Fluid Flow in Unconventional Gas ReservoirsView this Special Issue
Research Article | Open Access
Pengxiang Diwu, Tongjing Liu, Zhenjiang You, Ganggang Hou, Runwei Qiao, Lekun Zhao, "Study on Pulse Characteristic of Produced Crude Composition in CO2 Flooding Pilot Test", Geofluids, vol. 2018, Article ID 2968629, 5 pages, 2018. https://doi.org/10.1155/2018/2968629
Study on Pulse Characteristic of Produced Crude Composition in CO2 Flooding Pilot Test
It has been observed in many laboratory tests that the carbon number of the maximum concentration components (CNMCC) of produced oil varies monotonically with CO2 injection volume at the core scale. However, in CO2 flooding pilot test at the field scale, we find that the CNMCC is usually nonmonotonic function of CO2 injection volume, which is called “pulse characteristic” of CNMCC. To investigate the mechanism of this phenomenon, we analyze the physical process of CO2 flooding in heterogeneous reservoir and explain the reason of the pulse characteristic of CNMCC. Moreover, two 3D reservoir models with 35 nonaqueous components are proposed for numerical simulation to validate the conjecture. The simulation results show that pulse characteristic of CNMCC only occurs in the heterogeneous model, confirming that the pulse characteristic results from the channeling path between wells, which yields nonmonotonic variation of oil-CO2 mixing degree. Based on it, a new method can be developed to identify and quantify the reservoir heterogeneity.
In recent years, the gas drive technology has been rapidly developed and widely applied. It has become another important way to improve oil recovery, besides thermal recovery and chemical flooding [1, 2]. CO2 has several advantages on enhancing oil recovery, such as being easily soluble in oil, reducing oil viscosity , and reducing residual oil saturation . Therefore, CO2 drive is widely used in different types of oilfield, including sand and carbonate reservoirs [5–8].
The process of CO2 displacement differentiates different components in oil . This differentiation results in different flow rates among the compositions with different viscosities: light compositions are quickly carried away by CO2, while heavy compositions move significantly slower. Consequently, produced oil composition will change along with the increase of displacement time.
It is widely known that the carbon number of maximum concentration components (CNMCC) increases monotonically with CO2 injection volume, which has been confirmed by the experimental results [10, 11]. In 2011, Yang et al.  analyzed the produced oil components in a CO2 slime tube tests under conditions of immiscible and miscible flooding, respectively. The chromatograph analysis results of the two tests both show that the CNMCC of the produced oil increases monotonically with CO2 injection volume. In 2015, Zhou et al.  analyzed the contents of the fluid compositions under different pressures by an 18-meter slime tube made of quartz. The results show that, at 18.1 MPa, the CNMCC of produced oil increases from C9 to C14 monotonically and gradually with the increase of CO2 injection volume. At 25.2 MPa, the CNMCC of produced oil increases from C10 to C20 monotonically and quickly with the increase of CO2 injection volume.
Despite several laboratory studies on CNMCC variation at core scale, the researchers tend to pay more attention to asphalt deposition, rather than the CNMCC variation at reservoir scale [12–15]. In fact, the latter is still not well understood during CO2 flooding process in oil reservoirs. Generally, the differences in length and thickness between core and reservoir scales can result in different heterogeneity in horizontal and vertical directions. CO2 channeling in flow direction at core scale is weaker than that at reservoir scale. Meanwhile, in vertical direction, CO2 and oil at reservoir scale are mixed more thoroughly than at core scale. Based on these two main differences, it is reasonable to believe that the variation mechanism of CNMCC at reservoir scale would be quite different from that at core scale.
To investigate the CNMCC variation mechanism in CO2 field flooding, this paper introduces a new concept “pulse characteristic” to characterize the CNMCC change law in produced oil, based on the field data. The physical process of displacement in heterogeneous formation is analyzed, and a scientific conjecture on the reason of the pulse characteristic of CNMCC at reservoir scale is proposed. Moreover, two numerical components models are established to validate the conjecture successfully. The pulse characteristic of CNMCC may provide a new way to identify and quantify the heterogeneity of CO2 flooding reservoirs.
2. CNMCC Variation at Reservoir Scale
Based on the experimental results on CO2 flooding [10, 11], it is believed that CNMCC should increase monotonically with CO2 injection volume. However, CO2 flooding pilot test at reservoir scale give us different results.
Composition analysis of produced oil from production well W1 in oil field H1 (China) has been performed. Figure 1 illustrates the relative component concentrations of produced oil at different time (from 122 to 852 days), in which -axis shows components in oil phase and -axis presents relative concentration. The relative component concentrations are calculated as normalized concentrations of components C5 to C31.
There is strong heterogeneity in the oil field H1, resulting from microcracks. The main parameters of the reservoir and fluid are listed in Table 1.
The commissioning date of production well W1 is one month before the commissioning date of corresponding injection well, which has a continuous CO2 injection rate of 30 t/d. We set the commissioning date of corresponding injection well as the initial time; that is, injection time equals zero. The CNMCC variation of production well W1 can be divided into the following three stages (shown as three highlighted regions in Figure 1).
(1) From 122 to 304 days (pink and blue lines), the relative concentrations of compositions C5–C11 in produced oil decrease, while the relative concentrations of compositions C12–C16 increase. The corresponding CNMCC increases from C8 to C15. In this stage, the CNMCC has positive correlation with CO2 injection volume, which coincides with the core experiments [10, 11].
(2) From 304 to 457 days (blue and red lines), the relative concentrations of compositions C5–C10 in produced oil increase, while the relative concentrations of compositions C11–C16 decrease. In this stage, the CNMCC has negative correlation with CO2 injection volume. It indicates the nonmonotonic variation of CNMCC in the reservoir scale.
(3) From 457 to 852 days (red and green lines), while the relative concentrations of compositions C5–C10 in produced oil decrease, the relative concentrations of compositions C11–C23 increase. The corresponding CNMCC of produced oil increases again from C8 to C13.
As shown in Figure 1, the CNMCC of production well W1 changes nonmonotonically with the CO2 injection volume; that is, it increases at the beginning, then decreases later, and increases again. We propose the concept of “pulse characteristic” to describe the nonmonotonic variation of CNMCC at reservoir scale. To explain the mechanism of this phenomenon, we analyze the entire CO2-oil interaction process in heterogeneous reservoir and provide our explanation in the next section.
3. Mechanism of Pulse Characteristic
According to the composition characteristics of produced oil, there are three main stages in the process of CO2 flooding in homogeneous reservoirs [16–18]. The produced oil in the early stage is not in contact with CO2, so its compositions are the same as the original oil compositions. In the middle stage, the produced oil is from the leading displacement edge, where mainly the light oil compositions occupy, because of the oil-CO2 extraction and dissolution. The produced oil in the late stage mainly consists of the remaining oil and usually has heavy compositions. Since the characteristic of the middle stage is not significant , the CNMCC increases monotonically with CO2 injection volume at core scale.
However, the core scale usually exhibits lower heterogeneity than the reservoir scale. If there is plane heterogeneity, that is, CO2 channeling-paths in the formation, the above three stages of CO2 flooding would be different.
The main difference occurs in the late stage when CO2 has occupied the whole CO2 channeling-paths. The produced oil in this stage is mainly from outside of the channeling-paths, since CO2 has displaced almost all the oil in the channeling-paths, as shown in Figure 2. This “fresh oil” has not been extracted thoroughly and still has quite a few light compositions. Therefore, the light compositions in produced oil will increase, which results in CNMCC decrease in the late stage. To sum up, the interaction degree between oil and injected CO2 is not monotonic with time anymore under heterogeneous condition.
(a) Plane map
(b) Vertical map
Based on the above analysis, we propose a conjecture to explain the CNMCC pulse characteristic at reservoir scale: the CNMCC will show back-and-forth pulse characteristic if there is plane heterogeneity in the reservoir, that is, channeling-paths. In the next section, this conjecture will be validated by two fully compositional models numerically.
4. Validation by Numerical Simulation
4.1. Fully Compositional Model Setting
To validate the conjecture in Section 3, we establish two fully compositional numerical models from Eclipse E300 module to simulate homogenous and heterogeneous scenarios, respectively. The heterogeneous model has an interwell channeling path which permeability is 100 mD. The corresponding formation and fluid parameters of the models are the same as those listed in Table 1. To investigate the components variation, we set 35 nonaqueous components in the models, including CO2, N2, C1–C32, and C32+. The oil compositions are shown in Table 2. The corresponding relative component concentrations are shown in Figure 3 (blue line), illustrating that initial CNMCC is C9.
According to the field flow regime, the development processes are depletion drive until formation pressure reaches 12 MPa, then water drive, and finally CO2 drive development. In CO2 flooding, the injection and production rates are constant (7500 sm3/day and 2.5 rm3/day, resp.). Based on the MMP in Table 1, it is miscible flooding near injection well and immiscible flooding away from injection well.
4.2. Simulation Results
Figure 3 shows the produced oil composition of the homogeneous model at different time. During CO2 flooding process, the CNMCC increases monotonically from C9 to C12 with CO2 injection volume. Although the CNMCC of produced oil are constant at C12 from injection time 3803 to 7650 days, we still observe the increasing relative concentration of heavy compositions and decreasing relative concentration of light compositions.
Figure 4 shows the produced oil composition of the heterogeneous model at different time. The CNMCC of produced oil increases from C9 to C30 during CO2 injection time 0–210 days. After 210 days, the CNMCC of produced oil decreases from C30 to C12 until 788 days. Then the CNMCC of produced oil remains constant at C12. The nonmonotonic variation of CNMCC only occurs in heterogeneous scenario, which is also observed in the field case shown in Figure 1.
The CNMCC increases monotonically from C9 to C12 with CO2 injection volume in the homogenous model, exhibiting the same change law with the experimental results [10, 11]. Based on the analysis in Section 3, it results from the monotonic variation of oil-CO2 mixing degree in homogenous model.
The variation of CNMCC shows pulse characteristic in the heterogeneous model (Figure 4), which validates the conjecture in Section 3. Figure 5 presents the gas-oil ratio (GOR) and the oil saturation inside and out of the channel at the half of interwell distance.
Before injection time 210 days, the produced oil is mainly from the channeling path, since the oil saturation in the path decreases rapidly (blue line). The mixing degree between oil and injected CO2 is increased monotonically. Therefore, the CNMCC increases from C9 to C30 monotonically as well.
In the period of injection time 210–788 days, the produced oil is mainly from outside of the path, since the oil saturation out of the path decreases rapidly (red line). The interaction degree between the oil and injection CO2 becomes nonmonotonic because the new oil enters and contributes to the produced oil. Therefore, the CNMCC decreases from C30 to C12 in this period.
The current work proposes a sufficient condition for pulse characteristic of CNMCC at reservoir scale. More intensive research needs to be conducted on the effects of other factors, including formation pressure, oil composition, and water saturation.
To investigate the CNMCC variation mechanism at reservoir scale, the present paper introduces the concept of pulse characteristic to describe the CNMCC change law in CO2 flooding pilot test. Two fully compositional numerical models are established to validate the conjecture on the pulse characteristic. The main conclusions are drawn as follows.
(1) The CO2 flooding field data exhibits pulse characteristic; that is, the CNMCC of produced oil increases in the early stage, then decreases, and increases again in the late stage. It is different from the monotonic increasing characteristic at core scale.
(2) The pulse characteristic of CNMCC only occurs in heterogeneous formation, resulting from the nonmonotonic change of interaction degree between oil and injected CO2. It is validated by simulation results of the two fully compositional models.
(3) A new method may be developed to identify and quantify the reservoir heterogeneity, using the idea of pulse characteristic of CNMCC, since it contains the information of CO2 channeling path.
Conflicts of Interest
The authors declare no conflicts of interest regarding the publication of this paper.
This work was supported by the National Science and Technology Major Projects (2017ZX05009004) and Beijing Natural Science Foundation (2173061).
- H. Lei, S. Pingping, J. Ying, Y. Jigen, L. Shi, and B. Aifang, “Prediction of asphaltene precipitation during CO2 injection,” Petroleum Exploration and Development, vol. 37, no. 3, pp. 349–353, 2010.
- M. Dong, S. S. Huang, and R. Srivastava, “A laboratory study on near-miscible CO2 injection in Steelman reservoir,” Journal of Canadian Petroleum Technology, vol. 40, no. 2, pp. 53–61, 2001.
- Y. M. Hao, Q. W. Bo, and Y. M. Chen, “Laboratory investigation of CO2 flooding,” Petroleum Exploration and Development, vol. 32, no. 2, pp. 110–112, 2005.
- M. T. Li, W. W. Shan, X. G. Liu et al., “Laboratory study on miscible oil displacement mechanism of supercritical carbon dioxide,” Acta Petrolei Sinica, vol. 27, no. 3, pp. 80–83, 2006.
- E. H. Luo, Y. L. Hu, and B. Z. Li, “Practices of CO2 EOR in China,” Special Oil and Gas Reservoirs, vol. 20, no. 2, pp. 1–7, 2013.
- J. Qin, H. Han, and X. Liu, “Application and enlightenment of carbon dioxide flooding in the United States of America,” Shiyou Kantan Yu Kaifa/Petroleum Exploration and Development, vol. 42, no. 2, pp. 209–216, 2015.
- H.-Q. Chen, Y.-L. Hu, and C.-B. Tian, “Advances in CO2 displacing oil and CO2, sequestrated researches,” Oil-Field Chemistry, vol. 29, no. 1, pp. 116–127, 2012.
- C. Qiao, L. Li, R. T. Johns, and J. Xu, “Compositional modeling of dissolution-induced injectivity alteration during CO2 flooding in carbonate reservoirs,” SPE Journal, vol. 21, no. 3, pp. 809–826, 2016.
- M. A. Kelingsi, Carbon Dioxide Enhanced Oil Recovery Mechanism and Engineering Design, Petroleum Industry Press, Beijing, China, 1989, Translated by S. J. Cheng.
- Z. M. Yang, L. G. Tang, and S. Zhang, “Microcosmic mechanisms of CO2 flooding in low permeability reservoirs,” Intelligent Information Technology Application Association, vol. 5, 2011.
- T. Zhou, X. Liu, Z. Yang, X. Li, and S. Wang, “Experimental analysis on reservoir blockage mechanism for CO2 flooding,” Petroleum Exploration and Development, vol. 42, no. 4, pp. 548–553, 2015.
- R. J. Hwang and J. Ortiz, “Effect of CO2 flood on geochemistry of McElroy oil,” Organic Geochemistry, vol. 29, no. 1-3, pp. 485–503, 1998.
- N. A. E. da Silva, V. R. D. R. Oliveira, M. M. S. Souza, Y. Guerrieri, and G. M. N. Costa, “New method to detect asphaltene precipitation onset induced by CO2 injection,” Fluid Phase Equilibria, vol. 362, pp. 355–364, 2014.
- J. Vuillaume, I. Akervoll, and P. Bergmo, “CO2 injection efficiency, synthesis of conceptual chalk model: incremental oil recovery and CO2 storage potential,” in Proceedings of the Brazil Offshore Conference 2011, pp. 578–589, June 2011.
- E. Huang, “The effect of oil composition and asphaltene content on CO2 displacement,” in Proceedings of the SPE/DOE Enhanced Oil Recovery Symposium, Tulsa, Oklahoma, 1992.
- F. L. Zhao, H. D. Hao, J. R. Hou et al., “CO2 mobility control and sweep efficiency improvement using starch gel or ethylenediamine in ultra-low permeability oil layers with different types of heterogeneity,” Journal of Petroleum Science and Engineering, vol. 133, pp. 52–65, 2015.
- A. Alizadeh, M. H. Ghazanfari, V. Taghikhani, and A. Badakhshan, “Experimental investigation of water alternating CH4-CO2 mixture gas injection in light oil reservoirs,” International Journal of Oil, Gas and Coal Technology, vol. 8, no. 1, pp. 31–40, 2014.
- M. Bickle, N. Kampman, H. Chapman et al., “Rapid reactions between CO2, brine and silicate minerals during geological carbon storage: Modelling based on a field CO2 injection experiment,” Chemical Geology, vol. 468, pp. 17–31, 2017.
Copyright © 2018 Pengxiang Diwu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.