With the Google Cloud-Westinghouse Electric partnership, AI can address these challenges and shorten delays in real time, said Scott Sidener, chief engineer for data, digital & artificial Intelligence at Westinghouse Electric.
He provided an example demonstrating how Westinghouse can use AI to leverage 75 years of the company’s nuclear data and reduce the time and cost of construction. He focused on how AI works to optimize construction work for the Westinghouse Electric AP1000.
For example, he shared a 3-D model of an air handling equipment room where AI optimizes 345 construction tasks.
AI estimated that the tasks in this room will cost $3.8 million and will take 160 days to complete, he said. AI takes into account supply chain disruptions, workforce disruptions and other issues such as the availability of specialized workers, including nuclear electricians.
A “day in the life of construction” shows how AI can address delays
As a “day in the life of construction” example, 14 construction tasks must be performed, but three are delayed or on hold. In the model, it’s possible to look at any of the tasks–pipe fitting or welding, for instance.
Without AI, it’s too late to respond to the disruptions, schedules get pushed back and workers–welders and electricians who are critical to getting power plants up and running–come in and sit all day.
In the example described by Sidener, with AI, Westinghouse Electric workers know which tasks are disrupted before they come into work. AI has optimized the best sequence of the 345 required tasks, given the delays, and identifies the tasks that crews can do that day that will yield the maximum impact on the cost and construction timeframe.
AI re-estimates the total cost based on these changes. It eliminates the delayed tasks and instead includes the most valuable tasks to perform that day, Sidener explained.
The human element is critical
Supervisors review these recommendations, and once they approve them, the schedules are automatically updated and communicated to workers.
“You can do this in seconds or minutes as opposed to waiting a week for a new schedule,” Sidener said. “I hope you can see a picture of how this can literally transform the overall construction process.”
Lou Martinez Sancho, chief technology officer and executive vice president of R&D and innovation at Westinghouse said, “We have technical teams working together with Westinghouse and Google Cloud to help us oversee the huge amounts of data we need to analyze and make sure the final decision is made by an engineer on the human side.”
Big gains in nuclear expected by 2030
With AI, by 2030, nuclear power will be positioned to provide reliable power at the gigawatt scale, Raiford said. Not only does AI help Westinghouse Electric improve construction timelines, but is also boosting grid efficiency, he noted.
According to the International Energy Agency, AI can improve how grid operators integrate renewables and reduce curtailment and emissions. It can identify faults, reducing the length of outages by 30% to 50%. With remote sensors, AI can boost the capability of transmission lines.
Taking advantage of such AI-enabled advances, Westinghouse Electric has plans to deploy its new AP 1000 reactors in the US and abroad.
The company plans to install three AP 1000 reactors in Poland and two in Bulgaria. Ukraine has contracted for the construction of nine of the reactors. In October, Westinghouse, along with joint owners Brookfield and Cameco, signed a partnership with the US Government to accelerate the deployment of nuclear power.
“I don’t just see AI as an optimization tool. It’s a catalyst in the innovation of operating nuclear plants,” Raiford said.
