This paper describes a computational model, together with preliminary results, relating to the investigation of computational modelling strategies appropriate for the prediction of chalk instability and static liquefaction in hydrocarbon applications. The computational model incorporates a number of advanced procedures including; rate independent and rate dependent constitutive models, a finite element formulation that is able to represent the large strain associated with the liquefied state, automatic adaptive remeshing procedures to resolve the gross changes in geometry subsequent to liquefaction and a fully coupled procedure for the mechanical and porous flow equation systems. The model is calibrated using both rate independent and rate dependent experimental tests and then applied to high loading rate experiments which were specifically designed to initiate chalk liquefaction.
1. INTRODUCTION
A constitutive model that is able to represent the rate dependent stress vs. strain response of chalk, including static liquefaction.
A finite element formulation that is able to represent the large strain associated with the liquefied state.
An automatic adaptive remeshing procedure to resolve the gross changes in geometry subsequent to liquefaction.
A fully coupled procedure for the mechanical and porous flow equation systems.
Wellbore instability leading to chalk production is a relatively common occurrence in chalk reservoirs. These instabilities are occasionally manifested by massive bursts of chalk production which are potentially due to chalk liquefaction [1]. The produced volume of solid particles may range from small volume fractions to rapid production of cubic meters of chalk. This behaviour has been observed in the Valhall field in the Norwegian sector of the North Sea, where the chalk reservoir includes regions of extremely high porosity chalk (ca. 50%). In order to mitigate the potential damage caused by massive chalk production one of several completion systems (e.g. gravel packs, expandable sand screens, cased and perforated completions), are often deployed as part of the well management plan. These completion systems are expensive and a fuller understanding of the conditions leading to significant chalk production, and hence a more accurate quantification of the risk, has the potential to have a significant impact on the well cost and production efficiency. In this paper a computational model for the investigation of chalk instability is presented which is able to represent both the onset of liquefaction and the post-liquefaction state. The model incorporates a number of advanced procedures including:The rate dependent non-associated constitutive model, derived by extending the ideas of other researchers [ 2, 3, 4], represents both the short-term high strain rate and long-term creep of chalk. The method of characterizing chalk materials using this model is outlined and the performance of the model is compared against experimental evidence from rate dependent hydrostatic creep tests and undrained triaxial compression tests with creep for Liege chalk (Flatebø, 2005). The model is then applied to two experiments [1], specifically designed to investigate liquefaction of chalk due to high loading rates. It is shown that when the loading rate is sufficiently high liquefaction, resulting in rapid extrusion of chalk, is correctly predicted.
