The Investigation of an Efficient and Effective Proactive Pipeline Integrity System

Abstract

In this paper, we propose the investigation of an efficient and effective proactive pipeline integrity system that uses multi-sensors for monitoring and inspection of the pipeline. Although the focus is on the pipeline, the methodologies will also be applicable for other systems including other energy infrastructure elements. The investigation focuses on the integration of multi-sources of data as well as multi-types of data for the same location (section) of the pipeline. The data include imaging, acoustic signals, electromagnetic flux system, and others. Moreover, the data could be obtained using continuous monitoring of the pipeline sections or through time intervals. In both cases, the data may also include censored observations. We develop several approaches to filter the data for noise and identify outliers, integrate the data streams into one degradation path and determine the optimum time to maintain or replace the pipeline sections being monitored. It will utilize extensive mathematical modeling for condition-based maintenance and repair, which will be developed by the investigators of this proposal. Following the suggestions of one of the reviewers of the last year's submission, we plan to validate and modify the model using standard or artificially developed defects such as corrosion, cracks, stress cracking, etc. to understand the output of the proposed methodology and validate the models accordingly.