A Measurement-Based Approach to Reliability-Target-Design Of Pressure Vessels, Piping, Pumps, and Valves in New Types of Powerplants Where Failure Probabilities Based on Historical Records Do Not Exist

Abstract The American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (BPVC) Committee has recently developed a new Section XI (Nuclear Components Inspection) Division 2 Code named “Reliability and Integrity Management (RIM).” RIM incorporates a new concept known as “System-Based Code (SBC)” originally due to Asada and his colleagues (2002), where an integrated approach to safety assessment from design to service inspection is introduced using three reliability-based statistical quantities: (1) “System Reliability Index,” or “System co-Reliability Target” for any system consisting of structures and components, (2) “Structural Failure Probability (FP),” or “Structural co-Reliability (coR),” for any structure in the system, (3) “Component Failure Probability (FP),” or “Component co-Reliability (coR),” for any component in the system. Implicit in the SBC concept is the requirement that items (2) and (3) listed above, i.e., the failure probabilities, or co-reliabilities, of all structures and components such as pressure vessels, piping, pumps, and valves, be available as inputs to a probability risk analysis (PRA), for risk assessment and regulatory compliance, whereby item (1) is estimated to meet the system co-reliability target. For new types of powerplants, for which historical failure data of structures and components do not exist, this requirement cannot be fulfilled because it is not possible to estimate items (2) and (3) before any structure or component exists and operates for a period. However, using laboratory test data of fatigue, fatigue crack growth, creep, and creep crack growth, it is possible to estimate an upper bound (UB) of item (2) and that of item (3), as shown recently by Fong, et al. (PVP2021-62169, PVP-2024-123443), so for design and regulatory compliance purposes, this upper bound approach is more than adequate. In this paper, we summarize the results of the failure probability upper bound (FPUB) approach by first describing the model, and then illustrate its application in two examples: Example 1. Failure Probability Upper Bound (FPUB) of an AISI 4340 steel pipe in fatigue and fracture at 20 °C with an operating stress amplitude of 282 MPa. Example 2. Failure Probability Upper Bound (FPUB) of a 2-1/4 Cr 1 Mo ferritic steel pipe in creep and fracture at 565 °C with a creep stress of 73 MPa. The significance and limitations of this failure probability modeling methodology are presented and discussed.