New types of sensors, artificial intelligence and big data analysis are forms of technologies that are now applied to hydropower. Sira-Kvina own and operates multiple hydro units, totaling 1760 MW of installed capacity that on average generate 6.3 TWh annually. With most units installed in the 1960s and 1970s,the need for maintenance is increasing, and so is the value of optimal timing of maintenance. Maintenance carried out too early is expensive and inefficient while maintenance carried out too late is potentially catastrophic.
Given enough sensors, data, and smart enough algorithms, it is tempting to believe artificial intelligence and data science alone will give reliable answers when a unit fails, or when recommending optimal timing for maintenance. However, this assumption severely underestimates the complexity of condition monitoring and maintenance planning for hydro units.