Search

Gas flooding is a promising way to enhance oil recovery in unconventional oil reservoirs, but accurate identification of the distributions of movable and residual oil during gas flooding is difficu...

Stable gas gravity drainage is considered an effective method to enhance oil recovery, especially suitable for deep buried, large dip angle, and thick oil reservoirs. The influence of reservoir heterogeneity on controlling the gas–oil interface and sweep characteristics of injected gas is particularly important to design reservoir development schemes. In this study, according to the interlayer characteristics of Donghe carboniferous oil reservoirs in the Tarim Basin, NW China, 2D visual physical models are established, in which the matrix permeability is 68.1 mD and average pore throat radius is 60 nm. Then, hydrocarbon gas gravity drainage simulation experiments are carried out systematically, and a high-speed camera is used to record the process of gas–oil flow and interface movement. In this experiment, the miscible zone of crude oil and hydrocarbon gas is observed for the first time. The interlayer has an obvious shielding influence, which can destroy the stability of the gas–oil interface and miscible zone, change the movement direction of the gas–oil interface, and reduce the final oil recovery after gravity drainage. The remaining oil mainly is distributed near the interlayers. The higher displacement pressure leads to increased stability of the gas–oil displacement front and later gas breakthrough, which leads to higher oil recovery. The lower gas injection rate contributes to a slower front velocity and wider miscible zone, which could delay gas breakthrough. For the immiscible gas gravity drainage, there is a critical gas injection rate, with which the oil recovery factor is the highest.

This paper provide improved phase behavior models, trying to mitigate the problem that phase behavior of gas-crude system is difficult to describe in L block with low permeability and high water cut in China. This situation leads to a series of problems in CO2 flooding process and lower recovery up to expectations. The models is evaluated to possess both high calculation speed and accuracy compared with existing others.Characteristics of CO2-crude systems had been considered into repulsion-attraction type EOS (equation of state) based on the analysis of repulsion parameter and attraction parameter in EOS, and the improved EOS had been applied in developing calculation method of MMP (minimum miscible pressure). No ideality of CO2-crude systems had been considered into mixing rules of CO2-crude systems based on analysis of mixing rules of repulsion-attraction type EOS. Promotion had also been put into the obtain methods of parameters in phase behavior, including density, viscosity, MMP, critical parameters of plus components etc. All these methods are applied in L Block.The phase behavior models of CO2-crudes system promoted in this paper mainly include EOS, mixing rule and viscosity model and have been applied in CO2 flooding process in T Reservoir. The relative error of density calculation is reduced from 7% ∼ 20% to less than 1%and the modified EOS is applied to predict the MMP of the CO2-crude systems from 8 different blocks in T reservoir. The modified EOS also works well for the relative error of MMP prediction is reduced from 20% ∼ 70% to less than 5%. Compared with the existing mixing rules, the modified mixing rule is with higher calculation speed and accuracy. The relative error of components mole fraction calculation is reduced from 30% ∼ 80% to less than 10%. Compared with the existing viscosity models, there are large improvements of the modified viscosity model in accuracy. The relative error of viscosity simulation is reduced from more than 50% to about 5%. According to the simulation results, C2∼C15 are the key hydrocarbons with positive effect on the miscibility of CO2-crude systems, while C16+ are the key hydrocarbons with negative effect. The recovery of the pilot has increased by 23% by these methods.The improved phase behavior models provided in this paper possess as good performance as existing models in calculation speed, and accustom a big step forward in simulation accuracy. The modified components of the models also partially complete physical meaning in describing phase behavior of CO2-crude system. All the models mentioned above are finally applied in L block with HP/HT and high water cut and obtained an increase in recovery by 19.2%.

This paper provides a case study for the application of gas assisted gravity drainage (GAGD) EOR in the Donghe I carboniferous reservoir, which is ultra-deep (>16,000 ft), thick (150-400 ft) and has high temperature, pressure, and salinity values of 410 K, 6500–8500 psi and > 200,000 ppm, respectively. The study aims to mitigate the problems faced by this reservoir such as high degree of reserve recovery (40%), high water cut (> 82%), low reservoir pressure (pressure maintenance level < 0.65), and low production rate. The techniques described in this paper are based on natural gas miscible flooding in which the GAGD technique occupies the dominant position, with the plane displacement technique being assisted. A reservoir description focusing on fine characterization of the interlayer has been applied combined with a compositional ten-million-grid-scale reservoir simulation model to make the filtration mechanism be straightforwardness. Laboratory experiments were conducted to reveal the phase behavior of reservoir fluids after gas flooding. Several key methods have been established including gas breakthrough indication by composition, well pattern reconstruction, injection/production matching, typical treatment for typical wells, and injection/production ratio control. The aforementioned techniques and methods mentioned have been carried out in Donghe I carboniferous reservoir as part of a geology-reservoir-engineering-economics-HSE integral project. The daily gas injection rate at the pilot site is around 1.4 MMCF/D using 4 horizontal wells as gas injectors and the injection/production ratio is over 1.5, which is considered a satisfactory result. As of 2018, there are 13 response producers of 25 wells in this reservoir, of which 6 wells showed doubling or tripling of daily production and 5 wells transferred from artificial lifting to natural flow, with production of two of the producers exceeding 650 BPOD. The cumulated oil production and gas injection have respectively reached 1.6 MMbbl and 10.2 BCF. The injected gas serves as an effective supplement to reservoir energy, causing the pressure maintenance level to climb up to 0.8. The gas injection has also had a positive effect on the composite production decrement and composite water cut of this reservoir. The composite production decrement has declined from 15.5%/A before gas flooding to about zero after gas flooding till date, and the production still has a trend to change from constant rate to rising rate. Simultaneously, the water cut which was previously increasing has now started dropping down. The promoted GAGD EOR technique showcased in this paper effectively combines a series of technical and theoretical methods including natural gas miscible flooding, GAGD with plane gas displacement, gas breakthrough indication by composition, injection/production matching, typical treatment for typical well, injection/production ratio controlling, ten-million-grid-scale compositional-model simulation, and laboratory experiments on phase behavior. All these aspects combined have led to a favorable field test performance in the Donghe I reservoir.

Studying influence factors of movable fluid distribution is of significance for the development of clastic reservoirs. In this paper, the characteristics of movable fluid distribution in clastic core samples from Tarim Oilfield were studied with nuclear magnetic resonance (NMR) tests and computed tomography (CT) scanning. The movable fluid saturations of each core sample under centrifugal pressure of 3 psi, 10 psi, 50 psi, 100 psi, and 500 psi were measured, respectively. The main factors influencing the movable fluid saturation were analysed. The results show that the movable water saturation is strongly dependent on the centrifugal pressure used. When the centrifugal pressure increased to 100 psi, the movable fluid saturation increased sharply. The CT scanning indicates that the movable fluid saturation is mainly dominated by pore sizes of core samples, and the larger the average pore size is, the higher the fluid saturation will be.

Water injection is essential during secondary recovery process. For multi-layer reservoir, determining the water absorbing capacity of stratification is a necessary step when we investigate the reservoir's production status and analyze the distribution of remaining oil. It is also the precondition of reasonable measures to boost the production, and to improve the ultimate recovery. In view of the reservoir actual situation is multitudinous, such as water is injected at the bottom while the upper produces, and insulation between layer and layer is very poor which can lead to Interlayer channeling, conventional reservoir engineering methods have been unable to effectively forecast recovery performances. This paper set up a methodology of injection allocation in multi-layers water flooding reservoir with support vector machine (SVM) as the core. Because of the credibility of injection profile and liquid producing profile is the highest of all the materials, the method sets the injection profile data and fluid producing profile as the goal, the permeability, reservoir thickness, injection-production well spacing and injection speed and liquid producing as influencing factors, building support vector, then we establish forecasting model, and allocate the water injection rate to each layer. By comparing the injection profile data which is not involved in the prediction process, we found that the prediction of the support vector machine (SVM) method is close and practical, with high credibility. This paper also compared the results with numerical simulation results of commercial numerical simulation software. It is can be found that the method can accurately reveal the inherent law of oilfield development; it has good adaptability and sufficient accuracy. As the forecasting process taking too long, this paper gives the optimization of SVM algorithm. Due to the parameter optimization process of conventional SVM algorithm carried out by searching grid one by one, it ashows these characteristics: low efficiency, large memory overhead. In this paper, we replace the grid searching parameter optimization process by genetic algorithm in the process of machine learning, Results show that the computation time can be reduced more than 70%.

In order to obtain polymer microsphere for profile control and water shutoff with good suspension, high temperature resistance and high salinity tolerance, micrometer DCA (divinylbenzene-co-acrylamide) microsphere which had a three-layer structure was synthesized by the monomer of acrylamide, initiator of azobisisobutyronitrile, crosslinking agent of divinylbenzene and emulsifier of ethoxylated lauryl alcohol sulfate. Basic performance of DCA microsphere was evaluated, including temperature resistance, plugging capacity, effect of profile control and water shutoff and method optimization. The results showed that the particle size of DCA microsphere ranged from 2 μm to 339 μm and the shell had a Nano porous structure. The temperature resistance of DCA microsphere was up to 300°C. The residual resistance coefficient of each segment along artificial homogeneous cylindrical core was still more than 2 with a strong deep plugging efficiency. After water flooding of oil in two-dimensional heterogeneous cores was carried out, the slug of DCA microspheres was injected. The amplitude of enhanced oil recovery was up to 10.18% in stage of the subsequent water flooding. On the basis of the subsequent water flooding, the recovery of carbon dioxide flooding was 19.64%. Feasibility methods which were the plugging of DCA microspheres and carbon dioxide flooding on deep profile control technology in high permeability of high temperature and high salinity reservoirs have been roposed.