Influence of cracks and microcracks on flow and storage capacities of gas shales at core-level

Advisors

Anthony R. Kovscek, primary advisor
Khalid Aziz, co-advisor
Louis M. Castanier, co-advisor

Abstract

Economical shale gas production is only possible via hydraulic fracturing in which a fluid is injected in the subsurface at a pressure large enough to penetrate the near wellbore region. Hydraulic fracturing activates subsets of cracks and microcracks that facilitate transfer of gas to exit the matrix. This process creates stress sensitive flow dynamics between various media with their respective flow paths. This complex interplay of transport in media with different size scales poses challenges for modeling. Understanding the mechanisms of flow as the reservoir fluid pressure increases during fracturing or decreases during flow back is key for quantifying the production forecast and the volume of reserves. Gas transport in shales is believed to be driven by complex physical mechanisms including Darcy flow, Knudsen diffusion, and adsorption through fractures, microcracks, and matrix. Understanding the interplay of gas transport in these media at increased pore pressure and constant net effective stress, as well as at increased net effective stress (i.e. during depletion) is the goal of this work. This research work investigates the interplay of gas transport in the cracks, microcracks, and matrix of shale using He and CO₂ via transient upstream pressure pulse decay experiments. The cracks and microcracks are natural in the sense no lab pretreatment created them. A novel setup of the pressure pulse decay experiment was used to determine the storage and flow capacities simultaneously. Experimentally, the pressure signals are used to define time-dependent pore volume partitioning between the microcracks and the matrix. A dual continuum simulator is constructed to decouple the flow and storage capacity at early-time pressure where the gas flows simultaneously in both media. This is done via a history-matching process that quantifies the pore volume partitioning of the microcracks and matrix, the permeability in the microcracks, and the diffusivity of gas in the matrix. A series of experiments were conducted at constant net effective stress (500 psi) and increasing net effective stress (500, 1000, 2000 psi) to study the evolution of the pore volume partitioning, the permeability in the microcracks, and the diffusion in the matrix. The experiments were conducted on samples from the Eagle Ford and Haynesville shale plays. Results have shown that the pore volume partitioning in the microcracks in sample 180Ha (Eagle Ford) was between 8 - 24.1% of the total pore volume in all experiments at constant and variable net effective stresses. Sample TWG 1-3 (Haynesville) had a percentage of microcracks between 84.5 - 87.0% of the total pore volume. Sample TWG 3-3 had a microcracks portion between 68.8 - 83.1% of the total pore volume. These proportions were found to be related to the magnitude of permeability and void volumes in the system. The greater the permeability and the smaller the void volumes, the smaller is the microcracks pore volume. Samples 180Ha, TWG 1-3, TWG 3-3 had a liquid permeability of 396.3, 1.63, and 16.59 micro Darcy. The gas transport in the matrix was expressed volumetrically via a "diffusional transport group" parameter. This history-match parameter was found to increase generally with pore pressure at constant net effective stress and decrease with increased net effective stress. The role of adsorption was also investigated using CO₂ in the same experimental apparatus and conditions. It was found that the influence of pore volume partitioning is suppressed by the large adsorption capacity in the shale samples. The adsorption capacity was a history-match parameter in the analysis. It was found that adsorption occurs in the microcracks in permeabilities less than 50 micro Darcy along with adsorption in the matrix. For permeabilities greater than that value, adsorption had no importance on microcracks. Adsorption in both media required different adsorption-pressure functions. The CO₂ permeability was found to be smaller than the He permeability by a 2-3 factors.