Modification of Dent-Weld Interaction Factors and Dent Corrosion-Gouge Discrimination Methodology Incorporated in API RP 1183

Abstract The presence of coincident features like welds, corrosion or gouges within a dent can affect dent fatigue life. API Recommended Practice (RP 1183) incorporates several different approaches to consider the impact of weld, corrosion and gouges on dent integrity. This paper discusses further development and modifications proposed to the dent-weld interaction methodology and dent corrosion-gouge discrimination approach presently included in API RP 1183. A fatigue life reduction factor of 10 is incorporated in API RP 1183 when a dent is interacting with a weld. This reduction factor was developed based on dent fatigue full scale tests where the weld was positioned at the critical location within the dent, from a cyclic fatigue point of view. The fatigue life reduction factor of 10 is very conservative, especially when a weld is located a distance away from the dent peak. The presence of a dent in a pipe results in a large stress range gradient as the distance increases from the dent peak. Therefore, the effect of the weld on the dent fatigue life should decrease as the distance between the weld and the dent peak increases. This paper discusses the results of a project carried out on behalf of Pipeline Research Council International (PRCI), where the stress range gradients within a dent were utilized to modify the dent weld fatigue life reduction factor for single peak dents. Over 1,000,000 dent finite element simulations were evaluated, and their corresponding stress range gradients captured to modify the reduction factor as a function of the location of weld relative to the dent peak. Based on this study a range of fatigue life reduction factors, ranging between 3 and 10, have been proposed depending upon the location of the weld with respect to the dent peak, the dent geometry, the dent restraint condition and the pipe geometry. The corrosion-gouge discrimination approach described in Annex H of API RP 1183, Figure H2 was developed to determine the type of metal loss (i.e., corrosion or gouge) coincident with a dent. The Annex H approach utilizes four queries regarding the dent geometry, metal loss geometry and position, and pressure condition during an in-line inspection (ILI) run to make a determination. The Annex H approach was evaluated in the current project using real-world data provided by pipeline operators, which consisted of field inspection results and ILI feature listings of dents coincident with metal loss features. The supplied ILI data was evaluated using the Annex H approach to determine whether each metal loss feature was corrosion or gouge-like. Work was also carried out to implement a logistic regression model to discriminate between a gouge and a corrosion feature, which shows promise. The logistic regression model incorporated pipe outer diameter (OD), wall thickness (t), metal loss clock position, axial and circumferential distances between the dent peak and metal loss, metal loss aspect ratio and dent strain as input variables. The Annex H approach and the logistic regression model were tested and optimized using in ditch non-destructive evaluation (NDE) results and then implemented using ILI data. This paper presents the results based on the ILI data and the results confirmed against the field inspection reported observations.