Practical Approach to Address Pitfalls of High-Resolution Numerical Simulation Modeling in Advanced Completion Design Optimization Today

Abstract

Application of downhole flow control device (DFCV) has proven to be a successful strategy to mitigate early water or gas breakthrough in many fields in Asia. A conventional numerical modeling workflow is often applied in such studies. However, the full potential from the downhole installation is often not explored due to several numerical oversights in failure to adapt to the unique operation requirements and DCFV hardware design representation. This paper presents a practical approach to address pitfalls involved in high-resolution numerical simulation DFCV design optimization. In this paper, a multi-stage procedure is highlighted involving a numerical simulation model prepared for DFCV design and optimization. The first step is to investigate the grid resolution from the 3D model. This is to ensure the effect of grid-to-well resolutions from coarse scale to finer scales to capture the device behavior along the open hole (OH) length of horizontal wells and to capture the gas and water influx from contact. The second step is to design and optimize the packer placement based on permeability contrast as primary reference using a practical approach by setting the number of packers as a sensitivity variable with uniform DFCV setting design. It is then followed by an unbiased design workflow is to optimize benefits of all kinds of DFCV such as nozzle-based and viscosity dependent inflow control devices of zonally varying setting or optimal configuration designs. The practical approach is demonstrated on a synthetic simulation model with a horizontal well to address the oversights in modelling prior to DFCV design and optimization process. Based on this work, vertical grid resolution to oil thickness ratio exceeding 1:32 amplified the differences in results due to numerical dispersion problem. For packer location optimization, several sensitivities of different packer placements and number of packers were performed to compare the oil cumulative incremental. The optimum number of packers with uniform DFCV design is 19 packers, however the oil gain will be decreased once the number of packers is reduced. Finally, the practicality of applying a global optimization algorithm to such studies during real-time operations was investigated. A unique practical approach is presented to address pitfalls involved in the needs of high-resolution numerical simulation in DFCV optimization. This approach captures the complex physics and resolution involved while ensuring the design loop efficient enough to perform fine-tuning during run-in-hole or on-the fly design onsite. In addition, this design optimization workflow is possible due to the availability of a new standard advanced reservoir simulator, that is cloud-compliant and has efficient multi-core parallel processing which otherwise would take days if not weeks conventionally to complete the task, deemed unsuitable for near real-time design or fine-tuning needs.