Sensors aim to promote power plant safety, energy savings

Clemson University researchers are developing several new sensors designed to withstand two harsh environments, the intense heat inside power plants and the tremendous pressure at the bottom of hydraulic fracturing wells.

The research, led by Professor in Electrical Computer Engineering Hai Xiao, aims at helping advance the technology behind fossil fuels. The researchers expect the sensors to help plants generate power more efficiently while lowering emissions and using fewer resources.

Reducing dependence on fossil fuels has long been a goal of researchers and policy makers. About 63 percent of energy generation in the United States last year came from fossil fuels, including coal, natural gas, petroleum, and other gases, according to the U.S. Energy Information Administration.

The researchers said that the two power-plant sensors under development could help detect equipment failures, with one sensor monitoring boiler tubes and the other sensor monitoring turbine blades. They are designed to help utilities reduce maintenance costs and could be ready to test in real-world power plants within three years, according to Xiao. 

The well sensor would be all digital and would monitor pressure and be less expensive than current technology. The well sensor would eliminate the need for electronics that go into the well, as crews extract natural gas from hard-to-reach deposits, Xiao said.

A significant part of the research, funded largely by the U.S. Department of Energy’s Office of Fossil Energy, is conducted by a group of collaborators Xiao began assembling when he arrived at Clemson six years ago, he said. The group has grown to about seven faculty members and calls itself the Clemson University Center for Intelligent Systems for Extreme Environments (CU-ISEE).

“We’re now going back to the mode where we look at what will happen in the next five years,” Xiao said.  “We’re talking about artificial intelligence, machine learning and optimization. We’re talking about how we build intelligence into the machine and our research.”