News & Views, Volume 44 | Data Driven Solutions for the Most Difficult Problems
By: Andrew Crompton and Mark Jaeger
In recent years, SI has observed an increasing trend in the use of specialty instrumentation to solve “impossible” problems or answer “indecipherable” questions. This shift was particularly apparent within commercial nuclear, where data-driven solutions have long been perceived as challenging due to short outage windows, personnel dose concerns, and a significant paperwork burden, among other factors. Widespread adoption of instrumentation-based solutions creates new paths to tackling difficult/persistent problems, and shifts the industry focus for critical assets from reactionary to more of a predictive approach. In 2017, SI assisted numerous clients with deployment of specialty instrumentation in this fashion, comprising two general scenarios: 1) new designs/modifications, and 2) repeat failures. Each application requires different sensors and varying analytical methods, but the approach used to leverage the resultant data to solve the problem is generically applicable throughout the energy sector. The text below details important considerations for both scenarios and highlights a successful application of the underlying process for management of thermal fatigue in reactor coolant system branch piping.