Sage Journals: Discover world-class research

Abstract

When injecting polymer into an oil reservoir, polymer adsorption onto the grain surface may seriously clog the pore throat and decrease the permeability. This phenomenon may give rise to a severe cross-flow problem in the case of a layered reservoir with a large permeability contrast. Therefore, we proposed a remedial method, which injects polymer solution with different concentrations under the same condition of the frontal velocity at each layer, to overcome the cross-flow problem. In order to perform this task, the experimental system was built as a two-dimensional two-layered sandstone system with a large permeability contrast of 10. From the results of the experimental investigation, conventional polymer flooding was found to be inefficient in oil recovery, because of the cross-flow problem that occurred almost from the beginning of injection. Meanwhile, in the case of the suggested remedial method of polymer flooding, the oil production was significantly enhanced without showing the cross-flow problem, which was accomplished by injecting the polymer solution with different concentrations under the conditions of the same frontal velocity at each layer. Overall, the suggested injecting method was found to be sufficient to overcome the cross-flow problem in a layered reservoir.

Keywords

Polymer flooding layered reservoir permeability contrast polymer adsorption cross-flow

Introduction

Stratified sandstone reservoirs with a large permeability contrast are quite typical for many North Sea fields, in which 95% of the crude oil reserves of Western Europe and 90% of the crude oil production coming from the deposits lying under the North Sea reservoirs (Clifford and Sorbie, 1985). When polymer flooding is applied to the layered reservoir with porosity difference, the frontal velocity of fluid injected for each layer varies even when the fluid is injected at the same rate. This disparity leads to an imbalance in the lateral pressure gradient in each layer, and the pressure discrepancy between layers increases the cross-flow problem. The field data show that injected fluids tend to travel quickly through highly permeable layers and bypass less permeable layers in a two-layered system. In particular, polymer retention can be critical in low permeability formation due to the polymer adsorption-induced permeability reduction, which yields negative effects on oil recovery (Park et al., 2015). Even when both injectors and producers are completed in a low permeable zone, injected fluids quickly cross through the boundary to a highly permeable zone, and the oil in the low permeable zone tends to be bypassed, resulting in high water-cut and poor sweep of low permeable zone (Lee et al., 2011; Sorbie and Seright, 1992).

For such reservoirs, one of the important factors that affects sweep efficiency is cross-flow from one layer into another layer in the direction perpendicular to the main flow path (Fitzmorris et al., 1992). One possible remedy for cross-flow control is simultaneous injection of fluids with different properties (Rossen et al., 2006; Stone, 2004; Surgchev et al., 1996). Masalmeh and Blom (2011) showed the better performance in oil recovery and vertical sweep efficiency, as a simulation study, by controlling the mobility of the injected fluids in high and low permeable zones to prevent mixing of the two streams, so that the injected fluid was confined in each layer.

Therefore, we proposed a remedial method to overcome the cross-flow problem. This method injects a polymer solution with different concentration for each layer. At the same time, the frontal velocity was also designed to be the same for each layer, by considering the difference of the porosity in each layer. In order to perform this task, the experimental system was built as a two-dimensional two-layered sandstone system with a relatively large permeability contrast.

Experimental setup

Rock and fluid properties

The dimensions of Castlegate in the upper layer and Berea sandstones in the lower layer were 0.2 m × 0.1 m with 0.02 m thickness. The measured permeabilities of those sandstones were 7.5E-13 m² (750 mD) and 7.5E-14 m² (75 mD) in the upper and lower layers, respectively (Table 1). The measured porosities were 28% in the upper layer and 20% in the lower layer. Due to this porosity disparity between the two layers, the injection velocity (namely injection rate) was different from the frontal velocity which was divided by the porosity. The results of X-ray CT to investigate internal sections of sandstones showed that the pore size of Castlegate in the upper layer was larger than that of Berea in the lower layer, representing a disparity in the flowing capability (Figure 1).

Table 1.

Experimental setup for two-dimensional vertical layered system.

Two-dimensional vertical layered reservoir system
Permeability (mD)
Castlegate sandstone (Upper)	750
Gray Berea sandstone (Lower)	75
Porosity (%)
Castlegate sandstone (Upper)	28
Gray Berea sandstone (Lower)	20
Reservoir temperature (℃)	25
Viscosity at 25℃ (mPa s)
2% Brine (NaCl)	1.0
Oil (Shell Morlina S2BL10)	8.8
Flopaam 3330s at 1200 ppm	5.1
Confining pressure (MPa)	4.83
BPR at production (MPa)	2.07
Initial oil saturation (%)	100
Total injected amount (PV)	2.1
Color of upper/lower injected fluids	Red/Blue

Figure 1.

X-ray CT scanning images showing pore sizes of Castlegate and Berea sandstones.

The oil used in this study was Shell Morlina S2BL 10, showing a viscosity of 8.8E-3 kg/m s (8.8 mPa s) at 25℃. The 2% salinity brine with 1.0E-3 kg/m s (1.0 mPa s) of viscosity at 25℃ was applied in the system using NaCl. The polymer was Flopaam 3330S™ (SNF Floerger) which is a commonly used HPAM with a molecular weight of 8 million Daltons. During polymer flooding, fluids in porous media flow under isothermal conditions. Therefore, the viscosity of the HPAM solution in this study was affected by the injected polymer concentration, not temperature, as shown in Figure 2. The densities of oil, water, and polymer solution used in this experiment were 0.88, 1.00, and 1.03 g/cm³, respectively. This density difference between the fluids was only expected to have a small impact on the cross-flow problem. The injection fluids were dyed red with Rhodamine B and blue with FD & C Blue No. 1 for visual investigation of the sweeping appearance.

Figure 2.

Viscosity of the polymer solution for concentration and shear rate.

Experimental apparatus and procedure

A flooding apparatus for the experiment was setup as shown in Figure 3. The experimental system was set for reservoir conditions by using typical main components including a newly designed cylindrical plate holder for two-dimensional sandstone slabs, fluid storage vessels, displacement pumps, backpressure regulators, and a data acquisition system. The porosity and permeability of sandstone slabs were measured, and then the slabs were sufficiently dried. The two-dimensional sandstone slabs, in this research, were stacked as a two-layered reservoir with a glass fiber filter between the two slabs as shown in Figure 4. A glass fiber filter having mesh size of 1.6E-6 m was inserted between the slabs to obtain even contact and to provide good capillary contact (Eilertsen et al., 1996). Then, the layered system with micro fiber filter papers was thoroughly tested, so it did not form a high permeable flow path between the layers. The two-layered sandstone slabs were coated with silicon, wrapped with Teflon and aluminum foil, and covered with Viton rubber to prevent channeling through the walls (Han et al., 2016). Then, the assembled two-layered system was vertically inserted into the 4.6E-1 m cylindrical vessel. The sandstone system was fully saturated with oil using various injection rates for a sufficiently long time. In order to inject the fluids into each layer individually, two syringe pumps were used. The effluents produced were collected from each layer separately.

Figure 3.

Schematic presentation of flowing experiment.

Figure 4.

Two-dimensional vertical flowing core holder with layered reservoir system.

Results and discussion

In order to investigate the cross-flow problem, several experiments were conducted with regards to the polymer concentration (Table 2): Polymer concentrations in the upper and lower layers were 0 and 0 ppm as PF-1 (i.e. water flooding (WF)), and 1200 and 1200 ppm as PF-2 (i.e. conventional polymer flooding: PF), 1200 and 0 ppm as PF-3 (i.e. cross-flow-controlled polymer flooding (CCPF)). In these three cases, the injection rates for the upper and lower layers with different porosities were individually designed so that the frontal velocities would be the same as 2.5 × 10⁻⁴ m/s. This condition was the same in PF-1, PF-2, and PF-3. The experiment was set as a two-layered reservoir containing a high permeable upper layer and low permeable lower layer (permeability contrast was relatively large as 10.0) as shown in Figure 2.

Table 2.

Operating conditions of concentration based cross-flow-controlled polymer flooding.

Case	PF-1	PF-2	PF-3	PF-4	PF-5
Case	(WF) V_f,U = V_f,L	(PF) V_f,U = V_f,L	(CCPF) V_f,U = V_f,L	(CCPF) V_f,U < V_f,L	(CCPF) V_f,U > V_f,L
Upper
Concentration (ppm)	0	1200	1200	1200	1200
Injection rate (m³/s)	1.40E-8	1.40E-8	1.40E-8	1.67E-9	3.33E-9
Frontal velocity (m/s)	2.50E-4	2.50E-4	2.50E-4	2.97E-5	5.95E-5
Lower
Concentration (ppm)	0	1,200	0	0	0
Injection rate (m³/s)	1.00E-8	1.00E-8	1.00E-8	1.67E-9	8.33E-10
Frontal velocity (m/s)	2.50E-4	2.50E-4	2.50E-4	4.27E-5	2.13E-5

From the result of the investigation shown in Figure 5(a), the oil recoveries in the high permeable upper layer were relatively high and similar within the range of 45.5–47.1% for the cases of WF (PF-1), PF (PF-2), and CCPF (PF-3). This was thought to be mainly because of high permeability system (750 mD) in upper layer. On the other hand, the results showed a noticeable difference from 22% of WF to 33% of CCPF in the low permeable lower layer (Figure 5(b)). This result could be explained as follows: First, the flow paths were observed through the visual photos at the ending time of injection (2.1 PV injected) as shown in Figure 6. In this figure, red and blue colors denote the fluids injected into the upper and lower layers, respectively. As a result, in the cases of WF (PF-1) and PF (PF-2), the fluid injected into the lower layer was moved upward almost from the beginning of injection, rather than displacing the oil of the lower layer. Thus, most of the oil in the low permeable lower layer remained unswept. Whereas, the fluid injected into the lower layer displaced the oil mostly in the horizontal direction within the layer, without showing cross-flow problem, in CCPF (PF-3) case, with the same frontal velocity applied in both layers.

Figure 5.

Oil recovery for different injection methods. (a) Upper layer, (b) Lower layer.

Figure 6.

Pictorial presentation of flow paths at the end of injection (2.1 PV injected). (a) WF (PF-1), (b) PF (PF-2), (c) CCPF (PF-3).

Second, this result could also be confirmed from the pressure data recorded during the experiment. ΔP_V shown in Figure 7 represents the vertical difference in the average pressures of the upper and lower layers $({\bar{P}}_{U} - {\bar{P}}_{L})$ . From the results of WF (PF-1) and PF (PF-2) in Figure 7(a), a negative ΔP_V denotes that the pressure of the lower layer was greater than the pressure of the upper layer, implying that the fluid injected into the lower layer moved upward (i.e. cross-flow). Meanwhile, ΔP_V was almost maintained as zero during the entire time of injection in the case of CCPF (PF-3), and this is shown schematically in Figure 7(b). Moreover, the ΔP_V of PF (PF-2) was found to be greater than that of WF (PF-1). This result was due to the polymer retention, which reduced the effective permeability (Figure 9). ΔP_H shown in Figure 8 represents the horizontal pressure difference between the inlet and outlet in the upper layer and lower layer.

Figure 7.

Vertical pressure difference between upper and lower layers.

Figure 8.

Horizontal pressure difference between inlet and outlet in the upper and lower layers.

Figure 9.

SEM images of adsorbed polymer on the pore surface and effective permeability reduction by polymer retention in the lower layer. SEM: scanning electron microscope.

In addition, we described the actual flow path using vectorial presentation, based on the analysis of pressure differences in the vertical direction (Figure 7) and horizontal direction (Figure 8). In these figures, the flow path was represented by the angle of flowing direction, and it was calculated using the arctangent of ΔP_V/ΔP_H. As depicted in Figure 10(a), the actual flowing vectors were estimated to be 45.2° upward for the WF (PF-1) case and 17.0° upward for the PF (PF-2) case, while the flowing direction in CCPF (PF-3) was 0.9° upward, which indicated horizontal flow. As can be seen in Figure 10(b), the situation noted above occurred almost from the beginning of injection.

Figure 10.

Vectorial presentation of the flowing path and flowing direction of injected fluid in the lower layer.

Next, the following experiments were conducted to investigate the effects of frontal velocity in each layer (Table 2). In the first case, the same frontal velocity was applied in both layers (CCPF: PF-3). In the second case, the same injection rate was applied in both layers, which yielded higher frontal velocity in the lower layer (PF-4). In the third case, higher frontal velocity in the upper layer (PF-5).

From the results shown in Figure 11, the oil recoveries from upper layer were relatively similar in the range of 40–47.1%. However, the recovery from the lower layer showed a large difference from 3.6 to 32.9% among the cases. In the case of PF-4, when the injection rate was equally set for both layers, the frontal velocity of the lower layer was higher due to its low porosity, resulting in only 3.6% oil recovery from the lower layer. In the case of higher frontal velocity in the upper layer (PF-5), it showed a slight improvement in oil recovery, in comparison to the result noted above (PF-4). Meanwhile, the oil recovery was significantly increased when the same frontal velocity was applied in both layers in CCPF (PF-3). These trends could be explained as follows: First, the flow paths were observed through the visual photos taken after the completion of the experiments (2.1 PV injected) as shown in Figure 12. The results indicated that the CCPF, i.e. the same frontal velocity in both layers, yielded successful displacement of oil within the layer, without showing the cross-flow problem (Figure 12(a)). Meanwhile, when the injection rate was the same in both layers (PF-4), the fluid injected into the lower layer was moved upward almost directly as soon as it was injected (Figure 12(b)). A similar pattern to that in Figure 12(b) was also observed in Figure 12(c), in which a higher frontal velocity in the upper layer was set (PF-5).

Figure 11.

Oil recovery from for different injection methods. (a) Upper layer, (b) Lower layer.

Figure 12.

Pictorial presentation of flow paths at the end of injection (2.1 PV injected). (a) v_f,U = v_f,L (PF-3), (b) v_f,U < v_f,L (PF-4), (c) v_f,U > v_f,L (PF-5).

Second, the results noted above could also be confirmed from the pressure data analysis. In Figure 13, ΔP_V represents the vertical pressure difference of the average pressures of the upper and lower layers $({\bar{P}}_{U} - {\bar{P}}_{L})$ . Based on the results of Figure 13(a), the vertical pressure difference between the layers was negative almost from the beginning of injection when the polymer was injected at the same rate into both layers (PF-4). The case of higher frontal velocity in the upper layer (PF-5) also showed a negative value for the vertical pressure difference. On the other hand, when the frontal velocity was the same in both layers (CCPF: PF-3), the resulted vertical pressure difference was almost positive during the entire time of injection. These phenomena could also be schematically illustrated as shown in Figure 13(b). In Figure 14, ΔP_H represents the horizontal pressure difference between the inlet and outlet in the upper and lower layers.

Figure 13.

Vertical pressure difference between upper and lower layers.

Figure 14.

Horizontal pressure difference between inlet and outlet in the upper and lower layers.

In addition, the actual flow path was described using the vectorial presentation with the pressure differences in the vertical and horizontal directions. As shown in Figure 15(a), the actual flowing direction when the same injection rate was applied in both layers (PF-4) was estimated to be 32° upward. Whereas, the flow was almost in the horizontal direction in the case of the same frontal velocity in both layers. As can be seen in Figure 15(b), the flowing direction noted above occurred almost from the beginning of injection. This analysis was consistent with the flow paths shown in the results of the visual photos (Figure 12). As an overall result, the proposed injection method (CCPF) which injects polymer solution with different concentrations into each layer under the same condition of the frontal velocity in each layer remarkably enhanced the oil production during polymer flooding in a layered reservoir with large permeability contrast.

Figure 15.

Vectorial presentation of the flowing path and flowing direction of injected fluid in the lower layer.

Conclusion

In this study, we proposed a remedial method to inject polymer solution with different concentrations under the same condition of the frontal velocity in each layer, in order to overcome the cross-flow problem. In order to perform this task, the experimental system was built as a two-dimensional two-layered sandstone system with a relatively large permeability contrast.

From the results of the investigation of a two-layered reservoir containing a low permeable layer, conventional polymer flooding was found to be inefficient in oil recovery due to the cross-flow problem that occurred almost from the beginning of injection. However, the CCPF method suggested in this study remarkably enhanced the oil production from the low permeable lower layer by preventing the cross-flow problem, which was achieved by injecting polymer solution with different concentration. This result was achieved when the frontal velocity was the same in both layers having different porosities.

In the analysis of flow paths from visual photos and vectorial presentation of the pressure data, the results demonstrated that the injected fluid moved almost directly upward as soon as it was injected, in the case of conventional polymer flooding. Therefore, most of the oil in the low permeable lower layer remained unswept. Meanwhile, the oil moved mostly in the horizontal direction in the case of CCPF with the same frontal velocity in both layers. Accordingly, the cross-flow problem did not occur, implying successful displacement of oil within the layer.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research,authorship,and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research,authorship,and/or publication of this article: This work was supported by a grant funded as a part of the project “Development of IOR/EOR technologies and field verification for carbonate reservoir in UAE” by the Korean Government Ministry of Trade,Industry and Energy (MOTIE) (No. 20152510101980).

References

Clifford PJ and Sorbie KS (1985) The effects of chemical degradation on polymer flooding. In: SPE international symposium on oilfield chemistry, Society of Petroleum Engineers, Phoenix, AZ, 9–11 April, pp.1–23.

Eilertsen

Borresen

Bertin

et al. (1996) Experiments and numerical simulations of fluid flow in a cross layered reservoir model. Journal of Petroleum Science Engineering 18(1): 49–60.

Fitzmorris ER, Kelsey FJ and Pande KK (1992) Effect of crossflow on sweep efficiency in water/oil displacement in heterogeneous reservoirs. In: SPE Sixty-seventh annual technical conference and exhibition, Society of Petroleum Engineers, Washington, DC, 4–7 October, pp.1–14.

Han

Lee

et al. (2016) Effect of gravity segregation on CO₂ sequestration and oil production during CO₂ flooding. Applied Energy 161: 85–91.

Lee T, Jang Y, Park H, et al. (2011) Investigation of water BT phenomenon in the fractured basement reservoir contacted with bottom water aquifer. In: SPE Asia Pacific oil and gas conference and exhibition, Society of Petroleum Engineers, Jakarta, Indonesia, 20–22 September, pp.1–14.

Masalmeh SK and Blom CPA (2011) Simultaneous injection of water and polymer (SIWAP) to improve oil recovery and sweep efficiency from layered carbonate reservoirs. In: SPE enhanced oil recovery conference, Society of Petroleum Engineers, Kuala Lumpur, Malaysia, 19–21 July, pp.1–22.

Park

Han

Sung

(2015) Effect of polymer concentration on the polymer adsorption-induced permeability reduction in low permeability reservoirs. Energy 84: 666–671.

Rossen WR, van Dunjin CJ, Nguyen QP, et al. (2006) Injection strategies to overcome gravity segregation in simultaneous gas and liquid injection into homogeneous reservoir. In: SPE/DOE symposium on improved oil recovery, Society of Petroleum Engineers, Tulsa, OK, 22–26 April, pp.1–16.

Sorbie KS and Seright RS (1992) Gel placement in heterogeneous systems with crossflow. In: SPE/DOE eighth symposium on enhanced oil recovery, Society of Petroleum Engineers, Tulsa, OK, 22–24 April, pp.1–18.

10.

Stone HI (2004) A simultaneous water and gas flood design with extraordinary vertical gas sweep. In: SPE international petroleum conference, Society of Petroleum Engineers, Puebla Pue, Mexico, 8–9 November, pp.1–17.

11.

Surgchev LM, Hanssen JE, Johannessen HM, et al. (1996) Modelling injection strategies for a reservoir with an extreme permeability contrast: IOR qualification. In: European 3-D simulation modelling conference, Society of Petroleum Engineers, Stavanger, Norway, 16–17 April, pp.1–6.

A remedial method of polymer flooding to prevent cross-flow problem in a layered reservoir