A KNN BASED COLLAPSE METHODOLOGY AND RECENT QUALIFICATION OF FLEXIBLE PIPES IN DEEPWATER APPLICATION

摘要

As the flexible pipe industry targets more on deepwater applications, collapse performance of flexible pipes becomes a key challenge due to the huge hydrostatic pressure during installation and service. The collapse strength of flexible pipes largely depends on the structural characteristics of carcass, pressure sheath and pressure armor layers. Therefore, the collapse prediction methodology involving a sound modeling of these layers is essential. Over the years, Baker Hughes have collected a large amount of collapse testing data. The prediction tool needs to be validated and calibrated against all the collapse tests for best accuracy. In this paper, the latest progress of the collapse prediction methodology and qualification tests are presented. A generalized collapse model was developed to predict the collapse pressure of flexible pipes. This model incorporates the advantages of both the weighted kNN regression technique and an analytical collapse model. It is able to reproduce the exact collapse pressure on the pipes tested and can predict the collapse pressure of other pipe designs not tested. As part of the qualification process, the capacity to prevent collapse must be demonstrated. Several flexible pipes were designed based on this generalized prediction methodology for deep water application, and pipe samples were manufactured using industrial production facilities for collapse tests. The results show that flexible pipes following current design guidelines are suitable for deepwater applications.