Dynamic fmuid connectjvity during steady- state multjphase fmow in a - PowerPoint PPT Presentation

Dynamic fmuid connectjvity during steady- state multjphase fmow in a sandstone Catriona Reynolds, Hannah Menke, Matuhew Andrew, Martjn Blunt & Sam Krevor Proceedings of the Natjonal Academy of Science (2017) 114:31, 8187-8192 Department of Earth Science & Engineering, Imperial College London, UK SPE London Evening Meetjng, 30th January 2018 1

Importance of fmuid fmow for geologic CO2 storage 2

Conceptual models of two-phase fmow Capillary number increasing Ganglion dynamics and connected pathway fmow : Avraam & Payatakes (1995) JFMech, 293: 207-236 Conceptual models of two-phase fmow have been used to justjfy • u i = - k r , i k the multjphase extension to Darcy’s Law Observatjons from simple, model pore spaces suggest two phase ÑF i • fmow occurs via separate, stable channels m As the capillary number increases, the non-wetng phase breaks • i into discrete ganglia which are advected through the pore space by the non-wetng phase 3

Improvements in micro X-ray CT scanning are providing observatjons of new fmow behaviour `Droplet fragmentatjon’ Drainage event during a capillary pressure change Pak et al. (2015), PNAS 112(7): Andrew et al. (2015), Trans Porous Med 110(1): 1-24 1947-1952 4

Experimental Aim: observe rate-dependent, two-phase, steady-state fmow and the pore scale Low capillary numbers (10-8 <Ca<10-5) • representatjve of subsurface fmow Ganglia conditjons (Ca<10-6), mostly below the fmow threshold for pore-scale ganglion motjon Steady-state co-injectjon • Constant fractjonal fmow • Constant fmuid propertjes (viscosity, • Connected interfacial tension pathway Cross Caw-Canw space using total fmow • rate. Look for a transitjon in fmow behaviour • at a critjcal capillary number Datua et al. (2014), Physics of Fluids, 26. 5

Experimental set-up Fluid Propertjes Non-wetng phase N2 Wetng phase Brine (25 wt% KI) Conditjons 50°C, 10 MPa Rock Propertjes Rock type Bentheimer sandstone (>98% quartz) Permeability 2000 mD Porosity 20% 6

Imaging using fast synchrotron X-ray CT at Diamond Light Source Non-destructjve • imaging technique 3D images with a • voxel size of 3.6 microns were acquired every 43 s 7

Image processing 300 microns 8

Steady state saturatjon qT Exp tjme Pore Saturatjon [ml/min] [mins] volumes of N2 0.04 29 15 35.8% 0.3 29 51 34.0% 1.0 29 172 35.4% 0.02 44 15 27.4% 0.1 35 21 22.5% 0.3 45 80 30.2% 1 scan every 45 seconds 0.5 42 124 32.1% • 30-60 3D images per experiment • 9

Time sequence profles of saturatjon confrm steady state FLOW Increasing fmow rate/capillary number 10

Ganglia size distributjon 3.5 mm 150 μm 4mm 40 μm 11

Ganglia size distributjon Percentage of small ganglia increases with fmow rate • Percentage of large ganglia decreases with fmow rate •  N2 becomes less connected 12

Stable connected pathway? FLOW 13

Stable connected pathway? FLOW 14

Stable connected pathway? 0.02 ml/min 0.04 ml/min 15

Ganglia at high fmow rate/ high Ca FLOW 16

Dynamic connectjvity 17

Dynamic connectjvity: a single fmow regime? Disconnected ganglia only No connected Connected N2 pathways from inlet to outlet Intermituent connected pathways 18

Local rearrangements of the pore occupancy allow fmow sporadically, analogous to the stop-and-start of cars on roads controlled by traffjc lights. Dynamic Connected pathway Ganglia fmow connectjvity Connected non-wetng phase pathways form only at low capillary numbers but are not stable • Transient connectjons transport the non-wetng phase between statjc ganglia where the rate and • number of connectjons and disconnectjons increases with increasing fmow rate or capillary number At the contjnuum scale the conductance of a phase is governed by the entjre volume that it • occupies, even intermituently,. 19

Thank you 20

Multjphase fmow characteristjcs of heterogeneous rocks from CO2 storage reservoirs in the United Kingdom Catriona Reynolds, Martjn Blunt & Sam Krevor Water Resources Research (2018), Accepted Manuscript , doi:10.1002/2017WR021651 Department of Earth Science & Engineering, Imperial College London, UK SPE London Evening Meetjng, 30th January 2018 21

Reservoir conditjons afect relatjve permeability because of the varying role of rock heterogeneity Reynolds & Krevor (2015) . Characterizing fmow behavior for gas Replotued from Bachu & Bennion (2008) . Efects of in- injectjon: Relatjve permeability of CO2-brine and N2-water in situ conditjons on relatjve permeability characteristjcs heterogeneous rocks. Wat. Res. Res. 51, 12, 9464-9489 of CO2-brine systems. Environ Geol 54: 1707–1722

All rocks are heterogeneous

Small heterogeneitjes can lead to large variatjons in saturatjon And thus control relatjve • permeability [ kPa ] Larger fmow potentjal can • support larger capillary pressure gradients Virnovsky et al. (2004) Transport in Porous Media 54 : 167- 192

Test the impact by performing experiments under high fmow potentjal (viscous limited) and low fmow potentjal (capillary limited) During drainage: Capillary number decreases at higher fractjonal fmow of CO2 During imbibitjon: Capillary number increases at higher fractjonal fmow of water Reynolds, Krevor (2015) Wat. Res. Res., doi:10.1002/2015WR018046

Applied to real reservoir systems from the UK North Sea Bunter sandstone • Reservoir conditjons (53°C, • 13 MPa, 1 mol NaCl kg-1 Two CO2-brine steady state • drainage and imbibitjon relatjve permeability tests One at low fmow rate • (capillary limited) and one high fmow rate (viscous limited) 26

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

Low flow rate High flow rate S w qT = 0.2 ml min-1 qT = 20 ml min-1 0.8 Imbibition Drainage 0.6 Imbibition Drainage 0.4 1 B3 S w B4 B5 0 B6 0.2 0.3 3 30 Nc

For buoyantly driven CO2 plumes or other low potentjal fmows, the impacts are present in the fjeld Efects persist at larger scales: Li and Benson (2015) AWR; Meckel, Bryant, Ganesh (2015) IJGGC Upscaling techniques established for efgectjve relatjve permeability down to the core scale: Rabinovich, Ithisawatphan, Durlofsky (2015); Lohne, Virnovsky, Durlofsky (2006) SPEJ ; Pickup and Sorbie (1996) SPEJ Reynolds, Krevor (2015) Wat. Res. Res., doi:10.1002/2015WR018046

Small heterogeneitjes have a large impact at fmow rates observed in the reservoir 0.6 10 0 10 0 Residual High fmow rate k r CO 2 B7 saturation B8 10 – 1 10 -1 0.4 10 – 2 10 -2 B1, B3 B2, B4 0.2 B5 B6 10 – 3 10 -3 Low fmow rate 0 10 – 4 10 -4 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 S w Initial CO 2 saturation How do we characterise this? Can we incorporate it into modeling?

Limitatjons to performing core fmoods on heterogeneous rocks Pressure, temperature, fmuids must match reservoir 1. Multjple and low fmow velocitjes needed to cover 2. parameter space Heterogeneity orientatjon in the core should afect 3. fmow the same as in the reservoir