1 / 30

Pore-scale modelling and multiphysics phenomena Martin Blunt

Pore-scale modelling and multiphysics phenomena Martin Blunt Department of Earth Science and Engineering Imperial College London. 1-100 m m. 0.1 - 1 m. 100s m - kms. Introduction. Design of improved oil recovery schemes in reservoirs

kemal
Télécharger la présentation

Pore-scale modelling and multiphysics phenomena Martin Blunt

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Pore-scale modelling and multiphysics phenomena Martin Blunt Department of Earth Science and Engineering Imperial College London

  2. 1-100 mm 0.1 - 1 m 100s m - kms Introduction • Design of improved oil recovery schemes in reservoirs • Groundwater resources, pollutant transport and clean-up • Geological storage of carbon dioxide Pore-to-core-to-reservoir: digital rocks to continuous time random walks. Transformative approach.

  3. From rock to network to predictions • Starting point is a voxel image of the rock • Obtained from micro-CT, object-based or statistical methods • Use this directly for single-phase transport. • From this a representative network of pores and throats is constructed for multiphase flow computations. • Use a maximal ball algorithm 3 mm Dong and Blunt, Physical Review E 2009

  4. Example scans - carbonates 5 mm

  5. Carbonate at two resolutions Voxel size 2.7 um Voxel size 0.9 um Edward limestone from Texas: Lab measure K 20 mD vs. 19 mD from pore network

  6. Pore-space images of carbonates and networks Estiallades Ketton Mount Gambier

  7. Flow and dispersion – single-phase transport Direct simulation on the pore-space images. Stokes solver, streamline tracing, random motion for diffusion. Sandpack Sandstone (Bentheimer) Carbonate (Portland)

  8. Predicted propagators – concentration profiles Probability, y(t).dt, that a particle will arrive at a nearest neighbour in a time t to t+dt. Compute from block-to-block travel times. y(t)~t-(1+b); <t> divergesb<1; <t2> divergesb<2. b=0.7 b=1.8 Carbonate Sandstone Bijeljic, Mostaghimi and Blunt, Physical Review Letters 2011

  9. Predicted dispersion coefficient Power-law dependence on Pe related to travel time exponent Carbonate Sandstone Sandpack

  10. Towards truly multiscale simulation CTRW – continuous time random walks. Probability, y(t).dt, that a particle will arrive at a nearest neighbour in a time t to t+dt Truncated power-law form at the pore-scale. Generalized network model from pore to field scale. Rhodes, Bijeljic and Blunt, Advances in Water Resources 2008

  11. Pore-scale heterogeneity at the field scale Effectively model a trillion-cell model with effects of uncertainty at the small scale. Reservoir is SPE10. Delayed transport as solute is ‘trapped’ in slow flow regions. Not seen in other simulations, since they cannot capture the pore-scale heterogeneity properly. Rich extensions to reactive transport. Rhodes, Bijeljic and Blunt, Advances in Water Resources 2008

  12. Carbon capture and storage How much can be stored? 920 Gt – 45% of emissions to 2050 in oil and gas fields. 400-10,000 Gt in aquifers 20-500% of emissions to 2050 IEA estimates. 700 Gt in North Sea alone (DTI) ≈ CO2 produced by all UK population for >50 years

  13. Trapping background How can you be sure that the CO2 stays underground? Dissolution, chemical reaction, cap-rock and capillary trapping. Capillary trapping is rapid (decades): CO2 as pore-scale bubbles surrounded by water. host rock Juanes et al, Water Resources Research 2006

  14. Pore-to-core scale experiments • Novel micro-flow design in collaboration with Shell. Also core-scale experiments. • Carbon fibre core holder. Experiments at 9 MPa, 50oC.

  15. After primary drainage (CO2 displaces brine)Blue is brine and white is CO2 Iglauer et al, Geophysical Research Letters 2011 15

  16. Doddington at residual CO2; 50 PV brine injected Iglauer et al, Geophysical Research Letters 2011 16

  17. Trapped CO2 clusters – colour indicates sizeResidual saturation of 25% Pentland et al, Geophysical Research Letters 2011

  18. Producer Injector Design of CO2 storage A case study on a highly heterogeneous field representative of an aquifer below the North Sea. Inject brine and CO2 together and then use chase brine to trap CO2 1D results are used to design a stable displacement Simulations are used to optimize trapping SPE 10 reservoir model, 1,200,000 grid cells (60X220X85), 7.8 Mt CO2 injected. Qi, LaForce and Blunt, SPE 109905; IJGGC 2009

  19. Advancing CO2 front Dissolution Chase front brine front 0.3 0.2 Trapped Sg Mobile CO2 CO2 0.1 0 0 400 800 1200 Distance (m) Simulation Analytical The CO2-phase fractional flow fg as a function of CO2 (gas) saturation, Sg. solution ID results at the field scale with trapping Design to have both injection fronts stable Qi, LaForce and Blunt., SPE 109905; IJGGC 2009

  20. 170m 170m Z Z Y Y X X 3200m 2280m 3200m 2280m Trapped CO2 saturation Mobile CO2 saturation 3D results for aquifer storage 20 years of water and CO2 injection followed by 2 years of water injection in realistic geology 95% of CO2 trapped after 4 years of water injection Qi, LaForce and Blunt, SPE 109905; IJGGC 2009

  21. Water-wet two-phase predictions Experimental data from Berea sandstone cores (Oak ‘90) • No tuning of network (Øren and Bakke 2003) necessary • The fluids are water and oil • Water-wet data – predictions made with θa = [50°, 80°] Primary drainage Secondary waterflooding Relative permeability Relative permeability Valvatne and Blunt, Water Resources Research 2004

  22. Low permeability Middle Eastern Carbonate Synchrotron image at a resolution of 8 mm and associated network.

  23. Relative permeability predictions. K2 Waterflooding relative permeability curves. Network model prediction, with fractional wettability f=0.9, contact angles are ranging from 1000 to 1600 for oil-wet parts (blue lines) matched to steady state relative permeability measurements for sample K2 (black and red points). Also shown is a completely oil-wet case, f=1 (dashed blue).

  24. Relative permeability predictions. K3 Waterflooding relative permeability curves. Network model prediction, with fractional wettability f=0.9, contact angles ranging from 1000 to 1600 for oil-wet parts (blue lines) matched to unsteady state relative permeability measurements for sample K3 (black points). Also shown is a completely oil-wet case, f=1(dashed blue).

  25. Water-wet three-phase data • Water-wet experimental data on Berea cores (Oak ’90) • Gas injection into oil/water • Incorporate double displacement • Few experimental three-phase data sets

  26. At the field scale • Performed a field study on Maureen in the North Sea – wettability trend with initial water saturation. • Representative, fine-scale geological model. 3 km

  27. Recovery curves Much higher recoveries than current state-of-the-art hysteresis models. Combination of low residual saturation (mixed-wet and oil layer drainage) and low water relative permeability giving stable flooding (poorly connected water). Jackson and Blunt, JPSE 2003; SPEJ 2003

  28. Multiscale/multiphysics simulation? • Challenge of scale • From sub-micron to core. Lots of scans and curves. • From core to field: gigacell simulation? • Three-phase flow • Correct expressions for displacement processes • Develop unique, robust code and experimental validation • Characterization of mixed-wet media • Miniaturisation of steady-state experiments in conjunction with micro-CT will allow for pore-scale visualisation • Coupling of three-phase network model with large-scale simulation • Relative permeability is a function of displacement path • Extension of travel-time approach to multiphase flow?

  29. Conclusions • Have methods to image and reconstruct real rocks and extract networks • Predictions of single phase transport are good • Extensions to reactive transport is a rich topic for future work. • Two-phase experimental data are well predicted for a wide range of rocks • Wettability characterisation is not sufficiently well understood • More work is needed to assess impact of pore scale heterogeneity on flow response • Some success with three-phase flow • Where next? New scientific challenges and commercial application.

  30. Acknowledgements Branko Bijeljic, Tara LaForce, Ran Qi (Chevron), Stefan Iglauer (Curtin), Christopher Pentland (Shell), Hu Dong (iRock), Yukie Tanino, Rehab Al-Magharby, Ali Raeini, Peyman Mostaghimi, Oussama Gharbi Imperial College Consortium on Pore-Scale Modelling (Statoil, BG, BP, Saudi Aramco, BHP) Shell under the Grand Challenge on Clean Fossil Fuels Qatar Petroleum, Shell and the Qatar Science and Technology Park under the Qatar Carbonates and Carbon Storage Research Centre

More Related