Upload
vocong
View
212
Download
0
Embed Size (px)
Citation preview
Copyright 2008, International Petroleum Technology Conference This paper was prepared for presentation at the International Petroleum Technology Conference held in Kuala Lumpur, Malaysia, 3–5 December 2008. This paper was selected for presentation by an IPTC Programme Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper, as presented, have not been reviewed by the International Petroleum Technology Conference and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of the International Petroleum Technology Conference, its officers, or members. Papers presented at IPTC are subject to publication review by Sponsor Society Committees of IPTC. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of the International Petroleum Technology Conference is prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgment of where and by whom the paper was presented. Write Librarian, IPTC, P.O. Box 833836, Richardson, TX 75083-3836, U.S.A., fax +1-972-952-9435.
Abstract The recent discovery of the giant Jansz/Io Gas Field in the Carnarvon Basin, which is considered to be mature from an exploration perspective, has challenged explorers to reconsider exploration strategies in mature basins. The field was discovered in 2000 by the Jansz-1 well and lies 250 km offshore from the north-west coast of Australia, in water depths ranging from 1100 to 1400 m. Jansz/Io is a structural/stratigraphic trap with the gas-bearing reservoir in an Upper Jurassic, mud-rich sandstone, up to 65 m thick, deposited in a shallow-marine depositional setting. Most shallow-marine hydrocarbon reservoirs produce from the upper- and lower-shoreface depositional facies. However, in Jansz/Io the highest quality reservoir is mainly in distal lower-shoreface to offshore depositional facies. Mud-rich offshore facies are generally considered to be non-prospective by explorers but in the right geological setting at a shallow burial depth, can be highly prospective for gas. At Jansz/Io, reservoir quality is lithofacies-dependent and controlled by a proximal or distal location within the depositional environment, with the highest permeability (greater than 100 md) in proximal sandstones with lowest clay content (less than 20%). Jansz/Io gas is a key component of the Greater Gorgon deepwater gas assets, and is a focus for development activity to meet an expanding global liquefied natural gas (LNG) market. The reservoir description for Jansz/Io is a key subsurface input to field development planning, requiring the integration of seismic and well data collected during the exploration and appraisal program conducted from 2000 to 2006. The scope of the study included the structural framework, field extent, reservoir architecture, depositional environment, reservoir properties, fluid composition, hydrocarbon contacts and uncertainty analysis. Geological models were constructed to capture the range of uncertainty in structural framework, reservoir properties and original gas-in-place (OGIP) for the Upper Jurassic reservoir. The model-based uncertainty analysis indicates the OGIP range for Jansz/Io has a low-side of 320 Gm3 and a high-side of 950 Gm3 (11 Tcf to 33 Tcf), with a p50 value of 630 Gm3 (22 Tcf). The geological setting of the Jansz/Io Gas Field posed a number of technical challenges, including depth conversion, mapping depositional facies and reservoir extent, thickness of seismically tuned reservoir, and prediction of reservoir properties away from well control. The challenges were solved by the subsurface team using a fully integrated multi-disciplinary approach, employing a wide range of geological and geophysical concepts and techniques to characterize the reservoir.
Introduction The Jansz/Io Gas Field, discovered in 2000 by the Jansz-1 well, lies in the Carnarvon Basin 250 km offshore from the north-west coast of Australia, in water depths ranging from 1100 to 1400 m. The recent discovery of this giant gas field in a basin considered to be mature from an exploration perspective, has challenged explorers to reconsider exploration strategies in mature basins. Jansz/Io gas is a key component of the Greater Gorgon deepwater gas assets, and a focus for development activity to meet an expanding global LNG market (Fig. 1). The field lies within the WA-18-R, WA-25-R and WA-26-R retention leases. The joint
IPTC 12461
Reservoir Definition at the Jansz/Io Gas Field, NW Shelf, Australia: A Case Study of an Integrated Project From Exploration to Development C.C. Jenkins, R.M. Chiquito, P.N. Glenton, A.A. Mills, J.G. McPherson, M.C. Schapper, and M.A. Williams, ExxonMobil
2 IPTC 12461
venture participants in WA-18-R are ExxonMobil (25%, operator), Chevron (50%) and Shell (25%). For WA-25-R and WA-26-R, the joint venture consists of Chevron (50%, operator), ExxonMobil (25%), Shell (12.5%) and BP (12.5%). The reservoir description for Jansz/Io is a key subsurface input to field development planning, requiring the integration of the seismic and well data collected during the exploration and appraisal program conducted from 2000 to 2006. The scope of the study included the structural framework, field extent, reservoir architecture, depositional environment, reservoir properties, fluid type, contacts, and uncertainty analysis. The Jansz/Io Gas Field has a number of geological features that distinguish it from more conventional gas fields in the Carnarvon Basin and, indeed, from many other producing fields around the world. Jansz/Io is a giant gas field in a structural/stratigraphic trap, with a large areal extent of 2,000 km2, and facies variation across the field. The exploration and appraisal wells have been drilled relatively high on structure and have not intersected the free water level (FWL). Conventional depth conversion methods have been unable to remove velocity artifacts caused by the complex overburden, masking the structural style for the field. The depth artifacts are severe and when combined with the FWL uncertainty, affect reservoir properties such as water saturation (because height above FWL affects water saturation), which adds further uncertainty to the original gas-in-place estimates (OGIP). While there are many hydrocarbon fields worldwide with reservoirs in a shallow-marine depositional setting, the difference between these fields and Jansz/Io is that the others usually contain hydrocarbons in high quality, clean sandstones from upper-shoreface and lower-shoreface facies. Typical examples include a number of Central North Sea oil fields such as the Fulmar Field (Johnson et al., 1986; Gowland, 1996) and the Wytch Farm Field in the Wessex Basin, onshore UK (Colter and Harvard, 1981; Morris et al., 2006). By contrast, mud-rich offshore facies are the primary reservoir at Jansz/Io, whereas these deposits have low quality and are non-productive in other fields. As for what makes this muddy sandstone reservoir work in Jansz/Io, this has been a key focus of the reservoir characterization study. The reservoir description of this fine-grained, mud-rich system required a multi-disciplinary and fully integrated approach. A variety of geophysical and geological methods, and geological concepts were applied, including depth conversion, reservoir thickness below seismic resolution, field extent, depositional setting, reservoir quality away from well control, and uncertainty analysis. The details of this integrated approach are the subject of this paper.
Database The subsurface database for the Jansz/Io Gas Field includes seismic and well data (Fig. 2). A total of five wells have been drilled through the Oxfordian gas reservoir at Jansz/Io, beginning in 2000 with the Jansz-1 discovery well. This was followed by Io-1 (2001), Jansz-2 (2002), Jansz-3 (2003) and Io-2 (2006). The wells were drilled using synthetic oil-based mud systems resulting in minimal borehole washout. High quality wireline-logs were recorded, including gamma ray, resistivity/induction, neutron, density and cross-dipole sonic logs, with checkshots or a vertical seismic profile (VSP) to provide velocity control. Pressure data and fluid samples were collected using wireline tools and magnetic resonance logs were recorded in all of the wells, with the exception of Jansz-1. Full-hole conventional core was cut in the reservoir section in each well and a total of 200 m of core was recovered from the field for reservoir characterization studies (Fig. 3 and Fig. 4). The core was initially sampled for routine and special core analysis and then cut into slabs for core description and photography. A separate set of core plugs was submitted for special core analysis to determine electrical properties, relative permeability measurements and capillary-pressure behaviour. A cased-hole production test was conducted at Jansz-3 in the highest quality reservoir (Fig. 3). The Jansz/Io 3D seismic survey (2892 km2) recorded in 2004 extends over most of the field (Fig. 2), including all of the core area. The 3D seismic was acquired and processed using anisotropic pre-stack time-migration (APSTM) to provide high temporal resolution (6 to 90 Hz band-width) and spatial resolution (18.75 m cross-line binning) suitable for detailed horizon and fault interpretation, and seismic inversion for rock properties (Hefti et al., 2006). The 3D survey was not recorded over the north-west margin of the field; however reprocessed 2D seismic data (3 km by 1.5 km grid) compliments the 3D survey to provide complete seismic coverage. The processing flow for the Jansz/Io 3D seismic survey is described in detail by Hefti et al. (2006). The near-angle stacks were used for well ties, horizon and fault framework mapping. The mid- and far-angle stacks were corrected for normal move-out stretch and relative Q attenuation, using the deterministic method of Lazaratos and Finn (2004), in order to be suitable for AVO analysis. A subset of the APSTM volume (350 km2 image area) within the core area of the field (Fig. 2) was processed using anisotropic pre-stack depth-migration (APSDM). Depth-migration was used to enhance the seismic image of the reservoir. It was primarily required for future planning of high-angle development wells, but also proved useful for validation of the depth-structure model derived from the APSTM data.
IPTC 12461 3
Overview The regional geology and the discovery history for Jansz/Io are discussed in Jenkins et al. (2003). The Jansz/Io hydrocarbon trap has an areal extent of 2,000 km2 with both structural (faulted anticline) and stratigraphic (reservoir pinchout) components. The structural component of the trap is defined by a north-west to south-east trending faulted anticline. The faults are extensional, were initiated by Early Jurassic rifting of the Carnarvon Basin (Veevers, 1988), and remained active until the Early Cretaceous (Valanginian). The fault throws are up to 25 to 30 m at the top-of-porosity marker with fault traces up to 4 km long. Seismic interpretation and core analysis indicates fault density and continuity is low at the macro- and micro-scale suggesting that fault segmentation may not be an issue for the development. The stratigraphic component of the trap is defined by the reservoir extent, which is limited by depositional downlap to the north-west, and erosional truncation by the Tithonian and base-of-Cretaceous Unconformities to the south-east. The reservoir is Upper Jurassic (Oxfordian) (Fig. 5) and comprises shallow-marine, muddy sandstones, deposited in distal lower-shoreface, offshore and open shelf settings. The marine mudstones of the Lower Cretaceous Barrow Group and Upper Jurassic Dingo Claystone provide the top-seal and base-seal to the reservoir, respectively (Fig. 6). The gas source is interpreted to be in the underlying deep, mature fluvial-deltaic section of the Upper Triassic Mungaroo Formation. The gas composition exhibits a high methane content (89 to 94 mol%), with a low condensate yield of 25 m3/Mm3 (4.4 bbl/MMscf) and a low CO2 content (0.1 to 0.3 mol%). The Jansz/Io play type contrasts with the nearby gas discoveries at Geryon-1, Urania-1, Orthrus-1 and Maenad-1A (Fig. 1), which are small, fault-dependent structural traps, with gas-bearing sandstones in the fluvial-deltaic reservoir section of the Mungaroo Formation (Korn et al., 2003). The reservoir is informally subdivided into the Upper Wedge (high quality) and Lower Wedge (low quality) reservoir units using the base-of-high-permeability chronostratigraphic marker to separate them (Fig. 3). The top-of-porosity marker defines the top of the Upper Wedge reservoir and the Oxfordian Unconformity defines the base of the Lower Wedge reservoir. The core area of the field is defined by the 15 m isochore contour surrounding the Jansz-1, Jansz-3, Io-1 and Io-2 wells. Outside the core area, the isochore is calculated using a de-tuning algorithm based on the product of seismic amplitude and apparent isochron. This area includes the Jansz-2 well (Fig. 4) and most of the north-west part of the field, and typically has a thickness less than 15 m (Fig. 7). The 3D seismic line (near-angle stack) that ties the Jansz-3 well is shown in Fig. 8 with the polarity convention annotated. The base-of-high-permeability marks the top of a progradational parasequence set that dips at a low angle (less than 0.5 degrees) to the north-west and subcrops under the base-of-Cretaceous Unconformity, to the south-east of the well. The second-prograde marker also dips to the north-west and subcrops the base-of-Cretaceous Unconformity to the south-east of the well. The Upper Wedge reservoir has a maximum thickness of 50 m in the core development area of the field (Fig. 7), has the highest reservoir quality and is the primary gas reservoir at Jansz/Io. The Lower Wedge reservoir reaches a maximum of 25 m, generally has poor reservoir quality and is restricted to the core area of the field. The reservoir comprises a shallow-marine sequence that prograded to the north-west during the Oxfordian. Erosion of this sequence by the Tithonian and base-of-Cretaceous Unconformities removed the coastal plain and most of the shoreface sections leaving an erosional remnant of the distal lower-shoreface, offshore and open shelf deposits. The shoreline had a north-east to south-west orientation, with the Jansz-3, Io-1 and Io-2 wells being in a more proximal setting than Jansz-2 (Fig. 9). The reservoir thins to the north-west past Jansz-2 and changes facies to siltstones and mudstones of the open shelf. The reservoir is composed of three primary lithofacies types as identified from core data. The highest quality reservoir and principal sand is designated lithofacies S42, a fine-grained sandstone with 10 to 25% clay, porosities greater than 25% and permeabilities from 10 to 900 md (Fig. 10). A secondary reservoir component is lithofacies S43, a very fine-grained sandstone with 20 to 60% clay content, porosities less than 25% and permeabilities less than 10 md. A minor component includes poorly sorted, coarse-grained sandstones of lithofacies S1. This lithofacies is commonly cemented and stratigraphically restricted to the top of the Io-1, Io-2 and Jansz-3 reservoirs.
Structural Framework Depth Conversion – Structure Maps
A layer-based depth-conversion model is favoured in areas such as Jansz/Io where the overburden exhibits large changes in stratigraphic thickness, with pronounced vertical and lateral changes in velocity. The effect of high-seismic-velocity channel deposits in the overlying section had to be accounted for to remove time-structure artifacts from the final depth-converted structure surfaces. The initial approach to depth conversion used the regionally continuous time-structure markers, the smooth APSTM migration velocity field and the available well control to build a multi-layer depth conversion model. The time-structure map at the base-of-Cretaceous Unconformity marker is shown in Fig. 11. The Jansz-3 well is near the two-way time-structural crest and Io-2 is the most downdip well on the field.
4 IPTC 12461
The Tertiary overburden section above the top-of-Gearle horizon consists of a complex set of shelf-slope canyons with variable sedimentary fill, comprising marls and carbonates. The base-of-Miocene canyon marker provides an example of the Tertiary overburden (Fig. 12). The map depicts the two-way time-structure at the base of the canyon and the thalweg extends across the Jansz-1 and Jansz-3 well locations. The seismic line in Fig. 13 shows a profile through the canyon near the Jansz-2 well location. The velocity log indicates high-seismic-velocity sedimentary fill in the canyon, significantly faster than the background velocity trend established within the underlying mudstones. The seismic time horizons (e.g., base-of-Cretaceous Unconformity) beneath the canyon are locally distorted and two-way time pull-up occurs in response to the high-velocity overburden.
The impact of the high-velocity Tertiary carbonates on the seismic velocity field at the reservoir level is shown
schematically in Fig. 14. The seismic spread length used for the Jansz/Io 3D acquisition is 5500 m, similar to the width of the Miocene canyon. A schematic seismic spread with gathers (at the reservoir level) is shown in Fig. 14, with near-offset and far-offset raypaths. When the seismic spread overlies the canyon, the near-offset raypaths travel through the high-velocity canyon fill twice and arrive back at the receiver with a faster travel time, compared to near-offsets outside the canyon. The far-offset raypaths do not pass through the canyon fill and have unaltered travel times. Consequently, the normal move-out hyperbola is distorted and has more curvature. This results in a slower velocity to stack the data.
The opposite effect occurs outside the canyon. The near-offset raypaths do not pass through the high-velocity canyon fill
and the reflections have normal travel times, while the far-offset raypaths do pass through the canyon fill and have faster travel times. The normal move-out hyperbola is distorted but now has less curvature and a higher velocity is required to stack the data. This is shown in the iso-velocity schematic at the bottom of Fig. 14.
The Jansz/Io depth conversion model has four layers, separated by the water bottom, base-of-Miocene canyon, top-of-
Muderong Shale and base-of-Cretaceous Unconformity markers, to account for the complex velocity structure in the overburden (Fig. 15). It is apparent from the depth-structure map at the base-of-Cretaceous Unconformity (Fig. 16), that the layer-based depth conversion model has not compensated for the distorted seismic velocity field. The outline of the base-of-Miocene canyon (Fig. 12) is imprinted onto the deeper base-of-Cretaceous Unconformity depth horizon and is clearly an artifact inherent in the seismic velocity field. The artifact could not be removed by varying the interval velocity within the Miocene canyon, without invoking geologically unreasonable interval velocities.
A different approach using a geological constraint proved to be a novel method for removing the velocity artifact from the
depth map. The geological constraint was based on the structural history of the Muderong Shale, which is a marine mudstone unit widespread throughout the Carnarvon Basin (Fig. 5). The key assumption for the geological constraint is that the Muderong Shale at Jansz/Io has experienced a relatively simple structural history. The top-of-Muderong Shale is Lower Cretaceous (Aptian), and post-dates the Late Jurassic to Early Cretaceous (Valanginian) structural evolution of the Jansz/Io trap. The only significant structural event at Jansz/Io that post-dates the top-of-Muderong Shale is the regional uplift of the Exmouth Plateau (100 km west of Jansz/Io), during the Late Cretaceous to Recent (Barber, 1988). This caused a regional tilt across Jansz/Io, and the top-of-Muderong Shale now plunges down to the north-east. On this basis, the top-of-Muderong Shale marker should be a regionally dipping surface with no high-frequency complex folding at Jansz/Io. If high frequency folding is evident at this marker, it can be considered an artifact.
Using this reasoning, the depth conversion process was now applied to only the first three layers of the model to output
depth to the top-of-Muderong Shale marker. The map surface was examined and high-frequency folding (imprint of the Miocene canyon) was carefully edited from the map according to the approach outlined in Fig. 17. Once the velocity artifacts were removed, the depth map to the base-of-Cretaceous Unconformity was constructed by adding the fourth layer (top-of-Muderong Shale to base-of-Cretaceous Unconformity isochore) to the corrected top-of-Muderong Shale depth map (Fig. 15). A correction factor that was on average 97% of the depth from seismic velocities was required to correct the maps to actual well depths. The resulting depth map is shown in Fig. 18 and the artifacts noted in the preliminary depth map (Fig. 16) no longer mask the structural style. The Jansz/Io structure is a north-west to south-east trending faulted anticline.
The validity of the final depth map in Fig. 18 was tested and confirmed using depth-migration over the core area of the
field (Fig. 2). The tomographic velocity model that replicated the corrected top-of-Muderong Shale and base-of-Cretaceous Unconformity depth-structure maps in Fig. 16 and Fig. 18, respectively, was used for the depth-migration and produced a flat-gather solution, which indicated the geological constraint produced a plausible depth model. While this is not the only model that produced flat gathers (depth migration models are non-unique), it was the most geologically reasonable. The velocity field from the depth-migration also indicated that a number of previously unrecognized high-velocity canyons (from the APSTM velocity field) were “nested” beneath the Miocene canyon. This explains why the conventional layer-based depth-conversion model using geologically reasonable interval velocities could not remove the structural artifacts (Fig. 19).
IPTC 12461 5
Reservoir Framework Isochore Maps
An isochore map for the key reservoir unit, the Upper Wedge, was required to build the internal reservoir architecture, distribute reservoir properties in the geological model, and for OGIP calculations. The isochron map for the Upper Wedge reservoir zone was converted to an isochore using two methods, depending on whether it was above or below peak-tuning thickness. If the isochron was greater than peak-tuning, a well-based interval-velocity map was used for depth conversion. This method applied to the core area of the field which is above 15 m thick. If the isochron was equal to peak-tuning then a de-tuning algorithm based on the product of seismic amplitude and apparent isochron was applied. This was the case for most of the north-west part of the field in the vicinity of Jansz-2. By merging the maps derived from both methods, a final reservoir isochore map was produced (Fig. 7).
Peak tuning is a function of the band-width of the seismic wavelet and was calculated by ray tracing a forward model of a
wedge, using a wavelet extracted from the Jansz/Io 3D near-angle stack at the reservoir level, as shown in Fig. 20. The model has average gas sandstone and mudstone properties derived from the Jansz-2 well logs. It is evident from Fig. 20 that the apparent isochron of the wedge becomes a constant (11 ms) at peak tuning, when the amplitude is also at a maximum. This corresponds to sandstone thickness of 14 m using the average reservoir velocity at the Jansz-2 well, which excludes the 3 m of silty mudstone at the base of the Oxfordian section in this well.
A linear de-tuning algorithm, based on the product of the apparent isochron and the maximum amplitude on the zero phase
trough (top-of-porosity), was defined to calculate sandstone thickness below peak tuning. The method uses a similar approach to Connolly (2007), with the key assumptions that fluid type (gas) remains constant and seismic amplitude is proportional to sandstone thickness, below peak tuning. The seismic model was calibrated to the actual seismic using the 14 m thick sandstone and maximum seismic amplitude extracted from the top-of-porosity marker at the Jansz-2 well. The variation in amplitude from line to line for the 2D seismic rendered it unsuitable for use in the de-tuning algorithm. Beyond the 3D survey in the north-west part of the field, the de-tuned isochore was linearly extrapolated to the field zero edge, defined by the amplitude vs offset (AVO) response on the 2D seismic data. The isochore for the Upper Wedge reservoir is shown in Fig. 7. This is the primary reservoir at Jansz/Io and the 15 m isochore line defines the core area of the field.
Field Extent and Free Water Level
The reservoir extent and contacts have been interpreted by integrating seismic AVO response, structure maps, wireline-pressure measurements, water saturation from wireline-logs and cores, and capillary-pressure measurements. None of the exploration and appraisal wells have penetrated the FWL. The reservoir is well defined in the core area of the field where the isochore has a pronounced zero edge due to erosion by the Tithonian and base-of-Cretaceous Unconformities (Fig. 8). However, the gas-bearing reservoir section that is below tuning thickness in the north-west part of the field is not well defined from the isochore. Also, there is a progressive facies change in the thin reservoir to the north-west, from more proximal muddy sandstones into distal siltstones and mudstones, because of depositional downlap of the reservoir.
The problem of mapping reservoir extent and gas/water contact (GWC) was solved using the seismic AVO response of the
gas reservoir which distinguishes gas-bearing sandstones from siltstone, mudstone and water-bearing sandstones (Jenkins et al., 2003). In the north-west part of the field at Jansz-2, the cemented coarse-grained sandstones (distal lower-shoreface deposits) drilled at Jansz-3, Io-1 and Io-2 are not present and the top of the reservoir is defined by the boundary between high-impedance mudstones overlying low-impedance gas sandstones. This interface produces either a prominent Class 3 or Class 2 AVO response (according to the scheme of Rutherford and Williams, 1989) at Jansz/Io. A comparison of peak amplitude at the top-of-porosity marker for the band-width balanced near-angle stack and far-angle stack for Jansz/Io is shown in Fig. 21 and Fig. 22, respectively. The extent of the gas reservoir is defined by the far-offset AVO response for the 2D and 3D seismic at the top-of-porosity marker and the far-offset amplitude dim was used to define the detectable edge of the gas sand (1 to 2 m thick), which approximates the field limit (Fig. 22).
All wells have been drilled high on structure and have intersected gas-on-rock. Io-2 is the most downdip well on the
structure and has established a lowest known gas (LKG) at -2970 mss true vertical depth (TVD), indicating at least a 195 m gross gas column relative to Jansz-3, which is at a near-crestal location. The reservoir downdip of Io-2, is typically below seismic peak-tuning thickness but exhibits a Class 2 AVO response, consistent with a thin gas-bearing sandstone. This suggests that a GWC marked by a seismic flat spot is unlikely, but an amplitude termination at a common depth point could be expected to mark the contact. The most likely depth for a potential GWC is marked by the far-offset amplitude dim at -3140 mssTVD, at the south-west end of the field (Fig. 23), resulting in a potential 365 m gross gas column.
Pressures recorded during wireline-logging runs and the cased-hole test are shown in Fig. 24 and indicate all Jansz/Io wells
lie on a common gas pressure gradient. This is consistent with, but does not prove, a common (FWL) for the field. The pressure data shown in Fig. 24 indicate the Mungaroo Formation provides a regional aquifer to the nearby gas fields and may
6 IPTC 12461
provide a field-wide FWL at -3153 mssTVD for Jansz/Io, based on extrapolation of the gas pressure gradient to intersect the Mungaroo aquifer water pressure gradient (Jenkins et al, 2003). This extrapolated FWL depth (-3153 mssTVD) from pressure data corresponds closely with the GWC depth (-3140 mssTVD) determined from the far-offset dim at the south-west end of Jansz/Io (well within the depth conversion error range of +/- 40 m), and is interpreted to be a potential FWL for the field. The potential GWC determined from AVO analysis of gas-bearing sandstones approximates the FWL depth, the difference being a function of the capillary entry pressure for the reservoir.
Within the Jansz/Io Field, it is likely that perched water-bearing sandstones are present in the closed synclines. Where the
thin Oxfordian reservoir section is present in a structural “sump”, it is probably water-bearing and perched at a higher pressure above the FWL of the regional aquifer at -3153 mssTVD (Fig. 25), because the gas fill would have been unable to expel the water. The amplitude on the far-offsets is dim in the structural syncline in Fig. 25, consistent with water-bearing sandstone. It is likely that a number of FWLs exist across the field, depending on the distribution of potential perched water-legs. Work is ongoing to determine where alternative FWLs are likely to be. One approach that is showing promise uses the base of reservoir (Oxfordian Unconformity) depth-structure map, the wireline-log pressure data, capillary-pressure data and log-based saturation measurements. Capillary-pressure and log-based saturation data can be plotted to define a J-function to test different combinations of possible FWLs. The degree to which the wireline-log data can be collapsed to a common J-function may indicate a plausible combination of FWLs. The FWLs derived from capillary-pressure and log-based measurements must also be geologically reasonable according to the base-of-reservoir map (e.g., separate FWLs on opposite sides of a structural ridge), and consistent with the gas gradients derived from wireline-log pressure measurements. This study is ongoing.
Reservoir Characterisation Sedimentology
A total of 200 m of full-hole conventional core has been recovered from the Oxfordian reservoir at Jansz/Io. The location of the cored intervals for each well is shown on the cross sections (Figs. 3 and 4). The wireline-log and core analyses indicate that the reservoir comprises a gross coarsening- and shoaling-upward sequence of predominantly highly bioturbated, muddy, very fine to fine-grained sandstones. A detailed sedimentologic analysis of the core data indicates that the Oxfordian reservoir comprises three major lithofacies types, S1, S42 and S43, with the latter two as the dominant components (Figs. 26–28).
Lithofacies S1 is a minor component of the reservoir and is stratigraphically restricted to the top of the Io-1, Io-2 and Jansz-
3 reservoir sections. The lithofacies is composed of poorly sorted, coarse-grained (occasionally granular) sandstones, having minor (5 to 10%) detrital clay (Fig. 26). The sands are internally unstratified, but commonly display well developed size grading at the top. They are quartzose but with common iron-rich oolites and clay clasts, many of which are phosphatic. Mollusc and wood fragments are common and glauconite is a minor constituent. The S1 sands are of three types: 1) those weakly cemented by siderite, clays, and minor pyrite; 2) those tightly cemented by early authigenic carbonate (calcite and siderite) and; 3) those tightly cemented by a pervasive, early authigenic cement of iron-rich amorphous clay, berthierine/chamosite and siderite. With the exception of the small volume of uncemented sands, most S1 sands have low reservoir quality due to the total occlusion of porosity by early authigenic cements.
Lithofacies S42 is the primary reservoir rock of the Jansz/Io Gas Field. It is characterised as a highly bioturbated (churned),
glauconitic, muddy (10 to 25%), fine-grained sandstone (Fig. 27). Intense bioturbation by a variety of archetypal Cruziana ichnofacies, including abundant Paleophycus, Thalassinoides and common Teichichnus, Helminthopsis and Phycosiphon, has thoroughly homogenised thin-bedded sand and mud units with only rare remnants of the depositional bedding. Skolithos and Cylindrichnus traces are present in some of the more proximal facies. Ammonites, belemnites, bivalves (some articulated), and gastropods are common throughout this lithofacies. The sand fraction is quartzose, with common (10 to 15%) feldspar and minor lithics. The lithofacies is moderately cemented by quartz overgrowths, clays, and local calcite, siderite and pyrite. Reservoir quality is moderate overall, with high to very high porosity but moderate permeability because of the matrix clay.
Lithofacies S43 is a finer grained variant of the S42 lithofacies, and is characterised as a highly bioturbated (churned),
glauconitic, muddy (20 to 60%), very fine-grained sandstone (Fig. 28). Like lithofacies S42, the intense bioturbation of sand and mud interbeds was by a variety of Cruziana ichnofacies, but with a slightly different assemblage indicative of a more distal and deeper water depositional setting. The ichnofacies include abundant Planolites, Helminthopsis and Phycosiphon, common Chondrites and Paleophycus, and rare Teichichnus, Terebellina and Zoophycos. There are no Skolithos traces. Ammonites, belemnites, bivalves and gastropods are common in this lithofacies. The sand fraction is quartzose, with common (5 to 10%) feldspar and rare lithics. The lithofacies is moderately to well cemented by quartz overgrowths, clays, and local calcite, siderite and pyrite. Reservoir quality is low to very low, with moderate to high porosity but very low permeability because of the matrix clay and cement.
IPTC 12461 7
Depositional Setting The depositional setting of the Jansz/Io reservoir was assessed using a variety of data sets, including regional
paleogeography, seismic interpretation, wireline-log analysis, biostratigraphy and core sedimentology. The conclusions are that the sand lithofacies of the Jansz/Io reservoir are the deposits of a nearshore and inner-shelf setting (Fig. 9). Specifically, they are traction, suspension and sediment-gravity deposits of the distal lower-shoreface and offshore zone (sensu Coates and MacEachern, 2000), defined by a distinctive and diagnostic Cruziana ichnofacies assemblage, the macrofossils, the glauconite and the characteristic sedimentologic features, including the stacking pattern, of the sandstones. Sediment was largely fed to the offshore as a result of storm-induced geostrophic flow and surge-ebb underflows. It was then thoroughly mixed and homogenised by feeding and grazing organisms during the prolonged quiescent periods between storms. The intensity and uniformity of the bioturbation, coupled with the rarity of non-bioturbated sand beds with primary sedimentary structures, suggests that the setting was at or below fair-weather wavebase, and that the frequency and intensity of storms was probably not excessive. It suggests that, with the exception of the thick S1 beds at the most proximal localities, the storm beds were thin enough and infrequent enough not to wipe out the ichnofauna. Deposition was slow but continuous, with the organisms able to mix and homogenize the sediment at the rate at which it was introduced.
Overall, the reservoir interval in Jansz/Io is progradational but with a high degree of aggradation (Fig. 4). The sequence
coarsens, thickens, and “cleans” upward, but with small breaks in the sequence defining two depositional cycles (Upper and Lower Wedge). The coarsest sands, namely the S1 lithofacies, occur at the top of the reservoir and represent the most proximal facies. They are interpreted as storm-generated sandy sediment-gravity flows (turbidites) feeding the distal lower-shoreface. The common occurrence of non-abraded, ferruginous ooids in this lithofacies suggests that the sediment source included back-beach ponds and swamps in the paralic zone. The lithostratigraphy is time transgressive and this is evident from the change in lithofacies type along a common time marker (W. spectabilis A/B) from predominantly S1 at the more proximal Io-2 well, to S42 at the more distal Jansz-1 well (Fig. 4). Furthermore, the oldest S1 lithofacies in the Io-2 well equates to lithofacies S42 at Jansz-3.
There are many analogues for the Jurassic, shallow-marine depositional setting at Jansz/Io, including a number of Central
North Sea oil fields such as the Fulmar Field (Johnson et al., 1986; Gowland, 1996) and the Wytch Farm Field in the Wessex Basin, onshore UK (Colter and Harvard, 1981; Morris et al., 2006), which have the same distal lower-shoreface and offshore facies present. The difference between these fields and Jansz/Io is that they have the upper- and lower-shoreface section present at the top of the reservoir and they produce from high quality clean, sandstones. The mud-rich facies are present but mostly below the field hydrocarbon/water contact or if hydrocarbon bearing, they have less than 10% total porosity and are tight reservoir (Colter and Harvard, 1981). At Jansz/Io, the high quality upper- and lower-shoreface reservoirs were removed by erosion at the Tithonian and base-of-Cretaceous Unconformities. The lack of high quality upper- and lower-shoreface reservoir facies is compensated for by a shallow burial depth (1500 m below mudline) for the lower-quality preserved-reservoir section. For example, the S42 lithofacies exhibit total porosities from 25% to 35%, which are capable of gas production at reasonable rates (Fig. 10).
The depositional setting for Jansz/Io was modeled and tied to the well and seismic data within the structural framework of
the geological model (Fig. 29). Four discrete depositional facies belts were recognised within the model and represent a proximal to distal facies tract. The depositional facies are part of a progradational sequence and are time transgressive, as established from well and seismic data (Fig. 9). They include distal lower-shoreface, upper-offshore, lower-offshore, and shelf zones. Each zone was assigned a lithofacies association specifically calibrated to the well logs and core. The distal lower-shoreface is dominated by lithofacies S1 and S42, the upper-offshore by lithofacies S42, the lower-offshore by lithofacies S43, and the shelf by lithofacies S43 and mudstone. Reservoir Quality
The Jansz/Io reservoir is unlike most other shallow-marine reservoirs in that it is entirely a mud-rich sand system. Although mud-rich, there are almost no true mudstone interbeds, other than a 3 m silty mudstone at the base of the Jansz-2 well. Conventional petrophysical cutoffs based on porosity and Vshale are difficult to implement at Jansz/Io due to the bioturbated, mud-rich reservoir. The shallow burial depth has allowed for high total porosities generally greater than 15% throughout the reservoir (Fig. 10), except in thin calcite, siderite and berthierine/chamosite cemented sandstones. The notional permeability cutoff to define net/gross (N/G), was based on the mobilities measured during wireline pressure tests, which suggested limited drawdown below about 0.2 md. The net-pay and reservoir properties are based on a petrophysical cutoff for permeability of greater than 0.2 md (Figs. 3 and 4).
Geological models were built without a net/gross cutoff. The preference was to assign the appropriate rock properties to the
entire gross reservoir and use the physical properties of the reservoir in the dynamic simulation model to determine recoverable gas. Clearly, reservoir with low porosity and permeability and high Vshale would have high total water saturation associated with the microporosity. The resulting low relative permeability for gas would result in very little recovery of gas from this rock in a dynamic simulation model.
8 IPTC 12461
Jansz-3 and Io-1 have the highest quality reservoir seen in the field. The wireline-log data was analysed to determine which reservoir properties could potentially be predicted away from well control by inversion of the 3D seismic data. A good correlation was found for the S42 and S43 lithofacies, between acoustic impedance and total porosity (calibrated to core) from well logs (Fig. 30). This suggested that inversion of the 3D seismic near-angle stack to acoustic impedance could be a useful method for predicting total porosity away from well control. However, the histogram in Fig. 30 indicates that acoustic impedance is a poor indicator of lithofacies type for Jansz/Io. For wells at a common depth and compaction state, the acoustic impedance generally correlates with the lithofacies type, with only a small overlap between S42 and S43 groups at 5500 AI units. However, the Io-2 well, which is 200 m downdip of the other wells, has the same S42 lithofacies but higher acoustic impedance (6000 to 6500 AI units), similar to the S43 lithofacies. This overlap precludes the AI from being used directly for lithofacies prediction.
For “clean” sandstones, rock properties such as lithofacies and Vshale can commonly be predicted using elastic parameters
(Connolly, 1999; Lazaratos, 2006) such as the ratio of the compressional velocity to shear velocitiy (Vp/Vs), or linear combinations of acoustic impedance and shear impedance (Ip-2Is). However, in muddy sand systems such as Jansz/Io, Vshale correlates poorly with Vp/Vs for the primary reservoir (S42 and S43 lithofacies) (Fig. 31). The abundance and variety of clay minerals at Jansz/Io, combined with the uncertainty in deriving Vshale from log analysis, results in the scatter (Fig. 31) for this reservoir property. The rock property study indicated that elastic inversion of the seismic is unlikely to be a reliable method for predicting lithofacies or Vshale away from well control, however acoustic impedance should be useful for the prediction of total porosity. Porosity Model using Acoustic Impedance
The Jansz/Io 3D seismic near-angle stack was inverted to acoustic impedance (AI) using the Constrained Sparse Spike Inversion (CSSI) method (Pendrel, 2006, Buxton-Latimer et al., 2000). On well-log data from the Jansz/Io reservoir interval, a strong correlation is observed between AI and total porosity, indicating that broadband AI could be used to condition a geological model for total porosity (Fig. 30).
Within the seismic frequency band (6 to 90 Hz for the Jansz/Io 3D), inversion uncertainty is principally driven by seismic
wavelet variations related to transmission effects in the complex overburden (Fig. 19). A more significant source of error is the low-frequency model used to constrain the seismic inversion. It is evident from Fig. 32 that the correlation between AI and total porosity is substantially dependant upon low-frequency information, which is not recoverable from the bandlimited (6 to 90 Hz) seismic data. Spectral component analysis of acoustic impedance, measured at well locations, revealed a low-frequency variation in AI resulting from lateral variations in the overburden, due to Tertiary shelf-slope canyons and changes in carbonate and clastic sedimentary fill. Seismic velocity analysis indicated an additional potential for variation with depth below mudline. Extrapolation of the low-frequency model, away from well control, was achieved by merging a stratigraphic model based on well data with calibrated seismic velocity information to provide broadband (0 to 90 Hz) acoustic impedance.
As a comparison with the CSSI seismic inversion result, bandlimited spectral shaping inversion of the near-, mid- and far-
angle stacks (Lazaratos, 2006) was also carried out over the core area of the field. The output AI curves from the two inversion methods have a common low-frequency model and are compared against the filtered (0-90Hz) well logs in Fig. 33. It is apparent that both methods produce a reasonable match to the filtered well logs but the thin Oxfordian reservoir results in overprediction of AI near the top of the reservoir and underprediction of AI near the base. A linear function derived from the correlation in Fig. 30 was applied with a cloud transform approach to the broadband AI to transform it to total porosity for use in reservoir prediction. The resultant average total porosity attribute derived using broadband AI is shown in Fig. 34, for the core area of the field. The colours show average total porosity and contours show the thickness of the Upper Wedge isochore. It is evident that the highest average total porosity occurs between the Jansz-3 and Io-1 wells.
Porosity Model using Depth Below Mudline
Due to the sensitivity of the seismically derived AI models to the seismic wavelet and low-frequency component, a second independent method of porosity prediction was used in order to test the porosity model derived from broadband AI. This method is based on the geological concept that for a given lithofacies, porosity changes with depth of burial below the mudline, because of compaction and diagenetic effects.
The crossplot of depth below mudline and total porosity for the S42 and S43 lithofacies is shown in Fig. 35. A linear
regression was applied to the average porosity for the crestal Jansz-3 and downdip Io-2 wells for both S42 and S43 lithofacies. The porosity variation with depth was assumed to be linear over the 200 m depth range from Jansz-3 to Io-2; the rates of porosity reduction were 3.3 porosity units per 100 m for the S42 lithofacies and 4.7 porosity units per 100 m for the S43 lithofacies. These rates are unusually high for a depth-compaction gradient alone in a clastic reservoir compared to examples from deepwater clean, turbidite reservoirs in West Africa which have compaction rates of about 1 porosity unit per 100 m (McPherson, pers. comm., 2007).
IPTC 12461 9
The high clay content in the Jansz/Io reservoirs would lead to a higher compaction rate but it is likely that the high rates of porosity degradation also reflect the diagenetic history of the reservoir. It is not possible to extrapolate these high rates back to the surface using a simple compaction gradient. The main uncertainties for the depth-below-mudline method are whether the wells have adequately sampled the variability of porosity due to compaction and diagenesis in the reservoir, and the distribution of the lithofacies. Nevertheless, although there may be some uncertainty regarding the exact gradient, the concept is robust that for a given lithofacies, porosity decreases with depth, and this method does permit a good tie to the wells.
The average porosity derived from the depth-below-mudline model, for the Upper Wedge reservoir in the core area of the
field is shown in Fig. 36, and this can be compared with the same property from broad band AI in Fig. 34. Both methods confirm highest average porosities between Jansz-3 and Io-1, however local differences are apparent. The impact of the variability was tested using the uncertainty analysis process.
Permeability
The crossplot of porosity and permeability data in Fig. 10 is derived from wireline-log data calibrated to the core plug data at overburden conditions. The higher permeability S42 lithofacies are gradational from but can be readily distinguished from the lower permeability S43 lithofacies. It is evident from Fig. 10 that the S42 lithofacies have porosities greater than 25% with permeabilities from 10 to 900 md. The S43 lithofacies have porosities less than 25% and permeabilities less than 10 md. While the two types of lithofacies are generally separated in porosity and permeability space, there is an overlap between the two clusters as S42 lithofacies changes transitionally into S43 lithofacies. A single linear function was fitted to the log permeability versus linear porosity crossplot for the S42 and S43 lithofacies and a separate function was generated for the S1 lithofacies. The porosity and permeability data from routine core analysis are not representative for the berthierine/chamosite cemented S1 lithofacies. The apparent porosity and permeability is caused by shattering the “glassy“ berthierine/chamosite cement during pressure reduction and dehydration of the core as it is brought to the surface (Fig. 26)
The cross sections in Fig. 37 compare porosity and permeability effects in the AI-based and the depth-below-mudline
porosity models. Both models indicate a decrease in reservoir quality from Jansz-3 to Jansz-1, and downdip of Jansz-1 towards Io-2. The average porosities are comparable at the locations between the wells for both porosity models, however the range of porosity is different. This affects permeability thickness (kh), which is based on the same reservoir thickness in each model. The AI-based porosity model has about half the kh (md.m) of the depth-below-mudline based porosity model for the same average porosity. The smaller kh in the AI-based porosity model may be a function of the bed boundary effects on acoustic impedance, shown in Fig. 33. Alternatively, the AI-based porosity model may be indicating a diagenetic change in the reservoir that was not sampled by the wells used for the depth-below-mudline based porosity model. Furthermore, the pressure transient analysis from the Jansz-3 drill-stem test indicated an increase in kh several hundred meters from the well-bore. This uncertainty in kh will be addressed in a future appraisal drilling program in the core area of the field.
Uncertainty Analysis The most likely reservoir description provides a reference case for the field but does not itself account for the uncertainty in the subsurface inputs. An uncertainty analysis was used to capture the range of uncertainty around the reference case static and dynamic models. The approach taken for the Jansz/Io assessment was to use a model-based uncertainty analysis that included experimental design (ED). The model-based approach has the advantage over conventional (e.g., Monte Carlo) assessment techniques in that it can be used to understand complex interactions between static and dynamic factors and can output a range of useful responses, including OGIP, recoverable gas and work program timing (e.g., drilling to bring new wells online, timing to installation of compression). The first step for this approach was to establish a list of factors (e.g., gross-rock-volume, porosity and permeability) that could influence a particular outcome or response and to order them according to priority while controlling the resulting two-factor interactions. The key factors that influence the static resource (OGIP) are gross-rock-volume, depositional facies, porosity, permeability and water saturation (Fig. 38). Permeability is included with the static parameters because it is used to derive water saturation in the static model from the J-function. The maximum number of models required to investigate all of the possible high and low outcomes is equal to 2n, where n is the number of factors; this rapidly becomes unmanageable, even for a modest list of factors. The ED software is used to identify the minimum number of runs needed to capture the main uncertainties without aliasing useful interactions between factors, using a fractional factorial design method (LiBong Lee, pers.comm., 2005). It was necessary to choose suitable low-case and high-case values for the parameters in order to capture the range of uncertainty in the assessment around the reference case, which is based on best-estimate parameters. Each model is a unique combination of low-side (-1), high-side (1), and reference (0) values as shown in the matrix in Fig. 38. The outcomes of interest (responses) were captured from the simulation runs and response equations established using multi-linear regression techniques. The variables in the response equation were passed through a Monte Carlo simulation to test the impact of all of the factors together and individually. The outputs are exceedance-probability and uncertainty tornado plots which summarise the range of each response (e.g., OGIP) and show which factors (e.g., porosity) are the major drivers of the uncertainty. The exceedance-
10 IPTC 12461
probability is the inverse of cumulative probability and expresses the chance of the field size being greater than a specific value of OGIP. For example, the p90 OGIP implies there is a 90% chance of the field being greater than that value; thus, the p90 represents the low-side and the p10 represents the high-side. Only the OGIP range will be discussed here to quantify the uncertainty relating to the static model; the dynamic responses are beyond the scope of this paper.
Original Gas-in-Place (OGIP)
One of the key responses output from the model-based Uncertainty Analysis is the OGIP range, expressed as an exceedance-probability plot in Fig. 39. The static model OGIP for the Jansz/Io Oxfordian reservoir has a low-side of 320 Gm3 and a high-side of 950 Gm3 (11 to 33 Tcf). The reference case models were based on the two independent porosity models. The first was based on the depth-below-mudline model and OGIP = 620 Gm3 (21 Tcf), which compares closely with the p50 value of 630 Gm3 (22 Tcf) on the exceedance-probability plot. The second reference case was based on the AI-based porosity model and OGIP = 610 Gm3 (22 Tcf), which also compares closely with the p50 value, giving some confidence that both deterministic reference case models are valid representations of the p50 case.
The uncertainty tornado in Fig. 39 shows porosity, saturation and gross-rock-volume are the key parameters with a similar
impact on the OGIP. The saturation data has high uncertainty because of the issue of multiple FWLs.
Conclusions Reservoir characterization of the gas-bearing Oxfordian sandstones at Jansz/Io was achieved by the full integration of the exploration and appraisal well data and the 3D seismic. The geological setting of the Jansz/Io Gas Field posed a number of technical challenges to the subsurface team including the depth conversion, mapping of reservoir extent, thickness in seismically tuned areas, depositional facies mapping, and prediction of reservoir properties away from well control. The configuration of the structural/stratigraphic trap and the distribution of reservoir properties in the depositional setting required analysis of a variety of geological and geophysical factors, and an understanding of the geological concepts that influenced the trap and the reservoir system. The 3D seismic was processed to be suitable for AVO analysis and seismic inversion, and proved to be essential for definition of the structural framework, reservoir architecture, field extent and for reservoir property prediction. Geological concepts from large (regional) and small (reservoir) scales were essential inputs for the reservoir characterization. At a regional scale, the editing of high-frequency artifacts from the top-of-Muderong Shale depth map was a novel approach for the removal of severe velocity artifacts in the reservoir depth maps. At the small scale, conventional core data were essential for a full understanding of the unusual reservoir facies found in Jansz/Io. The ground-truthing of wireline-logs to the core was a key aspect of the field assessment. The variability in overburden at Jansz/Io resulted in high uncertainty in the low-frequency model, and the band-limited seismic amplitude data used for the inversion. This uncertainty in acoustic impedance translated to porosity uncertainty in the reservoir model. Prediction of porosity using depth below mudline provided an independent test of the AI-based porosity model, and indicated similar average porosity in the core area of the field but a wider range in permeability. This difference in permeability will be investigated with a future appraisal well. The possibility of multiple FWLs will also be further investigated, and scenarios tested for impact on development planning. Geological modeling software is an ideal tool to integrate the geoscience data and concepts to produce a range of different models to describe various reservoir scenarios. Model-based uncertainty analysis is a powerful method using experimental design and analysis of a large number of dynamic simulations of reservoir scenarios captured in the geological models, to output a probabilistic range of static and dynamic responses. The reference cases for Jansz/Io were based on the AI-porosity model and the depth-below-mudline porosity model, and provided useful deterministic representations of the p50 outcome from the probabilistic range. The uncertainty analysis allowed effective planning of the development work program and timing of investments. The field study provided the key subsurface inputs to the Jansz/Io field development plan.
Acknowledgements The authors wish to thank the Jansz Joint Venture in WA-18-R (ExxonMobil, Chevron & Shell) and the Io Joint Venture in WA-25-R and WA-26-R (Chevron, ExxonMobil, Shell & BP) for permission to publish this paper. The views expressed in this paper reflect those of the authors and are not necessarily the views of the respective Joint Venture participants. We especially thank Henryk Wojcik for his assistance with graphics work.
Nomenclature APSDM anisotropic pre-stack depth–migration APSTM anisotropic pre-stack time–migration AI acoustic impedance
IPTC 12461 11
AVO amplitude versus offset CSSI constrained sparse spike inversion ED experimental design FWL free water level Gm3 volume expressed in giga-cubic meters GWC gas/water contact Ip-2Is acoustic impedance – 2 x shear impedance J-function J = Pc/σgw x Cosθgw x (kair/Ф)0.5 where Pc is capillary-pressure, σgw is the gas/water interfacial tension, θgw is the gas/water angle of contact, kair is air permeability and Ф is porosity LKG lowest known gas LNG liquefied natural gas mss meters below sealevel datum MMcf/day flow rate expressed in million cubic feet per day Mm3/day flow rate expressed in million cubic meters per day OGIP volume of original gas-in-place S1 lithofacies: fine to coarse-grained, poorly sorted sandstone S42 lithofacies: fine-grained, well sorted, bioturbated sandstone with 10 to 25% clay S43 lithofacies: very fine-grained, well sorted, bioturbated sandstone with 20 to 60% clay Std/Res ratio of standard pressure (14.7 psia) and temperature (60º F) to reservoir pressure and temperature Tcf volume expressed in units of trillion cubic feet TVD true vertical depth Vshale volume of shale Vp/Vs ratio of compressional velocity to shear velocity
References Barber, P. 1988. The Exmouth Plateau Deep Water Frontier: A Case History. In Purcell, P. and Purcell, R. (eds) The North West Shelf Symposium Proceedings, PESA, 173-187. Buxton-Latimer, R., Davison, R. and Van Riel, P. 2000. An interpreter’s guide to understanding and working with seismic-derived acoustic impedance data. The Leading Edge, 19, (3): 242-248. Coates, L., and MacEachern, J.A. 2000. Integrating ichnology and sedimentology to differentiate between river-dominated deltas, wave-dominated deltas and shorefaces: examples from the Cretaceous of Western Canada. Proc.,Geological Society of America, Cordilleran Section, 96th Annual Meeting, Vancouver, British Columbia, 32, p.A7 Colter, V and Harvard, D. 1981. The Wytch Farm Oil Field, Dorset. From Illing, I. and Hobson, G. (eds), Petroleum Geology of the Continental Shelf of North-West Europe. Proc., Second conference, 494-503. London: Heydon & Son. Connolly, P. 1999. Elastic Inversion. The Leading Edge, 18, (4): 438-452. Connolly, P. 2007. A simple robust algorithm for seismic net pay estimation. The Leading Edge, 26, (10): 1278-1281. Gowland, S. 1996. Facies characteristics and depositional models of highly bioturbated shallow-marine siliciclastic strata: an example from the Fulmar Formation (Late Jurassic), UK Central Graben, From Hurst, A. et al. (eds), Geology of the Humber Group: Cental Graben and Moray Firth, UKCS, Geological Society special Publication No. 114, 185-214. Hefti, J., Dewing, S., Jenkins, C., Arnold, A. and Korn, B. 2006. Improvements in seismic imaging, Io-Jansz Gas Field North West Shelf, Australia. The APPEA Journal, 46, (1): 135-160. Jenkins, C., Maughan, D., Acton, J., Duckett, A., Korn, B. and Teakle, R. 2003. The Jansz Gas Field, Carnarvon Basin, Australia. The APPEA Journal, 43, (1): 303-324. Johnson, H., Mackay, T. and Stewart, D. 1986. The Fulmar Oil Field (Central North Sea): geological aspects of its discovery, appraisal and development. Marine and Petroleum Geology, 3: 99-125.
12 IPTC 12461
Korn, B., Teakle, R., Maughan, D. and Siffleet, P. 2003. The Geryon, Orthrus, Maenad and Urania Gas Fields, Carnarvon Basin, Western Australia. The APPEA Journal, 43, (1): 285-301. Lazaratos, S. and Finn, C. 2004. Deterministic spectral balancing for high fidelity AVO. SEG International Exposition and 74th annual Meeting, Denver, Colorado, Expanded Abstracts, 23, 219. Lazaratos, S. 2006. Spectral shaping inversion for elastic and rock property estimation. Research Disclosure, Issue 511. Morris, J., Hampson, G. and Johnson, H. 2006. A sequence stratigraphic model for an intensely bioturbated shallow-marine sandstone: The Bridport Sand Formation, Wessex Basin, UK. Sedimentology: 1229-1263. Pendrel, J. 2006. Seismic Inversion – The best tool for reservoir characterization. CSEG Recorder, 31, (1): 7-12. Rutherford, S. and Williams, R. 1989. Amplitude versus offset variations in gas sands. Geophysics, 54: 680-688. Veevers, J. 1988. Morphotectonics of Australia’s Northwestern margin- A review. In Purcell, P. and Purcell, R. (eds) The North West Shelf Symposium Proceedings, PESA, 19-29.
MAENAD
URANIA
GERYON
ORTHRUS
GORGON
SCARBOROUGH
W TRYAL ROCKS
JUPITER-1
MERCURY-1
ATLAS-1
TITANIA-1
DIONYSUS
CHRYSAOR
JANSZ-2
BELLATRIX-1
BLUEBELL-1
WA-269-P
WA-205-P
WA-23-R
WA-25-R
WA-22-R
WA-2-R
WA-18-R
WA-26-RWA-24-R
WA-5-R
JOHN BROOKES
MAITLAND
SPAR
WA-374-P
WA-370-P
PLUTO
CHANDON-1
EURYTION
WA-401-P
WA-404-P
WA-392-P
WA-390-PWA-383-P
EM operated
EM non-operated
JANSZ-3JANSZ-1
IO-2
IO-1
CLIO-1
GLENCOE-1
IXION-1
WA-268-P
DampierDampier
Western Australia
JANSZJANSZ
100 km
60 miles
WA-268-P
DampierDampier
Western Australia
JANSZJANSZ
100 km
60 miles
1000
1500
1000
500
200
Jansz/Io gas field
25 km (15 miles)0 25 km (15 miles)0
Fig. 1—Jansz/Io gas field location map. ExxonMobil (EM) interests are highlighted.
IPTC 12461 13
MAENAD
URANIA
GERYON
ORTHRUS
GORGON
W TRYAL ROCKS
DIONYSUS
CHRYSAOR
JANSZ-2
BLUEBELL-1
WA-23-R
WA-25-R
WA-22-R
WA-2-R
WA-18-R
WA-26-RWA-24-R
WA-5-R
JOHN BROOKES
MAITLAND
SPAR
PLUTO
CHANDON-1
EURYTIONIO-2
CLIO-1
1000
50020
0
Jansz/Io 3D
Jansz/Io Database
5 Wells• Wireline logs• 200 m core• Well test
2892 km2 3D seismic
3 by 1.5 km 2D seismic
Core area > 15 meters thick
Jansz 2D
Jansz 2D
Depth cube
JANSZ-3JANSZ-1
IO-1
EM operated
EM non-operated
25 km 0
EM operated
EM non-operated
25 km 0 25 km 0
Fig. 2—Seismic and well database.
7 kmJansz–1 Io–1
2900
2850
Jansz–311 km
2900
2800
2850
Io–2 10 km
2975
Oxfordian U/C
Base
25 m
eter
s
Cretaceous
Max. flow Q = 72.6 MMcf/day (2.06 Mm3/day) 140/64” (56 mm) choke
3025
Oxfordianreservoir
2975
Top porosity
47.1 m gross31.2 m net (N/G = 66%)*
24% ave. Ø*44% ave. Sw*
62.6 mg gross49.7 m net (N/G = 79%)
30% ave. Ø50% ave. Sw
64.5 m gross52.2 m net (N/G = 81%)
31% ave. Ø44% ave. Sw
55.5 m gross
Base high perm.
Second prograde
Oxfordian PSB
* Net pay cutoff permeability > 0.2 mdФ = total porosity Sw = total water saturation
Upperwedge
Lowerwedge
W. spectabilis A/B Boundary
GammaDensitym RKB
Resis Neutron GammaDensitym RKB
Resis Neutron GammaDensitym RKB
Resis Neutron GammaDensitym RKB
Resis Neutron
U/C Datum
Seismic markersBiostratigraphic markerLog marker
Cemented S1 lithofacies
Lower quality gas sand
Higher quality gas sandCore
DST–1
Fig. 3—Stratigraphic cross section along depositional-strike. The datum is the base-of-Cretaceous Unconformity. See Fig. 1 for well locations.
14 IPTC 12461
16.7 m gross6.5 m net (N/G = 40%)*
24% ave. Ø*60% ave. Sw*
25 meters
2900
2850
2800
2850
2850
47.1 m gross31.2 m net (N/G = 66%)
24% ave. Ø44% ave. Sw
64.5 m gross52.2 m net (N/G = 81%)
31% ave. Ø44% ave. Sw
W. spectabilis A/B Boundary
3025
2975
55.5 m gross
21 kmJansz–1 Io–2Jansz–311 km18 kmJansz–2
Max. flow Q = 72.6 MMcf/day (2.06 Mm3/day) 140/64” (56 mm) choke
Seismic markersBiostratigraphic markerLog marker
Cemented S1 lithofacies
Lower quality gas sand
Higher quality gas sandCore
DST–1
GammaDensitym TVD
Resis Neutron GammaDensitym RKB
Resis Neutron GammaDensitym RKB
Resis Neutron
GammaDensitym RKB
Resis Neutron
Upperwedge
Lowerwedge
2900
Top porosity
Second prograde
Oxfordian PSB datum
Base high permeability
BarrowGroup
Lower Dingo/Athol
Oxfordian U/C
Base Cretaceous U/C
* Net pay cutoff permeability > 0.2 mdФ = total porosity Sw = total water saturation
Fig. 4—Stratigraphic cross section along depositional-dip. The datum is an Oxfordian age parasequence boundary in the Lower Wedge reservoir. See Fig. 1 for well locations.
Fig. 5—Carnarvon Basin Mesozoic stratigraphy. The Jansz/Io Oxfordian reservoir is highlighted.
IPTC 12461 15
Jansz–3Jansz–1
2500
Depth (mss)
3000
3500
NW SE
Depositional edge Marine mudstonebase seal
Marine mudstonetop seal
Fluvio-deltaicgas source
Erosional edge
Oxfordian reservoir
MAENAD
URANIA
GERYON
ORTHRUS W TRYAL
ATLAS-1
DIONYSUS
CHRYSAOR
JANSZ-2
BELLATRIX-1
WA-23-R
WA-25-R
WA-22-R
WA-18-R
WA-26-RWA-24-R
WA-370-P
CHANDON-1
EURYTIONJANSZ-3
JANSZ-1
IO-2
IO-1
IXION-1
MAENAD
URANIA
GERYON
ORTHRUS W TRYAL
ATLAS-1
DIONYSUS
CHRYSAOR
JANSZ-2
BELLATRIX-1
WA-23-R
WA-25-R
WA-22-R
WA-18-R
WA-26-RWA-24-R
WA-370-P
CHANDON-1
EURYTIONJANSZ-3
JANSZ-1
IO-2
IO-1
IXION-1
Fig. 6—Jansz/Io hydrocarbon play (schematic).
Io-1
Io-2
Oxfordian reservoirerosional edge
Jansz-1
meters
Jansz–3
WA-18-R
WA-26-R
WA-25-R
WA-374-P
WA-369-P
Jansz 3D
Jansz APSDM
5 km0 5 km0
UPPER WEDGEISOCHORECI = 5 meters
0 10 km0 10 km0 10 km
LINE ALINE A
Fig. 7—Upper Wedge reservoir isochore in the core area of the field. The contour interval is 5 meters.
16 IPTC 12461
NW
Tim
e, s
econ
ds
SE
Line A (APSTM near angle stack)
Jansz–3
Top Muderong
Athol marker
Top Gearle
Base Cretaceous U/C
3.0
GR
Top porosity
Lower wedge
Oxfordian U/CUpper wedge
Second prograde
Base high permeability
3.4
3.0
3.4
Zero edge
Zero phasepolarity convention
Shale(high impedance)
Shale(high impedance)
Gas sand(low impedance)
Zero phasepolarity convention
Shale(high impedance)
Shale(high impedance)
Gas sand(low impedance)
Shale(high impedance)
Shale(high impedance)
Gas sand(low impedance)
LOW HIGH
1 km1 km00 1 km1 km00
Fig. 8—Seismic two-way time section (APSTM) showing the reservoir geometry at the Jansz-3 well. Note the polarity convention; the boundary between a high impedance unit and an underlying low impedance unit yields a negative reflection coefficient, which is displayed as a zero phase trough on a variable area/wiggle trace display or a red event on the variable intensity colour display. The location of Line A is shown in Fig.7.
Lower shoreface
Upper shoreface
Middle shoreface
Upper offshoreLower offshore
Fairweather wavebaseStorm wavebase
Foreshore
10-20 km10-20 km
Open shelf
Jansz–2
Chronostratigraphic markersFacies boundaries
Jansz–3
Io–1
Io–2Jansz–1
View: Looking to the North East
Fig. 9—Schematic block diagram for Jansz/Io showing the shallow-marine depositional setting established during the Oxfordian and prior to erosion by the Tithonian and base-of-Cretaceous Unconformities. The shoreline classification is after Coates and MacEachern (2000).
IPTC 12461 17
Core porosity, %
Cor
e pe
rmea
bilit
y, m
d
S43 low quality reservoir
S42 high quality reservoir
S1
S42
S42 cementedS43
S43 cemented
S1
S42
S42 cementedS43
S43 cemented
Fig. 10—Crossplot of core porosity and core permeability at overburden conditions for the S1, S42 and S43 lithofacies. The cemented lithofacies are also shown.
3400
3300
3000
3200
3100
3500
ms
BASE CRETACEOUS U/CTIME STRUCTURE
CI = 50 ms
WA-18-R
WA-26-R
WA-25-R
WA-24-RWA-22-R
Jansz 3D
Jansz–2
Jansz–1
Jansz–3
Io–1
Maenad–1A
Orthrus–1
Field outline
Dionysus–1
WA-268-P
Geryon–1
Callirhoe–1
WA-269-P
Io–2
WA-374-P0 20 km0 20 km
Chandon–1
Fig. 11—Base-of-Cretaceous unconformity two-way time-structure map. The contour interval is 50 milliseconds (ms).
18 IPTC 12461
ms
2800
2600
2400
2200
2000
1800
WA-18-R
WA-26-R
WA-25-R
WA-24-RWA-22-R
Jansz 3D
Jansz–1
Jansz–3Io–1
Maenad–1A
Orthrus–1
Field outline
Dionysus–1
WA-268-P
Geryon–1
Callirhoe–1
WA-269-P
Io–2
WA-374-P
Chandon–1 Canyon Thalweg
BASE MIOCENE CANYONTIME STRUCTURE
CI = 50 ms0 20 km0 20 km
LINE
-B
Jansz–2
Fig. 12—Base-of-Miocene canyon two-way time-structure map. The contour interval is 50 milliseconds (ms).
Base Cretaceous U/C
SW NE
Line B (APSTM migrated full stack)
Base Miocene canyon
Top Gearle
Muderong
Jansz–2 (Projected)
High velocity canyon fill
25002000
BACKGROUNDTREND
LOW HIGH
2.0
2.5
3.0
3.5
1.5
Tim
e, s
econ
ds
2.0
2.5
3.0
3.5
1.5
Vint (m/sec)
Fig. 13—Seismic time profile (APSTM) across the Miocene canyon and Jansz-2 velocity log. The location of Line B is shown in Fig. 12.
IPTC 12461 19
Base Cretaceous U/C
SW NE
Base Miocene canyon
Muderong
Jansz–2 (Projected)
LOW HIGH
2.0
2.5
3.0
3.5
1.5
2.0
2.5
3.0
3.5
1.5
SEISMIC SPREAD LENGTH
GATHER 1 GATHER 2
TIME SAG
TIME PULL UP
Offset
Tim
e
GATHER 2
Early arrival on nears that pass through high velocity
canyon require slowervelocities to stack
Offset
Tim
e
GATHER 2
Early arrival on nears that pass through high velocity
canyon require slowervelocities to stack
ISOVELOCITYFast
Slow
Fast
Offset
Tim
e
GATHER 1
Early arrival on fars that pass through high velocity canyon fill require faster
velocities to stack
Offset
Tim
e
GATHER 1
Early arrival on fars that pass through high velocity canyon fill require faster
velocities to stack
Offset
Tim
e
GATHER 1
Early arrival on fars that pass through high velocity canyon fill require faster
velocities to stack
High velocity canyon fill
Line B (APSTM migrated full stack)
Tim
e, s
econ
ds
Fig. 14—Seismic profile from Fig. 13 with schematic gathers, illustrating seismic velocity distortion caused by the Miocene canyon. Gather-1 is from the seismic spread partially overlapping the canyon (left) and gather-2 is from the seismic spread overlapping the canyon (right).
Water bottom
Base Miocenecanyon
Oxfordian U/C
Top Mungaroo
Jansz–1Layer 1
Layer 2
Layer 3
Layer 4
Jansz–3 Io–1
Top Muderong shale
Base Cretaceous U/C
0 5 km0 5 km
Fig. 15—Schematic of the layer model for depth–conversion.
20 IPTC 12461
3300
3200
3400
3100
2900
2800
3000
meters
WA-18-R
WA-26-R
WA-25-R
WA-24-RWA-22-R
Jansz 3D
Jansz–2
Jansz–1
Jansz–3
Io–1
Maenad–1A
Orthrus–1
Field outline
Dionysus–1
WA-268-P
Geryon–1
Callirhoe–1
WA-269-P
Io–2
WA-374-P
Chandon–1
0 20 km0 20 km
BASE CRETACEOUS U/CDEPTH (UNCORRECTED)
CI = 50 meters
Miocene canyon artifacts
Fig. 16—Base-of-Cretaceous unconformity depth-structure map prior to correction for velocity artifacts. The contour interval is 50 meters.
Edit “gull wings” toremove artifacts
Jansz–2
Jansz–1Jansz–3
Io–1
Geryon–1
Callirhoe–1Io–2
JANSZ 3D
Jansz 2D
Miocene canyonoutline
2500
meters
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
TOP MUDERONG DEPTHBEFORE EDITING
CI = 50 meters0 10 km
TOP MUDERONG DEPTHBEFORE EDITING
CI = 50 meters0 10 km0 10 km
TOP MUDERONG DEPTHAFTER EDITINGCI = 50 meters0 10 km
TOP MUDERONG DEPTHAFTER EDITINGCI = 50 meters0 10 km0 10 km
Artifacts removed
Jansz–2
Jansz–1 Jansz–3
Io–1
Geryon–1
Callirhoe–1Io–2
Fig. 17—Top-of- Muderong Shale depth-structure map before (left) and after (right) the editing procedure. The contour interval is 50 meters.
IPTC 12461 21
3300
3200
3400
3100
2900
2800
3000
meters
WA-18-R
WA-26-R
WA-25-R
WA-24-RWA-22-R
Jansz 3D
Jansz–2
Jansz–1
Jansz–3
Io–1
Maenad–1A
Orthrus–1Dionysus–1
WA-268-P
Geryon–1
Callirhoe–1
WA-269-P
Io–2
WA-374-P
Chandon–1
0 20 km0 20 km
Miocene canyonartifacts removed
BASE CRETACEOUS U/CDEPTH (CORRECTED)
CI = 50 meters
LINE C
Field outline
Fig. 18—Base-of-Cretaceous unconformity depth-structure map after correction for velocity artifacts. The map is tied to the wells and the contour interval is 50 meters.
Line C depth (Vint)
Athol marker
Canyon-1
Canyon-2
Base Miocene
WB
Mungaroo
Io–1
Base high vel.
Vint (m/sec)1500
3000
4500Line C time (APSTM)
Io–1
LOW HIGHLine C depth (APSDM)
Io–1
A B C
Base Cret. U/C
NW NW NWSE SE SE
Muderong
Athol marker
Canyon-1
Canyon-2
Base Miocene
WB
Mungaroo
Base Cret. U/C
Muderong
Athol marker
Canyon-1
Canyon-2
Base Miocene
WB
Mungaroo
Base high vel.
Base Cret. U/C
Muderong
2km0 2km0 Fig. 19—A: APSTM seismic two-way time profile at the Io-1 well showing local time pull-up at the top-of-Muderong Shale (black arrows). B: Interval velocity for the time profile in (A) derived from the tomographic velocity model used for the depth-migration. High-seismic-velocity sediments in the Miocene canyon and below the Canyon-2 marker (black arrows) cause time pull-up in A. C: APSDM seismic depth profile through the Io-1 well with velocity artifacts removed. The location of Line C is shown in Fig. 18.
22 IPTC 12461
Distance, km
0
50
100
150
0
50
100
150
Tim
e, m
sTh
ickn
ess,
m
0 2 4 6 8
Normal incidence ray trace model
Depth model (Jansz–2 average)
Dominantfrequency 32 Hz
VELOCITY = 2700 m/secDENSITY = 2.36 g/cc
GAS SANDSTONE
MUDSTONE BASE SEAL
MUDSTONE TOP SEAL
VELOCITY = 2577 m/secDENSITY = 2.26 g/cc
VELOCITY = 2664 m/secDENSITY = 2.29 g/cc
App
aren
t iso
chro
n, m
s
80
60
40
20
0
Am
plitu
de x
10-
4
-4
-2
-6
-8
Peak tuning 11ms(14m, Jansz–2 average properties)
Top of gas sandstone
Base of gas sandstone
Distance, km0 2 4 6 810 10
Fig. 20—Tuning wedge depth-model and zero offset ray-traced time-section. The apparent isochron between top- and base-of-gas sandstone and amplitude extracted from the top-of-gas-sandstone event in the seismic wedge-model are shown.
IPTC 12461 23
WA-18-R
Amplitude
Jansz 3D
WA-26-R
WA-25-R
NEAR ANGLE STACK0 20 km
NEAR ANGLE STACK0 20 km0 20 km
Jansz–1
Jansz–3Io–1
Maenad–1A
Orthrus–1Dionysus–1
Geryon–1Callirhoe–1
Io–2
Jansz–2
Chandon–1
-15000
-14000
-13000
-12000
-11000
-10000
-9000
-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
-0
-15000
-14000
-13000
-12000
-11000
-10000
-9000
-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
-0
Field outline
Fig. 21—Maximum amplitude at top-of-porosity from the band-width balanced near-angle stack.
WA-18-R
Amplitude
Jansz 3D
WA-26-R
WA-25-R
FAR ANGLE STACK0 20 km
FAR ANGLE STACK0 20 km0 20 km
Jansz–1
Jansz–3Io–1
Maenad–1A
Orthrus–1Dionysus–1
Geryon–1Callirhoe–1
Io–2
Jansz–2
Chandon–1
-15000
-14000
-13000
-12000
-11000
-10000
-9000
-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
-0
-15000
-14000
-13000
-12000
-11000
-10000
-9000
-8000
-7000
-6000
-5000
-4000
-3000
-2000
-1000
-0
Field outline
Bright amplitudeon far angle stacksdefines field outline
Fig. 22—Maximum amplitude at top-of-porosity from the band-width balanced far-angle stack.
24 IPTC 12461
meters
WA-26-R
Traverse
TOP POROSITY DEPTHCI = 10 meters
Field outline
GWC -3140mss
NEAR ANGLE STACK (Time)
FAR ANGLE STACK (Time)
Top porosity
Oxfordian U/C
GWC -3140mss
S E
AVO increaseAVO decrease
0 2 km0 2 km
LOW HIGH
3100
3110
3120
3130
3140
3150
3160
3170
3180
3190
3200
Jansz–2
Jansz–1Jansz–3 Io–1
Io–2
Jansz–2
Jansz–1Jansz–3 Io–1
Io–2
Fig. 23—Top-of-porosity depth map showing a possible GWC at the south-west margin of the field defined by the far-offset amplitude dim at the top-of-porosity (see Fig. 22) and Oxfordian Unconformity markers. The contour interval is 10 meters.
Jansz/Io
Geryon/Callirhoe Tithonian & Brigadier
Geryon/Callirhoe Mungaroo AA
IO-2Geryon Brigadier
Orthrus Mungaroo
Maenad Mungaroo AA
Maenad Tithonian
Chrysaor Mungaroo A
Dionysus Mungaroo AA AAAA
AA
AA
kPa360003300032000 3400031000 35000 37000
FWL -3153 mss TVD
LKG -2970 mss TVD
500047004600 48004500 4900 53005100 5200
Formation pressure, psia
Dep
th, m
ss T
VD
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
Fig. 24—Jansz/Io pressure data relative to the Mungaroo Formation regional aquifer gradient defines a possible FWL at -3153mss, which compares closely with the amplitude cutoff depth of -3140mss at the south-west margin of the field.
IPTC 12461 25
Jansz–1
TRAVERSE
IO-2
Possible perched water
Break over point3100
3025
3000
FWL
GWC
Perched water
WA-18-RWA-26-R
Jansz 3D
FAR ANGLE STACK (Time)
Top porosity
Oxfordian U/C
SW NE
Possible perched water
Jansz–1
GWC?AVO increase
OXFORDIAN U/C DEPTHCI = 25 meters
0 5 km
OXFORDIAN U/C DEPTHCI = 25 meters
0 5 km
NEAR ANGLE STACK (Time)
LOW HIGH
AVO increase
AVO decrease
Fig. 25—Local closed structural sumps at the base-of-Oxfordian reservoir provide potential for perched water with a local FWL structurally higher than the FWL for the field. The far-offset amplitude dim is consistent with either a water-leg or non-reservoir section in the syncline to the south-west of Jansz-1. The contour interval is 25 meters.
Scale bar = 10 cm
S1b
S1c
Ø = 33%, k = 37 md*
Ø = 9%, k = 0.01 md* Apparent Φ & k due to dehydration & decompaction of clay
Quartz
Siderite
Fe Si grain
Fe clay cement
Siderite
QuartzPhosphaticclay grain
Calcitecement
Calcitecement
0.2 mm0.2 mm
2800
2850
2825
BASE CRETACEOUS U/C
OXFORDIAN U/C
Jansz–3Gamma
densitym RKBResis Neutron
0.2 mm0.2 mm
Fig. 26—S1 lithofacies from Jansz-3; wireline-logs, core photos and micrographs.
26 IPTC 12461
S42
19% detrital clay
S42
13% detrital clay
Ø = 33%, k = 118 md
Quartz
Clay matrix
Ø = 34%, k = 376 md
Quartz
Detrital claymatrix
Glauconite
K-feldspar
Glauconite
2800
2850
2825
BASE CRETACEOUS U/C
OXFORDIAN U/C
Jansz–3Gamma
densitym RKBResis Neutron
Scale bar = 10 cm
0.2 mm0.2 mm
0.2 mm0.2 mm
Fig. 27—S42 lithofacies from Jansz-3; wireline-logs, core photos and micrographs.
Ø = 25%, k = 0.3 md
Clay matrix
Quartz
Shell
Pyrite
Ø = 16%, k = 0.02 md
Quartz
Clay matrix
Calcite
Glauconiticclay pellet
Glauconite
Calcite
GlauconitePyrite
S43
S43c
50% detrital clay
26% detrital clay
2800
2850
2825
BASE CRETACEOUS U/C
OXFORDIAN U/C
Jansz–3Gamma
densitym RKBResis Neutron
Scale bar = 10 cm
0.2 mm0.2 mm
0.2 mm0.2 mm
Fig. 28—S43 lithofacies from Jansz-3; wireline-logs, core photos and micrographs.
IPTC 12461 27
A’
A A’
A A’Jansz–3
Io–1 Proj.
Vertical exaggeration 25:1
UPPER WEDGE
Jansz–1Io–1
Jansz–3Io–2
Proxim
al
Distal
Jansz–2
LOWER WEDGE0 1 km0 1 km
Depositional facies &lithofacies associations
Distal lower shoreface(S1 & S42)
Shelf (Mdst & S43)
Cemented sst
Upper offshore (S42)
Lower offshore (S43)
Depositional facies &lithofacies associations
Distal lower shoreface(S1 & S42)
Shelf (Mdst & S43)
Cemented sst
Upper offshore (S42)
Lower offshore (S43)
Fig. 29—The nearshore and shelf setting of the geological model with the depositional facies belts shown, and the lithofacies associations within each belt. The top surface of the model is displayed.
S42
S42c
S43
S43c
S1b
S1
S1c
Lithofacies
Acoustic impedance, (g/cc*m/sec)
Freq
uenc
y
Downdip Io-2 S42Lithofacies > 5500 AI units
Crestal wellsS42 Lithofacies < 5500 AI unitsS43 Lithofacies > 5500 AI units
Corr’n coeff = 0.94S43 lithofacies
S42 lithofacies
Acoustic impedance, g/cc*m/sec
Poro
sity
, fra
ctio
n
0
4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000 10500 110004000
10
20
30
40
50
60
70
80
90
100
Fig. 30—Crossplot of total porosity and acoustic impedance from wireline-log data for S42 and S43 lithofacies illustrates a high correlation coefficient. Histogram of acoustic impedance sorted by lithofacies type for wireline-log data indicates the depth dependence of acoustic impedance. Note the S42 lithofacies at Io-2 have acoustic impedance values similar to the S43 lithofacies.
28 IPTC 12461
Corr’n coeff = 0.51S43 lithofacies
S42 lithofacies
Corr’n coeff = 0.51S43 lithofacies
S42 lithofacies
Vp/Vs1.5
0.01.6 1.7 1.8 1.9 2.0 2.1 2.2
0.1
0.2
0.3
0.4
0.5
0.6
V sh
ale,
frac
tion
Fig. 31—Crossplot of Vshale and the elastic rock property Vp/Vs illustrates a low correlation coefficient.
Unfiltered P-impedanceCorr’n coeff. = 0.94
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec3000
0.00
4000 5000 6000 7000 8000
0.15
0.20
0.25
0.30
0.35
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec
6 to 90 Hz bandpass filtered P-impedanceCorr’n coeff. = 0.57
-1500 -1000 -500 0 500 1000 1500
0.00
0.15
0.20
0.25
0.30
0.35
Unfiltered P-impedanceCorr’n coeff. = 0.94
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec3000
0.00
4000 5000 6000 7000 8000
0.15
0.20
0.25
0.30
0.35
Unfiltered P-impedanceCorr’n coeff. = 0.94
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec3000
0.00
4000 5000 6000 7000 8000
0.15
0.20
0.25
0.30
0.35
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec
6 to 90 Hz bandpass filtered P-impedanceCorr’n coeff. = 0.57
-1500 -1000 -500 0 500 1000 1500
0.00
0.15
0.20
0.25
0.30
0.35
Tota
l por
osity
, fra
ctio
n
Acoustic impedance, g/cc*m/sec
6 to 90 Hz bandpass filtered P-impedanceCorr’n coeff. = 0.57
-1500 -1000 -500 0 500 1000 1500
0.00
0.15
0.20
0.25
0.30
0.35
Jansz–1Jansz–2Io–1Jansz–3
Well logsJansz–1Jansz–2Io–1Jansz–3
Well logsJansz–1Jansz–2Io–1Jansz–3
Well logsJansz–1Jansz–2Io–1Jansz–3
Well logs
Fig. 32—Comparison of broadband (left) and band-limited (right) acoustic impedance crossplotted against total porosity (fraction). Note the correlation decreases for the band-limited crossplot.
IPTC 12461 29
Res
ervo
irIo–1
Base Cretaceous U/C
Oxfordian U/C
Res
ervo
ir
Res
ervo
ir
Res
ervo
ir
Io–2 Jansz–1 Jansz–2 Jansz–3Acoustic impedance (AI)
Dep
th, m
Sparse spikeinversion
Spectral shapinginversion
Well log AI(0-90 Hz)
Well log AI
Predicted AItoo high
Predicted AItoo low
Base Cretaceous U/C
Oxfordian U/C
Base Cretaceous U/C
Oxfordian U/C
Base Cretaceous U/C
Oxfordian U/C
Base Cretaceous U/C
Oxfordian U/C
Fig. 33—Comparison of broadband acoustic impedance using constrained sparse spike inversion (red) and spectral shaping inversion (green) techniques with the filtered and unfiltered well logs (blue). Thin bed effects are evident at the top and base of the reservoir where AI is overpredicted at the top or underpredicted at the base.
Io-1
Io-2
Edge of Oxfordianreservoir
Jansz-1
Jansz–3
WA-18-R
WA-26-R
WA-25-R
WA-374-P
WA-369-P
Jansz 3D
Jansz PSDM
5 km0 5 km0
Upper wedge averageTotal porosity from AIwith isochore overlay
CI = 5 meters
Ave. Ф
Fig. 34—Average total porosity for the Upper Wedge reservoir using acoustic impedance to predict total porosity. The isochore contours for the Upper Wedge reservoir are overlayed and the contour interval is 5 meters.
30 IPTC 12461
OB core porosity, %
Dep
th, m
bm
l
Jansz 3–5 PU/100 m
Jansz–3
Io–2Porosity Perm.
PredictedActual
Jansz–3
Porosity Perm.
PredictedActual
Jansz–3
10 15 20 25 30 35 401000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
S42
S43
S42 average Jansz–3/Io–23.3 PU/100mS43 average Jansz–3/Io–24.7 PU/100m
S42
S43
S42 average Jansz–3/Io–23.3 PU/100mS43 average Jansz–3/Io–24.7 PU/100m
Fig. 35—Crossplot of depth below mudline and total porosity for Jansz-1, 2, 3 Io1 and Io-2 core plugs by lithofacies type.
Io-1
Io-2
Edge of Oxfordianreservoir
Jansz-1
Jansz–3
WA-18-R
WA-26-R WA-374-P
WA-369-P
Jansz 3D
Jansz PSDM
5 km0 5 km0
Upper wedge averageTotal porosity from
depth BML withisochore overlay
CI = 5 meters
Ave. Ф
WA-25-R
Fig. 36—Average total porosity for the Upper Wedge reservoir using depth-below-mudline and lithofacies to predict total porosity. The isochore contours for the Upper Wedge reservoir are overlayed and the contour interval is 5 meters.
IPTC 12461 31
2750
3000
2750
3000
Jansz-1 Jansz-3SW
Io-1NE
2750
3000
2750Top Oxfordian porosity
Dep
th (m
etre
s)
Upper wedgeLower wedge
Upper wedgeLower wedge
Oxfordian U/CBase high
permeability
Top Oxfordian porosity
Ave. Ф = 0.32
Ave. Ф = 0.30
Ave. Ф = 0.23
Ave. Ф = 0.32
Ave. Ф = 0.30
Ave. Ф = 0.23
Ave. Ф = 0.28kh = 7068
Ave. Ф = 0.28kh = 3008
Ave. Ф = 0.28kh = 1947
Ave. Ф = 0.28kh = 4227
Faul
tOxfordian U/C
Base highpermeability
Reservoir dip section
Depth bml vs total porosity trend
Reservoir strike section
0 2km0 2km
Broadband AI vs total porosity
Jansz-1 Jansz-3 Io-1
Ф (frac.)0.30
0.100.150.200.25
Ф (frac.)0.30
0.100.150.200.25
Fig. 37—Cross sections through the core area of the field showing porosity distribution according to the broadband AI-porosity model (top) and the depth-below-mudline porosity model (bottom). Although average porosities for the Upper Wedge are similar between models the actual porosity range is different, resulting in different kh (md.m) values at the same location. The differences in kh will be investigated in a future appraisal well.
Factor GroupRank
Geomodel or Simulation
factor? GRV / Framework 1 Geomodel
EOD 2 Geomodel Porosity 3 Geomodel
Permeability 4 Geomodel
Water Saturation 5 Geomodel
Reservoir Compaction for collapse (rate of decline)
6 Simulation
Gas Water Contact depth 7 Geomodel
Reservoir Connectivity across Progrades
8 Simulation
Fault Baffles/Segmentation 9 Simulation
Perched Water/Juxtaposed Water/Thief Zones
10 Simulation
Completion Factors (Rscale) 11 Simulation
Model SettingsPorosity Perm Sw GRV EoD
ED01 -1 -1 -1 -1 1ED02 1 -1 -1 -1 -1ED03 -1 1 -1 -1 -1ED04 1 1 -1 -1 1ED05 -1 -1 1 -1 -1ED06 1 -1 1 -1 1ED07 -1 1 1 -1 1ED08 1 1 1 -1 -1ED09 -1 -1 -1 1 -1ED10 1 -1 -1 1 1ED11 -1 1 -1 1 1ED12 1 1 -1 1 -1ED13 -1 -1 1 1 1ED14 1 -1 1 1 -1ED15 -1 1 1 1 -1ED16 1 1 1 1 1ED17 0 0 0 0 0 Center PointED18 0 0 0 0 0 Seed SensitivityED19 0 0 0 0 0 Seed SensitivityED21 -1 0 0 0 0ED22 1 0 0 0 0ED23 0 -1 0 0 0ED24 0 1 0 0 0ED25 0 0 -1 0 0ED26 0 0 1 0 0ED31 -0.66 0 0 0 0ED32 0.66 0 0 0 0ED33 1 1 -1 1 1ED34 1 1 1 1 -1
Cas
e N
umbe
r
Porosity Sensitivities
Permeability Sensitivities
Water Saturation Model Sensitivities
Porosity Sensitivities for Narrower Range
Cases to increase improve resolution
Fig. 38—List of factors used for uncertainty analysis (left) and matrix (right) showing settings for individual model sensitivities. A (-1) denotes low-side, (1) denotes high-side and (0) indicates the reference case parameters. Perm = permeability, Sw = total water saturation, GRV = gross rock volume and EoD = depositional facies.
32 IPTC 12461
Tcf
POROSITY*GROSS ROCK VOLUME
FACIES
PERMEABILITY
STANDARD DEVIATION
GROSS ROCK VOLUME
WATER SATURATION
POROSITY
ALL
0 5 10 15 20 25 30 35 40 45 50 55
MAXP10P50P90MIN
600 800 1000 1200400200 1400Gm3
POROSITY*GROSS ROCK VOLUME
FACIES
PERMEABILITY
STANDARD DEVIATION
GROSS ROCK VOLUME
WATER SATURATION
POROSITY
ALL
0 5 10 15 20 25 30 35 40 45 50 55
MAXP10P50P90MIN
MAXP10P50P90MIN
600 800 1000 1200400200 1400Gm3
Exce
edan
cePr
obab
ility
OGIP, Gm3
OGIP, Tcf
0.1
0.0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 5 10 15 20 25 30 35 40 45
OGIP Gm3 (Tcf)P90 = 320.1 (11.3)P50 = 631.7 (22.3)P10 = 946.2 (33.4)Depth/Porosity = 594.7 (21.0)AI/Porosity = 611.6 (21.6)
OGIP Gm3 (Tcf)P90 = 320.1 (11.3)P50 = 631.7 (22.3)P10 = 946.2 (33.4)Depth/Porosity = 594.7 (21.0)AI/Porosity = 611.6 (21.6)
600 800 1000 1200400200
Fig. 39—Exceedence-probability plot (top) for Jansz/Io OGIP derived from uncertainty analysis, with the two reference cases (Depth/Porosity and AI/Porosity) shown. Depth/Porosity and AI/Porosity refer to the deterministic OGIP models based on depth below mudline vs porosity and acoustic impedance vs porosity relationships, respectively. The tornado plot (bottom) indicates which parameters that define OGIP have the greatest impact on uncertainty.