Vestas Uses Big Data for Better Wind Turbine Placement
By Aditi Bhat
With the help of IBM Big Data, Vestas is now capable of performing 150 Trillion calculations per second on the data gathered from 35,000 meteorological stations around the world. Such a magnitude of calculations has solved major problems faced by the company. It is considered a role-model for the renewables industry. Founded in 1945, Vestas Wind Systems A/S is a Danish Manufacturer and servicer of wind turbines. It is the largest wind turbine company in the world that operates in 6 continents and has about 21000 employees globally. The company has done ground-breaking work in making wind a competitive source of energy. Data Science played a crucial part in the success of this turbine-maker.
Growing the libraries from 18 to 24 Petabytes of data and gaining knowledge from that data
Reviving the $1.2 billion “Fosen-area-wind-turbines” project that was shelved
Placing wind-turbines at optimal locations to increase the ROI. Such optimal locations would reduce the irreversible wind-damage caused to high-end turbine-equipment
Expanding the Vestas wind libraries - the data management systems – by more than 10 folds for including a larger range of weather-data captured from 35,000 meteorological stations around the world
Reducing the feedback time (regarding potential optimal sites for placing wind turbines) from 3 weeks to just a few hours
Apache Hadoop chosen to power all the functional systems provided to Vestas by IBM
IBM have provided Vestas with IBM® InfoSphere® BigInsights software running on an IBM System x® iDataPlex® system that serves as the core infrastructure for analysing and managing weather data in innovative ways
Vestas engineers can now analyse wind-data-grids for 3X3 Kilometres area. Earlier, engineers could not analyse anything less than 27X27 Kilometres of grid area for finding potential locations for turbine placements
InfoSphere BigInsights system helps Vestas predict valuable insights that help their customers get increased ROI on their investments
The response time for wind forecasting is now 15 minutes instead of 3 weeks.
Wind turbines can now be placed with enhanced accuracy.
Considerable reduction in IT footprint and related costs.
Total energy consumption reduced by 40% while computational power of forecasting systems increased manifolds.