Joint development of "Dam optimal operation system"that uses AI

Hokuriku Electric Power Company

Hokuriku Electric Power Company and JFE Engineering Corporation have advanced the development of “Dam optimal operation system”.
It was confirmed that this system can forecast the inflow into dams with a high　accuracy and that further increase of hydropower generation could be expected.
We make effort to increase hydropower generation additionally by expanding use of this system to other dams.

1 Outline of verification test

Testing site: Asaida Dam of Hokuriku Electric Power Company located upstream of Jinzu-gawa river system in Takayama-City, Gifu
The system used for forecasting inflow: AI engine WinmuSe🄬 that JFE Engineering Corporation developed uniquely
We have let AI learn past rainfall and inflow related to Asaida Dam and have worked to improve forecast accuracy of inflow into the dam while taking into forecast know-how of Hokuriku Electric Power Company.
It was confirmed by the test that the forecast accuracy of inflow into Asaida Dam greatly improved.
It was also confirmed that optimization of dam operation enables to recover the dam water level earlier by eliminating wasted discharge and to decrease wasted discharge and electricity^※¹ by about 5GWh per year.

※1：electricity not used for power generation during flood

2 Future Expansion

We aim to maximize hydropower generation by expanding use of “Dam optimal operation system” to other dams in Jinzu-gawa river system.
We make effort to increase hydropower generation greatly that does not emit CO2 by expanding use of this system to other dams in other river systems and further upgrading of this system.