December 2010 Vol. 237 No. 12

Features

CO-2 Injection Studied For Deepwater Heavy Oil Reservoir

Simulation of deepwater, cold heavy oil production (CHOP) using the captured carbon dioxide (CO-2-EOR) technique has been investigated as part of the Well Engineering Research Group’s unconventional oil reservoir management studies being undertaken at The Robert Gordon University (RGU) in Scotland.
The modeling considered transportation of CO-2 from an onshore compression station using a 240-km subsea pipeline and injected into a heavy oil reservoir (2 km below the seabed) via a vertical injection well.

A production system designed to facilitate the recovery of the heavy crude to the topside separator (2 km water depth) was connected to the reservoir.

The Integrated Injection and Production Systems modeled to operate as a single module have demonstrated interesting results. Sensitivity analysis showed that the productivity of the reservoir and performance of the CO-2-EOR was significantly influenced by the reservoir characteristics, production history and the injection pressure. Both miscible and immiscible displacements were evaluated based on the reservoir pressure criteria available in the public domain. Despite the supercritical state of the transported CO-2 at high injection pressure, the integrated modeling results showed no specific requirement for intermediate compression or booster pump along the system, particularly for miscible conditions with minimum injection pressure of 3,000 psig.

Overall, the results indicated that CO-2-EOR for CHOP will initiate (for unloading reservoir) or boost (loading reservoir) production in miscible conditions. However, at low reservoir pressure, although recovery of the resources was achievable, a severe delay in the production forecast (i.e. years of zero recovery) at continuous CO2 injection revealed that immiscible condition was perhaps not as viable as miscible process for the conditions investigated.

Heavy oil development is progressively becoming a way forward to mitigate the decline in worldwide conventional crude. Despite older literature claiming that heavy oil is only found in shallow water [1], recent publication [2] indicates that Petrobras has approved the first offshore heavy oil development project for its Siri field in the Campos Basin. The Siri field is known to have recoverable reserves of 270 million barrels of heavy oil, at around 12.3 gravity API and Petrobras recognizes that the exploitation of the resources will rely on special and emerging technology.

Heavy oil recovery was previously investigated using some of the tertiary recovery techniques such as the water alternating gas (WAG) chemical process, gas injection and microbial enhanced oil recovery. About 13 methods were theoretically evaluated on two heavy oil (18-24 API) fields in Africa consisting of four reservoirs in total [3]. Pure CO-2 was reported to be the best recovery agent [4] following core-flood laboratory investigation using three injection gases (flue gas containing 15 mol% CO-2 in N2, a produced gas containing 15 mol% CO-2 in CH4, and pure CO-2) for heavy oil recovery (~14 degrees API collected from the Senlac reservoir located in the Lloydminster area, Saskatchewan, Canada).

With sensitivity of water alternating CO-2 carried out [5], a reduction in either the waterflood or the CO-2 injection rate resulted in an increase in oil recovery and showed the interference of viscous, capillary and diffusive forces. The simultaneous injection of CO-2 and steam increased recovery, reduced injection temperatures and reduced the heat input required, following a high pressure displacement on the recovery of West Sak heavy crude oil (19.2 degrees API,) using Steam/CO-2 in a 1-D laboratory experimental test conducted in an unconsolidated sand-pack that was two inches in diameter and four feet long [6].

On the other hand, space and weight constraint on most offshore platforms (new or existing) is generally a major challenge and enormous difficulties are foreseen with regard to accommodating additional facilities such as those required for CO-2-EOR. Consequently, direct injection from an onshore source could be a good way forward and this study looked into an integrated injection and production system interaction with CO-2 injected from a remote onshore source.

Model Description
Deepwater heavy oil recovery using the CO-2-EOR technique was investigated using the Petroleum Experts suite of software GAP/PROSPER/MBAL. The investigation was entirely simulation-based and the surface and subsurface facilities were integrated together to properly assess the effect of injecting CO-2 from a remote location. The total pipeline length was 250 km covering both onshore (10 km) and offshore sections (240 km). The water depth was 2 km as also was the depth below the seabed. The transported CO-2 was injected into the heavy oil reservoir via a vertical injection well.

The injection system comprised the subsea pipeline transporting CO-2 from the surface facility and connected to the subsea structure ready for injection, while the production system was connected to the topside separator via a subsea wellhead having 6-inch tubing size and 8-inch casing diameter. The schematic representation of the integrated configuration system is shown in Figure 1.

The reservoir was modeled using typical data. The reservoir thickness and radius were 300 feet and 2,500 feet respectively, the reservoir temperature was 120 oF and the original oil in place (OOIP) was 500 million stock tank barrels (MMSTB). Different production histories with initial pressure varying from 1,000 psig to 4,000 psig were used. A comprehensive sensitivity analysis was carried out for a wide range of reservoir production history to appreciate the importance of the integrated injection and production systems. Production forecast was performed for different reservoir conditions under both miscible and immiscible conditions using the following criteria presented in [7] and in Table 1.

Table 1: Miscibility and Immiscibility Criteria Based on CO2 Critical Temperature and Pressure.

Results of the following cases are discussed next: 1) miscible process and the influence on the reservoir production trend; 2) immiscible process and the influence on the reservoir productivity; 3) varying reservoir pressure at constant gas-oil ratio (GOR) for different CO2 injection pressures; 4) constant reservoir pressure at various GOR for different CO2 injection pressures; and 5) sensitivity of GOR, viscosity, heavy oil API and injection pressure.

High pressure reservoir (> 1,000 psig) is known to be suitable for CO-2 miscible process by enhancing the flow performance. However, this study has demonstrated that under certain conditions such as that of non-Newtonian heavy crude with high viscosity, the reservoir pressure will probably need to be as high as 4,000 psig to create an instantaneous impact on the productivity. Meanwhile, with CO-2 immiscible process occurring at reservoir pressure below 1,000 psig, the production forecast has demonstrated that heavy oil recovery was achieved by compensating the low reservoir pressure using high injection pressure to force the heavy oil toward the production well. Table 2 shows the key findings and differences between the two techniques.

Miscible Process
The reservoir pressure was kept constant at 4,000 psig and the injection pressure at the pipeline inlet (onshore) ranged from 2,000-7,000 psig. Other parameters remained as follow: heavy oil 20o API specific gravity, GOR (500 scf/STB). The integrated system results showed that heavy crude extraction was easily enhanced at injection pressure as high as 3,000 psig, provided that the reservoir pressure was around 4,000 psig. The production forecast (year 2000-2020) presented in Figure 2 showed the variation in maximum heavy oil production and the averaged CO-2 injection rates at different injection pressure.
Immiscible Process

The immiscible process was investigated using a lower reservoir pressure (1,000 psig) with the injection pressure at the pipeline inlet varying from 800 psig-7,000 psig. The heavy oil specific gravity was 20o API. Different production trends compared to the results obtained with high initial reservoir pressure (4,000 psig) were experienced. Even at injection pressures as high as 7,000 psig, the production forecast showed no recovery until early in the third year (01/05/2003). The recovery was initiated by a rapid production leading to the peak, followed by a curvy decline which gradually led to a steady state production.

Figure 3a and Figure 3b illustrate the gas injection and the total gas production trend as well as the heavy oil production when the injection pressure was 3,000 psig. The reservoir pressure was 1,000 psig and the GOR 100 scf/STB. The variation in heavy oil peak production was between 800-1,000 STB/D as the injection pressure reduced from 7,000 psig to 4,000 psig in increments of 1,000 psig. The difference in maximum production was approximately 2,000 STB/D as the injection reduced from 3,000 psig to 2,000 psig, and much lower (3,000 STB/D) when the injection pressure reduced from 2,000 psig to 1,000 psig.
This indicates that the production was significantly influenced by the injection pressure and that – under immiscible conditions – significant injection pressure will be required to compensate for the low reservoir pressure. Although production was possible at 1,000 psig and 900 psig injection pressure, production could only start from year 2016 and 2023 respectively, leading to about 16 and 23 years of continuous CO-2 injection with zero recovery.

Heavy crude displacement by CO-2 injection is known to rely on the phase behaviour of CO-2 and the interaction within the reservoir. The reservoir temperature and pressure can significantly affect the miscibility of the two components (CO-2 and heavy oil). At low reservoir pressure, recovery of the heavy oil to the surface was significantly delayed until mobility of the heavy crude was possible. Due to low pressures hindering the fluids immiscible or delaying the mobility of fluids, swelling and heavy oil viscosity reduction were the prerequisites prior to the fluids displacement mechanism becoming possible. The long period of no production at continuous CO-2 injection was either caused by the lack of sufficient energy (low reservoir and injection pressures) to push the fluids out of the reservoir or the inefficient fluids interaction to create the required swelling and viscosity reduction, or a combination of the two effects.

On the other hand, at 900 psig injection pressure – which was below the CO-2 critical pressure – the heavy oil displacement was possible despite more than two decades of zero gas or heavy oil production. A much smaller return on investment (ROI) and longer payback time seem to prove that the immiscible process for heavy oil recovery in some conditions may not be a viable option.

A sensitivity analysis was carried out to assess the influence of the reservoir pressure, GOR, heavy oil viscosity and specific gravity, on the reservoir productivity. The results are shown in sub-plot format in Figure 4.

The productivity of the reservoir was further investigated by varying the reservoir pressure and the injection pressure while keeping the GOR constant at 100 scf/STB. With the injection pressure below 4,000 psig, the production forecast indicated that the recovery was unlikely when the reservoir pressure varied between 2,000-4,000 psig.

However, at 1,000 psig reservoir pressure, the heavy oil recover was achievable when the injection pressure was below 3,000 psig. The period of zero production was shortened as the reservoir pressure increased; however, no heavy crude displacement occurred before the reservoir pressure was sufficient to push the heavy crude toward the production well up to the surface facilities located 4 km above the reservoir.

In another case, the production forecast is shown in Figure 4 for a constant reservoir pressure (4,000 psig) and various GORs. The maximum heavy oil production occurred at the lowest GOR (100 scf/STB), and the peak production decreased as the GOR reduced. At 3,000 psig injection pressure, the heavy oil recovery – although very unstable – was also possible, but only when the GOR was 400-500 scf/STB.

By considering different heavy oil specific gravities varying from 10 degrees API to 18 degrees API, a reasonably good range of most common types of heavy oil was taken into account. Heavy oil viscosity is known to vary between 100 cP and 10,000 cP. The viscosity effect has been assessed covering from 10-10,000 cP. The results are also shown in Figure 4, and reveals that the production was spontaneous as soon as the injection was initiated for 10 cP. Production started two months after the injection was initiated when the viscosity was 100 cP and production started a year later when the viscosity was 1,000 cP and 10,000 cP. This suggests that the heavy oil recovery is appreciably influenced by the reservoir properties, the fluids’ interaction and the mixing process, and other thermodynamic effects that effectively enable the dynamic of the fluids within the reservoir up to the surface.

Conclusions
On the basis of the results of this investigation, offshore cold heavy oil recovery is certainly achievable using the CO-2-EOR technique, with CO-2 injected from a remote onshore source. Under such conditions, CO-2 is transported in dense phase because a high injection pressure is required to account for the frictional losses along the transmission line and to create the enhancement required that will move the heavy crude from the reservoir to the topside surface facilities. Both immiscible and miscible conditions were evaluated and it was clear under the conditions investigated that the miscible process was more efficient and pragmatic than the immiscible process.

Low heavy oil reservoir pressure with low GOR can be very costly to optimize due to production “hold back” and the tremendous energy (injection pressure) required to initiate recovery. The higher the viscosity, the longer the mixing process within the reservoir, and the higher the required injection pressure to facilitate the movement of the heavy oil within the reservoir to the production platform. When the initial reservoir pressure was as high as 4,000 psig with 500 scf/STB (GOR), heavy oil production was instantaneous as soon as CO-2 injection was initiated.

It is believed that during cold heavy oil production using the CO-2-EOR technique, part of the injected CO-2 must be trapped in the heavy oil reservoir by various means, while a considerable volume of the injected CO-2 must undoubtedly return with the produced heavy oil to the topside production facilities. The subject of CO-2 sequestration is being investigated as part of the research interests of the Well Engineering Research Group at The Robert Gordon University.

Authors
Eric Tchambak, Babs Oyeneyin and Gbenga Oluyemi are with the Well Engineering Research Group in the School of Engineering at The Robert Gordon University, Aberdeen, UK.

References
1 Renfro, J.J., “Sheep Mountain Production Facilities – A Conceptual Design”, J. Petr. Tech., November 1979, p1462-1468.

2 Kolodziejski, J., “ Petrobras Approves First Offshore Heavy Oil Development,” Dow Jones Newswires. June 17, 2009, http://www.rigzone.com/news/article.asp?a_id=77336&rss=true.

3 Hon Vai Yee, Nor Idah Kechut, and Wan Nurul Adyani W Razak, (2007) “Enhanced-Oil-Recovery Potential of Heavy-Oil Fields in Africa” PETRONAS Group Research SPE 108513-MS 2007.

4 Srivastava, Raj K., Sam S. Huang and Mingzhe Dong, (1999) “Comparative Effectiveness of CO2 Produced Gas and Flue Gas for Enhanced Heavy-Oil Recovery,” Saskatchewan Research Council SPE 56857-PA 1999 95.80 % Relevance.

5 Srivastava, R.K., S.S. Huang, S.B. Dyer and F.M. Mourits, ‘‘Heavy Oil Recovery by Subcritical Carbon Dioxide Flooding,’’ paper SPE 27058 presented at the 1994 Third Latin American and Caribbean Petroleum Engineering Conference ~LACPEC, Buenos Aires, 26–29 April.

6 Hornbrook, M.W., (1991) “Effects of CO2 Addition to Steam on Recovery of West Sak Crude Oil,” Dehghani, Kaveh, U. of Alaska; Qadeer, Suhail, Stanford U.; Ostermann, R.D., Ogbe, D.O., U. of Alaska. SPE 18753-PA 1991.

7 Ahmed T., (1997) “A Generalized Methodology for Minimum Miscibility Pressure,” SPE 39034-MS. Latin American and Caribbean Petroleum Engineering Conference, 30 August-3 September 1997, Rio de Janeiro, Brazil.