Abstract
Creep damage in low alloy steel plates welded using carbon steel filler can be of concern for petroleum and petrochemical industry due to differential thermal strain and creep behavior. An aggravated scenario can occur when a flaw is present at the weld. The current work considers a catalytic reactor vessel in steady-state operation whose design life is consumed by 80%. It is assumed that after few years of initial operation a low alloy steel patch (similar to the vessel plate) was welded to the vessel using carbon steel filler. A crack like flaw is assumed to be present currently and was sized using non-destructive testing. However, uncertainty remains regarding the time when the crack first appeared, which is often a representative case for a vessel in operation for many years.
The objective of the current study is to, by following API 579-1/ASME FFS-1 code procedures, assess the creep damage at the end of the design life considering the postulated scenario as described above. Two cases are considered. For the first case, the currently observed flaw was assumed to be present as is since the beginning of time when the welded patch was made. No flaw growth is assumed for this case from initial time till present. This represents a conservative upper bound case. For the second case, it is assumed that a smaller initial flaw grew to the current size during normal operation. The initial flaw size was determined iteratively such that after growth it matches the currently determined flaw size. For the ease of calculations, the time from initiation of the flaw to the present was discretized into three time intervals during which it is assumed that the flaw size remains constant for the time interval. This removes some of the conservatism inherent in the first case. For both the cases, additional accumulated creep damage is determined considering crack growth. A Finite Element Analysis (FEA) is performed for the reactor vessel patch consisting of a single crack at the weld-patch interface to assess the accumulated creep damage from normal operation till the end of design life. Due to inherent uncertainties in the parameters, sensitivity studies on creep damage due to Adjustment Factors were also performed, based on which Adjustment Factors for Creep Strain Rate (material scatter) and Creep Ductility were chosen appropriately. Steady-state crack growth is considered for creep damage using analytical approach.
The current work shows a practical approach combining FEA and analytical calculations to determine accumulated creep damage where a crack appears to have initiated sometime during the past normal operation.
