You have access to this full article to experience the outstanding content available to SPE members and JPT subscribers.
To ensure continued access to JPT's content, please Sign In, JOIN SPE, or Subscribe to JPT
Although polymer flooding has become a promising enhanced oil recovery (EOR) technique, no field tests have been performed to date in Alaska’s underdeveloped heavy-oil reservoirs. The overall objective of the research outlined in the complete paper is to perform a field experiment to validate the use of polymer flooding in this challenging environment. At the time of writing, no unexpected injectivity issues or polymer breakthroughs have been encountered and the two horizontal producers are showing positive response to the polymer injection, resulting in incremental increase in oil production rate.
Introduction
The Alaska North Slope (ANS) contains vast resources of heavy oils, primarily concentrated in the West Sak (also called Schrader Bluff) and Ugnu reservoirs. There are currently six fields producing heavy oil in Alaska: Orion, Polaris, Milne Point, Tabasco, Kuparuk, and Nikaitchuq. Tabasco is a Kuparuk River Unit (KRU) satellite, and the Ugnu formation overlies the West Sak/Schrader Bluff formation across the North Slope fields. The estimated total oil in place within these reservoirs is approximately 20 billion–25 billion bbl, with approximately two-thirds of the heavy oil lying under the KRU. At present, oils from the West Sak/Schrader Bluff formation are being developed.
Despite the fact that heavy oil represents approximately a third of known ANS original oil in place (OOIP), the development pace has been slow, with cumulative production contributing only 1% of the OOIP slopewide. High development costs, significant logistical and environmental challenges, and low oil recovery resulting from conventional techniques have been the major factors for underdevelopment of these vast resources. Thermal recovery methods, such as steam injection, are impractical on the ANS because of the high cost and, more importantly, the concerns associated with thawing the nearly 2,000 ft of permafrost, which could cause massive environmental damage.
From a technology standpoint, preliminary laboratory and simulation studies have indicated that polymer flooding has great potential to enhance oil recovery from the Schrader Bluff reservoirs, but these have yet to be tested because of the lack of field tests. In fact, no large-scale polymer flood of heavy-oil or other unconventional resources has occurred to date in the United States, although it has been tested and implemented in other countries such as Canada and China. Initial scoping studies suggest that successful implementation of polymer flooding could increase heavy oil recovery by 50% on the ANS.
Description of the Polymer Field Pilot Area and Test Wells
The project area, the Milne Point Unit (MPU), is between the Prudhoe Bay Unit (PBU) and the KRU. The MPU is approximately 30 miles northwest of the PBU and 15 miles northeast of the KRU. The current working interest owners of the MPU are the operator, 5Hilcorp Alaska (50%), and BP Alaska (50%).
The pilot area contains two horizontal injectors and two horizontal producers drilled into the Schrader Bluff. The reservoir characteristics in the pilot area are favorable because they are amenable to polymer injection with formation porosity in the range of 30 to 35%, permeability from 100 to 3000 md, a low reservoir temperature of 70°F, oil API of approximately 15°, and in-situ oil viscosity of approximately 300 cp. The oil is undersaturated.
Results and Discussion
Implementation of Polymer Field Pilot. Since the end of August 2018, polymer injection has continued as planned for the most part. However, two separate operational events—as might be expected in a field—in late September and November caused interruptions in the injection. One was a temporary shutdown caused by the presence of more hydrocarbon gas found in the source water than expected. The second was the result of injection pump repairs and shut-in for pressure-falloff tests.
In order to achieve a target viscosity of 45 cp, polymer concentration has varied between 1,600 to 1,800 ppm. In considering the pre- and post-polymer injection response of the two producers, J-27 and J-28, it should be noted that J-27 is supported by two injectors, whereas J-28, drilled close to a sealing fault to the south side, is supported by only one injector. A decrease in total liquid production rate was caused by the decline in reservoir pressure caused by a lower injection than the production voidage. The positive effect of polymer injection was reflected mainly by the decrease in water cut from approximately 65% to 40–45% in both producers. The oil rate increased by 200 to 300 B/D post-polymer injection, although the total liquid rate was trending down.
Finally, there is no indication of polymer presence in the producers to date, which has been confirmed by both clay flocculation and nitrogen-fluorescence water-composition analyses.
Laboratory Corefloods. If polymer retention is greater than 50 µg/g, polymer retention can be more important to the economics of the process than polymer concentration or bank size. Therefore, in order to assess this factor for the subject set of rock and fluids and the polymer, five different retention experiments have been carried out at the time of writing. These tests are on 6- to 11‑darcy sandpacks created from representative sand from the MPU.
Results from the first three tests using native sand (stored for years without preservation) and freshly cleaned sand, using the conservative and perhaps more-accurate nitrogen-analysis method, revealed polymer-retention values in the range of 153 to 290 µg/g, considered very high. Ongoing experimentation is focused on understanding this high retention and monitoring the polymer pilot that allows the determination of field-based polymer-retention values. The authors’ most recent two tests, with cleaned reservoir sand saturated with fresh oil, yielded polymer retention values of only 28 µg/g—a much more palatable value.
Numerical Reservoir Simulation. Reservoir-modeling efforts have focused on assessing the polymer-retention effects using a 1D homogenous and 2D heterogeneous models in areal plane on laboratory sandpack experiments and simulations geared toward field-scale analyses. Capillary-pressure effects on polymer retention also were simulated using the experimentally determined polymer-retention values in oil sand, which indicate that, for the target heavy-oil reservoir with highly permeable oil zones and large-pore-size sand, no significant effects on polymer retention or fluid saturation using the 1D sandpack model were recorded. A simulation comparison using the retention values obtained from laboratory studies demonstrated that the polymer-absorption behavior agreed with the range of delayed polymer slugs from laboratory experimental results.
The 3D grid system of the initial reservoir-simulation model, as shown in Fig. 1, has been generated on the basis of the geological model, which was developed by combining seismic data, well logs, core data, and wellbore trajectories. Because the injector/producer pairs have been used for waterflooding, one of the primary objectives of reservoir simulation is to perform a waterflood history match and to update the reservoir model continually as sustained long-term production data from the polymer-flood pilot becomes available.
Conclusions
- After nearly 5 months of polymer injection in the two horizontal injectors, no injectivity issues or polymer breakthrough have been encountered.
- Considering the encouraging response of both producers, the authors state that they are cautiously optimistic with regard to the performance of the polymer field pilot.
- Initial polymer-retention laboratory tests in sandpacks indicated high values on cleaned sand. However, tests on cleaned sand resaturated with fresh oil yielded moderate polymer-retention values on the order of 28 µg/g.
- Low-salinity polymer flooding in a tertiary mode carried out on two sandpacks have demonstrated an improvement in the oil-recovery factor over a range of low- to high-salinity waterfloods.
- A good waterflood history match using the permeability strip model and power-law-estimated relative permeabilities has been achieved that establishes the reservoir-simulation model for future applications.