|Producing Light Oil from a Frozen Reservoir: Reservoir and Fluid Characterization of Umiat Field, National Petroleum Reserve, Alaska
||Last Reviewed 3/27/2013
The goal of this project is to develop a robust reservoir model to test possible oil recovery methods that do not use steam or a liquid capable of freezing for the Umiat and similar frozen reservoirs. The results will provide important information concerning production methods for this and similar frozen reservoirs in northern Alaska and other arctic regions.
University of Alaska Fairbanks, Fairbanks, AK 99775-7320
Linc Energy Alaska LLC (Linc)
The Umiat oil field in the Brooks Range foothills of northern Alaska contains light oil in a shallow, frozen reservoir. The Umiat field was discovered in the 1940s but was never considered viable because it is shallow, in the permafrost, relatively small, and far from any transportation infrastructure. Initial estimates of recoverable reserves in the Umiat field ranged from 30 to over 100 million bbl, with an average of about 70 million bbl. However, recent reserve estimates by private industry suggest that the accumulation may be considerably larger than originally thought, and modern horizontal drilling techniques now enable development of shallow reservoirs. This makes Umiat and similar fields in northern Alaska attractive exploration and production targets.
Little is known about how to produce conventional oil from a frozen reservoir. Most prior work has been on developing production techniques for heavy oil in unconsolidated but unfrozen sands, or for gas hydrates. There is no information available describing the behavior of a rock/ice/light oil system at low pressure. This information, along with a robust reservoir model, is needed to accurately model and evaluate the effectiveness of different production methods.
The Umiat field contains significant albeit unconventional (i.e., frozen) potential energy resources. This is a particularly attractive target considering the decline in the production of conventional oil resources from Alaska’s North Slope. Development of production methods and strategies for these shallow, unconventional resources will promote economically viable resource extraction. This project will encourage involvement of smaller exploration and production companies in Alaska by providing critical information not currently available to the public and by evaluating the applicability of existing production techniques. This information will increase the likelihood of successfully bringing smaller fields into production.
Develop sedimentologic and stratigraphic model of the reservoir interval
Description of core taken during the original exploration program in the 1940s and ’50s was completed during the summer of 2010. Two weeks of field work during June and July 2010 focused on describing outcrops of the Umiat reservoir in the vicinity of Umiat field. Samples were also collected for diagenetic studies and pore structure analysis. Horizontal and vertical permeability was measured in major lithofacies in both core and outcrop.
Results of this work indicate that the Umiat reservoir is a complex system of shallow marine and distributary mouth bar sands with a strong vertical permeability anisotropy. Permeability patterns can be linked to the identified lithofacies and facies associations, enabling prediction of permeability and permeability anisotropy across the field. This information was incorporated into the reservoir model used for simulation of the reservoir under proposed production scenarios.
Develop structural model of reservoir
Examination of the existing Umiat core during the summer of 2009 showed that natural fractures occur in the reservoir, but were previously not recognized because they are sparse, widely spaced and/or steeply dipping. To address this issue, more emphasis was placed on developing a structural model for the distribution of fractures by examining their structural context in more detail. Fieldwork conducted during June and July 2010 examined both the structural geometry and the distribution of fractures in the Umiat structure and in a similar structure that is nearby and well-exposed, the Big Bend anticline.
Results of this work indicate that there are at least two major fracture networks present in the Umiat structure: one approximately perpendicular to the structure and a second approximately parallel to the structure. These fractures could play a major role in enhancing vertical permeability and vertical connectivity in a reservoir that is otherwise highly compartmentalized.
Characterization of reservoir and fluid properties
An experimental apparatus was designed and constructed to conduct relative permeability experiments of oil and gas in the presence of ice. Experiments on both Berea Sandstone and samples of Umiat core over a range of temperatures and salinities demonstrated a significant reduction of oil relative permeability at irreducible water saturation caused by freezing of connate water. For samples of the Umiat reservoir saturated with deionized water and frozen, the average decline was 61%. These samples were subsequently determined to be representative of the less permeable facies of the reservoir; more permeable facies may not experience such a dramatic decline in relative permeability. These conclusions were supported by experiments on Berea sandstone samples with permeabilities similiar to that of the main reservoir facies. In these samples, continuous reduction in relative permeability was observed when the temperature was decreased from 23°C to -10°C, with an average reduction of 43.2% when saturated with deionized water and 32.5% when saturated with water with salinities approximating those in the Umiat reservoir. This suggests that the reduction of the relative permeability is strongly dependent on both the reservoir temperature and salinity of connate water.
The phase behavior of the Umiat fluid needs to be well understood in order to produce an accurate reservoir simulation. However, only a small amount of Umiat oil was available; this oil was collected in the 1940s and was severely weathered. The composition of this ‘dead’ Umiat fluid was characterized by gas chromatography and then compared to a theoretical Umiat composition derived using the Pedersen method and Umiat fluid properties published in the original reports. This comparison allowed estimation of the ‘lost’ light hydrocarbon fractions. A ‘pseudo-live’ reservoir oil sample was then physically created by adding the lost light ends to the weatherized Umiat dead oil sample and used for experimental PVT and phase behavior studies to determine fluid properties over the range of reservoir pressures and temperatures. The experimental results were used to tune a Peng-Robinson equation of state model so that the phase behavior of the reconstructed Umiat oil at reservoir conditions could also be simulated.
Analyses also indicated that the total asphaltene content measured for Umiat dead oil is not significant, which suggests that Umiat oil does not have a high propensity for asphaltene precipitation. Lower asphaltene precipitation will lead to smooth production, no pipeline deposition and plugging (resulting in less pressure drop along the pipeline), and no change in composition.
Desktop reservoir simulation of proposed production strategies
A reservoir property model was constructed based on published data from Umiat field and subsequently modified to incorporate the new geologic information regarding sand continuity and permeability anisotropy. These observations and petrophysical data indicate that the two major reservoir intervals, the Upper and Lower Grandstand, have distinct petrophysical properties and flow structures and should be treated separately.
A single realization of this geologic model using average observed porosity and Sw values yielded an estimated original-oil-in-place (OOIP) of approximately 1.52 billion barrels with 99 bcf of associated gas. A Monte Carlo simulation was conducted to evaluate the sensitivity of this OOIP value to a range of porosities, Swi, bulk volume, net-to-gross thickness ratio, formation volume factor, and gas content. The results of this simulation yielded OOIP estimations ranging from P10 of 750 million barrels, a P50 of 1550 million barrels, and a P90 of 2475 million barrels.
The geologic model was gridded and initialized with the observed rock and fluid property values. The two major reservoir intervals, the Upper and Lower Grandstand sands, are being modeled separately because of different reservoir properties and computational time. Based on the geologic results, OOIP estimates and engineering constraints, initial drilling by Linc will focus on the Lower Grandstand, and it was the focus of the simulation efforts.
A wagon-wheel pattern was used in the simulation as the most efficient means of accessing the maximum amount of the reservoir interval while minimizing the surface footprint. In the proposed pattern, one vertical well in the center along with two dual lateral injectors in the north and south at the top of the reservoir supports pressure for a combination of four dual lateral producers at the bottom of the interval, each one angled at 36 degrees in a square mile spike configuration. The wells have a total length of 3000 ft., with a 1500 ft. horizontal leg and a 2.5 inch open hole completion across the productive area. To reduce surface impact and the cost of infrastructure, only five pad locations were used in the simulation, with a total combination of 85 producers and 25 injectors. Three scenarios were evaluated in the simulation: no gas injection and cold gas injection with bottom hole injection pressures (bhip) of 400 and 600 psi.
Simulation results indicate that recovery will be very sensitive to injection pressure and permeability anisotropy. The higher the injection pressure, the higher the average reservoir pressure and the more oil production. Recovery over 50 years of gas injection at 400 psia bhip is ~12%; 14% for 600 psia bhip; and 8% for no gas injection. Higher permeabilitity anisotropies reduce the effectiveness the cold gas injection, reducing recovery.
Throughout this project, UAF researchers worked closely with industry partners (initially Renaissance Alaska and later Linc Energy), sharing both data and interpretations. Teleconferences involving the entire research team were held quarterly.
During fall 2011, all UAF faculty and students participated in a weekly one hour seminar focusing on the implications of the ongoing geologic and engineering research on Umiat reservoir performance, drilling optimization, and production. Renaissance Alaska personnel participated in the seminars via teleconference.
Throughout the project, members of the team presented talks at the national conferences, including the American Association of Petroleum Geologists (AAPG) national meeting; the AAPG International Conference and Exhibition; the 2011 Arctic Technology Conference; and the 2012 Society of Petroleum Engineering (SPE) annual meeting.
As of December 2012, four M.S. theses have been completed. A Ph.D. dissertation and four additional M.S. theses are nearing completion.
Linc Energy is planning the first well in the Umiat field for the 2012/2013 drilling season.
Current Status (March 2013)
The project has been completed. The final report is available below under "Additional Information".
Project Start: October 1, 2008
Project End: December 31, 2012
DOE Contribution: $1,350,652
Performer Contribution: $1,530,000
NETL - Chandra Nautiyal (firstname.lastname@example.org or 281-494-2488)
UAF – Catherine Hanks (email@example.com or 907-474-5562)
If you are unable to reach the above personnel, please contact the content manager.
Final Project Report [PDF-47.4MB]