Equipment Health Monitoring and Damage Prediction Using Mechanical Stress Soft Sensing Through Data Analytics

Abstract

Fatigue damage due to structural stress is a common problem for equipment used in offshore drilling such subsea connectors and subsea riser joints. Fatigue damage is characterized by the weakening of a structure or component of the equipment due to cyclic loading. Consistently operating the equipment above the operational parameters can lead to premature failure of the equipment causing unplanned downtime and posing a safety risk to nearby workers. There are several methods currently being used to determine cumulative fatigue damage as a way of assessing the operational life of machines used for drilling. Linear cumulative fatigue damage analysis is one of the most used methods for life prediction of a structure and components of equipment subjected to cyclic loading. The model involves examining the operational stress ranges caused by cyclic loading and comparing them to an established fatigue curve to estimate the total utilization and predict equipment failure. However, the linear damage rule (Miner's rule) has several limitations namely: The damage model often depends on complex, and time consuming, stress analysis depicting exact geometry and operating conditions.Damage can only be assessed subsequently, making it difficult to forecast use and plan scheduled maintenance. This document presents the development of a mechanical stress soft sensing algorithm for determining real-time cumulative fatigue damage using finite element analysis with response surface methodology. The results in this document show that the new real-time cumulative damage determination approach could effectively help address the limitations of the current models by providing a means of determining real-time cumulative damage with little computational power.