Huff and puff process optimization in micro scale by coupling laboratory experiment and numerical simulation

Abstract

Huff and Puff Enhanced Oil Recovery method can be regarded as promising process to increase oil production rates from developed field. Worldwide experiences in the application for an industrial-scale of this technology has been extensively discussed for heavy oil and tight oil production, however, field unique does not guarantee success for technology transfer to different site. In this way reservoir simulation is used as a first approximation of the project efficiency. However, numerical simulation requires representative data from laboratory experiments. Furthermore, huff-and-puff should be considered as complex problem, where influences from injection rates, soaking time and production rates can not be neglected. On the other side, conducting laboratory investigations are expensive and time-consuming, therefore, these researches should provide the most valuable information. In the presented methodology, laboratory experiments were conjuncted with the numerical representation of a core sample, to generate trustworthy models which were used for the process optimization. The optimal huff-n-puff operational design was computed using a stochastic population-based particle swarm optimization (PSO) method. As a consequence of high computational cost of a single full physic numerical run, the genetic programming as a novel tool for the huff-and-puff process optimization was successfully implemented. The comparison of the optimized results between genetic programming data-drive model and the full-physic numerical run revealed the right approximation and significant computing time reduction.

Introduction

Different enhanced oil recovery (EOR) methods are widely applied to increase oil production rates. Definitively, miscible gas injection can be considered as the most promising in the scope of efficiency. CO₂ flooding has shown favorable recovery potential from laboratory investigations and field applications [1], [2]. Contrasting with conventional water flooding, CO₂ injection provided greater results, due to its favorable injectivity and possibility to reach the miscible displacement condition after exceeding minimum miscible pressure (MMP) during injection stage [3]. Main physical mechanisms for improving oil recovery by CO₂ injection include oil swelling, oil viscosity reduction, the formation of multiple contact, and miscible mixed [4]. However, a huge demand for the injected gas, limits its extensive application for the continuous flooding [5]. Thus, Huff-and-Puff CO₂ can be auspicious method for the small oil field or located out of the main infrastructures. In the Huff-and-Puff scenario, the producers are converted temporarily to CO₂ injectors to stimulate the surrounding well area after a shut-in period for soaking, later time, these wells are again reverted to the producers [6], [7], [8]. Recently, the CO₂ huff-n-puff process has received more attention in the literature, because it is considered as one of the most effective treatments for the oil field development. Chen and Gu investigated CO₂ huff-and-puff injection process for the shale oil reservoirs by utilizing numerical modeling in terms of developed process parameters [9]. They found that short injection and production periods were favorable, no matter what primary depletion period was used, because they increased the overall huff-and-puff cycle numbers and took advantage of the high initial injection and production rates during the huff-and-puff stage. Kanfar et al. investigated the feasibility of EOR in tight oil reservoir using cycling solvent injection with carbon dioxide as solvent [10]. Optimal process pattern was developed by genetic algorithm in term of maximizing project profitability. Optimized parameters included numbers of cycles, duration of injection, soaking and production as well as start time of huff-n-puff process. As a result, significant improvements of recovery factor were observed. Alfarge et al. took into account the effect of diffusion mechanism on huff-n-puff process efficiency [11]. In their work, the series of numerical experiments were conducted on dual-porosity simulation model of Bakken formation with and without molecular diffusion assumed. The parameters affecting molecular diffusion mechanism were analyzed. Based on the obtained results applied to match the full field case study production, the best fitting was achieved for the single porosity model with relatively low diffusivity coefficient. This means that CO₂ diffusion in reservoir was extremely slow or kinetics of oil recovery process in producing areas overcame the CO₂ diffusivity. Recovery factor for low permeability oil & gas reservoirs can be significantly improved by hydraulic fracturing. Zuloaga-Molero et al. provided the new insights into the understanding of the impacts of induced fracture geometry, reservoir permeability, and natural fractures on the performance of CO₂-EOR processes in tight oil reservoirs [12]. Near linear relationship between fracture area and increment of recovery factor can be observed. Two permeability case studies (0.01 and 0.1 mD) were conducted and results showed the improvement for huff-n-puff process for the lower permeability due to the reduced effect of gravity and viscous forces. Zuloaga et al. investigated diverse fracture aligned pattern where zipper fracture over-performed others [3]. In range 0.001 to 0.01 mD CO₂Huff-n-Puff scenario performed better than CO₂ flooding scenario. Increasing contact of fracture area should be also considered as major key point to improve process efficiency. Zhang et al. investigated the effect of capillary pressure fluid properties in nanopore confinement on huff-n-puff process by combining laboratory experiment with numerical simulation [13]. The complex study revealed the positive influence of capillary pressure and molecular diffusion on huff-n-puff overall efficiency. Yu et al. states that most important parameters in CO₂ huff-and-puff are CO₂ injection rate, injection time, and number of cycles [14]. Some laboratory attempts have been carried out to optimize operational parameters including soaking time and production pressure in order to maximize the recovery factor for CO₂ huff-and-puff processes [15]. Monger et al. tested huff-and-puff method in the aspect of CO₂ impurity, cycle numbers and working pressure on the oil recovery factor from highly permeable core sample [16]. Abedini and Torabi performed the core flooding experiments under miscible and immiscible conditions, concluding that pressure increase did not considerably promote oil recovery [17]. Ma et al. researched 2.3 mD rock samples denoting usefulness for the enhanced oil production. Some papers investigated technological factors for the huff-n-puff overall efficiency. Yu et al. discussed the role of soaking time and pressure drop. A ”soak” period is crucial to recover oil effectively, but an extra longer soaking time has no effect on the improvement of recovery factor (RF) [18], [19]. Sheng state that injection time should last so long until the pressure near the wellbore reach the maximum value. Furthermore, the number of huff-and-puff cycles should be determined when an economic rate cut-off is reached [20]. Wei et al. performed a series of laboratory experiment of huff-n-puff process on Jimsar sag reservoir core samples to develop an improved process scenario and production dynamics [21]. The interactions between CO₂ and the crude oil were also examined. Process performance was carried out on low permeability core sample (0.6 mD) under reservoir conditions. Besides of recovery factor, efficiency of CO₂ sequestration in tight formation was assessed. Laboratory experiment provided detailed description of influence of CO₂ injection timing, slug size, injection rates and cyclic schemes on cyclic gas injection process efficiency.

In the presented paper, effect of injection rate, soaking time as well as production rate on the huff-and-puff process was investigated in term of maximizing recovery factor by coupling laboratory measurements with numerical simulation. Core flooding test for the CO₂-oil binary system was carried out to study the relative permeability. Thereafter, the right huff-and-puff experiments were carried out. Obtained data were used to create and calibrate numerical representation of core sample. The numerical simulation data set allowed to assess the process efficiency. Applying the genetic programming enabled to develop a data-driven model correlations between investigated variables and total oil recovery from the rock sample. Based on a created response surface, particle swarm optimization was involved to determine the best production and injection strategy. Finally, obtained results were compared each other to full simulation run, revealing the right solution quality of response surface. To best author’s knowledge, the presented approach is the novel one in terms of coupling numerical simulation of micro scale laboratory experiment with state-of-the-art optimization method.