We present a model for well inflow control devices (ICDs) that includes the effects of an annulus in which the flow between the ICDs is open or partially obstructed by the presence of packers, and we describe the application of this model in a full-field simulator. Flow in an open or partially obstructed annulus requires looped flowpaths to be modeled within the well. We describe the extension to the formulation of the well model together with considerations to ensure that the resulting equations have a Jacobian matrix that is invertible and explain the necessary modifications to the techniques used to solve the resulting linear system. The effect of simulating flow in both the annulus and the tubing was investigated in two case studies involving ICDs: a synthetic case and a sector of a North Sea field model. Results showing significant differences between the inflow profiles of horizontal wells with and without packers in the annulus are presented. Introduction Advanced well completion solutions are becoming increasingly common in both onshore and offshore hydrocarbon reservoirs. Two main advanced well installations are inflow control devices (ICDs) and the flow control valves (FCVs). An ICD is a screen which passively regulates inflow so that high-velocity flow regions are choked back, resulting in a more uniform inflow profile along the well. The screen acts as a flowpath between the annulus and the tubing; flow from the reservoir enters the annulus and passes through the screen into the tubing. An FCV allows active and remote control of inflow or outflow in different zones along the wellbore or in the individual branches of multibranch wells. To optimize the design and operation of wells with these installations, their behavior must be represented in reservoir simulation tools. Simulation tools that focus purely on the near-well region and wellbore flow can be used to design this type of advanced well. In these tools the well models can be very sophisticated, taking account of flow in the annulus for example, but their capability to model reservoir flow is often limited. Full-field reservoir simulators allow a much more detailed representation of the reservoir flow, and some of these simulators contain powerful well modeling tools such as multisegment wells (Holmes et al. 1998), which allow the representation of multilateral topology, the presence of inflow control devices, multiphase flow, wellbore storage, cross flow and friction effects. For an example of the multisegment well model being used to represent ICDs in field case reservoir simulations see Henriksen et al. (2006). A former restriction of the multisegment well model was that it could not represent looped flowpaths within a well. Thus, while flow in an annulus could be modeled, the well segment topology had to be such that there was only one flowpath from each section of the annulus into the well tubing (Fig. 1a). This restriction limited the usefulness of the multisegment well model when investigating wells without packers in the annulus, or the effects of different packer installations, or leakage through packers. In this work we model devices requiring looped flowpaths by extending the multisegment well model in a full-field simulator (Schlumberger 2008a, 2008b). In the previous formulation of the model, each segment could have only one outlet. This restriction meant that only device models with a "gathering tree" topology could be represented, and loops were not permitted. This paper describes how the formulation has been extended, effectively to allow any number of outlets from a segment, thus enabling loops to be incorporated in the ICD model. We apply the extended model to investigate the effect of simulating flow in the annulus in two case studies involving horizontal wells with ICDs: a synthetic case and a sector of a North Sea field model. We compare cases with no flow in the annulus, flow in an open annulus (Fig. 1b), and flow in an annulus with packers at locations that isolate sands of different qualities.

Open hole gravel packing of reservoir sections drilled with oil based fluid is traditionally performed with an aqueous carrier fluid. This typically involves displacing oil based fluid to aqueous fluid once the gravel pack screen is in place. In reservoirs with swelling or unstable shale this approach reduces the risk associated with open hole exposure to aqueous fluid over time. However experience has shown that instability can still occur resulting in an incomplete, or even an aborted gravel pack. In addition, mixing of incompatible oil and water based fluids downhole has the potential to generate very viscous emulsions that negatively impact gravel pack efficiency and well productivity. The objective of the new technology was to maintain borehole stability, eliminate fluid incompatibility and enable a complete gravel pack.An oil based carrier fluid has been developed and qualified using laboratory and yard scale testing. The fluid is a solids free invert emulsion that exhibits near Newtonian rheological behavior; thereby promoting settling of proppant during gravel packing. The density of the fluid is controlled by adjusting the volume fraction and density of the brine phase. The fluid has been qualified up to a density of 1.25 SG with further potential to achieve a density of 1.63 SG.The oil based carrier fluid has been introduced on a mature field, with a long history of gravel pack completions. Progressive reservoir depletion has created operational challenges, resulting in inconsistent gravel pack performance. Consequently, several procedural changes have been implemented over time. Reservoir inclination is typically up to 50°, open hole length up to 200 meters, and bottom hole static temperature around 90°C. Gravel packs were most recently performed with a 1.10 SG aqueous carrier fluid.The new carrier fluid has exhibited stable properties during implementation on multiple well completions. Gravel pack efficiencies have been consistently good, at 100% or higher. As a result, well productivity expectations have consistently been achieved or exceeded. The operational time for installing the lower completion compares well with the traditional approach with aqueous fluid.The implementation of oil based gravel packs in multiple wells, allows a comparison with brine based gravel packs in the same field. It is therefore considered to be an industry first.

The production experience from the Statfjord Field on the Norwegian Continental Shelf is one of the greatest adventures in modern oil and gas history. After achieving very high oil recovery factor using a predominant drainage strategy with pressure maintenance by water and gas injection, the drainage strategy in the field has since 2007/2008 been changed to reservoir depressurization. Prior to depressurization start-up, the field has produced about 652 million Sm3 (4.1 billion bbl) oil and 187 billion Sm3 gas. Currently, the field is producing at an oil rate of approximately 5 300 Sm3/d and a gas rate of about 11 million Sm3/d. Estimates indicate that successful implementation of the new drainage strategy will continue and lead to an ultimate oil recovery of higher than 67% and a significant additional gas production, as a result of the depressurization process. In addition, the field life will be extended from 2009 to 2025, and this will contribute to lifetime extension of the attached satellite fields. The main purpose of this paper is to provide a description of the multidisciplinary approach used for evaluation and planning of the Statfjord Late Life (SFLL) with reservoir depressurization, share learnings from depressurization start-up and address challenges, uncertainties and opportunities.