Improving Fracture Conductivity by Developing and Optimizing Channels within the Fracture Geometry: CFD Study

Abstract

Abstract A conventional proppant pack can lose up to 99% of its conductivity due to gel damage, fines migration, multiphase flow, and non-Darcy flow. Consequently, pillar fracturing was developed to generate highly conductive paths for hydrocarbon flow. This paper describes experimental results and numerical models of a new method of generating stable proppant pillars. The proposed treatment method depends on fingering phenomena observed when a less-viscous fluid, which does not carry proppant, is injected to displace a more-viscous fluid that carries proppant. The low-viscosity fluid channels through the high-viscosity fluid and creates isolated proppant pillars. This method promises to reduce proppant costs, pumping horsepower, and gel damage, when compared to conventional treatments. A computational fluid dynamics (CFD) model using commercial CFD software was constructed to simulate the fluid flow inside a full-scale fracture dimensions. The objective of this study was to further evaluate the treatment design parameters during the generation of stable pillar-propped fractures. This study focused on gravity effects on the created channels' characteristics. The study also performed detailed investigation of the channel pattern as a function of treatment design (injection rate (1 to 120 bpm), and pulse stage time (5 seconds to 5 minutes), viscosity ratio (2 to 200)), and fracture geometry (width and height). Finally, horizontal proppant covering efficiency was calculated. Numerical modeling results confirmed that the gravity effect can be minimized either by increasing the injected fluid viscosity or by increasing the fluid injection rate. A strong linear relation was found between the minimum required viscosity to eliminate gravity effect and the density ratio between the two injected fluids. As an example, for a fracture width of 0.2 in. and injection rate of 4 bpm, the minimum viscosity needed to eliminate the gravity effect was 50 cP, 150 cP and 300 cP for density ratios (ρR) of 1.05, 1.25 and 1.5, respectively. The optimum channel pattern has small channel sizes, remain opened under closure stress, more channels throughout the entire fracture area and good communication between unpropped areas. Increasing fracture height or width during injection shifted the created channel pattern toward an optimum shape. Reducing the injection rate and/or stage pulse time moved the created channel pattern toward an optimum shape. A new dimensionless term, Dimensionless Stage Volume (VSD), is presented to describe the channel pattern inside the fracture. Smaller the VSD number resulted in the smaller and more distributed channels. Therefore, it is highly recommended to select and design a proppant pillar fracture treatment to achieve the lowest VSD possible and create the optimal channel pattern. For each viscosity ratio, the horizontal proppant coverage efficiency was independent of the VSD. However, a power-law relationship was defined between the horizontal proppant coverage efficiency and viscosity ratio between the two injected fluids. Finally, data and results obtained in this paper can be used as guideline to design, develop and optimize the channel fracturing treatment.