Risk Analysis of Well Completion Systems

Abstract

Summary The techniques of risk analysis have been applied to alternative offshore completions for a tension leg platform (TLP) to determine the probability of a leak of well fluids to the environment. For each completion type, logic diagrams were drawn, failure rates were assigned to the input events, and the reliability for production and workover was calculated. The sources of these data are described and the importance of proper ranking of equipment failure rates is discussed. The results of the analysis have determined that reliability of a well completion system during workover significantly affects the lifetime reliability of the well. Application of the risk-analysis techniques has played an informative role in selecting the equipment arrangement providing the safest well completion for a TLP. Introduction Central to the evaluation of new well completion techniques is the question of risk: How is the safety of offshore completion, production. workover, and recompletion affected by different equipment and practices? Without a history of events on which to base estimates of the systems safety, selection of the most reliable completion is subject to conjecture. An alternative to guesswork on the safety of the entire system would be to reduce a novel well completion to its component parts, which have an operational or testing history, to permit consistent assigning of reliability values. The safety of the system then can be estimated by proper combination of the component reliabilities by a fault tree or logic diagram method. This paper describes how reliability analysis was applied to alternative offshore completions for a TLP to determine their relative safety. For each completion type, logic diagrams were drawn with the aid of an interactive graphics computer program created for this analysis. The logic diagram represents the logical relationship of the well completion equipment necessary to prevent leaks of production fluid to the environment. The four completion types are as follows. Case 1 is a conventional offshore platform completion as might be found in the Gulf of Mexico or the North Sea. This system includes a tubing retrievable surface controlled subsurface safety valve (SCSSV). Case 2 is a TLP completion identical to Case 1 except for an additional anchor seal and assembly, a tubing hanger, a seating nipple, and a riser. Case 3 is a TLP completion like Case 2 with two additional side pocket mandrels, a surface controlled subsurface annulus valve (SCSAV), and a wireline retrievable SCSSV. Case 4 is a subsea tree completion. The master valve and annulus access valve are on the sea bed. A tubing retrievable SCSSV is included in this completion. Before any safety analysis is initiated, the event to be investigated must be determined. For the analysis performed here, this event is no leak of well fluids from the well system to the environment during production and workover. "Well system" means the production and workover system upstream of, and including, the Christmas tree (Xtree). The combination of events that prevent the desired event have been limited to those brought about by improper installation or mechanical failure. Events occurring outside the well system have not been included because they would be common to all TLP completions. In the final analysis, system safety must be a relative measure. This is because decision-makers may be unaccustomed to judging reliability expressed as an absolute number, but are comfortable comparing the reliability of a familiar system with an unfamiliar one. In addition, a relative measure of system safety tends to be more accurate than an absolute measure, as long as comparable consistent methods of analysis are used. JPT P. 713^