In recent years, CO₂ flooding has emerged as an efficient method for improving oil recovery. It also has the advantage of storing CO₂ underground. As one of the promising types of CO₂ enhanced oil recovery (CO₂-EOR), CO₂ water-alternating-gas injection (CO₂-WAG) can suppress CO₂ fingering and early breakthrough problems that occur during oil recovery by CO₂ flooding. However, the evaluation of CO₂-WAG is strongly dependent on the injection parameters, which in turn renders numerical simulations computationally expensive. So, in this work, machine learning is used to help predict how well CO₂-WAG will work when different injection parameters are used. A total of 216 models were built by using CMG numerical simulation software to represent CO₂-WAG development scenarios of various injection parameters where 70% of them were used as training sets and 30% as testing sets. A random forest regression algorithm was used to predict CO₂-WAG performance in terms of oil production, CO₂ storage amount, and CO₂ storage efficiency. The CO₂-WAG period, CO₂ injection rate, and water–gas ratio were chosen as the three main characteristics of injection parameters. The prediction results showed that the predicted value of the test set was very close to the true value. The average absolute prediction deviations of cumulative oil production, CO₂ storage amount, and CO₂ storage efficiency were 1.10%, 3.04%, and 2.24%, respectively. Furthermore, it only takes about 10 s to predict the results of all 216 scenarios by using machine learning methods, while the CMG simulation method spends about 108 min. It demonstrated that the proposed machine-learning method can rapidly predict CO₂-WAG performance with high accuracy and high computational efficiency under conditions of various injection parameters. This work gives more insights into the optimization of the injection parameters for CO₂-EOR.

Summary: CO2 huff ‘n’ puff is a promising enhanced oil recovery (EOR) technique for tight/shale reservoirs, also enabling CO2 geological storage. However, the effectiveness of this method can be significantly affected by the aqueous phase resulting from connate water and hydraulic fracturing. The mechanism underlying the influence of the aqueous phase on oil recovery during CO2 huff ‘n’ puff, as well as the corresponding stimulation methods in such scenarios, remain unclear and warrant further study. To investigate this, we utilized a nuclear magnetic resonance (NMR) instrument to track the movement of fluids during CO2 huff ‘n’ puff under water invasion conditions. The impact of the invaded aqueous phase on oil recovery was examined, and the impact of different treatment parameters was explored. The results show that the aqueous barrier formed by water invasion alters the pathway of CO2 diffusion to matrix oil. This alteration leads to a diminished concentration of CO2 in the oil phase, which, in turn, results in a substantial reduction in oil recovery. Consequently, the performance of CO2 huff ‘n’ puff is highly sensitive to the water phase. Nevertheless, the oil recovery dynamics in cyclic CO2 huff ‘n’ puff under water invasion exhibit distinctive patterns compared with those without water invasion. These differences manifest as notable low oil recovery in the first cycle, followed by a rapid increase in the second cycle. This behavior primarily arises from the expulsion of a significant portion of the invaded water from the macropores after the first cycle. However, the effectiveness of this mechanism is limited in micropores due to the challenging displacement of trapped water in such pores. Raising the injection pressure mainly boosts oil recovery in macropores, with minimal response in micropores. Yet, the achievement of miscibility does not lead to a substantial improvement in the CO2 huff ‘n’ puff performance, primarily due to the constraints imposed by the limited CO2 dissolution through molecular diffusion Additionally, we have proposed three stimulation mechanisms achieved by lengthening the soaking time under water invasion conditions. First, the prolonged soaking time increases the concentration of CO2 molecules that diffuse into the matrix oil. Second, it promotes the imbibition of the trapped water on the fracture surface into the deeper matrix to alleviate water blockage. Finally, the invaded water in macropores displaces oil in micropores by capillary force during the soaking period.

Natural gas hydrates (NGHs) are regarded as a new energy resource with great potential and wide application prospects due to their tremendous reserves and low CO₂ emission. Permeability, which governs the fluid flow and transport through hydrate-bearing sediments (HBSs), directly affects the fluid production from hydrate deposits. Therefore, permeability models play a significant role in the prediction and optimization of gas production from NGH reservoirs via numerical simulators. To quantitatively analyze and predict the long-term gas production performance of hydrate deposits under distinct hydrate phase behavior and saturation, it is essential to well-establish the permeability model, which can accurately capture the characteristics of permeability change during production. Recently, a wide variety of permeability models for single-phase fluid flowing sediment have been established. They typically consider the influences of hydrate saturation, hydrate pore habits, sediment pore structure, and other related factors on the hydraulic properties of hydrate sediments. However, the choice of permeability prediction models leads to substantially different predictions of gas production in numerical modeling. In this work, the most available and widely used permeability models proposed by researchers worldwide were firstly reviewed in detail. We divide them into four categories, namely the classical permeability models, reservoir simulator used models, modified permeability models, and novel permeability models, based on their theoretical basis and derivation method. In addition, the advantages and limitations of each model were discussed with suggestions provided. Finally, the challenges existing in the current research were discussed and the potential future investigation directions were proposed. This review can provide insightful guidance for understanding the modeling of fluid flow in HBSs and can be useful for developing more advanced models for accurately predicting the permeability change during hydrate resources exploitation.

A single paragraph of about 200 words maximum. For research articles, abstracts should give a pertinent overview of the work. We strongly encourage authors to use the following style of structured abstracts, but without headings: (1) Background: Place the question addressed in a broad context and highlight the purpose of the study; (2) Methods: briefly describe the main methods or treatments applied; (3) Results: summarize the article’s main findings; (4) Conclusions: indicate the main conclusions or interpretations. The abstract should be an objective representation of the article and it must not contain results that are not presented and substantiated in the main text and should not exaggerate the main conclusions.