Time-to-Failure Estimation of Barrier Systems in Permanently Plugged and Abandoned Wells

Abstract

Abstract With the increasing number of aging fields and wells worldwide, a large number of wells will have to permanently plugged and abandoned in the coming decades. Today’s technical solutions on P&A design are primarily driven by legislations or recognized standards such as NORSOK D-10 or the Oil & Gas UK Well Decommissioning Guidelines. The NORSOK D-010 say that the well should be sealing to “eternity” without providing any link between the recommended solution and time-to-failure. During the last few years, there has been a drive towards a risk-based approach to P&A design. With such an approach, the goodness of a P&A design can be formulated in terms of the associated leakage risk, which consists of two components; i) the time-to-failure of the barrier system and ii) the leakage rate given the barrier system has failed. When failure data are available, there is a wide range of statistical tools available for establishing the time-to-failure probability distribution. However, barrier failure data on permanently plugged and abandoned wells are scarce in the North Sea region. In order to estimate the time-to-failure for wells in the North Sea region, all relevant information should be taken into account; survival data, expert input and physiochemical degradation models. In this paper, we will show how this can be accomplished by means of a Bayesian reliability approach. The paper will first describe the general framework for how to perform Bayesian time-to-failure estimation. Thereafter, information pertaining to barrier system lifetime for wells on the Norwegian Continental Shelf (NCS) and relevant assumptions will be discussed. Finally, the methodology will be applied on a synthetic case.