Dynamic Modeling and Design Optimization of Cyclonic Autonomous Inflow Control Devices

Abstract

Autonomous inflow control devices (AICDs) have recently been introduced in the petroleum industry to restrict the production of unwanted fluids, namely water and gas, much more effectively than conventional inflow control devices (ICDs). As with ICDs, AICDs are installed downhole along the completion string to first delay water/gas coning and then restrict their influx, without well intervention, if/when coning such occurs. Unlike ICDs, AICDs selectively choke back water and gas significantly more so than oil. A novel cyclonic AICD was recently developed using computational fluid dynamics (CFD) driven design optimization. The cyclonic AICD's unique internal geometry increases the flow resistance to unwanted fluids based on how their viscosities and densities differ from oil, as initially predicted using CFD and subsequently validated by extensive, carefully controlled single- and two-phase flow tests. The resulting excellent match obtained between CFD and such laboratory tests yielded accurate mathematical models for predicting flow performance over a broad range of flow rates and oil, water and gas properties. The flow performance models were then incorporated into a state-of-the-art dynamic reservoir simulator with multi-segmented wellbore capability to compare the production performance over time for the same well but completed with no ICDs, conventional ICDs, and cyclonic AICDs. A synthetic but realistic three- dimensional (3-D) reservoir model has used that allowed oil, gas and water production. Multiple sensitivity runs were initially performed to optimize the number of compartments using packers for annular isolation, and the number of ICDs per compartment. Once these parameters were optimized, only the ICD type was varied for performance comparison. The results of this systematic, multi-step process, as presented herein, demonstrate that the cyclonic AICD adds significant value to the improvement of oil production by controlling unwanted fluids, such as water and gas, and by preserving the reservoir energy.

In-Depth Understanding of ICD Completion Technology Working Principle

“…These results were found to have a comparable trend to other numerical experiments reported in the literature. For example, a reduction of 15% (Gurses et al, 2019) or 90% (Eltaher et al, 2014) has been reported for gas production. Alternatively, with water, there have been reported reductions of 82% [21] or 12% [13].…”

“…Additionally, the model was optimized after evaluating several parameters such as the number of inlets, the slope of the inner funnel chamber, and even the installation position. Chochua et al [15] and Gurses et al [16] proposed an optimized diode AICD design with CFD. The hydrodynamic profile was characterized for oil, gas, and water and even tested for different oils, and the performance curve was obtained by numerical means.…”

Application of mathematical and machine learning models to predict differential pressure of autonomous downhole inflow control devices

“…After breakthrough, AICDs significantly reduce unwanted fluid production (e. g., water and gas), which increases oil production and also reduces costs of separation and re-injection of unwanted fluids produced along with oil to the surface (Aakre et al, 2013;Andreas et al, 2019;Lei et al, 2020). Cochua et al (2018) and Gurses et al (2019) introduced a cyclone type AICD which is similar to Equiflow AICD and performs based on the viscosity and density of flowing fluids. Zhang et al (2019) designed a new Cyclone Inflow Control Device using the principle of the cyclone to deal with fine particles like sand.…”

A semi-implicit approach for the modeling of wells with inflow control completions

“…In recent years both ICDs and ICVs have gained popularity and are being applied to a wide range of fields. Their efficiency has been studied through a variety of field cases (see, e.g., Al-Enezi et al, 2012;Al-Khelaiwi and Davies 2007;Al-Thuwaini et al, 2009;Broni-Bediako et al, 2019;Das and Al-Enezi, 2014;Dimitios et al, 2009;Gurses et al, 2019;Javid et al, 2018;Raffn et al, 2007).…”