Prognostic Methods for Predicting Remaining Useful Life of Nuclear Plant Equipment and Components

Although nuclear power plants have provided a reliable source of power for the U.S. over the last four decades, unexpected equipment failures have resulted in losses of power production and millions of dollars of revenue, which may have been avoided if the warning signs of impending failures had been recognized. Likewise, utilities have spent millions of dollars and lost countless days of power production replacing equipment and components that showed no signs of degradation, yet had reached the end of the manufacturers specified design life. These types of problems and resource inefficiencies have prompted the nuclear industry to explore alternatives to traditional time-based maintenance practices in favor of condition-based maintenance strategies that employ advanced surveillance, diagnostic, and prognostic techniques. The area of prognostics, which concerns methodologies for predicting the remaining useful life (RUL) of equipment and components, has only recently been given consideration by research organizations and other experts as a promising technology for the nuclear industry. The work for this Phase II project proposes to expand the research in prognostic technologies and demonstrate its application in nuclear power plants. This goal will be accomplished in this project by developing and validating practical prognostics tools in terms of algorithms and software products for RUL estimation of selected nuclear power plant equipment. The Phase I successfully established the feasibility of using prognostics for nuclear plant equipment and components through hands-on laboratory work, prototype software development, and industry research. Nuclear plant applications were also identified that would be amenable to the use of prognostics. Additionally, partnerships were formed with a number of nuclear utilities to allow for implementation and demonstration of these technologies in a Phase II effort. This R & amp;D project will employ a comprehensive, hands-on approach required to move prognostics in the nuclear industry from feasibility towards full commercialization. This project will have four primary goals to include: 1) identify critical components and systems in nuclear power plants that are most immediately amenable to the use of prognostics, 2) develop software modules that integrate a variety of prognostic methods for estimating system RUL, 3) validate the prognostic methods with plant and/or laboratory data, and 4) demonstrate the application of prognostics in the asset lifecycle management practices currently employed in the nuclear power industry. Commercial Applications and Other Benefits: The product of this work has immediate application in all operating nuclear plants both in the U.S. and worldwide. Products that can accurately determine the RUL of equipment and components would be of immediate benefit for predictive maintenance in all industrial processes.