Sensitivity Analysis of Diffusion-Based Hydrocarbon Gas Huff-n-Puff Injection in Eagle Ford Shale

Abstract

Abstract The utilization of hydrocarbon gas in enhanced oil recovery (EOR) processes offers two significant advantages: an increase in the recovery factor and a reduction in net emissions. Through core-scale experimental and numerical investigations, effective diffusion coefficients for single-phase and cross-phase behavior were determined by Fu et al. (2021), enabling their application in larger-scale predictions. [1] The primary objectives of this study are to 1) better understand the impact of upscaling from core-scale to field-scale simulations; 2) verify the effect of diffusion mechanism during huff-n-puff by history matching a model for a single well pilot; and 3) conduct a comprehensive sensitivity analysis and optimization of the recovery factor for huff-n-puff schedule, taking into account fracture spacing and injection-production patterns in both the dead and live oil windows of the Eagle Ford formation. The fluids in place in the Eagle Ford shale show a wide range of GORs, with hydrocarbon maturities ranging from black oil to lean gas condensates, [2] therefore, both live and dead oil regions are investigated in this study. Two compositional models, incorporating dual porosity and dual permeability characteristics, were constructed using the Petrel software. The first model replicated a huff-n-puff field pilot study reported by Orozco et al. (2020) in the Eagle Ford [3] and consisted of one well with the well length of 6,240 ft and 26 hydraulic fracture stages. The second model encompassed a single stage of eleven horizontal wells, designed according to the field blueprint reported by Baldwin et al. (2020). [4] Within this model, six wells were allocated for injection and production during the huff-n-puff cycles, four were used as containment wells, and one functioned as a monitoring well at the center of all eleven wells. The well spacing was set at 1000 ft, with the first stage of each well measuring 220 ft in length, and each well containing 10 hydraulic fractures. These fractures were spaced 20 ft apart (cluster spacing), with a height of 100 ft, and a half-length of 500 ft. Once the pilot well's primary and huff-n-puff oil production rate was history matched, the same reservoir properties, including matrix and natural-fracture porosity, permeability, natural fracture spacing, and relative permeability, were applied to the eleven-well model. Both models employed history-matched effective diffusion coefficients and a tuned equation of state fluid model to fluid samples collected and analyzed for the Eagle Ford formation. [5, 1] Results show that models including the diffusion mechanism had a 2.2% higher oil recovery factor compared to those that did not include diffusion after five cycles of huff-n-puff. The sensitivity analysis on hydraulic fracture spacing showed that smaller fracture spacing creates larger contact surface area between the matrix and fracture, promoting the diffusion mechanism and facilitating higher oil recoveries. The sensitivity analysis also revealed that depletion level on the producer before starting Huff n Puff also had an impact on recovery efficiency. Producing a well on primary production for 6 years and then implementing huff-n-puff yielded the most oil cumulative produced. If the huff-n-puff cycle was delayed to 10 years after initial production, cumulative values were lower than at the 6-year mark due to depletion effects and difficulties in re-pressurizing the formation. The sensitivity analysis on the "puff" production period suggested that longer production times delayed the speed of oil production, but resulted in higher oil production after completing six cycles of huff-n-puff. Further sensitivity analysis on the length of the soaking period suggested that longer soaking times delayed oil production and did not contribute significantly to oil production. These parameters’ effects on cumulative oil production and reservoir pressure were analyzed to determine the optimal approach for field application. Investigations on using different injection gases such as CO2, y-grade, and lean gas for dead oil and live oil systems rank the best injectants for maximizing oil production in the following order: y-grade > CO2 ≈ hydrocarbon gas > lean gas. The findings of this study provide a deeper understanding of upscaling considerations and offer recommendations for huff-n-puff pilot designs in the Eagle Ford formation.