FCTM Esope

S2.15 - Development of the API 579 Methodology for the Assessment of Piping Vibration Fatigue

7 oct. 2021 | 15:30 - 16:00

Studio Photo

Description

Piping vibration failures are not uncommon across upstream, midstream, petrochemical, refining, and liquefied natural gas (LNG) processing industries. The majority of higher temperature piping designs are often governed by thermal code compliance rules in the ASME B31 series, which pushes the designer to increase flexibility, leading to more vibration prone systems. There is clearly a need to improve design rules and best practices, but there are inconsistencies in how in-service piping is evaluated when vibration problems are detected. In-service piping vibration has long been considered a fitness-for-service problem and solved using brute force cycle counting methods combined with fatigue curves or screened using a variety of vibration acceptance curves. Traditional fatigue evaluations, such as cycle counting, relies on appropriate use of fatigue curves intended for very high cycle life predictions. Moreover, commonly used vibration screening curves often lack a well-documented technical basis so that the risk of failure cannot be quantified. To remedy these problems, an API 579-1/ASME FFS-1 (API 579) Fitness-for-Service method was recently proposed for the assessment of piping vibration fatigue. The Level 1 method consists of a single vibration acceptance curve unique to either mainline or cantilevered branch connections and provides clear direction so that either root-mean-square (RMS) or 0-to-peak velocity units can be used. Level 2 is based on the popular ASME OM-3 method with slight modification, and now includes guidance on use of stress concentration factors approximated from ASME B31J or ASME Section III Stress Indices. In line with the API 579, Level 3 provides guidance on use of numerical methods such as dynamic piping stress models or finite element analysis to more precisely calculate alternating stress ranges specific to the system in question. This paper summarizes the technical basis for the proposed Level 1, 2, and 3 methods as well as presents a variety of case studies which highlight its use.

Présentée par