Well Integrity Flow Detection Using Novel Acoustic Pattern Recognition Algorithms

Abstract This paper discusses a new technology that helps to identify and differentiate multiple downhole well integrity events - tubing leaks, casing leaks, flow behind casing and overburden integrity, using unique algorithmic pattern recognition capability based on acoustic measurements from Distributed Acoustic Sensing (DAS) data. Here, we demonstrate application of this technology to identify flow behind casing and effectively build remediation strategies for Plug and Abandonment (P&A) of one of the wells in a BP operated field. Applying DAS for tubing leak detection has been discussed in prior publications. The approach has however, relied on simple data processing methods ‘listening’ for zones of increased noise when wells are shut in. In this paper, we discuss results from the application of a novel real time pattern recognition based approach to processing data from DAS that: Denoises data to decouple fluid movement ‘signals’ from background noise Uses signature recognition to constructs fluid ‘flow logs’ across the entire length of the well bore every second; Displays evolution of the fluid noise through depth and time The technology has been deployed across several assets for identifying well integrity events through different stages of well life – starting from drilling, production through to abandonment. In this paper, we demonstrate the application of the new capability for identifying flow behind multiple casing strings on one of the BP operated wells. The system was deployed to identify and pattern the evolution of fluid movement noise behind multiple casing strings through time and depth which was then used in conjunction with a tailored operational workflow to inform effective remediation strategies for plug and abandonment of the well. The fluid flow logs generated were used to conclude on the source of flow, flow propagation and pathway and facilitated an informed decision on extending the well’s first lateral barrier into the formation and adjustment of the position of the planned second barrier. The results showed the effectiveness of the capability in the field and demonstrated its unparalleled value for the operator, allowing for well abandonment. The technology offers a new method of acoustic data processing on DAS that extracts valuable and often hidden insights to identify the source of fluid flow and flow pathways, providing an ability of capturing events behind multiple casing strings – all fundamental challenges that the industry has been looking to solve for several decades.