OS43A-1790 Abstract Reference Number: 1460537 Fluid flow and assessment of the leakage potential in the Snøhvit reservoir and overburden in the Barents Sea Alexandros Tasianas (1), Melanie Darcis (2), Stefan Buenz (1), Jurgen Mienert (1) (1) Department of Geology, University of Tromso, N-9037 Tromsø, Norway. E-mail : [email protected] (2) Department of Hydromechanics and Modeling of Hydrosystems, University of Stuttgart, DE-70569 Stuttgart, Germany. Email : [email protected] 2) Aims 1) Background • The Snøhvit reservoir and overburden have been an important location for testing Carbon Capture and Storage (CCS) techniques. Fluid flow in the region is caused mainly by repeated glacial cycles and differential geographic uplift, which caused tilting and spilling of various structural traps in the area. Seismic data interpretation and geological modeling has allowed us to model the local stratigraphy and any potential fluid-flow features and pathways in order to determine how effective CCS would be in the area. • • To better understand the pathways and mechanisms related to fluid flow at Snøhvit. To propose potential leakage scenarios. To accurately simulate fluid flow with the aid of realistic geological models. 5) References 3) Methods 4) Study area • Existing well-log analysis and new well-log creation (using iMOSS software) • Reflector interpretation (using the autotracking method in Petrel) • Geological model building (via the “Structural framework” tool in Petrel) • Grid population with porosity and permeability values • Simulation of CO2 flow with the aid of Dumux software Seismic data 1) Linjordet, A., Olsen, R.G., 1992. The Jurassic Snohvit Gas-Field, Hammerfest Basin, Offshore Northern Norway. Giant Oil and Gas Fields of the Decade 1978-1988 54, 349-370 2) Kim, G.Y., Yi, B.Y., Yoo, D. G., Ryu, B.J., Riedel, M., 2011. Evidence of gas hydrate from downhole-logging data in the Ulleung Basin, East Sea. Marine Petroleum and Geology 28 (2011) 1979-1985 3) Chand, S., Mienert, J., Andreassen, K., Knies, J., Plassen, L., Fotland, B., 2008. Gas hydrate stability zone modeling in areas of salt tectonics and pockmarks of the Barents Sea suggests an active hydrocarbon venting system. Mar Petrol Geol 25, 625-636. 4) National, Oceanographic, Data, Center, NODC, 2009. World ocean database. http://www.nodc.noaa.gov/OC5/WOD09/pr_wod09.html. 5) Sloan, E.D., 1990. Clathrate hydrates of natural gases. M. Dekker, New York. 5) Results a) Potential leakage scenarios CO2 from the Tubåen Formation (Fm) can partially leak upwards to the Hekkingen Fm or less deep formations via faults (ref.1) and gas chimneys. i) Models with faults ii) Fig.1. Study area location map Models with gas Depending on the leakage mode, models of different types of domain size and grid resolution were created and populated with properties such as porosity (phi) and permeability (k). Variation in the porosity and permeability values allowed also for extreme cases to be considered; giving rise to MEDIUM, LOW and HIGH case scenarios. If leaking CO2 reaches the Top kvitting Fm it can continue migrating upwards via pipe structures, faults, gas chimneys or the clinoforms of the Torsk Fm and accumulate under the Upper Regional Unconformity (URU). The presence of pockmarks at the seabed could indicate further leakage between the URU and the seabed via vertical fluid flow structures underlying the pockmarks. c) Geological modelling chimney 3D conventional seismic data related to cube ST0306 from the Hammerfest Sedimentary Basin (HFB), covering the Snøhvit and Albatross fields, (water depth from -511 to -369 ms TWT) and P-Cable high resolution cubes from Snøhvit (water depth from -468 to -425 ms TWT) were used. d) Simulation results Fig.6. Modeled faults in Fig.7. Permeability model for the HIGH generic case scenario including impermeable continuous a stair stepped form and permeable faults (entire and planar view) (ref .1) Fig.2. Fluid Fluid-flow simulation results with the «MEDIUM» Scenario, without any faults or flow gas chimneys, indicate highest saturation values of CO2 in the interface between mechanisms reservoir and cap rock with no signs of any CO2 reaching the seabed. at Snøhvit 4 years 40 years ? Potential fluid migration pathways b) Pockmarks and fluid flow 1 Fig.3. Snøhvit After 2 years subsurface structure from inline 80 years 140 years 2706 (See The injection fig.1 for period is 30 years location on 3D seismic 6 Km block Fig.10. Generic permeability After 5 years ST0306) model for the MEDIUM case scenario Fig.11. CO2 saturation distribution with time in After injection the COyears 2 phase migrates upwards until it encounters the low Fig.12. CO2 plume evolution for permeability cap rock. CO2 flows within the reservoir alongside the the MEDIUM case models of interface between reservoir and cap rock. This is followed by a slow the faulted domain after y upward migration into the cap rock. No leakage (suitable storage site). number of years of injection. Large horizontal extension of the CO2 plume (3 Km in diameter after 140 (See fig.1 for location) years). e) GHSZ modeling results (ref.3) Fig.4. Pockmark density and Fig.5. Iceberg ploughmark form distribution along the seabed and extent Both Composition 2 (96% CH4, 3% C2H6, 1% C3H8) and 3 (87% Pockmarks occur in the form of small numerous circular ones, the CH4, 4.5% C2H6, 3.5% C3H8, 1.65% N2, 3.4% CO2) gas hydrates are “unit pockmarks” (which measure up to 20m wide and up to 1m deep) stable, providing a supplementary sealing effect that prevents any or much larger asymmetrical ones, the “normal pockmarks” (with potential leaking fluid reaching the seabed. The study area is diameters of several hundreds of meters and depths reaching 12m). however lying outside the Composition 1 (100% CH4) gas hydrate Pockmarks are often found within glacial ploughmarks having affected stability field. Any fluid with such composition leaking from the their internal structure. The distribution of pockmarks can be thus reservoir will thus not form gas hydrates. controlled by the orientation of the ploughmarks. Graph 1. Variation of gas hydrate stability zone thicknesses with varying gas Fig.8. P2 chimney characteristics Fig.9. Vertical k model with gas chimney k values varying both laterally and vertically (ref.2) Cross section Gas Bottom view The injection location is the real location for models including faults, but at a virtual location for the gas chimney models E W N N 3 Km 4 Km Black contour for the CO2 plume End of injection period (after 40 yrs) After 2 years E W N 4 Km 3 Km After 5 years After 30 years 6) Conclusions Postinjection (after 300 years) Fig.13. CO2 plume evolution for the MEDIUM case models of the gas chimney domain after x number of years from injection and for a k of 342 mD in the gas chimney. (ref.2) (See fig.1 for location) a. No indication of leakage of CO2 from fluid flow simulations where no faults and gas chimneys are present. b. There is leakage of CO2 both in the faulted and gas chimney models (with CO2 reaching the seabed only in the latter ones). c. The composition 2 and 3 gas hydrates that may form can provide a supplementary sealing effect that prevents any leaking fluid reaching the seabed. 7) Acknowledgements The 3D conventional data was acquired in 2003 by PGS Geophysical on request by Statoil ASA for which we are grateful. I would also like to thank CGG Norge for having processed the data sets. I also acknowledge the participants of the cruise carried out in july 2011 for helping acquire the 3D P-Cable high resolution data.