Fatigue Analysis Through Automated Cycle Counting Using ThermAND

Abstract

The potential for fatigue damage due to thermal transients is one of the degradation mechanisms that needs to be managed for plant components. The original design of CANDU® stations accounts for projected fatigue usage for specific components over a specified design lifetime. Fatigue design calculations were based on estimates of the number and severity of expected transients for 30 years operation at 80% power. Many CANDU plants are now approaching the end of their design lives and are being considered for extended operation. Industry practice is to have a comprehensive fatigue management program in place for extended operation beyond the original design life. A CANDU-specific framework for fatigue management has recently been developed to identify the options for implementation, and the critical components and locations requiring long-term fatigue monitoring. An essential element of fatigue monitoring is to identify, count and monitor the number of plant transients to ensure that the number assumed in the original design is not exceeded. The number and severity of actual CANDU station thermal transients at key locations in critical systems have been assessed using ThermAND™, AECL’s health monitor for systems and components, based on archived station operational data. The automated cycle counting has demonstrated that actual transients are generally less numerous than the quantity assumed in the design basis, and are almost always significantly less severe. This paper will discuss the methodology to adapt ThermAND for automated cycle counting of specific system transients, illustrate and test this capability for cycle-based fatigue monitoring using CANDU station data, report the results, and provide data for stress-based fatigue calculations.