Producing Light Oil from a Frozen Reservoir: Reservoir and Fluid Characterization of Umiat Field, National Petroleum Reserve, Alaska
Umiat oil field is a light oil in a shallow, frozen reservoir in the Brooks Range foothills of northern Alaska with estimated oil-in-place of over 1 billion barrels. Umiat field was discovered in the 1940’s but was never considered viable because it is shallow, in the permafrost, and far from any transportation infrastructure. The advent of modern drilling and production techniques has made Umiat and similar fields in northern Alaska attractive exploration and production targets. Since 2008 UAF has been working with Renaissance Alaska Inc. and, more recently, Linc Energy, to develop a more robust reservoir model that can be combined with rock and fluid property data to simulate potential production techniques. This work will be used to by Linc Energy as they prepare to drill up to 5 horizontal wells during the 2012-2013 drilling season. This new work identified three potential reservoir horizons within the Cretaceous Nanushuk Formation: the Upper and Lower Grandstand sands, and the overlying Ninuluk sand, with the Lower Grandstand considered the primary target. Seals are provided by thick interlayered shales. Reserve estimates for the Lower Grandstand alone range from 739 million barrels to 2437 million barrels, with an average of 1527 million bbls. Reservoir simulations predict that cold gas injection from a wagon-wheel pattern of multilateral injectors and producers located on 5 drill sites on the crest of the structure will yield 12-15% recovery, with actual recovery depending upon the injection pressure used, the actual Kv/Kh encountered, and other geologic factors. Key to understanding the flow behavior of the Umiat reservoir is determining the permeability structure of the sands. Sandstones of the Cretaceous Nanushuk Formation consist of mixed shoreface and deltaic sandstones and mudstones. A core-based study of the sedimentary facies of these sands combined with outcrop observations identified six distinct facies associations with distinctive permeability trends. The Lower Grandstand sand consists of two coarsening-upward shoreface sands sequences while the Upper Grandstand consists of a single coarsening-upward shoreface sand. Each of the shoreface sands shows a distinctive permeability profile with high horizontal permeability at the top getting progressively poorer towards the base of the sand. In contrast, deltaic sandstones in the overlying Ninuluk are more permeable at the base of the sands, with decreasing permeability towards the sand top. These trends impart a strong permeability anisotropy to the reservoir and are being incorporated into the reservoir model. These observations also suggest that horizontal wells should target the upper part of the major sands. Natural fractures may superimpose another permeability pattern on the Umiat reservoir that need to be accounted for in both the simulation and in drilling. Examination of legacy core from Umiat field indicate that fractures are present in the subsurface, but don't provide information on their orientation and density. Nearby surface exposures of folds in similar stratigraphy indicate there are at least three possible fracture sets: an early, N/S striking set that may predate folding and two sets possibly related to folding: an EW striking set of extension fractures that are parallel to the fold axes and a set of conjugate shear fractures oriented NE and NW. Analysis of fracture spacing suggests that these natural fractures are fairly widely spaced (25-59 cm depending upon the fracture set), but could provide improved reservoir permeability in horizontal legs drilled perpendicular to the open fracture set. The phase behavior of the Umiat fluid needed to be well understood in order for the reservoir simulation to be accurate. However, only a small amount of Umiat oil was available; this oil was collected in the 1940’s and was severely weathered. The composition of this ‘dead’ Umiat fluid was characterized by gas chromatography. This analysis was then compared to theoretical Umiat composition derived using the Pedersen method with original Umiat fluid properties published in the original reports. This comparison allowed estimation of the ‘lost’ light hydrocarbon fractions. An Umiat 'dead' oil sample then could be physically created by adding the lost light ends to the weatherized Umiat dead oil sample. This recreated sample was recombined with solution gas to create a 'pseudo-live' Umiat oil sample which was then used for experimental PVT and phase behavior studies to determine fluid properties over the range of reservoir pressures and temperatures. The phase behavior of the ‘pseudo-live’ oil was also simulated using the Peng- Robinson equations of state (EOS). The EOS model was tuned with measured experimental data to accurately simulate the differential liberation tests in order to obtain the necessary data for reservoir simulation studies, including bubble point pressure and oil viscosity. The bubble point pressure of the reconstructed Umiat oil is 345 psi, suggesting that maintenance of reservoir pressures above that pressure will be important for the any proposed production technique. A major part of predicting how the Umiat reservoir will perform is determining the relative permeability of oil in the presence of ice. Early in the project, UAF work on samples of the Umiat reservoir indicated that there is a significant reduction in the relatively permeability of oil in the presence of ice. However, it was not clear as to why this reduction occurred or where the ice resided. To explore this further, additional experimental and theoretical work was conducted. Core flood experiments were performed on two clean Berea sandstone cores under permafrost conditions to determine the relative permeability to oil (kro) over a temperature range of 23ºC to - 10ºC and for a range of connate water salinities. Both cores showed maximum reduction in relative permeability to oil when saturated with deionized water and less reduction when saturated with saline water. This reduction in relative permeability can be explained by formation of ice crystals in the center of pores. Theoretically, the radius of ice formed in the center of the pore can be determined using the Kozeny–Carman Equation by assuming the pores and pore throats as a cube with ‘N’ identical parallel pipes embedded in it. Using the values of kro obtained from the experimental work as input to the Kozeny–Carman Equation at -10ºC, the radius of ice crystals dropped from 0.145 μm to 0.069 μm when flooding-water salinity is increased to 6467 ppm. This explains the reduction of relative permeability with decreasing salinity but does not take into consideration other effects such as variations in pore throat structure. In addition, fluids like deionized water, saline water, and antifreeze (a mixture of 60% ethylene or propylene glycol with 40% water) were tested to find the best flooding agent for frozen reservoirs. At 0ºC, 9% greater recovery was observed with antifreeze was used as a flooding agent as compared to using saline water. Antifreeze showed 48% recovery even at -10ºC, at which temperature the rest of the fluids failed to increase production. Preliminary evaluation of drilling fluids indicate that the brine-based muds caused significantly less swelling in the Umiat reservoir sands when compared to fresh-water based muds. However since freezing filtrate is another cause of formation damage, a simple water-based-mud may not a viable option. It is recommended that new fluids be tested, including different salts, brines, polymers and oil-based fluids. These fluids should be tested at low temperatures in order to determine the potential for formation damage, the fluid properties under these conditions and to ensure that the freezing point is below that of the reservoir. In order to reduce the surface footprint while accessing the maximum amount of the Lower Grandstand interval, simulations used development from 5 surface locations with a wagon-wheel pattern of multilateral injectors and producers. There is no active aquifer support due to small peizometric head in the area and no existing gas cap, so an alternative method of pressure support is needed. Cold gas injection was used in the simulations as it is considered the most viable means of providing pressure maintenance while maintaining wellbore stability and reducing impact on the permafrost. Saline water injection may be a viable alternative, though this may have a detrimental effect on permafrost. In the short term, the results of this work are being incorporated into Linc Energy’s drilling and development plan. This project has also provided valuable information on the rock and fluid properties of low temperature reservoirs as well as the efficacy of potential production techniques for Umiat or similar shallow frozen reservoirs in the circum-Arctic.
- Research Organization:
- University Of Alaska Systems
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- NT0005641
- OSTI ID:
- 1080462
- Country of Publication:
- United States
- Language:
- English
Similar Records
Chemical Methods for Ugnu Viscous Oils
PH Sensitive Polymers for Improving Reservoir Sweep and Conformance Control in Chemical Flooring