The Algorithm Making Water More Efficient

Published on Friday, 28 July 2017

“We’ve named it KOOS (Kaplan Online Optimization System). Thanks to the algorithm we invented, we can significantly improve the efficiency of hydroelectric turbines.”

– Alessandro Quadrelli, Innovation Project Manager

A winning idea

In 2015, four colleagues (Roberto Suffredini, Stefano Sello, Alessandro Quadrelli and Matteo Masotti) came together to try to understand how to make the Kaplan turbines, which produce clean and renewable energy in our plants all around the world every day, more efficient.

After a lot of hard work, they came to a special algorithm able to monitor the activity of the turbines constantly and optimise the results of their work. The first test took place in 2016 at the Fabbriche hydroelectric plant, in the province of Lucca (Tuscany), with extremely positive results: the KOOS system improved the precise efficiency of the turbine up to several percentage points, avoiding useless waste of water and allowing a more responsible production of energy.

“We later decided to share this know-how with contractors building control systems for hydroelectric turbines, in order to have solutions that are immediately applicable to our plants.”

– Gianluca Gigliucci, Head of Innovation Hydro

An algorithm for all

Following the success of early experimentation, the KOOS algorithm was patented and shared with other large companies that can reap its benefits. The first supplier we started to collaborate with was the Brazilian company Reivax, which developed and tested our optimiser in the lab with excellent results. The algorithm will therefore be soon integrated into our plants at Sauzalito (Chile) and Cachoeira Dourada (Brazil) to verify its performance in the field.

“We are also working to integrate KOOS into the Riba-roja plant in Spain by the end of 2017, using other possible suppliers that have carried out automation of our plants, particularly in Europe”.”

– Alessandro Quadrelli

Once officially in operation, the KOOS project is designed to grow further. We will soon extend the algorithm to so-called “multi-group” hydroelectric plants, with the aim of improving the efficiency not only of individual turbines, but also of the whole system of machines present at the same plant, improving the overall generation efficiency, with even more promising results expected.