Revised reservoir model for the Paleocene mounds of the Utsira High, North Sea, Norway John Wild (1) & Nowell Briedis (2)

Transcription

1 Revised reservoir model for the Paleocene mounds of the Utsira High, North Sea, Norway John Wild (1) & Nowell Briedis (2) (1) Mobil North Sea LLC (2) Esso Exploration & Production Norway A/S (ExxonMobil Subsidiary Companies) DEVEX May 2009

2 Summary Up until 2005, ExxonMobil s geological model for the mounded Paleogene oil fields of the northern Utsira High consisted of sandy debrite and turbidite fan reservoirs of three distinct ages, linked by post-depositional sand injections. However, it was well recognised that some of the data did not fit this model. Recently improved seismic imaging does not show the expected geometries for the stratigraphic relationships. Stratigraphically-climbing injected sills are now recognised to be of much greater significance. In all the simple mounds, disruptions are imaged in the underlying Chalk that are interpreted to be large fluid escape features. Stratigraphically anomalous sands have been encountered by wells drilled through these features. We now re-interpret the Balder mounds to be extensive fluidised injection complexes. The mound morphology may be entirely secondary in origin, caused by fluid escape from the pre-chalk section underlying the Utsira High. A new simulation model based on these ideas is more closely matching production history at Balder than previous models. Page 2

3 Location and early Tertiary setting Top Paleocene (Base Balder Fm.) Structure Ringhorne Norway A A Grane A A' EAST SHETLAND PLATFORM VIKING GRABEN ca 90km UTSIRA HIGH Balder BALDER 1700 EOCENE 2000m PALEOC ENE PALEOCENE SAND SHALE NW 1800 CRET 1900 UJR 2000 JR TR SE 10km 6000m Page 3 Depth m

4 Previous geological model: Compensational stacking of reservoirs Base Balder Structure Ringhorne Grane A Balder A km Depth m A Five major Paleocene-Early Eocene sands were identified Heimdal and Hermod sand mounds were interpreted to be sandy debrites (shed from the East Shetland platform) Younger Balder sands were thought to be mostly turbidites All reservoirs were thought to be compensationally stacked, in a relatively unconfined deep-water setting. Much of the mounded relief was thought to be due to post-depositional sand re-mobilisation A Sand Re-mobilisation Conceptual Line of Section Page 4

5 Unusual characteristics of the sand Section A Balder M5 25/ /11-5 Balder 25/11-A5H 25/11-A5H Hermod Th5 100m 100m Heimdal Yp2 600m Th2 Ty PETM 25/11-A4BH 25/11-A4BH GAS SAND OIL SAND WET SAND CLAYSTONE TUFF CHALK Sands are not found at consistent stratigraphic levels The sands are clean, fine to medium grained and consistently high porosity (~33%) Massive sands in the Heimdal, Hermod and Balder are visually indistinguishable and mineralogically identical Angular shale intraclasts are seen in core No distal or off-axis facies encountered (in over 150 wells) Location Ringhorne Section B Section A Balder Page 5

6 Unusual characteristics of the shale Erosion of shale is minimal or absent For 25 years the uniform shale (thickness) model has been used to predict net sand prior to drilling Local variation in Paleocene isopach is due to variable sand thickness Net sand estimated from isopach Complete Section 25/ / / / /11-21S 25/11-18T2 Datum Balder Hermod Heimdal GAS SAND OIL SAND WET SAND CLAYSTONE TUFF CHALK Ty Chalk >200 m 200m 100m 0m 0m Page 6 Shale Only 25/ / /11-20

7 Beneath the mounds: intriguing anomalies in the Chalk TWT Chalk Rafts Seismic line with hole and raft Chalk Structure Base Balder Chalk Chalk Rafts Disrupted Chalk Reflector West 1,000 m ~5x Vertical Ex East Detached rafts of chalk are associated with all simple mounds Allocthonous glide blocks derived from the east? But - matching underlying depressions in the Chalk surface Map 1700 Line of section Mound structure Attribute showing rafts 2100 Depth m 10km 2500 Page 7

8 Balder raft penetration TWT 25/11-C-13 Chalk Raft N In-situ Chalk 1 Location North 500m ~2x Vert Exag South The raft is elevated about 10m Matching depression in the underlying Chalk High quality massive sand beneath raft Page 8

9 Ringhorne raft, Ty and Heimdal sand relationship 25/8-C3 25/8-15S 25/08-15S (pilot) 25/08-C3 (producer) TWT Repeat with Lithlog Repeat section South 1,000 m ~2x Vert Exag North Location Sand beneath the raft is in seismic continuity with Heimdal sands drilled 250m away Well 25/8-C3 shows 20m repeat shale section over chalk The raft is elevated by ~35m Size: ~ 1,100 x 600 x 15m Mass: ~ 16 Million tonnes Page 9

10 Cretaceous Chalk clasts in Paleocene sands 100mm 25/8-10ST2: 1778m 25/8-10S: 1737m 50mm Chalk clasts have been penetrated in core Larger 25+ cm clast is in massive well-sorted finemedium grained sand Smaller 5 cm clast is contained within angular intraclastic breccia, typical of thinner injection sands in area Many more wells show probable chalk clasts on wireline logs Location Page 10 Page 10

11 Chalk rips West 1,000 m ~2x Vert Exag East Progression seen from: + Short deflections in chalk (>100m) with dipping reflector above, to a + Rip a normal to reverse fault rarely exceeding 200m in Xl 1810 length, sometimes with a small partly-detached raft Location Page 11

12 Relationship of Heimdal, Hermod and Balder sands 1000 m West ~2x Vertical Ex Rip in Chalk under mound Reflectors cross-cut Base Balder, joining M3 to M4 Balder M1 East M2 M3 M4 (Eocene) Page 12

13 Relationship between Heimdal, Hermod and Balder sands West 1,000m ~2x Vertical Ex Reflectors cross-cut biostratigraphy on mound margins Higher continuity of sand is consistent with well performance Balder M1 East M2 M3 M4 (Eocene) Page 13

14 Distribution of Balder Fm. (Eocene) sand Base Balder Structure Base Balder Structure Attribute-based Balder Sand isopach 70m 1 2 M1 M2 40m 20m M3 0m TWT 2km Sum of neg amplitudes calibrated with well data Massive (~40m, 100% N/G) high quality (Ø~33%) Balder sand is unusual for a distal turbidite No source or feeder channels have ever been found No contemporaneous sands have been drilled in main basin to west East Shetland Platform undergoing transgression during Balder period - unlikely sediment source We now interpret the majority of the Eocene Balder sand to be intrudites and extrudites Page 14

15 Summary of Seismic-Stratigraphic relationships West 1000 m ~2x Vertical Ex East Most or all of the Hermod and Balder sands are interpreted to be sourced directly from the Heimdal mounds, rather than deposited compensationally around them All the mounds are underlain by disruptions in the Chalk We interpret there to be a genetic relationship between the two Page 15

16 Possible control on location of the Paleocene mounds Base Triassic Structure A A Balder M3 TWT Paleocene Mesozoic < Tr u/c Tr u/c > Zechstein Paleozoic Basement West 5,000m ~2x Vert Exag East The mounds overlie the flank of the Utsira High, a prominent tectonic high lying between the Viking Graben and Stord Basin The chalk rafts are generally confined to an area underlain by a subcrop of thick Paleozoic clastics below a Lower Triassic angular unconformity Page 16

17 Summary of the new model Hermod Heimdal Our observations are consistent with most of the sand in the Paleocene mounds and the overlying adjacent Eocene being continuous bodies injected across stratigraphy This model can explain the: + Minimal erosion of shale + Lack of stratigraphic consistency + Absence of transitional facies + Lack of vertical variations in porosity + Common, field-wide OWC and GOC + Excellent reservoir communication on a production time-scale The mounds are related to Chalk rips and rafts, which are thought to be fluid escape structures The location of mounds appears to be controlled by deeper structure The original distribution and geometry of depositional Paleocene sand is obscured by the extensive remobilisation The volume of injected pre-paleocene sand remains uncertain 25/8-1 25/8-1 25/8-C24 25/8-C24 25/8-3 25/8-10ST2 25/8-10ST2 25/8-4 Balder Hermod Heimdal Ty Chalk One sand body - injected across stratigraphy Balder M3 Page 17

18 Potential importance for field production Modified mound geometries in a new field simulation is improving history match to production Our model helps to explain: The high degree of communication between reservoirs of differing ages Observed common oil and gas contacts over large areas The shared aquifer over entire region...and we hope will lead to the optimal placement of future wells Page 18