Wellhead Fatigue Analysis : Surface casing cement boundary condition for subsea wellhead fatigue analytical models

Abstract

Material fatigue is a failure mode that has been known to researchers and engineers since the 19th century. Catastrophic accidents have happened due to fatigue failures of structures, machinery and transport vehicles. The capsizing of the semisubmersible rig Alexander L. Kielland in Norwegians waters in 1980 killed 123 people, and investigations pointed at the fatigue failure of a weld as one of the direct causes. This accident led to a number of improvements to the design of offshore structures. The noticeable safety principle ”No single accident should lead to escalating consequences” has since been adopted in a widespread manner. Since 1992 the Petroleum Safety Authority in Norway has enforced a risk based safety regime. Wells are designed to hold back reservoir pressures and avoid uncontrolled escape of hydrocarbons. In other words a well is a pressure containing vessel. Norwegian safety regulations require a dual barrier construction of wells. This safety principle ensures that one “barrier” is preventing an escalating situation should the other barrier fail. A wellhead is a heavy walled pressure vessel placed at the top of the well. The wellhead is part of the second well barrier envelope during drilling. The subsea wellheads are located at sea bottom and during subsea drilling the Blow Out Preventer (BOP) is placed on top of the subsea wellhead. The drilling riser is the connection between the BOP and the floating drilling unit. Waves and current forces acting on the drilling riser and drilling unit will cause dynamic movement. Flexible joints at top and bottom of the drilling riser protects the drilling riser from localised bending moments.

The subsea wellhead is both a pressure vessel and a structurally load bearing component resisting external loads transmitted from a connected riser. These external loads can be static and cyclic combinations of bending and tension (compression). Cyclic loads will cause fatigue damage to the well. The well can take a certain amount of fatigue damage without failing. A fatigue failure of a WH system may have serious consequences. Should the WH structurally fail its pressure vessel function will be lost and for this reason WH fatigue is a potential threat to well integrity. The structural load bearing function will also be affected. Wellhead fatigue analysis can be used as a tool to estimate the accumulated fatigue damage. Analysis results then compares to a safe fatigue limit. This thesis addresses selected aspects of fatigue damage estimations of subsea wellheads and surface casings. The presented work is a contribution to the fatigue analysis methodology currently being developed within the industry. The well cement role as a boundary condition for surface casings in analytical models is particularly addressed. The majority of research focuses on the casing shoe and formation sealing, which is the primary objective of well cementing. Recent research focus on the cement limits conditions e.g. elevated temperatures. The “near-seabed” conditions of lead cements have seen less scrutiny. Some researchers have shown interest in this issue related to deep water cementing. Deep water bottom temperature is low all year round regardless of location latitude. Low sea water temperatures will depress the normal thermal gradient of the upper parts of the soil. Subsea wells are typically cemented using a lead and tail cement system, and the lead top casing cement will be pumped all the way to seabed. This lead cement will then be left curing in a low temperature environment. Hydration of cement is an exothermic chemical reaction, and the reaction rate is dependent on temperature. Laboratory measurements of low temperature early compressive strength of typical lead cement slurries are presented herein. In the North Sea the duration between placement of surface casing lead cement and installation of BOP/drilling riser will typical be around 24 hrs. Then dynamic riser loads will start acting on the upper part of a subsea well. Bending of the well causes relative motions between the conductor and surface casing. The cement around these casings will experience these relative motions. The combination of delayed cement setting due to low temperature and surface casing motions will cause localized failure of cement bonding in the upper part of the well. In subsea wellhead fatigue analysis finite element models are used. Boundary conditions in analytical models are important in ensuring similar behaviour of model and reality. One boundary condition in wellhead models is the lateral cement support of the surface casing. Modelling this cement support as infinitely stiff with a discrete vertical transition is the existing solution. In this work a modified boundary condition is presented based on low curing temperatures in combination with “premature” loading of the supporting cement. An overall analysis methodology approach has been suggested. Using a detailed local model of the well to define the lower boundary condition for the global riser load analytical model is one of its features. The implementation of a modified cement boundary condition will change the global stiffness of the local well model. The possible effect on global riser load from variations to the lower boundary condition has been studied. The conclusion supports the suggested analysis approach. Overall well ultimate structural strength will be reduced by the presence of a fatigue crack in a non pressurised load bearing part of a subsea well. An analysis methodology with case results are presented and indicate that the location of a fatigue crack affects the reduction in ultimate strength. Cases of significant reduction are expected to impact normal operating limitations. To be able to include the wellhead fatigue failure mode in an overall risk management system, the failure probability needs to be estimated. This can be done by applying a structural reliability analysis methodology to the problem. A suggested structural analysis methodology approach is suggested and notational failure probabilities are presented. Future improvements to wellhead fatigue analysis may emerge from calibrations from measurements of the reality. A comparison between analytical fatigue loading and measured fatigue loading has been presented and results indicate that the analysis results are conservative. This is evidence that analytical estimate on acceptable fatigue limits can be trusted from a safety point of view. It also indicates the monetary potential that measurements can present to the well.