News Release

Intelligent Computing Could Lower Price of Power

Peer-Reviewed Publication

Penn State

University Park, Pa. -- "Smart" computer chips and other new, advanced, "intelligent" control approaches have the potential to make both nuclear and fossil fuel power plants last longer, operate more efficiently and supply electricity at lower costs, say Penn State nuclear engineers.

Dr. Robert M. Edwards, associate professor of nuclear engineering; Dr. Kwang Y. Lee, professor of electrical engineering; and Dr. Daniel E. Hughes, nuclear engineer, recently demonstrated the new techniques' potential, for the first time, using Penn State's Breazeale nuclear reactor. The demonstration and the use of the Penn State research reactor as a simulation for a commercial power plant are detailed in their paper, "Testbed for Nuclear Plant Instrumentation and Control Validation."

Edwards says, in the demonstration, a new intelligent control concept developed by the Penn State group was applied to estimating the temperature of the reactor fuel and coolant. He explained that it is very difficult to measure reactor fuel temperature in commercial power plants. Using the group's new advanced computer programs as controllers, the group accurately estimated the temperatures and improved the plant's performance.

Edwards adds, "The reactor fuel is designed to operate within certain temperature limits. If those limits are exceeded, it can shorten the lifetime of the fuel. Being able to keep the temperature within bounds, lengthens the life of the fuel, increases operating efficiency and saves money -- which could translate into lower costs for consumers."

While the new Penn State techniques focus mainly on improving performance, Edwards notes that they can enhance nuclear power plant safety as well. He explained that, now, during a potentially hazardous or "upset" situation, a commercial nuclear power plant's safety system can shut the plant down causing severe stress and strain on the system. With the aid of the new Penn State monitoring and control techniques, plant operators can prevent shutdowns by preventing fuel overheating, for example. The new control techniques can also allow operators to respond more quickly to potentially hazardous situations, or faults, and to develop routes around the problem.

Edwards says, "We demonstrated the fault-accommodating characteristic back in 1993 on an actual power producing reactor. The test case was a loss of steam supply. The automated decision making controls detected the problem immediately and enabled the system to adjust for it without the necessity of a shutdown."

The new techniques don't totally cede decision-making to the computer -- only the most basic level. By automating the lowest level of control, the control system can respond more quickly than humans could during the first few seconds of an "event". This immediate action "buys time" for the human controllers who can take more time selecting appropriate corrective action.

Edwards expects to complete development of automated controllers for nuclear reactor coolant flow rate by the end of the summer. Upset in coolant flow rates is one of the most frequent reasons for shutting a plant down.

Penn State's Intelligent Distributed Controls Research Laboratory, where the new techniques were developed, is the only center in the world demonstrating these advanced computer techniques on an operating nuclear reactor. Penn State was recently the site of the 1996 American Nuclear Society International Topical Meeting on Nuclear Plant Instrumentation Control and Human Machine Interface Technologies. Edwards, Lee and Hughes presented their paper at the meeting and also conducted workshops for plant engineers interested in applying the new techniques in their own facilities.

###

EDITORS:
Dr. Edwards is available at 814-865-0037 or via e-mail at rmenuc@engr.psu.edu; Dr. Lee at 814-865-2621, e-mail kyl@ecl.psu.edu; and Dr. Hughes at 814-865-6351 or e-mail dehnuc@engr.psu.edu



Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.