about this project