It’s Never Simple: Controls on the Complicated Flow Pathways of the Wahoo Reservoir (Early Pennsylvanian) Lisburne Field, North Slope Alaska, USA
James R. Markello¹, Richard J. Wachtman², Aaron R. Liesch³, Richard J. Paterson4, Hugh Nicholson5, David L. Boyer6, Mark Swanson5, Robert Krantz7, Ned Frost8, Anastasia Mironova7, and Melissa Hicks9
¹ExxonMobil Upstream Research Company, Houston, TX
²ExxonMobil Production Company, Houston, TX
³ExxonMobil Development Company, Houston, TX
4ExxonMobil International Limited, Epsom, UK
5BP Alaska, Anchorage, AK
6Petrotechnical Resources of Alaska, Chugiak, AK
7ConocoPhillips, Houston, TX
8Bureau of Economic Geology, University of Texas, Austin, TX
9Department of Geology Syracuse University, Syracuse, NY
Fluid flow in the Wahoo Reservoir of Lisburne Field is differentially controlled by matrix reservoir quality, reservoir architecture, and overprinting of faults and fractures across the field. To understand and predict specific fluid flow pathways and field performance, the multi-company team completed an updated reservoir characterization with a new sequence stratigraphic-based reservoir architecture and fracture assessment by reinterpreting cores, logs, and seismic within the context of Late Pennsylvanian regional and global controls of tectonics, climate, eustasy, ocean circulation and geologic history. The new interpretations were the bases for populating new sector and full-field geologic models with matrix and fracture properties. These geologic models were dynamically simulated and achieved reasonable performance history matches.
The multi-billion barrel OOIP Lisburne Field (27.5km max length; 14.2km max width; ~360km²) was discovered in 1968 by the Prudhoe Bay State #1 well. The field is an oil accumulation (27°API gravity) with gas-cap, and is operated by BP with ExxonMobil, ConocoPhillips, and Chevron as working interest owners. It consists of the Wahoo (Early Pennsylvanian) and Alapah (Late Mississippian) mixed carbonate-clastic-evaporite reservoirs. To date, six delineation wells and 95 development wells have been drilled from six surface pads (L1, L2, L3, L4, L5, and LGI-1). As of 2008, 17 wells are active producers and 3 wells are gas re-injectors. Field production is by pressure depletion and gas-cap expansion. Total withdrawals have not been volumetrically replaced by gas re-injection, so current field pressure is about half of original pressure (4300psi). First oil was produced in 1985. Field production peaked in 1988 at about 47KBD and declined to under 10KBD by 1997. Since then, production has varied from 5 to 13KBD. Well tests, spinner surveys, and temperature logs have been run in various wells, and tracer experiments were conducted between wells of the L2 drilling pad. L2 pad was the location for a water injection pilot from 1988-90, which was gave mixed results. Water injection pilots commenced in 2011 for pressure maintenance and secondary recovery in the L3 and L5 pads.
The updated 2009-2010 reservoir description included new depositional, diagenetic and stratigraphy syntheses that incorporate 1) known global-scale geologic constrains of Early Pennsylvanian-age Pangean-assembly tectonics, paleogeography-directed warm vs. cold-water ocean circulation, climate-driven high-amplitude high-frequency icehouse eustasy, and 2) concepts of sequence stratigraphy, the stratigraphic hierarchy, syn-tectonic differential subsidence on an actively deforming distally-steepened ramp, mixed carbonate-clastic depositional facies systems with reciprocal and coeval sedimentation styles, and tectonic-driven surface and burial episodes of diagenesis.
Within the Lisburne Field area, the Wahoo Formation unconformably overlies the Alapah Formation with minimal time gap. The Wahoo is dated to be Early Pennsylvanian (Bashkirian or Morrow/Atoka) age. It is unconformably overlain by the Late Permian Kavik Shale. This top unconformity represents a time gap of up to 50 million years, and is interpreted to be related to tectonic uplift and erosion associated with the development of the Late Paleozoic Pangean supercontinent. In the field area, the Wahoo Formation is subdivided into seven (7) lithostratigraphic carbonate units (ranging from 0 m thick due to post-depositional erosion to upwards of 30 m thick) that are simply numbered from 1 at base to 7 at top. The carbonate units are separated by field-wide to regionally extensive shale units (1 to 7 m thick) of green to light gray and medium to dark gray colors. The basal Wahoo shale is informally named the Green Shale (18 to 20 m thick). Across the Lisburne Field area, the Wahoo Formation ranges in thickness from 230 to 275 m. To the north, the Wahoo is absent due to pre-Permian erosion. To the east, the Wahoo is absent due to Early Cretaceous erosion. To the south toward the Brooks Range, the Wahoo thickens to more than 1500m due to Early Pennsylvanian differential subsidence and deposition in the basinward direction. Interestingly, the carbonate units thicken into the basin to the south; whereas, the shales thin and pinch-out into the basin.
Based on core descriptions from 17 wells across the field, a diverse assemblage of depositional facies was recognized and grouped into 11 major depositional environments: 1) green, gray and black shale, silty shale, and shaly skeletal mudstone/wackestone (Inner Ramp Subtidal Lagoon); 2) siltstone, skeletal siltstone, shaly/silty skeletal peloidal lithoclast grainstone to packstone (Inner Ramp Subtidal Channel); 3) mm-cm scale current and climbing rippled, flaser laminated, un-burrowed green shale, silty shale, calcareous shale (Inner Ramp Intra-Lagoon Tidal Bars); 4) bi-directional and current large-scale trough cross-bedded intraclast skeletal conglomerate, grainstone, mixed silty skeletal packstone/grainstone (Inner Ramp Tidal Pass); 5) fenestral, mm-laminated, desiccation cracked lime mudstone/wackestone (Inner Ramp Back-Barrier Tidal Flats); 6) Intraclast breccia (karst) and grainstone with rooted soil profiles, geopetally filled vugs and cracks (Barrier Island); 7) echinoderm, bryozoan, coral grainstone, ooid/coated-grain grainstone (Beach/Foreshore); 8) wave-rippled, trough cross laminated, diverse skeletal (echinoderms, bryozoa, fusulinid forams, brachiopods, corals, molluscs) grainstone to packstone (Upper Shoreface); 9) diverse skeletal grainstones, packstones, wackestones, and mudstones with hummocky cross lamination, firmgrounds, hardgrounds, horizontal and vertical burrows (Lower Shoreface); and 10/11) skeletal peloidal wackestone, skeletal mudstone with bryozoa, echinoderms and sponges, abundant burrows, firmgrounds, suspension lamination (Outer Ramp to Open Shelf). It must be noted that this facies compilation is from all wells and all zones. Importantly, not all facies occur within each zone of every well. As an example, in the L3 wells, there are no occurrences of exposure surfaces and of thick in-place ooid/coated grainstone shoals. Further the shales are thin, highly burrowed with significant carbonate skeletal content. Contrastingly, in L1, L2, L4, and L5 wells there are well defined exposure surfaces and karsts with local thick ooid/coated grainstone shoals and tidal flat mudstones. Further, some of the shales in the L1 and L5 wells display both current ripple and flaser lamination as well as rooted soil profiles. Although these facies were observed within the field area and EODs were interpreted for each facies, extrapolations were made for interpreted EODs outside the field area to complete the full depositional profile. For this entire profile, six depositional systems were defined: 1) provenance uplands, 2) fluvial, 3) inner ramp tidal lagoon/estuary – 0 to 10m water depth, 4) inner ramp – +10m elevation to 60m water depth, 5) outer ramp – 60 to 200m water depth, and 6) basin – >200m water depth. These six were subdivided further into 16 major depositional environments (11 are characterized above) that complete the profile.
Integrating these key facies details and interpreted local EODs with regional and global-scale paleogeography enabled development of a new Lisburne field area Wahoo depositional profile and paleogeographic setting. They are interpreted to be a syn-tectonic, asymmetrically crenulated, distally steepened ramp with updip meta-sedimentary basement highlands to the north (present-day direction) that were sources for the clastics and a rapidly subsiding basin to the south. The field area occupies the inner/outer ramp shallow marine transition location between the clastic provenance uplands to deep marine basin. Across the field, six “informally named” local depositional paleogeographic areas have been defined: central embayment (L3 pad area); west embayment flank (L2 pad area); east embayment flank (L4 pad area); northeast ramp (L5 pad area); northwest ramp (L1 and LGI pads area); and north central ramp (area between L1 and L5 pads). Spatial orientation and geometry of these local paleogeographic areas, interpreted water depths of carbonate facies across the constructed depositional profile, recognition that the field-wide shales cut across all EODs and were deposited in all water depths, coupled with principles of sequence stratigraphy and the global constraint of Bashkirian-age high-amplitude icehouse eustasy are critical for development of the new Wahoo reservoir architecture. Subtle reactivation of preexisting basement fault trends influenced the paleogeography and facies distributions.
Overall, the Wahoo reservoir consists of seven carbonate-dominated depositional units separated by shales. Palynology demonstrates that each shale interval is a separate time unit. Sedimentologically, the shales and associated siltstones are land-derived clastics deposited across a carbonate ramp, and thus represent a significant lowering of sea level. Re-establishment of the carbonate factory across the ramp and retreat of clastics to the hinterland represents significant sea level rise. Together, these observations are used to define a third-order depositional sequence: sequence boundary at base of shale, lowstand systems tract consisting of shale and associated siltstones and shaly carbonates, transgressive systems tract consisting of clastics interfingered with and overlain by carbonates, highstand systems tract consisting of carbonates, and top carbonate sequence boundary overlain by next youngest shale. The Wahoo reservoir is interpreted to consist of seven (7) third-order sequences: Zone 1 through Zone 7. Based on vertical trends of thickness, facies proportions, and exposure abundances, the seven third-order sequences are grouped into two second-order supersequences: Zones 1-4 and Zones 5-7. Zones 4 to 7 were divided into fourth-order high-frequency sequences (hfs): Zone 4 – 9 hfs; Zone 5 – 5 hfs; Zone 6 – 4 hfs; and Zone 7 – 4 hfs. Zones 1 to 3 were not subdivided because they are in the water leg.
Importantly, 21 of the 22 high-frequency sequences were subdivided into systems tracts (hfs 7.3 was not subdivided). Observations of numerous exposure surfaces within the carbonate intervals updip and outside of the central embayment and of none identified within the central embayment, plus the interpretation of 75 m amplitude high-frequency eustatic sea level oscillations result in the geometric construction of central embayment-restricted hfs lowstands with even conservative ramp dips of only 2°. Thus, transgressive and highstand units are best developed outside of the central embayment with only thin, mud-dominated basinal equivalents. Further, it appears that the west and northwest embayment flanks experienced a different subsidence rate than the east and northeast embayment flanks because interval thicknesses, facies types (including clastics), and number of exposures are different between them. This sequence stratigraphic analysis and architecture not only defines systems tracts that pinch-out, but as important, the beds within the parasequences that comprise the systems tracts also have limited areal extent. Depositional strike is not uniformly straight from east to west, but rather is curved and sinuous paralleling the central embayment rim. Depositional dip also is not unidirectional to the south, but rather appears to be convergent toward the center of the central embayment. The net results of the new depositional model and stratigraphic reservoir architecture are that the starting conditions for fluid flow are both heterogeneous and discontinuous across the field.
Post-depositional reservoir quality-destructing and enhancing diagenesis includes multiple episodes of cementation, meteoric leaching and dolomitization. Reservoir quality is heterogeneous across the field with most of the reservoir facies being limestones that have very poor to moderate quality (porosity range: 0.1 – 20% and avg: ~6%; permeability range: 0.01 – 20md and avg: ~0.2md); reservoir facies with porosities > 10% and permeabilities >20md are typically leached sucrosic dolostones. Dolostones with k>50md typically have 25-30% porosity and these are less abundant than the low permeability dolostones. Dolomitization is restricted to thin units within high frequency sequences, but is not restricted to specific systems tracts. Individual dolomite beds cannot be correlated field wide. Meteoric dissolution formed vugs and molds during times of high frequency exposure (e.g. Zone 4 Well L2-06). It is interpreted that larger-scale vugs and caves (e.g. Zone 7, Well L2-06) were formed during the 50 million year time gap of the top Wahoo unconformity. The high porosity and permeability features are also local. Lastly, further surface and subsurface aquifer-style dissolution occurred dominantly along the eastern side of the field during the time of Early Cretaceous erosion and exposure (e.g. Zones 7 and 6 in many L5 wells). Thus, these local to field-wide diagenetic effects were spatially overlapping, but time-separated. The net result has been to differentially enhance depositional continuity within high frequency sequences, and locally to establish new continuity between depositionally discontinuous units.
Superimposed on the depositional and diagenetic continuity and reservoir quality characteristics are post-depositional and syn- plus post-diagenetic tectonic folding, faulting and fracturing. The Ellesmerian tectonic event (Middle Pennsylvanian to Late Permian) created widely-spaced, low relief rift blocks across the central North Slope. The Beaufortian extensional event (Jurassic to Early Cretaceous) generated a rifted margin and rift shoulder uplift, defining the initial trap configuration for the Lisburne field (and overlying Prudhoe Bay field). Beaufortian extension also created the main east-west intra-reservoir normal fault set. The Brookian Orogeny (Late Cretaceous to Miocene) accommodated major thrusting in the Brooks Range to the south. Compressional pulses reached the Lisburne field, where Beaufortian extensional faults were re-activated during the Brookian Orogeny. Far field compression also created additional fracture sets. The current Wahoo reservoir subsurface structural and fault framework, based on seismic mapping, has east-west striking, discontinuous normal faults (some down-to-the north; others down-to-the-south) across the field. Subsurface core and image-log fracture and seismic attribute analyses document at least two populations of fractures within the Wahoo reservoir: ENE and NNW.
Fluid flow in the Lisburne Wahoo reservoir is clearly complex and controlled by a number of factors: 1) a complicated depositional facies mosaic with varying strike and dip directions, 2) overprinted diagenetic patterns that variably modify reservoir quality, 3) high frequency sequence stratigraphic architecture with beds and units of variable thickness and continuity, and 4) faults and fractures with varying direction and transmissibility. Results from (inter-well) pulse tests indicate that the NNW-SSE fracture trend is a preferred permeability direction in areas away from faults, reflecting flow through the aforementioned fracture set. There also seems to be improved E-W connectivity between wells close to or crossing E-W trending faults. The E-W faults also act as barriers/baffles in the opposite, N-S, direction, as a result of fault juxtaposition of stratigraphic units and/or cementation of the fault core. Thus, whilst there are some general conclusions that can be drawn from the field data, it is imperative to build drilling pad-specific sector flow models based on pulse test and tracer data that establish local connectivity between adjacent wells. Importantly, flow directions, connectivity lengths, and rates that can be defined for the Wahoo L-2 wells are not the same for those of the L-1, L-3, L-4, and L-5 wells. One size does not fit all. This reservoir must be developed with area specific strategies.
AAPG Search and Discovery Article #120034©2012 AAPG Hedberg Conference Fundamental Controls on Flow in Carbonates, Saint-Cyr Sur Mer, Provence, France, July 8-13, 2012